Practical Guide to Simulation in Delivery Room Emergencies 3031100662, 9783031100666

Citation preview

Gilda Cinnella Renata Beck Antonio Malvasi Editors

Practical Guide to Simulation in Delivery Room Emergencies

123

Practical Guide to Simulation in Delivery Room Emergencies

Gilda Cinnella • Renata Beck • Antonio Malvasi Editors

Practical Guide to Simulation in Delivery Room Emergencies

Editors Gilda Cinnella Professor of Anesthesia and Intensive Care University of Foggia Foggia, Italy

Renata Beck Department of Anesthesia and Intensive Care Riuniti University Hospital, University of Foggia Foggia, Italy

Head of Anesthesia and Intensive Care Department Riuniti University Hospital Foggia, Italy Antonio Malvasi Department of Biomedical and Human Oncological Science (DIMO) Unit of Obstetrics and Gynecology University of Bari Bari, Italy Santa Maria della Misericordia Hospital Perugia University of Perugia Perugia, Italy International Translational Medicine and Biomodelling Research Group Department of Applied Mathematics Moscow Institute of Physics and Technology (State University) Moscow, Russia The New European Surgical Academy (NESA) Berlin, Germany

ISBN 978-3-031-10066-6    ISBN 978-3-031-10067-3 (eBook) https://doi.org/10.1007/978-3-031-10067-3 © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

To my children Mariaelena and Vito Maurizio, who are the most precious part of my life, and who will certainly continue to follow my dedication to medical science and research. Prof. Antonio Malvasi To my children Nancy and Brian, and to my mother and father for giving me serenity, love, and strength now and then, to fulfill my dreams, study what I like, and do my best in my work. To the health of mothers and babies. Dr. Renata Beck To my beloved husband Luciano and my son Chanthit. Prof. Gilda Cinnella

Foreword

Obstetric emergencies are a major public healthcare problem that requires human and structural resources. The delivery room is a place where different professionals meet, and it is during emergencies that they demonstrate their skills and experience in the service of pregnant women and newborn babies. Triage represents the cornerstone of the structure and the pathway that the members of staff must follow to resolve the emergencies they are faced with. Emergencies, including obstetric emergencies, require trained and experienced staff members as part of the team. Staff members’ experience is gained in the delivery room and needs continuous updating. It is important to improve learning about possible emergencies through simulation training. For this purpose, recently many hospitals have equipped simulation laboratories, which are used for regular courses. As a consequence, simulation has also become increasingly important for medico-legal aspects. In fact, in the event of maternal and/or fetal complications, certification of the attendee courses of simulation can play a role in medico-legal litigation. This book covers obstetric emergencies with the expertise of leading experts in obstetrics, anesthesia, and neonatology, highlighting the value of training and simulation to improve the health of mothers and children. Prof. Ettore Cicinelli Department of Biomedical and Human Oncological Science (DIMO) Unit of Obstetrics and Gynecology University of Bari Bari, Italy Postgraduate School of Obstetrics and Gynecology University of Bari Bari, Italy

vii

Acknowledgments

Special thanks to the School of Specialization in Obstetrics and Gynecology, University of Bari “Aldo Moro,” (Italy) for the efforts made in the production of this book. Prof. Antonio Malvasi We would like to thank the University of Foggia (Italy), the instructors of the high-fidelity medical simulation center SimUMed, and the trainees of the Department of Anaesthesia and Intensive Care for their support and contribution. Prof. Gilda Cinnella and Dr. Renata Beck

ix

Contents

Part I Fundamentals of Simulation 1 Simulation  in Obstetric: From the History to the Modern Applications�������������    3 Reuven Achiron, Laura Adamo, and Tal Weissbach 2 The  Role of Simulation in Obstetric Schools in the UK ���������������������������������������   19 Sasha Taylor and Wassim A. Hassan 3 Ontologies,  Machine Learning and Deep Learning in Obstetrics �����������������������   29 Lorenzo E. Malgieri Part II Simulation and Management of Pathologic Pregnancy 4 Assisted  Reproductive Technologies: Complications, Skill, Triage, and Simulation���������������������������������������������������������������������������������������������   67 Maria Mina, Ioannis Tsakiridis, Styliani Salta, Themistoklis Dagklis, Apostolos Mamopoulos, Anastasia Vatopoulou, Angelos Daniilidis, Apostolos Athanasiadis, Minas Paschopoulos, Ioannis Kosmas, Antonio Malvasi, and Domenico Baldini 5 Acute  Abdomen of Non-obstetric Origin in Pregnancy�����������������������������������������   97 Giuseppe Piccinni, Christopher Clark, and Emanuela Cagnazzo 6 Eclampsia:  Skill, Triage, and Simulation���������������������������������������������������������������  113 Susan Leong-Kee, Brennan Lang, and Julia Lawrence 7 Renal  Failure in Pregnancy�������������������������������������������������������������������������������������  133 Lada Zibar and Katja Vince 8 Simulation  in Obstetric Patients with Cardiovascular Disorders �����������������������  141 Erkan Kalafat and Koray Gorkem Sacinti 9 Cardiac  Arrest in Pregnancy: Simulation and Skills���������������������������������������������  155 Daniele De Viti, Agostino Brizzi, Pierpaolo Dambruoso, Pasquale Raimondo, and Flavio Fiore 10 Aortic  Dissection in Pregnancy�������������������������������������������������������������������������������  179 Vito Margari and Domenico Paparella Part III Simulation and Management of Pathological Fetus 11 Twin-Twin  Transfusion Syndrome: Complications and Management ���������������  191 Sultan Seren Karakus

xi

xii

12 Intrauterine  Fetal Death: Management and Complications���������������������������������  219 Reyyan Gökçen İşcan and Antonio Malvasi 13 Abortion  an Obstetric and Anesthesiologic Emergency: Skills and Simulation �����������������������������������������������������������������������������������������������  245 Resul Karakuş and Önder Tosun Part IV Simulation of Normal and Abnormal Labour 14 Labor  Simulations: “Hard Drill Makes an Easy Battle” �������������������������������������  269 Chen Ben David, Yoav Paltieli, and Ido Solt 15 Intrapartum  Ultrasonographic Simulation in Dystocic Labor�����������������������������  279 Sasha Taylor and Wassim A. Hassan 16 Simulation  and Learning Curve of the Traditional and Sonographic Pelvimetry �����������������������������������������������������������������������������������������������������������������  289 Dominic Gabriel Iliescu, Smaranda Belciug, and Ioana Andreea Gheonea 17 Simulation  of Urgent Obstructed Delivery: Scenario and Triage �����������������������  309 Alexis C. Gimovsky 18 Twin Vaginal Delivery�����������������������������������������������������������������������������������������������  333 Miha Lučovnik, Lili Steblovnik, and Nataša Tul 19 Emergency  Delivery in Patients with Obesity �������������������������������������������������������  343 Haitham Baghlaf, Cynthia Maxwell, and Dan Farine Part V Simulation and Management of Pathologic Delivery 20 Breech  Delivery and Updates in Simulation for Breech Vaginal Delivery�����������  363 Joseph Bouganim, Fatima Estrada Trejo, and Kfier Kuba 21 Umbilical  Cord Prolapse: Simulation, Skills and Triage �������������������������������������  381 Antonella Vimercati, Antonio Malvasi, Raffaella Del Papato, Nico Picardi, Ilaria Ricci, Marta Spinelli, and Ettore Cicinelli 22 Unexpected  Placental Invasion: Scenario, Management, and Simulation����������  397 Giuseppe Calì, Francesco Labate, Francesca De Maria, Federica Calò, and Laura Messina 23 Abnormal  Invasive Placentation Simulation of Emergency Scenario: Low- and Full-Resource Setting �����������������������������������������������������������������������������  403 José M. Palacios-Jaraquemada, Nicolás Andrés Basanta, and Álbaro José Nieto-Calvache 24 Uterine  Rupture: A Rare Event But Terrible to Know How to Face�������������������  411 Andrea Tinelli, Antonio Malvasi, Marina Vinciguerra, Gianluca Raffaello Damiani, Miriam Dellino, Ilaria Ricci, and Antonella Vimercati Part VI Operative Delivery Simulation 25 Urgent  Cesarean Section with Misgav Ladach (Stark’) Method: Simple Cesarean Delivery and Learning Curve�����������������������������������������������������  441 Michael Stark, Andrea Tinelli, and Antonio Malvasi 26 Simulation  of Urgent Cesarean Delivery: Scenario and Triage���������������������������  457 Panos Antsaklis and Maria Papamichail

Contents

Contents

xiii

27 Forceps  Application: Training, Simulation, and Learning Curve �����������������������  469 José Antonio Sainz-Bueno, Rocio Garcia Jimenez, Laura Castro Portillo, Luis M. Pastor Colomer, Carlota Borrero González, and José A. García Mejido 28 Vacuum  Extractor: Skills, Education, Simulation, and Learning�����������������������  479 Sasha Taylor and Wassim A. Hassan Part VII Postpartum Haemorrhage Simulation 29 Abruptio  Placentae: Simulation and Scenario�������������������������������������������������������  499 Ingrid Marton and Dubravko Habek 30 Skills  Training and Multi-Professional Simulation Training on Postpartum Haemorrhage ���������������������������������������������������������������������������������  515 Signe Egenberg, Alemnesh Reta, Jette Led Sørensen, Anna af Ugglas, Shirley Nilsen, and Cherrie Evans 31 Combined  Management of Postpartum Obstetric Bleeding Using Zhukovsky Balloon Tamponade������������������������������������������������������������������������������  523 Sergey V. Barinov, Yulya I. Tirskaya, Tatyana V. Kadsyna, Oksana V. Lazareva, Irina V. Medyannikova, and Aleksander V. Bindyuk 32 PPH:  Triage, Scenario, and Simulation �����������������������������������������������������������������  533 Christian Bamberg 33 Postpartum Hemorrhage: Conservative Treatments���������������������������������������������  539 Antonio Simone Laganà, Jvan Casarin, Antonio Lembo, Elisa Ervas, and Antonella Cromi 34 The  Role of Noninvasive Uterine Mechanical Compression in the Golden Hour of Postpartum Hemorrhage���������������������������������������������������  557 Antonio Belpiede, Antonio Malvasi, Claudio Crescini, Giuseppe Trojano, and Ettore Cicinelli Part VIII Management of Puerperium and Simulation 35 Complicated Cesarean Hysterectomy���������������������������������������������������������������������  563 Pelin Özdemir Önder, Çetin Kılıççı, and Şafak Hatırnaz 36 Postpartum  Uterine Inversion: Skill, Simulation and Learning Curve���������������  577 Dragan Belci and Michael Stark 37 Emergency  and Urgency in Puerperium: Scenario and Complications �������������  585 Andrea Dall’Asta, Monica Minopoli, and Tullio Ghi Part IX Obstetric Anesthesia Emergencies 38 Fundamentals  of Emergencies in Obstetrics: Training and Simulation �������������  603 Kübra Taşkın and Cansu Ofluoglu 39 Simulation  of Difficult Airway Management in Obstetric Emergencies�������������  621 Renata Beck, Potito Salatto, Giuseppe Ferrara, Nancy Loco, Jadranka Pavičić Šarić, and Enrico Marinelli 40 Sonographic  Locating of the Lumbar Space in the Difficult Spine and Obese Parturient: Simulation and Skills���������������������������������������������������������  643 Ayse Gulsah Atasever and Marc Van De Velde

xiv

41 Amniotic  Fluid Embolism and the Role of Thromboelastometry. And What About Simulation?���������������������������������������������������������������������������������  653 Judith Gerets, Frederik Marynen, Ayse Gulsah Atasever, and Elke Van Gerven 42 Septic  Shock in Obstetric Emergency���������������������������������������������������������������������  671 Antonella Cotoia, Giulia Zingarelli, Renata Beck, and Gilda Cinnella 43 Transfusional  Optimization Using Viscoelastic Test Guided Therapy in Major Obstetric Hemorrhage: Simulation and Skills�����������������������  683 Lucia Mirabella, Marco Paolo Perrini, and Renata Beck 44 Thromboelastography  (TEG): Point of Care Test of Hemostasis for Emergency Postpartum Hemorrhage���������������������������������������������������������������  695 Pierpaolo Dambruoso, Pasquale Raimondo, Daniele De Viti, Antonio Malvasi, and Agostino Brizzi 45 Improving  of Hemodynamic and Hemostatic in the Golden Hour ���������������������  701 Antonella Cotoia and Giuseppe Ferrara Part X Neonatal Emergencies 46 Neonatal Resuscitation���������������������������������������������������������������������������������������������  713 Katarina Bojanić, Dora Jelinek, Ruža Grizelj, Nada Sindičić Dessardo, and Tomislav Ćaleta 47 Premature  Neonatal Life Support���������������������������������������������������������������������������  725 Katarina Bojanić, Nada Sindičić Dessardo, Ruža Grizelj, Tomislav Ćaleta, and Dora Jelinek 48 Umbilical  Venous Catheter Placement: A Step-by-Step Guide for Neonatologists�����������������������������������������������������������������������������������������������������  739 Ruža Grizelj, Katarina Bojanić, Tomislav Ćaleta, and Dora Jelinek 49 Simulation  of Newborn Thermoregulation and Temperature Preservation After Birth�������������������������������������������������������������������������������������������  751 Matteo Rinaldi, Annalisa Fracchiolla, and Gianfranco Maffei 50 Role  of Training in Neonatal Encephalopathy Prevention�����������������������������������  757 Matteo Loverro, Nicola Laforgia, Maria Teresa Loverro, Antonio Malvasi, and Edoardo Di Naro Part XI Covid 19 Pandemy and Obstetric Simulation 51 Voluntary  Termination of Pregnancy, Therapeutic, and Spontaneous Abortion: What Is Happening in Coronavirus Era? An Italian Experience�����������������������������������������������������������������������������������������������  781 Marina Vinciguerra, Marcella Lerro, Rosanna Zaccaro, Antonio Malvasi, Giuseppe Trojano, Bruno Lamanna, Giuseppe Lupica, and Giovanni Di Vagno 52 SARS-CoV-2-Related  Acute Respiratory Failure in Pregnant Women: What Role Can Simulation Play?�������������������������������������������������������������  801 Bénédicte Jeannin and Dan Benhamou 53 Acute  Abdomen in Pregnancy: Triage, Skills, and Simulator during COVID-19 Pandemic Situation�������������������������������������������������������������������  813 Sarah Gustapane, Andrea Tinelli, and Antonio Malvasi

Contents

Contents

xv

54 Postpartum  Hemorrhage in COVID-19 Patients: Instruction for Use ���������������  829 Giovanni Di Vagno, Antonio Malvasi, Giuseppe Lupica, Alessandra Ferrari, Giuseppe Trojano, Dragan Belci, Ospan Mynbaev, and Alessandro Savino 55 Urgent  Cesarean Delivery in COVID-19 Patients: Simulation, Skill, and Triage �������������������������������������������������������������������������������������������������������  849 Antonio Malvasi, Davide Campanelli, Luigi Liaci, Giovanni Di Vagno, Rosanna Zaccaro, Nico Picardi, Nancy Loco, and Michael Stark 56 Cardiopulmonary  Resuscitation of a Pregnant Woman During COVID-19 Pandemic �����������������������������������������������������������������������������������������������  867 Daniele De Viti, Pasquale Raimondo, Antonio Pipoli, Chiara Spina, and Assunta Stragapede Part XII Complications, Medico-Legal Issues and Importance of Simulation 57 Can  the Simulation of Delivery Prevent Perineal Trauma?���������������������������������  879 Rasha A. Kamel and Khaled M. Ismail 58 Shoulder  Dystocia and Simulation �������������������������������������������������������������������������  887 Nicola Caporale, Alessandro Svelato, Caterina De Luca, Emma Zucchelli, Sara D’Avino, and Antonio Ragusa 59 The  Role of Episiotomy in Emergency Delivery ���������������������������������������������������  893 Maddalena Falagario, Francesca Greco, Maristella De Padova, Maria Grazia Morena, Tea Palieri, Francesco D’Antonio, Michele Bollino, Felice Sorrentino, Lorenzo Vasciaveo, and Luigi Nappi 60 Obstetric  Errors: Sepsis and Shoulder Dystocia as Examples of Heuristic Thinking in Obstetrics�������������������������������������������������������������������������  915 Antonio Ragusa, Caterina De Luca, Sara D’Avino, Emma Zucchelli, and Alessandro Svelato 61 Importance  of Simulation to Avoid Childbirth Trauma���������������������������������������  927 Vasily V. Vlasyuk Part XIII Appendix 62 The  Role of Images in Obstetric Teaching: Past, Present, and Future ���������������  947 Antonio Dell’Aquila, Vito Maurizio Malvasi, Giuseppe Lupica, Mariaelena Malvasi, Renata Beck, and Antonio Malvasi 63 Importance  of Skill, Learning Curve, and Simulation in Endoscopy �����������������  977 Camran Nezhat 64 Fibroids  in Obstetric and Gynecology: Training and Skill in Myomectomy�������  981 Andrea Tinelli, Marina Vinciguerra, Radmila Sparić, Şafak Hatırnaz, Oğuz Güler, Ioannis Kosmas, Kyriaki Spyropoulou, and Michael Stark 65 Rupture  of the Uterus: A Dramatic Condition in a Genital Organ��������������������� 1027 Leonardo Resta, Gerardo Cazzato, Eliano Cascardi, and Roberta Rossi 66 Skills,  Learning Curve and Simulation in an Italian University Clinic��������������� 1031 Marica Falini, Simona Freddio, Antonio Malvasi, and Sandro Gerli

Contributors

Reuven Achiron, MD  Department of Obstetrics and Gynecology, Prenatal Diagnosis Unit, Chaim Sheba Medical Center, Tel Hashomer, Israel Sakler School of Medicine, Tel Aviv University, Tel Aviv, Israel Laura  Adamo, MD Department of Obstetrics and Gynecology, IRCCS Fondazione Policlinico San Matteo, University of Pavia, Pavia, Italy Panos  Antsaklis, MD Department of Obstetrics and Gynecology, University of Athens, Athens, Greece Ayse Gulsah Atasever, MD  Department of Anesthesiology and Intensive Care, Gaziosmanpasa Research and Training Hospital, Istanbul, Turkey Department of Anesthesiology, University Hospitals Leuven, Leuven, Belgium Apostolos  Athanasiadis, MD 3rd Department of Obstetrics and Gynecology, Ippokrateio Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece Haitham Baghlaf, MD  Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Mount Sinai Hospital, Toronto, ON, Canada Domenico Baldini  MOMO Fertilife-IVF Center, Bisceglie, Italy Christian  Bamberg, MD, PhD Department of Obstetrics and Fetal Medicine, University Medical Centre Hamburg-Eppendorf, Hamburg, Germany Sergey V. Barinov, MD, PhD  Department of Obstetrics and Gynecology, Omsk State Medical University, Omsk, Russia Nicolás Andrés Basanta, MD  Department of Obstetrics and Gynecology, Juan A. Fernández Hospital, Buenos Aires, Argentina School of Medicine, University of Buenos Aires, Buenos Aires, Argentina Renata Beck, MD  Department of Anesthesia and Intensive Care, Riuniti University Hospital, University of Foggia, Foggia, Italy Dragan Belci, MD  Department of Gynaecology and Obstetrics, General Hospital Pula, Pula, Croatia The New European Surgical Academy (NESA), Berlin, Germany Smaranda  Belciug Department of Computer Science, Faculty of Sciences, University of Craiova, Craiova, Romania Antonio Belpiede, MD  Apulia Region Birth Points Committee, Apulia, Italy Dan  Benhamou, MD, PhD Service d’Anesthésie Réanimation Médecine Péri Opératoire AP-HP, Hôpital Bicêtre, Université Paris Saclay, Le Kremlin Bicêtre Cedex, France

xvii

xviii

Aleksander V. Bindyuk, MD  Obstetric Physiological Department, Perinatal Centre, Budget Healthcare Institution of the Omsk Region, Regional Clinical Hospital, Omsk, Russia Katarina  Bojanić, MD, PhD Division of Neonatology, Department of Obstetrics and Gynecology, University Hospital Merkur, Zagreb, Croatia Michele  Bollino, MD Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Carlota Borrero González, MD  Department of Obstetrics and Gynecology, Valme University Hospital, Seville, Spain Joseph  Bouganim, MD Department of Obstetrics, Gynecology and Women’s Health, Montefiore Medical Center, Albert Einstein College of Medicine, New York, NY, USA Agostino  Brizzi, MD Department of Anesthesia and Intensive Care Unit, Santa Maria Hospital, GVM Care and Research, Bari, Italy Emanuela Cagnazzo  Faculty of Biology, University of Bari, Bari, Italy Tomislav Ćaleta, MD  Department of Pediatrics, University Hospital Centre Zagreb, School of Medicine, University of Zagreb, Zagreb, Croatia Giuseppe Calì, MD  Obstetrics and Gynecology Unit, V. Cervello Hospital, Palermo, Italy ARNAS Civico, Palermo, Italy Federica Calò, MD  Obstetrics and Gynecology Unit, V. Cervello Hospital, Palermo, Italy Davide  Campanelli, MD  Obstetrics and Gynecology Unit, Regional Hospital, San Paolo, ASL, Bari, Italy Nicola Caporale, MD  Department of Woman and Child Health and Public Health, Fondazione Policlinico Universitario A. Gemelli, IRCCS, Rome, Italy Jvan Casarin, MD  Department of Obstetrics and Gynecology, “Filippo Del Ponte” Women and Children Hospital, University of Insubria, Varese, Italy Eliano Cascardi, MD  Department of Medical Sciences, University of Turin, Turin, Italy Pathology Unit, FPO-IRCCS Candiolo Cancer Institute, Candiolo, Italy Laura  Castro  Portillo, MD  Department of Obstetrics and Gynecology, Valme University Hospital, Seville, Spain Gerardo Cazzato, MD  Dipartimento dell’Emergenza e dei Trapianti d’Organo (DETO), Sez. di Anatomia Patologica, Bari, Italy Università degli Studi “Aldo Moro”, Bari, Italy Ettore  Cicinelli, MD, PhD Department of Biomedical and Human Oncological Science (DIMO), Unit of Obstetrics and Gynecology, University of Bari, Bari, Italy Postgraduate School of Obstetrics and Gynecology, University of Bari, Bari, Italy Gilda Cinnella, MD, PhD  Professor of Anesthesia and Intensive Care, University of Foggia, Foggia, Italy Head of Anesthesia and Intensive Care Department, Riuniti University Hospital, Foggia, Italy Christopher  Clark, MD Department of Biomedical Science and Human Oncology, University of Bari, Bari, Italy Antonella Cotoia, MD, PhD  Department of Anesthesia and Intensive Care, Riuniti University Hospital, University of Foggia, Foggia, Italy Claudio Crescini, MD  AOGOI and Confalonieri Ragonese Foundation, Milan, Italy

Contributors

Contributors

xix

Antonella  Cromi, MD Department of Obstetrics and Gynecology, “Filippo Del Ponte” Women and Children Hospital, University of Insubria, Varese, Italy Francesco D’Antonio, MD  Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Sara D’Avino, MD  Department of Obstetrics and Gynecology, Fatebenefratelli Isola Tiberina Hospital, Gemelli Isola, Rome, Italy Themistoklis  Dagklis, MD 3d Department of Obstetrics and Gynecology, Ippokrateio Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece Andrea  Dall’Asta, MD, PhD Department of Medicine and Surgery, Obstetrics and Gynecology Unit, University of Parma, Parma, Italy Department of Metabolism, Digestion and Reproduction, Institute of Reproductive and Developmental Biology, Imperial College London, London, UK Pierpaolo Dambruoso, MD  Department of Anesthesia and Intensive Care Unit, Santa Maria Hospital, GVM Care and Research, Bari, Italy Gianluca  Raffaello  Damiani, MD Unit of Obstetrics and Gynaecology, Department of Biomedical Sciences and Human Oncology, Bari, Italy Angelos Daniilidis, MD  2nd Department of Obstetrics and Gynecology, Ippokrateio Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece Chen  Ben  David, MD Department of Obstetrics and Gynecology, Rambam Health Care Campus, Haifa, Israel Caterina  De Luca, MD  Department of Obstetrics and Gynecology, Fatebenefratelli Isola Tiberina Hospital, Gemelli Isola, Rome, Italy Francesca De Maria, MD  Obstetrics and Gynecology Unit, V. Cervello Hospital, Palermo, Italy Maristella De Padova, MD  Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Daniele De Viti, MD  Department of Cardiology, Cardiac Surgery and Intensive Care Unit, Santa Maria Hospital, GVM Care and Research, Bari, Italy Raffaella  Del Papato, MD  II UO of Obstetrics and Gynecology, Policlinico Universitario Bari, Bari, Italy Antonio Dell’Aquila  Pandosia Medical Graphic Academy, Bari, Italy Miriam  Dellino, MD  II UO of Obstetrics and Gynecology, Policlinico Universitario Bari, Bari, Italy Edoardo  Di Naro, MD Department Interdisciplinary Medicine, Unit of Obstetrics and Gynecology, University Hospital Policlinico of Bari, Bari, Italy University of Bari “Aldo Moro”, Bari, Italy Giovanni Di Vagno, MD  Obstetric and Gynecologic Unit, San Paolo Hospital, ASL, Bari, Italy Signe  Egenberg, MD Department of Obstetrics and Gynecology, Stavanger University Hospital, Stavanger, Norway Elisa Ervas, MD  Department of Obstetrics and Gynecology, “Filippo Del Ponte” Women and Children’s Hospital, University of Insubria, Varese, Italy

xx

Fatima Estrada Trejo, MD, FACOG  Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Women’s Health, Montefiore Medical Center, Bronx, NY, USA Albert Einstein College of Medicine, New York, NY, USA Cherrie Evans, MD  Technical Leadership Office, Jhpiego, Baltimore, MD, USA Maddalena  Falagario, MD Department of Obstetrics and Gynecology and Medical and Surgical Sciences, University of Foggia, Foggia, Italy Marica Falini  Department of Medicine and Surgery, Section of Obstetrics and Gynecology, Perugia University, Perugia, Italy Dan  Farine, MD, FRCSC, FACOG  Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Mount Sinai Hospital, University of Toronto, Toronto, ON, Canada Giuseppe  Ferrara, MD  Department of Anesthesia and Intensive Care, Riuniti University Hospital, University of Foggia, Foggia, Italy Alessandra  Ferrari, MD  Obstetrics and Gynecology Unit, Regional Hospital, San Paolo, ASL, Bari, Italy Flavio Fiore, MD  Department of Anesthesia and Intensive Care Unit, Anthea Hospital, GVM Care and Research, Bari, Italy Annalisa Fracchiolla, MD  Neonatal Intensive Care Unit, Policlinico Riuniti Foggia, Foggia, Italy Simona Freddio  Department of Medicine and Surgery, Section of Obstetrics and Gynecology, Perugia University, Perugia, Italy José A. García Mejido, MD  Department of Obstetrics and Gynecology, Valme University Hospital, Seville, Spain Judith  Gerets, MD  Department of Anesthesiology, University Hospitals Leuven, Leuven, Belgium Sandro  Gerli, MD Department of Medicine and Surgery, Section of Obstetrics and Gynecology, Perugia University, Perugia, Italy Ioana Andreea Gheonea  Department of Radiology and Imagistics, University of Medicine and Pharmacy Craiova, Craiova, Romania Tullio  Ghi, MD Department of Medicine and Surgery, Obstetrics and Gynecology Unit, University of Parma, Parma, Italy Alexis  C.  Gimovsky, MD, PhD Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Women & Infants Hospital of Rhode Island, Alpert Medical School of Brown University, Providence, RI, USA Reyyan  Gökçen  İşcan, MD Department of Obstetrics and Gynecology, Zeynep Kamil Women and Children’s Health Training and Research Hospital, İstanbul, Turkey Koray Gorkem Sacinti, MD  Department of Obstetrics and Gynecology, Ankara University, Ankara, Turkey Francesca  Greco, MD Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Ruža  Grizelj, MD, PhD Department of Pediatrics, University Hospital Centre, Zagreb, Croatia School of Medicine, University of Zagreb, Zagreb, Croatia Oğuz Güler, MD  Bilge Hospital, Istanbul, Turkey

Contributors

Contributors

xxi

Sarah  Gustapane, MD Department of Obstetrics and Gynecology, “Veris delli Ponti Hospital”, Scorrano, LE, Italy CERICSAL (CEntro di RIcerca Clinico SALentino), “Veris delli Ponti Hospital”, Scorrano, LE, Italy Dubravko Habek, MD, PhD  University Department of Obstetrics and Gynecology, Clinical Hospital “Sveti Duh”, Zagreb, Croatia School of Medicine, Catholic University of Croatia, Zagreb, Croatia Wassim  A.  Hassan, MD, PhD Fetal Medicine Unit, Department of Obstetrics and Gynaecology, Colchester Hospital, East Suffolk and North Essex Foundation Trust, Colchester, UK Department of Surgery and Cancer, Imperial College London, London, UK Şafak  Hatırnaz, MD, PhD IVF-IVM and Reproductive Endocrinology Unit, Medicinal Samsun International Hospital, Samsun, Turkey Dominic Gabriel Iliescu  Department of Obstetrics and Gynecology, University of Medicine and Pharmacy Craiova, Craiova, Romania 2nd Clinic of Obstetrics and Gynecology, University Emergency County Hospital Craiova, Craiova, Romania Khaled  M.  Ismail, MSc, MD, PhD  Obstetrics and Gynaecology, Faculty of Medicine in Pilsen, Charles University, Pilsen, Czech Republic Bénédicte  Jeannin, MD Service d’Anesthésie Réanimation Médecine Péri Opératoire AP-HP, Hôpital Bicêtre, Université Paris Saclay, Le Kremlin Bicêtre Cedex, France Dora  Jelinek, MD Division of Neonatology, Department of Obstetrics and Gynecology, University Hospital Merkur, Zagreb, Croatia Rocio  Garcia  Jimenez, MD  Department of Obstetrics and Gynecology, Valme University Hospital, Seville, Spain Tatyana  V.  Kadsyna, MD Department of Obstetrics and Gynecology, No. 2 Omsk State Medical University, Omsk, Russia Erkan  Kalafat, MD, PhD  Department of Obstetrics and Gynecology, Ankara University, Ankara, Turkey Rasha  A.  Kamel, MD, PhD Obstetrics and Gynaecology, Maternal-Fetal Medicine Unit, Department of Obstetrics and Gynecology, Cairo University, Cairo, Egypt Resul  Karakuş, MD, PhD Department of Obstetrics and Gynecology, Zeynep Kamil Maternity and Children’s Health Training & Research Hospital, İstanbul, Turkey Sultan Seren Karakus, MD, PhD  Zeynep Kamil Women’s and Children’s Disease Training and Research Hospital, İstanbul, Turkey Çetin  Kılıççı, MD, PhD, SBU  Department of Obstetrics and Gynecology, Zeynep Kamil Maternity Hospital, Istanbul, Turkey Ioannis  Kosmas, MD Department of Obstetrics and Gynecology, Ioannina State General Hospital G. Chatzikosta, Ioannina, Greece Kfier  Kuba, MD, FACOG Department of Obstetrics, Gynecology and Women’s Health, Montefiore Medical Center, Albert Einstein College of Medicine, New York, NY, USA Francesco  Labate, MD Obstetrics and Gynecology Unit, V.  Cervello Hospital, Palermo, Italy

xxii

Nicola Laforgia, MD  Neonatology and NICU, Azienda Ospedaliera Universitaria Policlinico di Bari, Bari, Italy University of Bari “Aldo Moro”, Bari, Italy Antonio Simone Laganà, MD, PhD  Unit of Gynecologic Oncology, ARNAS “Civico – Di Cristina –Benfratelli”, Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (PROMISE), University of Palermo, Palermo, Italy Bruno Lamanna, MD  Obstetric and Gynecologic Unit, San Paolo Hospital, ASL, Bari, Italy Brennan Lang, MD  Department of Obstetrics and Gynecology, Baylor College of Medicine, Houston, TX, USA Julia Lawrence, MSHP, RRT-NPS  Texas Children’s Hospital Simulation Center, Houston, TX, USA Oksana  V.  Lazareva, MD Department of Obstetrics and Gynecology, No. 2 Omsk State Medical University, Omsk, Russia Antonio Lembo, MD  Department of Obstetrics and Gynecology, “Filippo Del Ponte” Women and Children Hospital, University of Insubria, Varese, Italy Susan Leong-Kee, MD, PhD  Department of Obstetrics and Gynecology, Baylor College of Medicine, Houston, TX, USA Texas Children’s Hospital Simulation Center, Houston, TX, USA Marcella Lerro, MD  Obstetric and Gynecologic Unit, San Paolo Hospital, ASL, Bari, Italy Luigi Liaci, MD  Obstetrics and Gynecology Unit, Regional Hospital, San Paolo, ASL, Bari, Italy Nancy Loco  School of Medicine, Catholic University of Croatia, Zagreb, Croatia Maria Teresa Loverro, MD  Department of Obstetrics and Gynecology, University Hospital Policlinico of Bari, Bari, Italy School of Medicine, University of Bari “Aldo Moro”, Bari, Italy Matteo Loverro, MD, PhD  Department of Women and Child Health, Fondazione Policlinico Universitario Agostino Gemelli, IRCCS, Rome, Italy Miha  Lučovnik, MD, PhD Division of Obstetrics and Gynecology, Department of Perinatology, University Medical Center Ljubljana, Ljubljana, Slovenia Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia Giuseppe Lupica, MD  Department of Biomedical Sciences and Human Oncology, Unit of Obstetrics and Gynecology, Bari, Italy Gianfranco Maffei, MD  Neonatal Intensive Care Unit, Riuniti University Hospital, Foggia, Italy Lorenzo E. Malgieri  FIAT-ENI, Milan, Italy Environmental Companies, Milan, Italy Chief Innovation Officer in CLE, Bari, Italy Antonio  Malvasi, MD, PhD  Department of Biomedical and Human Oncological Science (DIMO), Unit of Obstetrics and Gynecology, University of Bari, Bari, Italy Santa Maria della Misericordia Hospital Perugia, University of Perugia, Perugia, Italy International Translational Medicine and Biomodelling Research Group, Department of Applied Mathematics, Moscow Institute of Physics and Technology (State University), Moscow, Russia The New European Surgical Academy (NESA), Berlin, Germany

Contributors

Contributors

xxiii

Mariaelena Malvasi, MD  Faculty of Medicine and Dentistry, Sapienza University of Rome, Rome, Italy Vito  Maurizio  Malvasi  Faculty of Medicine and Dentistry, Sapienza University of Rome, Rome, Italy Apostolos  Mamopoulos, MD 3d Department of Obstetrics and Gynecology, Ippokrateio Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece Vito  Margari, MD  Department of Cardiac Surgery, Santa Maria Hospital, GVM Care & Research, Bari, Italy Enrico  Marinelli, MD  Department of Anatomical, Histological, Forensic and Orthopedic Sciences, Sapienza University of Rome, Rome, Italy Ingrid Marton, MD, PhD  Department of Obstetrics and Gynecology, Clinical Hospital Sveti duh, Zagreb University School of Medicine, Zagreb, Croatia Croatian Catholic University, Zagreb, Croatia Frederik  Marynen, MD Department of Anesthesiology, University Hospitals Leuven, Leuven, Belgium Cynthia Maxwell, MD  Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Mount Sinai Hospital, Toronto, ON, Canada Irina V. Medyannikova, MD  Department of Obstetrics and Gynecology, No. 2 Omsk State Medical University, Omsk, Russia Laura Messina, MD  Obstetrics and Gynecology Unit, V. Cervello Hospital, Palermo, Italy Maria Mina, MD  Department of Obstetrics and Gynecology, Ioannina State General Hospital G. Chatzikosta, Ioannina, Greece Monica  Minopoli, MD  Department of Medicine and Surgery, Obstetrics and Gynecology Unit, University of Parma, Parma, Italy Lucia Mirabella, MD, PhD  Department of Anesthesia and Intensive Care, Riuniti University Hospital, University of Foggia, Foggia, Italy Maria Grazia Morena, MD  Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Ospan Mynbaev, PhD, ScD  Moscow Institute of Physics and Technology, National Research University, Dolgoprudny, Moscow, Russia Department of Traumatology, Orthopedics and Oncology, Khoja Akhmet Yassawi International Kazakh-Turkish University, Turkistan, Kazakhstan The New European Surgical Academy (NESA), Berlin, Germany Editorial Advisor, SCIRP Inc, Irvine, CA, USA Luigi Nappi, MD, PhD  Departments of Obstetrics and Gynecology and Medical and Surgical Sciences, Riuniti University Hospital, University of Foggia, Foggia, Italy Camran  Nezhat, MD, FACOG, FACS  University of California San Francisco School of Medicine, San Francisco, CA, USA Stanford University Medical Center, Stanford, CA, USA Medical University of Vienna, Vienna, Austria American Society of Reproductive Medicine (ASRM), Washington, DC, USA Society of Laparoscopic and Robotic Surgeons (SLS), Miami, FL, USA Center for Special Minimally Invasive and Robotic Surgery, Woodside, CA, USA Camran Nezhat Institute, Woodside, CA, USA

xxiv

Álbaro  José  Nieto-Calvache, MD Department of Obstetrics, Placenta Accreta Spectrum Clinic. Fundación Valle de Lili, University Hospital, Cali, Colombia Clinical Postgraduate Department, Universidad ICESI, Cali, Colombia Shirley  Nilsen Department of Obstetrics and Gynecology, Stavanger University Hospital, Stavanger, Norway Cansu  Ofluoglu, MD Department of Anesthesiology and Intensive Care, Fatih Sultan Mehmet Training and Research Hospital, University of Health Sciences, Istanbul, Turkey Pelin  Özdemir  Önder, MD Department of Obstetrics and Gynecology, Zeynep Kamil Maternity Hospital, Istanbul, Turkey José M. Palacios-Jaraquemada, MD, PhD  Obstetrics and Gynecology Department, CEMIC University Hospital, Buenos Aires, Argentina School of Medicine, University of Buenos Aires, Buenos Aires, Argentina Tea Palieri, MD  Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Yoav Paltieli, MD, PhD  Trig Medical Ltd, Haifa, Israel Maria Papamichail, MD  Department of Obstetrics and Gynecology, University of Athens, Athens, Greece Domenico Paparella, MD, PhD  Department of Medical and Surgical Science, University of Foggia, Foggia, Italy Cardiac Surgery Hospital Santa Maria, GVM Care & Research, Bari, Italy Minas Paschopoulos, MD  Department of Obstetrics and Gynecology, University Hospital of Ioannina, Ioannina, Greece Luis M. Pastor Colomer, MD  Department of Obstetrics and Gynecology, Valme University Hospital, Seville, Spain Jadranka  Pavičić  Šarić, MD, PhD Department of Anaesthesiology and Intensive Care, University Hospital Merkur, Zagreb, Croatia Marco Paolo Perrini, MD  Department of Anesthesia and Intensive Care, Riuniti University Hospital, University of Foggia, Foggia, Italy Nico Picardi  II UO of Obstetrics and Gynecology, Policlinico Universitario Bari, Bari, Italy Giuseppe Piccinni, MD, PhD  Interdisciplinary Department of Medicine, University of Bari, Bari, Italy Antonio Pipoli, MD  Stayin Alive Training and Simulation Center, Monopoli, Italy Antonio  Ragusa, MD Department of Obstetrics and Gynecology, Campus Bio Medico University, Rome, Italy Pasquale Raimondo, MD  Department of Anesthesia and Intensive Care, Azienda Ospedaliera Universitaria Consorziale Policlinico, Bari, Italy Leonardo Resta, MD, PhD  Division Dipartimento dell’Emergenza e dei Trapianti, University d’Organo (DETO), Bari, Italy University of Bari “Aldo Moro”, Bari, Italy Alemnesh Reta  Laerdal Global Health, Stavanger, Norway

Contributors

Contributors

xxv

Ilaria Ricci, MD  II UO of Obstetrics and Gynecology, Policlinico Universitario Bari, Bari, Italy Matteo Rinaldi, MD  Neonatal Intensive Care Unit, Riuniti University Hospital, Foggia, Italy Roberta Rossi, MD  Dipartimento dell’Emergenza e dei Trapianti d’Organo (DETO), Sez. di Anatomia Patologica, Bari, Italy Università degli Studi “Aldo Moro”, Bari, Italy José  Antonio  Sainz-Bueno, MD, PhD  Department of Obstetrics and Gynecology, Valme University Hospital, Seville, Spain Department of Obstetrics and Gynecology, University of Seville, Seville, Spain Potito Salatto, MD  Department of Anesthesia and Intensive Care, Riuniti University Hospital, University of Foggia, Foggia, Italy Styliani Salta, MD  University Hospitals of Leicester, Haemophilia Centre, Leicester Royal Infirmary, Leicester, UK Alessandro  Savino, MD Obstetrics and Gynecology Unit, Regional Hospital, San Paolo, ASL, Bari, Italy Nada  Sindičić  Dessardo, MD PhD  Department of Pediatrics, University Hospital Centre Zagreb, School of Medicine, University of Zagreb, Zagreb, Croatia Ido Solt, MD  Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Rambam Health Care Campus, The Rappaport Faculty of Medicine, Technion, Haifa, Israel Jette  Led  Sørensen Section 4074 Juliane Marie Centre, Rigshospitalet, Copenhagen University, Copenhagen, Denmark Felice  Sorrentino, MD Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Radmila Sparić, MD, PhD  Clinic for Gynecology and Obstetrics, Clinical Centre of Serbia, Belgrade, Serbia Medical Faculty, University of Belgrade, Belgrade, Serbia Chiara Spina, MD  Stayin Alive Training and Simulation Center, Monopoli, Italy Marta  Spinelli, MD II UO of Obstetrics and Gynecology, Policlinico Universitario Bari, Bari, Italy Kyriaki  Spyropoulou, MD Department of Obstetrics and Gynecology, Ioannina State General Hospital, Ioannina, Greece Michael Stark, MD, PhD  The New European Surgical Academy, Berlin, Germany ELSAN Hospital Group, Paris, France Lili  Steblovnik, MD, MSc Division of Obstetrics and Gynecology, Department of Perinatology, University Medical Center Ljubljana, Ljubljana, Slovenia Assunta  Stragapede, MD Department of Biomedical Sciences and Human Oncology, Section of Internal Medicine, University of Bari, Bari, Italy Alessandro  Svelato, MD  Department of Obstetrics and Gynecology, Fatebenefratelli Isola Tiberina Hospital, Gemelli Isola, Rome, Italy Kübra Taşkın, MD  Department of Anesthesiology and Intensive Care, Zeynep Woman and Children Training and Research Hospital, University of Health Sciences, Istanbul, Turkey

xxvi

Sasha Taylor, MD, PhD  Department of Obstetrics and Gynaecology, West Suffolk Hospital, West Suffolk NHS Foundation Trust, Bury St Edmunds, UK Andrea  Tinelli, MD, PhD Division of Experimental Endoscopic Surgery, Imaging, Technology and Minimally Invasive Therapy, Department of Obstetrics and Gynecology, Vito Fazzi Hospital, Lecce, Italy Laboratory of Human Physiology, Phystech BioMed School, Faculty of Biological & Medical Physics, Moscow Institute of Physics and Technology (State University), Moscow, Russia Yulya I. Tirskaya, MD  Department of Obstetrics and Gynecology, No. 2 Omsk State Medical University, Omsk, Russia Önder Tosun, MD  Department of Obstetrics and Gynecology, Zeynep Kamil Maternity and Children’s Health Training and Research Hospital, İstanbul, Turkey Giuseppe Trojano, MD  Department of Obstetrics and Gynecology, Madonna delle Grazie Hospital, ASM Matera, Matera, Italy Ioannis Tsakiridis, MD  3d Department of Obstetrics and Gynecology, Ippokrateio Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece Nataša Tul, MD, PhD  Women’s Hospital, Postojna, Slovenia Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia Anna af Ugglas  Laerdal Global Health, Stavanger, Norway Elke Van Gerven, MD  Department of Anesthesiology, University Hospitals Leuven, Leuven, Belgium Marc Van De Velde, MD, PhD, EDRA  Department of Anesthesiology, University Hospitals, Leuven, Belgium Lorenzo  Vasciaveo, MD  Department of Obstetrics and Gynecology, University of Foggia, Foggia, Italy Anastasia  Vatopoulou, MD 3d Department of Obstetrics and Gynecology, Ippokrateio Hospital, Aristotle University of Thessaloniki, Thessaloniki, Greece Antonella  Vimercati, MD, PhD II UO of Obstetrics and Gynecology, Department of Biomedical Sciences and Human Oncology, University of Bari “Aldo Moro”, Bari, Italy Katja  Vince, MD  Department of Gynecology and Obstetrics, University Hospital Merkur, Zagreb, Croatia Marina Vinciguerra, MD, PhD  Obstetric and Gynecologic Unit, San Paolo Hospital, ASL, Bari, Italy COVID Center, ASL, Bari, Italy Department of Obstetrics and Gynecology, School of Medicine, Azienda Ospedaliera Universitaria Policlinico di Bari, Bari, Italy University of Bari Aldo Moro, Bari, Italy Vasily V. Vlasyuk, MD, PhD  Department of Forensic Medicine, S. M. Kirov Military Medical Academy, Saint Petersburg, Russia Tal  Weissbach, MD Department of Obstetrics and Gynecology, Prenatal Diagnosis Unit, Chaim Sheba Medical Center, Tel Hashomer, Israel Sakler School of Medicine, Tel Aviv University, Tel Aviv, Israel

Contributors

Contributors

xxvii

Rosanna Zaccaro, MD  Obstetric and Gynecologic Unit, San Paolo Hospital, ASL, Bari, Italy Lada  Zibar, MD, PhD  Department for Nephrology, University Hospital Merkur, Zagreb, Croatia Faculty of Medicine, University Josip Juraj Strossmayer in Osijek, Osijek, Croatia School of Medicine, University of Zagreb, Zagreb, Croatia Giulia  Zingarelli, MD Department of Anesthesia and Intensive Care, Riuniti University Hospital, University of Foggia, Foggia, Italy Emma  Zucchelli, MD Department of Obstetrics and Gynecology, Campus Bio Medico University, Rome, Italy

Part I Fundamentals of Simulation

1

Simulation in Obstetric: From the History to the Modern Applications Reuven Achiron, Laura Adamo, and Tal Weissbach

1.1 Introduction

1.2 History of Obstetrical Simulation

The term simulation derived from the Latin word simulo, meaning to pretend or imitate. According to common dictionaries, the definition of the word “simulation” is “a situation in which a particular set of conditions is created artificially in order to study or experience something that could exist in reality” [1]. Simulation is integrated in many aspects of our lives: entertainment, military, financial to cite some, and of course medicine. Simulation in the medical field has advantages that are now widely recognized. Simulation can teach, test, and prepare for important clinical scenarios. Besides facilitating the acquisition of new skills and the maintaining of previously acquired skills, simulation assists in gaining experience in the management of emergencies and life-threatening conditions without inflicting harm to the patient. The Society for Simulation in Healthcare defined simulation as “an educational technique that replaces or amplifies real experiences with guided experiences that evoke or replicate substantial aspects of the real world in a fully interactive manner” [2]. Simulation in obstetrics has developed from a necessity to improve medical student’s education to an integral working tool for every level of expertise, recognizing the importance of maintaining high standard level of care, as an individual clinician or a team of caregivers.

A PubMed search of the words “simulation” and “obstetrics” led to more than 5000 articles, most of which have been published within the last 10 years. Despite recent interest, obstetric simulators have been already documented since the seventeenth century and, considering simulation in general medical education, its roots are dated millennia ago [3]. In 1027, the Chinese physician Wang Wei-Yi (987–1067), standardized the teaching of acupuncture with two life-size statues made of bronze, that had more than 300 holes to demonstrate to the students the locations of acupuncture points [4] (Fig. 1.1). In Europe, following the end of the Middle Ages, a new era of Renaissance emerged with a strong desire of knowledge and innovation. De humani corporis fabrica libri septum written by Andrea Vesalio (1514–1564) in 1543, revolutionized the interest of anatomy studies [5] (Fig. 1.2) and half a century later, Ludovico Cardi (1559–1613) produced the first wax anatomical model. These carved figures became a common teaching tool throughout Europe and typically presented either male or female organs. The female version was often featured as pregnant including a fetus attached to the mother by a red silk string as an umbilical cord [6]. Giovanni Antonio Galli (1585–1652), a surgeon based in Bologna, recognized the importance of increasing midwive’s education since often they were lacking essential knowledge and technical skills [3] (Fig. 1.3). He designed one of the first birth simulators which included a pelvis containing a glass uterus with a flexible fetus. Galli’s simulator and other obstetric teaching models can still be admired at Musei di Palazzo Poggi in Bologna.

R. Achiron · T. Weissbach Department of Obstetrics and Gynecology, Prenatal Diagnosis Unit, Chaim Sheba Medical Center, Tel Hashomer, Israel Sakler School of medicine, Tel Aviv University, Tel Aviv, Israel L. Adamo (*) Department of Obstetrics and Gynecology, IRCCS Fondazione Policlinico San Matteo, University of Pavia, Pavia, Italy

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 G. Cinnella et al. (eds.), Practical Guide to Simulation in Delivery Room Emergencies, https://doi.org/10.1007/978-3-031-10067-3_1

3

4

R. Achiron et al.

Fig. 1.1  Acupuncture statue used by Wang Wei-Yi in China during the eleventh century

The role of obstetric simulation became increasingly more popular throughout European countries and various models were created for teaching and developing skills. By the middle of seventeenth century, Angelique Marguerite Le Boursier du Coudray (1712–1794) was summoned by King Louis XV to educate midwives to ­ decrease intrapartum mortality in rural France. She developed innovative female pelvis simulators, with interchangeable cervices to assess different cervical dilations

and with different-sized fetuses. These high-fidelity mannequins were also able to emulate rupture of membranes and hemorrhage. Madame Du Coudray incorporated the practice of simulation with traditional frontal lectures, developing an instructional course of 40 lessons addressing management of labor and its complications [7] (Fig. 1.4). Contemporaries to madame Du Coudray, two surgeons, father and son, the Gregoires made their own obstetric simu-

1  Simulation in Obstetric: From the History to the Modern Applications

5

Fig. 1.2  De humani corporis fabrica libri septum consecrated Andrea Vesalio as the father of modern anatomy

lator known as the “phantom” using a human cadaver pelvis, a woven leather uterus, and deceased neonates as fetuses [8]. One of the most famous Gregoire pupil was the Scottish William Smellie (1697–1763), known for his studies on pelvis deformities and for vaginal assessment of the obstetric conjugate [9]. Once back in the United Kingdom, Smellie decided to design an improved version of the Gregoire phantom. In order to avoid using cadavers for training, his phan-

tom was composed of human bones covered in leather, a fetus made of wood and rubber with articulating limbs and a placenta [3, 10] (Fig. 1.5). In 1831, there was a big improvement in mannequins technology, when Doctor Gustave Ozenne (1822–1871) presented to the French Royal Academy of Medicine a very sophisticated whole-body patient simulator that he worked on for 6 years.

6

R. Achiron et al.

Fig. 1.3  In 1757, Giovanni Antonio Galli professor of obstetrics in Bologna, created the Galli’s “machine” as a teaching tool for midwives and medical students

The uterus was made of longitudinal and radial fibers to simulate uterine contractions, with the possibility of changing the strength, rate, and rhythm. There was an amniotic sac and the fetus skull had fontanelles, a moveable lower jaw, and a rump to allow students to recognize different fetal pre-

sentations and practice uncomplicated deliveries. They could even protect the perineum from tearing. Ozenne reported the benefits of teaching the management of a physiologic labor, but also the opportunity to recreate obstructed labors and possible interventions [11].

1  Simulation in Obstetric: From the History to the Modern Applications

7

Fig. 1.4  In the middle of seventeenth century, Angelique Marguerite Le Boursier du Coudray developed high-fidelity mannequins that were able to emulate different cervical dilatations, rupture of membranes, and hemorrhage

Few years later, following Ozenne’s footsteps, Pierre Budin (1846–1906) and Adolphe Pinard (1844–1934) developed an internal and external digital palpation simulator, enabling to determine the fetal presentation and position as well as performing forceps application (Figs. 1.6 and 1.7). Obstetric pelvimetry and the study of birth mechanisms was a push forward for European obstetric school’s simulations (Fig. 1.8). The understanding of the pathological pelvis led to research and simulation of the descent of the fetal head within the birth canal such as Selheim’s theory of asynclitism (Fig. 1.9).

Throughout the 1800s, simulation was widely used all around Europe with various model types, some of them even included detailed instruction manuals, like the one designed by Professor Schultze at the University Women’s Hospital in Jena, Germany. It had interchangeable pelvic floors and sacral promontories for an improved pelvic anatomy simulation for teaching clinical pelvimetry. Moreover, Shultes Medacta, founded by Prof. Shultze, began large-scale phantom manufacturing from 1890, making it the oldest existing supplier of medical simulators [3].

8

R. Achiron et al.

Fig. 1.5  The Scottish William Smellie designed an improved version of the Gregoire phantom, using human bones covered in leather, a fetus made of wood and rubber with articulating limbs and a placenta

Overseas, in the United States, simulation began to be widely used in medical schools around the country to compensate for the lack of births in hospitals [12]. In 1910, Flexner published a report named Medical Education in the United States and Canada, which led to a reform in American medical education [13]. In particular, he perceived the mannequin as a useful tool for teaching and preparing students for clinical practice, admonishing some schools for making poor use of them.

However, despite the great advancements made in the field of obstetrics over the previous 50  years, simulators were still designed based on seventeenth century models (Figs. 1.10 and 1.11). Additionally, the increasing number of hospital deliveries led to a higher exposure to clinical practice, which consequently made the role of the simulation wane.

1  Simulation in Obstetric: From the History to the Modern Applications

Fig. 1.6  A draft of Budin–Pinard mannequin made by Maison Matieu and sons in Paris

9

10

Fig. 1.7  Mannequin used to teach the operative childbirth with forceps during the nineteenth century

R. Achiron et al.

1  Simulation in Obstetric: From the History to the Modern Applications

11

Fig. 1.8  Jean-Louis Baudelocque, a pioneer of pelvimetry

Fig. 1.9  Graphic simulation of the descent of the fetal head in anterior asynclitism according to Selheim

12

Fig. 1.10  Professor Theophilus Parvin’s mannequin, anterior view

R. Achiron et al.

1  Simulation in Obstetric: From the History to the Modern Applications

Fig. 1.11  Professor Theophilus Parvin’s mannequin, posterior and lateral view

13

14

R. Achiron et al.

1.3 The Twentieth Century Became a “Dark Age” for Simulation

1.4 The Role of Obstetrical Simulation Today

Only towards the end of the twentieth century, contemporary changes in technology and the heightened concern for patient safety, renewed the interest in simulation once again. In 1968, the first modern high-fidelity medical mannequin, named Harvey, was designed. It was able to simulate vital signs and heart sounds, thanks to computerized technology [14]. Harvey was an innovation that paved the road for the development of the modern-day obstetric simulators. Simulators were progressively becoming more realistic. They enabled training practitioners to visualize the descent of the fetus through the birth canal, to place forceps, to practice shoulder dystocia maneuvers. Simulation was focused on improving specific and confined practical skills. What was still lacking was the ability to recreate obstetric emergency scenarios in order to assess and improve teamwork efficacy. An airplane crash provided significant insights on the importance of team training. Investigators identified a lack of communication between the pilot and crew, which led to a wrong management of a malfunctioning light and distracted the pilot from identifying a lack of fuel. The aviation industry was already using flight simulators, however, there was still a need of creating programs aimed to increase collaboration between pilots and crew in identifying problems. It was clear that every member of the team was essential and has an individual responsibility which, when synchronized, could contribute to optimal management of critical situations [15]. In 2001, the first international meeting on medical simulation met as part of an anesthesiology technology conference. Three years later, the Society for Simulation in Healthcare (SSH) was founded [16]. In addition to the SSH, individual medical specialties created specific simulation working groups. Likewise, principal bodies in the field of Obstetrics and Gynecology began to acknowledge the importance of simulation. The American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) began to offer hands-on simulation courses during their annual meetings. The ACOG Simulations Consortium was created in 2009 with the aim of “establishing [simulation] as a pillar in education for women’s health through collaboration, advocacy, research, and the development and implementation of multidisciplinary simulation-based educational resources and opportunities for Obstetrics and Gynecology” [17].

Nowadays simulations can take many forms. Simulation can be a towel folded over a chair to represent a perineal laceration (low fidelity) and can be a breathing and bleeding robot in an immersion data cave (high fidelity). Simulation can be a role play with live actors, can take place in virtual reality, or it can be a tutorial on a desktop computer [18]. To a degree, simulators should be realistic in order to create real-life situations and working conditions. However, overly sophisticated technology could have an opposite effect, by over guiding the training practitioner and creating a nonrealistic situation. Macedonia et al. pointed out this concept and created the ARRON rule (As Reasonably Realistic as Objectively Needed). Several studies demonstrated the efficacy of the ARRON rule showing, for example, that simulation of shoulder dystocia can be managed on a low-fidelity mannequin [19] and that medical students do not need high-fidelity simulators to understand vaginal birth [20]. There are countless advantages to simulation-based training such as an organized learning environment, the ability to control clinical parameters, providing immediate feedback, and an objective method for assessing performance. Through simulation, skills can be practiced until mastered, without inflicting harm to patients. Simulation helps to acquire and refine both cognitive and technical skills necessary to perform complex patient care activities. It can be used to train complex decision making, to practice rare or acute clinical emergencies, and to learn and practice skilled maneuvers [21, 22]. Several studies have demonstrated the expanding roles of simulation in undergraduate medical education in the field of obstetrics and gynecology. Compared to traditional lectures, simulation programs in pelvic examination, cervical dilatation, and vaginal delivery, have been shown to be superior in terms of confidence, knowledge, skills, workplace behaviors, and translation to patient care [23–29]. Simulation can be beneficial for mastering any newly learnt procedure during residency. While senior clinicians might have accumulated experience and knowledge in obstetrical emergencies over time, residents are inexperienced and rely on standard medical education. Therefore, it is not surprising that interventions such as simulation and drills have shown a positive impact on clinical management and skill acquisition among obstetrics residents. Commonly reported topics in residency include operative vaginal delivery, breech delivery, twin delivery, cord pro-

1  Simulation in Obstetric: From the History to the Modern Applications

Fig. 1.12  Simulation of cord prolapse on mannequin in the School of Obstetrics and Gynecology of the University of Bari, Italy

lapse (Fig. 1.12) management of shoulder dystocia (Figs. 1.13 and 1.14), third- and fourth-degree laceration repair, management of eclampsia, and hemorrhage. For example, the decline in the number of breech deliveries has gravely affected the ability of contemporary residents to manage this type of event, leading to the development of simulation training in these and other rare conditions. Deering et  al. [30] have reported an improvement in residents’ management of a vaginal breech delivery after simulation. Easter et al. [31] described that after simulating a breech extraction on a nonvertex second twin, residents personal comfort improved from 5.5% to 66.7%. One of the most common procedures required in obstetrics is an operative delivery. A systematic review of eight studies suggested that operative vaginal delivery simulation is a promising tool to increase trainee skills, knowledge, and confidence, while also improving maternal and neonatal outcomes [32]. Similar observations were found when comparing the performance of residents after simulating the management of shoulder dystocia and eclampsia to those receiving standard didactic education [33, 34].

15

Fig. 1.13  Shoulder dystocia simulation conducted by the School of Obstetrics and Gynecology of the University of Bari, Italy

Fig. 1.14 Rubin maneuver to solve Shoulder dystocia on a mannequin

Simulation has a crucial role in emergency scenarios that require a quick yet appropriate response from medical providers.

16

R. Achiron et al.

In fact, Ellis et  al. have demonstrated that simulation training of eclampsia drills has enhanced team performance, increased the rate of task completion, and shortened the time to magnesium sulfate administration [38]. Shoulder dystocia and perimortem cesarean delivery are additional examples of emergencies that have benefited from the use of simulation [39]. The advantages of simulation training in improving clinical skills are unquestionable, but does it improve the clinical outcomes? Most of the studies presented so far have focused on education, describing the learning curve of performance measured on the corresponding simulator. Even if large randomized studies are lacking, there is evidence of clinical outcome. Over a 12-year period, Crofts et  al. [40] have demonstrated how the implementation of simulation training for shoulder dystocia management has led to a reduction in newborn brachial plexus injuries. Gossett et  al. [41] reported that a program focusing on Fig. 1.15 Difficult airway management simulation in obstetrics forceps-assisted vaginal delivery has led to a 26% reduction emergencies in severe perineal lacerations while increasing the proper use Often, emergency situations require a team of caregivers of forceps in labor room. Umbilical cord prolapse is another obstetric emergency working in a coordinated manner. Anesthesiologists, as part of the obstetric team, should that has exhibited an improvement in management as a result be trained in difficult airway management, for example, by of simulation. After simulation drills, a shorter time to delivuse of videolaryngoscopy with a special curved d blade ery and higher likelihood of cord compression alleviation maneuvers were demonstrated [42]. (Fig. 1.15). A half-day simulation-based training program in Commonly, acquiring hands-on clinical experience Tanzania has proven a 38% reduction in postpartum hemorthrough real-life emergency cases is limited. First, life-­ rhages. The hemorrhage rate drop was associated with a betthreatening situations always require intervention by the most skilled caregiver nearby, limiting less experienced co-­ ter performance of basic delivery skills and appropriate use workers to take full charge of the situation. Second, rare con- of oxytocin [43]. A Cochrane Library review in developing countries ditions that require special skills are less likely to be demonstrated how a standardized neonatal resuscitation confronted in reality [35]. training program resulted in the reduction of early neonaSimulating emergency scenarios has shown to improve tal and 28-day mortality, compared with basic newborn efficiency, response time, appropriate order of actions, teamcare [44]. work skills, and reduction of errors. A study conducted by the Department of Defense showed There are numerous perilous conditions and procedures in obstetrics that impact the health of both the parturient and how teamwork training was able to reduce the time to incineonate, thus require prompt action. These include operative sion for an immediate cesarean delivery from 33 to 21 min delivery, emergency cesarean section, shoulder dystocia, [45]. There are is a wealth of evidence to show the clear concord prolapse, postpartum hemorrhage, eclampsia, and tribution of simulation to clinical management. maternal cardiac arrest. A review by Deering and Rowland has shown the various types of models used for different emergency situations and 1.5 Future Perspectives the benefit gained [36]. For example, practicing cesarean section on mannequins has led to a better understanding of Simulation helps to overcome the limitations of current forthe different steps and, thereby, to an increased sense of con- mal medical education. During the upcoming years, we expect to see a wider use fidence by the performing clinician [37]. of simulation in the assessment of residency programs and Likewise, eclampsia management has been shown to board examinations. improve after simulation [34].

1  Simulation in Obstetric: From the History to the Modern Applications

Standardizing simulation programs will allow a more homogeneous acquisition of clinical skills and emergency scenario management. Ennen et  al. [46] have based five keys components for establishing an effective simulation program: 1. Identifying the target trainee 2. Recognizing the skills required to be tested 3. Determining the appropriate frequency of simulations 4. Determining the location and required equipment (ARRON rule) 5. Debriefing and analysis of the performance Some societies have already designed programs that incorporate simulation, however, most medical schools and hospitals do not regularly use it. The Royal College of Obstetricians and Gynecologists developed a specific birth simulation training course, named ROBuST, which emphasizes operative vaginal delivery including manual rotation and vacuum- or forceps-assisted delivery [47]. The Society for Maternal-Fetal Medicine supported John Hopkins Hospital in developing a video library of critical care scenarios. These video scenarios include pulmonary embolism, maternal cardiac arrest, hypertensive emergencies, eclampsia, severe sepsis with shock, pulmonary edema with acute respiratory distress syndrome, hemodynamic monitoring and mechanical ventilation through the ARDS net protocol, myocardial infarction, diabetic ketoacidosis, amniotic fluid embolism, Advanced Cardiac Life Support in pregnancy, massive hemorrhage and perimortem cesarean delivery [48]. While standardizing skills among physicians, each unit should use simulation to enhance their experience and efficacy. Identifying preexisting or potential errors of team emergency management is essential for reducing malpractice. For example, Maslovitz et al. reported how, during a postpartum hemorrhage simulation, an underestimation of blood loss led to a tardive prostaglandin treatment for uterine atony, which, in turn, has delayed patient transfer to the operating room and administration of blood products [35]. Use of simulation is the perfect setting to introduce, reinforce, and practice effective team performance. Currently, there are no standardized obstetric simulation courses that have been associated with improvement of clinical outcomes. Large multicenter trials are necessary to determine best practice and understand where simulation resources should be implemented.

1.6 Conclusions Simulation is not a substitute for clinical experience, but nonetheless is an essential additive to training and patient safety improvement [48].

17

Acquiring new skills and managing rare complications or emergency scenarios are subjected to a learning curve, which varies among individuals. Simulation cannot alter the innate learning potential; however, it can improve the learning process by exposing the individual to an optimal amount of practice.

References 1. Oxford Learning Dictionaries. n.d.. https://www.oxfordlearnersdictionaries.com/definition/english/simulation. 2. Lopreiato JO.  Healthcare simulation dictionary. Rockville, MD: Agency for Healthcare Research and Quality; 2016. AHRQ Publication No. 16(17)-0043. 3. Owen H.  Early use of simulation in medical education. Simul Healthc. 2012;7(2):102–16. https://doi.org/10.1097/ SIH.0b013e3182415a91. 4. Schnorrenberger C. Anatomical roots of acupuncture and Chinese medicine. Schweiz Z Ganzheitsmed. 2013;25:110–8. https://doi. org/10.1159/000349905. 5. Orlandini GE, Paternostro F. Anatomy and anatomists in Tuscany in the 17th century. Ital J Anat Embryol. 2010;115(3):167–74. PMID: 21287970. 6. Ballestriero R.  Anatomical models and wax Venuses: art masterpieces or scientific craft works? J Anat. 2010;216(2):223–34. https://doi.org/10.1111/j.1469-­7580.2009.01169.x. 7. Marzano D. The history of simulation in obstetrics and gynecology. In: Deering S, Auguste TC, Goffman D, editors. Comprehensive healthcare simulation: obstetrics and gynecology. Cham: Springer International Publishing; 2013. p. 4–5. 8. Buck GH.  Development of simulators in medical education. Gesnerus. 1991;48(Pt 1):7–28. PMID: 1855669. 9. Forster FMC.  William Smellie and his contribution to obstetrics. Aust New Zeal J Obstet Gynaecol. 1963;3(3):132–4. 10. Boyd G. William Smellie. Ulster Med J. 1958;27(1):29–36. 11. Owen H. Simulation in obstetrics, gynecology and midwifery. In: Owen H, editor. Simulation in healthcare education: an extensive history. Cham: Springer International Publishing; 2016. p. 79–81. 12. Owen H, Pelosi MA. A historical examination of the Budin-Pinard phantom: what can contemporary obstetrics education learn from simulators of the past? Acad Med. 2013;88(5):652–6. https://doi. org/10.1097/ACM.0b013e31828b0464. 13. Flexner A. Medical education in the United States and Canada. From the Carnegie Foundation for the Advancement of Teaching, Bulletin Number Four, 1910. Bull World Health Organ. 2002;80(7):594– 602. PMID: 12163926. 14. Gordon MS, Evy GA, DeLeon AC, Waugh RA, Felner JM, Forker AD, Gessner IH, Mayer JW, Patterson D. “Harvey”, the cardiology patient simulator: pilot studies on teaching effectiveness. Am J Cardiol. 1980;45(4):791–6. https://doi. org/10.1016/0002-­9149(80)90123-­X. 15. Helmreich R.  Managing human error in aviation. Sci Am. 1997;276(5):62–7. 16. American College of Obstetricians and Gynecologists. Simulations Working Group. n.d.. https://www.acog.org/education-­and-­events/ simulations/about. 17. Rosen KR.  The history of medical simulation. J Crit Care. 2008;23(2):157–66. https://doi.org/10.1016/j.jcrc.2007.12.004. 18. Macedonia CR, Gherman RB, Satin AJ.  Simulation laboratories for training in obstetrics and gynecology. Obstet Gynecol. 2003;102(2):388–92. https://doi.org/10.1016/ s0029-­7844(03)00483-­6. 19. Szymanski L, Arnold C, Vaught AJ, LaMantia S, Harris T, Satin AJ. Implementation of a multicenter shoulder dystocia injury prevention program. Semin Perinatol. 2017;41:187–94.

18 20. Sabourin JN, Van Thournout R, Jain V, Demianczuk N, Flood C.  Confidence in performing normal vaginal delivery in the obstetrics clerkship: a randomized trial of two simulators. J Obstet Gynaecol Can. 2014;36(7):620–7. https://doi.org/10.1016/ S1701-­2163(15)30542-­9. 21. Andreatta PB, Bullough AS, Marzano D.  Simulation and team training. Clin Obstet Gynecol. 2010;53(3):532–44. https://doi. org/10.1097/GRF.0b013e3181ec1a48. 22. Everett EN, Forstein DA, Bliss S, Buery-Joyner SD, Craig LB, Graziano SC, Hampton BS, Hopkins L, McKenzie ML, Morgan H, Pradhan A, Page-Ramsey SM, Undergraduate Medical Education Committee, Association of Professors of Gynecology and Obstetrics, Crofton MD. To the point: the expanding role of simulation in obstetrics and gynecology medical student education. Am J Obstet Gynecol. 2019;220(2):129–41. https://doi.org/10.1016/j. ajog.2018.10.029. Epub 2018 Oct 25. PMID: 30696555. 23. Smith PP, Choudhury S, Clark TJ. The effectiveness of gynaecological teaching associates in teaching pelvic examination: a systematic review and meta-analysis. Med Educ. 2015;49(12):1197–206. https://doi.org/10.1111/medu.12816. 24. Duffy JMN, Chequer S, Braddy A, Mylan S, Royuela A, Zamora J, Ip J, Hayden S, Showell M, Kinnersley P, Chenoy R, Westwood OM, Khan KS, Cushing A. Educational effectiveness of gynaecological teaching associates: a multi-centre randomised controlled trial. BJOG. 2016;123:1005–10. 25. Nitsche JF, Shumard KM, Fino NF, Denney JM, Quinn KH, Bailey JC, Jijon R, Huang C, Kesty K, Whitecar PW, Grandis AS, Brost BC. Effectiveness of labor cervical examination simulation in medical student education. Obstet Gynecol. 2015;126(Suppl 4):13S– 20S. https://doi.org/10.1097/AOG.0000000000001027. 26. Holmström SW, Downes K, Mayer JC, Learman LA.  Simulation training in an obstetric clerkship: a randomized controlled trial. Obstet Gynecol. 2011;118(3):649–54. https://doi.org/10.1097/ AOG.0b013e31822ad988. 27. Dayal AK, Fisher N, Magrane D, Goffman D, Bernstein PS, Katz NT. Simulation training improves medical students’ learning experiences when performing real vaginal deliveries. Simul Healthc. 2009;4(3):155–9. https://doi.org/10.1097/SIH.0b013e3181b3e4ab. 28. Nitsche J, Morris D, Shumard K, Akoma U. Vaginal delivery simulation in the Obstetrics and Gynaecology clerkship. Clin Teach. 2016;13(5):343–7. https://doi.org/10.1111/tct.12458. 29. DeStephano CC, Chou B, Patel S, Slattery R, Hueppchen N.  A randomized controlled trial of birth simulation for medical students. Am J Obstet Gynecol. 2015;213(1):91.e1–7. https://doi. org/10.1016/j.ajog.2015.03.024. 30. Deering S, Brown J, Hodor J, Satin AJ.  Simulation training and resident performance of singleton vaginal breech delivery. Obstet Gynecol. 2006;107(1):86–9. https://doi.org/10.1097/01. AOG.0000192168.48738.77. 31. Easter SR, Gardner R, Barrett J, Robinson JN, Carusi D. Simulation to improve trainee knowledge and comfort about twin vaginal birth. Obstet Gynecol. 2016;128(Suppl 1):34S–9S. https://doi. org/10.1097/AOG.0000000000001598. 32. Bligard KH, Lipsey KL, Young OM. Simulation training for operative vaginal delivery among obstetrics and gynecology residents: a systematic review. Obstet Gynecol. 2019;134(Suppl 1):16S–21S. https://doi.org/10.1097/AOG.0000000000003431. 33. Deering S, Poggi S, Macedonia C, Gherman R, Satin AJ. Improving resident competency in the management of shoulder dystocia with simulation training. Obstet Gynecol. 2004;103(6):1224–8. https:// doi.org/10.1097/01.AOG.0000126816.98387.1c. 34. Fisher N, Bernstein PS, Satin A, Pardanani S, Heo H, Merkatz IR, Goffman D.  Resident training for eclampsia and magnesium toxicity management: simulation or traditional lecture? Am J

R. Achiron et al. Obstet Gynecol. 2010;203(4):379.e1–5. https://doi.org/10.1016/j. ajog.2010.06.010. 35. Maslovitz S, Barkai G, Lessing JB, Ziv A, Many A.  Recurrent obstetric management mistakes identified by simulation. Obstet Gynecol. 2007;109(6):1295–300. https://doi.org/10.1097/01. AOG.0000265208.16659.c9. 36. Deering S, Rowland J.  Obstetric emergency simulation. Semin Perinatol. 2013;37(3):179–88. https://doi.org/10.1053/j. semperi.2013.02.010. 37. Vellanki VS, Gillellamudi SB. Teaching surgical skills in obstetrics using a cesarean section simulator - bringing simulation to life. Adv Med Educ Pract. 2010;6(1):85–8. https://doi.org/10.2147/AMEP. S14807. 38. Ellis D, Crofts JF, Hunt LP, Read M, Fox R, James M. Hospital, simulation center, and teamwork training for eclampsia management: a randomized controlled trial. Obstet Gynecol. 2008;111(3):723–31. https://doi.org/10.1097/AOG.0b013e3181637a82. 39. Fransen AF, van de Ven J, Merién AE, de Wit-Zuurendonk LD, Houterman S, Mol BW, Oei SG. Effect of obstetric team training on team performance and medical technical skills: a randomised controlled trial. BJOG. 2012;119(11):1387–93. https://doi. org/10.1111/j.1471-­0528.2012.03436.x. 40. Crofts JF, Lenguerrand E, Bentham GL, Tawfik S, Claireaux HA, Odd D, Fox R, Draycott TJ.  Prevention of brachial plexus injury-12 years of shoulder dystocia training: an interrupted time-series study. BJOG. 2016;123(1):111–8. https://doi. org/10.1111/1471-­0528.13302. 41. Gossett DR, Gilchrist-Scott D, Wayne DB, Gerber SE. Simulation training for forceps-assisted vaginal delivery and rates of maternal perineal trauma. Obstet Gynecol. 2016;128(3):429–35. https://doi. org/10.1097/AOG.0000000000001533. 42. Siassakos D, Hasafa Z, Sibanda T, Fox R, Donald F, Winter C, Draycott T.  Retrospective cohort study of diagnosis-­ delivery interval with umbilical cord prolapse: the effect of team training. BJOG. 2009;116(8):1089–96. https://doi. org/10.1111/j.1471-­0528.2009.02179.x. 43. Nelissen E, Ersdal H, Mduma E, Evjen-Olsen B, Twisk J, Broerse J, van Roosmalen J, Stekelenburg J.  Clinical performance and patient outcome after simulation-based training in prevention and management of postpartum haemorrhage: an educational intervention study in a low-resource setting. BMC Pregnancy Childbirth. 2017;17(1):301. https://doi.org/10.1186/s12884-­017-­1481-­7. 44. Dempsey E, Pammi M, Ryan AC, Barrington KJ.  Standardised formal resuscitation training programmes for reducing mortality and morbidity in newborn infants. Cochrane Database Syst Rev. 2015;(9):CD009106. https://doi.org/10.1002/14651858. CD009106.pub2. 45. Nielsen PE, Goldman MB, Mann S, Shapiro DE, Marcus RG, Pratt SD, Greenberg P, McNamee P, Salisbury M, Birnbach DJ, Gluck PA, Pearlman MD, King H, Tornberg DN, Sachs BP.  Effects of teamwork training on adverse outcomes and process of care in labor and delivery: a randomized controlled trial. Obstet Gynecol. 2007;109(1):48–55. https://doi.org/10.1097/01. AOG.0000250900.53126.c2. 46. Ennen CS, Satin AJ. Training and assessment in obstetrics: the role of simulation. Best Pract Res Clin Obstet Gynaecol. 2010;24(6):747– 58. https://doi.org/10.1016/j.bpobgyn.2010.03.003. 47. Attilakos G, Draycott T, Gale A, Siassakos D, Winter C. ROBuST: RCOG operative birth simulation training: course manual. Cambridge: Cambridge University Press; 2013. https://doi. org/10.1017/CBO9781107445154. 144 p. 48. Gavin NR, Satin AJ.  Simulation training in obstetrics. Clin Obstet Gynecol. 2017;60(4):802–10. https://doi.org/10.1097/ GRF.0000000000000322.

2

The Role of Simulation in Obstetric Schools in the UK Sasha Taylor and Wassim A. Hassan

2.1 Introduction Simulation is a valuable training tool, which is becoming increasingly popular in medical training programmes and forms an essential component of the Royal College of Obstetricians and Gynaecology UK Training programme [1]. Medical education and training has become ever challenging for Doctors in training to gain competency and experience in practical skills in recent years. This is due to increases in training programme sizes, working-hour restrictions and a reduction in tolerance for medical errors. As with the aviation industry, educators looking for methods to overcome these challenges have found simulation training to be an appealing option in training in a vast range of clinical and non-clinical skills within the medical profession [2]. There is significant research to support the use of simulation-­based education in training, from teaching basic surgical and practical skills, improving teamwork and leadership skills in complex emergency scenarios [3–5]. Simulationbased training has the benefit of introducing new concepts in a safe artificial environment, removing the risk of causing harm to patients, whilst also providing a method to gain experience and evaluate required competencies. The use of simulation in medical education has become increasingly advanced from the use of animal and human cadavers, to low-fidelity in animate simulators, through to and high-­fidelity virtual reality simulation suites. Encouragingly, there is vast evidence to S. Taylor Department of Obstetrics and Gynaecology, West Suffolk Hospital, West Suffolk NHS Foundation Trust, Bury St Edmunds, UK W. A. Hassan (*) Fetal Medicine Unit, Department of Obstetrics and Gynaecology, Colchester Hospital, East Suffolk and North Essex Foundation Trust, Colchester, UK Department of Surgery and Cancer, Imperial College London, London, UK e-mail: [email protected]

support that the skills gained from simulation translate into measurable improvements in a trainee’s real-life performances [3, 6, 7].

2.2 The History of Obstetric Simulation Training Many will link the implementation of simulators in UK healthcare with their historical use and success in the Aviation industry. Simulators were created for flight training pilots in the 1920s, and the ability to reproduce clinical scenarios has been utilised in Obstetrics and Gynaecology (O&G) training. Standardised training mannequins such as “Resusci Anne” and “Harvey Cardiology Mannequin” became popular in the 1960s and continued to evolve with the development of simulation software in the 1980s [8, 9]. In the last decade, the success of non-technical skills and human factors training in reducing Aviation errors has been further implemented into Medical simulation training, in an attempt to improve healthcare providers situational awareness, team working and communication skills. Obstetric simulators have been reported to date back as early as the eighteenth century. During the 1700s, France was facing a public health crisis owing to an excessive intrapartum stillbirth rate, especially in the countryside. With the majority of midwives in the countryside ill-trained, the care for women using “unproven” clinical methods resulted in many stillbirths and birth injuries. In 1751, experienced midwife Angélique Marguerite Le Boursier du Coudray (Fig. 2.1) was inspired to revolutionise childbirth education. Madame du Coudray crafted birthing mannequins to simulate childbirth and the potential Obstetric complications that may arise. Surgeons from each city came to study with Madame du Coudray and used the mannequin to train their local midwives in her methods (Fig.  2.2a–c). Du Coudray wrote of untrained surgeons and birthing attendants “When difficulties arise, they are absolutely unskilled, and until long

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 G. Cinnella et al. (eds.), Practical Guide to Simulation in Delivery Room Emergencies, https://doi.org/10.1007/978-3-031-10067-3_2

19

20

S. Taylor and W. A. Hassan

Fig. 2.1  French Midwife Madame du Coudray crafted a birthing mannequin to simulate Obstetric complications and revolutionised childbirth education in the early eighteenth century. (https://litfl. com/wp-­c ontent/uploads/2020/01/Angelique-­M arguerite-­L e-­ Boursier-­du-­Coudray-­1712-­1789.jpeg)

a

Fig. 2.2 (a–c) Madame du Coudray’s birthing mannequins produced in the early eighteenth century are not dissimilar to the modern simulation training models used to train today’s UK Doctors in

Obstetrics and Gynaecology. (http://10unknownpleasures.blogspot. com/2019/05/angelique-­marguerite-­le-­boursier-­du.html)

2  The Role of Simulation in Obstetric Schools in the UK

b

21

c

Fig. 2.2 (continued)

experience instructs them, they are the witness or the cause of many misfortunes”. She asserted that “one learns on the machine in little time to prevent such accidents” [10].

2.3 Simulation in UK Obstetrics and Gynaecology Training Programme Medical training has traditionally relied on a ‘see one, do one, teach one’ approach. This inevitably exposes patients to inexperienced healthcare practitioners, and the potential dangers and harm associated with this are increasingly unacceptable in modern medicine. It is no longer acceptable for doctors to gain experience by trial and error, and simulation-­based training is now an accepted method within the UK O&G training programme to overcome this. The Royal College of Obstetricians and Gynaecologists (RCOG) is responsible for the specialty training programme and curriculum in Obstetrics and Gynaecology training within the UK.  The programme requires a minimum of 7 years training (ST1–ST7) which can be commenced following completion of foundation training. The programme is divided into basic, intermediate and advanced levels of training (Fig. 2.3). More recently, experienced Doctors who have

satisfied basic training competencies in O&G outside of the UK training programme can apply to enter at ST3 level. On successfully completing the programme, the Doctor is awarded a Certificate of Completion of Training (CCT), qualifying entry onto the Specialist Register in the UK, which is a requirement to practice as a Consultant in the NHS [11] (RCOG: Trainees Guide to O&G Curriculum and Specialty Training). The RCOG Training Matrix outlines key curriculum competencies that trainees must satisfy in order to progress to the next level of training. There are a number of simulation courses outlined in the Training Matrix for which attendance is mandatory, such as PROMPT, ALSO, STEP-UP, third-­ degree tear repair and ROBuST [1]. Trainees are encouraged to try to access as much hands-on clinical exposure wherever possible to facilitate their learning during the 7-year programme. However, it is acknowledged by the RCOG that certain conditions are not commonly encountered in clinical practice, and for such situations competencies in their training logbook may be signed off by a trainer using ‘other methodologies’ (OM). Trainees may choose to provide evidence from simulation courses they have attended, in order to gain experience and evidence of competency in areas that they may not commonly encounter in

S. Taylor and W. A. Hassan

22 Fig. 2.3  The diagram above demonstrates the different stages of the RCOG training programme. (https://www. rcog.org.uk/en/careers-­ training/about-­specialty-­ training-­in-­og/ introduction-­to-­specialty-­ training-­in-­og/)

Royal College of Obstetricians & Gynaecologists

Specialty training and education programme Foundation training

FY1

FY2

Basic training

ST1

ST2

Intermediate training

ST3

ST5

ST6

ST7

Core Training

Training in women’s health

NTN

ST4

Advanced training

Part 1 MRCOG

Part 2 MRCOG Part 3 MRCOG

ATSMs

Part 1 MRCOG to be completed in ST1 or ST2. Required for progression to ST3 Part 2 MRCOG to be completed in ST3, ST4 or STS. Required for progression to ST6

Subspecialty

Part 3 MRCOG to be completed in ST3, ST4 or STS. Part 2 MRCOG. Required for progression to ST6

practice, for example, cervical cerclage insertion or management of eclampsia. There are many additional non-mandatory simulation courses available for trainees to attend in order to practice their surgical skills, such as hysteroscopy and laparoscopy prior to carrying out procedures on real patients under senior supervision. The majority of UK training hospitals will have clinical skills labs and or simulation models for trainees to practice surgical skills and emergency procedures and manoeuvres to further their knowledge and skills base. In order to support the delivery of the RCOG’s ‘Simulation Strategy’, The RCOG Simulation Network ‘RCOG Sim-Net’ was established. Consultants are invited to apply as ‘local simulation leads’ to join the Network, providing representation for their local hospitals to promote and support the delivery of simulation training throughout O&G training for Junior doctors [12]. There are also more recently opportunities for doctors in training to apply for simulation fellowship positions within their deaneries to further support this drive.

2.4 Simulation Training in Practice Simulation training is a technique to replace or amplify real-­ patient experiences with guided experiences, artificially contrived, evoking or replicating substantial aspects of the real world in a fully interactive manner [13]. In Obstetrics, simulation training is applied to multiple procedural skills and emergencies, including assisted vaginal delivery, management of shoulder dystocia, breech vaginal delivery, delivery of a deeply engaged foetal head at caesarean section and ultrasound training (Fig. 2.4). There are an increasing number of simulation courses and models available to teach Gynaecology skills, including hysteroscopy, operative laparoscopy, open surgery, suturing skills and post-operative complications. Simulation training in the domain of technical competence enables learners to practice, making mistakes in a safe environment, with opportunities for feedback and learning from those mistakes, in order to achieve a level of proficiency [14].

2  The Role of Simulation in Obstetric Schools in the UK

23

Fig. 2.4  The Desperate Debra® simulator has been designed to facilitate the training and implementation of correct clinical procedures when confronted with impaction of the foetal head—a serious and

potentially life-threatening event. (https://www.adam-­rouilly.co.uk/ products/clinical-­skills-­simulators/impacted-­fetal-­head/ar58-­desperate-­ debra-­impacted-­fetal-­head-­simulator)

2.5 Low-Fidelity Simulation

have significant benefits; simple and inexpensive low-fidelity synthetic models provide the operator with the opportunity to use real instruments to refine their skills [15]. Many doctors will have learned and practiced suturing on cadaveric or animal-derived tissue prior to suturing a person in the operating theatre, and these practices are still used throughout UK medical schools and for doctors in O&G training.

Low-fidelity simulation models are often less costly, relatively easy to implement and transport in comparison to highfidelity models (Fig. 2.5). They tend to be the least real of the simulation modalities and may not provide learners with the experience of working in a real-life situation. However, they

24

S. Taylor and W. A. Hassan

Fig. 2.5  The SIMULAB Perineal Repair trainer is an example of a low-fidelity synthetic model which can be used for incision training, training to administer a pudendal block, and repair of first to fourth perineal tears. (https://www.simulab.com/products/perineal-­repair-­trainer-­0)

2.6 High-Fidelity Simulation The use of high-fidelity virtual reality simulation has further increased the role of simulation training to a more advanced level, enabling training opportunities in a more realistic clinical environment. This can be produced through blended simulation combining actors with low-fidelity training m ­ odels, however, with technological advances, simulators themselves are becoming increasingly high-tech. SimMom™ (Fig. 2.6) is an advanced, full body high-tech mannequin which can simulate antenatal, intrapartum and postnatal events and emergencies, including spontaneous vaginal delivery (Fig. 2.7), shoulder dystocia, breech vaginal delivery and post-partum haemorrhage [16]. The simulator software can be controlled wirelessly by a trainer, enabling

the mannequin to talk and express pain and other symptoms. Vital observations can be measured and altered to reflect a changing clinical picture. Not only is SimMom™ useful in teaching and assessing clinical and practical skills; she provides opportunities for enhanced in-situ training with multi-­ disciplinary teams working collaboratively as they would do in real life. There are also simulation Cardiotocography (CTG) programmes available, which can be used to replicate a real-­time foetal heart rate monitor on tablet computers to increase the authenticity of an Obstetric emergency in the mock labour ward (Fig. 2.8). Many UK hospitals and medical schools now have access to such high-fidelity programmes and mannequins, using them to run emergency skills training sessions and assess competency in assessments in clinical skills.

2  The Role of Simulation in Obstetric Schools in the UK Fig. 2.6 SimMom mannequin demonstrating a simulation birth. (https:// laerdal.com/us/products/ simulation-­training/ obstetrics-­pediatrics/ obstetric-­solution-­simmom-­ and-­mamabirthie/)

Fig. 2.7 SimMom mannequin demonstrating a spontaneous vaginal delivery for a baby in direct occipito-­ posterior position. (https:// laerdal.com/us/products/ simulation-­training/ obstetrics-­pediatrics/ obstetric-­solution-­simmom-­ and-­mamabirthie/)

25

26

S. Taylor and W. A. Hassan

Fig. 2.8  CTG is a highly advanced and realistic foetal heart rate monitor simulation package which can be used in the high-fidelity simulation of Obstetric emergencies. (https://www.isimulate.com/ ctgi/)

2.7 The Application of Simulation Training Whilst we know that simulation training is favourable for removing risks to training on real-life patients, it is reasonable to question the transferability to real patients. However, many studies have investigated the educational value of simulation and found them to be effective for trainees and healthcare professionals in developing their skills and knowledge. A wide variety of technologies have been studied including simulated patients, basic and interactive mannequins, and confirmed transfer of training to patient care settings. In Obstetrics, Draycott et al., demonstrated improved neonatal outcomes of births complicated by shoulder dystocia after the implementation of simulation training [6]. Birch et  al., conducted a study to determine the most effective method of delivering training to staff on the management of an Obstetric emergency. All teams improved in their performance and knowledge following teaching via three different methods. The teams taught using simulation only were the only group to demonstrate sustained improvement in clinical management, confidence, communication skills and knowledge; although the study did not reach statistical significance [3]. Moreover, within the surgical field, a number of studies have proven that simulator-trained individuals are faster and more accurate, performing fewer errors during their first real surgical case [17, 18]. This is encourag-

ing for simulation training, and further supported by two studies measuring the effect of the learning curve in the clinical environment: the curve was found to be shorter and flatter than for groups trained in the standard manner without simulation [19, 20]. The mandatory courses outlined in the RCOG training curriculum combine simulation training using both low-­ fidelity and high-fidelity equipment and scenarios. Studies have generally not favoured low-fidelity or high-fidelity training, with no differences in performance and skills assessed on either method [21–23]. Different levels of fidelity may be needed for different objectives and levels of trainees [24, 25].

2.8 Beyond the Technical Skills Previous MBRRACE-UK (Mothers and Babies: Reducing Risk through Audits and Confidential Enquiries across the UK) reports have identified serious gaps in clinicians’ human factors skills, including communication, leadership and teamwork which contributed to maternal death [26]. The most recent MBRRACE report [27] called for the need for multidisciplinary training to improve team working and reduce maternal death, the rationale being ‘those who work together should train together’. In addition to the development of technical skills, simulation training also covers human factors and team working

2  The Role of Simulation in Obstetric Schools in the UK

skills involved in managing critically ill patients. Members of the multi-disciplinary team, working together a simulated environment enable ‘team training’ to explore communication, decision-making, judgement and leadership skills [19, 27]. Simulation training is recommended as the most appropriate and safe modality to train and teach both the technical and human factors skills involved in managing deteriorating pregnant women [26, 27]. Individual UK hospitals in which trainees are placed, run regular ‘skills and drills’ training, in order to maintain competence in managing obstetric emergencies, and ensuring good multidisciplinary team working within this setting. We are seeing increasing use of high-fidelity and ‘in-situ simulation’ in UK hospitals. In-situ simulation involves bringing the real team, i.e. those involved in the operating theatre or delivery suite to the real environment, but with a computerised manikin as the patient. Hunt et al. report improved individual and team response in acute and ambulatory care settings following in-situ simulation [28].

2.9 Conclusion Simulation training contributes to a major part of medical education and training, including the UK Obstetric training curriculum. UK Obstetric schools and hospitals have invested considerably in the use of high-fidelity simulation-based education in order to train in clinical skills as well as assessing human factors and team working skills involved in managing Obstetric emergencies. Whilst there is some contention that there may be little advantage in learning success achieved by high-fidelity over low-fidelity simulators, simulation training is recommended as the most appropriate and safe modality to train and teach both the technical and human factors skills involved in managing the deteriorating pregnant women. MBRRACE strongly advocates that ‘those who work together should train together’ in order to improve multi-disciplinary training to improve team working and reduce maternal death. Whilst some early experience suggests that such training can be effective, connections to improved Obstetric outcomes have yet to be made and further research in this area is needed. Traditional apprentice style of training in Obstetrics and Gynaecology may no longer be possible, due to the increasing difficulties for trainees in the UK to gain access training opportunities. Simulation training seems to be a plausible adjunct to training but should not be a replacement for hands­on clinical exposure. With the rising popularity for simulation training in the last decade and ever advancing technology, we predict that simulation training in Obstetrics will continue to contribute to a major portion of the training curriculum in the future.

27

References 1. RCOG.  Training matrix. n.d.. https://www.rcog.org.uk/globalassets/documents/careers-­and-­training/curriculum/curriculum2019/ matrix/2020-­21-­matrix-­covid19-­st6-­7-­who-­switched-­to-­2019-­core. pdf. 2. Sanders A, Wilson RD. Simulation training in obstetrics and gynaecology residency programs in Canada. J Obstet Gynaecol Can. 2015;37(11):1025–32. S1701-2163(16)30053-6 [pii]. 3. Birch L, Jones N, Doyle PM, Green P, McLaughlin A, Champney C, et al. Obstetric skills drills: evaluation of teaching methods. Nurse Educ Today. 2007;27(8):915–22. S0260-6917(07)00018-4 [pii]. 4. Thompson S, Neal S, Clark V. Clinical risk management in obstetrics: eclampsia drills. BMJ. 2004;328(7434):269–71. https://doi. org/10.1136/bmj.328.7434.269. 5. Cro S, King B, Paine P. Practice makes perfect: maternal emergency training. Br J Midwifery. 2001;9:492–6. https://doi.org/10.12968/ bjom.2001.9.8.7936. 6. Draycott TJ, Crofts JF, Ash JP, Wilson LV, Yard E, Sibanda T, et al. Improving neonatal outcome through practical shoulder dystocia training. Obstet Gynecol. 2008;112(1):14–20. https://doi. org/10.1097/AOG.0b013e31817bbc61. 7. Deering S, Poggi S, Macedonia C, Gherman R, Satin AJ. Improving resident competency in the management of shoulder dystocia with simulation training. Obstet Gynecol. 2004;103(6):1224–8. https:// doi.org/10.1097/01.AOG.0000126816.98387.1c. 8. Rosen KR.  The history of medical simulation. J Crit Care. 2008;23(2):157–66. https://doi.org/10.1016/j.jcrc.2007.12.004. 9. Cooper JB, Taqueti VR. A brief history of the development of mannequin simulators for clinical education and training. Postgrad Med J. 2008;84(997):563–70. https://doi.org/10.1136/qshc.2004.009886. 10. JSTOR Daily. Madame du Coudray. n.d.. https://daily.jstor.org/how-­ a-­french-­midwife-­solved-­a-­public-­health-­crisis/. Accessed 12 Apr 2021. 11. RCOG.  Trainees guide to O&G curriculum and specialty training. n.d.. https://www.rcog.org.uk/globalassets/documents/ careers-­and-­training/core-­curriculum/trainees-­guide-­to-­the-­og-­ curriculum-­2019-­2020.pdf. Accessed 12 Apr 2021. 12. RCOG.  Simulation network reference. n.d.. https://www.rcog. org.uk/en/careers-­training/resources-­and-­support-­for-­trainers/job-­ descriptions-­for-­rcog-­educational-­roles/local-­simulation-­lead/. Accessed 12 Apr 2021. 13. Aggarwal R, Mytton OT, Derbrew M, Hananel D, Heydenburg M, Issenberg B, et al. Training and simulation for patient safety. Qual Saf Health Care. 2010;19(Suppl 2):34. https://doi.org/10.1136/ qshc.2009.038562. 14. Aggarwal R, Ward J, Balasundaram I, Sains P, Athanasiou T, Darzi A. Proving the effectiveness of virtual reality simulation for training in laparoscopic surgery. Ann Surg. 2007;246(5):771–9. https://doi. org/10.1097/SLA.0b013e3180f61b09. 15. Fried GM.  FLS assessment of competency using simulated laparoscopic tasks. J Gastrointest Surg. 2008;12(2):210–2. https://doi. org/10.1007/s11605-­007-­0355-­0. 16. Laerdal. n.d.. https://laerdal.com/us/products/simulation-­training/ obstetrics-­pediatrics/obstetric-­solution-­simmom-­and-­mamabirthie/. Accessed 12 Apr 2021. 17. Sedlack RE, Kolars JC, Alexander JA. Computer simulation training enhances patient comfort during endoscopy. Clin Gastroenterol Hepatol. 2004;2(4):348–52. S1542356504000679 [pii]. 18. Seymour NE, Gallagher AG, Roman SA, O’Brien MK, Bansal VK, Andersen DK, et  al. Virtual reality training improves operating room performance: results of a randomized, double-­ blinded study. Ann Surg. 2002;236(4):458–4. https://doi. org/10.1097/00000658-­200210000-­00008.

28 19. Aggarwal R, Undre S, Moorthy K, Vincent C, Darzi A. The simulated operating theatre: comprehensive training for surgical teams. Qual Saf Health Care. 2004;13(Suppl 1):27. 20. Ahlberg G, Enochsson L, Gallagher AG, Hedman L, Hogman C, McClusky DA, et al. Proficiency-based virtual reality training significantly reduces the error rate for residents during their first 10 laparoscopic cholecystectomies. Am J Surg. 2007;193(6):797–804. S0002-9610(07)00071-2 [pii]. 21. Matsumoto ED, Hamstra SJ, Radomski SB, Cusimano MD.  The effect of bench model fidelity on endourological skills: a randomized controlled study. J Urol. 2002;167(3):1243–7. S0022-­5347(05)65274-3 [pii]. 22. Cheng A, Lockey A, Bhanji F, Lin Y, Hunt EA, Lang E.  The use of high-fidelity manikins for advanced life support training--a systematic review and meta-analysis. Resuscitation. 2015;93:142–9. https://doi.org/10.1016/j.resuscitation.2015.04.004. 23. Nimbalkar A, Patel D, Kungwani A, Phatak A, Vasa R, Nimbalkar S.  Randomized control trial of high fidelity vs low fidelity simulation for training undergraduate students in neonatal resuscitation. BMC Res Notes. 2015;8:636–9. https://doi.org/10.1186/ s13104-­015-­1623-­9.

S. Taylor and W. A. Hassan 24. Grady JL, Kehrer RG, Trusty CE, Entin EB, Entin EE, Brunye TT. Learning nursing procedures: the influence of simulator fidelity and student gender on teaching effectiveness. J Nurs Educ. 2008;47(9):403–8. https://doi.org/10.3928/01484834-­20080901-­09. 25. Rudolph JW, Simon R, Raemer DB.  Which reality matters? Questions on the path to high engagement in healthcare simulation. Simul Healthc. 2007;2(3):161–3. https://doi.org/10.1097/ SIH.0b013e31813d1035. 26. Knight M, Nair M, Brocklehurst P, Kenyon S, Neilson J, Shakespeare J, et al. Examining the impact of introducing ICD-MM on observed trends in maternal mortality rates in the UK 2003-­ 13. BMC Pregnancy Childbirth. 2016;16(1):178-z. https://doi.org/10.1186/ s12884-­016-­0959-­z. 27. MBRRACE. n.d.. https://www.npeu.ox.ac.uk/assets/downloads/ mbrrace-­uk/reports/MBRRACE-­UK%20Maternal%20Report%20 2018%20-­%20Web%20Version.pdf. Accessed 12 Apr 2021. 28. Hunt EA, Heine M, Hohenhaus SM, Luo X, Frush KS. Simulated pediatric trauma team management: assessment of an educational intervention. Pediatr Emerg Care. 2007;23(11):796–804. https://doi. org/10.1097/PEC.0b013e31815a0653.

3

Ontologies, Machine Learning and Deep Learning in Obstetrics Lorenzo E. Malgieri

3.1 Integrated Care Pathways 3.1.1 Introduction This work intends to take a survey of Artificial Intelligence for a possible use in Diagnostic and Therapeutic Care Pathways (DTCP), which favors real integration between the various professionals involved in the management of the treatment process and the real sharing of results in terms of clinical, social, organizational, and governance aspects. Digital Health technologies [1], connectivity, software, and sensors for health care and related uses include categories such as Mobile Health (mHealth), Health Information Technology (IT), Wearable Devices, Telehealth, Telecare and Telemedicine, and Personalized Medicine. From applications in general wellness to applications as a medical device, they include technologies used to develop or study medical products, for use as a medical product, in a medical product, as companion diagnostics, or as an adjunct to other medical products (devices, drugs, and biologics). Modern medicine has shifted from developing treatments after the fact, to preventing, personalizing, and delivering precision care. This requires vast amounts of data to increase available knowledge through all the Diagnostic and Therapeutic Care Pathways and eliminates the discontinuity between the three classic levels of care (primary care, territorial specialist care, hospital stay), giving rise to a continuum that includes the identification of specific “products” (clinical and non-­ clinical) by each actor (or team of which he is a member) in relation to the prefixed health objective. In fact, e-Health technologies and Artificial Intelligence can support the implementation of a network operating mode, facilitating the L. E. Malgieri (*) FIAT-ENI, Milan, Italy Environmental Companies, Milan, Italy Chief Innovation Officer in CLE, Bari, Italy e-mail: [email protected]

integration between the various figures in charge of care and service delivery in the health and social health. The goal of this survey is to help address this central challenge.

3.1.2 Artificial Intelligence and SaMD Artificial Intelligence has been and still is the subject of many different definitions; in our virtual tour, we would like to start with the definition given by J. McCarthy, in 2007, and also shared by the FDA [2]: “Artificial Intelligence, the science and engineering of making intelligent machines, especially intelligent computer programs, and Machine Learning, an artificial intelligence technique that can be used to design and train software algorithms to learn from and act on data” (McCarthy [3]).

3.1.2.1 Software as a Medical Device As a software, Artificial Intelligence in Medicine is subject to the global approach of Software as a Medical Device [4]. The term Software as a Medical Device (SaMD) is defined by the International Medical Device Regulators Forum (IMDRF) as “software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device” [5]. The IMDRF [6] is a voluntary group of medical device regulators from around the world who have come together to reach harmonization on medical device regulation. IMDRF develops internationally agreed upon documents related to a wide variety of topics affecting medical devices. In 2013, IMDRF formed the Software as a Medical Device Working Group (WG) to develop guidance supporting innovation and timely access to safe and effective Software as a Medical Device globally. Chaired by the FDA, the Software as a Medical Device WG agreed upon the Key definition [7] for Software as a Medical Device, framework for risk categorization for Software as a Medical Device [5], the Quality management for Software as

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 G. Cinnella et al. (eds.), Practical Guide to Simulation in Delivery Room Emergencies, https://doi.org/10.1007/978-3-031-10067-3_3

29

30

L. E. Malgieri

a Medical Device [8], and the Clinical Evaluation [9] of Software as a Medical Device. The International Medical Device Regulators Forum (IMDRF) Software as a Medical Device Working Group (WG) published a possible risk categorization framework [7] for Software as a Medical Device (SaMD). The recommendations allow manufactures and regulators to more clearly identify risk categories of Software as a Medical Device based on how the output of a Software as a Medical Device is used for healthcare decisions in different healthcare situations or conditions. The Software as a Medical Device risk categorization framework [7] has four categories (I, II, III, and IV). These categories are based on the levels of impact on the patient or public health where accurate information provided by the Software as a Medical Device to treat or diagnose, drive, or inform clinical management is vital to avoid death, long-term disability or other serious deterioration of health, mitigating public health. The Level IV category is Software as a Medical Device with the highest impact on the patient or public health, and Level I is the lowest (Fig. 3.1). The IMDRF Quality Management System for Software as a Medical Device (SaMD) framework [8] helps manufacturers and international regulators to attain a common understanding and vocabulary for the application of medical device quality management system requirements to Software as a Medical Device. Healthcare decisions increasingly rely on information provided by Software as a Medical Device output, where these decisions can impact clinical outcomes and patient care. The FDA issued the Software as a Medical Device Clinical Evaluation [10] final guidance to describe an internally agreed upon understanding of clinical evaluation and principles for demonstrating the safety, effectiveness, and performance of Software as a Medical Device among regulators in the International Medical Device Regulators Forum. The guidance focuses on the activities Software as a Medical Device manufacturers can take to clinically evaluate their Software as a Medical Device. Software as a Medical Device [5] ranges from software that allows a smartphone to view images obtained from a magnetic resonance imaging (MRI) medical device for diagnostic purposes to Computer-­ Aided Detection (CAD) software that performs image post-­ processing to help detect breast cancer. If the software is part Fig. 3.1  Risk categories of software as a medical device

of a hardware medical device, then it does not meet the definition of Software as a Medical Device. Examples include Software used to “drive or control” the motors and the pumping of medication in an infusion pump, or Software does not have a medical purpose [11].

3.1.2.2 Software as a Medical Device: Digital Therapies One of the new trends in digital healthcare is the so-called Digital Therapies. Digital therapies are technological solutions—mainly Apps, but also video games (EndeavorRx®), the first-and-only prescription treatment delivered through a video game FDA cleared [12]—that must be clinically certified and authorized by regulatory bodies and that help patients to take a drug, increase adherence to the therapy, and/or modify their behavior. According to the Digital Therapeutics Alliance [13] “Digital therapeutics (DTx) deliver evidence-based therapeutic interventions that are driven by high quality software programs to prevent, manage, or treat a medical disorder or disease. They are used independently or in concert with medications, devices, or other therapies to optimize patient care and health outcomes.” In September 2017, the U.S.  Food and Drug Administration permitted marketing of the first mobile medical application to help treat substance use disorders (SUD). The application, Reset [14], is intended to be used with outpatient therapy to treat alcohol, cocaine, marijuana, and stimulant SUDs [15]. The European regulatory framework in this sector is undergoing a major overhaul with the new Regulations on medical devices (MDR): Regulation 2017/745 of the European Parliament and of the Council. Although the new EU Regulations entered into force on May 25, 2017, they will apply, with certain exceptions, from May 26, 2021  in respect of Regulation (EU) 2017/745 [16]. The first app in Europe is Tinnitracks [17]. 3.1.2.3 Artificial Intelligence and Software as a Medical Devices The FDA’s traditional paradigm of medical device regulation was not designed for adaptive artificial intelligence and machine learning technologies. The FDA has cleared or

Significance of information provided by SaMD to health care decision State of Health care situation or condition Treat or diagnose Critical Serious Non-serious

IV III II

Drive clinical management III II I

Inform clinical management II I I

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

approved several medical devices using “locked” algorithms. They define a “locked” algorithm as an algorithm that provides the same result each time the same input is applied to it and does not change. However, many recent medical devices, especially when AI/ML based, use algorithms that change and can adapt over time; these are described by the FDA as adaptive algorithms, for which current regulatory frameworks were not designed [18]. The power of these AI/ ML-based algorithms lies within the ability to continuously learn, where change to the algorithm might only be realized after the device or software has been distributed for use and could learn from real-world experience. On April 2, 2019, the FDA published a discussion paper “Proposed Regulatory Framework for Modifications to Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD)—Discussion Paper and Request for Feedback” [19] that describes the FDA’s foundation for a potential approach to premarket review for artificial intelligence and machine learning-driven software modifications. The ideas described in the discussion paper leverage practices from the current premarket programs and rely on IMDRF’s risk categorization principles, the FDA’s benefit-­ risk framework, risk management principles described in the software modifications guidance, and the organizationbased total product lifecycle approach (also envisioned in the Digital Health Software Precertification (Pre-Cert) Program) [20]. In the framework described in the discussion paper, the FDA envisions a “predetermined change control plan” in premarket submissions. This plan would include the types of anticipated modifications—referred to as the “Software as a Medical Device Pre-Specifications”—and the associated methodology being used to implement those changes in a controlled manner that manages risks to patients—referred to as the “Algorithm Change Protocol.”

31

The AI/ML-Based Software as a Medical Device Action Plan outlines five actions that the FDA intends to take, including: • further developing the proposed regulatory framework, including through issuance of draft guidance on a predetermined change control plan (for software’s learning over time); • supporting the development of good machine learning practices to evaluate and improve machine learning algorithms; • fostering a patient-centered approach, including device transparency to users; • developing methods to evaluate and improve machine learning algorithms; and • advancing real-world performance monitoring pilots. The AI/ML Action Plan is a response to stakeholder feedback received from the April 2019 discussion paper, Proposed Regulatory Framework for Modifications to Artificial Intelligence/Machine Learning-Based Software as a Medical Device [19]. The FDA welcomes continued feedback in this area and looks forward to engaging with stakeholders on these efforts. The agency will also continue to collaborate across the FDA to build a coordinated approach in areas of common focus related to AI/ML. Launched in September of 2020, the CDRH Digital Health Center of Excellence [22] is committed to strategically advancing science and evidence for digital health technologies within the framework of the FDA’s regulatory and oversight role. The goal of the Center is to empower stakeholders to advance health care by fostering responsible and high-quality digital health innovation.

The State of Artificial Intelligence-Based FDA-­ Approved Medical Devices and Algorithms: FDA Artificial Intelligence/Machine Learning Action An Online Database Plan Benjamens et al. [23] provided an insight into the currently In January 2021, the U.S.  Food and Drug Administration available AI/ML-based medical devices and algorithms that released the agency’s first Artificial Intelligence/Machine have been approved by the US Food & Drugs Administration Learning (AI/ML)-Based Software as a Medical Device (FDA). 29 AI/ML-based medical technologies listed in the (SaMD) Action Plan [21]. This action plan describes a multi-­ study are mentioned algorithm, but the application of AI/ML pronged approach to advance the Agency’s oversight of AI/ has not been confirmed by the official FDA announcements ML-based medical software. but by other online sources [23]. The plan outlines a holistic approach based on total prodThe first FDA approval was granted in the year 2016, with uct lifecycle oversight to further the enormous potential that three approvals at the end of the year 2017. Most FDA these technologies have to improve patient care while deliv- approvals were granted in the year 2018, with 13 (44.8%) ering safe and effective software functionality that improves approvals, while 10 (34.4%) and 2 (6.9%) approvals were the quality of care that patients receive. To stay current and granted in 2019 and 2020 up until February, respectively. address patient safety and improve access to these promisThe two main medical specialties with AI/ML-based ing technologies, this action plan will continue to evolve medical innovations are Radiology and Cardiology, with 21 over time. (72.4%) and 4 (13.8%) FDA-approved medical devices and

32

algorithms, respectively. The remaining medical devices and algorithms can be grouped as focusing on internal medicine/ endocrinology, neurology, ophthalmology, emergency medicine, and oncology.

3.1.3 Pathology Innovation Collaborative Community (PICC) Mission of PICC, the Pathology Innovation Collaborative Community [24], is to bring together a broad range of stakeholders to accelerate the development and delivery of regulatory science initiatives in the pre-competitive space that modernize the clinical practice of pathology. Specifically, this will include digital pathology and its major enabling fields of machine learning and artificial intelligence, imaging informatics, engineering, and computational and data sciences. Similar to other innovative diagnostic technologies, digital pathology faces the dilemma of how best to demonstrate its essential value to stakeholders. The Payor working group of the Alliance for Digital Pathology has been active in developing a paper about reimbursement and its landscape in Digital Pathology, published in the Journal of Precision Medicine before the end of the year 2020 [25]. The paper discusses the introduction of new technologies such as Artificial Intelligence in Pathology and the adoption thereof, and how they might impact clinical innovation, payment reform, value-based care, and coverage and reimbursement decisions. Benefits of digital pathology to pathologists have been documented in the many studies that have correlated its use with greater accuracy and precision in pathology results [26, 27]. Any variability in these studies tends to be not due to the skill level, or lack thereof, between pathologists, but rather the lack of good, comparable consensus standards. On the other hand, the reference for measuring accuracy and precision for digital pathology devices is often individual pathologists. For this reason, FDA aims to develop key performance indicators to monitor and communicate progress on how pre-competitive regulatory science can drive innovation in pathology.

3.1.4 Standard and Healthcare Bringing together medical, software, and artificial intelligence expertise requires pooling terminology definitions that belong to different expertise, application domains, standards in healthcare, and software programming languages.

3.1.4.1 The Clinical Element Model (CEM) The Clinical Element Model (CEM) [28] provides an architecture for representing information in EHRs. CEM includes two models: Abstract instance model for representing indi-

L. E. Malgieri

vidual instances of collected data, and Abstract constraint model for representing constraints on the data instances. These two models are abstract specifications and can be implemented using different programming languages. The main purpose of such an abstract implementation is to provide a way to normalize different data from EHRs. Clinical models and clinical data are the basis for the SaMD. Detailed clinical models are the basis for retaining computable meaning when data is exchanged between heterogeneous computer systems. Detailed clinical models are also the basis for shared computable meaning when clinical data is referenced in decision support logic. The current well-established point in all the Information Systems currently in use in healthcare is based mainly on several reasons. The goal is classification of clinical data to be able then data can be exchanged between different computers within a facility or enterprise to make information available to clinicians at the point of care, with the goal of improving clinical decision making [28].

3.1.4.2 Electronic Medical Records (EMR) Electronic medical records (EMRs) are the digital version of the patient chart having information stored on the computer system. Every paper of the patient such as his medical history, lab tests, and diagnoses is stored in the information system rather than in the form of bulky paper files. In general, it is a closed system within a hospital or research center. 3.1.4.3 Electronic Health Records (EHR) Electronic health records (EHRs) are the digital health information system of the person that covers many aspects of healthcare, including the management of clinical records. EHRs include vital signs, past medical history, diagnoses, progress notes, medications, allergies, lab data, immunization dates, and imaging reports. EHRs allow this information to travel inside or outside the premises of the organization, as well as exchange of patient data between healthcare institutions, integration of medical devices, interfaces with clinical decision support systems, etc. [29]. Providers can use EHRs with various tools to make the important decisions about the treatment of the patient whereas all the tools provided by medical records system are limited to the diagnosis of the patient. 3.1.4.4 openEHR openEHR [30] is an open standard specification in health informatics that describes the management and storage, retrieval, and exchange of health data in electronic health records (EHRs). In openEHR, all health data for a person are stored in a “one lifetime,” vendor independent, person-­ centered EHR [31]. In Fig. 3.2, shows the multi-level modeling basis of the openEHR platform for e-health that differentiates the reference model defining data representation from the archetypes that contain knowledge.

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

33

AQL Queries Match required content

Reporting Apps

? ? ? ?

Reference Model

Archetypes

Defines data representation

Define possible content

Analytics

?

?

APls

+ Portable (archetype-based) + developed by specialist users + basis of data use and CDS

OpenEHR Data

Terminology

+ 150 classes + developed by IT experts + basis of syntactic interop

+ O(50K) data points for medicine + developed by domain experts + basis of semantic interop

Templates Define the actual content

+ unlimited data sets for systems + developed by local HIT devs + basis of systems development

EMR

RIS

LIS

Fig. 3.2  The semantic framework of openEHR

3.1.4.5 Health Level Seven (HL7) Health Level Seven (HL7) [32] is a non-profit organization that provides a framework (and related standards) for the exchange, integration, sharing, and retrieval of electronic health information. These standards have been developed to prescribe common building blocks of EHRs. HL7 standards are grouped into reference categories such as Primary Standards for system integrations, interoperability, and compliance and Clinical Document Architecture (CDA®) Products—the CDA uses eXtensible Markup Language (XML) for the specification of appropriate clinical information—functional models, and profiles that enable the constructs for management of electronic health records. These standards aim to support clinical practice and the management, delivery, and evaluation of health services. The current version of the standard suite is HL7 Version 3 (V3), a suite of specifications based on HL7’s Reference Information Model (RIM) [33]. Other standards are Arden Syntax, a formalism for representing procedural clinical knowledge in order to facilitate the sharing of computerized health knowledge bases among personnel, information systems and institutions, and CCOW—HL7 Clinical Context Management Specification (CCOW), aimed at facilitating the integration of applications at the point of use. As a ­standard for health care data exchange for both internal applications programming and runtime environment infrastructure, HL7 integrates the traditional emphasis on data interchange and enterprise workflow: Cross-paradigm/ Domain Analysis Models - Cross-paradigm/Logical

Level Standards, e.g. Domain Analysis Models. The Clinical Document Architecture (CDA®) components are defined using the HL7 RIM and HL7 DT by referencing a shared medical terminology such as The Systematized Nomenclature of Medicine Clinical Terms (SNOMED-CT) [34]. Other standards in HL7 include Clinical and Administrative Domains, Implementation Guides for implementation guides, and/or support documents created to be used in conjunction with an existing standard, and technical specifications, programming structures, and guidelines for software and standards development.

3.1.4.6 Unified Medical Language System (UMLS) The Unified Medical Language System® (UMLS®) [35] integrates and distributes key terminology, classification and coding standards, and associated resources to promote creation of more effective and interoperable biomedical information systems and services, including electronic health records. The Three UMLS Knowledge Sources are: • Metathesaurus [36]: terms and codes from many vocabularies, including CPT, ICD-10-CM, LOINC, MeSH, RxNorm, and SNOMED-CT, hierarchies, definitions, and other relationships and attributes. • Semantic Network [37]: Broad categories (semantic types) and their relationships (semantic relations). • SPECIALIST Lexicon and Lexical Tools [38]: a large syntactic lexicon of biomedical and general English and

34

L. E. Malgieri

tools for normalizing strings, generating lexical variants, and creating indexes. The lexicon has been released as one of the UMLS Knowledge Sources since 1994. The SPECIALIST lexical tools and LexAccess that provides access to information from the SPECIALIST LEXICON are software developed in Java (programming language) [39].

3.1.4.7 CEN/ISO EN13606 ISO 13606:2019 [40] is a standard from the International Standardization Organization (ISO) [41], originally designed by the European Committee for Standardization (CEN) [42]. The overall objective of the ISO 13606 [43] standard is to define a rigorous and stable information architecture for communicating part or all the electronic health record (EHR) of a single subject of care (patient) between EHR systems, or between EHR systems, and a centralized EHR data repository. It may also be used for EHR communication between an EHR system and clinical applications or middleware components (such as decision support components) that need to access EHR data, or as the representation of EHR data within a distributed (federated) record system. The Parts most relevant for a basic implementation of the standard are a reference model specifying statements that are applied to all entities of the same class, and knowledge level which is represented through archetypes that are statements about specific entities, archetype interchange specification, and reference archetypes and term lists.

3.1.5 Artificial Intelligence is the Way Forward in Obstetrics The aim of this chapter is to help tackle what is widely considered to be the central challenge of our time, which sees us already now and increasingly in the coming years involved with software and medical devices assisted in various ways by Artificial Intelligence. The explanation of symbolic Artificial Intelligence, as well as non-symbolic Artificial Intelligence up to deep neural networks, is necessarily high level, synthetic but systematic, to provide a panorama of easy reading also to those who do not often have the time to read the numerous scientific and research contributions on Artificial Intelligence techniques, which are more and more numerous and compelling, continuously moving the frontier of technological possibilities. Wearable sensors that can continuously monitor all vital signs—including blood pressure, heart rate and rhythm, blood oxygen saturation, respiratory rate, and temperature—there is the potential to preempt a large number of patients being hospitalized or not in the future.

Doctors reading this chapter can rest assured that from a learning perspective, they will not have to learn to program, nor will they have to become data scientists. We want to provide a brief overview of all that AI offers today and that are being lost, and at the same time, we want to highlight those new professional figures, from data scientist to have emerged and different fields of knowledge are introducing and using them profitably. In previous decades, new professional figures have emerged from biomedical engineers to data scientists. Today, it is time for data scientists or machine learning engineers to join the medical team, to support the entire team of medical professionals to evolve using techniques, algorithms, and software that are already well established and, working together, can increase knowledge and techniques. Let us hope that the pandemic has taught us to look up: we do not have to wait for directives from world organizations: a simple Chinese doctor realized before anyone else what was going on. Let us try to think of making available to all doctors like the young Chinese doctor—like him, thank God, there are many, but many in the world—a series of tools that allow them to do what they know how to do, to learn from what happens, to reduce the risks for everyone, doctors [44], and patients, in this adventure: “Medicine is a great adventure … especially for patients” (Professor Gilbert Aimard, Claude-Bernard Lyon I University) [45]. Learning to formulate a medical diagnosis has traditionally taken years. Even for professionals, formulating a diagnosis is often a long and complex process. Moreover, in many areas, the demand for these skills outstrips supply, putting pressure on the healthcare system. However, where it is possible to digitize diagnostic information, machines can help ease the burden. The advantage of an algorithm is that it can draw conclusions from data in a fraction of a second. Moreover, unlike a “flesh and blood” expert, machine learning (ML) skills can theoretically be reproduced indefinitely or extended even where there are no flesh and blood people. Machine skills can enhance human skills, not to limit them. Professor du Sautoy [46] uses the figure of the data curator as an example, calling him “almost a new kind of artist.” “Algorithms learn from data,’ he says, ‘if you provide certain data, the algorithm goes in one direction, but if you provide other data, it might go in the opposite direction. It is therefore necessary to be able to understand this path and how to shape it so that it leads us where we want to go”. This kind of attitude is destined to be a new category of competence. “Every day we interact with dozens of AI tools, which shape our decisions and preferences, our values and relationships,” continues Kevin Roose [47], technology columnist for The New York Times. “Virtually most of our choices— from the TV shows we watch, to the politicians we vote for—are modulated in the background by AI.”

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

35

3.2 Ontologies

Vocabulary standards are used to describe clinical problems, terms, categories, procedures, medications, and allerThe use of ontologies has not yet entered in medical practice. gies. Various medical vocabulary standards exist and in order We are not yet living in a semantically driven world, building to implement a usable healthcare standard interoperability, ontologies and semantic networks getting access to all of the and these vocabularies must be also taken into the information sources to allow automatic or at least semi-­ consideration. automatic integration of healthcare Information Systems, The Unified Medical Language System (UMLS) (see object-oriented programming languages, formats, and syn- Sect. 3.1.4.6) aims to alleviate the problem that exists when tax of the information, mining information out of structured using multiple vocabularies in a healthcare informatics sysand free-text sources, and normalizing the semantic content tem. UMLS comprises several controlled vocabularies in the of the resulting information are all complex and challenging biomedical sciences including SNOMED-CT, ICD, RxNorm problems [48]. We will describe some of the most significant [49], etc. It provides a mapping structure among these vocabresults achieved by researchers and will present the state of ularies, and it may also be viewed as a comprehensive theart and several questions that are still open. saurus and ontology of biomedical concepts. Therefore, it is often referred as the UMLS meta-thesaurus. Although it provides a mapping structure, it does not make semantically 3.2.1 Lists, Thesauri, and Taxonomies integrated terminology interoperable. However, it provides enough information about term relations to be used in an The simplest form of knowledge representation is a Lists, integration process. Additionally, UMLS provides facilities which is usually an alphabetical catalog of the names of all for Natural Language Processing (NLP). known concepts in each field with no implicit or explicit SNOMED-CT [50], the world leading product of relationships between them. Thesauri are derived from lists SNOMED International, consists of sets of terms naming but have an important additional component in that they descriptors in a hierarchical structure that permits searching associate synonyms with the items in the lists, synonyms at various levels of specificity and is accepted as the global which are very useful in improving the accuracy and com- common language for clinical terms. It is not an ontology per pleteness of keyword searches. Taxonomies build on se, but it is referenced from a vast majority of ontologies as Thesauri by adding relationships between concepts to pro- to provide classification and categorization of the biomedical vide a parent-child organization to Lists (Fig. 3.3). terms. Therefore, it should be considered in all integration

List, Thesauri and Taxonomies THESAURI Targets

Diseases

TAXONOMIES Anatomy

Targets

LISTS Targets

Diseases Synonyms

Anatomy

Diseases Synonyms

Synonyms

ONTOLOGY Anatomy

Diseases

Targets

Anatomy

IS EXSPRESSED IN

Synonyms

ED

IN

IS EXSPRESSED

IN

IS U

PR

EGU

L AT

EFFECTS

IS

UP

IS A TA

RE GU

RGET

IS DOWN REGU

CTS EFFE LA TE

D

IN

FOR

LATED IN

EFFECTS

Fig. 3.3  The progression of data representation from simple lists, to thesauri, to taxonomies, and then multi-relational ontologies

36

approaches as it provides structure and hierarchical information about the medical categories. Medical Subject Headings (MeSH) [51] thesaurus is a controlled and hierarchically organized vocabulary produced by the National Library of Medicine. It is used for indexing, cataloging, and searching of biomedical and health-related information. MeSH includes the subject headings appearing in MEDLINE/PubMed, the NLM Catalog, and other NLM databases. MeSH is often used in conjunction with RxNorm [52], a pharmaceutical vocabulary used for e-prescribing, medication history, government reporting, and drug compendium mapping, and Logical Observation Identifiers Names and Codes (LOINC) [53, 54], a database and universal standard for identifying medical laboratory observations. Another classification commonly referenced from ontologies is International Statistical Classification of Diseases and Related Health Problems (ICD) [55]; it is a medical classification and contains codes for diseases, signs and symptoms, abnormal findings, complaints, social circumstances, and external causes of injury or diseases. Another thesaurus is the National Cancer Institute thesaurus (NCIt) [56] that provides reference terminology for many NCI and other systems. It covers vocabulary for clinical care, translational and basic research, and public information and administrative activities.

3.2.2 How Ontologies Work Ontologies are part of “symbolic” Artificial Intelligence, “the science and engineering of making intelligent machines, especially intelligent computer programs”, as opposed to “Machine Learning, an artificial intelligence technique that can be used to design and train software algorithms to learn from and act on data” (McCarthy [3]). Ontologies have become common on the World Wide Web. According to the WWW Consortium (W3C) [57], the trend is to use the word “ontology” for more complex, and possibly quite formal collection of terms, whereas “vocabulary” is used when such strict formalism is not necessarily used or only in a very loose sense. Ontologies contain much richer and more descriptive relationships between concepts and provide a true semantic network, where having more information in the network means more connections, and thus, more directions and perspectives through which relevant information can be found. By linking all the knowledge derived from various sources, Ontologies can synthesize a very rich and powerful map of all the information known about an entire domain, something we usually do in our heads, reconciling slight differences in nomenclature and data representation to form a coherent plan of action [48].

L. E. Malgieri

The idea of the semantic web is therefore aimed at assigning a meaning to the data on the network, and the meaning involves the relationship between language and reality. The relationship between language and reality is called semantic. According to the experience common to all, the relationship between language and reality is not direct, but it is mediated by concepts: each of us is able to connect the term “drug” to the drugs present, just because we have in our mind the concept of drug. This means that the term “drug” can be used to describe in general the concept of “drug,” and at the same time indicates the set (or class) of all drugs, or the term drug can be used to describe a single drug (or individual instance) that is part of the concept or class of drugs. Many disciplines develop standardized ontologies that domain experts can use to share and annotate information in their fields. Medicine, for example, has produced large standardized structured vocabularies such as SNOMED-CT and the Unified Medical Language System (UMLS) semantic network. Due to the lack of a common vocabulary, healthcare ontologies often have reference terms from various existing vocabularies. Classes, sometimes called concepts, are at the core of most ontologies. A class can have a class hierarchy: subclasses that represent concepts that are more specific than the superclass. Classes correspond to and describe concepts in a particular knowledge domain. Each class can be subdivided into subclasses that represent more specific concepts of a superclass. There are several possible approaches in developing a class hierarchy: • A top-down development process that begins with the definition of the most general concepts or classes in the domain and the subsequent specialization of the concepts. • A bottom-up development process that begins with the definition of the most specific classes, the leaves of the hierarchy, with the subsequent grouping of these classes into more general concepts. • A combined development process that is a combination of top-down and bottom-up approaches: we first define the most salient concepts, then generalize, and specialize them as appropriate. None of these three methods is inherently better than any of the others. The combined approach is often the easiest for many since the concepts “in the middle” tend to be the most descriptive concepts. Once we have defined some of the classes, we must describe the internal structure of concepts: properties of each concept describing various features and attributes of the concept (slots, sometimes called roles or properties), and restrictions on slots (facets, sometimes called role restrictions) [58].

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

Modeling an ontology is a joint effort between application domain experts, physicians in our case, and software engineers. Building an ontology requires being clear about what the ontology will be used for, how detailed or general it should be, what kinds of questions the information in the ontology should answer, who will use the ontology, and who will be responsible for maintaining it. An ontology together with a set of individual instances of classes constitutes a Knowledge Base (KB). Formalizing of knowledge in logic is the possibility of treating it as a tangible object to which certain operations can be applied. In Description Logic (DL), the object is the Knowledge Base (KB) (set of axioms) and the operations are reasoning tasks which attempt to extract new knowledge from it. In reality, there is a fine line where the ontology ends, and the Knowledge Base (KB) begins. The term ontology is used to a conceptualization in a specific knowledge. The aspect that clearly distinguishes a knowledge base from a database is the ability to conduct reasoning automatically about the information it contains. In the context of logic, when we talk about reasoning, we always refer to deductive procedures, or more simply deductions (inferences). Logic is a system for deducing new statements from previously asserted statements with statements constructed by means of the propositional connectives: not, and, or, if–then, and, etc. First-order logic (FOL), often used to formalize mathematical statements, extends propositional logic with the notions of variables, constants, functions, predicates, and quantifiers. There are three kinds of statements, in which Description Logic (DL) is also called axioms. These are grouped into boxes: • an Assertional Box (A-Box) that is the formal representation of the concrete model of the chosen fragment of reality; it includes information on the individuals that populate the reality and on the existing relationships among them; • a Terminological Box (T-Box), also called ontology, which is the formal representation of the conceptual model of the chosen fragment of reality; • a RBox, available only in very expressive description logics, such as SROIQ.  RBox contains statements that describe characteristics of roles and interdependencies between roles. All DL-based ontologies have a T-Box and most of them an A-Box [59]. There are still further extensions, such as second-order logic, and logics which adopt slightly different approaches, such as modal and fuzzy logics. Ontologies can be accessed for querying and/or modification purposes, and they can be implemented using the core technologies of the Semantic Web [60]. The Resource Description Framework (RDF)-based languages remain the

37

principal standards of the Semantic Web. RDF provides a vocabulary and constructors for describing sets of triples and has the form composed of a subject, a predicate, and an object, using XML as the syntax option for writing expressions: RDF/XML. Each triple stands for a directed edge in the graph. The subject of the triple becomes the source of the edge, the object becomes its target, and the predicate becomes the edge label (e.g., RDF graphs). RDF graphs consist of four nodes and three edges. They need specialized databases, called graph databases and triple stores, designed to efficiently keep and process RDF data. In computer science, to interact with a database based on the relational model (RDBMS), one uses a standardized SQL (Structured Query Language) language, designed to create and modify database schemas (DDL = Data Definition Language), insert, modify, and manage stored data (DML  =  Data Manipulation Language), query stored data (DQL = Data Query Language), and create and manage data access and control tools (DCL = Data Control Language) [61]. SPARQL (SPARQL Protocol and RDF Query Language) is the language for graph databases and triple stores what SQL is to relational databases. It is a declarative interface for querying and manipulating the contents of the database/graph [62]. The Ontology Web Language (OWL) [63] is a description language that extends RDF with cardinality constraints, enumeration, and axioms, enabling the creation of richer and more accurate models. There are three sublanguages of OWL, also called species: OWL Full, OWL DL, and OWL Lite. OWL Full contains both OWL DL and OWL Lite, and OWL DL contains OWL Lite. OWL 2 extends OWL with additional features, including extended datatype support and annotation capabilities. However, OWL remains the prevalent ontology language, with many supporting editors. The information from OWL models can be queried using an RDF-based query language such as SPARQL [64]. In addition, SPARQL Update [65] can be used for retrieving and updating ontological models. Some applications might need to represent relations and allow derivations which go beyond what can be expressed and achieved with pure DL. Rule-based languages such as the Semantic Web Rule Language (SWRL), pronounced “swirl,” can be employed within ontologies.

3.2.3 Particularities of Ontologies in the Medical Domain Ontologies, as described in the previous paragraph, can also be applied to parts of the knowledge domain. It is not necessarily necessary to start by covering the whole treatment of a disease, and one can also start in small steps according to the mode we have called bottom-up; the single ontologies, thus constructed can be joined to cover a wider range for the treat-

38

ment of a disease. It is therefore important to always have in mind what are the questions to which the information contained in the ontology should respond, who will be the user of the ontology, and who will be responsible for the maintenance of the ontology itself [58]. In medicine, knowledge is based on evidence as much as possible, i.e., knowledge comes from a proof, typically a significant difference observed during a clinical trial and involving a statistical test. This is known as Evidence-Based Medicine (EBM) [66]. On the other hand, the growing demand for personalized medicine raises several questions about both therapy and greater emphasis on the drug-safety monitoring. Ontologies provide identifiers for classes and relations that represent phenomena within a domain, thereby enabling integration of data. Ontologies provide labels for classes and relations, thereby providing a domain vocabulary. Ontologies provide metadata associated with classes and relations that allow human users to understand their meaning and contribute to consistent use in annotation and other applications. Ontologies provide axioms and formal definitions that enable computational access to some aspects of the meaning of classes and relations. Combining the four main features of ontologies facilitates semantic integration of heterogeneous, multimodal data within and across domains, and enables novel data-mining methods that span traditional boundaries between domains and data types [67]. Classes and individuals are the basic elements to build ontology, so medical concepts, both for disorders and drug therapies can be represented through classes, creating within them various levels: metaclasses, classes, and subclasses. The relationship between these classes and, thus, the level of granularity is usually represented using is-a relationships, in the sense that if a class B is related to a metaclass A, and if C is a subclass of B, it is also related to metaclass A, and so on. Another aspect is the contraindications, for which one can intervene in the construction of the ontology with ad hoc classes, or with individuals within the class, or by using role-­ filler constructs. As far as individuals and their data and properties are concerned, it should be kept in mind that the level of granularity allowed can be applied by specializing the ontology for a particular disease and a particular phase within an Integrated Care Pathways, proving to be a useful support to the decision-making process in a particular diagnosis without representing medical knowledge as highlighted in the literature [68]. An important characteristic of the reasoning tasks in Description Logic (DL) is that they follow the open-world assumption (OWA): facts which cannot be deduced from a KB are assumed to be unknown but not necessarily false [59]. The opposite of the OWA is the closed-world assumption (CWA), which considers that everything that is not contained in the database is false.

L. E. Malgieri

In a medical reasoning, only evidence-based or expert-­ validated knowledge should be considered as true. The open-­ world assumption is not appropriate for medical knowledge, because the reasoning should not make new hypotheses about medical knowledge but is interesting when the reasoning is applied to the patient data as decision support in diagnostic systems or diagnosis phase. It allows the reasoner making hypotheses about yet-unknown patient disorders or clinical conditions. To checks the contraindications of a drug, the closed-world assumption (CWA), which considers that all contraindications are known and listed in official texts and drug databases, is appropriate even if this might not be entirely true. Through Diagnostic and Therapeutic Care Pathways by building ontologies which can be customized for each patient and enriching drug database, ontologies can support a personalized medicine. In ontologies, major medical concepts, such as disorders and drug therapies, can only be represented by classes, and subsequent levels of granularity are represented using is-a relationships. Disorders can be represented as classes and subclasses of Disorders. It is true that classes are usually more complex to manipulate than individuals, and therefore, this could complicate the use of ontologies in the biomedical domain, which can be overcome by appropriately creating concepts, classes, subclasses, and categories so as to be able to collect individual patient data in the form of individuals to which the inferential methods, SPARQL, and SWRL query can be applied.

3.2.4 Ontologies in Healthcare, Medical Data Collection Systems, and Their Use with Ontology-Based Symbolic AI Methods We do not yet live in a semantically driven world, so building ontologies and semantic networks is no small feat. Accessing and interconnecting all information sources, reconciling information systems, formats and syntaxes, extracting information from structured and unstructured data sources such as free text, and normalizing the semantic content of the resulting information are all complex and challenging problems. At the same time, the construction of an ontology is an interdisciplinary work between medical researchers and software engineers. On the one hand, physicians and researchers are used to being experts in their field, and in several cases believing that their opinion is incontrovertible; on the other hand, software engineers have to deal with the difficulties inherent in programming languages. This contextual experience over time may become biased and even dogmatic if one does not also take into account the changing nature of knowledge that evolves as new information becomes available and therefore requires “evolutionary maintenance” of both ontologies and software systems [58].

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

In recent years, there have been several approaches that have developed methodologies on how to use ontologies in healthcare and in clinical research. Several healthcare ontologies were developed. Many authors and research works have used ontologies to classify knowledge, literature, and medical dictionaries. An example is MeSH [69], the vocabulary of Medical Subject Headings developed and maintained by the US National Library of Medicine for the indexing and retrieval of literature. Health ontologies such as SNOMED-CT or MeSH provide a representation of the clinical contents to be used by the information systems. Open Biomedical Ontologies (OBO)  [70] is a set of ontologies developed and maintained by the scientific community to allow easier representation and integration of biomedical data. To clarify the terminology, the biomedical domain is broader than just the healthcare domain as it comprises other knowledge not only specific to patient medical care and EHRs. OBO Library,  the Open Biological and Biomedical Ontologies (OBO) library [71] consists of a number of ontologies that have been developed according to a set of agreed principles including complementarity and collaborative development. OntoBee  [72] is an ontology repository in which ontologies are presented as Linked Data and provides information about the classes and relations used by the OBO project. These ontologies are focused on some specific parts of EHR, but it could be beneficial, in the context of globally transferable healthcare data, to integrate systems that are using these ontologies. Human Disease Ontology (DO)  [73] is a part of OBO repository and expands the utility of the ontology for the examination and comparison of genetic variation, phenotype, protein, drug, and epitope data through the lens of human disease. Do integrates disease concepts from ICD, the National Cancer Institute (NCI) Thesaurus, SNOMED-CT, and MeSH extracted from the Unified Medical Language System (UMLS) based on the UMLS Concept Unique Identifiers for each disease term. DO also includes disease terms extracted directly from Online Mendelian Inheritance in Man (OMIM), an Online Catalog of Human Genes and Genetic Disorders [74], and the Experimental Factor Ontology (EFO) [75], which provides a systematic description of many experimental variables available in EBI databases and from whose site it is possible to download the latest version of EFO in OWL format (inferred) or the OBO format version, and Orphanet, the portal for rare diseases and orphan drugs [76].

39

Another ontology form the OBO repository is the Gene Ontology (GO) [77] that provides structured, controlled vocabularies and classifications used in the annotation of genes, gene products, and sequences. BioPortal  [78] is the world’s most comprehensive repository of biomedical ontologies. It contains 890 ontologies with a total of >13  million classes, 36,286 properties, and >55  million mapping. BioPortal can be used basing the search on both classes and ontologies. The Foundational Model of Anatomy Ontology (FMA)  [79] is an evolving computer-based knowledge source for biomedical informatics; it is concerned with the representation of classes or types and relationships necessary for the symbolic representation of the phenotypic structure of the human body in a form that is understandable to humans and is also navigable, parsable, and interpretable by machine-­ based systems. Specifically, the FMA is a domain ontology that represents a coherent body of explicit declarative knowledge about human anatomy. We have enumerated several projects and sites that have produced different types of ontologies, which we recommend analyzing before starting to build new ontologies. Some authors suggest that existing ontologies need to be harmonized [80]. The other fundamental aspect to be able to use ontologies is the classic and proper repositories of medicine, those that contain medical data, which we described in the previous chapter. Given the ontologies formulated in OWL/ RDF, many works have focused mainly on the interactions between ontologies expressed in OWL/RDF and communication standards: Health Level Seven (HL7) suite, openEHR, and CEN/ISO EN13606 standards, The Clinical Element Model (CEM) [81]. Another aspect is related to the structure of the databases containing medical data: the data contained, in order to be used with ontologies, must have a logical structure that can support reasoning methods and therefore designed to contain all the data of individuals and their properties defined through the use of an ontology-based medical data warehouse [82]. OntoNeo  [83] is an interesting ongoing initiative to build in the obstetric and neonatal domain, a formal ontology. The Obstetric and Neonatal Ontology is a structured controlled vocabulary to provide a representation of the data from ­electronic health records (EHRs) involved in the care of the pregnant woman, and of her baby. The development of OntoNeo is following the OBO Foundry principles, which aims to develop a set of interoperable ontologies for representation of biological and biomedical reality. The ongoing project employed Basic Formal Ontology (BFO) version 2.0 as top-­level ontology of OntoNeo, which is a large accep-

40

tance and use in medical and biological domains [84]. There are plans for making OntoNeo effective in the daily information processing routine, and they concern the application of the OntoNeo as a terminological reference for systems working on the continuity of women’s care and newborns in the Brazilian Health Unified System (SUS). The e-SUS strategy consists of two systems, the Citizen Electronic Health Record and the Simplified Data Collection. Although all these systems were planned to gather and unify information about prenatal care from health care facilities at all levels— national, provincial, and municipal levels—the healthcare facilities themselves are autonomous, and they have been establishing many different systems to deal with their EHR information. In practice, this situation requires a great deal of effort to unify EHRs at the national levels to ensure continuity of care in a complex network of systems present in the current Brazilian reality. Ontology-based decision support systems  In [85], the authors have developed a new knowledge-based intelligent system for obstetrics and gynecology ultrasound imaging, based on an ontology and a reference image collection. The study evaluates the new system to support accurate annotations of ultrasound images. The resulting ectopic pregnancy ontology consisted of 1395 terms, and 80 images were collected for the reference collection. The observers used the knowledge-based intelligent system to provide a total of 1486 sign annotations.

3.2.5 Ontology Software Language, Ontology Editor, and Ontology Reasoner Several approaches have been proposed for accessing an ontology in a computer program. The software used in the emPhasys [86] Project is Apache Stanbol [87], a set of reusable components for semantic content management, and the MeSH and SNOMED-CT as medical vocabularies. Many projects have used traditional API (Application Programming Interface) for OWL, such as OWL API [88] in Java. Ontology-oriented programming tries to unify the ontology with the object model of the programming language. This approach exploits the similarities between ontologies and the object-oriented programming paradigm [89]: classes, properties, and individuals in ontologies correspond to classes, attributes, and instances in object models like Python. Programming in Python also allows you to spend about a third of the time needed to program in Java, which also means a clear reduction in the percentage of bugs in your code. Another characteristic is the dynamic approach. Using dynamic programming languages, it is possible to generate classes and instances from the ontology at run time.

L. E. Malgieri

Vocabularies are the basic building blocks for inference techniques on the Semantic Web. Inference [90] means that automatic procedures can generate new relationships based on the data or discovering possible inconsistencies in the data. W3C offers a large palette of techniques to describe and define different forms of vocabularies in a standard format; these include RDF (Resource Description Framework) and RDF Schemas [91], Simple Knowledge Organization System (SKOS), Web Ontology Language (OWL), and the Rule Interchange Format (RIF). The most popular editor for ontologies is Protégé-2000 [92], developed by Mark Musen’s group at Stanford Medical Informatics. We generated some of the figures with the OWLViz [93] plug-in on Protégé-2000. Protege contains internal reasoners: Pellet [94], an Open-Source OWL DL reasoner for Java, HermiT [95], and others. Owlready2 [88] appears to be one of the most advanced approaches. It is able to perform automatic classification not only on individuals but also on classes, it is able to perform local closed-world reasoning, and it proposes a high-level syntax for defining “role-filler” constraints. These points are crucial when working on medical ontologies.

3.2.6 New Frontiers for Ontology Reasoning from Symbolic AI to Non-symbolic AI Traditionally, logic-based symbolic methods from the field of knowledge representation and reasoning have been used to equip agents with capabilities that resemble human logical reasoning qualities. More recently, however, there has been an increasing interest in using machine learning rather than logic-based symbolic formalisms to tackle these tasks. Most ontologies in the life sciences are encoded using the OWL language that is a part of the Semantic Web and based on using the common syntax of Description Logics which uses symbols for expressions (Symbolic Artificial Intelligence). Syntactic constructs assign an interpretation in a mathematical structure that resembles a world in which these constructs are true. For example, if we want to assert that class A is-a part of class B we can symbolically write A ⊑ B. New studies are bringing mathematical methods of deep learning applied to ontologies [96], and DeepGO [97], a novel method for predicting protein functions using protein sequences and protein–protein interaction (PPI) networks. It uses deep ­neural networks to learn sequence and PPI network features and hierarchically classifies it with GO classes. PPI network features are learned using a neuro-symbolic approach for learning knowledge graph representations [98]. In the use of ontologies, the borderline between symbolic and non-­ symbolic artificial intelligence is blurring: more and more researchers are developing models of graphical neural networks which we will discuss later [99].

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

3.3 Machine Learning The definition given by J. McCarthy, in 2007, and also shared by the FDA: “Machine Learning, an artificial intelligence technique that can be used to design and train software algorithms to learn from and act on data” (McCarthy [3]) not only can it still be considered up to date, but in recent years, it has become considerably richer in terms of algorithms thanks to the use of software algorithms known as deep learning and neural networks. Compared to classical machine learning algorithms, the growth of neural network algorithms and models has been due to the great attention paid by researchers and companies to computer vision, image processing and classification, and natural language processing for both machine translation and speech recognition, accompanied by a rapid growth in computational resources in terms of processors and hardware for the enormous amount of data that these fields of application produce. In the following, we aim to provide a comprehensive but accessible overview of many of the fundamental concepts of machine learning and deep learning, and to provide references and pointers for further investigation. The algorithms that we will describe in this paragraph are mainly used on existing databases, often created for other purposes. Many works have been published and continue to be published on the use of machine learning algorithms; some of them also contain references or evidence of the use of methods more proper to statistical science. It is true that the methods of investigation, whether of machine learning or not, have parameters of “self-regulation,” which generally measure the quality of the algorithms’ outputs. It is also true that an algorithm is not always appropriate for the purpose, but on the other hand, it is also true that very often the available databases were born for other purposes. The relatively young deep learning, which we will deal with in the next paragraph, it is as if, compared to the machine learning, it had another advantage, apart from that of being more powerful and more performing for some purposes: the data that is taken as input is “born” together with the algorithms. It is as if the data and the algorithms were born almost at the same time, allowing the full potential of the algorithm is to be exploited. Machine Learning Algorithms helps computer system to learn without being explicitly programmed for mapping inputs to predicted outputs. The first step, data analysis, is often understanding—identifying the underlying mechanisms that give rise to patterns or hidden properties in the data and, exploring and visualizing the data, the dimensionality reduction. Many studies compared the performance of Machine Learning Algorithms for the development of diagnostic or prognostic clinical prediction models for binary outcomes based on clinical data [100], other [101]: the robustness and consistency of a variety of machine learning and statistical

41

models on individual risk prediction and the effects of censoring on risk predictions and what to expect from Machine Learning in fetal cardiology in [102].

3.3.1 Supervised Machine Learning Algorithms This is the most used machine learning algorithm. It is called supervised because the process of algorithm learning from the training dataset can be thought of as a teacher supervising the learning process. In this kind of ML algorithm, the possible outcomes are already known, and training data are also labeled with correct answers. It can be understood as follows. Suppose we have input variables x and an output variable y, and we applied an algorithm to learn the mapping function from the input to output such as Y = f(x). The goal is to approximate the function so well that when we have new input data (x), we can predict the output variable (Y) for that data. Mainly supervised learning problems can be divided into the following two kinds of problems:

3.3.1.1 Classification Classification, when the goal is the categorized output such as “cured,” “sick,” and “chronic.” After implementing a machine learning algorithm, we need to find out how effective the model is. The criteria for measuring the effectiveness may be based upon datasets and metric. For evaluating different machine learning algorithms, we can use different performance metrics. For example, suppose if a classifier is used to distinguish between images of different objects, we can use the classification performance metrics such as average accuracy and AUC. In one or other sense, the metric we choose to evaluate our machine learning model is very important because the choice of metrics influences how the performance of a machine learning algorithm is measured and compared. Following are some of the metrics. Confusion Matrix Basically, it is used for classification problem where the output can be of two or more types of classes. It is the easiest way to measure the performance of a classifier. A confusion matrix is basically a table with two dimensions namely “Actual” and “Predicted.” Both the dimensions have “True Positives (TP),” “True Negatives (TN),” “False Positives (FP),” and “False Negatives (FN)” (Fig. 3.4). In the confusion matrix above, 1 is for positive class and 0 is for negative class. Following are the terms associated with Confusion matrix: • True Positives—TPs are the cases when the actual class of data point was 1 and the predicted is also 1.

42

L. E. Malgieri

• True Negatives—TNs are the cases when the actual class of the data point was 0 and the predicted is also 0. • False Positives—FPs are the cases when the actual class of data point was 0 and the predicted is also 1. • False Negatives—FNs are the cases when the actual class of the data point was 1 and the predicted is also 0.

Recall or Sensitivity

It may be defined as how many of the positives do the model return. Following is the formula for calculating the recall/ sensitivity of the model: Recall = TP(TP + FN) Specificity

Accuracy

The confusion matrix itself is not a performance measure as such but almost all the performance matrices are based on the confusion matrix. One of them is accuracy. In classification problems, it may be defined as the number of correct predictions made by the model over all kinds of predictions made. The formula for calculating the accuracy is as follows: Accuracy = (TP + TN)/(TP + FP + FN + TN) Precision

It is mostly used in document retrievals. It may be defined as how many of the returned documents are correct. Following is the formula for calculating the precision: Precision = TP(TP + FP)

1 1

ACTUAL

0

True Positives (TP)

False Positives (FP)

False Negatives (FN)

True Negatives (TN)

PREDICTED 0

Fig. 3.4  Confusion matrix

It may be defined as how many of the negatives do the model return. It is exactly opposite to recall. Following is the formula for calculating the specificity of the model: Specificity = TN(TN + FP) Class Imbalance Problem

Class imbalance is the scenario where the number of observations belonging to one class is significantly lower than those belonging to the other classes. For example, this problem is prominent in the scenario where we need to identify the rare diseases, fraudulent transactions in bank, etc. Ensemble Techniques This methodology basically is used to modify existing classification algorithms to make them appropriate for imbalanced datasets. In this approach, we construct several two-stage classifiers from the original data and then aggregate their predictions. Random forest classifier is an example of ensemble-based classifier (Fig. 3.5).

3.3.1.2 Regression Regression, when the goal is the real value output, there is one of the most important statistical and machine learning tools and may be defined as the parametric technique that allows to make decisions based upon data or to make predictions based upon data by learning the relationship between DATA

Classifiers

Fresh Sample of Data

Fig. 3.5  Ensemble techniques

C1

C2

Total Votes

Cn

Fresh Sample of Data

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

input and output variables. The output variables dependent on the input variables and are real numbers. In regression, the relationship between input and output variables matters and it helps us in understanding how the value of the output variable changes with the change of input variable. Regression is frequently used for prediction of prices, economics, variations, and so on. Decision tree, Random Forest, kNN, and Logistic Regression are the examples of supervised machine learning algorithms. An evaluation of multivariable logistic regression analyses in the obstetrics and gynecology literature are in [103].

43

• Multiple linear regression—A linear regression algorithm is called multiple linear regression if it is having more than one independent variable.

Linear regression is mainly used to estimate the real values based on continuous variable(s). For example, the total sale of a shop in a day, based on real values, can be estimated by linear regression. The logistic regression, measuring the relationship by estimating the probabilities using a logistic function, is a classification algorithm that is used to estimate the discrete values like 0 or 1, true or false, yes or no, based on a given set of independent variables, predicting the probability; 3.3.1.3 Supervised Learning hence, its output lies in between 0 and 1. Decision Trees, Random Forests, and Gradient Boosting Regression can be used to predict disease presence (diagare a popular class of methods in supervised learning. These nosis) or disease outcomes (prognosis) [107]. methods can also work with regression/classification tasks These methods may be a starting point to test out a very and are well suited to model nonlinear relations between the simple version of the full problem. Due to their simplicity, input features and output predictions. Random forests, which linear and logistic regressions are highly interpretable and ensemble decision trees, can often be preferred to deep learn- provide straightforward ways to perform feature attribuing methods in settings where the data has a low signal-to-­ tion. Some variants are Count Regression, specializes in noise ratio. These methods can typically be less interpretable target variables with Poisson distributions, and Gamma than linear/logistic regression, but recent work [104] has Regression, specializes in target variables with Gamma looked at developing software libraries [105] to address this distributions. challenge. In [106], it is possible to find code for Linear Here, if we talk about dependent and independent variModels, Linear and Quadratic Discriminant Analysis, Kernel ables, then dependent variable is the target class variable that ridge regression, Support Vector Machines, Stochastic we are going to predict, and on the other side, the indepenGradient Descent, k-Nearest Neighbors, Gaussian Processes, dent variables are the features that we are going to use to Naive Bayes, Decision Trees, Ensemble methods, Multiclass predict the target class. and multioutput algorithms, Feature selection, Semi-­ supervised learning, Isotonic regression, Probability calibra- Decision Tree and Random Forest Classifier tion, and Neural network models (supervised). Decision Tree  is a binary tree flowchart where each node splits a group of observations according to some feature variLinear Regression and Logistic Regression (and Variants!) able. It does not require scaling or normalization and is not They are easy to interpret, quick runtime, may be particu- sensitive to missing values or outliers. It is a classifier larly useful when there are limited data, and a clear set of expressed as recursive partition based on the independent features. Linear regression requires linear relationships variables. Decision tree has nodes which form the rooted between variables and its poor for imbalanced datasets. tree. Rooted tree is a directed tree with a node called “root.” Logistic regression [100], estimating the probabilities Root does not have any incoming edges and all the other between dependent variables and independent variables, nodes have one incoming edge. These nodes are called leaves means to predict the likelihood occurrence of the event, use or decision nodes. the sigmoid curve, a logistic function to build the function with various parameters. Mainly linear regression is a linear Random Forest  is a collection of decision trees. It is better model that assumes a linear relationship between the input than single decision tree because while retaining the predicvariables say x and the single output variable say y. In other tive powers, self-selective for features, prone to overfitting words, we can say that y can be calculated from a linear com- and can reduce overfitting by averaging the results, same bination of the input variables x. The relationship between benefits as Decision Tree, more difficult to explain than variables can be established by fitting a best line. Linear Decision Trees. regression is of the following two types: Naïve Bayes Classifier • Simple linear regression—A linear regression algorithm Naïve Bayes is a classification technique used to build clasis called simple linear regression if it is having only one sifier using the Bayes theorem. The assumption is that the independent variable. predictors are independent. In simple words, it assumes that

L. E. Malgieri

44

the presence of a particular feature in a class is unrelated to the presence of any other feature. There are three types of Naïve Bayes models named Gaussian, Multinomial, and Bernoulli under scikit learn package. Support Vector Machines (SVM) Support vector machine (SVM) is a supervised machine learning algorithm that can be used for both regression and classification, but mainly it is used for classification problems. The main concept of SVM is to plot each data item as a point in n-dimensional space with the value of each feature being the value of a particular coordinate. Here, n would be the features we would have. Following is a simple graphical representation to understand the concept of SVM (Fig. 3.6). In the above diagram, we have two features. Hence, we first need to plot these two variables in two-dimensional space where each point has two coordinates, called support vectors. The line splits the data into two different classified groups. This line would be the classifier. Here, we are going to build an SVM classifier by using scikit learn and iris dataset. Scikit learn library has the sklearn.svm module and provides sklearn.svm.svc for classification. The SVM classifier to predict the class of the iris plant based on four features is shown below. K-Nearest Neighbors (KNN) It is used for both classification and regression of the problems. It is widely used to solve classification problems. The main concept of this algorithm is that it is used to store all the available cases and classifies new cases by majority votes of its k neighbors. The case is being then assigned to the class which is the most common among its k-nearest neighbors, measured by a distance function. The distance function can

Support Vectors

Fig. 3.6 SVM

be Euclidean, Minkowski, and Hamming distance. Consider the following to use KNN: • Computationally KNN are expensive than other algorithms used for classification problems. • The normalization of variables needed otherwise higher range variables can bias it. • In KNN, we need to work on pre-processing stage like noise removal. Finding Nearest Neighbors. If we want to build recommender systems such as a movie recommender system, then we need to understand the concept of finding the nearest neighbors. It is because the recommender system utilizes the concept of nearest neighbors. The concept of finding nearest neighbors may be defined as the process of finding the closest point to the input point from the given dataset. The main use of this KNN (k-nearest neighbors) algorithm is to build classification systems that classify a data point on the proximity of the input data point to various classes. The Python code given below helps in finding the k-­ nearest neighbors of a given dataset using the Nearest Neighbors module from the scikit learn package [108, 109] and provides functionality for unsupervised and supervised neighbors-based learning methods. Unsupervised nearest neighbor is the foundation of many other learning methods, notably manifold learning and spectral clustering.

3.3.2 Unsupervised Machine Learning Algorithms As the name suggests, these kinds of machine learning algorithms do not have any supervisor to provide any sort of guidance. That is why unsupervised machine learning algorithms are closely aligned with what some call true artificial intelligence. It can be understood as follows: Suppose we have input variable x, then there will be no corresponding output variables as there is in supervised learning algorithms. In simple words, we can say that in unsupervised learning, there will be no correct answer and no teacher for the guidance. Algorithms help discover interesting patterns in data. Unsupervised learning problems can be divided into the following two kinds of problem: clustering and association.

3.3.2.1 Clustering Clustering—In clustering problems, we need to discover the inherent groupings in the data. For example, grouping customers by their purchasing behavior. Basically, it is a type of unsupervised learning method and a common technique for statistical data analysis used in many fields. Clustering mainly is a task of dividing the set of observations into sub-

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

sets, called clusters, in such a way that observations in the same cluster are similar in one sense, and they are dissimilar to the observations in other clusters. In simple words, we can say that the main goal of clustering is to group the data on the basis of similarity and dissimilarity. Each clustering algorithm [110], k-means, Affinity Propagation, Mean Shift, Spectral clustering, Hierarchical clustering, DBSCAN, and OPTICS BIRCH come in two variants: a class that implements the fit method to learn the clusters on train data, and a function that, given train data, returns an array of integer labels corresponding to the different clusters. Clustering of unlabeled data can be performed with the module sklearn cluster [111]. For example, the following diagram shows similar kind of data in different clusters (Fig. 3.7). Measuring the Clustering Performance The real-world data is not naturally organized into number of distinctive clusters. Due to this reason, it is not easy to visualize and draw inferences. That is why we need to measure the clustering performance as well as its quality. It can be done with the help of silhouette analysis. Silhouette Analysis This method can be used to check the quality of clustering by measuring the distance between the clusters. Basically, it provides a way to assess the parameters like number of clusters by giving a silhouette score. This score is a metric that measures how close each point in one cluster is to the points in the neighboring clusters. Analysis of Silhouette Score The score has a range of [−1, 1]. Following is the analysis of this score

45

• Score of +1—Score near +1 indicates that the sample is far away from the neighboring cluster. • Score of 0—Score 0 indicates that the sample is on or very close to the decision boundary between two neighboring clusters. • Score of −1—Negative score indicates that the samples have been assigned to the wrong clusters. Calculating Silhouette Score In this section, we will learn how to calculate the silhouette score. Silhouette score can be calculated by using the following formula silhouette score = (p − q)/ max (p, q) Here, 𝑝 is the mean distance to the points in the nearest cluster that the data point is not a part of. And 𝑞 is the mean intra-cluster distance to all the points in its own cluster.

3.3.2.2 Association Association—A problem is called association problem because such kinds of problem require discovering the rules that describe large portions of our data, for example, finding the customers who buy both x and y. k-means for clustering and apriori algorithm for association are the examples of unsupervised machine learning algorithms. 3.3.2.3 Unsupervised Learning K-Means Algorithm K-means clustering algorithm is one of the well-known algorithms for clustering the data. We need to assume that the numbers of clusters are already known. This is also called flat clustering. It is an iterative clustering algorithm. The steps given below need to be followed for this algorithm: • K-means picks k number of points for each cluster known as centroids. • Now each data point forms a cluster with the closest centroids, i.e., k clusters. • Now, it will find the centroids of each cluster based on the existing cluster members. • We need to repeat these steps until convergence occurs. As this is an iterative algorithm, we need to update the locations of k centroids with every iteration until we find the global optima, or in other words, the centroids reach at their optimal locations (Fig. 3.8) [112].

Fig. 3.7 Clustering

Mean Shift Algorithm It is another popular and powerful clustering algorithm used in unsupervised learning. It does not make any assumptions; hence, it is a non-parametric algorithm. It is also called hierarchical clustering or mean shift cluster analysis. Followings would be the basic steps of this algorithm

46

• First of all, we need to start with the data points assigned to a cluster of their own. • Now, it computes the centroids and update the location of new centroids. • By repeating this process, we move closer the peak of cluster, i.e., toward the region of higher density. • This algorithm stops at the stage where centroids do not move anymore. • Code implementing Mean Shift clustering algorithm in Python is available in [113] (Fig. 3.9). Fig. 3.8  K-Means Algorithm

Fig. 3.9  Mean Shift Algorithm

L. E. Malgieri

3.3.3 Reinforcement Machine Learning Algorithms These kinds of machine learning algorithms are used very less. These algorithms train the systems to make specific decisions. Basically, the machine is exposed to an environment where it trains itself continually using the trial-and-­error method. These algorithms learn from past experience and tries to capture the best possible knowledge to make accurate decisions [114]. Markov

47

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

Decision Process is an example of reinforcement machine learning algorithms. This type of learning is used to reinforce or strengthen the network based on critic information. That is, a network being trained under reinforcement learning, receives some feedback from the environment. However, the feedback is evaluative and not instructive as in the case of supervised learning. Based on this feedback, the network performs the adjustments of the weights to obtain better critic information in future. This learning process is similar to supervised learning, but we might have very less information. Figure 3.10 gives the block diagram of reinforcement learning.

3.3.3.1 Building Blocks: Environment and Agent Environment and Agent are main building blocks of reinforcement learning in AI [115]. 3.3.3.2 Agent The principles underlying almost all intelligent systems are simple reflex agents, model-based reflex agents, goal-based agents, and utility-based agents. An agent is anything that can perceive its environment through sensors and acts upon that environment through effectors. • A human agent has sensory organs such as eyes, ears, nose, tongue, and skin parallel to the sensors, and other organs such as hands, legs, mouth, for effectors. • A robotic agent replaces cameras and infrared range finders for the sensors, and various motors and actuators for effectors. • A software agent has encoded bit strings as its programs and actions.

3.3.3.3 Agent Terminology The following terms are more frequently used in reinforcement learning in AI

• Performance Measure of Agent—It is the criteria, which determines how successful an agent is. • Behavior of Agent—It is the action that agent performs after any given sequence of percepts. • Percept—It is agent’s perceptual inputs at a given instance. • Percept Sequence—It is the history of all that an agent has perceived till date. • Agent Function—It is a map from the precept sequence to an action.

3.3.3.4 Environment Some programs operate in an entirely artificial environment confined to keyboard input, database, computer file systems and character output on a screen. In contrast, some software agents, such as software robots or softbots, exist in rich and unlimited softbot domains. The simulator has a very detailed, and complex environment. The software agent needs to choose from a long array of actions in real time. For example, a softbot designed to scan the online preferences of the customer and display interesting items to the customer works in the real as well as an artificial environment. 3.3.3.5 Properties of Environment The environment has multifold properties as discussed below: • Discrete/Continuous—If there are a limited number of distinct, clearly defined, states of the environment, the environment is discrete, otherwise it is continuous. For example, chess is a discrete environment and driving is a continuous environment. • Observable/Partially Observable—If it is possible to determine the complete state of the environment at each time point from the percepts, it is observable; otherwise, it is only partially observable.

Fig. 3.10  Basics of Reinforcement Learning X (input)

Neural Network

Y (Actual Output)

Error Signal Error Signal Generator

48

L. E. Malgieri

• Static/Dynamic—If the environment does not change while an agent is acting, then it is static; otherwise, it is dynamic. • Single agent/Multiple agents—The environment may contain other agents which may be of the same or different kind as that of the agent. • Accessible/Inaccessible—If the agent’s sensory apparatus can have access to the complete state of the environment, then the environment is accessible to that agent; otherwise, it is inaccessible. • Deterministic/Non-deterministic—If the next state of the environment is completely determined by the current state and the actions of the agent, then the environment is deterministic; otherwise, it is non-deterministic. • Episodic/Non-episodic—In an episodic environment, each episode consists of the agent perceiving and then acting. The quality of its action depends just on the episode itself. Subsequent episodes do not depend on the actions in the previous episodes. Episodic environments are much simpler because the agent does not need to think ahead (Fig. 3.11).

3.3.3.6 Constructing an Environment with Python For building reinforcement learning agent, we will be using the OpenAI Gym [116] package which can be installed with the help of the following command: pip install gym

There are various environments in OpenAI gym which can be used for various purposes. Few of them are Cartpole-v0, Hopper-v1, and MsPacman-v0. They require different engines. The detail documentation of OpenAI Gym can be found on [117]. The cartpole can balance itself (Fig. 3.12) [118].

Sensors

Fig. 3.12  Constructing a learning agent with Python

3.4 Deep Learning The definition given by J. McCarthy, in 2007, and also shared by the FDA [119]: “Machine Learning, an artificial intelligence technique that can be used to design and train software algorithms to learn from and act on data” (McCarthy [3]) does not explicitly contain the term Deep Learning. Indeed, Deep Learning is considered in the literature to be a subset of machine learning in which algorithms are inspired by the structure and function of human brain.

3.4.1 Introduction to Deep Learning Deep Learning is considered by all to be a subset of Machine Learning and has recently been growing considerably in terms of both, methodologies and types and methods of use, applicable to any type of input, be it an image, a sound clip, or an unordered collection of features: whatever their dimensionality, their representation can always be flattened into a vector before the transformation, which can be done by using some mathematical methods as follows:

Percepts

Effectors Environment Actions

Fig. 3.11  Properties of environment

• the matrix calculus to understand the training of deep neural networks [120]; • partial derivatives; and • the affine transformations: a vector is received as input and is multiplied with a matrix to produce an output, to which a bias vector is usually added before passing the result through a non-linearity [121].

49

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

Neural networks consist of many units, organized into multiple collections of neurons called layers: input layer, hidden(s) layer, and output layer. In each layer, the number of neurons may change depending on the data and algorithms used, as well as the number of hidden layers. There is nothing mysterious about the term “hidden layer”: these are all the mathematical steps involved in a classic resolution, which can be rearranged according to the problem posed and the problem itself [122]. For example, in a single neuron, the function is y(x) = ax + c. But this is just one neuron, and neural networks must train the weights and biases of all neurons in all layers simultaneously. Because there are multiple inputs and (potentially) multiple network outputs, we really need general rules for the derivative of a function with respect to a vector and even rules for the derivative of a vector-valued function with respect to a vector (Fig. 3.13). Training this neuron means choosing weights a and bias c so that we get the desired output for all N inputs x from one layer to the next layer. Computation unit in a neural network is typically calculated using the dot product (from linear algebra) of an edge weight vector w with an input vector x plus a n scalar bias (threshold): z ( x ) = ∑ i wi xi + b = w ⋅ x + b . Function z(x) is called the unit’s affine function and is followed by a rectified linear unit, which clips negative values to zero: max(0, z(x)). Such a computational unit is sometimes referred to as an “artificial neuron” and looks like (Fig. 3.14). Neural networks consist of many of these units, organized into multiple collections of neurons called layers. The activation of one layer’s units becomes the input to the next layer’s Input layer

Input layer 1

Fig. 3.13  Deep Neural network with three hidden layers

units. The activation of the unit or units in the final layer is called the network output. Training this neuron means choosing weights w and bias b so that we get the desired output for all N inputs x. To do that, we minimize a loss function that compares the network’s final activation (x) with the target (x) (desired output of x) for all input x vectors. To minimize the loss, we use some variation on gradient descent, such as plain stochastic gradient descent (SGD), SGD with momentum, or Adam. All of those require the partial derivative (the gradient) of activation (x) with respect to the model parameters w and b. Our goal is to gradually tweak w and b so that the overall loss function keeps getting smaller across all x inputs. If we are careful, we can derive the gradient by differentiating the scalar version of a common loss function (mean-­ squared error) (Fig. 3.15): Handwritten digit recognition is a classic problem in the field of image recognition. The shape of the digits and its features helps identify the digit from the strokes and boundaries. There have been great achievements in recent years in the field of pattern recognition, particularly in the field of Handwritten digit recognition problem. Handwriting recognition is the ability of a device to take handwriting as input from sources. The handwriting taken as input can be used to verify signatures, used to interpret text and OCR (optical character recognition) to read the text and transform it into a form which can be manipulated by computer [123, 124]. The handwritten digit recognition system uses the MNIST dataset [125]. It has 70,000 images that can be used to train and evaluate the system. The train set has 60,000 images, pattern

Input layer 2

Input layer 3

output layer

50

L. E. Malgieri

training set contained examples from approximately 250 writers, and the test set has 10,000 images; and the sets of writers of the training set and test set were disjoint. It is the subset of NIST dataset (National institute of standards and technology), having 28 × 28 size input images and ten class labels from 0 to 9. Therefore, the size of image is 28 × 28 pixel square, i.e., 784 pixels.

X1 X2

W1

The handwritten digits are not always of the same size, width, orientation, and justified to margins as they differ from writing of person to person, so the general problem would be while classifying the digits due to the similarity between digits such as 1 and 7, 5 and 6, 3 and 8, 2 and 5, 2 and 7, etc. This problem is faced more when many people write a single digit with a variety of different handwritings. Lastly, the uniqueness and variety in the handwriting of different individuals also influence the formation and appearance of the digits (Fig. 3.16). When a computer or system takes an image, it just sees an array of pixel values (see Fig. 3.17). Each of these numbers is assigned with a value of 0–255 as pixel intensities at that point as from RGB color code. Background as white (0 value from RGB) and foreground as black (255 value from RGB). The algorithm [126] breaks the image into small image tiles, similar to sliding window, feeding each tiny tile into the smaller size neural network, and saving the results from each small tile into a new array. The multiple occurring of these layers, grouped by purpose, and the tasks entrusted to them define how deep and the name of the neural network.

activation function

W2 Wn

+

b

activation (X)

Xn affine function Fig. 3.14  Computational unit referred to an “artificial neuron”

Fig. 3.15  Gradient mean-­ squared error

Fig. 3.16  Sample of handwritten digits in MNIST dataset

_ 1 N

_ (target(x) - activation(x)) 2 = 1 N

|x|

(target(x) - max(x0, wi xi + b)) 2 i

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

51

Fig. 3.17  Image and pixel

• Multilayer Perceptrons the basic feed-forward neural networks, with just one hidden layer, are considered the universal approximators and they can represent combinations of step functions, allowing to approximate any continuous function with arbitrary precision [127]. • Artificial Neural Networks (ANNs) with multilayer perceptron utilize matrix multiplications interleaved with a nonlinear transform. For their simplicity, they are used in many researches for problems where the data might consist of a set of distinct features for problems of the classification, logistic/linear regression, and are often used and compared in terms accuracy, time, specificity, sensitivity, and other parameters’ comparison with Machine Learning algorithms (see Sect. 3.3) such as Support Vector Machine, Random Forest, k-Nearest Neighbors, Naive Bayes, Decision Tree, or Logistic or Logistics Regression. • Recurrent Neural Networks (RNNs) are the model successfully used on many fields and application for sequential data and prediction problems where the tasks consist in transforming one sequence to another or determining

sequence similarities and each token cause an update of the internal cell state. Updated and output emitted are controlled by gating functions. RNNs have found several scientific applications for data with sequential structure, such as in genomics and proteomics. • Convolutional Neural Networks (CNNs) are become famous among the recent times: have spread as Facebook for their automatic tagging algorithms, Google for photo search, Amazon for their product recommendations, Pinterest for their home feed personalization, and Instagram for search infrastructure. Each neural network is built for a specific purpose, and it is easy to deduce that a training dataset is needed for the large number of parameters and functions involved in the mathematical formulations mentioned above. Image classification or object recognition consists of passing an image, including video prediction, action recognition, and style transfer [128], transformed by computer vision from image in a matrix of pixel as a parameter and predicting whether a condition is satisfied or not (cat or not,

52

L. E. Malgieri

8

Input

8x8

12x12

24x24

x2

28

Fully Connected Layer

Pooling Layer

20 20

10

Log Softmax

10

Flatten

10 filters

Convolution (5x5 kernel)

10

Output

Convolution Layer

Dropout (0.25)

Pooling Layer

4x4

Convolution Layer

Max Pooling (2x2)

Convolution (5x5 kernel)

Max Pooling (2x2)

Fig. 3.18  CNN architecture in MNIST dataset

dog or not), or the probability or most satisfying condition for an image. The performance or efficiency of a classifier is shown by various features which tells how well working the particular classifier. Confusion matrix, also same as error matrix, shows that what percent of predictions made by the classifier was correct and where it was difficult for the classifier to predict the actual classification. If T is a total of image to be classified, TP (True positive) is a correct identification of positive labels, TN (True negative) is a correct identification of negative labels, FP (False positive) is incorrect identification of positive labels, and FN (False negative) is incorrect identification of negative labels. Statistical parameters are Accuracy, (TP + TN)/T, Sensitivity, (TP)/(TP + FN), Specificity  =  (TN)/(FP  +  TN), Prevalence  =  (TP  +  FN)/T. Others statistical parameters are Positive predicted values, Negative predicted values, Detection rate, Expected accuracy, Kappa statistic, or a value that compares an observed accuracy with an expected accuracy. They are all parameters for a comparison among different classifiers used in the training set (Fig. 3.18).

3.4.2 Image Classification and Object Detection Image classification, object detection, image segmentation, super resolution, and image registration of medical imaging supported by neural networks have been the subject of numerous research papers and publications in recent years in various medical specialties. Image classification is a common application of deep learning: for the image in input, the output is a set of labels for that image. There are many different types of high-performing and well-established Convolutional Neural Network CNN models for classification and to predict disease labels analyzing brain scans, PET, and fMRI Functional Magnetic Resonance Imaging [129, 130]. Other examples of Artificial Intelligence

for clinicians: Radiology/neurology, Pathology, Dermatology, Ophthalmology, Gastroenterology, Cardiology, and a review of publications of AI algorithms compared with doctors is given by [131]. Object detection represents a deepening and looks at identifying, localizing, and categorizing different objects or the kind of object they contain in the image: within the image, for example the brain, I want to locate tumor cells across different imaging modalities [132] or automated fracture detection and localization on wrist radiographs [133] or intervertebral disk detection in X-ray images using faster r-CNN [134]. A source of code and models is in [135].

3.4.3 Image Segmentation A further step is image segmentation. The purpose of image segmentation is to characterize each pixel into which an image is transformed with additional information relating to the position in which the pixel was located in the original figure, thus, remembering both the spatial coordinates and a subgroup of pixels (or voxel) in which it was allocated. In image segmentation, the most appropriate is the Convolution Neural network (CNN), because it is necessary to categorize every pixel (or voxel) and preserve spatial information about the image. Some architectures for scientific applications are the FCNs (Fully Convolutional Networks) [136], the SegNet [137], and the Object-Contextual Representations for Semantic Segmentation [138]. The U-net [139], a sliding-window convolutional network for segmentation of neuronal structures in electron microscopic stacks, in some of the scientific applications has also been successful for super resolution. Super resolution is a technique for transforming low-resolution images to high-­ resolution images [140]. Segmentation with a Convolutional Neural Network in MR brain Neonatal images, in total 22 images of preterm infants, was acquired in accordance with standard clinical

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

practice in the neonatal intensive care unit of the University Medical Center Utrecht (UMCU), The Netherlands [141]. Segmentation in MR is important for quantitative analysis in large-scale studies with images acquired at all ages, is applied to five different data sets: coronal T2-weighted images of preterm infants acquired at 30  weeks postmenstrual age (PMA) and 40  weeks PMA, axial T2-weighted images of preterm infants acquired at 40 weeks PMA, axial T1-weighted images of aging adults acquired at an average age of 70  years, and T1-weighted images of young adults acquired at an average age of 23 years. The neonatal images were manually segmented by experts into eight classes: cerebellum (CB), myelinated white matter (mWM), basal ganglia and thalami (BGT), ventricular cerebrospinal fluid (vCSF), unmyelinated white matter (uWM), brain stem (BS), cortical gray matter (cGM), and extracerebral cerebrospinal fluid (eCSF). To ensure that the method obtains accurate segmentation details as well as spatial consistency, the Convolutional Neural Network uses multiple patch sizes and multiple convolution kernel sizes to acquire multi-scale information about each voxel. The method obtained the following average Dice coefficients over all segmented tissue classes for each dataset, respectively: 0.87, 0.82, 0.84, 0.86, and 0.91. The results demonstrate that the method obtains accurate segmentations in all five sets and hence demonstrates its robustness to differences in age and acquisition protocol. Several algorithms for brain segmentation in preterm born infants have been published [142]. In the NeoBrainS12 study [143], three different image sets of preterm born infants were set up to provide such a comparison. These sets are (1) axial scans acquired at 40  weeks corrected age, (2) coronal scans acquired at 30  weeks corrected age, and (3) coronal scans acquired at 40 weeks corrected age. Each of these three sets consists of three T1- and T2-weighted MR images of the brain. The task was to segment cortical gray matter, non-myelinated and myelinated white matter, brainstem, basal ganglia and thalami, cerebellum, and cerebrospinal fluid in the ventricles and in the extracerebral space separately. Convolutional Neural Network has been used for automatic segmentation of neuroanatomy. In [144], the authors introduce DeepNAT, a 3D Convolutional Neural Network for the in T1-weighted Magnetic Resonance images. This approach to brain segmentation that jointly learns an abstract feature representation and a multi-class classification do not only predict the center voxel of the patch but also neighbors, which is formulated as multi-task learning. They arrange two networks hierarchically, where the first one separates foreground from background, and the second one identifies 25 brain structures on the foreground. Since patches lack spatial context, they augment them with coordinates and introduce a novel intrinsic parameterization of the brain vol-

53

ume, formed by eigenfunctions of the Laplace-Beltrami operator. The CNN architecture uses three convolutional layers with pooling, batch normalization, and nonlinearities, followed by fully connected layers with dropout. The final segmentation is inferred from the probabilistic output of the network with a 3D fully connected conditional random field, which ensures label agreement between close voxels. The roughly 2.7  million parameters in the network are learned with stochastic gradient descent. The results show that DeepNAT compares favorably to state-of-the-art methods. Finally, the purely learning-based method may have a high potential for the adaptation to young, old, or diseased brains by fine tuning the pre-trained network with a small training sample on the target application, where the availability of larger datasets with manual annotations may boost the overall segmentation accuracy in the future. The results demonstrated the high potential of convolutional neural networks for segmenting neuroanatomy. Network definitions and trained networks are available for download: https:// tjklein.github.io/DeepNAT/. Other application in image segmentation with CNN is to identify key regions of cells in different tissues, segmenting and classifying epithelial and stromal regions in histopathological images [145], segmentation of cervical cytoplasm and nuclei [146], segmentation Nuclei in Feulgen-Stained Images [147] and even studying bone structure, 3D-deformable approach for tissue segmentation in musculoskeletal magnetic resonance imaging [148].

3.4.4 Pose Estimation The goal of pose estimation is to predict the pose of a human in a given image by identifying this pose through the localization of certain human anatomical keypoints such as knee, elbow, wrist, head of the person in the image. 2D human pose estimation is a core problem in computer vision. It has many applications, including human action recognition, human–computer interaction, animation, etc. and many architectures of CNN have been developed. A recent paper [149] presents a deep high-resolution representation in single-­person pose estimation, which is the basis of other related problems, such as multi-person pose estimation. The performance and diagnostic utility of magnetic resonance imaging (MRI) in pregnancy are fundamentally ­constrained by fetal motion, which are unpredictable and rapid on the scale of conventional imaging times. In the [150], the authors propose and demonstrate methods that exploit a growing repository of MRI observations of the gravid abdomen that are acquired at low spatial resolution but relatively high temporal resolution and over long durations (10–30 min). Little is known about the characteristics of fetal motion during MRI, and here, they estimate fetal

54

pose per frame in MRI volumes of the pregnant abdomen using a 3D Convolution Neural Network that detects key fetal landmarks. Evaluation of the proposed method shows that the framework achieves quantitatively an average error of 4.47  mm and 96.4% accuracy (with error less than 10  mm). Fetal pose estimation in MRI time series yields novel means of quantifying fetal movements in health and disease and enables the learning of kinematic models that may enhance prospective mitigation of fetal motion artifacts during MRI acquisition. Other authors [151] propose a fetal pose estimation with Adaptive Variance and Conditional Generative Adversarial Network. MR acquisitions are largely limited to so-called single-shot techniques in an attempt to “freeze” fetal motion through fast imaging, while the problem due to motion occur between slices still exists. Conditional Generative Adversarial Network consists of two networks, a generator and a discriminator. The generator is responsible for estimating key point heatmaps from input MRI, and the discriminator tries to learn the features of plausible fetal pose and distinguish ground-truth heatmaps from generated ones. With this adversarial training scheme, the generator can robustly produce realistic heatmaps for fetal pose inference using adaptive variance to model the difference in intensity of motion of different keypoints. Evaluation shows that the proposed method can improve the performance of pose estimation in 3D MRI, achieving quantitatively an average error of 2.64  mm and 98.31% accuracy (with error less than 10  mm). The proposed method can process volumes with latency less than 300 ms, potentially enabling low-latency online tracking of fetal pose during MR scans. Twin-to-twin transfusion syndrome (TTTS) is a placental defect occurring in monochorionic twin pregnancies, and it is associated with high risks of fetal loss and perinatal death. Current tools and techniques face limitations in case of more complex cases of fetoscopic Elective Laser Ablation (ELA) of placental anastomoses, established as the most effective therapy for TTTS. Visualization of the entire placental surface and vascular equator maintaining an adequate distance and a close to perpendicular angle between laser fiber and placental surface are central for the effectiveness of laser ablation and procedural success. A convolutional neural network (CNN) is trained to predict the relative orientation of the placental surface from a single monocular fetoscope camera image [152] and automatic placental pose estimation in fetoscopic images. Robot-assisted technology could address these challenges, offer enhanced dexterity, and ultimately improve the safety and effectiveness of the therapeutic procedures. The authors propose a ‘minimal’ robotic TTTS approach whereby rather than deploying a massive and expensive robotic system, a compact instrument is ‘robotized’ and endowed with ‘robotic’ skills so that operators can quickly and efficiently use it. The trained network

L. E. Malgieri

shows promising results on unseen samples from synthetic, phantom, and in vivo patient data. The performance of the network for collaborative control purposes was evaluated in a virtual reality simulator in which the virtual flexible distal tip was autonomously controlled by the neural network. Improved alignment was established compared to manual operation for this setting, demonstrating the feasibility to incorporate a CNN-based estimator in a real-time shared control scheme for fetoscopic applications. Other scientific examples of Pose estimation: • a CNN-cascaded architecture [153] specifically designed for learning part relationships and spatial context, and robustly inferring pose even for the case of severe part occlusions. Code can be downloaded from [154]; • a toolbox called DeepLabCut [155] that can achieve human accuracy with only a few hundred frames of training data; • a real-time pose estimation of a free-hand ultrasound (US) image without any position sensor for diagnostics and image guidance [156]; • a learning procedure called probabilistic boosting network (PBN) for real-time object detection and pose estimation (Fig. 3.19) [157].

3.4.5 Image Registration Image registration is the alignment of two input images from different imaging modalities be to each other. In [104], the authors present a fast learning-based algorithm for deformable, pairwise, 3D medical image registration. Current registration methods optimize an objective function in dependently for each pair of images, which can be time consuming for large data. The paper presents an unsupervised learning-based approach to medical image registration that requires no supervised information such as ground-truth registration fields or anatomical landmarks, using a CNN and a spatial transform layer to reconstruct one image from another while imposing smoothness constraints on the registration field. The method promises to significantly speed up medical image analysis and processing pipelines, while facilitating novel directions in learning-based registration [158]. Deformable templates play an important role in image analysis tasks. Others [159] develop a learning framework for automatically learning such templates from data for building deformable templates, which play a fundamental role in many images analysis and computational anatomy tasks. This is particularly useful for clinical applications where a pre-existing template does not exist or creating a new one with traditional methods can be prohibitively expensive.

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

55

Fig. 3.19  Pose estimation

ht-1

x

+

x

+ tanh

tanh

x

x

tanh

Xt-1

x

+

tanh

x

ht+1

ht

x

x

tanh

tanh

Xt

x

Xt+1

Fig. 3.20  Diagram of a Recurrent Neural Network model, specifically a LSTM (Long-Short-Term-Memory)

3.4.6 Natural Language Processing A common attribute for data is to have a sequential structure. In NLP context, in a sequence of words, trough the presence of a word in position x, it is possible to predict the word in position x + 1. The same is in translation from one language to another, for question answering in specialized domains, and for capturing data sequence dependencies in scientific applications (Fig. 3.20). LSTM is a type of Recurrent Neural Network. The input xt at each timestep also informs the internal network state and consists of sigmoid and tanh functions, which transforms and recombine the input for an updated internal state and emit an output. The output of the cell considers, through the two functions sigmoid and tanh, what has happened previously in the cell. We are immersed in these algorithms: all well-known virtual assistants as well as simultaneous translators are based on LSTM algorithms.

In [160], the study describes a combined use of Machine learning and Deep Learning: Natural Language Process (LSTM) is used to extract clinically relevant information from free-text EHR notes to accurately predict a patient’s diagnosis, and when processed with logistic regression classifiers (Machine learning algorithm) to establish a diagnostic system based on anatomic divisions organ-based approach, divided into organ subsystems and/or into more specific diagnosis groups, to be able to achieve high association between predicted diagnoses and initial diagnoses determined by a human physician (Fig. 3.21).

3.4.7 Geometric Deep Learning: Ongoing and Next Steps Graph Neural Networks (GNNs) are a framework to model complex structural data containing elements (nodes) with

56

L. E. Malgieri Acute app respiratory infection Acute sinusitis

Enterovirus vesicular stomatitis with exanthem Varicella without complication

Sinusitis

Acute recurrent sinusitis

Upp er respiratory diseases Acute laryngitis

Systematic generalized diseases

Influenza Infectious mononucleosis Sapsis

Acute pharyngitis

Respiratory diseases

Acute bronchitis Bronchiolitis

Exanthema subitem

Bronchitis

Acute bronchitis due to Mycoplasma pneumonia Bronchopneumonia Bacterial pneumonia elsewhere

Tic disorder Bacterial pneumonia

Pneumonia

Mycoplasma infection

Asthma uncomplicated Cough variant asthma

Lower respiratory diseases

Neuropsychiatric diseases

Attention deficit hyperactivity disorder Encephalitis Convulsions

Asthma

Asthma with acute exacerbation Acute tracheiti

Genitourinary diseases

Diarrhea Mouth- related diseases

Gastrointestinal diseases

Fig. 3.21  Hierarchical logistic regression classifier used to establish a diagnostic system based on anatomic divisions

associated relationships (edges) between them. Not only chemical molecules have a graph structure consisting of vertices connected by edges, but also a variety of domains such as social networks, computer programs, and chemical and biological systems can be naturally represented as graphs. There are also spatio-temporal Graph Neural Networks where the graph structures evolve over time. Graph neural networks are operated by propagating data between neighboring nodes. At every propagation step, the network of nodes computes each node’s sent data; every node aggregates its received messages, and each node updates its representation by combining the aggregated incoming messages with its own previous representation [161]. Since the proposal of the Graph Neural Networks (GNNs), one of the major problems in training GNNs is the different ranges of interaction between nodes: a node’s correct prediction mostly depends on its local neighborhood (short-range information) and the contribution of distant nodes (long-range information) in the graph is very limited. Using the “umbrella” of Geometric Deep Learning, Michael M.  Bronstein, Joan Bruna, and Taco Cohen Petar Veličković [127] give an overview of deep learning methods providing a common mathematical framework to study the most successful neural network architectures, such as CNNs, RNNs, GNNs, and Transformers, and giving a constructive procedure to incorporate prior physical knowledge into neural architectures and providing principled way to build future architectures yet to be invented.

3.5 Other Examples of AI in Obstetrics 3.5.1 Cardiotocography Electronic fetal monitoring (EFM) or Cardiotocography (CTG) [162] is a technical means of continuously monitoring and recording the fetal heart rate (FHR), based on Doppler ultrasound principle, and recently divided into three classes as normal, suspicious and pathological, and uterine contraction (UC) signals, based on pressure transducers. This surveillance technique has also become a standard of care in many developed countries [163] and is used widely by obstetricians during antepartum and intrapartum periods for the biophysical assessment of fetal well-being condition [142], and it is employed to detect dangerous situations for the fetus and plays important role in reducing mortality and morbidity rate. Other fetal monitoring devices can be used: Intrauterine pressure catheter (IUPC), placed into amniotic space to measure strength of uterine contractions, Fetal scalp electrode (FSE), to monitor fetus electrocardiogram during labor, and Abdominal electrocardiogram (aECG or NIfECG). CTG and NIfECG are both noninvasive. In the clinical practice, CTG interpretation is one of the parameters used for intervention, like caesarian section, and a recent study highlights accuracy limitations of CTG and proposes more accurate noninvasive fetal ECG (NIfECG) as data acquisition methods to acquire FHR and electro hysterogram (EHG) to capture UC and propose to use ST waveform analysis to

3  Ontologies, Machine Learning and Deep Learning in Obstetrics

improve results. In this context, Artificial Intelligence can be useful [164]. According to the standards of the National Institute of Child Health and Human Development, and using 21 attributes in the measurements of FHR and uterine contractions (UCs) on CTG [165], the numerous works published on this subject for prediction of fetal risk are mainly based on machine learning algorithms defined in Sect. 3.3.1 Supervised Machine Learning algorithms, i.e., Random forest, Logistic regression, Gaussian Naïve Bayes, and Decision tree. Others, considering a baseline and the number of acceleration and deceleration patterns, and variability recommended by International Federation of Gynecology and Obstetrics (FIGO) taken into account during CTG analysis, were applied to classify FHR patterns as the input to classical Artificial Neural Network (ANN) defined in Sect. 3.4 Deep Learning and a Neural Network called Extreme Learning Machine (ELM) [166], a Neural Network does not require gradient-based backpropagation to work [167]. The results of the studies show that the accuracy of classification of some algorithms obtained is interesting and satisfactory for a study [163]; however, none of them have been universally adopted. Artificial intelligence can be used to predict through classification: the classification is, however, limited to the data sample used, and consequently, the predictions also take into account the dataset used. Some authors have suggested using Artificial intelligence also for cleaning data from UC. The dataset used in the studies did not include other clinical characteristics, such as primiparity, maternal nutritional status, anemia, gestational age, fetal wellbeing, which may affect the intrapartum course of events and could potentially contribute toward further refinement of the AI model and differences in sociodemographic characteristics of pregnant women [165].

3.5.2 Preterm Labor and Birth When formulating a problem, it is important to consider what kind of algorithm, machine learning or deep learning is most appropriate. From this point of view, the case of preterm labor and birth is very interesting: to predict perinatal outcomes in asymptomatic pregnant women with short cervical length CL, which is a common and significant risk factor for prematurity. The electrohysterography signals have been used to detect preterm births [168], because electrohysterography signals provide a strong basis for objective prediction and diagnosis of preterm birth. In this case, the authors are used machine learning algorithm for the prevision based on signals. Three different machine learning algorithms were used to identify these records. The results illustrate that the Random Forest

57

performed the best in predicting preterm labor with an accuracy of 97% [169]. Other authors are analyzed data coming from Anam Hospital in Seoul, Korea, with 596 obstetric patients during March 27, 2014–August 21, 2018 [170]. In this study, the class label (or dependent variable) was spontaneous preterm labor and birth (or preterm birth, i.e., birth between 20 and 37 weeks of gestation, coded as “no” vs. “yes”), and the following attributes (or independent variables) were included demographic factor, health-related determinants, and obstetric variables, i.e., cervical length measured between 18 and 24  weeks of gestation (cm), in  vitro fertilization (no, yes), myomas and adenomyosis (no, yes), parity, prior cone biopsy (no, yes), pelvic inflammatory disease history (no, yes), prior preterm birth (no, yes), and prior placenta previa (no, yes). Data on 596 participants were divided into training and validation sets with a 50:50 ratio; this means that the data input, 4.172, comes from the multiplication of 14 variables, 298 patients, the numbers of attributes, and observations in the training set. The ANN results put more focus on indirect determinants of preterm birth whereas their Random Forest counterparts place more emphasis on direct factors for preterm birth. On the other hand, shortened cervical length is reported to be a strong predictor of preterm birth and is consistent with the findings of the random forest in this study. An innovative research studied how to predict perinatal outcomes in asymptomatic women with short cervix length with the combination of ANN and amniotic fluid (AF) proteomics and metabolomics [171]. Of the 32 patients included in the study using combined omics, demographic and clinical data, ANN displayed good to excellent performance, with an AUC (95% CI) for delivery 50 hysteroscopies (“experts”) and 36 ≤ 50 (“novices”). 4/60 (6.6%) responding participants judged the overall impression as “7–absolutely realistic,” 40/60 (66.6%) as “6–realistic,” and 16/60 (26.6%) as “5–somewhat Realistic.” Novices rated the overall training capacity notably higher than experts (6.48 vs. 6.08), and high-grade acceptance was found in both groups. In addition, 95.2% believe that HystSim allows procedural training of diagnostic and therapeutic hysteroscopy, and 85.5% suggest that

69

HystSim training should be available for all novices before performing real surgery [6]. The same author [7] introduced a new multimetric scoring system (MMSS) for the HystSim virtual-reality (VR) hysteroscopy training simulator and evaluated learning curves for both novices and experienced surgeons. They tested it on two different typical diagnostic procedures. The 15 relevant metrics were grouped into 4 modules: visualization, ergonomics, safety, and fluid handling. At first, 24 novice medical students and 12 experienced gynecologists went through a self-paced teaching tutorial and, then, performed 5 repeated trials on HystSim. After 2 weeks, 23 of the novices

Fig. 4.2  Simulation has a large impact on the knowledge and technical skills of novices for a wide range of hysteroscopic procedures, including performance time, overall performance scores, and improvements in self-confidence, comfort, and overall competence

70

returned for a second training session. While construct validity was shown for two of four scoring modules (ergonomics and fluid handling), the experienced group did not score much higher in the visualization module than novices. Also, the safety module showed a significant but inverse difference, with novices scoring higher than the experienced group. Overall, all subjects improved notably during the training on HystSim, with clear indication that the second training session was beneficial for novices [7]. Rackow et  al. [8] in a prospective, comparative, multicenter trial reported data from a supervised, deliberate, dry lab practice in hysteroscopy for junior obstetrics-gynecology residents. They compared Objective Structured Assessment of Technical Skills (OSATS) performance of 2 groups: 19 postgraduate year (PGY)-1/PGY-2 and 18 PGY-3/PGY-4 Ob-Gyn residents. All participants completed a simulated hysteroscopic polypectomy OSATS using uterine models. They reported that PGY-1 and PGY-2 residents who had completed OSATS training performed at or above the level of untrained PGY-3 and PGY-4 residents. Junior residents had better assembly times and scores, resection scores, and global skills scores. This curriculum was effectively implemented at three institutions and generated comparable results, suggesting generalizability [8]. Janse et al. [9] in a survey assessed whether hysteroscopy training in the Dutch gynecological residency program is judged as sufficient in daily practice. This was achieved by review of the opinion on hysteroscopy training and current performance of hysteroscopic procedures by postgraduate years 5 and 6 residents in OB/GYN and gynecologists who finished residency (within 2008–2013  in the Netherlands). They received an online survey concerning performance and training of hysteroscopy, self-perceived competence, and hysteroscopic skills acquirement. Most residents felt sufficiently prepared for basic hysteroscopic procedures (86.7%) but notably less believed the same for advanced procedures (64.5%). Residents demonstrated a 10% higher appreciation of the training curriculum compared to their peers in 2003, but their selfperceived competence did not increase, except for diagnostic hysteroscopy. At last, the lack of simulation training is thought to be the most important factor during residency that could be enhanced for optimal acquirement of hysteroscopic skills [9]. Neveu et al. [10] developed a detailed curriculum including knowledge, technical skills, and simulator-based hysteroscopy training, using the Delphi method. Twenty hysteroscopy experts participated in Delphi rounds. The rounds were to be continued until 80–100% agreement was obtained for at least 60% of items. During the first round, 51 items were selected. During the second round, only 25 (49%) of them had 80–100% agreement. So, during the third round, 31 (61%) items achieved 80–100% agreement and were used to create the curriculum. Then, it was evaluated in residents. All 14 residents felt that a simulator training session was adequate and helped them to improve their skills [10].

M. Mina et al.

4.2.2.3 Hysteroscopy Training for Gynecologic Residents Elessawy et al. [11] evaluated the construct validity of the hysteroscopic simulator HystSim, which attempts to integrate hysteroscopy simulation into the training curriculum and tried to determine whether simulation training can improve the acquisition of hysteroscopic skills despite level of experience of the participants, and, finally, analyzed the learning curves of novice and expert groups. The novice group consisted of 42 medical students and residents with no prior experience, while the expert group consisted of 15 participants with ≥2 years of experience. They showed that all participants achieved significant improvements between their pre-test and post-tests, regardless of their level of experience. Regarding visualization and ergonomics, the novices showed a better pre-test value than the experts, but the experts were able to improve radically during the post-test. Overall, 66.6% evaluated the HystSim as a realistic simulator, 6.6% as “absolutely realistic,” and 26.6% as “somewhat realistic.” Generally, a high grade of acceptance was recorded: 95.2% believed that the simulator allows procedural training of diagnostic hysteroscopy, and 85.5% suggest that HystSim training should be offered to all novices before performing surgery on real patients [11]. Bassil et al. [12] suggested an innovative training program combining theoretical courses and hands-on training with bovine models and virtual simulation for diagnostic and operative hysteroscopy. The study included 25 end-year residents, and General Points Averages (GPAs) were calculated before and after the workshop. The biggest observed difference between the GPAs concerned the technical knowledge part. The GPA for operative knowledge was also higher in post-test (0.55 vs. 0.28 in pre-test). Furthermore, complication management knowledge was respectively 0.46  in pre-­test and 0.7 post-test. The overall feedback was satisfactory, and all participants applauded the realism of the simulation training [12]. Savran et al. [13] developed a standardized, simulation-­ based test of competence in hysteroscopy and established arguments of validity evidence for the test. They used the virtual-reality simulator Hyst Mentor. A total of 43 participants—15 medical students, 17 residents, and 11 experienced gynecologists—completed the test. Validity evidence was explored for all five sources of evidence which were content, response process, internal structure, relations to other variables, and consequences of testing. Inter-case reliability was high for four out of five metrics. Participants’ clinical experience was significantly correlated to their simulator test score. So, they developed this virtual-reality simulation-­based test in order to guarantee competency in a mastery learning program [13]. Munro et al. [14] conducted a multicenter prospective controlled pilot study to assess the Essentials in Minimally Invasive Gynecology (EMIG)—Fundamentals of Laparoscopic Surgery Laparoscopic Simulation System and the EMIG Hysteroscopy Simulation System for face validity

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

71

and functionality. The study enrolled 27 residents and gynecologists. First, every participant completed a screening survey and was exposed to a structured orientation to the two simulation systems. Then, they were tested with proctor supervision on the five laparoscopic and two hysteroscopic exercises. In order to determine face validation, a short 5-point questionnaire was handed to each subject. Face validity was high for all seven exercises and participants thought that the instructions were clear. The recorded exercise times and exercise errors generally reduced with increasing levels of training. Overall, the systems, including the devices and recording mechanisms, performed well, and proctor evaluation and training were satisfactory. So, they showed that the EMIG laparoscopic and hysteroscopic systems have good face validity and performed notably well, thus allowing a large-scale multicenter construct validation trial to take place [14].

4.2.2.4 Surgical Hysteroscopy Training for Fibroid Resection Munro et  al. [15] in a pilot prospective comparative study evaluated the face validity and educational utility of a virtual reality uterine resectoscopic training system (Fig. 4.3). They compared the performances of novice and expert hysteroscopists on targeting exercise and myomectomy with the virtual loop electrode. The first practiced each exercise a total of nine times with the tenth recorded as the training outcome. They compared their results both to baseline and to those of the experts. They reported that all experts achieved 4 targets in a mean of 33 s with no perforations, whereas the 11 novices achieved 2 in a mean of 57 s with one perforation. In 3 min, the experts removed a mean of 97.3% of the virtual myoma and the novices 66.1%. On the10th time, novices touched a mean of 4 targets in a mean of 23 s: an improvement from baseline and improved to 89% on the myoma resection exercise 36.3% over baseline. So, it seems that virtual reality resectoscopic + systems may improve the technical skills of novices [15]. Faurant et  al. [16] in a prospective study examined the value of hysteroscopic simulator for the resection of myoma by novice surgeons. They used the Virta Med HystSim simulator. The study enrolled 20 medical students who received a short demonstration of myoma resection, and then they were evaluated before and after a specific training in hysteroscopic resection of 60 min long. The practice required the resection of a submucous myoma type 0. The main outcome criteria were time for the resection before and after training, which were significantly reduced (170 s vs. 335 s). The second criteria were fluid quantity used, number of contact between optic and uterine cavity, and uterine perforation. The results for the latter showed reduction in fluid quantity used (335  mL vs. 717  mL) and in the number of contact between optic and uterine cavity (0.2 contact vs. 3). No perforation occurred in the simulation. Therefore, they

Fig. 4.3  Surgical hysteroscopy training for fibroid resection

demonstrated that the simulator assists and enhances hysteroscopic skills for novices [16]. Neis et al. [17] evaluated the HystSim TM simulator in a clinical training setting as a tool to assess the learning curve of both experienced and inexperienced hysteroscopic surgeons in three training rounds of both a polyp and a myoma resection. In this study, 15 inexperienced and 24 experienced surgeons participated. Primary outcomes were improvement in resection time, cumulative resection path length, and distention media use. They concluded that virtual reality ­simulator was helpful for both experienced and inexperienced surgeons and improved their skills notably. Mainly, there was improvement in resection time and required distention medium, thus increasing patient safety [17].

72

M. Mina et al.

note, Weiss et al. [22] managed to model the behavior of the human uterus under an intracavital pressure of 150 mmHg and found that the application of such an intrauterine pressure (hydrometra) corresponds to the procedure which is performed at the beginning of hysteroscopy. A three-dimensional, finite element model was constructed, while the calculated results are in agreement with other in vivo experiments [22].

Fig. 4.4 Hystero-trainer

4.2.2.5 HysteroTrainer Wallwiener et al. [18] supported that complication rates may be minimized by a well-structured training program which includes the recently developed in  vitro simulation trainer, the HysteroTrainer, based on results from a meta-analysis of the literature (n = 10,000), as well as their own experience (n = 200) [91]. The HysteroTrainer was developed to provide in vitro simulation training for diagnostic and operative hysteroscopy, including laser and high frequency electrosurgery [18] (Fig. 4.4). 4.2.2.6 Development of a Model of Hydrometra Simulation The modeling of the uterus is described by Niederer et  al. [19] as it is to be implemented in a simulator for minimally invasive gynecological procedures. The methods used to construct this model were magnetic resonance diffusion tensor imaging (DTI), the pressure–volume relationship of the cavum was determined, the behavior of the human myometrium under compression was examined with hysterectomy samples, and then the construction of the finite element model of each sample was developed. In view of mechanical accuracy of hydrometra, anatomical characteristics including the fiber architecture along with the mechanical deformation properties can be adequately simulated [19]. Harders et  al. [20] provided an extension to the statistical shape model of the uterus to include also uterine deformation states during hydrometra. So they modeled the uterus deformation process due to different pressure settings [20]. Moreover, Sierra et  al. [21] first described two different methods of hydrometra simulation: the uterine muscle distension and the liquid flow simulation in the cavity, while in parallel, this model computes the deformation of uterine shape [21]. Of

4.2.2.7 Hierarchical Task Decomposition for Hysteroscopy Tuchschmid et al. [23] presented strong methods for objective, structured, and automated assessment of surgical performance in virtual diagnostic hysteroscopy. A hierarchical task decomposition, on which surgery measurements were based, helped them to build their simulation setup. These measurements include visible surface quantification, fluid consumption, and indicators for safety and economy of movement [23].

4.3 Oocyte Retrieval 4.3.1 Oocyte Pick-Up Simulator Another study [24] aimed to assess an oocyte pick-up (OPU) simulation training program for residents using the high-­ fidelity PickUp Sim TM simulator. The oocyte retrieval rate between residents was 87%, much more than the ≥70% rate that defined as a base. A mean time of 3.4 ± 1.1 min was the time for OPU completion. The study results found that all the residents found training beneficial and the majority of them (87%) were in favor of having simulation-based training programs for OPU, in their training clinics [24]. Published evidence from a similar study from Soave et al. [25] identified a significant improvement in efficiency (mean number of follicles correctly aspired), speed (mean time to aspirate 1 follicle), and accuracy (mean percentage of follicles correctly aspirated in 1 min) among 44 clinicians (both novices and experts) [25].

4.3.2 Assessment Methods to Improve Safety During Oocyte Retrieval 4.3.2.1 Transvaginal Ultrasound Guided Oocyte Retrieval plus Doppler Transvaginal ultrasound-guided oocyte retrieval with the use of Doppler and the appropriate ultrasound-guided simulation training techniques has been proposed as a method to increase safety and avoid bleeding by Porter in 2008 [26] (Fig. 4.5). Risquez et al. in 2010 [27] evaluated the use of Doppler ultrasound during oocyte pick-up, to predict moderate peritoneal bleeding. Authors measured the fluid pockets to predict hemoperitoneum (Fig. 4.6). Pockets measured less

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

Fig. 4.5  Ultrasound guided oocyte retrieval in which the needle that has entered the follicle is seen Fig. 4.6  Moderate peritoneal bleeding after the oocyte pickup

73

than 2  cm have been considered as mild hemoperitoneum while if at least one pocket with maximum diameter 2–4 cm existed, then moderate hemoperitoneum was defined. If one pocket measured more than 5 cm, then bleeding was considered. Unfortunately, a large percentage (45%) of patients with bleeding was not predicted. The incidence of vaginal bleeding was I% with the use of Doppler ultrasound, while the expected incidence is 2.8% (Fig. 4.7). Authors raise certain questions about the validity of this method to minimize hemoperitoneum. The visualization of vascularized ovarian areas during OPU mandates the clinician who takes a decision whether to risk hemorrhage in order to harvest an extra number of oocytes. When the ovaries are located far from their normal location or are very close to the iliac vessels, a long training is required from the operator to reduce the risk of complications.

74

M. Mina et al.

In fact, an expert operator can use the Trendelenburg position from the patient to move the ovaries away from the iliac vessels (Fig.  4.8), or in case of difficult pelvic access, the oocyte retrieval can be performed transabdominally (Fig. 4.9). Sometimes transvesical egg retrieval can be performed, which is not without risk of complications (Fig. 4.10); therefore, it requires the operator’s experience, while at other times, the bladder can be accidentally involved in complications during the pick-up (Figs. 4.11, 4.12, and 4.13, Videos 4.1 and 4.2).

Fig. 4.7  The incidence of vaginal bleeding was I% with the use of Doppler ultrasound, while the expected incidence is 2.8%

Fig. 4.8  When the ovaries are located far from their normal location or very close to the iliac vessels, extensive training is required from the operator to reduce the risk of complications. In fact, an expert operator can use the Trendelenburg position from the patient to move the ovaries away from the iliac vessels

4.3.2.2 SIPS Technique Shah and Walmer in 2010 [28] tried to evaluate the pelvic anatomy with a technique called saline intraperitoneal sonogram. They included ten women with unexplained infertility and normal hysterosalpingogram while five of them had

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

75

Fig. 4.11  Injury of the bladder during the ovarian pick-up

Fig. 4.9  The expert operator in case of difficult pelvic access, the oocyte retrieval can be performed transabdominally

Fig. 4.10  Hematoma of the superior bladder wall after the oocyte pick-up

known adhesive disease. The technique involved two steps. First of all, they performed a sonohysterogram with saline and observed for a pocket of fluid in the peritoneal cavity. Second, they directed at 17-g oocyte retrieval needle into the pocket of peritoneal fluid and infused 600  mL of normal saline. At this step, they evaluated with 3D and 4D ultrasound the pelvic anatomy. More saline was infused (up to 1500 mL) in case tubes and ovaries did not float easily. At the end of the procedure, saline was aspirated from the peritoneal cavity, and antibiotics have been administered. Then they evaluated the images and considered as normal if the uterus, fallopian tubes, and ovaries were completely surrounded from saline and no abnormal pathology was observed . In case, the structures were not completely surrounded by saline, then the images were considered non­diagnostic. From the other side, if these images have been suspicious for pelvic adhesions or hydrosalpinges, then they were considered abnormal. Diagnostic laparoscopy was offered for patients that included in the last two groups. The average time for this procedure was 45 min. Eventually, with this technique, all patients finished the procedure. Eight of them requested sedation and pain management, and one developed signs of abdominal infection. This patient was admitted and treated with IV antibiotics. For patients with known risk factors for adhesions, SIPS technique was very accurate finding pelvic adhesions, evidence of prior peritonitis (Fitz-Hugh-Curtis adhesions), and peritoneal adhesions involving 75% the peritoneal cavity after myomectomy. Of the five subjects without risk factors for intraperitoneal disease, four had studies that were judged to be normal and one had findings with unilateral hydrosalpinx and bilateral pelvic adhesions not seen in a previous hysterosalpingogram. This technique is considered feasible and

76

M. Mina et al.

4

3 ASPIRATION

2

1

mmHg

5

6

1. Vacuum Pump 2. Disposable Vacuum Line with hydrophobic filter 3. Aspiration line 4. Mono lumen Needle 5. test tube

Fig. 4.12  Video tutorial about pick-up technique during oocyte retrieval

Fig. 4.13  Video tutorial about complications during oocyte retrieval

cost effective for patients that would like to avoid laparoscopy while in parallel evaluating the intraperitoneal cavity for pathology that affects fertility treatment.

4.4 Embryo-Transfer Simulation Embryo transfer (ET) remains the critical step in assisted reproduction techniques. With regard to the fact that the proper ET technique is clearly an operator-dependent vari-

able and as such it should be objectively standardized, evidence from a retrospective comparative analysis found that the operators’ performance did not improve with additional transfers. However, this was a single-center study mainly focused on experienced performers [29]. Another ­retrospective study proved that ET simulation improved fellow’s pregnancy rates in their first transfers and led to a more rapid ET proficiency. Half fellows analyzed before ET trainer and half after ET trainer. Pregnancy rates were similar in the 2 groups after 20 ETs, and collective terminal pregnancy rates were >50% after 40 ETs [30]. Considering the live birth rates, McQueen et al. [31] concluded that under the appropriate supervision, there is no difference between ETs performed by fellows and experienced performers. Authors alarming the small number of ETs performed by fellows and ask for a minimum number [31]. Of note, Kresowik et al. [32] compared the clinical pregnancy and live birth rates following ET performed by REI fellows after a prolonged lapse (18 m) and found no significant differences. Clinical pregnancy and live birth were similar between the two time periods both for individual fellows and for the overall group [32]. Intrauterine insemination (IUI) may serve as a treatment modality before starting ETs by REI fellows (Fig.  4.14).

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

Fig. 4.14  Intrauterine insemination (IUI) may serve as a treatment modality before starting ETs by REI fellows

The clinical pregnancy rates for the first 100 ETs performed were unchanged after implementing an IUI training requirement, according to the retrospective cohort study of Shah et al. Authors pointes that ET training is mandatory [33]. Moreover, Ceccaldi et al. [34] pointed out the contribution of 3D printing and digital simulators, will facilitate teamwork in assisted reproductive technology and enable an enhancement of knowledge within the specialty. The increase of educational possibilities due to the easy access to the digital stimulator and the 3D printing, lead forwards the digitalization of care and the dematerialization of the patient and his care [34].

4.5 Ovarian Hyperstimulation (OHSS) Triage and Complications Avoidance 4.5.1 Evaluating High-Risk Patients Lainas et al. [35] conducted a retrospective cohort study of 319 women undergoing IVF and were at high risk for OHSS following administration of hCG for the triggering of final

77

oocyte maturation. High risk for OHSS was defined as the presence of at least 19 follicles ≥11 mm on the day of triggering of final oocyte maturation. Their aim was to evaluate ultrasound and hematological changes during the early luteal phase (until day 5 post-oocyte retrieval). Severe OHSS was diagnosed when at least moderate ascites was present and ≥2 of the following: maximum ovarian diameter (MOD) >100 mm, hematocrit (Ht) >45%, white blood cell count (WBC) >15,000/mm3, hydrothorax, dyspnea, and oliguria (Fig.  4.15). All of them, except for MOD, increased notably during the early luteal phase, whether or not they developed OHSS.  MOD was the only variable which increased between day 3 and 5, only in patients who developed severe early OHSS, but not in those who did not. They also reported that Ht is directly associated with the quantity of ascetic fluid accumulation; thus, it may be used as a marker of hemoconcentration, and probably as a new additional criterion in a modern OHSS classification system [35]. Soon after (2020), they assessed 321 women at high risk for severe OHSS. They examined the same parameters and used cut-offs or ascites grade >2, Ht >39.2%, WBC >12,900/mm3 and MOD >85 mm on day 3 post-oocyte retrieval (Fig. 4.16). Their aim was to predict subsequent severe OHSS development on Day 5. They reported that the probability of severe OHSS with no criteria fulfilled on Day 3 was 0%, with 1 criterion 0.8%, with 2 criteria 13.3%, with 3 criteria 37.2%, and with 4 criteria 88.9%. The predictive model of severe OHSS had an area under the curve (AUC) of 0.93 with a sensitivity of 88.5% and a specificity of 84.2%.The algorithm may be useful to warn doctors when severe OHSS is likely to happen. Based on this, they could advise women on their treatment and manage them properly at an early stage [36].

4.5.2 Ovarian Hyperstimulation and Pregnancy Outcomes Lainas et al. [37] in a prospective cohort study compared live birth rates between high-risk patients who develop severe early OHSS and received low-dose GnRH antagonist in the luteal phase and in high-risk patients who do not develop severe early OHSS and do not receive GnRH antagonist in the luteal phase. The study included 192 IVF patients; those who were diagnosed with severe early OHSS on day 5 post-­ oocyte retrieval and were administered 0.25 mg of ganirelix for 3 days, from day 5 till day 7 (OHSS + antagonist group, n = 22) and high-risk patients who did not develop the severe early OHSS and did not receive the same intervention (control group, n = 172). All patients underwent embryo transfer on day 5. They reported that live birth rates (40.9% vs. 43.6%), ongoing pregnancy rates (45.5% vs. 48.8%), clinical

78

M. Mina et al.

Fig. 4.15  Severe OHSS was diagnosed when at least moderate ascites was present and  ≥2 of the following: maximum ovarian diameter (MOD) >100 mm, hematocrit (Ht) >45%, white blood cell count (WBC) >15,000/mm3, hydrothorax, dyspnea, and oliguria

pregnancy rates (50% vs. 65.1%), positive hCG (72.7% vs. 75%), duration of gestation (36.86  ±  0.90  weeks vs. 36.88 ± 2.38 weeks), and neonatal weight (2392.73 ± 427.04 vs. 2646.56 ± 655.74 g) were all similar in the OHSS + antagonist and control group, respectively. All infants born were healthy. This suggests that low-dose luteal GnRH antagonist administration during the preimplantation period may be

safe. So, they concluded that low-dose luteal GnRH antagonist administration is associated with a favorable IVF outcome, comparable to control high-risk patients without severe OHSS and not receiving the intervention, but this protocol should be used with caution until larger studies needed with children follow-up [37].

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

79

Fig. 4.16  Ascites are the first sign of OHSS and when severe it requires paracentesis in some cases

4.6 Ectopic Pregnancy Triage 4.6.1 Ectopic Pregnancy Risk and Triage and Complications Avoidance Jin et al. [38] in a retrospective study examined the possible factors affecting the incidence of ectopic pregnancy (EP) in IVF/ICSI.  They found that the percentage of patients with secondary infertility or tubal infertility (Fig.  4.17) and the percentage of cycles involving transfer with a cleavage embryo were significantly higher in EP group than that in the non-EP group. There was, also, significant difference in endometrial combined thickness (ECT) and previous history of EP that affect EP, but not in type of transfer (fresh/frozen-­ thawed) or number of embryo transferred [38].

Fig. 4.17  The ectopic tubal pregnancy in patient with tubal infertility

intrauterine pregnancies (IUPs): age, prior history of EP, prior history of miscarriage, bleeding, and hCG concentration at presentation. They categorized women into three groups based on acuity of outpatient surveillance: low acuity for low risk (−2 to −1), standard for intermediate risk (0 to +4), and high-acuity surveillance for those with high risk (≥+5). Therefore, a scoring system based on symptoms at presentation is important to assess risk and define the intensity of outpatient surveillance but does not serve as a diagnostic tool. The benefit of it is to triage and organize the follow-up evaluations, thus, potentially decrease the number of visits and time required for diagnosis [39].

4.6.3 Triage Protocols for Ectopic Pregnancy 4.6.2 Clinical Risk Scoring System In a multicenter retrospective study, Barnhart et al. [39] evaluated a scoring system to triage women with a pregnancy of unknown location (PUL). This scoring system was based on five factors obtained at initial evaluation of a woman with a PUL, distinguishing EPs and miscarriage from ongoing

4.6.3.1 Serum hCG at 48 H Krause et al. [40] in a retrospective study tried to determine if the Bayes classifier can be used to distinguish between an EP and IUP after ET based on early hCG levels (measured at 12 and 20 days after OPU). They analyzed singleton intrauterine (n = 91) and ectopic gestations (n = 14). Using the Bayes classifier, an hCG value 75%) of EP. There was no statistically significant difference regarding endometrial thickness (p = 0.77), fresh or frozen embryo transfer (p  =  0.53), number of embryos transferred (p = 0.13), donor or autologous oocytes (p = 0.76), or the day of hCG draw (p = 0.13 and 0.43 for first and second measurement). So, the Bayes classifier may be a useful tool to forewarn of a possible EP [40].

M. Mina et al.

a failed PUL, IUP, or EP. The protocol classifies women with PUL as low risk if the estimated risk of EP is 5%, otherwise as high risk. They found that the progesterone protocol based on levels of 10 nmol/L classified 24% of failed PUL, 95% of IUP, and 76% of EP as high risk. The M4 protocol classified 14 of failed PUL, 37% of IUP, and 84% of EP as high risk. The hCG ratio cut-offs classified 10% of failed PUL, 15% of IUP, and 63% of EP as high risk. Overall, the M4 model 4.6.3.2 Two-Step Triage Protocol performs better than both other protocols as it misclassifies Van Calster et al. [41] in an observational cohort study devel- significantly fewer EPs, while correctly categorizing the oped a two-step triage protocol based on presenting serum majority of IUP and failed PUL to the low-risk group [43]. progesterone (step 1) and hCG ratio 2 days later (step 2) to Later on, Fistouris et al. [44] evaluated four different protoselect possible EPs. They evaluated 2753 PUL (301 EP), cols. In “Protocol A,” a PUL was classified as low risk if involving a secondary analysis of prospectively and consecu- >13% hCG level decline or  >66% hCG level rise was tively collected PUL.  They divided chronologically 1449 achieved; otherwise, the PUL was classified as high risk. PUL for development and 1304 of them for validation. In “Protocol B” classified a PUL as low or high risk using cut-­ step 1, low-risk PUL were selected at presentation using a offs of 35–50% hCG level decline and of 53% hCG level serum progesterone threshold. The remaining PUL were cat- rise. “Protocol C” used hCG level cut-offs published by egorized using a novel logistic regression risk model based NICE, 50% for declining and 63% for rising hCG levels. on hCG ratio and initial serum progesterone (step 2), defin- Finally, in “Protocol M4,” if the calculated risk of being an ing low-risk as an estimated EP risk 2 nmol/L or a published by NICE, and M4 model correctly classify more measurement had not been taken, hCG level was measured PULs as high risk, compared to two other protocols [44]. again at 48  h and results were entered into the M6 model. Bobdiwala et al. [45] in a prospective multicenter cohort They were then classified as low or high risk. Women of high study tried to define the adverse outcomes associated with risk were reviewed clinically within 48 h. In step 1, 15.5% of the M4 model in women with PUL.  They examined 835 patients were classified as low risk. In step 2, 1038 women women in total. The M4 model categorized 70% of PUL as with a PUL were classified as low risk and 901 as high risk. low risk. Adverse events were reported in 26 PUL and 1 seriOverall, 85.9% EPs were correctly classified as high risk and ous adverse event. A total of 17/26 (65%) adverse events 1445/2625 PUL (55.0%) were classified as low risk, of were misclassifications of a high-risk PUL as low risk by the which 15 (1.0%) were EP. They demonstrated that the two-­ M4 model. The other 5/26 (19%) adverse events were related step protocol incorporating the M6 model is an efficient and to incorrect clinical decisions, as well as the serious adverse safe way of categorizing women with a PUL [42]. event. The remaining 4/26 adverse events were secondary to unscheduled admissions for pain/bleeding. Therefore, they 4.6.3.3 M4 Decision Support System concluded that expectant management of PUL using the M4 Guha et al. [43] in a retrospective diagnostic accuracy study, prediction model is safe due to the low number of adverse compared the performance of serum progesterone-based tri- events (2.0%) [45]. age with the M4-based triage protocol using both the initial hCG level and serial hCG measurements, in women with a 4.6.3.4 M6 Decision Support System PUL. M4 is a multinomial logistic regression model based Bobdiwala et al. [42] in a prospective multicenter intervenon the initial serum hCG level and the hCG ratio (48 h/0 h). tional study evaluated the triage performance of the two-step Entering these values into the M4 formula (see Supplementary protocol using an initial progesterone level of ≤2 nmol/L, to Videos 4.1 and 4.2) returns estimated risks of the PUL being identify probable failing pregnancies of unknown location

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

(Step 1), followed by the M6 model (Step 2). They examined 3272 women with a PUL.  If the initial progesterone level was ≤2 nmol/L, patients were discharged. If the progesterone level was >2  nmol/L or a measurement had not been taken, hCG levels were measured again at 48 h, and results were entered into the M6 model. Then, they were classified as low or high risk. Women classified as high risk were reviewed clinically (with hCG plus ultrasound) within 48 h. In Step 1, 15.5% of patients were classified as low risk. In Step 2, 1038 women with a PUL were classified as low and 901 as high risk. Overall, 85.9% EPs were correctly classified as high risk and 55.0% were classified as low risk, of which 1% were EP. They demonstrated that the two-step protocol incorporating the M6 model is an efficient and safe way of categorizing women with a PUL [42].

4.6.4 Ectopic Pregnancy and Biological Factors 4.6.4.1 Ectopic Pregnancy and Ovarian Reserve Lin et al. [46] in a retrospective study of 2.061 women compared the incidence of EP in women with normal (NOR) and decreased ovarian reserve (DOR). DOR is defined as FSH >10 IU/L and NOR as FSH ≤10 IU/L. They reported that the clinical pregnancy rate was significantly higher in patients with NOR than in those with DOR (47.52% vs. 39.10%). Also, the incidence of EP in clinical pregnancies was significantly higher in the DOR than in the NOR group (5.51% vs. 2.99%) [46]. Kim et  al. [47] in an observational study assessed the association between ovarian reserve and the incidence of ectopic pregnancy (EP) following IVF/ET.  Women with AMH 10 mIU/mL were classified into the decreased ovarian reserve (DOR) group, and the remaining patients were classified into the normal ovarian reserve (NOR) group. Although there is no significant difference for intrauterine and chemical pregnancy between the two groups, they reported that EP occurred more often in the DOR than in the NOR group (10.7% vs. 2.5%). Thus, more intense monitoring may be necessary for pregnant women with DOR [47]. 4.6.4.2 Ectopic Pregnancy and Frozen Embryo Transfers (FET) Decleer et  al. [48] in a retrospective study, compared the incidence of EP in fresh and frozen/thawed cycles of IVF. The Belgian legislation was followed for the number of transferred embryos. They reported that the incidence of EPs per established clinical pregnancy was 1.92% for the fresh vs. 1.28% for the frozen/thawed cycles. So, they concluded that there was no significant difference between the two, in a large cohort of patients [7]. In another retrospective study,

81

Huang et  al. used data from 31,925 women undergoing IVF-ET from January 2006 to December 2013. They reported that the incidence of EP per clinical pregnancy was 4.62% for the fresh transfer group vs. 2.22% for the frozen-thawed cycle group. Also, day 3 embryo FET cycles had a higher risk of EP than blastocyst FET cycles. The differences were all statistically significant [48].

4.6.4.3 Ectopic Pregnancy and BMI Cai et al. [49] in a retrospective cohort study examined the association between body mass index (BMI) and ectopic pregnancy following fresh or frozen ET.  They analyzed 16.378 pregnancies using the generalized estimating equation (GEE). According to the WHO criteria, the number of cycles with low ( 15 mm. There were significant differences in ectopic pregnancy rates (10.0%, 4.3%, 2.1%, and 2.2%, respectively). Logistic regression analyses indicated EMT on day of hCG administration, as one of the independent predictive variables for EP [51], negatively correlated with EP rate. Also, progesterone levels on the same day and number of embryo transferred have been positively correlated by the same analysis [51]. Ding et  al. [52] in a prospective study investigated the correlation between endometrial thickness and clinical pregnancy outcomes in frozen-thawed embryo-transfer cycles. They analyzed 1.475  cycles. They measured endometrial thickness on the ovulation day (in natural cycles) or the point of endometrial transformation (on hormone replacement cycles). Patients with thin endometrium have been considered when endometrial thickness was ≤6  mm. The overall clinical pregnancy rate, embryo implantation rate, abortion rate, multiple birth rate, and live birth rate were 54.4%, 35.7%, 23.3%, 24.1%, and 43.9%, respectively. The ectopic pregnancy rate in the study was 0.6% [52]. In a retrospective 1:4 matched case–control study, Liu et al. [53] evaluated 225 EP patients undergoing IVF-ET, to identify the risk factors that are associated with EP, using conditional logistic regression for the analysis. They found that, because endometrial receptivity plays an important role in embryo implantation, a thinner endometrium indicates poorer endometrial receptivity. Also, the increased risk of EP could be due to uterine peristalsis wave frequency, which may dislodge the uterus. So, compared with an endometrial thickness 12  mm before embryo transfer was a strong protective factor against EP (OR 0.27) [53], along with the number of transferable embryos [53]. In another retrospective cohort study, Liu et al. [54] examined if endometrial thickness (EMT) influences the incidence of EP in frozen embryo-transfer (FET) cycles. They used 17,244 FET cycles that resulted in pregnancy. Compared to women with an EMT of ≥14 mm, the ORs for women with EMT of 7–7.9, 8–9.9, and 10–11.9 mm were 2.70, 2.06, and 1.66, respectively. Furthermore, they noticed that hormonal replacement FET was associated with higher EP risk com-

M. Mina et al.

pared to modified natural or stimulated cycles. So, they concluded that EMT is inversely proportional to EP rate in FET cycles. EMT seems to be a potential quantitative marker of endometrial receptivity and uterine contractibility in an FET cycle [54].

4.6.4.5 Ectopic Pregnancy and Endometrial-­ Embryonal Synchronization Murtinger et al. [55] conducted a retrospective single-center study of 12,429 blastocyst transfers (8182 fresh and 4247 frozen ETs) in order to determine the main risk factors responsible for EP.  They noticed that neither female age (36.7 vs. 35.8  years), BMI, quality of transfer, nor stimulation protocol affected the EP rate. Conversely, history of previous EP (OR 3.26), tubal surgery, or both (OR 6.20) did. Furthermore, the incidence of EP was higher in women with uterine malformations (OR 3.85) or pathologies (OR 5.35), uterine surgeries (OR 2.29), or sub-optimal endometrial build-up (OR 4.46–5.31). The risk of EP was notably increased when slow-developing blastocysts were transferred (expressed by expansion) (OR 2.59), thus, indicating the importance of proper embryonalmaternal synchronization. The overall EP rate after blastocyst transfer was low, comparable with reported EP rates in spontaneous conceptions [55].

4.6.5 Ectopic Pregnancy Triage After a Previous Operation 4.6.5.1 Prediction Rule for Ectopic Pregnancy After Laparoscopic Salpingostomy In a retrospective cohort study, Morse et  al. [56] tried to identify a persistent ectopic pregnancy after linear salpingostomy using hCG serum concentration. They assessed 854 laparoscopic salpingostomies. The most clinically useful prediction rule was calculated by dividing the difference between the first and second post-operative hCG levels by the first post-operative HCG level (i.e., [HCG1-HCG2]/ HCG1). When this ratio surpassed 0.75, it reliably ruled out persistent EP with a NPV of 99%. When this ratio was 24 h, pain during coughing, brown disdivided into 2 groups: 344 patients in derivation group and charge, and unilateral pelvic pain. Two-thirds of patients were 172 in validation group. Vomiting, sudden onset of pain, and included in the derivation group for the SAQ-GE EP score

84

which was built on multiple logistic regressions. One-third of them were used for internal validation. The SAQ-GE EP score was based on these 5 variables (score 0–100). The lowrisk group of EP (score 70) had a specificity of 97.4 and a LR+ of 12.3. So, they concluded the SAQ-GE EP score might prove helpful [59].

M. Mina et al.

transfers of frozen cleavage embryo (1.7%), transfers of fresh blastocyst (1.3%), and transfers of frozen blastocyst (0.8%). The likelihood of EP was 30% higher for fresh cleavage stage ETs when compared to fresh blastocyst transfer. So, they concluded that the lowest risk of EP was associated with the transfer of blastocyst, frozen, and single embryo compared to cleavage stage ET, fresh, and multiple embryos [62]. In a retrospective cohort study, Acharya et  al. [63] 4.6.6.2 Combination of All Tools assessed the impact of controlled ovarian stimulation on EP In a prospective observational study, Varas et al. [60] tried to rate as a function of the number of oocytes retrieved, using identify the best combination of non-invasive diagnostic donor IVF cycles as a control. They included 91,504 autolotools for the diagnosis and treatment of potentially life-­ gous cycles and 17,636 donor cycles in patients with non-­ threatening gynecological emergencies (G-PLEs). They ana- tubal infertility. They concluded that, in autologous IVF lyzed 5 diagnostic tools by logistic regression: (1) the Triage cycles, increasing oocyte yield is associated with a signifiProcess model included systolic blood pressure 7, (2) the history-taking model recipients, indicating that the increased EP rate may be due included a history of ectopic pregnancy, shoulder pain, to the supra-physiologic hormone levels achieved with conunbearable pain, pain during movement, pain on abdominal trolled ovarian hyperstimulation [63]. palpation, and vomiting during pain, (3) the physical examination model included abdominal guarding and rebound tenderness, (4) the ultrasonography model comprised fluid in 4.7 Fertility Preservation the Morison pouch, pelvic fluid reaching the uterine corpus, abnormal adnexal mass, and ovarian cyst larger than 50 mm, Noyes et al. [64] reviewed reproductive principles such as ovarand (5) the biological exams model was composed of the ian reserve, uterine function, cervical competence, and early urine hCG test or serum hCG assay among patients without obstetrical management, as well as available fertility preservaa known pregnancy and a leukocyte count >10  G/L.  They tion (FP) methods for gynecologic cancer patients. The impleconcluded that the use of a standardized self-assessment mentation of protocols for the early diagnosis and treatment of questionnaire for history taking and ultrasound examination gynecologic cancer has resulted in improved prognosis. Hence, was the most successful combination for the diagnosis of FP has become a primary need in cancer management. They G-PLEs. Other tools, such as physical examination, do not noted that there are reliable ovarian reserve measures and cliniadd any substantial diagnostic value [60]. cians should assess potential candidates for FP according to them. Particularly, oocytes cryopreservation (OC) has improved the success rates associated with the techniques available. 4.6.7 How to Lower the Risk of Ectopic Currently, OC provides reproductive autonomy and appears to Pregnancy in IVF be the preferred method for single women with a gynecologic malignancy. Another alternative is embryo cryopreservation In a retrospective study, Cheng et al. [61] reviewed women using donor gamete. Therefore, as the field of oncofertility who achieved a clinical pregnancy after IVF-ET, during a expands, FP consultation should be of great importance [64]. 15-year period. They found that the incidence of EP after In another systematic literature review, Srikanthan et al. fresh ET was 1.5%, similar to that of natural conception. In [65] proffered a handful resource for oncology providers addition, EP rates were similar for Day 3 (1.2%) and Day 5 with practical information about FP.  All cancer patients (1.7%) ETs. Furthermore, the incidence of EP in thawed should be informed of potential fertility risk and could be compared to fresh ET cycles was not notably reduced (0.6% referred to a FP clinic for further information on the availvs. 1.5%). Finally, the strategy used for ET may affect the able interventions. Fertility sparing surgery should be offered incidence of EP; tubal ET (TET) and ET under full bladder to patients needed. Also, oocyte and embryo cryopreservadistention had a significant effect on EP, but this needs fur- tion should be of standard practice. Furthermore, oncologic ther evaluation [61]. treatment delays are minimized by independent-of-­ Li et al. [62] attempted to define which type of transferred menstrual-­cycle-day start protocols, but these protocols for embryo is associated with a lower rate of ectopic pregnancy. ovarian stimulation are different and center specific. They reported an overall rate of EP in women following ART Stimulation protocols are unlikely to increase the risk of cantreatment of 1.4%. Pregnancies after single embryo transfer cer recurrence in women with hormone-sensitive cancers. At presented with 1.2% EPs, significantly lower than double ETs last, GnRH analogs should not be used in all cancer patients (1.8%). Furthermore, the highest EP rate was 1.9% for preg- but may be considered in high-risk cases where definitive FP nancies from transfers of fresh cleavage embryo, followed by is not possible or feasible [65].

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

4.8 Ovarian Complication 4.8.1 Ovarian Torsion Moore et al. [66] reviewed cases of surgically proven ovarian torsion at their institution during a 20-year period, assessing retrospectively CT findings. CT was obtained in 33% of the 167 patients. All reports described an enlarged ovary, ovarian cyst, or adnexal mass of the involved ovary. So, they claimed that ovarian torsion is ruled out by a CT scan when the ovaries are well visualized and appear normal [66]. Huchon et  al. [67] in a retrospective study, developed a score, based on multiple logistic regression after a jackknife procedure, for assisting in the pre-operative diagnosis of adnexal torsion (AT) in women with acute pelvic pain. Then, they validated the score in a prospective cohort of 35 women with acute pelvic pain. They evaluated 5 criteria: unilateral lumbar or abdominal pain, pain duration, 8 h at first presentation, vomiting, absence of leucorrhoea and metrorrhagia, and ovarian cyst larger than 5 cm by U/S. Low-risk and high-­ risk groups were a result from values of the score (probability of AT 3.7% and 69%, respectively). The diagnostic accuracy of the score indicated adnexal torsion probability of 0% and 75% in the low-risk and high-risk groups, respectively [67]. Bannas et al. [68] in a review signified that MR imaging provides excellent image quality of the abdominal cavity in the setting of the nontraumatic acute abdomen. It, also, contributes in the diagnosis of a broad spectrum of diseases in the setting of the nontraumatic abdomen, comparable in many ways to CT. MR is of great worth when examinig pregnant women or pediatric population. It can detect diseases, such as acute apendicitis, cholocystitis, pancreatitis, bowel obstruction, ovarian torsion, pelvic inflamatory disease, urolithiasis, and many more. So, MR imaging provides for detailed evaluation in cases of nontraumatic acute abdomen that is comparable to CT scan [68]. Melcer et al. [69] in a retrospective study investigated the correlation between clinical, laboratory, and ultrasound findings with the surgical diagnosis of AT.  The study included 199 women of reproductive age. Adnexal torsion was surgically diagnosed in 55.8% of the cases. The associated various parameters with AT using a multivariate logistic regression analysis and odds ratios (OR) ± 95% confidence intervals (CI). So, complaints of nausea/vomiting (OR 4.5), peritoneal irritation signs (OR 100.9), elevated WBC >11,000 cells/mL (OR 3.7), the presence of free pelvic fluid on ultrasound (OR 34.4), ultrasound findings suggestive of ovarian edema (OR 4.2), ultrasound findings suggestive of benign cystic teratoma (OR 7.8), and location of the ultrasound pathology on the right side (OR 4.7) were positively associated with adnexal torsion. Conversely, ultrasound findings suggestive of hemorrhagic corpus luteum cyst (OR

85

0.04) were negatively associated with adnexal torsion. By combining these eight parameters, the ROC curve was calculated, yielding an area under the curve of 0.93 [69].

4.8.2 Gynecological Ultrasound in the Context of Ovarian Torsion Glanc et al. [70] reviewed the importance of sonography in the evaluation of acute abdomen during pregnancy (Figs. 4.20 and 4.21). They demonstrated that sonography is the first-­ line screening test in pregnant patients presenting with acute abdomen. If it is inconclusive or inadequate, further imaging might be required, such as MRI [70].

Fig. 4.20  Ultrasonographic image of dermoid cyst torsion

Fig. 4.21  Left tubo-ovarian abscess

86

a

Fig. 4.22 (a) The International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) has assembled the following expert-opinion-­ based guidance for the rationalization of ultrasound investigations for

M. Mina et al.

b

gynecological indications, during the current coronavirus (SARS-­ CoV-­2) pandemic. (b) Rupture of ovarian endometriosic cyst

4.8.3 Ovarian Abscess Ultrasound Diagnosis

ovarian torsion or ruptured hemorrhagic ovarian cyst (Fig. 4.22b), OHSS, abnormal uterine bleeding with anemia, Peces Rama et al. [71] in a prospective study evaluated the pelvic inflammatory diseace, and/or tubo-ovarian abcess. (2) IOTA three-step strategy and compared it with subjective Ultrasound assessments that can be delayed for 2–4 weeks assessment by an expert. Non-expert examiners examined 81 (SOON): abnormal uterine bleeding (postmenopausal or patients using the simple descriptors (SD) and simple rules postcoital), abdominopelvic “mass” or signs of malignancy (SR) and then an expert assessed the inconclusive cases (IOTA ≥10), signs of recurrent gynecological malignancy, or (SA). Surgery was performed for 30 (8 malignant and 22 family history of this type of malignancy. (3) Ultrasound benign) of the 81 masses, while 51 of them were considered assessments that can be delayed for theduration of the panas benign and managed expectantly. The inexperienced demic (LATER): chronic pelvic pain, cyclical dyschezia, examiners were able to classify 69 (85.2%) patients using dysmenorrhea, dyspareunia, infertility and recurrent pregSD and SR, while 12 (14.8%) of them were referred to an nancy loss, and review of previously noted pelvic pathology expert. The overall sensitivity and specificity of the three-­ (leiomyoma, endometriosis, and adenomyosis) [72]. step strategy were 87.5% and 100%, respectively. Conversely, when using only subjective assessment, the sensitivity was 87.5%, but specificity was 98.6%. Therefore, the three-step 4.9 Thrombosis strategy seems to be a useful method in triaging adnexal 4.9.1 Acuity Scale: VELTAS masses [71]. Parsi et al. [73] instituted the Venous and Lymphatic Triage and Acuity Scale (VELTAS), in order to provide a universal 4.8.4 Ultrasound Triage at the COVID-19 Era triage of patients with venous and lymphatic disorders or The International Society of Ultrasoundin Obstetrics and vascular anomalies and urgency. VELTAS aims to improve Gynecology (ISUOG) has assembled the following expert-­ patient safety and increase triage reliability by providing a opinion-­based guidance for the rationalization of ultrasound standardized framework for the management of several coninvestigations for gynecological indications, during the cur- ditions categorized as follows: (a) venous thromboemborent coronavirus (SARS-CoV-2) pandemic. They suggest lism, (b) chronic venous disease, (c) vascular anomalies, (d) that ultrasound appointments for gynecological indications venous trauma, (e) venous compression, and (f) lymphatic should be triaged based on the clinicalscenario, as follows: disease. Triage urgency is classified as follows: (1) medical (1) Ultrasound assessments that should be performed with- emergencies (requiring immediate attendance), (2) urgent (to out delay (NOW): acute pelvic pain (Fig.  4.22a) such as be seen as soon as possible), (3) semi-urgent (to be attended

4  Assisted Reproductive Technologies: Complications, Skill, Triage, and Simulation

to within 30–90 days), and (4) discretionary/non-urgent (to be seen within 6–12  months). This scale not only may be used during epidemics such as the current COVID-19 pandemic, but can also be used as a general framework to classify urgency of the above-mentioned conditions [64]. However, the same team concluded that VELTAS does not substitute clinical judgment and individual care. Particularly, some conditions can be managed differently, and a multidisciplinary approach may be followed or can be delayed, if chronic, for definitive procedural management and in some cases can be treated conservatively [73]. Ohta et  al. [74] aimed to develop a diagnostic scale (GAI2AA—including three factors) to predict the occlusion and, thus, effectively administer immediate mechanical thrombectomy. He included 429 patients with acute ischemic stroke, developed the scale and then applied this scale to 259 patients. Gaze palsy, aphasia, and inattention are hemispheric symptoms and complement each other. They take a score of 2, while arm paresis and atrial fibrillation take a score of 1, respectively. This study concluded that with the appropriate cutoff (≥3), the sensitivity was 88% and the specificity was 81% [74]. Bernardini et al. [75] described several clinical scales for the prediction of large vessel occlusion and they concluded that no scale is perfect so far. They included the GAI2AA scale score and compared to those of the PASS, CPSSS, 3I-SS, FAST-ED, and RACE scales. Cutoff values of the GAI2AA scale ≥3 were associated with false-negative and false-positive rates of 5% and 21%, respectively. In contrast, CT angiography has 100% sensitivity and specificity to detect large vessel occlusion in a timely fashion and was used in all major stroke trials. To date, the use of any stroke scale with lower sensitivity and specificity than CT angiography as its replacement is of uncertain added value. Resources (interventionalist, angiography suite, and staffing) are less cost effective and should be used appropriately [75]. Lemma et al. [76] in a single retrospective study assessed possible modifiable factors for acute mesenteric ischemia, with special focus on the pathways to treatment. The study found that the most important prognostic factor for patients with acute mesenteric ischemia was the type of emergency room at which the patients were initially examined [76].

4.9.2 Pulmonary Embolism Triage Zaleski et  al. [77] performed an analysis on the ability of Wells score in detecting deep vein thrombosis (DVT) or pulmonary embolism (PE) in athletes (Fig.  4.23). The Wells score for DVT includes the factors: active cancer (treatment ongoing, within the last 6 months, or receiving current palliative treatment), paralysis, paresis, or recent cast immobili-

87

zation of the lower extremities, recently bedridden for ≥3 days or major surgery within 12 week requiring general or regional anesthesia, localized tenderness along the distribution of the deep venous system, entire leg swelling, calf swelling by ≥3 cm when compared with the asymptomatic leg (measured 10  cm inferior to tibial tuberosity), pitting edema greater in the symptomatic leg, the presence of collateral superficial veins (nonvaricose), previously documented DVT, and alternative diagnosis at least as likely or greater than DVT. The same score for PE includes clinical signs and symptoms of DVT (minimum of leg swelling and pain with palpitation of the deep veins), an alternative diagnosis that is less likely than PE, heart rate >100 bpm, immobilization for >3 days or surgery in the previous 4 week, previous DVT or PE, hemoptysis, and active cancer (treatment ongoing, within the last 6 months, or receiving current palliative treatment). They found that the Wells score had a 100% failure rate for triaging patients with known DVT/ PE.  When performed, D-dimer adequately improved the diagnostic accuracy of a timely diagnosis of DVT/PE in athletes. Hence, improving awareness of an atypical presentation of thrombosis has the potential to reduce the widespread underestimation of DVT/PE [77]. Obviously such a score cannot be used after ovarian hyperstimulation. Another study [78] evaluated the use of pulmonary CT angiography for acute PE in patients categorized as low, moderate, and high probability according to the Wells score. Patients with high probability presented with higher rate of positive PE with CTA. They found that the rates of positive test for PE by pulmonary CT angiography in patients with suspected acute PE were approximately 33%. Sign of DVT S1Q3T3 pattern and enlarged right pulmonary artery were significant clinical predictors for positive pulmonary CT angiography [78]. Hendriksen et al. [79] assessed the cost-effectiveness of point-of-care D-dimer tests (“Simplify,” “Cardiac,” “Triage,” and “Nycocard”) compared with typical laboratory tests, for diagnosis of DVT in primary care. They found that D-dimer tests resulted in similar health outcomes as laboratory-based testing procedures but can be performed more easily and were most cost-effective [79]. In another study [80], the Alere D-dimer assay in combination with the Wells score and the central laboratory standard assay were compared with regard to the detection of DVT or PE, and they were proved to show similar results. Thus, the Alere D-dimer assay may be used in primary care to reduce referrals to the emergency unit [80]. Maskell et  al. [81] assessed the use of automated strain gauge plethysmography to detect DVT and triage to anticoagulation treatment. They found that plethysmography produced a negative predictive value of 97% and a sensitivity of 90% for proximal DVT. Regarding distal DVT, the sensitiv-

88

M. Mina et al.

Fig. 4.23  Pulmonary embolism post-ovarian pick-up

ity was 66%, the specificity 80%, the positive and negative predictive values were 36% and 93%, respectively. They conclude that this test may serve as a quick, easy, non-­ invasive initial screening test, but cannot be used alone but together with clinical assessment [81]. When it comes to guidelines for DVT in pregnancy [82], D-dimer tests have a low specificity. Moreover, a compression ultrasound should be used to evaluate pregnant women for DVT, followed by magnetic resonance imaging of the pelvis for cases with strong clinical suspicion. As for PE, an initial chest X-ray should be used to triage the patient to either a ventilation/perfusion study after a normal X-ray or a CT pulmonary angiogram after an abnormal one [82]. In another review, Wan et  al. [83] reported that neither D-dimer alone nor clinical prediction rules should be used to rule out VTE in pregnant women without objective testing. Also, in cases of clinically suspected PE, echocardiography and ultrasound examination of the lower extremities should be considered as initial imaging methods to substantiate the suspicion of PE and confirm the diagnosis of VTE, respectively [83].

4.10 Preeclapsia 4.10.1 Community-Level (CLIP) Intervention Bellad et al. [84] in a large prospective cohort study evaluated if task-sharing care could reduce adverse pregnancy outcomes, such as pregnancy hypertension, related to delays intriage, transport, and treatment of women in northwest Karnataka, India. The Indian Community-Level Interventions for Preeclampsia (CLIP) open-label cluster randomized controlled trial recruited pregnant women in 12 clusters: 6  in intervention and 6 in control. The intervention clusters consisted of community engagement, community health workers (CHW) who provided mobile health (mHeath)-guided clinical assessment, initial treatment, and referral to facility either urgently (120°, the probability of either a successful vacuum extraction or spontaneous vaginal delivery was 90% or higher [37] (Fig. 17.12). The use of intrapartum ultrasound is particularly useful in case of posterior occiput position associated with asynchlitism to avoid the failures of vaginal operative delivery (Fig. 17.13).

A. C. Gimovsky

17.5.4 Manual Rotation Manual rotation of a fetus in persistent occiput posterior with only minimal caput or molding and a normal fetal heart rate may increase the chance of a successful vaginal delivery during the second stage. Rotation may not be an option in the case of obstructed labor as rotation tends to be more successful earlier on in the second stage of labor [38]. A manual rotation is performed by inserting the hand into the vagina with the palm upward. Digital rotation is attempted by placing the fingertips of the index and middle fingers in the lambdoid suture near the posterior fontanelle. The fingers are used to rotate the head to the occiput anterior by rotation of the hand and forearm. The thumb may be used to exert pressure anteriorly on the parietal bone to assist in rotation (Fig. 17.14). If rotation is unsuccessful, operative delivery can be attempted from the occiput posterior position if the lowest part of the fetal skull is at 2+ station or below.

Fig. 17.12  Sonographic translabial longitudinal section of fetal head in occiput posterior position

Fig. 17.13  Sonographic evaluation of the angle of progression in prolonged labor

Fig. 17.14  Manual rotation in occiput posterior position

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

17.5.5 Operative Vaginal Delivery According to the WHO, only 10% of women with obstructed labor are managed by operative vaginal delivery (OVD) [2]. Vacuum or forceps assisted delivery can be attempted in cases of mild disproportion and low fetal station (Fig. 17.15). If the fetus is in occiput posterior position, there is a significantly higher risk of failure of operative vaginal delivery than with a fetus in occiput anterior position [39] (Fig. 17.16).

17.5.6 Symphysiotomy Symphysiotomy is a procedure where the ligaments connecting the pubic symphysis are divided to allow joint separation and facilitate vaginal delivery by enlarging the dimensions of the maternal pelvis. This can be a life-saving maneuver in rural areas or low resource settings for the mother and fetus when a CD cannot be performed in a timely manner [40]. It must be noted that this procedure is rarely used except in remote areas of resource-limited countries where operating rooms are unavailable [41]. During this procedure, after injection with local anesthesia, a Foley catheter is inserted and the urethra is displaced Fig. 17.15 Forceps-assisted delivery can be attempted in cases of mild disproportion, low fetal station, and non-internal rotation (arrows)

317

laterally using the index and middle fingers placed against the posterior aspect of the symphysis. An incision is made through the cartilaginous portion of the symphysis to allow for permanent separation of the pubic bones (Fig.  17.17). Complications include lacerations of the bladder, urethra, and/or vagina; urinary incontinence; vesicovaginal fistula; and long-term pelvic pain and instability [42].

17.5.7 Cesarean Delivery CD is the most common treatment for obstructed labor and is estimated to be the treatment in 90% of cases [2]. CD after prolonged second stage is often associated with an impacted fetal head, i.e., a fetal head that is very difficult to deliver even via CD (Fig. 17.18). CD performed after prolonged second stage has been associated with long surgery time, increased postoperative fevers, increased blood loss, maternal intraoperative trauma including higher risk of cystotomy (Fig. 17.19), extensions of the uterine incision (and higher composite maternal morbidity. Similar perinatal outcomes are reported when compared with CD performed before prolonged second stage [43–45]. In a prospective multicentric study, maternal and

318

A. C. Gimovsky

Fig. 17.16  If the fetus is in occiput posterior position, there is a significantly higher risk of failure of operative vaginal delivery than with a fetus in occiput anterior position

Fig. 17.18  Cesarean delivery of a G1P0 patient at week 41, 5 hours after complete cervical dilatation with the fetal head in occiput posterior position resulting in right uterine laceration and transfer of the newborn to the neonatal intesive care unit

Fig. 17.17  Symphysiotomy during obstructed labor with shoulder dystocia

Fig. 17.19  Accidental bladder injury with scissors during bladder flap detachment by cesarean delivery in a patient with myomas

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

319

neonatal outcomes were compared when primary cesarean delivery was performed in the second stage vs the first stage of labor [44]. Cesarean deliveries performed in the second stage were associated with longer operative times. The maternal composite index (containing at least one of the following: endometritis, intraoperative surgical complication, blood transfusion, or wound complication) was slightly increased in women undergoing cesarean delivery in the second stage of labor, primarily due to uterine atony, uterine incision extension (Fig.  17.20), and incidental cystotomy. The neonatal composite (which included at least one of the following: Apgar score of 3 or less at 5 min, neonatal death, neonatal intensive care unit admission, seizure, delivery room intubation in the absence of meconium, or fetal injury) did not differ significantly between groups [44].

Fetal risks during a second stage CD include an increased risk of NICU admission and perinatal asphyxia and a decreased risk of scalp and facial bruising compared to operative vaginal delivery [46, 47].

Fig. 17.20  Extension of the incision of the lower uterine segment during cesarean section in prolonged and obstructed labor. The surgeon shows a tear underneath the hysterotomy with forceps

Fig. 17.21  A cesarean delivery for cord prolapse in which the assistant pushes the fetal head from the vagina to avoid compression of the umbilical cord between the fetal head and the pubic symphysis

17.5.8 Cesarean Delivery Positioning Prior to performing a CD for an impacted fetal head, consideration should be given for patient positioning in low-­lithotomy for better access in case assistance from a vaginal hand is necessary. Alternatively, the patient could be placed in “froglegged” position. An assistant may also push the head up and hold the fetal head in place prior to hysterotomy. This should be done utilizing multiple digits on vaginal examination as using one digit is associated with higher pressure which can cause increased trauma. This mechanism elevates the fetal head, thereby creating additional space between the bony pelvis and the fetal presenting part (Fig. 17.21) for the provider’s hand, which consequently helps minimize injury to the fetus and to the maternal soft tissues at the time of cesarean delivery (Fig. 17.22). This furthermore enables the delivery attendant to

320

Fig. 17.22  Cesarean delivery in prolonged labor with the fetal head in marked anterior asynchlitism (the surgeon's finger indicates the squint sign already diagnosed with intrapartum ultrasound). The surgeon must rotate the head and extract it without causing damage to the fetus and maternal tissues before extracting the fetal head

avoid the unlikely situation of having the fetus deliver vaginally before it can be pulled out through the abdominal incision. Although rare, this has been known to happen since the dense regional anesthesia further relaxes the pelvic floor musculature, leading to flexion and rotation of the fetal head, which can then descend and deliver. Lastly, it provides additional information about the extent to which the fetal head is impacted in the pelvis and may influence decision making at the time of cesarean delivery. For example, if the fetal head were deeply impacted in the pelvis and could not be disimpacted vaginally, the surgeon may choose to make a different uterine incision (such as a low vertical hysterotomy) (Fig. 17.23), administer a uterine relaxant (an inhaled anesthetic agent or nitric oxide), ask for additional instrumentation, and/or ask an assistant to be ready to elevate the fetal head vaginally should this be necessary.

17.5.9 Cesarean Delivery Incision Either an incision higher on the uterus than a low transverse incision or a low vertical incision is suggested due to the

A. C. Gimovsky

Fig. 17.23  Vertical incision of the lower uterine segment during cesarean section (arrows)

increased risk of uterine extensions during the delivery. Careful consideration of anatomical landmarks needs to be addressed due to increased distortion of anatomy after an obstructed second stage of labor. A standard incision may risk injury to the vagina, bladder, or cervix (Fig.  17.24). Lower uterine incision extensions are at an increased risk of extension and may be difficult to repair. An inverted T-shaped incision may be necessary if one of the maneuvers is performed as described below (Fig. 17.25).

17.5.10 Cesarean Delivery Medical Adjuncts The provider should recognize that the patient may have ongoing contractions; oxytocin should be discontinued prior to operative intervention if labor has been augmented. Additional consideration should be given to uterine relaxants. Several agents such as nitroglycerine and terbutaline have been reported to be used as uterine relaxants during CD. A systematic review of 13 randomized controlled trials found that the most commonly used agent, nitroglycerin, is not

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

321

superior to placebo for uterine relaxation during CD [48]. Furthermore, no evidence has been found demonstrating the benefit of uterine relaxants during CD at full dilatation [48]. It is also important to note that use of uterine relaxants can

impair postpartum uterine contraction and therefore is likely to increase the risk of postpartum hemorrhage.

17.5.11 Bandl’s Ring at the Time of Cesarean Delivery At the time of CD, a Bandl’s ring can be noted by observing a thick constriction ring separating the upper and lower segments of the uterus, visible as a thick band of tissue (Fig. 17.26). It needs to be excised in order to access the fetus and usually is resolved with a vertical uterine incision. Transecting the ring alone may not be adequate, and the fetus may still be difficult to deliver. Extending the incision to an inverted T shape or a J shape may be necessary for ease of delivery.

Fig. 17.24  Uterine rupture was detected during a cesarean delivery with a transverse incision on the lower uterine segment

Fig. 17.25  Inverted T incision during a cesarean section (arrow): made after a transverse incision of the lower uterine segment to widen the longitudinal hysterotomy and facilitate breech extraction

Fig. 17.26  Bandl’s ring can be noted by observing a thick constriction ring separating the upper and lower segments of the uterus, visible as a thick band of tissue (in this figure Bandl’s ring is indicated by “X”)

322

17.6 Cesarean Delivery Techniques 17.6.1 Alternate Hand Technique When performing a CD, a right-handed surgeon typically stands on the patient’s right side. This enables the surgeon to use their right hand for the majority of the operation, including use of the scalpel and delivery of the head (Fig. 17.27). In the alternative hand technique, the surgeon uses their left hand, therefore crossing their body with their left hand and

A. C. Gimovsky

reaching into the hysterotomy to deliver the fetal head via cephalad traction (Figs. 17.28 and 17.29, Video 17.1). The surgeon thus has an angle for delivery of the fetal head that is more perpendicular to the fetal head than what is available in the traditional right-handed maneuver. Theoretically, this may decrease the risk of uterine extensions, but has not been studied thoroughly.

17.6.2 Pull Vs Push Techniques The “pull” method of delivery during CD involves the use of the “reverse breech extraction” technique done through the hysterotomy incision. This method involves clasping both fetal feet at the fundus of the uterus and applying steady traction in the downward direction, parallel to the fetal lie. Buttocks then follow, and flexion of the spine occurs at the thoracolumbar region. This method allows for more space in which to deliver the fetal head (Fig. 17.30, Video 17.2). The “push” method of delivery via CD involves the use of assistance of a vaginal hand pushing the fetal head upwards to disimpact it (Fig.  17.31, Video 17.3). It is important that an experienced technician be employed to spread equal pressure over the fetal head. Pressure at a single point is more likely to cause fetal trauma such as a skull fracture. If possible, the head should be flexed to narrow the diameter and ease delivery. Several studies have evaluated the “pull” versus “push” techniques for delivery at CD with obstructed labor in the second stage. The evidence supports that the “push” method increases blood loss, operative time, uterine extensions, endometritis and neonatal morbidity, compared to the “pull” method [49–53].

Fig. 17.27  When performing a CD, a right-handed surgeon typically stands on the patient’s right side. This enables the surgeon to use their right hand for the majority of the operation, including the use of the scalpel and delivery of the head. The right-handed surgeon extracts the fetus from the right (not with the left hand that is indicated by the cross), and the left-handed surgeon extracts it from the left. In prolonged labor with the fetal head fixed in the pelvis, the extracting hand must leverage and let the fetal head slip (arrow)

Fig. 17.28  Video tutorial shows alternative hand techniques

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

323

a

b

c

d

Fig. 17.29  Reverse breech technique: (a) This drawing shows clasping of both fetal feet at the fundus of the uterus; (b) operator applies steady traction in the downward direction, parallel to the fetal lie; (c)

this figure shows the buttocks that then follow, and flexion of the spine that occurs at the thoracolumbar region; (d) that method allows more space to deliver the fetal head [52]

17.6.3 Shoulder First Technique

holding the trunk of the baby parallel to the spine and with fundal pressure given by the surgical assistant), the buttocks, and then finally lifting the head out of the pelvis (Figs.  17.32 and 17.33, Video 17.4). In one study, the shoulder first technique was associated with less uterine extensions and blood transfusions than in use of the pull or push techniques [54].

Otherwise known as the Patwardhan technique, the shoulder first maneuver is when the surgeon delivers both fetal shoulders through the incision. If needed, the anterior arm and shoulder can be hooked by the operator’s fingers under the axillae or elbow. This is followed by the trunk (via

324

A. C. Gimovsky

Fig. 17.31 The Video tutorial shows the push method technique

Fig. 17.30 The Video tutorial shows the pull technique

Fig. 17.32  Video tutorial shows shoulder first technique

a

Fig. 17.33  Shoulder first technique is also known as the Patwardhan technique: (a, b) These two drawings show the shoulder first maneuver when the surgeon delivers both fetal shoulders through the incision; (c)

b

this drawing shows delivery of the trunk and the buttocks of the baby; (d) this drawing shows lifting the head out of the pelvis [52]

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

c

325

d

Fig. 17.33 (continued)

spoons, including the Murless head extractor, the Sellheim spoon, and the Coyne spoon. All spoons have a rounded edge that is placed in the uterus between the fetal head and the lower uterine segment; the spoon is elevated directly cephalad to the maternal spine out of the pelvis in order to prevent bladder injury and severe uterine extensions.

17.7.2 C-Snorkel

Fig. 17.34  Video technique

tutorial

shows

abdominovaginal

delivery

17.6.4 Abdominovaginal Delivery Technique This technique, which is similar to the “push method,” except done by the surgeon without an assistant, is when the surgeon uses one hand in the vagina to push the fetal head up and the other hand in the pelvis to deliver the fetal head out of the maternal abdomen [55] (Fig. 17.34, Video 17.5). This technique is usually utilized when there is a lack of skilled assistants at CD.

17.7 Medical Devices 17.7.1 Obstetric Spoon Obstetric spoons have been in use for decades and are employed during delivery of an impacted fetal head since the space taken up by the spoon device is less than that of the surgeon’s hand [56]. There are many types of obstetrical

The C-Snorkel is a disposable device that was developed to release an impacted fetal head by alleviating the vacuum that is created between the fetal head and the maternal tissue [57]. The C-Snorkel device has ventilation ports at the tip of the device to increase airflow and therefore release the vacuum created by the deeply engaged fetal head. It is similarly shaped to an obstetric spoon, with the advantage of being both disposable and ventilated in order to release the pressure of the fetal head that is deep-seated in the pelvis.

17.7.3 Pillow A medical device called the “Fetal Pillow” is a silicone balloon which is placed vaginally prior to CD of a fetus with a suspected impacted fetal head (Fig. 17.35, Video 17.6). The balloon is inflated with 180  mL of normal saline prior to delivery and is placed in a manner to manipulate the fetal head in an upward direction to displace the fetal head cephalad (Figs. 17.36 and 17.37, Video 17.7). Two trials have evaluated this device and have shown that the device reduced difficulty of delivery, reduced incision to delivery time and reduced uterine extensions [58–60].

326

A. C. Gimovsky

Fig. 17.35  Video tutorial shows a medical device called fetal pillow

Fig. 17.36 This Animation Video tutorial explained the use of fetal pillow. The balloon is inflated with normal saline prior to delivery to displace the fetal head cephalad

Fig. 17.38  Cesarean hysterectomy after prolonged and obstructed labor

17.7.4 Uterine Rupture In the case of a suspected uterine rupture, a laparotomy should be performed with delivery of the fetus and placenta. Hemostasis should be attained and consideration should be given for uterine artery ligation or internal iliac ligation prior to proceeding with hysterectomy if necessary (Fig. 17.38).

17.7.5 Fetal Death Fig. 17.37  Fetal pillow (From: Ref. [58])

In the case of fetal death prior to delivery, destructive delivery options may be considered depending on: (a) if the

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

327

maternal condition is critical and (b) legal options based on geographical location.

labor. The model can be used to simulate varying degrees of fetal head flexion, rotation and impaction into the pelvis. In fact, even at the beginning of the twentieth century, mathematical models of normal and abnormal labor (Fig.  17.39) and devices to simulate childbirth were proposed (Fig. 17.40). Research evaluating the use of simulation is currently underway to assess whether the incorporation of this model into training practice could be beneficial [61].

17.8 Obstetrical Trainer for Second-Stage Cesarean Delivery An impacted fetal head simulator facilitates the training of clinicians when confronted with the challenging obstetrical dilemma of delivery of the fetus late in the second stage of

Fig. 17.40  Device to simulate the internal rotation of the fetal head proposed by Italian obstetricians at the beginning of the twentieth century Fig. 17.39  Mathematical model to simulate the internal rotation of the fetal head proposed by Italian obstetricians at the beginning of the twentieth century

328

17.9 Conclusion The intrapartum difficulties surrounding a diagnosis of obstructed labor result in clinical complexity. Both malposition and prolonged second stage of labor and the combination thereof result in increased morbidities for both the mother and the neonate. Therefore, managing this labor abnormality is clinically significant. Ultrasound can be a useful objective tool to assess the success of a vaginal or operative vaginal delivery in this setting (Fig. 17.41). Morbidities include postpartum hemorrhage, stillbirth, maternal sepsis, dehydration, oliguria, ketoacidosis, obstetric fistula, uterine rupture, fetal

Fig. 17.41  Ultrasound can be a useful objective tool to assess the success of a vaginal or operative vaginal delivery in this setting. (The figure shows a prolonged labor in nullipara without labor analgesia diagnosed with intrapartum ultrasonography)

A. C. Gimovsky

injuries (Fig. 17.42), neonatal sepsis, neonatal seizures, neonatal asphyxia, and both neonatal and maternal death. The difficulty with management is most likely resolved at the time of CD. Techniques at CD include proper fluid resuscitation, patient positioning, uterine incision, pull versus push method, shoulder first technique, and abdominopelvic delivery. Several medical devices have been developed to aid the surgeon at the time of cesarean delivery. Knowledge of and practice via simulation of these techniques can aid the clinician for ultimately encountering this challenging patient scenario. Overall, obstructed delivery is a severe, but potentially manageable, complication.

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

329

Fig. 17.42  Intracranial hemorrhage caused by misapplications of forceps in dystocia and prolonged labor Neonatal Med. 2015;28(16):1890–4. https://doi.org/10.3109/14767 058.2014.972925. 12. Pattinson RC, Cuthbert A, Vannevel V. Pelvimetry for fetal cephalic 1. National Institute for Health and Clinical Excellence. Intrapartum presentations at or near term for deciding on mode of delivery. care for health healthy women and babies. Nice. 2014;1–58. Cochrane Database Syst Rev. 2017;3(3):CD000161. Published 2. Dolea C, AbouZahr C. Global burden of obstructed labor in the year 2017 Mar 30. https://doi.org/10.1002/14651858.CD000161.pub2. 2000. Geneva: World Health Organization; 2000. 13. Anshelevich A, Osterhoudt KC, Introcaso CE, Treat JR.  Picture 3. Gimovsky AC, Ghi T. Abnormal second stage. In: Berghella V, ediof the month—quiz case. Halo scalp ring. Arch Pediatr tor. Evidence based labor and delivery. 1st ed. London: Informa Adolesc Med. 2010;164(7):673. https://doi.org/10.1001/ Healthcare; 2018. archpediatrics.2010.109-­a. 4. American College of Obstetricians and Gynecologists. Operative 14. Tinelli A, Di Renzo GC, Malvasi A.  The intrapartum ultrasonovaginal delivery. ACOG Practice Bulletin No. 17. Obstet Gynecol. graphic detection of the Bandl ring as a marker of dystocia. Int J 2012;17(2):1–8. Gynaecol Obstet. 2015;131(3):310–1. https://doi.org/10.1016/j. 5. Brown SJ, Gartland D, Donath S, MacArthurc C.  Effects of proijgo.2015.06.030. longed second stage, method of birth, timing of caesarean section 15. Laughon SK, Berghella V, Reddy UM, Sundaram R, Lu Z, and other obstetric risk factors on postnatal urinary incontinence: an Hoffman MK.  Neonatal and maternal outcomes with prolonged Australian nulliparous cohort study. BJOG. 2011;118(8):991–1000. second stage of labor. Obstet Gynecol. 2014;124(1):57–67. https:// 6. Spong CY, Berghella V, Wenstrom KD, Mercer BM, Saade doi.org/10.1097/AOG.0000000000000278. Erratum in: Obstet GR.  Preventing the first cesarean delivery: summary of a joint Gynecol. 2014 Oct;124(4):842. PMID: 24901265; PMCID: Eunice Kennedy Shriver National Institute of Child Health and PMC4065200. Human Development, Society for Maternal-Fetal Medicine, and 16. Altman M, Sandström A, Petersson G, Frisell T, Cnattingius S, American College of Obstetricians and Gynecologists Workshop. Stephansson O. Prolonged second stage of labor is associated with Obstet Gynecol. 2012;120(5):1181–93. low Apgar score. Eur J Epidemiol. 2015;30(11):1209–15. https:// 7. Gimovsky AC, Guarente J, Berghella V.  Prolonged second stage doi.org/10.1007/s10654-­015-­0043-­4. Epub 2015 May 26. PMID: in nulliparous with epidurals: a systematic review. J Matern Fetal 26008749. Neonatal Med. 2017;30(4):461–5. https://doi.org/10.1080/1476705 17. Gimovsky AC, Berghella V.  Randomized controlled trial of pro8.2016.1174999. Epub 2016 May 5. PMID: 27050812. longed second stage: extending the time limit vs usual guidelines. 8. Altman MR, Lydon-Rochelle MT. Prolonged second stage of labor Am J Obstet Gynecol. 2016;214(3):361.e1–6. and risk of adverse maternal and perinatal outcomes: a systematic 18. Arrowsmith S, Hamlin EC, Wall LL. Obstructed labor injury comreview. Birth. 2006;33(4):315–22. plex: obstetric fistula formation and the multifaceted morbidity of 9. Stitely ML, Gherman RB.  Labor with abnormal presentation and maternal birth trauma in the developing world. Obstet Gynecol position. Obstet Gynecol Clin North Am. 2005;32(2):165–79. Surv. 1996;51:568. 10. Ghi T, Dall’Asta A, Kiener A, Volpe N, Suprani A, Frusca 19. Berhan Y, Berhan A.  Causes of maternal mortality in Ethiopia: a T.  Intrapartum diagnosis of posterior asynclitism using two-­ significant decline in abortion related death. Ethiop J Health Sci. dimensional transperineal ultrasound. Ultrasound Obstet Gynecol. 2014;24(Suppl):15–28. https://doi.org/10.4314/ejhs.v24i0.3s. 2017;49(6):803–4. 20. Menticoglou SM, Manning F, Harman C, Morrison I. Perinatal out11. Malvasi A, Barbera A, Di Vagno G, et  al. Asynclitism: a literacome in relation to second-stage duration. Am J Obstet Gynecol. ture review of an often forgotten clinical condition. J Matern Fetal 1995;173(3 Pt 1):906–12.

References

330 21. Dupuis O, Ruimark S, Corinne D, Simone T, André D, René-­ Charles R.  Fetal head position during the second stage of labor: comparison of digital vaginal examination and transabdominal ultrasonographic examination. Eur J Obstet Gynecol Reprod Biol. 2005;123(2):193–7. 22. Akmal S, Kametas N, Tsoi E, Hargreaves C, Nicolaides KH. Comparison of transvaginal digital examination with intrapartum sonography to determine fetal head position before instrumental delivery. Ultrasound Obstet Gynecol. 2003;21:437–40. 23. Sherer DM, Miodovnik M, Bradley S, Langer O. Intrapartum fetal head position I: comparison between transvaginal digital examination and transabdominal ultrasound assessment during the active stage of labor. Ultrasound Obstet Gynecol. 2002;19:258–63. 24. Sherer DM, Miodovnik M, Bradley KS, Langer O.  Intrapartum fetal head position II: comparison between transvaginal digital examination and transabdominal ultrasound assessment during the second stage of labor. Ultrasound Obstet Gynecol. 2002;19:264–8. 25. Souka AP, Haritos T, Basayiannis K, Noikokyri N, Antsaklis A.  Intrapartum ultrasound for the examination of the fetal head position in normal and obstructed labor. J Matern Fetal Neonatal Med. 2003;13:59–63. 26. Kreiser D, Schiff E, Lipitz S, Kayam Z, Avraham A, Achiron R. Determination of fetal occiput position by ultrasound during the second stage of labor. J Matern Fetal Med. 2001;10:283–6. 27. Akmal S, Tsoi E, Nicolaides KH. Intrapartum sonography to determine fetal occipital position: interobserver agreement. Ultrasound Obstet Gynecol. 2004;24:421–4. 28. Chou MR, Kreiser D, Taslimi MM, Druzin ML, El-Sayed YY.  Vaginal versus ultrasound examination of fetal occiput position during the second stage of labor. Am J Obstet Gynecol. 2004;191:521–4. 29. Ramphul M, Kennelly M, Murphy DJ.  Establishing the accuracy and acceptability of abdominal ultrasound to define the foetal head position in the second stage of labour: a validation study. Eur J Obstet Gynecol Reprod Biol. 2012;164(1):35–9. 30. Ramphul M, Ooi PV, Burke G, Kennelly MM, Said SA, Montgomery AA, Murphy DJ. Instrumental delivery and ultrasound: a multicentre randomised controlled trial of ultrasound assessment of the fetal head position versus standard care as an approach to prevent morbidity at instrumental delivery. BJOG. 2014;121(8):1029–38. 31. Sherer DM, Abulafia O.  Intrapartum assessment of fetal head engagement: comparison between transvaginal digital and transabdominal ultrasound determinations. Ultrasound Obstet Gynecol. 2003;21(5):430–6. 32. Dietz HP, Lanzarone V.  Measuring engagement of the fetal head: validity and reproducibility of a new ultrasound technique. Ultrasound Obstet Gynecol. 2005;25:165–8. 33. Ghi T, Farina A, Pedrazzi A, Rizzo N, Pelusi G, Pilu G. Diagnosis of station and rotation of the fetal head in the second stage of labor with intrapartum translabial ultrasound. Ultrasound Obstet Gynecol. 2009;33:331–6. 34. Tutschek B, Torkildsen EA, Eggebo TM.  Comparison between ultrasound parameters and clinical examination to assess fetal head station in labor. Ultrasound Obstet Gynecol. 2013;41(4):425–9. 35. Duckelmann AM, Bamberg C, Michaelis SA, Lange J, Nonnenmacher A, Dudenhausen JW, Kalache KD. Measurement of fetal head descent using the ‘angle of progression’ on transperineal ultrasound imaging is reliable regardless of fetal head station or ultrasound expertise. Ultrasound Obstet Gynecol. 2010;35:216–22. 36. Masturzo B, De Ruvo D, Gaglioti P, Todros T. Ultrasound imaging in prolonged second stage of labor: does it reduce the operative delivery rate? J Matern Neonatal Med. 2014;27(15):1560–3. 37. Kalache KD, Dückelmann AM, Michaelis SA, Lange J, Cichon G, Dudenhausen JW. Transperineal ultrasound imaging in prolonged

A. C. Gimovsky second stage of labor with occipitoanterior presenting fetuses: how well does the ‘angle of progression’ predict the mode of delivery? Ultrasound Obstet Gynecol. 2009;33(3):326–30. 38. Le Ray C, Serres P, Schmitz T, Cabrol D, Goffinet F. Manual rotation in occiput posterior or transverse positions: risk factors and consequences on the cesarean delivery rate. Obstet Gynecol. 2007;110(4):873–9. https://doi.org/10.1097/01.AOG.0000281666.04924.be. 39. Ben-Haroush A, Melamed N, Kaplan B, Yogev Y.  Predictors of failed operative vaginal delivery: a single-center experience. Am J Obstet Gynecol. 2007;197(3):308.e1–3085. https://doi. org/10.1016/j.ajog.2007.06.051. 40. Monkjok E, Okokon I, Opiah M, Ingwu J, Ekabua J, Essien E.  Obstructed labour in resource-poor settings: the need for revival of symphisiotomy in Nigeria. Afr J Reprod Health. 2012;16(3):93–100. 41. Ersdal HL, Verkuyl DA, Björklund K, Bergström S. Symphysiotomy in Zimbabwe; postoperative outcome, width of the symphysis joint, and knowledge, attitudes and practice among doctors and midwives. PLoS One. 2008;3(10):e3317. https://doi.org/10.1371/journal.pone.0003317. 42. Wilson A, Truchanowicz EG, Elmoghazy D, MacArthur C, Coomarasamy A.  Symphysiotomy for obstructed labour: a systematic review and meta-analysis. BJOG. 2016;123(9):1453–61. https://doi.org/10.1111/1471-­0528.14040. 43. Sung JF, Daniels KI, Brodzinsky L, El-Sayed YY, Caughey AB, Lyell DJ.  Cesarean delivery outcomes after a prolonged second stage of labor. Am J Obstet Gynecol. 2007;197(3) 44. Alexander JM, Leveno KJ, Rouse DJ, Landon MB, Gilbert S, Spong CY, Varner MW, Moawad AH, Caritis SN, Harper M, Wapner RJ, Sorokin Y, Miodovnik M, O’Sullivan MJ, Sibai BM, Langer O, Gabbe SG, National Institute of Child Health and Human Development (NICHD) Maternal-Fetal Medicine Units Network (MFMU). Comparison of maternal and infant outcomes from primary cesarean delivery during the second compared with first stage of labor. Obstet Gynecol. 2007;109(4):917–21. 45. Govender V, Panday M, Moodley J. Second stage caesarean section at a tertiary hospital in South Africa. J Matern Fetal Neonatal Med. 2010;23(10):1151–5. https://doi.org/10.3109/14767051003678002. PMID: 20233130. 46. Murphy DJ, Liebling RE, Verity L, Swingler R, Patel R.  Early maternal and neonatal morbidity associated with operative delivery in second stage of labour: a cohort study. Lancet. 2001;358(9289):1203–7. https://doi.org/10.1016/S0140-­ 6736(01)06341-­3. PMID: 11675055. 47. Allen VM, O’Connell CM, Baskett TF.  Maternal and perinatal morbidity of caesarean delivery at full cervical dilatation compared with caesarean delivery in the first stage of labour. BJOG. 2005;112(7):986–90. https://doi.org/10.1111/j.1471-­0528.2005.00615.x. PMID: 15958005. 48. Morgan PJ, Kung R, Tarshis J. Nitroglycerin as a uterine relaxant: a systematic review. J Obstet Gynaecol Can. 2002;24(5):403–9. https://doi.org/10.1016/s1701-­2163(16)30403-­0. 49. Fasubaa OB, Ezechi OC, Orji EO, Ogunniyi SO, Akindele ST, Loto OM, Okogbo FO. Delivery of the impacted head of the fetus at caesarean section after prolonged obstructed labour: a randomised comparative study of two methods. J Obstet Gynaecol. 2002;22(4):375–8. 50. Veisi F, Zangeneh M, Malekkhosravi S, Rezavand N. Comparison of “push” and “pull” methods for impacted fetal head extraction during cesarean delivery. Int J Gynecol Obstet. 2012;118(1):4–6. 51. Waterfall H, Grivell RM, Dodd JM. Techniques for assisting difficult delivery at caesarean section. Cochrane Database Syst Rev. 2016;(1):CD004944. 52. Lenz F, Kimmich N, Zimmermann R, Kreft M. Maternal and neonatal outcome of reverse breech extraction of an impacted fetal head

17  Simulation of Urgent Obstructed Delivery: Scenario and Triage

331

during caesarean section in advanced stage of labour: a retrospective cohort study. BMC Pregnancy Childbirth. 2019;19(1):98. Published 2019 Mar 27. https://doi.org/10.1186/s12884-­019-­2253-­3. 53. Jeve YB, Navti OB, Konje JC. Comparison of techniques used to deliver a deeply impacted fetal head at full dilation: a systematic review and meta-analysis. BJOG. 2016;123(3):337–45. 54. Saha PK, Gulati R, Goel P, Tandon R, Huria A. Second stage caesarean section: evaluation of Patwardhan technique. J Clin Diagn Res. 2014;8(1):93–5. https://doi.org/10.7860/JCDR/2014/6709.3782. 55. Landesman R, Graber EA.  Abdominovaginal delivery: modification of the cesarean section operation to facilitate delivery of the impacted head. Am J Obstet Gynecol. 1984;148(6):707–10. https:// doi.org/10.1016/0002-­9378(84)90551-­9. PMID: 6702937. 56. Weisman AI, Carrabba SR.  Experiences with the Murless head extractor in cesarean section. J Am Med Assoc. 1952;150(12):1209– 12. https://doi.org/10.1001/jama.1952.63680120012010e.

57. C-Snorkel impacted fetal head release device. Clinical Innovations website. http://www.performance-­mastermedical.com/Product_ Line/OB-­GYN/C-­Snorkel%20Brochure.pdf. Accessed 31 Aug 2020. 58. https://www.safeob.com/fetalpillow. Accessed 31 Aug 2020. 59. Seal SL, Dey A, Barman SC, Kamilya G, Mukherji J, Onwude JL. Randomized controlled trial of elevation of the fetal head with a fetal pillow during cesarean delivery at full cervical dilatation. Int J Gynecol Obstet. 2016;133(2):178–82. 60. Lassey SC, Little SE, Saadeh M, et al. Cephalic elevation device for second-stage cesarean delivery: a randomized controlled trial. Obstet Gynecol. 2020;135(4):879–84. https://doi.org/10.1097/ AOG.0000000000003746. 61. Impacted Fetal Head Simulator. https://www.adam-­rouilly.co.uk/products/clinical-­skills-­simulators/impacted-­fetal-­head/ar58-­desperate-­ debra-­impacted-­fetal-­head-­simulator. Accessed 24 Aug 2020.

Twin Vaginal Delivery

18

Miha Lučovnik, Lili Steblovnik, and Nataša Tul

18.1 Introduction Incidence of twin pregnancies has risen over the last decades in developed countries from the natural occurrence rate, which is just below 1%, to between 2% and 4% of all gestations (Fig.  18.1) [1–4]. This increase has been associated with the expanded use of fertility therapies (in vitro fertilization and ovulation induction techniques) and older maternal age [5, 6]. An increase in the incidence of twin pregnancies in the last decades has been accompanied by an increase in the cesarean delivery rate for twins worldwide [7–10]. In the US, the rate of cesarean deliveries among women with twin gestations increased from 53% in 1995 to 75% in 2008 [10]. According to the Slovenian National Perinatal Information System (NPIS), cesarean delivery rates for twin pregnancies in Slovenia rose from 53% in 2005–2009 to 60% in 2015– 2019. Cesarean deliveries of twins increased in obstetric units with 2000 deliveries yearly (Fig.  18.2). Nevertheless, cesarean delivery rates for twins continue to be Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/978-­3-­031-­10067-­3_18. M. Lučovnik Division of Obstetrics and Gynecology, Department of Perinatology, University Medical Center Ljubljana, Ljubljana, Slovenia Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia L. Steblovnik Division of Obstetrics and Gynecology, Department of Perinatology, University Medical Center Ljubljana, Ljubljana, Slovenia e-mail: [email protected] N. Tul (*) Women’s Hospital, Postojna, Slovenia Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia e-mail: [email protected]

higher in smaller as compared to larger maternity units (Fig. 18.2). Cesarean deliveries in twin pregnancies currently account for approximately 5–7% of the overall cesarean rate and represent around 7% of primary cesarean sections [9, 11]. The rise in cesarean deliveries of twins occurred despite evidence showing no neonatal benefit of the planned cesarean section when the first twin is in cephalic presentation [12–14]. In 1987, a small randomized trial compared the policy of planned cesarean delivery vs. planned vaginal delivery for twins at >35  weeks with a cephalic/non-cephalic presentation [15]. Rabinovici et al. found no difference in neonatal outcomes and an increase in maternal febrile morbidity in the planned cesarean delivery group (40% vs. 11%) [15]. The trial was, however, underpowered to detect a potential small increase in fetal/neonatal risk associated with vaginal birth [15]. Consequently, a large, multicenter randomized trial (the Twin Birth Study) was conducted comparing planned vaginal delivery and planned cesarean delivery of twins between 32 weeks 0 days and 38 weeks 6 days of gestation, with the first twin in cephalic presentation [16]. Planned cesarean delivery did not reduce the risk of fetal/ neonatal mortality or serious neonatal morbidity. There was a higher risk of adverse perinatal outcome for the second twin, as already reported in several observational studies, but the risk was not affected by the planned mode of delivery [16–20]. As a result of the Twin Birth Study findings, most Ob/Gyn societies issued statements encouraging the trial of vaginal delivery in the setting of a cephalic-presenting first twin [12– 14, 21]. A trial of labor could be recommended for the majority of twins, as approximately 40% of twins will be in a cephalic/cephalic presentation, 35% in a cephalic/non-­ cephalic presentation, while twin A will be in a non-cephalic presentation only in the remaining 25% at the time of delivery (Fig. 18.3a–f). In an attempt to reduce the rate of primary cesarean delivery rate, the American College of Obstetricians and

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 G. Cinnella et al. (eds.), Practical Guide to Simulation in Delivery Room Emergencies, https://doi.org/10.1007/978-3-031-10067-3_18

333

334 1,8 1,6 1,4 1,2 1

Dichorionic twins

0,8

Monochorionic twins

0,6 0,4 0,2 0

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

Proportion of twins among all pregnancies (%)

Fig. 18.1  Proportion of dichorionic twins (upper curve) and monochorionic twins (lower curve) among all pregnancies in Slovenia between 1987 and 2019

M. Lučovnik et al.

Year

Fig. 18.2  Changes in cesarean delivery rates for twins in Slovenia between 2005 and 2019

62,00

% of cesarean deliveries in twins

60,00

58,00

56,00

54,00 OB units with < 2000 deliveries per year 52,00

OB units with > 2000 deliveries per year

50,00 2005-2009

Gynecologists and the Society for Maternal-Fetal Medicine made a joined recommendation, that women with either cephalic/cephalic presenting twins or cephalic/non-cephalic presenting twins should be counseled to attempt vaginal delivery [11]. They also stated that “in order to ensure safe vaginal delivery of twins, it is important to train residents to

2010-2014 Time period

2015-2019

perform twin deliveries and to maintain experience with twin vaginal deliveries among practicing obstetric care providers” [11]. Lack of provider skills is often cited as a barrier to providing twin vaginal delivery [22–25]. Active management of second twin delivery is the crucial component of safe twin

18  Twin Vaginal Delivery

335

a

b

c

d

e

f

Fig. 18.3 Various presentations of twins at the time of delivery. Approximately 40% of twins will be in a cephalic/cephalic presentation (a), 35% in a cephalic/non-cephalic presentation (b), and the remaining 25% of twin A will be in a non-cephalic presentation (c–e, or f).

(Adapted with permission from Tul N. Multiple pregnancy. In: Takac I,Gersak K, editors. Gynecology and Perinatology. University of Maribor, 2016: 461–73)

vaginal birth [26]. This may include an internal podalic version and breech extraction for the delivery of the unengaged non-cephalic second twin. Acquaintance with these obstetric procedures is not only required for vaginal delivery of cephalic/non-cephalic presenting twins but also in cephalic/ cephalic twins as the second twin can spontaneously vert to

a transverse lie after delivery of the first twin. Newer generations of obstetric care providers may, however, lack the experience and manual dexterity required for these maneuvers [22, 27]. This is the result of decreasing numbers of vaginal twin deliveries but also steeply rising cesarean rates in breech singletons as a result of the 2000 Term Breech Trial [28].

M. Lučovnik et al.

336

18.2 Simulation-Based Training of Twin Vaginal Delivery Simulation of high-acuity, low-frequency clinical events provides opportunities for training in a realistic way without harming patients [29–38]. It is, therefore, well suited to address current challenges in providing safe twin vaginal deliveries. Simulation training is a complex educational intervention. It can comprise several techniques, from purely technical skill training to simulated scenarios involving multidisciplinary teams. Devices used for simulation training also vary significantly. For example, twin vaginal deliveries can be trained using high-tech full-scale mannequins or much simpler plastic models. Finally, simulation-based training of twin vaginal delivery can be applied in various settings. It can be conducted as off-site simulation in simulation centers or as in-situ simulation occurring in obstetric units with healthcare professionals in their own working environment. Several simulation-based training modalities can, therefore, be used targeting different aspects of twin vaginal delivery. They can mainly focus on individual care provider’s manual dexterity but can also address non-­ technical skills required for the functioning of multidisciplinary teams involved in twin deliveries. In addition, simulation can even aim for organizational learning, such as changes in equipment, guidelines or the physical clinical environment.

Table 18.1  The Peyton four-step approach to teach and train technical skills Step 1 2 3 4

Instructor Performs Shows and explains Performs Observes

Trainee Observes Observes Instructs Explains and performs

(simulation instructors/facilitators) should teach and train technical skills. We use the Peyton four-step approach for this purpose during our simulation training. This is a widely accepted method to teach technical surgical and medical skills, though evidence of its effectiveness is limited [39] (Table 18.1). During the first step, the instructor shows how to perform the skill (e.g. breech extraction of the second twin) without explanation. This enables the trainee to focus on the skill performed without distracting verbal information. During the second step, the instructor shows and explains specific parts of the skill. In the third step, the instructor performs the skill following the instructions provided by the trainee. This step is crucial, as it provides the teacher/instructor insight on trainee’s understanding of how the skill should be performed and avoid any potential misunderstandings. In the fourth step, the trainee first verbalizes and then performs parts of the skill. At this point, the instructor should check if his/her instructions have been correctly understood and if the skill is being performed properly. Misunderstandings and mistakes should be corrected to avoid the wrong automation of the skill. The frequency of feedback during step four should, 18.2.1 Technical Skills Teaching and Training however, be tailored to the level of the trainee’s knowledge and skills. Highly frequent feedback during step four faciliOne aspect of simulation-based training of twin vaginal tates the initial acquirement of the skill in the short term, but deliveries concerns technical skills, i.e. acquisition of dexter- less frequent feedback helps retention of the skill in the long ity required for managing the delivery of the second non-­ term [39]. It is worth remembering that in more advanced cephalic twin. Both internal podalic version and breech trainees, who are able to evaluate and improve their skills extraction can be taught using birth simulators. At our insti- themselves, too frequent feedback may actually hamper tution, we use the Sophie and Sophie’s Mum Birth Simulator learning [40]. (MODEL-med, Portland, OR, USA). The video accompanying this chapter presents the teaching of internal podalic version and breech extraction in our courses of simulation-based 18.2.2 Non-technical Skills Teaching training in obstetric emergencies. and Training Technical skill teaching contains demonstrations, instructions and immediate feedback (especially in the case of nov- In addition to the technical skills required for the safe vaginal ice trainees) to prevent the wrong acquisition and automation delivery of twins, the obstetrician should also possess the of skills [39]. There is little evidence on how exactly teachers non-technical skills needed to lead multidisciplinary teams

18  Twin Vaginal Delivery

involved in twin deliveries. In the last decades, it became clear that training in non-technical skills is crucial for patient safety [41, 42]. Adverse events in all fields of medicine are more often caused by deficiencies and errors in non-­technical than technical skills [41, 42]. Simulation-based approach to teaching and training non-­ technical skills has been described as one of the most effective [43, 44]. During scenario simulations, trainees apply and train non-technical skills, such as situational awareness, leadership skills and communication skills, in a safe but realistic and multidisciplinary setting. Debriefing sessions immediately after the simulation scenario are most commonly used to teach non-technical skills. In order to facilitate this, scenarios must be well-designed. Such scenarios usually have the following three components: 1. beginning (setting the scene, e.g. A 30  year old G2P1 pregnant with dichorionic twins in cephalic/cephalic presentation presents to labor and delivery ward at 36 weeks in active labor) 2. middle (problem needs to be solved, e.g. after vaginal delivery of the first twin, the second twin spontaneously verts to a transverse lie making internal podalic version and breech extraction of the second twin necessary) 3. end (debrief and feedback). Debrief and feedback are challenging but essential parts of simulation-based training. At this point, the instructor brings participating trainees together to review and discuss/critique their actions. During debrief, trainees should receive feedback not only on their theoretical knowledge and technical skills but also on their non-technical skills. Feedback should be based on the instructor’s observation of the trainee’s performance and should address strong points as well as weaknesses with suggestions for improvement [43–45]. It should always be provided in a respectful way. Instructors should try to understand trainee’s perceptions and not merely focus on their own frames. This will stimulate a deeper reflection among trainees. It is, in fact, most important that trainees reflect on their performance during this phase of simulation training. Several methods of debriefing have been developed in order to enable this, but no one has been proven to be superior to others so far [40]. Training of

337

novel instructors by more experienced colleagues is currently the most effective way to enable valuable analysis and teaching of non-technical skills during debriefing after simulation scenarios [45].

18.2.3 Simulation Setting; Off-Site Vs. In-Situ Several non-randomized studies examining simulation-based training of obstetric and non-obstetric emergencies suggested that in-situ simulation may be more effective compared to off-site simulation as it is conducted in a more authentic environment [46–50]. Conversely, randomized trials comparing off-site vs. in-situ simulation-based training of obstetric emergencies, showed that choice of setting does not seem to influence an individual or team learning significantly [51–53]. The in-situ simulation does, however, seem to be superior to off-site simulation in terms of learning at the organizational level [46–48, 53–55]. Our experience confirms this. Although we have not yet used in-situ simulation to train twin vaginal delivery in our labor ward, we ran several simulations-based training of other obstetric emergencies (e.g. eclampsia, postpartum hemorrhage, maternal cardiac arrest, etc.) in our high dependency obstetric unit. Based on our experience, in-situ simulation can be very effectively used to test equipment, emergency response systems within the unit, room layouts, emergency trolley setups, etc. (Figs.  18.4 and 18.5, Video 18.1). A potential drawback of in-situ simulation is that it can, when not planned properly, compromise patients’ safety. In-situ simulation involves the use of equipment and medications from the clinical site. Simulation instructors, who are usually also busy healthcare providers involved in clinical work, must schedule a time to organize mannequins and equipment, run the simulation, and appropriately clean and store the equipment used. They must also be prepared to cancel or postpone simulation training in the event of heavy patient loads or a shortage of staff. The reported rate of cancellation for in–situ simulation ranges from 28% to 67% but seems to decrease as training matures [47, 56, 57]. In-situ simulation can, therefore, be very useful from the organizational standpoint. It does, however, require strong management support.

338 Fig. 18.4 Organizational level learning through in-situ simulation. Simulation training in real working environments can be used to test equipment and physical environments. At our institution, in-situ simulation was successfully used to test medication and equipment storage as well as room layouts after the renovation of the high dependency obstetric unit

M. Lučovnik et al.

18  Twin Vaginal Delivery

339

training in managing obstetric emergencies and correct use of personal protective equipment should not be viewed as a luxury we do not have in pandemic conditions but rather as an essential part of our strive to provide high-quality obstetric care even during these challenging times. Acknowledgement  The authors would like to thank prof. Ziva Novak Antolic, MD, PhD, Trainer the Trainers Course leader at the European Board and College of Obstetrics and Gynaecology (EBCOG), for her support and guidance in developing the simulation-based training at our institution. Fig. 18.5  Video tutorial about twin vaginal delivery

References 18.3 Simulation-Based Training During the Coronavirus Disease 2019 (COVID-­19) Pandemic COVID-19 pandemic has overwhelmed hospital infrastructure and demanded remodeling of healthcare systems worldwide. While most non-urgent clinical care can be temporarily postponed, obstetricians do not have this option. Deliveries and delivery complications occur regardless of the pandemic. As a result, obstetric units had to adapt and develop measures to minimize transmission risks to non-infected patients and protect healthcare workers. This goal can be achieved through several strategies with synergistic effects. Such strategies include physical distancing, careful environmental cleaning, strict adherence to hand hygiene, screening and testing all pregnant women at admission (case identification), isolating and managing SARS-­CoV-­2 positive patients within specifically designated areas of the hospital, and ultimately vaccination of healthcare personnel as well as pregnant women. Correct and consistent use of personal protective equipment is another critical cornerstone for reducing SARS-CoV-2 transmission. Current recommendations from the Centers for Disease Control and Prevention (CDC) for the care of persons with known or suspected COVID-19 include an N95 respirator, if available (or surgical facemask if a respirator is not available), eye ­protection, gown, and gloves [58]. Healthcare personnel should receive proper training of donning (putting on) and especially doffing (removing) personal protective equipment. This is most effectively done through simulation. Simulation can also be used for training and transferring infected patients within different areas of the hospital in order to successfully implement and potentially improve the institution’s protocols. Most importantly, simulation

1. Hamilton BE, Martin JA, Osterman MJK, Curtin SC, Matthews TJ. Births: final data for 2014. Natl vital Stat reports from Centers Dis Control Prev Natl Cent Heal Stat. Natl Vital Stat Syst. 2015;64(12):1–64. 2. Ward Platt MP, Glinianaia SV, Rankin J, Wright C, Renwick M. The North of England Multiple Pregnancy Register: five-year results of data collection. Twin Res Hum Genet. 2006;9(6):913–8. 3. Lucovnik M, Malvasi A, Tinelli A, Tul N.  Twin pregnancies. Management and therapy of late pregnancy complications: third trimester and puerperium. 2017. 4. Tul N, Lucovnik M, Verdenik I, Druskovic M, Novak Z, Blickstein I.  The contribution of twins conceived by assisted reproduction technology to the very preterm birth rate: a population-based study. Eur J Obstet Gynecol Reprod Biol. 2013;171(2) 5. Reynolds MA, Schieve LA, Martin JA, Jeng G, Macaluso M. Trends in multiple births conceived using assisted reproductive technology, United States, 1997-2000. Pediatrics. 2003;111(5 Pt 2):1159–62. 6. Kulkarni AD, Jamieson DJ, Jones HWJ, Kissin DM, Gallo MF, Macaluso M, et  al. Fertility treatments and multiple births in the United States. N Engl J Med. 2013;369(23):2218–25. 7. Schmitz T, Prunet C, Azria E, Bohec C, Bongain A, Chabanier P, et  al. Association between planned cesarean delivery and neonatal mortality and morbidity in twin pregnancies. Obstet Gynecol. 2017;129(6):986–95. 8. Kyvernitakis A, Kyvernitakis I, Karageorgiadis AS, Misselwitz B, Papaspyrou G, Kalder M, et  al. Rising cesarean rates of twin deliveries in Germany from 1990 to 2012. Z Geburtshilfe Neonatol. 2013;217(5):177–82. 9. Rossen J, Lucovnik M, Eggebø TM, Tul N, Murphy M, Vistad I, et  al. A method to assess obstetric outcomes using the 10-Group Classification System: a quantitative descriptive study. BMJ Open. 2017;7(7) 10. Lee HC, Gould JB, Boscardin WJ, El-Sayed YY, Blumenfeld YJ. Trends in cesarean delivery for twin births in the United States: 1995-2008. Obstet Gynecol. 2011;118(5):1095–101. 11. Caughey AB, Cahill AG, Guise J-M, Rouse DJ.  Safe prevention of the primary cesarean delivery. Am J Obstet Gynecol. 2014;210(3):179–93. 12. ACOG Practice Bulletin No. 144: multifetal gestations: twin, triplet, and higher-order multifetal pregnancies. Obstet Gynecol. 2014;123(5):1118–32. 13. Gibson JL, Castleman JS, Meher S, Kilby MD.  Updated guidance for the management of twin and triplet pregnancies from the National Institute for Health and Care Excellence guidance, UK:

340 what’s new that may improve perinatal outcomes? Acta Obstet Gynecol Scand. 2020 Feb;99(2):147–52. 14. Visintin C, Mugglestone MA, James D, Kilby MD. Antenatal care for twin and triplet pregnancies: summary of NICE guidance. BMJ. 2011;343:d5714. 15. Rabinovici J, Barkai G, Reichman B, Serr DM, Mashiach S.  Randomized management of the second nonvertex twin: vaginal delivery or cesarean section. Am J Obstet Gynecol. 1987;156(1):52–6. 16. Barrett JFR, Hannah ME, Hutton EK, Willan AR, Allen AC, Armson BA, et al. A randomized trial of planned cesarean or vaginal delivery for twin pregnancy. N Engl J Med. 2013;369(14):1295–305. 17. Hoffmann E, Oldenburg A, Rode L, Tabor A, Rasmussen S, Skibsted L. Twin births: cesarean section or vaginal delivery? Acta Obstet Gynecol Scand. 2012;91(4):463–9. 18. Smith GCS, Shah I, White IR, Pell JP, Dobbie R. Mode of delivery and the risk of delivery-related perinatal death among twins at term: a retrospective cohort study of 8073 births. BJOG. 2005;112(8):1139–44. 19. Smith GCS, Fleming KM, White IR.  Birth order of twins and risk of perinatal death related to delivery in England, Northern Ireland, and Wales, 1994-2003: retrospective cohort study. BMJ. 2007;334(7593):576. 20. Armson BA, O’Connell C, Persad V, Joseph KS, Young DC, Baskett TF.  Determinants of perinatal mortality and serious neonatal morbidity in the second twin. Obstet Gynecol. 2006;108(3 Pt 1):556–64. 21. Silver RM, Landon MB, Rouse DJ, Leveno KJ, Spong CY, Thom EA, et al. Maternal morbidity associated with multiple repeat cesarean deliveries. Obstet Gynecol. 2006;107(6):1226–32. 22. Blickstein I.  Delivery of vertex/nonvertex twins: did the horses already leave the barn? Am J Obstet Gynecol. 2016;214:308–10. 23. Easter SR, Taouk L, Schulkin J, Robinson JN. Twin vaginal delivery: innovate or abdicate. Am J Obstet Gynecol. 2017;216(5):484– 488.e4. 24. Easter SR, Gardner R, Barrett J, Robinson JN, Carusi D. Simulation to improve trainee knowledge and comfort about twin vaginal birth. Obstet Gynecol. 2016;128(Suppl):34S–9S. 25. Frenken MWE, de Wit-Zuurendonk LD, Easter SR, Goossens SMTA, Oei SG.  Simulation-based training of vaginal twin delivery for experienced gynaecologists: useful or not? Eur J Obstet Gynecol Reprod Biol. 2020;251:89–97. 26. Fox NS, Silverstein M, Bender S, Klauser CK, Saltzman DH, Rebarber A. Active second-stage management in twin pregnancies undergoing planned vaginal delivery in a U.S. population. Obstet Gynecol. 2010;115(2 Pt 1):229–33. 27. Goossens SMTA, Roumen FJME, Derks JB, Kessels FG, Dirksen CD, Nijhuis JG. Planning the mode of delivery for twin pregnancies: a web-based questionnaire. J Obstet Gynaecol. 2016;36(2):172–7. 28. Hannah ME, Hannah WJ, Hewson SA, Hodnett ED, Saigal S, Willan AR.  Planned caesarean section versus planned vaginal birth for breech presentation at term: a randomised multicentre trial. Term Breech Trial Collaborative Group. Lancet. 2000;356(9239):1375–83. 29. Cheng A, Lang TR, Starr SR, Pusic M, Cook DA.  Technology-­ enhanced simulation and pediatric education: a meta-analysis. Pediatrics. 2014;133(5):e1313–23. 30. Langhan TS, Rigby IJ, Walker IW, Howes D, Donnon T, Lord JA.  Simulation-based training in critical resuscitation procedures improves residents’ competence. CJEM. 2009;11(6):535–9. 31. Mileder LP, Urlesberger B, Szyld EG, Roehr CC, Schmölzer GM. Simulation-based neonatal and infant resuscitation teaching: a systematic review of randomized controlled trials. Klin Padiatr. 2014;226(5):259–67.

M. Lučovnik et al. 32. Wagner M, Mileder LP, Goeral K, Klebermass-Schrehof K, Cardona FS, Berger A, et  al. Student peer teaching in paediatric simulation training is a feasible low-cost alternative for education. Acta Paediatr. 2017;106(6):995–1000. 33. Wagner M, Heimberg E, Mileder LP, Staffler A, Paulun A, Löllgen RM.  Status quo in pediatric and neonatal simulation in four central European regions: the DACHS survey. Simul Healthc. 2018;13(4):247–52. 34. Buljac-Samardzic M, Dekker-van Doorn CM, van Wijngaarden JDH, van Wijk KP. Interventions to improve team effectiveness: a systematic review. Health Policy. 2010;94(3):183–95. 35. Merién AER, van de Ven J, Mol BW, Houterman S, Oei SG.  Multidisciplinary team training in a simulation setting for acute obstetric emergencies: a systematic review. Obstet Gynecol. 2010;115(5):1021–31. 36. Lepage J, Ceccaldi PF, Remini SA, Plaisance P, Voulgaropoulos A, Luton D.  Twin vaginal delivery: to maintain skill—simulation is required. Eur J Obstet Gynecol Reprod Biol. 2019;234:195–9. 37. Nielsen PE, Goldman MB, Mann S, Shapiro DE, Marcus RG, Pratt SD, et al. Effects of teamwork training on adverse outcomes and process of care in labor and delivery: a randomized controlled trial. Obstet Gynecol. 2007;109(1):48–55. 38. Siassakos D, Crofts JF, Winter C, Weiner CP, Draycott TJ.  The active components of effective training in obstetric emergencies. BJOG. 2009;116(8):1028–32. 39. Nicholls D, Sweet L, Muller A, Hyett J.  Teaching psychomotor skills in the twenty-first century: revisiting and reviewing instructional approaches through the lens of contemporary literature. Med Teach. 2016;38(10):1056–63. 40. Alken A, Luursema J-M, Weenk M, Yauw S, Fluit C, van Goor H.  Integrating technical and non-technical skills coaching in an acute trauma surgery team training: is it too much? Am J Surg. 2018;216(2):369–74. 41. Gordon M, Darbyshire D, Baker P.  Non-technical skills training to enhance patient safety: a systematic review. Med Educ. 2012;46(11):1042–54. 42. Gordon M. Non-technical skills training to enhance patient safety. Clin Teach. 2013;10(3):170–5. 43. Hull L, Sevdalis N. Advances in teaching and assessing nontechnical skills. Surg Clin North Am. 2015;95(4):869–84. 44. Dedy NJ, Bonrath EM, Zevin B, Grantcharov TP.  Teaching nontechnical skills in surgical residency: a systematic review of current approaches and outcomes. Surgery. 2013;154(5):1000–8. 45. Spanager L, Dieckmann P, Beier-Holgersen R, Rosenberg J, Oestergaard D. Comprehensive feedback on trainee surgeons’ non-­ technical skills. Int J Med Educ. 2015;6:4–11. 46. Riley W, Davis S, Miller KM, Hansen H, Sweet RM.  Detecting breaches in defensive barriers using in situ simulation for obstetric emergencies. Qual Saf Health Care. 2010;19(Suppl 3):i53–6. 47. Patterson MD, Geis GL, Falcone RA, LeMaster T, Wears RL.  In situ simulation: detection of safety threats and teamwork training in a high risk emergency department. BMJ Qual Saf. 2013;22(6):468–77. 48. Walker ST, Sevdalis N, McKay A, Lambden S, Gautama S, Aggarwal R, et  al. Unannounced in situ simulations: integrating training and clinical practice. BMJ Qual Saf. 2013;22(6):453–8. 49. Stocker M, Burmester M, Allen M. Optimisation of simulated team training through the application of learning theories: a debate for a conceptual framework. BMC Med Educ. 2014;14:69. 50. Kobayashi L, Dunbar-Viveiros JA, Sheahan BA, Rezendes MH, Devine J, Cooper MR, et  al. In situ simulation comparing in-­ hospital first responder sudden cardiac arrest resuscitation using semiautomated defibrillators and automated external defibrillators. Simul Healthc. 2010;5(2):82–90.

18  Twin Vaginal Delivery 51. Ellis D, Crofts JF, Hunt LP, Read M, Fox R, James M. Hospital, simulation center, and teamwork training for eclampsia management: a randomized controlled trial. Obstet Gynecol. 2008;111(3):723–31. 52. Crofts JF, Ellis D, Draycott TJ, Winter C, Hunt LP, Akande VA.  Change in knowledge of midwives and obstetricians following obstetric emergency training: a randomised controlled trial of local hospital, simulation centre and teamwork training. BJOG. 2007;114(12):1534–41. 53. Sørensen JL, Navne LE, Martin HM, Ottesen B, Albrecthsen CK, Pedersen BW, et al. Clarifying the learning experiences of healthcare professionals with in situ and off-site simulation-based medical education: a qualitative study. BMJ Open. 2015;5(10):e008345. 54. Rosen MA, Hunt EA, Pronovost PJ, Federowicz MA, Weaver SJ.  In situ simulation in continuing education for the health care professions: a systematic review. J Contin Educ Health Prof. 2012;32(4):243–54.

341 55. Wheeler DS, Geis G, Mack EH, LeMaster T, Patterson MD. High-­ reliability emergency response teams in the hospital: improving quality and safety using in situ simulation training. BMJ Qual Saf. 2013;22(6):507–14. 56. Sørensen JL, Lottrup P, van der Vleuten C, Andersen KS, Simonsen M, Emmersen P, et al. Unannounced in situ simulation of obstetric emergencies: staff perceptions and organisational impact. Postgrad Med J. 2014;90(1069):622–9. 57. Bullough AS, Wagner S, Boland T, Waters TP, Kim K, Adams W.  Obstetric team simulation program challenges. J Clin Anesth. 2016;35:564–70. 58. (CDC) C for DC and P.  Interim infection prevention and control recommendations for healthcare personnel during the coronavirus disease 2019 (COVID-19) pandemic. https://www.cdc.gov/ coronavirus/2019-­ncov/hcp/infection-­control-­recommendations. html.

Emergency Delivery in Patients with Obesity

19

Haitham Baghlaf, Cynthia Maxwell, and Dan Farine

19.1 Introduction and Epidemiology Multiple methods to assess obesity in clinical settings exist, including waist circumference of more than 88 cm, waist to hip ratio of >0.85, and body mass index (BMI) of 30 kg/m2 or more, calculated by dividing the individual’s weight in kilogram by the height in meter squared. The enlarged gravid uterus makes the two former methods less likely to provide meaningful clinical information or estimation of obesity in pregnancy. Moreover, the simplicity of estimating one’s BMI and its replicability in clinical settings make it the most used method to assess obesity. Furthermore, maternal obesity using a BMI cutoff of 30 kg/m2 or more has been associated with an increased risk of maternal mortality even after adjusting for plausible confounding factors [1–4]. However, this method is not without limitations, such as its inability to differentiate between fat and lean body mass indexes. Obesity is subdivided into classes I, II, and III, defined as BMI of 30–34.9, 35–39.9, and 40  kg/m2 or more, respectively [5]. Globally, the obesity rate has increased tremendously in the last four decades, and it has been estimated that the rate has

tripled in men, from 3.2% to 10.8%, and more than doubled in women, from 6.4% to 14.9% [6]. In 2014, at least 350 million adult women had obesity, and around 40 million women had a BMI of 40 kg/m2 or more worldwide [6, 7]. The association between obesity and long-lasting health implications such as diabetes mellitus (DM), dyslipidemia, and hypertension (HTN) is well documented in the literature (Fig. 19.1). The diseases mentioned above contribute to the increased incidence of stroke and coronary heart disease in people with obesity [8, 9]. Obesity places a substantial economic burden on healthcare systems [10, 11]. Worldwide efforts have started to lower the prevalence of obesity; however, the obesity rate in adults is unlikely to change in the near future giving the eight times increased in childhood and adolescent obesity rate in the last four decades [12]. The reasons mentioned above highlight the importance of familiarity with obesity’s effects on individuals’ health as this group would require special care and preparation during antenatal, intrapartum, and postpartum care. This chapter will address emergency delivery in pregnant persons with obesity and its management during the intrapartum period.

H. Baghlaf · C. Maxwell Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Mount Sinai Hospital, Toronto, ON, Canada e-mail: [email protected] D. Farine (*) Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, Mount Sinai Hospital, University of Toronto, Toronto, ON, Canada © The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 G. Cinnella et al. (eds.), Practical Guide to Simulation in Delivery Room Emergencies, https://doi.org/10.1007/978-3-031-10067-3_19

343

344

H. Baghlaf et al.

Fig. 19.1  The association between obesity and long-lasting health implications such as diabetes mellitus (DM), dyslipidemia, and hypertension (HTN) is well documented in the literature

19  Emergency Delivery in Patients with Obesity

19.2 Prenatal Assessment People planning a pregnancy with a BMI of 30  kg/m2 or more would benefit from preconception counseling and multidisciplinary care during their prenatal period to optimize the pregnancy, delivery, and postpartum care. Moreover, people with obesity class III should be at least co-followed by obstetrical providers with expertise in pregnancies complicated by a BMI of 40 kg/m2 or more. Hospitals should be prepared to deliver optimal care to patients with BMI ≥30  kg/m2. Clinics, labor and delivery rooms (L&D), and operating rooms should be established in a way to accommodate these patients and ensure everyone’s safety. Specific pieces of equipment should be available all the time, including but not limited to large blood pressure cuffs in clinics, L&D and operating rooms, appropriately-­ sized gowns, bariatric operating beds, wheelchairs, spaced corridors in L&D and operating theaters, sufficient personnel to aid with patient transfer and large seats in waiting areas. Obstructive sleep apnea (OSA) is more common in people with obesity as compared to normal weight. Moreover, the rise in estrogen and progesterone level during pregnancy increases the risk of OSA due to physiological hypervolemia and upper airway congestion [13]. In a population database study, OSA was associated with a significantly increased risk of preeclampsia, cardiomyopathy, pulmonary embolism, and

345

a fivefold increased risk of maternal mortality [14]. Moreover, it is vital to diagnose OSA in pregnant people with obesity and refer them to anesthesia services in a timely fashion as this group is at high risk of failed intubation [15, 16]. Screening methods have been developed to assess the risk of OSA.  Berlin questionnaire is the most commonly used method in pregnancy; however, its validity in the pregnant population has been challenged. Its sensitivity and specificity compared to polysomnography were 35% and 64%, respectively [17]. In the light of lacking validated method for optimal screening for OSA in pregnant people with obesity, we suggest all pregnant people with HTN or DM and a history of snoring be referred for in-laboratory polysomnography. Suggested treatment modalities include minimizing gestational weight gain and continuous positive airway pressure.

19.3 Emergency Delivery 19.3.1 Risk Factors for Emergency Delivery Pregnant people with a BMI of 30 kg/m2 or more are likely to have their pregnancy complicated with hypertensive disorders and pre-gestational and gestational diabetes [8, 9, 18]. These co-morbidities place the pregnant person at higher risk of macrosomia (Fig.  19.2), intrauterine growth restriction,

Fig. 19.2  Obesity place the pregnant person at higher risk of diabetes and fetal macrosomia

346

emergency induction of labor, cesarean section (C-section), and stillbirth [19–23]. The factors mentioned above and the nature of delayed spontaneous labor in this population led to an increased labor induction rate [24]. A Canadian study found that almost half of the pregnant people with obesity underwent induction of labor compared only to 29% in those without obesity [22]. Body mass index of 30 kg/m2 or more is associated with an altered labor curve leading to prolonged labor and increased risk of C-section [25]..

19.3.2 Anesthesia Prospective 19.3.2.1 Epidural Analgesia Generally speaking, regional anesthesia minimizes the cardiorespiratory system’s alteration, and it is the preferred method in all pregnant people regardless of BMI in labor or undergoing a C-section. Normal anatomic changes related to pregnancy should be considered during the placement of regional anesthesia. These include narrower interspace between the lumbar spinous processes, shifting in lumbar lordosis and thoracic kyphosis levels resulting in high sensory blockage levels, and the change of the imaginary line (Tuffier’s line) from L4-L5 to L3-L4 interspace (Fig. 19.3) [26, 27]. Moreover, the hormonal changes in pregnancy affect the consistency of the ligamentum flavum, resulting in loss of sensation as the needle passes it and increases the risk of dural puncture [28]. Furthermore, patients with painful contractions might face a challenge in maintaining an ideal position to place the regional anesthesia catheter. Two different positions are usually used to place the epidural catheter; Fig. 19.3  Normal anatomic changes related to pregnancy should be considered during the placement of regional anesthesia. These include narrower interspace between the lumbar spinous processes, shifting in lumbar lordosis and thoracic kyphosis levels resulting in high sensory blockage level, and the change of the imaginary line (Tuffier’s line) from L4-L5 to L3-L4 interspace

H. Baghlaf et al.

these include the left lateral decubitus and the sitting positions (Fig.  19.4). In the former position, anesthesiologists generally face difficulty identifying the midline in a pregnant person with obesity, and it is deemed that sitting posture is the most preferred position. Some commercially available positioning devices can help the pregnant patient during this procedure. Placement of regional analgesia catheter is more likely to be unsuccessful at the first attempt in more than 40% of pregnant people with class III obesity compared to 6% in those without obesity [29]. This number reached 75% in a study by Perlow and colleagues, and the authors reported that at least three attempts for a successful placement in 14% of the pregnant population with class III obesity were needed [30]. Numerous complications result from improper placement of the epidural catheter, and it can be as severe as spinal hematoma or injury to neuronal structures [31, 32]. In recent years, different techniques and specialized equipment have at least doubled epidural placement success in this population [32, 33]. The use of ultrasound to allocate the midline and estimate the epidural space depth has proven effective [32] (Fig. 19.5).

19.3.2.2 Epidural Anesthesia for C-Section In a population study from the United Kingdom, at least 80% of C-sections were unplanned, and 40% were secondary to fetal distress [34]. Several medical societies recommend a maximum of 30  min interval from decision to delivery time in emergency C-sections [35–37]. Therefore, being prepared for such emergency deliveries can prevent drastic sequelae and complications. One of the most advan-

19  Emergency Delivery in Patients with Obesity

347

Fig. 19.5  In the obese pregnant patient, the use of ultrasound to allocate the midline and estimate the epidural space depth has proven effective

Fig. 19.6  Several medical societies recommend a maximum of 30 min interval from decision to delivery time in emergency C-sections. Therefore, being prepared for such emergency deliveries can prevent drastic sequelae and complications. One of the most advantages of placing an epidural in labor is using the catheter for a top-up in emergency cesarean delivery and avoiding general anesthesia (GA)

Fig. 19.4  The sitting positions of the obese patient during the epidural anesthesia. In the former position, anesthesiologists generally face difficulty identifying the midline in a pregnant person with obesity, and it is deemed that sitting posture is the most preferred position

tages of placing an epidural in labor is using the catheter for a top-up in emergency cesarean delivery and avoiding general anesthesia (GA) (Fig. 19.6) [38]. This was clearly demonstrated in Lim and colleagues’ study, as no difference between top-up epidural and GA regarding the decision to delivery interval time was detected [39]. Also, it has been noted that an epidural with adequate block has a lower rate of conversion to GA than an epidural with inadequate block, 1.5% versus 14.5%, respectively [40]. The top-up can be given in the L&D room before transferring the patient to the

operating theater. This technique will help to reduce the time interval from decision to incision [41]. In the circumstances with an inadequate epidural, spinal anesthesia can be considered; however, this can result in a high spinal block (Fig.  19.7) [42]. In an audit of 115 pregnant people with inadequate epidural analgesia, the use of lower spinal anesthesia dose with the addition of opioids has achieved adequate anesthesia without any high spinal block cases [43]. However, a study from Canada showed that spinal anesthesia extended the decision to incision interval by 20  min compared to GA, functioning epidural, or converting to GA [44]. In conclusion, it is crucial to establish an early epidural, confirm the catheter’s proper placement, and obtain frequent assessments of its functionality during labor to decrease the conversion risk to GA when emergency cesarean delivery rises [29, 45].

348

H. Baghlaf et al.

Fig. 19.7  In the circumstances with an inadequate epidural, spinal anesthesia can be considered; however, this can result in a high spinal block

Fig. 19.8  Combined spinal-epidural catheter placement may convey an easier placement than spinal and provide continuous anesthesia during the procedure. This type of anesthesia can be used in an unplanned C-section deemed not urgent

19.3.2.3 Combined Spinal-Epidural Versus Spinal Anesthesia for C-Section A pregnant person requiring an emergency C-section without established regional anesthesia represents a challenge to the healthcare team. Usually, the procedure is longer in people with obesity, and the use of combined spinal-epidural (CSE) anesthesia offers multiple advantages over the use of single-shot spinal anesthesia. Combined spinal-epidural catheter placement may convey an easier placement than spinal and provide continuous anesthesia during the procedure [46]. This type of anesthesia can be used in an unplanned C-section deemed not urgent (Fig.  19.8). However, this might not be feasible on rare occasions, and anesthesiologists might need to use GA as the only valid option for a C-section.

60  min from the administration [48]. This is an important aspect to consider in the pregnant population with obesity at the time of extubation, as they typically undergo a lengthy C-section. Difficult intubation was reported in 5.2% of pregnant people with class II obesity undergoing GA, and this is a 35% increase over the rate in people with normal BMI [49]. Moreover, difficult intubation is more profound with increasing BMI, and it can be encountered in at least one-third of the pregnant population with class III obesity [29]. It is attributed to positioning difficulty, increased neck soft tissue, and enlarged tongue [50, 51]. The operating theater should be well equipped with different intubation devices such as laryngoscope blades, different endotracheal tube sizes, and fiberoptic intubation devices (Fig. 19.9) to help facilitate intubation in this population [52]. Awake fiberoptic intubation can help patients anticipated to have difficult intubation; however, this method can be time-consuming and can lengthen the decision to delivery interval time [53]. The presence of two anesthesiologists with expertise in managing difficult airways is recommended [54]. In emergency cases, a rapid sequence induction is indicated if the intubation appeared not to be dif-

19.3.2.4 General Anesthesia for C-Section The use of pharmacological aspiration prophylaxis medications minimizes the risk of aspiration in people with obesity. The administration of two medications can have an additive effect [47]. One of the most commonly prescribed medications is sodium citrate, and its effect usually wanes off after

19  Emergency Delivery in Patients with Obesity

Fig. 19.9  Moreover, difficult intubation is more profound with increasing BMI, and it can be encountered in at least one-third of the pregnant population with class III obesity. It is attributed to positioning difficulty, increased neck soft tissue, and enlarged tongue. The operating theater

349

should be well equipped with different intubation devices such as laryngoscope blades, different endotracheal tube sizes, and fiberoptic intubation devices to help facilitate intubation in this population

ficult. Chiron et al. found eight deep breaths in 1 min were as effective as the 3-min tidal volume of denitrogenation [55]. This method is essential for a pregnant person with obesity before administrating GA. The use of GA with opioid administration in pregnant persons with a BMI of 30 kg/m2 or more and known to have OSA increases the risk of opioid-induced respiratory depression. Thus, fully awake extubation is highly recommended in this scenario [56, 57].

19.3.3 Emergency C-Section Delivery A systematic review and meta-analysis by Poobalan and colleagues showed an increased C-section rate in people with obesity [21]. Although pregnant people with obesity represented only 35–44% of the pregnant population, an emergency C-section was performed half of the time in people with obesity [58–60]. Moreover, it is estimated that one in every five patients with obesity will deliver by an emergency C-section [61]. Surgical site infection (SSI) is a common devastating complication of C-section, and the rate of SSI is five-fold increased in people with obesity [62] (Fig.  19.10). Prophylactic antibiotic usage reduced the incidence of wound infection by 60%. In the general population, the recommended dose of Cefazolin is 2 g to achieve the above SSI rate reduction [63]. Several studies have questioned the dosage level in patients with obesity. Swank and colleagues reported a higher adipose concentration in pregnant people with obesity if 3 g of Cefazolin was used instead of the 2 g [64]. However, a study from two tertiary hospitals in the United States had failed to show a statistical difference in the SSI rate between the 2 and 3 g dosing of Cefazolin in patients

Fig. 19.10  Surgical site infection in a patient in puerperium after Pfannenstiel incision at cesarean section. The surgical site infection is a common devastating complication of C-section, and the rate of SSI is five-fold increased in people with obesity

350

with class III obesity [65]. In a case-control study, patients were twice likely to develop SSI if they underwent an emergency C-section compared to elective C-section despite receiving appropriate antibiotic prophylaxis [66]. The very urgent C-section might encounter a higher SSI rate if proper sterilization was not done prior to skin incision. The addition of Azithromycin 500  mg IV to Cefazolin has proven to reduce the risk of wound infection by 64% compared to placebo in a randomized clinical trial [67]. Although this study did not perform a sub-analysis in the population with a BMI of 30 kg/m2 or more, patients with obesity represented 73% of the study population. Based on the previous evidence, we recommend using Azithromycin in addition to the 3  g Cefazolin in an emergency C-section. Furthermore, a recent randomized clinical trial in patients with obesity showed a 60% reduction in SSI rate in pregnant person who received Metronidazole 500 mg combined with Cephalexin 500 mg every 8 h for a total of 48 h in the postpartum period compared to placebo [68]. Different cesarean skin incisions have been used in people with obesity. Most obstetricians prefer Pfannenstiel incision [69] (Fig.  19.11). The vertical incision has been associated with a lower risk of low umbilical artery PH and a lower rate of low 5-min APGAR score in patients with a BMI

H. Baghlaf et al.

of 40 kg/m2 or more [70]. However, it was associated with a higher risk of wound complications in the elective C-section, which might be more profound in emergency cases. Supraumbilical transverse incision might be the right route to enter the abdomen in emergency cases as it has the advantage of easy access to the lower uterine segment and delivering the fetus over the other techniques [71]. Moreover, intraoperative ultrasound has been used in elective cases to determine the shortest distance between the skin and the gravid uterus [72]. This might be beneficial in case of an emergency. However, the operator should outweigh this technique’s benefit against the risks of delaying the C-section if the ultrasound machine was not running and readily available in the operating room. Once the fetus is delivered, the procedure should not be rushed, and it should follow the same steps as an elective C-section. The closure of subcutaneous tissue with a depth of 2  cm or more has reduced wound complications by 34% [73]. In some patients with obesity class III, this might require closure with multiple layers (Fig.  19.12). Neither subcutaneous drain placement (Fig. 19.13) nor prophylactic negative pressure wound therapy (NPWT) showed benefit in systematic review and meta-analysis studies of randomized clinical trials [74, 75]. However, in 2018, a meta-analysis including randomized and non-randomized studies of NPWT showed a reduced risk of SSI by 32%; this reduced risk was not seen in the last two randomized clinical trials since the mentioned meta-analysis [76–78]. Furthermore, the healthcare team should obtain an X-ray imaging of the abdomen at the end of the C-section if instruments and sponge counts were not obtained at the beginning of the procedure or if the count was incorrect.

19.3.4 Implications of COVID-19 on Emergency Delivery

Fig. 19.11  Different cesarean skin incisions have been used in people with obesity. Most obstetricians prefer Pfannenstiel incision

It has been suggested that pregnant people with higher BMI require a longer time from decision to delivery in emergency C-section situations. A study by Turner and Warshak did not show a statistically significant difference in terms of the mean interval from the decision to delivery time in patients with obesity [79]. However, it was noted that the functioning regional analgesia rate was 64%, 75%, and 80% in patients with BMI 4000  g, footling breech presentation, an extended neck, or an existing indication for cesarean deliv-

372

ery [1]. RCOG guidelines also make note of characteristics that increased the probability of a poor outcome in the PREMODA cohort, which will be discussed later in this chapter [16], including the presence of a hyperextended neck on ultrasound, estimated fetal weight 3800 g, non-frank breech presentation, and evidence of antenatal fetal compromise [3]. Both ACOG and RCOG guidelines discuss the importance of having trained personnel available for the delivery of a fetus in breech presentation as one of the most important modifiable factors affecting the neonatal outcome, as well as creating an institutional protocol for breech vaginal delivery to decrease the risk of neonatal morbidity [17].

20.3.2.1 Labor Induction or Augmentation There is limited data on induction or augmentation of labor for fetuses in breech presentation. It has been generally believed that labor augmentation should be avoided for fetuses in breech presentation, as inadequate labor progress is the best evidence of fetopelvic disproportion. Of note, the PREMODA study included fetuses undergoing both labor induction and augmentation; as such, RCOG and SOGC do not exclude the use of labor augmentation for cases of uterine hypocontractility [3, 18]. 20.3.2.2 Potential Maternal and Neonatal Complications of Breech Vaginal Delivery Historically, fetuses in breech presentation have had a poorer outcome than those in cephalic presentation. The reasons for these findings are varied and include both fetal and maternal etiologies. Fetal etiologies include a higher rate of anomalies that are associated with fetal malpresentation such as hydrocephaly or macrocephaly, stillbirth, prematurity, and oligohydramnios. Maternal labor and delivery complications contributing to poorer outcomes include cord prolapse and traumatic birth injury including humeral fracture, clavicular fracture, or dislocation of the neck, shoulder or hip joints, and internal organ trauma. Neonatal risks also include brachial plexus injury, depressed APGAR scores, intracranial hemorrhage (due to head compression during delivery), and overall increased perinatal mortality when compared to planned cesarean delivery for breech presentation [19]. Maternal complications described after vaginal breech delivery include higher order perineal lacerations and obstetric anal sphincter injuries (OASIS, mainly due to delivery of the aftercoming head) bladder injury, pelvic hematoma formation, and cervical laceration with extension to the lower uterine segment, all of which can be associated with an

J. Bouganim et al.

increased risk of infection, and hemorrhage requiring transfusion of blood products [20].

20.3.2.3 Counseling The Term Breech Trial (TBT) published in October 2000 was the impetus for a change in practice worldwide. As documented by Lee et al. [21], in 1992, most singleton breech fetuses in the United States at that time were already delivered by cesarean; however, the rate of cesarean still increased from 83.8% to 85.1% in the 6 years following the publication of the TBT.  The increase in cesarean deliveries was more impressive in other countries such as the Netherlands, where the cesarean rate increased from 50% to 80% within 2 months of publication. This change of practice was due to a shift in the language of guidelines provided by ACOG, RCOG, and other governing obstetrical societies, indicating that a planned term vaginal breech delivery was not advisable. Over 20 years since the TBT’s publication, however, its conclusions have been questioned in light of follow-up data in which the long-term outcomes appear to be similar for both groups [16, 17, 22]. Most current guidelines for the management of breech-­ presenting fetuses at term now highlight both the importance of a detailed discussion between provider and patient regarding the options for management (including planned cesarean delivery, external cephalic version, and planned breech vaginal delivery) and the factors predicting safe breech vaginal birth. Discussion regarding a planned breech vaginal delivery at term should include information regarding maternal and neonatal outcomes. Most data regarding neonatal outcomes, with exception of the TBT, is retrospective. The TBT did not report a difference in perinatal mortality between planned cesarean delivery and a planned vaginal breech delivery in countries with a baseline low perinatal mortality rate but did report a significant increase in serious short-term neonatal morbidity of 5.1% as opposed to 0.4% in this population. Serious neonatal morbidity was defined as significant birth trauma, seizures 400 mg/dL [6]. Cortet et al. found that the odds of severe PPH (hemoglobin decrease ≥4 g/dL, red cell transfusion, arterial emboliza-

tion or emergency surgery, admission to intensive care, or death) for patients with fibrinogen between 200 and 300 mg/ dL were almost doubled and increased 12-fold for fibrinogen less than 200  mg/dL compared to patients with fibrinogen >300 mg/dL [7].

43.3 Looking for a Goal-Directed and Point-of-Care Transfusional Approach in MOH Point-of-care coagulation monitoring methods, whenever present should be considered (Table 43.2). It is a daily practice and there is nowadays enough evidence on the use of thromboelastography and rotational thromboelastometry in parturients for correction of coagulopathy effectively and quickly. A targeted goal-directed transfusional therapy based on such viscoelastic methods is associated with less incidence of allogenic blood transfusions and thromboembolic event [15, 16]. Table 43.2  Indication of the use of VHA by some anaesthesiology societies Postpartum haemorrhage

ESA (European society for Anaesthesiology)

The association of Anaesthetists of Great Britain and Ireland ISTH SSC (International society on thrombosis and haemostasis scientific and standardization committee) SFAR (Société Française d’Anesthésie et de Réanimation)

Viscoelastic haemostasis assays are recommended to identify obstetric coagulopathy In severe PPH, a VHA-­ guided intervention protocol is recommended [16] There should be equipment to enable bedside estimation of coagulation such as TEG or ROTEM analysers [17] Recommend monitoring haemostasis with either PT/ aPTT and Clauss fibrinogen or POCTs using thromboelastometry during PPH [18] The French Working group on Perioperative Haemostasis on viscoelastic tests [19]: Fibrinogen concentration should be rapidly evaluated in the event of PPH and viscoelastic tests may be useful in this regard

aPTT activated partial thromboplastin time, POCT point-of-care test, PPH postpartum hemorrhage, PT prothrombin time, VHA viscoelastic hemostatic assays

43  Transfusional Optimization Using Viscoelastic Test Guided Therapy in Major Obstetric Hemorrhage: Simulation and Skills

43.4 Thromboelastography (TEG® 5000 and TEG® 6s Hemostasis Analyzers with TEG Manager® Software) TEG allows a point of care and near real-time assessment of the clotting capacity, providing the interpreter informations regarding the shear elasticity and the dynamics of clot development, strength, stability, and dissolution. Static measurements of hemostasis such as PT, aPTT, INR, fibrinogen level, and fibrin degradation products haven’t this kind of capability. TEG provides information about all components of hemostasis: coagulation, platelet function through the platelet mapping too, fibrinolysis and functional fibrinogen test but offers a particular advantage in diagnosing fibrinolysis. Through the graphic interpretation of TEG, we can easily assess any kind of variation in the coagulation pattern given by coagulopathies such as thrombocytopenia, factor deficiency, heparin effect, hypofibrinogenemia, and hyperfibrinolysis.

43.5 How TEG Works In order to perform a TEG, a citrated-sample of whole blood is placed into a heated sample cup with calcium chloride (to overcome the effects of the citrate), kaolin (a negatively charged molecule known to initiate the intrinsic pathway), and cephalins-phospholipids (required for optimal functioning of the extrinsic pathway). As the sample cup oscillates in a limited arc, the formation of clot results in the generation of rotational forces on a pin suspended from a torsion wire. Forces translated to the torsion wire are then transmitted to an electrical transducer and displayed on a computer screen, creating a characteristic waveform with numerical measurements pertaining to the kinetics of fibrin formation, fibrinolysis and the strength of the resulting fibrin clot. Heparinase cups are commonly paired with plain cups to identify a heparin effect (h-TEG). The TEG® 6s device is used with disposable cartridges that contain all of the assay reagents, avoiding the need for reagents to be pipetted manually.

43.6 Rapid TEG (r-TEG) Rapid TEG (r-TEG) can be completed within 15 min as compared to 30–45 min. For a standard TEG. In contrast to a TEG, whole blood samples for an r-TEG can be performed with citrated or non-citrated samples.

685

Samples utilized for r-TEG are combined with celite (tissue factor activating the extrinsic pathway), and kaolin (activating the intrinsic pathway) +/− calcium chloride as applicable.

43.7 Functional Fibrinogen The Functional Fibrinogen assay measures the contribution of fibrinogen to the strength and stability of the cloth. It is performed with a platelet glycoprotein IIb/IIIa receptor antagonist, which prevents platelets from binding and contributing to clot formation. As a result, the formed clot is based on fibrin activity. The level of fibrinogen is estimated from the MA of the assay. The contribution of platelets to the clot strength can be directionally comparing the MA in the Functional Fibrinogen assay and the Kaolin assay, or through comparison with the Rapid TEG.

43.8 Platelet Mapping Assays TEG Platelet Mapping assay is a platelet function test and it is performed using four separate. Assays: Kaolin, Activator F assay, Arachidonic acid plus Activator F assay, Adenosine diphosphate plus Activator F assay. It allows to measure the effects of antiplatelet medication (e.g., ticagrelor, clopidogrel) or aspirin or to assess platelet function.

43.9 Rotational Thromboelastometry (ROTEM®) Unlike traditional TEG, which utilizes a sample cup rotating in a limited arc, ROTEM employs a static sample cup with an oscillating pin/wire transduction system. ROTEM is a more complex diagnostic test than TEG, as it has four channels with different reagents to detect abnormalities in different components involved in coagulation: • INTEM: Phospholipids for Intrinsic pathway activation. • EXTEM: Tissue factor for Extrinsic pathway activation. • HEPTEM: (Heparinase enz. + Phospholipids) for neutralization of heparin. • FIBTEM: (Cytochalasin D) to inhibit platelet activity to differentiate between hypofibrinogenemia and platelet deficiency. • APTEM: (Aprotinin + Tissue factor) predicts clinical effect of fibrinolysis inhibitors in case of hyperfibrinolysis. • NATEM: Native whole blood without reagent.

L. Mirabella et al.

686

The values of analogous TEG and ROTEM parameters are not interchangeable but provide similar interpretations.

43.10 Parameters of TEG and ROTEM (Table 43.3) • R (Reaction time, min) or CT (Clotting Time, s): Is the period of time from initiation of the test until clot firmness reaches an amplitude of 2 mm, normal range 5–7 min. • K (Kinetic time, min) or CFT (Clot Formation Time, s): Is a measure of time from the beginning of clot formation until the amplitude reaches 20 mm, and represents the dynamics of clot formation, normal range 1–3 min. • α-angle (degree): Is an angle between the line in the middle of the TEG tracing and the line tangential to the developing “body” of the TEG tracing. The alpha angle represents the acceleration (kinetics) of fibrin build-up and cross-linking, normal range 53–67 degrees. • MA (Maximum Amplitude, mm) or MCF (Maximum Clot Firmness, mm): reflects the strength of the clot which is dependent on the number and function of platelets and its interaction with fibrin, with a normal range of 59–68 mm. • CL30 or LY30 (%) (A30/MA*100): Clot lysis is measured as the decay in MA over 30 min., normal range of 0–8%. • CL60 or LY60 (%) (A60/MA*100): Clot lysis is measured as the decay in MA over 60  min., normal range 12 mm. • During major PPH, fibrinogen replacement may improve clinical hemostasis if FIBTEM A5 is 1000 ml cesarean Blood loss>1500 ml severe

Consider early TXA administration (within 3 h) according to the WOMAN trail

YES

A5 ECTEM < 35 mm or CR FIBTEM>600 a (falthine) or ML≥10% (within 60 min)

YES

Tranexamic acid 1-2 g as a single bolus (repeat if indicated)

YES

HYPERFIBRINOLYSIS

DONE NO

A5 FIBTEM < 12 mm YES

Fibrinogen concentrate or cryoprecipirate (dose calculation according to the table3) Target: A5 FIBTEM> 16 mm

CLOT FIRMNESS EVALUATION

NO A5 EXTEM < 35mm And A5 FIBTEM > 12 mm

Platelet concentrate 1 pooled oraphersis YES

NO

CT EXTEM < 80 s And A5 FIB ≥ 12 mm

YES

4F-PCC 10-15 IU/kg bw Or FFP 10-15 ml/kg bw (FFP may dibute fibrinogen)

Consider FFP 10 ml/kg bw Or 90 µg/kg rFVIIa

NO

CT INTEM < 240 s YES

Ongoing bleeding YES

Fig. 43.1  Transfusion management in MOH

CT INTEM / CT HEPTEM ≥1.25

YES

NO

THROMBIN GENERATION EVALUATION

Consider protamine 25-50 mg (2.5-5 ml)

Re-check after 10-15 min using a new blood sample

CLINICAL AND POINT OF CARE REASSESSMENT

43  Transfusional Optimization Using Viscoelastic Test Guided Therapy in Major Obstetric Hemorrhage: Simulation and Skills

Fig. 43.2  TEG guided transfusion algorithm

689

690

L. Mirabella et al. THROMBOELASTOMETRY (ROTEM) GUIDED TRANSFUSION ALGORITHM

CLINICAL BLEEDING EVALUATION

MOH A5 algorithm Blood loss>500 ml vaginal Blood loss>1000 ml cesarean Blood loss>1500 ml severe

Consider early TXA administration (within 3 h) according to the WOMAN trail

YES

A5 ECTEM < 35 mm or CR FIBTEM>600 a (falthine) or ML≥10% (within 60 min)

YES

Tranexamic acid 1-2 g as a single bolus (repeat if indicated)

YES

HYPERFIBRINOLYSIS

DONE NO

A5 FIBTEM < 12 mm YES

Fibrinogen concentrate or cryoprecipirate (dose calculation according to the table3) Target: A5 FIBTEM> 16 mm

CLOT FIRMNESS EVALUATION

NO A5 EXTEM < 35mm And A5 FIBTEM > 12 mm

Platelet concentrate 1 pooled oraphersis YES

NO

CT EXTEM < 80 s And A5 FIB ≥ 12 mm

YES

4F-PCC 10-15 IU/kg bw Or FFP 10-15 ml/kg bw (FFP may dibute fibrinogen)

Consider FFP 10 ml/kg bw Or 90 µg/kg rFVIIa

NO

CT INTEM < 240 s YES

CT INTEM / CT HEPTEM ≥1.25

YES

NO

Ongoing bleeding YES

THROMBIN GENERATION EVALUATION

Consider protamine 25-50 mg (2.5-5 ml)

Re-check after 10-15 min using a new blood sample

CLINICAL AND POINT OF CARE REASSESSMENT

Fig. 43.3  ROTEM guided transfusion algorithm

• Clinical bleeding evaluation: check for basic conditions (Temp. >35  °C; pH  >7.3; Cai2+  >1  mmol/L; Hb ≥7 g/dL).

• Hyper fibrinolytic stage: apply antifibrinolytic therapy [29, 42–46] Prophylactic administration of Tranexamic acid can be given within 3 h after trauma, delivery or any

43  Transfusional Optimization Using Viscoelastic Test Guided Therapy in Major Obstetric Hemorrhage: Simulation and Skills

•

•

•

•

kind of major obstetric bleeding [29, 30, 44]. CTFIB >600 s represents a flat-line in FIBTEM. Clot firmness evaluation: according to FIBTEM A5 results give fibrinogen using the dose calculation (Table 43.4): Fibrinogen dose (g)  =  targeted increase in A5FIB (mm)  ×  body weight (kg)/160. Correction factor (140– 160 mm kg/g) depends on the actual plasma volume. 10  U Cryoprecipitate ≈ 2  g Fibrinogen concentrate. Fibrinogen dose calculation is based on the targeted increase in FIBTEM A5 (A10) in mm. In case of severe bleeding, high plasma volume (e.g., in pregnancy, significant hemodilution, or TACO) and/or factor XIII deficiency, the achieved increase in FIBTEM A5 (A10) may be lower than the calculated increase. –– Platelet concentrate transfusion: Check platelet function with ROTEM platelet (ADPTEM and TRAPTEM) or Multiplate, if available. –– Consider compensation by increased A5FIB  ≥12  mm. Consider TXA (25  mg/kg) and/or desmopressin (DDAVP; 0.3  μg/kg) in patients with dual antiplatelet therapy and/or ADPTEM 5 cm, could be a common symptom seen mostly in the second and third trimesters (Fig. 64.58). In such a period, the possible increase in size happened during the first trimester adds up to the larger volume of pregnant uterus, causing myoma impaction and thus pain (Fig. 64.59), less or more localized up to involve the entire abdomen, often with a gradual onset, through a mechanical process [81, 82]. A pain, which is more severe, with a rapid onset, constant over time, basically much more localized, is instead experienced by pregnant women affected by myoma undergone so called “red degeneration” [80]. This latter is the most common form of degeneration of myoma in preg-

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

1007

Fig. 64.59  Transabdominal ultrasonographic scan of subserosal/intramural anterior fibroid of the body, located under the bladder, in gravida at 26  weeks of pregnancy. The location of the fibroid causes marked pain in the paravesical area

nancy and owes its name to the characteristically hemorrhagic appearance, which fibroids assumes because of the thrombosis within the periphery of the lesion or because of the rupture of intralesional artery [82–84]. In particular during pregnancy, both uterus growth and myoma growth led to a reduction of blood supply to the lesion due to kinking of uterine blood vessels and insufficient compensatory perilesional angiogenesis, respectively, causing ultimately anoxia, ischemia, and necrosis and thus release of pain mediators, such as prostaglandins [82–84]. In this case ultrasound examination helps to detect the characteristic heterogeneous echogenic pattern with or without cystic area, which is typical of myoma red degeneration and which is found in up to 70% of pregnant women affected by fibroid with acute abdominal pain [75]. It is very important to do a comprehensive physical examination, because the presence of abdominal tenderness without any signs of peritonism and normal vital signs makes rational to proceed with caution, procrastinating or totally excluding a surgical treatment. In a stable pregnant woman with this latter clinical condition should be wiser to adopt a conservative management, which consists in hospitalization until the symptoms disappear followed by a protected discharge [76]. In this circumstance the pain relief is guaranteed through a supportive analgesia therapy, chosen by a pain specialist team in order to not harm fetus wellness. As Royal Collage Obstetricians and Gynecologists (RCOG) in agreement with Medicines and Healthcare products Regulatory Agency (MHRA) and European Medicines Agency (EMA) specifies that any pain relief medication should be taken at the lowest effective dose and for the shortest possible duration, choosing the safer drug classes for pregnancy and, if possible, ­trying to avoid them up to

Fig. 64.60  Multiloculate fibroma of 12  cm in diameter, removed in laparotomy, with areas of necrotic degeneration

12  weeks of pregnancy. Paracetamol has the best safety profile in pregnancy and it is the first-line medical treatment option, but, in case of poorly responsive abdominal pain. It is possible to employ opioids, instead Non-steroidal anti-inflammatory drugs (NSAIDs) are absolutely contraindicated after 30 weeks and highly discouraged before [85]. Analgesic therapy could be associated, as reported in several case reports in literature, with antibiotic, antispastic, and tocolytic drugs in order to improve maternal and fetal outcome. In particular, a prophylactic use of progesterone or tocolytics, even with a relaxed uterus on palpation of the fundus during clinical examination, seems to contrast the uterine irritability and contractility and so activation due to all the mediators and/or the myoma-related mechanical effect [79]. However, sometimes the conservative management cannot be the first-line treatment option and cannot be enough for symptomatic pregnant women with degenerated myoma as happened in cases of severe acute abdomen. The evidence of peritonism signs, like a distinctly positive Blumberg sign, and/or abnormal vital parameters are been reported in literature in cases of massive hemoperitoneum linked to an actively bleeding ruptured vessels in the contest of myoma in necrotic degeneration (Fig. 64.60) [86]. A

1008

A. Tinelli et al.

Fig. 64.61  Pedunculated fibroma of 18 cm in diameter, twisted on its axis, removed in laparotomy

diagnostic and eventually operative surgical procedure results mandatory and a wise option for these, however rare, case of heavily bleeding myoma during pregnancy with intra-abdominal hemorrhage [86]. A similar clinical presentation of acute abdominal pain even with Blumberg positive sign could be also due to myoma torsion during pregnancy (Fig. 64.61). This eventuality typically happens when the peduncle of a subserosal fibroid turns on its own axis (Fig.  64.62), causing the compression of its venous vessels first and its arterial vessels then and thus hindering the blood support to the myoma, which therefore undergoes ischemia and necrosis. The mechanisms of pain are the same illustrated above and the entire process gradually led to reactive peritonitis with the corresponding clinical findings. The diagnosis of myoma torsion could be reach, when possible, through ultrasound scanning, which shows a uterine lesion sensibly bigger than the previous controls and a positive whirlpool sign at the PD (Power Doppler), meant as a spiral-like pattern given by the characteristically torsion of the peduncle. The clinical presentation is the key of interpretation for ultrasound findings, which are poorly sensitive and specific in myoma torsion diagnosis, which have to be confirmed intraoperative, mostly during pregnancy when the larger pregnant uterus can hamper imaging ultrasound view [87]. However, uterine fibroid torsion dur-

Fig. 64.62  Subserous/pedunculated fibroma of 14  cm in diameter removed in laparotomy for worsening pain, caused by the compression of its venous vessels first and its arterial vessels

ing pregnancy not always represents a surgical emergency and in the first place deserves the attempt to conservative management [88], as the treatment strategy mentioned above. Less frequent but still to be mended are other types of myoma degeneration, which could happen during pregnancy causing abdominal discomfort and pain, like the cystic and the hyaline degeneration. These latter differ from red degeneration by a histopathology point of view, but the mechanisms at the base of the painful symptoms are the same, as explained above [89]. An extremely rare occurrence is instead the finding during pregnancy of a suppurative leiomyoma, also called pyomyoma, whose clinical presentation, as described by Kobayashi et al. in the only case published in literature, is characterized by the increase of inflammatory markers (white blood cells, c-reactive protein, procalcitonin, etc.) and concomitant peritonitis [90]. Moreover, as reported in several case reports in literature, all these complications are more likely to happen and to be clinically severe when more than one fibroid are present (Fig.  64.63) and the gravid

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

1009

Fig. 64.63  Uterus strewn with numerous fibroids, located in every part of the uterus and of various sizes Fig. 64.65  Large fibroid of about 16 cm in diameter that occupies the entire uterine volume and grows progressively on the uterine fundus, causing intense compressive pain symptoms

Fig. 64.64  Large progressively growing intramural/subserosal fundal uterine fibroid, removed for pain resistant to drug therapy

uterus are a polymyomatous nature. The coexisting increasing size of uterus more likely and sooner could indirectly cause myomas complications, due to the less space and the less vascular supply available to the growing lesions (Fig. 64.64) [88]. Considering the lack of high-quality evidences on myoma complication management in pregnancy, according to the opinion of the authors of this chapter, a basic role has in all

Fig. 64.66  Huge uterine fibroid removed urgently for acute abdomen and pain refractory to any medical therapy

these cases a good clinical evaluation, the capacity to discriminate between an abdominal discomfort in stable patient, who is candidate to a clinical and instrumental follow-up, and an acute abdomen (Fig.  64.65) with increasing peritonism and abnormal vital parameter, which represents a real surgical emergency (Fig. 64.66).

1010

64.11 Myoma-Related Obstetric Complications During Pregnancy Uterine fibroids can harm pregnancy by worsening obstetric outcomes. It has been estimated that in a 10–30% of pregnant women affected by myoma occur obstetric complication attributable to the benign lesion itself [82]. In this regard it is appropriate to specify that the findings reported in literature refer to an extreme variability of obstetric outcomes, in low-quality studies often burdened by several bias, confounding variables, and too wide inclusion criteria. All of these elements make only hypothetical and poorly found the relationship between myoma and obstetric outcomes, as supported by the lack of a defined pathogenetic link. Some authors claim that the common thread among the several obstetric complications would be the myoma ability to promote a premature activation of uterus by biochemical factors in association with the mechanical distortion and c­ ompression of myometrium and so of uterine cavity and gestational chamber [91]. Among the possible obstetric complications myoma would predispose to: • early miscarriage, 14% versus 7.6% of controls [65], mostly in women with multiple, corporal and intramural or submucosal fibroids, through a mechanical effect and a reduction placental blood supply [79]; • preterm labor and premature rupture of membranes, 16% vs. 8.7% and 10.8%, respectively, of controls [46], more likely when multiple fibroids are contacting placenta, but this relationship has been strongly questioned recently by several experts [46]; • placental abruption, the risk is threefold increased, mostly with retroplacental myoma, which would cause a reduction of placenta blood flow supply with a consequent decidual necrosis in the placental area overlaying the fibroid [81]; • placenta previa, the risk is twofold increased regardless of previous surgery [92]; • fetal growth restriction, not yet a clear correlation, but the risk seems to be slightly increased, although there are not high-quality evidences [93]; • fetal deformities, an extremely rare event caused by the effect of large myoma-related uterine cavity distortion on fetal skull (dolichocephaly), neck (torticollis), or limb (reduction defects) [94, 95]. In addition to that myomas can promote antenatal hemorrhage, which have to be meant not only as linked to myoma degeneration but also as a worsening of bleeding linked to other obstetric causes. In case of placenta abruption, placenta previa or miscarriage, all causes of antenatal bleeding mostly in third trimester, the coexistent presence of myoma (especially if multiple, large, intramural, and/or underlying

A. Tinelli et al.

decidua), could hinder myometrium contraction and its hemostatic effect, worsening hemorrhage [46, 79, 81]. However, a series of studies in the last decades attempted to quantify the real burden of myomas in relation to adverse obstetric events and in spite of the variety of complications it would appear to be involved in, many authors, such as Klatsky et al., Poovathi and Ramalingam, and Saleh, more recently, claimed that fibroids, in a general and comprehensive view of all the obstetric adverse events mentioned above, do not significantly worse obstetric outcomes either maternal or neonatal [46, 96, 97] and this also applies to twin pregnancies, as reported by Stout et al. [98]. Therefore, these obstetric adverse effects have to be kept in mind by specialist, but generally myomas do not need a treatment in pregnancy because of them.

64.12 Myomectomy During Pregnancy As said at the beginning of this chapter the treatment of fibroids can be pharmacological and surgical, depending on the impact of the fibroids on the woman and depending on the severity of the problem. This is partly true for pregnant women; as a matter of fact in this specific pool of patient the treatment of fibroids has to be done only in certain cases and in particular when the lesion becomes the cause of complications during pregnancy, becoming the patient symptomatic up to need of hospitalization. Moreover, the pharmacological treatment existing for myomas cannot be used in pregnant women and the options available in pregnancy are, first of all, the conservative management (as described above) and, if it is necessary according to the clinical conditions, the surgical approach. In both cases it is suggested to consider the pregnancy at risk and to plan a close follow-up in time till delivery, including clinical evaluation and sonography in order to detect any lesion’s changes and/or significant alterations of fetal biophysical profile and Doppler assessment [88]. Focusing on surgical treatment, all the cases in which the specialist decides to bring the pregnant patient to the operating room are substantially emergency situations or very close to being so, therefore is an oxymoron that to date the myomectomy during pregnancy still represents an open debate, not yet well defined in all its aspects. The past preconception that a myomectomy during pregnancy always represented a danger to be avoided in every way in the obstetric patient, also for the limited number of cases reported and a lack of high-quality evidences in literature. Moreover, the absence of worldwide guidelines on myomectomy during pregnancy limited the small number of surgeons with an adequate training and skill to do the procedure at the best and so with the lowest probability of complications. As a matter of fact, almost all of the publications on myomectomy in pregnancy refer to case report [88, 99], whose analysis is

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

burdened by three main problems: (a) the indefinite categorization of the lesions preoperatively and postoperatively; (b) the heterogeneity data related to the description of the surgical procedure done; and (c) the missing and unreported data [88, 99]. In the large part of the cases reported the relation of myoma removed with placenta is not specified, and it would have been important in order to stratify the possibility to occur obstetric complications after surgery, which, as mentioned in this chapter, seem to be more associated with fibroids in contact with placental and mostly the retroplacental ones [46, 99]. Since the beginning of ultrasound era, which allowed to promote a better detection and characterization of fibroids in pregnancy (Figs. 64.67 and 64.68) and overall a better risk stratification of patients for a more

Fig. 64.67  Small subserosal fibroma of the uterine body occasionally found during transvaginal ultrasound at 9 weeks

1011

appropriate therapeutic choice, only two studies have been published: the retrospective one of De Caroli et al. [83] and the prospective one of Lolis et al. [89], with a sample, respectively, of 18 and 13 pregnant women affected by myoma and undergone to its removal during pregnancy. For that reason, all the consideration we can do to date have to be extrapolated only from these two studies and the series of case reports with all the limits associated. As general principle what emerges from the literature is that the surgical treatment has to be a choice also for pregnant women and the procedure have to be done in order to improve mother condition without further compromising her safety and minimizing fetal risk at all [100]. Surely, once the indication for intervention has been established, the first choice to do is the abdominal access mode to perform myomectomy: laparoscopy or laparotomy? The less invasive surgical access is obviously the laparoscopic one, but in case of pregnancy this choice cannot be done always and, in the past, it was strongly avoided and discouraged. However, the emerging evidences in the last two decades reveal the safety and feasibility of a laparoscopic myomectomy and in general of the laparoscopy in pregnancy if it pays attention to the gestational period and to the technique to execute it. Considering that the first non-­ obstetric causes of severe abdominal pain in pregnancy are acute appendicitis and acute cholecystitis typically gallstones related, most of the many of the current evidences derive from the work of general surgeons as well as gynecologists [101, 102]. In fact, in 2017 American Gastrointestinal and Endoscopic Surgeon (SAGE) presented the first indications [100] and then in 2019 the union of British Society for Gynecological Endoscopy (BSGE) and Royal College of Obstetricians & Gynecologists (RCOG) finally published the definitive “Evidence-Based Guidelines on Laparoscopy in Pregnancy,” without an explicit reference to myomectomy but still applicable to this intervention [103]. These guidelines state that laparoscopy during pregnancy: • has to be preferred to open routes of surgery according to available expertise, infrastructure, background history, gestation, and the women preference; • guarantees a faster recovery, a shorter hospitalization, and a lower rate of wound infection rather than laparotomy; • has to be done in general anesthesia with endotracheal intubation; laryngeal mask airway is no recommended; and • could be feasible at any gestational period [103].

Fig. 64.68  Intramural fibroma of the uterine body found occasionally during a transvaginal ultrasound at 5 weeks

This latter indication could be questionable according to some expert point of view because of the augmentation of the mechanical obstacles faced in the third trimester, when the uterus occupies the largest part of abdominal cavity,

1012

making practically almost impossible to set up a laparoscopic access [101]. Spyropoulou et al., in fact, highlight that almost all laparoscopic myomectomies were performed in the early part of gestation and no later than the end of second trimester [99]. Surgeons should be aware of specific technical differences that are essential to make laparoscopic surgery easier and safer in pregnant patients, as it will be specified immediately below. In a way that is directly proportional to the gestational period the enlargement of the uterus determines a reduction of visual field, an obstacle to the operator’s dexterity, and thus a major risk of vascular and organ trauma interesting not only the uterus, up to causing its perforation, but also the other abdominal organs [100, 103]. Therefore, with the aim to facilitate the operator in carrying out the procedure by obtaining more space and visibility, the port location should be modulating in relation to the level of the uterine fundus, which can be evaluated preoperatively by palpation or transabdominal ultrasound. Substantially the more weeks of gestation, the larger is the uterus, the higher the ports must be positioned upper on the abdominal quadrants rather than in the traditional position. The primary port could be placed at umbilical point as usual, but, according to the case in exam, also at supraumbilical and subxiphoid point, also known as Lee-Wang point, or alternatively at Palmer’s point, which is at the left upper quadrant in the mid-­ clavicular line. These latter two and the corresponding one on the right of Palmer’s point are suggested in the end of the second trimester and in the third trimester [100, 101, 103]. Likewise, the secondary port, which is traditionally placed in the left and right iliac fossa, can be located upper along the same longitudinal lateral line for each side [100, 103]. There is no increased risk of herniation at the port site rather than non-pregnant women and just like in these latter it is mandatory to close the fascia when the port site is major than 10 mm [103]. The only entry technique considered feasible and safe for every period of pregnancy is the Hasson’s one, which in pregnant patient can be done inside or above the umbilicus, however almost 3–4 cm above the uterine fundus [100, 103]. Veress needle is discouraged in pregnancy because of the increased risk of intestine, aorta or uterus injury. In fact, in literature as reported a rate of abdominal entry-related complication in pregnant patients of 2.8% in Veress’ group versus 0% in Hasson’s group [104]. A safe alternative entry technique could be the direct ore glassless entry approach, nevertheless to date there are no enough data on the second part of the gestational period and the operator has to be confidential with it [103]. The creation of pneumoperitoneum should be gradual and the carbon dioxide (CO2) insufflation pressure should not exceed 10–12  mmHg in order to contain the risk of fetal acidosis [100, 104]. About that also an intraoperative CO2 monitoring by capnography is suggested [100]. Melgrati et  al. described the only one case of isobaric laparoscopic myomectomy during preg-

A. Tinelli et al.

nancy in order to avoid the adverse effects of CO2 [105]. Moreover, considering that the enlarged pregnant uterus can more likely compresses the vena cava in supine decubitus, expert opinion suggests a left lateral tilt of pregnant patient up to 15° and 30° during the laparoscopic procedure. In this way the uterine blood supply is guaranteed and so the placental one, avoiding fetal hypoxia and risk of premature labor [106]. This measure can be considered also in laparotomic myomectomy. In lights of all this laparoscopy should be consider the standard care option in pregnancy; nevertheless, Spyropoulou et  al. have found that of all myomectomies, which were performed in pregnant women and reported in literature, as much as 78.4% were performed laparotomically [99]. Most likely that decision might be explained by the fact that many surgeons, mostly in the past, have greater confidence, experience, and surgical skill with laparotomy and this represents a wise way to regardless minimize surgical complications. Therefore, the present and future goal should be a good and complete surgical training, which will give the possibility to have much more surgeons experienced in both laparotomy and laparoscopy on pregnant women and thus to offer, even in an emergency situation, both options for each patient. As regard myoma removal, the pre-operatory evaluation of the lesion is at the base to forecast the complexity of procedure and the best technique to adopt. The cases reported in literature evidences that there are not specific characteristics prohibitive for the removing, but by and large the majority of the myomectomies during pregnancy have been executed on a single lesion (76.28%), with a median largest diameter of 13  cm, subserous, mostly pedunculated, and located at the uterine fundus [99]. Surely among the several myoma characteristics what really is condition the grade of difficulty of the procedure and the eventual complications related is the intramural component of the fibroid. Not by chance the best peri- and postoperative outcomes have been reported by myomectomy during pregnancy in case of pedunculated myoma [89, 107] and in fact at the beginning of the 1990s the first criteria for myomectomy during pregnancy suggested by Burton et  al. and also Exacoustos and Rosati tended to limited the procedure to this type of fibroid or at most to the subserosal one [81, 108]. The removal of pedunculated myoma is performed first through the identification of the base of the peduncle and consequently through its section, having previously strangled with a no-needle suture or with diathermocoagulation by bipolar forceps or monopolar hook or “caught” by endoloop. Usually, the remaining breach on the uterine serous is so small and superficial that no need any closure [109, 110]. On the contrary, in case of serous sessile myoma or myoma with intramural component the surgical approach is different. Unfortunately, the technique employed for myomectomy is not specified in the largest part of the cases published and these are big missing data in

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

Fig. 64.69  Intracapsular laparotomy myomectomy with preservation of the myoma pseudocapsule, by pulling it down with bipolar scissors during myoma enucleation

order to evaluate the occurrence of complication related to the surgical procedure performed. However, in order to respect the innovative findings on the biological relevance of myoma pseudocapsule (evaluated by Tinelli et al.), myomectomy should be always performed by the intracapsular approach [111]. The myoma enucleation with the contextual preservation of the surrounding pseudocapsule should be performed similarly to what has already been described in this chapter for nonpregnant patient patients (Figs.  64.69 and 64.70). Nevertheless, the pseudocapsule role in physiological wound healing is more significant on a gravid uterus, enlarged and supplied by a larger amount of blood and, consequently, at risk for possible scar dehiscence and bleeding during uterine contractions in labor. Moreover, the intracapsular myomectomy technique is found on the idea to limit as far as possible the use of electrified devices, promoting a cold enucleation of fibroids (Fig. 64.71). This allows to preserve not only the integrity of the pseudocapsule but also the surrounding myometrium (Fig. 64.72), which could be collaterally damaged by electrosurgical tools, causing an increased risk of bleeding and/or dystocic labor. Anyhow to date there are not absolute contraindications to the use in

1013

Fig. 64.70  Intracapsular laparotomy myomectomy with preservation of the myoma pseudocapsule, by sectioning it with bipolar scissors, during myoma enucleation

Fig. 64.71  Fundal fibroid enucleation by cutting the pseudocapsule with cold blade scissors

pregnancy of electrosurgical devices, although, as the data reported in literature and expert opinion suggest, it should prefer bipolar device rather than monopolar ones in contact

1014

Fig. 64.72  Cutting the pseudocapsule with cold blade scissors allows the surrounding myometrium to be preserved from damage from electrocoagulation

with uterine wall, in order to avoid the risk of resistive heating and thermal damage on it [83, 99, 100, 103]. One of the most complicated postoperative courses has been described in a case of laparoscopic myoma removal during pregnancy employing monopolar diathermy: the patient underwent to an abdominal laparotomy because of the onset of myometrial necrosis with consequent abscess formation, uterine dehiscence with procidence of the amniotic sac from the breach [112]. No indications exist on the best way to reconstruct uterine wall, but in the largest part of the cases it was performed a double-layered sutures with Vicryl [99], instead the monofilament suture has been reported in only two cases, with PDS [55] and Monocryl [113]. No data in pregnant women undergone myomectomy are available on the use of barbed suture, which on the contrary is more and more employed and suggested for myometrium breach closure in non-pregnant women. Myoma extraction in case of laparoscopy have to be done in a safe way because of the impossibility to exclude an occulted sarcoma without a histological evaluation. In fact, the open (uncontained) morcellation in abdominal cavity of an occulted sarcoma drastically worsens the prognosis of the patient. Therefore, in order to avoid the spread of unsuspected sarcoma the lesion removed has to be extract previous of containment system, as an endobag, which, once closed, is pushed out of the trocar breach and then opened outside of the abdomen; after a hand-assisted morcellation with cold blade and inside endobag the specimen fragments are recovered. Anyway, Food and Drug Administration (FDA) has highly discouraged the

A. Tinelli et al.

use of laparoscopic power morcellator for myomectomy [114]. No data are available on “safer” containment system plus power morcellation for myomectomy in pregnant women, but it could be cumbersome and risky in an abdominal cavity with an enlarged gravid uterus. Intraoperative antibiotic prophylaxis is indicated, opting for drug classes which are safe for fetus, like broad spectrum beta lactam (i.e., ampicillin) or cephalosporins (i.e., cefazolin), and, in case of allergy, macrolides (i.e., azithromycin, erythromycin) [99, 100, 103]. As a matter of fact, the risk, albeit quite rare, of developing a chorioamnionitis with eventual consequent miscarriage and/or preterm labor has to always keep in mind [83]. The added risk for venous thromboembolism present in pregnancy in consideration of the typical hypercoagulability state makes mandatory a prophylaxis after surgery using anticoagulant, like enoxaparin or seleparin, whose dosage is modulated in relation with a preoperative stratification of risk for each patient [103, 115]. Myomectomy is one of the gynecological procedures at major risk of bleeding in the imminent postoperative time and so anticoagulant treatment should be started 12 h after surgery and, according to author’s opinion, after evaluation of hemoglobin and clinical condition. In all the great majority of the myomectomy during pregnancy reported had a regular postoperative course with an uneventful remaining course of pregnancy. The median intraoperative blood loss ranges around 200–400 cm3 and only a minority of women, which is approximately inferior to 5%, needed blood transfusion [99]. The cases of miscarriage are few, inferior to 5%, and almost all after multiple myomectomy [81, 89] or in particular case, as the fetal loss described by Kilpatrick et al. due to a premature rupture of membrane 2  weeks later vaginal removal of a submucosal myoma [116]. Finally, it is important to note and at the same time to inform the patient preoperative that it is far more likely to give birth by CS rather than vaginal route, which would expose to the risk of uterine rupture during labor because of the shortest interval available to wound healing. This evidence has been represented a constant over time and it is not affected by the surgical technique employed, which seems not to change the occurrence of cesarean delivery [107, 117]. Generally, in light of what is reported by the most recent data, it is confirmed that for women affected by myoma the CS has a definitely higher prevalence rather than the vaginal delivery (VD), 58.76% versus 29.89%, respectively. This discrepancy correlates in a way that is directly proportional to the number of myomas removed, having in the case of single myomectomy and one in the case of multiple myomectomy, being been reported 51.35% CS versus 36.48% VD after a single myomectomy and, instead, 82.6% versus 8.69% in the multiple myomectomy sample. Moreover, a good 44.32% of these CSs were elective, in particular 37% of the patients of the single myomectomy group and 65% of the multiple myomectomy group [99].

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

64.13 Myoma-Related Obstetric Complications Affecting Childbirth and Postpartum Uterine fibroids can condition the time and way to carry out the birth, leading the specialist to plan a correct strategy in order to guarantee the best assistance to the mother and her baby, reaching the best perinatal outcomes for each other [96]. The main obstetric complication reported in literature in case of women affected by one or more uterine myomas are the following: • fetal malpresentations, mostly breech presentation, 13% versus 4.5% of controls [46, 93], mostly in case of large, multiple and/or located in the lower segment myomas [79]; • labor dystocia, twofold increased risk [46, 93]; • CS, risk increased nearly four times [79]; • postpartum hemorrhage, 2.5% versus 1.4% of controls [79]; • retained placenta, infrequent eventuality to be differentiated at the ultrasound scanning from the myoma, which instead more frequently slightly prolapses in uterine cavity after delivery [118]. The distortion of the uterine cavity shape and the reduced distensibility of uterine walls can consider the mechanical effects through which fibroids promote the onset of fetal malpresentation or abnormal labor. Instead, the increased occurrence of postpartum massive blood loss should be explained by the characteristically pattern growth of fibroids, which develop inside myometrium, centrifugally displacing the surrounding muscle fibers (Fig. 64.73). In this way the subverted architecture of the myometrium does not manage to efficiently contract after delivery and so also the hemostatic effect is reduced [119]. This also explained the slightly increased risk of puerperal hysterectomy as hemostatic procedure in patient affected with myoma [79]. Moreover, one of the main etiopathogenetic factors common to dysfunctional labor, postpartum hemorrhage, and placental retention is a dysfunctional

1015

retroplacental myometrium, meant as a muscular tissue incapable to contract adequately in terms of power, duration, frequency, and coordination; thus, it is easy to understand how the presence of a myoma in contact with placenta and/or underlying placenta can promote these complications in the same way [118, 120]. More scientific and high-quality data are needed to better investigate this relationship and to find the clinical-therapeutic implications. However, according to all these eventualities mentioned it is evident that the presence of uterine fibroids could significantly reduce the chance of the pregnant women to give birth to their baby with an uncomplicated VD; in fact both fetal malpresentations, like breech, transverse, or oblique presentation, and obstructed labor represent an indication to cesarean delivery. Anyway, a pregnant woman affected by a myoma, even if large (maximum diameter >5 cm), has not to date an absolute contraindication to a trial of labor [79, 92, 93]. Overall, fibroids could influence childbirth in different ways. In example, just about that, another interesting and relative rare fibroid-related condition affecting delivery is worth mentioning, the myoma previa, a term which indicates a lesion located at the isthmus level immediately behind the internal uterine orifice, so as to determine an obstacle to access to the birth canal. It this case the cesarean delivery is mandatory and, as reported in literature, operators usually choose to perform a corporeal hysterotomy upper than the site of fibroid, leaving this latter in place [121]. No absolute indications exist to date in reference to the technique to employ to perform a cesarean removal of a previa myoma; in that case it should be considered that the hysterotomy performed at the isthmic level and so at the site of myoma would be used for the myomectomy itself as well as for the delivery. In that way the healing process and the patient recovery would be easier, faster, and better rather than performing two separate uterine incision sites during cesarean myomectomy or also procrastinating myomectomy after CS. Although as early as in the 1990s Michalas et al. reported an interesting successful and uncomplicated cesarean myomectomy with the removal of eight fibroids obstructing the lower part of the uterus [122], new and more consistent evidences on this management issue are actually needed in order to establish the better conduct to be adopt.

64.14 Cesarean Myomectomy Rationale and Technique

Fig. 64.73  Cesarean intracapsular myomectomy, showing pseudocapsule fibers clamped by surgical forceps

In all cases in which, for the above reasons and for any obstetric event that foresees it in the guideline, a caesarean section is arranged for a patient suffering from myoma, nowadays it is essential at least to consider the option of performing simultaneously the cesarean delivery and also the removal the myoma itself [47]. This surgical procedure is known as cesarean myomectomy and nevertheless has been discouraged and avoided in the past because of misleading preconceptions and opinions;

1016

to date its feasibility is being demonstrating day by day by the increasing data reported literature and more and more sustained [47]. But why should we consider cesarean myomectomy safe and why should it be preferred to a simple CS followed by a planned myoma removing after the complete uterus involution? The underlying rationale is simple and intuitive. The detractors of this surgical procedure support their ideas substantially in ­relation to two considerations. First of all, they remarked that, because it has been demonstrated that myomas are dynamic lesions which use to constantly dislocate muscle fibers in order to move toward points of least resistance, such as the abdominal cavity or uterine cavity, immediately after delivery the enlarged empty uterine cavity make myoma with intramural component displace. Adding to it there is the trend of myoma to shrink in concomitance with the uterine involution, which is complete after 6 months, up to nullifying the need of any treatment. The second issue is the most conditioning in the rejected for long-time cesarean myomectomy and it is the assumption of a constant and significant increase of risk of massive bleeding associated to this procedure due to the physiological implementation of uterine vascularization during pregnancy, which adds to the basic bleeding risk of a simple myomectomy. For long time, this fear was supported by some reports of literature on the first cases of cesarean myomectomy, which were not performed according to the technical measures known to date in order to minimize perioperative complication; thus, also cases of hemostatic hysterectomies were executed contextually [108, 123, 124]. Considering these reasons, the only case in which myoma removing was allowed was when the lesion was pedunculated subserous and “with a small peduncle,” as “Te Linde” stated, because the risk of myometrium intraoperative damage or postoperative dysfunction and bleeding was almost null and this type of fibroids does not tend to shrink at all after pregnancy [123]. Currently, recent scientific evidences highlight that the natural history of fibroids is not indeed characterized by significant changes in size during and after pregnancy [60, 71, 73, 76, 78]. Moreover, till now many are the increasing cases reported of successful cesarean myomectomy with an eventful perioperative period. Nine are the main and more recent studies published in literature, conducted on a large overall sample of patients, in fact, and they all agree in terms of results that the myomectomy contextually CS exposes the patient to a more likely bleeding and surgeon had to always keep in mind it. However, at the same time the same results clarify that this risk is only slightly increased, not statistically significant, and thus does not implies any relevant negative impact on maternal or fetal outcome or in general any detriment to the patient’s general condition [47, 125]. The same consideration has to be done for the slightly lengthening of operating times and duration of hospitalization in case of caesarean myomectomy rather than caesarean delivery alone, on average 5–11 additional minutes and 0.5–1 additional day respectively, which, however, does not interfere with the good

A. Tinelli et al.

clinical course and with the general well-being of the patient [125–127]. No additional risks in terms of wound infections, increased inflammation indices, or sepsis are observed in patients undergoing caesarean myomectomy [126, 127]. During this procedure, we remind you, like what happens in caesarean section or in short gynecological surgeries, an intraoperative prophylactic antibiotic therapy is always indicated as mentioned previously in this chapter. A general consideration, which emerges from the most recent data, is that cesarean myomectomy is a safe and feasible procedure, but this is not true regardless of everything; it is mandatory to evaluate several variables which influence the results of the intervention [47], as follows: • • • •

the patient’s characteristics; the myoma characteristics; the surgical technique employed to remove myoma; the operator experiences.

Many authors state that the main and independent variable which is correlate to the safety of cesarean myomectomy more than any other one is the gynecologist’s surgical ability and skills [125, 126, 128]. Surely to date there are not many surgeons confident with this type of procedure, thus more training is needed for the future, but the fact that the large part of the CS during which myomas are removed are elective can represent an opportunity to better plan the surgery, choosing almost one operator with the with the experience necessary to guarantee the best results while minimizing risks [71]. In the case of an emergency–urgency caesarean section, if there are not permissive conditions, we have to remind that cesarean myomectomy is not an obligatory choice but should be a choice in a safe background, always providing the well-being of the mother and baby the main goal. This is true also during the preoperative evaluation, which is also fundamental in terms of stratification of perioperative risk cesarean myomectomy-related in relation to the patient and myoma characteristics. As a matter of fact, inherited, congenital, or acquired disorders by hemostasis, which implies a major risk of bleeding, would make patient unfit for myoma removal during CSs, because it could be associated with a major risk of uncontrollable bleeding, blood transfusion, and also hysterectomy [108, 129]. Nevertheless, this could be partly questionable if we consider that the same additional risk, however, would add to the risk of bleeding related to leaving fibroid in situ after a caesarean section, as mentioned above. On the other hand, also the myoma characteristics could be of some importance in perioperative bleeding, because they are linked to the complexity of the surgical procedure. Data on the role of fibroid size are controversial, if Zaho et al. correlated the augmentation risk of bleeding to the removing during CS of a myoma larger than 5 cm [130], the more recent studies prove that there are not any statistically significant differences in blood loss in relation to any myoma size [131, 132]. In particular idea that leaving in situ fibroid during

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

CS is always a safe choice has been unhinged by Dedes et al. study, which found that in case of fibroids larger than 5  cm there is a similar higher ­incidence of massive blood loss regardless of whether the myoma was removed during the CS or not [133]. The same study highlights the higher incidence of hemorrhage, although with the low threshold of 500 mL, in case of removing of multiple fibroids during CS rather than a single one (OR 4.7% versus 1.4%, respectively) [133]. There are not available other studies, mostly prospective ones, on the bleeding risk in multiple cesarean myomectomies, which still remains a field of research. Definitely more investigated is, instead, the impact of myoma relationship with uterine wall on safety of cesarean myomectomy. Established for some time that the removal of a subserous sessile and, mostly, pedunculated myoma is safe and feasible, it took many years to reach the same conclusion as regards intramural lesions. The hypothesis sustained for long time was that the deeper and larger the intramural component of the myoma was, the deeper and larger the breach would be and therefore the more complex the repair process with a consequent greater probability of atony and peripartum hemorrhage. On the contrary the data reported by all of the observational studies available to date demonstrate that this correlation is not statistically proved, definitively showing the feasibility cesarean myomectomy also in intramural lesions [134]. Therefore, a key role has the surgical technique regardless myoma’s characteristics, although the procedure has to been always tailored on each patient [47, 131]. Although there are not guidelines and official indications shared worldwide, in this chapter we try to give some technical tips on how to perform at best cesarean myomectomy based on the evidences of the scientific literature, focusing on the additional precautions which make this procedure different both from a myomectomy and from a cesarean section. First of all, a common practice is the postponing of the fibroid removal after the extraction of the baby and the placental expulsion (Fig. 64.74) with the aim to guarantee better neonatal outcomes and also to eventually face first complication related to placenta without getting worse maternal outcomes [47]. This could be difficult to respect in case of large myoma previa of the anterior uterine wall, as stated above. Hysterotomy for fetal extraction, even if in the first cases reported was executed longitudinally in line with the surgical habits of the past, nowadays have to be performed at the lower uterine segment like in any other traditional cesarean delivery. This choice is assuming a fundamental importance in the preliminary assessment phase for a correct preoperative planning; as a matter of fact, the aim of the surgeon should be to try to enucleate and remove all the myoma designated using the uterine incision just done for the delivery (Fig. 64.75). Fibroids’ mapping is not always feasible and a corrected preoperative mapping of the myoma help to predict the incisions on the uterine wall necessary for myomectomy and thus the risk of blood loss, who’s the number of scars is the major predictor.

1017

Fig. 64.74  Uterus extracted from the abdomen after a transverse cesarean section on the lower uterine segment. The fibroid is posterior and is addressed by the surgeon after hysterorrhaphy

Fig. 64.75  Uterus extracted from the abdomen after a transverse cesarean section on the lower uterine segment. The fibroid is left anterior pedunculated and is addressed by the surgeon after hysterorrhaphy

1018

A. Tinelli et al.

Therefore, the anterior isthmic myomas are the most eligible to cesarean myomectomy (Fig. 64.76) [134]. When an alternative incision is needed the approach depends on the fibroid location and position has to be removed. As a matter of fact, in case of serous myomas the only choice is an incision on the external face of uterine wall is performed over the site of myoma; on the contrary in case of intramural myomas, mostly when they occupy the inner side of myometrium (Fig.  64.77), this approach is not the only one possible. Recently Hatırnaz et al. proposed the “endometrial myomectomy” (Fig.  64.78), an alternative approach according to which the myoma is reached by a small transendometrial incision at the site of myoma, which is performed only after a palpatory evaluation of the location of myoma (Fig. 64.79). After removal and enucleation of fibroid, according to its inventor the uterine breach has to be sutured only if the site defect is larger than 3  cm [135]. Hence, according to this innovative approach there are not additional scars on the uterine surface besides to that of hysterotomy, which is used

Fig. 64.77  Caesarean myomectomy with removal of myoma deeply located in the anterior uterine wall

Fig. 64.78  Cesarean endometrial myomectomy. The myoma is hooked with Backhaus forceps and pulled out of the uterus

Fig. 64.76 Anterior isthmic myomas enucleated during cesarean myomectomy

for carrying out childbirth, thus significantly reduced the adhesion development and all the possible complications in the imminent and late postoperative period linked to them. The transendometrial incision should be very smaller than the eventual incision by surface needed for the same myoma removing; therefore, this significantly reduced the intraop-

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

1019

Fig. 64.80  Intracapsular caesarean myomectomy, with incision of the fibroconnective shoots of the myoma pseudocapsule during its enucleation from the uterus

Fig. 64.79  Cesarean endometrial myomectomy. The myoma is hooked with Backhaus forceps and detached from its bed, in an atraumatic way, by the surgeon’s finger

erative bleeding rather than serosal myomectomy (209 mL versus 375  mL, p  =  0.001). Moreover, the hysteroscopic inspection did not show a significant risk of development intracavity adhesions or Asherman syndrome [135]. Obviously, before establishing this approach as the reference one for intramural myoma in case of cesarean myomectomy, new studies are needed in order to clarify some doubts raised by some experts, such as the possible risk of affection of endometrial role in reproductive process or the consequent risk of abnormal placentation [136]. One of the main differences to highlight between transendometrial myomectomy and serous myomectomy during cesarean section is that in the first case the hysterorrhaphy at the lower uterine segment is done after myoma removal and eventual breach repair,

because it is used to access to the site of myoma; on the contrary according to the traditional fashion the uterine closure is always prior to myomectomy in order to minimize blood loss [134, 136]. Both in case of traditional approach and transendometrial one the myoma, as demonstrated by Huang et al. [137], has to be enucleated saving up the pseudocapsule, the surrounding structure rich in fiber, vessels, and mediators involved in intracellular signaling, which, when preserved, make more safer the procedure at all, promoting a correct, rapid, and valid myometrium wound healing by a structural and functional point of view, as stated by Tinelli et al. first in 2014 [30]. This technique, as better explained above in this chapter with all its facets, as known as “intracapsular myomectomy,” and also in case of cesarean myomectomy represent to date the gold standard in order to reach the best surgical results, reducing all the risk linked to the procedure its self [47, 111, 127]. Having first identified the correct cleavage plane between myoma and the surrounding myometrium, the “intracapsular myomectomy” during cesarean section should be performed through a sharp dissection of pseudocapsule, which allows to break the fibrous bridges anchoring the lesion to the surrounding tissue (Fig. 64.80), and, after myoma enucleation and removal, it should be accurately approximated the edges of the myometrium and completely closed the residual dead space [30] (Figs. 64.81 and 64.82, Videos 64.1 and 64.2). Moreover, the other fundamental point is just the way to repair the residual myometrial defect after myomectomy, which is essential for the correct wound healing process, because, if a hemostatic effect is not reached at this site through an adequate suture, the space continues to be gradually supplied by the surround-

1020

A. Tinelli et al.

Fig. 64.81  Video tutorial 1: The video showing a double myomectomy during a caesarean section: a fundal pedunculated myoma and an anterior intramural one. The first is removed with diathermocoagulation of the base of the myoma, with suture of the myometrial serosa, the second by intracapsular myomectomy with preservation of the pseudocapsule. Both are always removed after fetal and placenta extraction and hysterorrhaphy

Fig. 64.83  Video tutorial 3: The video showing the myoma enucleation technique with transendometrial myomectomy, with removal of a myoma from inside the uterine cavity after extraction of the fetus and placenta

Fig. 64.82  Video tutorial 2: The video highlighting the result of the caesarean section and myomectomy during the caesarean section; both sutures are properly affixed, with no muscle bleeding

ing myometrial vessels. This condition could potentially evolve to the formation of a hematoma with possible superinfection and abscessualization or to the onset of a dehiscence of the suture up to uterus rupture with all the related obstetric consequences [30, 47, 111, 127, 137]. The technique of suture the myometrium defect is not standardized to date and it can be chosen according to the surgeon’s experience. The authors of this chapter, in agreement with many expert opinions, suggest to just oppose the two sides of fibroid bed starting from the bottom through one or more layers of interrupted Vicryl suture, opting for the size of needle more adapt to the dimension of the breach. After completely closed the myometrial defect and achieved hemostasis,

in case of serous myomectomy, the superficial layer can be sutured in several manners, through an introflecting or not continuous suture or alternatively through simple X stiches [30, 47, 111, 127, 137] (Figs. 64.83, 64.84, and 64.85, Videos 64.3, 64.4, and 64.5). In conclusion we have to focus on the possible strategies existing and described in literature in order to prevent and contain the hemorrhage risk so feared performing cesarean myomectomy. In regard to the pharmacological strategy, surely the intravenous oxytocin infusion in higher dose rather than the ones indicated for the peripartum hemorrhage prophylaxis (20 UI in 500 cm3 of physiological saline solution) represents nowadays the best solution which the best statistical results [131], differently from the intra-­ pseudocapsule injection of diluted oxytocin strategy, which did not significantly reduce bleeding risk [138]. Promising according to the results published but not yet sufficiently investigated especially with regard to possible related side effects is the technique based on local infiltration of diluted adrenaline at the fibroid bed [139]. Moreover, the oxytocin infusion or administrated by intramuscular way can be suggested for the entire time interval between surgery and dis-

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy

Fig. 64.84  Video tutorial 4: Posteriorly located intramural myoma removed by transendometrial myomectomy during cesarean section, without meaningful blood loss

1021

charge according to the surgeon’s opinion and the basic risk of peripartum hemorrhage estimated for each patient. On the other hand, it always mandatory to consider among the surgical skills of the operator who perform cesarean myomectomy the ability to adopt surgical technique of bleeding control, which presupposes a good knowledge of the main anatomical landmarks. The uterine blood supply could be temporary interrupt by several techniques, for example, bilaterally positioning a tourniquet around uterine arteries or ligating or clamping them with a soft-ended instrument. In particular it is suggested that the tourniquet placement is performed passing through the broad ligament and at the cervico-­isthmic level, similarly the preventive arteries uterine ligation has not to be done at their origin, but caudally at along the ascending tract at the Mackenroth ligament level. However, Sapmaz et al. demonstrated that these two strategies do not significantly contain the perioperative blood loss, but anyway suggest the bilateral ascending uterine artery ligation mostly because of the best results reported in terms of blood loss control in the postoperative period [140]. Another possible option is the employed of electrosurgery for the myoma removal alone [141] or together with tourniquet and oxytocin [142]. In spite of the good results achieved the use of electrosurgery in relation with the possible tissue damage, as explained above, has to be always consider with caution. Concluding it is clearly evident that cesarean myomectomy could be a feasible and valid treatment option, whose safety, however, cannot be guaranteed if the operator does not correctly plan the procedure, preliminarily evaluate all the possible variable increasing the risk procedure-related, choose the best technical precautions, always in relation to the own surgical experience, skill and training and, first of all, in agreement with a well-informed patient.

References

Fig. 64.85 Video tutorial 5: A 10  cm cornually located myoma removed by transendometrial myomectomy during cesarean section without any bleeding

1. Sparic R, Mirkovic L, Malvasi A, Tinelli A. Epidemiology of uterine myomas: a review. Int J Fertil Steril. 2016;9(4):424–35. 2. Laganà AS, Vergara D, Favilli A, La Rosa VL, Tinelli A, Gerli S, Noventa M, Vitagliano A, Triolo O, Rapisarda AMC, Vitale SG.  Epigenetic and genetic landscape of uterine leiomyomas: a current view over a common gynecological disease. Arch Gynecol Obstet. 2017;296(5):855–67. 3. Tinelli A, Malvasi A, Rahimi S, Negro R, Cavallotti C, Vergara D, Vittori G, Mettler L. Myoma pseudocapsule: a distinct endocrino-­ anatomical entity in gynecological surgery. Gynecol Endocrinol. 2009;25(10):661–7. 4. Tinelli A, Sparic R, Kadija S, Babovic I, Tinelli R, Mynbaev OA, Malvasi A.  Myomas: anatomy and related issues. Minerva Ginecol. 2016;68(3):261–73.

1022 5. Malvasi A, Cavallotti C, Morroni M, Lorenzi T, Dell’Edera D, Nicolardi G, Tinelli A. Uterine fibroid pseudocapsule studied by transmission electron microscopy. Eur J Obstet Gynecol Reprod Biol. 2012;162(2):187–91. 6. Mettler L, Tinelli A, Hurst BS, Teigland CM, Sammur W, Dell’Edera D, Negro R, Gustapane S, Malvasi A. Neurovascular bundle in fibroid pseudocapsule and its neuroendocrinologic implications. Expert Rev Endocrinol Metab. 2011;6(5):715–22. 7. Walocha JA, Litwin JA, Miodonski A. Vascular system of intramural leiomyomata revealed by corrosion casting and scanning electron microscopy. Hum Reprod. 2003;18:1088–93. 8. Walsh PC.  The discovery of the cavernous nerves and development of nerve sparing radical retropubic prostatectomy. J Urol. 2007;177(5):1632–5. 9. Forssman L. Blood flow in myomatous uteri as measured by intraarterial 133-Xenon. Acta Obstet Gynecol Scand. 1976;55:21–4. 10. Wei JJ, Zhang XM, Chiriboga L, Yee H, Perle MA, Mittal K. Spatial differences in biologic activity of large uterine leiomyomata. Fertil Steril. 2006;85:179–87. 11. Malvasi A, Tinelli A, Cavallotti C, Morroni M, Tsin DA, Nezhat C, et al. Distribution of substance P (SP) and vasoactive intestinal peptide (VIP) in pseudocapsules of uterine fibroids. Peptides. 2011;32:327–32. 12. Malvasi A, Cavallotti C, Nicolardi G, Pellegrino M, Dell’edera D, Vergara D, et al. NT, NPY and PGP 9.5 presence in myometrium and in fibroid pseudocapsule and their possible impact on muscular physiology. Gynecol Endocrinol. 2013;29:177–81. 13. Malvasi A, Cavallotti C, Nicolardi G, Pellegrino M, Vergara D, Greco M, et al. The opioid neuropeptides in uterine fibroid pseudocapsules: a putative association with cervical integrity in human reproduction. Gynecol Endocrinol. 2013;29:982–8. 14. Tinelli A, Hurst BS, Mettler L, Tsin DA, Pellegrino M, Nicolardi G, et  al. Ultrasound evaluation of uterine healing after laparoscopic intracapsular myomectomy: an observational study. Hum Reprod. 2012;27:2664–70. 15. Delgado AV, McManus AT, Chambers JP. Exogenous administration of Substance P enhances wound healing in a novel skin-injury model. Exp Biol Med (Maywood). 2005;230:271–80. 16. Jiang MH, Chung E, Chi GF, Ahn W, Lim JE, Hong HS, et  al. Substance P induces M2-type macrophages after spinal cord injury. Neuroreport. 2012;23:786–92. 17. Malvasi A, Tinelli A, Cavallotti C, Bettocchi S, Di Renzo GC, Stark M. Substance P (SP) and vasoactive intestinal polypeptide (VIP) in the lower uterine segment in first and repeated cesarean sections. Peptides. 2010;31:2052–9. 18. Collins JJ, Usip S, McCarson KE, Papka RE. Sensory nerves and neuropeptides in uterine cervical ripening. Peptides. 2002;23:167–83. 19. Tingaker BK, Ekman-Ordeberg G, Facer P, Irestedt L, Anand P.  Influence of pregnancy and labor on the occurrence of nerve fibers expressing the capsaicin receptor TRPV1 in human corpus and cervix uteri. Reprod Biol Endocrinol. 2008;6:8. 20. Tingaker BK, Ekman-Ordeberg G, Forsgren S. Presence of sensory nerve corpuscles in the human corpus and cervix uteri during pregnancy and labor as revealed by immunohistochemistry. Reprod Biol Endocrinol. 2006;4:45. 21. Cormio L, Gesualdo L, Maiorano E, Bettocchi C, Palumbo F, Traficante A, et al. Vasoactive intestinal polypeptide (VIP) is not an androgen-dependent neuromediator of penile erection. Int J Impot Res. 2005;17:23–6. 22. Rodríguez R, Pozuelo JM, Martín R, Arriazu R, Santamaria L. Stereological quantification of nerve fibers immunoreactive to PGP 9.5, NPY, and VIP in rat prostate during postnatal development. J Androl. 2005;26:197–204. 23. Takenaka A, Murakami G, Soga H, Han SH, Arai Y, Fujisawa M.  Anatomical analysis of the neurovascular bundle supplying penile cavernous tissue to ensure a reliable nerve graft after radical prostatectomy. J Urol. 2004;172(3):1032–5.

A. Tinelli et al. 24. Hernandez DJ, Epstein JI, Trock BJ, Tsuzuki T, Carter HB, Walsh PC. Radical retropubic prostatectomy. How often do experienced surgeons have positive surgical margins when there is extraprostatic extension in the region of the neurovascular bundle? J Urol. 2005;173(2):446–9. 25. Shah SJ, Goyal V, Sachar R, Nath AK, Jain N, Kapadia K. Seminal vesicle sparing laparoscopic radical prostatectomy using a low-­ energy source: better continence and potency. Indian J Urol. 2009;25(2):199–202. 26. Han M, Kim C, Mozer P, Schäfer F, Badaan S, Vigaru B, Tseng K, Petrisor D, Trock B, Stoianovici D.  Tandem-robot assisted laparoscopic radical prostatectomy to improve the neurovascular bundle visualization: a feasibility study. Urology. 2011;77(2):502–6. 27. Walz J, Burnett AL, Costello AJ, Eastham JA, Graefen M, Guillonneau B, Menon M, Montorsi F, Myers RP, Rocco B, Villers A. A critical analysis of the current knowledge of surgical anatomy related to optimization of cancer control and preservation of continence and erection in candidates for radical prostatectomy. Eur Urol. 2010;57(2):179–92. 28. Tinelli A, Malvasi A, Hurst BS, Tsin DA, Davila F, Dominguez G, Dell’edera D, Cavallotti C, Negro R, Gustapane S, Teigland CM, Mettler L. Surgical management of neurovascular bundle in uterine fibroid pseudocapsule. JSLS. 2012;16(1):119–29. 29. Tinelli A, Mettler L, Malvasi A, Hurst B, Catherino W, Mynbaev OA, Guido M, Alkatout I, Schollmeyer T.  Impact of surgical approach on blood loss during intracapsular myomectomy. Minim Invasive Ther Allied Technol. 2014;23(2):87–95. 30. Tinelli A, Malvasi A, Mynbaev OA, Barbera A, Perrone E, Guido M, Kosmas I, Stark M.  The surgical outcome of intracapsular cesarean myomectomy. A match control study. J Matern Fetal Neonatal Med. 2014;27(1):66–71. 31. Wamsteker K, Emanuel MH, de Kruif JH.  Transcervical hysteroscopic resection of submucous fibroids for abnormal uterine bleeding: results regarding the degree of intramural extension. Obstet Gynecol. 1993;82(5):736–40. 32. Tinelli A, Favilli A, Lasmar RB, Mazzon I, Gerli S, Xue X, Malvasi A. The importance of pseudocapsule preservation during hysteroscopic myomectomy. Eur J Obstet Gynecol Reprod Biol. 2019;243:179–84. 33. Casadio P, Youssef AM, Spagnolo E, Rizzo MA, Talamo MR, De Angelis D, et  al. Should the myometrial free margin still be considered a limiting factor for hysteroscopic resection of submucous fibroids? A possible answer to an old question. Fertil Steril. 2011;95(5):1764–8.e1. 34. Mazzon I, Sbiroli C. Miomectomia. In: Mazzon I, Sbiroli C, editors. Manuale di Chirurgia Resettoscopica in Ginecologia. Torino: UTET; 1997. p. 191–217. 35. Mazzon I, Favilli A, Grasso M, Horvath S, Bini V, Di Renzo GC, Gerli S. Risk factors for the completion of the cold loop hysteroscopic myomectomy in a one-step procedure: a post hoc analysis. Biomed Res Int. 2018;2018:8429047. 36. Mazzon I, Bettocchi S, Fascilla F, DE Palma D, Palma F, Zizolfi B, DI Spiezio Sardo A.  Resectoscopic myomectomy. Minerva Ginecol. 2016;68(3):334–44. 37. Mazzon I, Favilli A, Grasso M, Horvath S, Bini V, Di Renzo GC, Gerli S. Predicting success of single step hysteroscopic myomectomy: a single centre large cohort study of single myomas. Int J Surg. 2015;22:10–4. 38. Mazzon I, Favilli A, Grasso M, Horvath S, Di Renzo GC, Gerli S.  Is cold loop hysteroscopic myomectomy a safe and effective technique for the treatment of submucous myomas with intramural development? A series of 1434 surgical procedures. J Minim Invasive Gynecol. 2015;22(5):792–8. 39. Magos AL, Bournas N, Sinha R, Richardson RE, O’Connor H.  Vaginal myomectomy. Br J Obstet Gynaecol. 1994;101(12):1092–4.

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy 40. Faivre E, Surroca MM, Deffieux X, Pages F, Gervaise A, Fernandez H.  Vaginal myomectomy: literature review. J Minim Invasive Gynecol. 2010;17(2):154–60. 41. Tinelli A, Mynbaev OA, Sparic R, Vergara D, Di Tommaso S, Salzet M, Maffia M, Malvasi A.  Angiogenesis and vascularization of uterine leiomyoma: clinical value of pseudocapsule containing peptides and neurotransmitters. Curr Protein Pept Sci. 2017;18(2):129–39. 42. Di Tommaso S, Massari S, Malvasi A, Bozzetti MP, Tinelli A. Gene expression analysis reveals an angiogenic profile in uterine leiomyoma pseudocapsule. Mol Hum Reprod. 2013;19(6):380–7. 43. Stewart EA, Laughlin-Tommaso SK, Catherino WH, Lalitkumar S, Gupta D, Vollenhoven B. Uterine fibroids. Nat Rev Dis Primers. 2016;2:16043. 44. Awataguchi K. Studies on the angioarchitecture of uterine myoma. Nippon Ika Daigaku Zasshi. 1982;49:225–32. 45. Hanna KR, Katz AJ. An update on wound healing and the nervous system. Ann Plast Surg. 2011;67(1):49–52. 46. Klatsky PC, Tran ND, Caughey AB, et al. Fibroids and reproductive outcomes: a systematic literature review from conception to delivery. Am J Obstet Gynecol. 2008;198(4):357. 47. Vinciguerra M, Sparic R, Hartinaz S, Trojano G, Tinelli A. Cesarean myomectomy technique: a critical review. JDREAM. 2020;4(1):25–35. 48. Laughlin SK, Baird DD, Savitz DA, Herring AH, Hartmann KE. Prevalence of uterine leiomyomas in the first trimester of pregnancy an ultrasound-screening study. Ob Gyn. 2009;113(3):630–5. 49. Malvasi A, Tinelli A, Rahimi S, D’Agnese G, Rotoni C, Dell’edera D, Tsin DA, Cavallotti C.  A three-dimensional morphological reconstruction of uterine leiomyoma pseudocapsule vasculature by the Allen-Cahn mathematical model. Biomed Pharmacother. 2011;65(5):359–63. 50. Mettler L, Schollmeyer T, Tinelli A, Malvasi A, Alkatout I. Complications of uterine fibroids and their management, surgical management of fibroids, laparoscopy and hysteroscopy versus hysterectomy, haemorrhage, adhesions, and complications. Obstet Gynecol Int. 2012;2012:791248. 51. Vikhareva Osser O, Valentin L.  Risk factors for incomplete healing of the uterine incision after caesarean section. BJOG. 2010;117(9):1119–26. 52. Kubinova K, Mara M, Horak P, Kuzel D, Dohnalova A. Reproduction after myomectomy: comparison of patients with and without second-look laparoscopy. Minim Invasive Ther Allied Technol. 2012;21(2):118–24. 53. Tinelli A, Malvasi A, Guido M, Tsin DA, Hudelist G, Hurst B, Stark M, Mettler L.  Adhesion formation after intracapsular myomectomy with or without adhesion barrier. Fertil Steril. 2011;95(5):1780–5. 54. Casini ML, Rossi F, Agostini R, et al. Effects of the position of fibroids on fertility. Gynecol Endocrinol. 2006;22(2):106–9. 55. Saccardi C, Visentin S, Noventa M, Cosmi E, Litta P, Gizzo S.  Uncertainties about laparoscopic myomectomy during pregnancy: a lack of evidence or an inherited misconception? A critical literature review starting from a peculiar case. Minim Invasive Ther Allied Technol. 2015;24:189–94. 56. Suwandinata FS, Gruessner SEM, Omwandho COA, Tinneberg HR.  Pregnancy-preserving myomectomy: preliminary report on a new surgical technique. Eur J Contracept Reprod Health Care. 2008;13:323–6. 57. Laughlin SK, Schroeder JC, Baird DD. New directions in the epidemiology of uterine fibroids. Semin Reprod Med. 2010;28:204–17. 58. Wallach EE, Vlahos NF. Uterine myomas: an overview of development, clinical features, and management. Obstet Gynecol. 2004;104(2):393–406. 59. Tamaya T, Fujimoto J, Okada H.  Comparison of cellular levels of steroid receptors in uterine leiomyoma and myometrium. Acta Obstet Gynecol Scand. 1985;64(4):307–9.

1023

60. Maruo T, Ohara N, Wang J, Matsuo H. Sex steroidal regulation of uterine leiomyoma growth and apoptosis. Hum Reprod Update. 2004;10(3):207–20. 61. Moro E, Degli Esposti E, Borghese G, Manzara F, Zanello M, Raimondo D, Gava G, Arena A, Casadio P, Meriggiola MC, Seracchioli R. The impact of hormonal replacement treatment in postmenopausal women with uterine fibroids: a state-of-the-art review of the literature. Medicina Kaunas. 2019;55(9):549. 62. Parazzini F, Negri E, La Vecchia C, Chatenoud L, Ricci E, Guarniero P.  Reproductive factors and risk of uterine fibroids. Epidemiology. 1996;7(4):440–2. 63. Samaldi AR, Lee NC, Flanders WD, Boring JR III, Parris EB. Risk factors for self-reported uterine fibroids: a case control study. Am J Public Health. 1996;86(6):858–62. 64. Marshall LM, Spiegelman D, Barbieri RL, Goldman MB, Manson JE, Colditz GA, Willett WC, Hunter DJ. Variation in the incidence of uterine leiomyoma among premenopausal women by age and race. Obstet Gynecol. 1997;90(6):967–73. 65. Benson CB, Chow JS, Chang-Lee W, Hill JA III, Doubilet PM. Outcome of pregnancies in women with uterine leiomyomas identified by sonography in the first trimester. J Clin Ultrasound. 2001;29(5):261–4. 66. Adisso S, Hounsossou H, Alle IR, Adisso EL, Takpara L, Alihonou E. Quelle issue pour la grossesse jeun dans un uterus myomateux? JBC. 2014;21:13–7. 67. Eze CU, Odumeru EA, Ochie K, Nwadike UI, Agwuna KK.  Sonographic assessment of pregnancy co-existing with uterine leiomyoma in Owerri, Nigeria. Afr Health Sci. 2013;13(2):453–60. 68. Egbe TO, Badjang TG, Tchounzou R, et  al. Uterine fibroids in pregnancy: prevalence, clinical presentation, associated factors and outcomes at the Limbe and Buea Regional Hospitals, Cameroon: a cross-sectional study. BMC Res Notes. 2018;11:889. 69. Noor S, Fawwad A, Sultana R, Bashir R, Qurat-ul-ain, Jalil H, Suleman N, Khan A.  Pregnancy with fibroids and its and its obstetric complication. J Ayub Med Coll Abbottabad. 2009;21(4):37–40. 70. Fuchs A, Dulska A, Sikora J, Czech I, Skrzypulec-Plinta V, Drosdzol-Cop A.  Symptomatic uterine fibroids in pregnancy  wait or operate? Own experience. Ginekol Pol. 2019;90(6):320–4. 71. Chill HH, Karavani G, Rachmani T, et al. Growth pattern of uterine leiomyoma along pregnancy. BMC Womens Health. 2019;19:100. 72. Vitagliano A, Noventa M, Di Spiezio Sardo A, Saccone G, Gizzo S, Borgato S, Vitale SG, Laganà AS, Nardelli GB, Litta PS, Saccardi C.  Uterine fibroid size modifications during pregnancy and puerperium: evidence from the first systematic review of literature. Arch Gynecol Obstet. 2018;297(4):823–35. 73. Aharoni A, Reiter A, Golan D, Paltiely Y, Sharf M.  Patterns of growth of uterine leiomyomas during pregnancy. A prospective longitudinal study. Br J Obstet Gynaecol. 1988;95(5):510–3. 74. Neiger R, Sonek JD, Croom CS, Ventolini G. Pregnancy-related changes in the size of uterine leiomyomas. J Reprod Med. 2006;51(9):671–4. 75. Lev-Toaff AS, Coleman BG, Arger PH, Mintz MC, Arenson RL, Toaff ME.  Leiomyomas in pregnancy: sonographic study. Radiology. 1987;164(2):375–80. 76. Benaglia L, Cardellicchio L, Filippi F, et al. The rapid growth of fibroids during early pregnancy. PLoS One. 2014;9(1):e85933. 77. La Marca A, Giulini S, Orvieto R, De Leo V, Volpe A.  Anti-­ Müllerian hormone concentrations in maternal serum during pregnancy. Hum Reprod. 2005;20(6):1569–72. 78. Sarais V, Cermisoni GC, Schimberni M, Alteri A, Papaleo E, Somigliana E, Vigano’ P.  Human chorionic gonadotrophin as a possible mediator of leiomyoma growth during pregnancy: molecular mechanisms. Int J Mol Sci. 2017;18(9):2014. 79. Lee HJ, Norwitz ER, Shaw J.  Contemporary management of fibroids in pregnancy. Rev Obstet Gynecol. 2010;3(1):20–7.

1024 80. Cerdeira AS, Tome M, Moore N, Lim L. Seeing red degeneration in uterine fibroids in pregnancy: proceed with caution. Lancet. 2019;394(10212):37. 81. Burton CA, Grimes DA, March CM. Surgical management of leiomyomata during pregnancy. Obstet Gynecol. 1989;74(5):707–9. 82. Katz VL, Dotters DJ, Droegemeuller W. Complications of uterine leiomyomas in pregnancy. Obstet Gynecol. 1989;73(4):593–6. 83. De Carolis S, Fatigante G, Ferrazzani S, Trivellini C, De Santis L, Mancuso S, Caruso A. Uterine myomectomy in pregnant women. Fetal Diagn Ther. 2001;16(2):116–9. 84. Parker WH. Etiology, symptomatology, and diagnosis of uterine myomas. Fertil Steril. 2007;87(4):725–36. 85. Bisson DL, Newell SD, Laxton C, Royal College of Obstetricians and Gynaecologists. Antenatal and postnatal analgesia: scientific impact paper no. 59. BJOG. 2019;126(4):e114–24. 86. Kasum M.  Hemoperitoneum caused by a bleeding myoma in pregnancy. Acta Clin Croat. 2010;49(2):197–200. 87. Le D, Dey CB, Byun K.  Imaging findings of a torsed pedunculated uterine leiomyoma: a case report. Radiol Case Rep. 2019;15(2):144–9. 88. Basso A, Catalano MR, Loverro G, Nocera S, Di Naro E, Loverro M, Natrella M, Mastrolia SA. Uterine fibroid torsion during pregnancy: a case of laparotomic myomectomy at 18 weeks’ gestation with systematic review of the literature. Case Rep Obstet Gynecol. 2017;2017:4970802. 89. Lolis DE, Kalantaridou SN, Makrydimas G, Sotiriadis A, Navrozoglou I, Zikopoulos K, Paraskevaidis EA. Successful myomectomy during pregnancy. Hum Reprod. 2003;18(8):1699–702. 90. Kobayashi F, Kondoh E, Hamanishi J, Kawamura Y, Tatsumi K, Konishi I. Pyomayoma during pregnancy: a case report and review of the literature. J Obstet Gynaecol Res. 2013;39:383–9. 91. Rice JP, Kay HH, Mahony BS. The clinical significance of uterine leiomyomas in pregnancy. Am J Obstet Gynecol. 1989;160(5 Pt 1):1212–6. 92. Vergani P, Locatelli A, Ghidini A, Andreani M, Sala F, Pezzullo JC.  Large uterine leiomyomata and risk of cesarean delivery. Obstet Gynecol. 2007;109(2 Pt 1):410–4. 93. Coronado GD, Marshall LM, Schwartz SM.  Complications in pregnancy, labor, and delivery with uterine leiomyomas: a population-­based study. Obstet Gynecol. 2000;95(5):764–9. 94. Chuang J, Tsai HW, Hwang JL.  Fetal compression syndrome caused by myoma in pregnancy: a case report. Acta Obstet Gynecol Scand. 2001;80(5):472–3. 95. Romero R, Chervenak FA, DeVore G, Tortora M, Hobbins JC.  Fetal head deformation and congenital torticollis associated with a uterine tumor. Am J Obstet Gynecol. 1981;141(7):839–40. 96. Saleh HS, Mowafy HE, El Hameid AAA, Sherif HE, Mahfouz EM.  Does uterine fibroid adversely affect obstetric outcome of pregnancy? Biomed Res Int. 2018;2018:8367068. 97. Poovathi M, Ramalingam R. Maternal fetal outcome in pregnancy with fibroids: a prospective study. Int J Sci Study. 2016;3(11) 98. Stout MJ, Odibo AO, Shanks AL, Longman RE, MacOnes GA, Cahill AG. Fibroid tumors are not a risk factor for adverse outcomes in twin pregnancies. Am J Obstet Gynecol. 2013;208(1):68–e5. 99. Spyropoulou K, Kosmas I, Tsakiridis I, Mamopoulos A, Kalogiannidis I, Athanasiadis A, Daponte A, Dagklis T.  Myomectomy during pregnancy: a systematic review. Eur J Obstet Gynecol Reprod Biol. 2020;254:15–24. 100. Pearl JP, Price RR, Tonkin AE, Richardson WS, Stefanidis D.  SAGES guidelines for the use of laparoscopy during pregnancy. Surg Endosc. 2017;31(10):3767–82. 101. Kocael PC, Simsek O, Saribeyoglu K, Pekmezci S, Goksoy E. Laparoscopic surgery in pregnant patients with acute abdomen. Ann Ital Chir. 2015;86(2):137–42.

A. Tinelli et al. 102. Zachariah SK, Fenn M, Jacob K, Arthungal SA, Zachariah SA.  Management of acute abdomen in pregnancy: current perspectives. Int J Women’s Health. 2019;11:119–34. 103. Ball E, Waters N, Cooper N, Talati C, Mallick R, Rabas S, Mukherjee A, Sri Ranjan Y, Thaha M, Doodia R, Keedwell R, Madhra M, Kuruba N, Malhas R, Gaughan E, Tompsett K, Gibson H, Wright H, Gnanachandran C, Hookaway T, Baker C, Murali K, Jurkovic D, Amso N, Clark J, Thangaratinam S, Chalhoub T, Kaloo P, Saridogan E. Evidence-based guideline on laparoscopy in pregnancy: commissioned by the British Society for Gynaecological Endoscopy (BSGE) endorsed by the Royal College of Obstetricians & Gynaecologists (RCOG). Facts Views Vis Obgyn. 2019;11(1):5–25. 104. Donkervoort SC, Boerma D.  Suspicion of acute appendicitis in the third trimester of pregnancy: pros and cons of a laparoscopic procedure. JSLS. 2011;15(3):379–83. 105. Melgrati L, Damiani A, Franzoni G, Marziali M, Sesti F. Isobaric (gasless) laparoscopic myomectomy during pregnancy. J Minim Invasive Gynecol. 2005;12:379–81. 106. Curet MJ, Vogt DA, Schob O, Qualls C, Izquierdo LA, Zucker KA. Effects of CO2 pneumoperitoneum in pregnant ewes. J Surg Res. 1996;63(1):339–44. 107. Rothmund R, Taran FA, Boeer B, et  al. Surgical and conservative management of symptomatic leiomyomas during pregnancy: a retrospective pilot study. Geburtshilfe Frauenheilkd. 2013;73(4):330–4. 108. Exacoustòs C, Rosati P.  Ultrasound diagnosis of uterine myomas and complications in pregnancy. Obstet Gynecol. 1993;82(1):97–101. 109. Ardovino M, Ardovino I, Castaldi MA, Monteverde A, Colacurci N, Cobellis L. Laparoscopic myomectomy of a subserous pedunculated fibroid at 14 weeks of pregnancy: a case report. J Med Case Rep. 2011;5:545. 110. Luxman D, Cohen JR, David MP.  Laparoscopic myomectomy during pregnancy. Gynaecol Endosc. 1998;7:105–7. 111. Tinelli A, Malvasi A, Hudelist G, Cavallotti C, Tsin DA, Schollmeyer T, Bojahr B, Mettler L.  Laparoscopic intracapsular myomectomy: comparison of single versus multiple fibroids removal. An institutional experience. J Laparoendosc Adv Surg Tech A. 2010;20(8):705–11. 112. Sentilhes L, Sergent F, Verspyck E, Gravier A, Roman H, Marpeau L.  Laparoscopic myomectomy during pregnancy resulting in septic necrosis of the myometrium. BJOG. 2003;110:876–8. 113. Domenici L, Di Donato V, Gasparri ML, Lecce F, Caccetta J, Panici PB. Laparotomic myomectomy in the 16th week of pregnancy: a case report. Case Rep Obstet Gynecol. 2014;2014:154347. 114. ACOG Committee. Opinion no. 770: uterine morcellation for presumed leiomyomas. Obstet Gynecol. 2019;133(3):e238–48. 115. Fogerty AE. Challenges of anticoagulation therapy in pregnancy. Curr Treat Opt Cardiovasc Med. 2017;19(10):76. 116. Kilpatrick CC, Adler MT, Chohan L.  Vaginal myomectomy in pregnancy: a report of two cases. South Med J. 2010;103:1058–60. 117. Celik C, Acar A, Ciçek N, Gezginc K, Akyürek C. Can myomectomy be performed during pregnancy? Gynecol Obstet Investig. 2002;53(2):79–83. 118. Gerome JM, Church TL. Puerperal complications of a retroplacental uterine leiomyoma. J Am Osteopath Assoc. 2017;117:660–3. 119. Szamatowicz J, Laudanski T, Bulkszas B, Akerlund M. Fibromyomas and uterine contractions. Acta Obstet Gynecol Scand. 1997;76(10):973–6. 120. Agathoklis C, Arulkumaran S. Uterine contractions. Best practice in labour and delivery. Cambridge: Cambridge University Press; 2016. p. 60–73.

64  Fibroids in Obstetric and Gynecology: Training and Skill in Myomectomy 121. Krimou Y, Erraghay S, Guennoun A, Mamouni N, Bouchikhi C, Banani A.  Myoma praevia and pregnancy. Pan Afr Med J. 2019;33:216. 122. Michalas SP, Oreopoulou FV, Papageorgiou JS.  Myomectomy during pregnancy and caesarean section. Hum Reprod. 1995;10:1869–70. 123. Mattingly RF.  Te Linde’s operative gynecology. 5th ed. Philadelphia, PA: JB Lippincott Co; 1977. p. 219. 124. Scott JR, Disaia PJ, Hammond CB, Spellacy WN.  Danforth’s obstetrics and gynecology. Philadelphia, PA: JB Lippincott Co; 1994. p. 936. 125. Ghaemmaghami F, Karimi-Zarchi M, Gharebaghian M, Kermani T. Successful myomectomy during cesarean section: case report and literature review. Int J Biomed Sci. 2017;13(2):119–21. 126. Hsieh TT, Cheng BJ, Liou JD, Chiu TH. Incidental myomectomy in cesarean section. Changgeng Yi Xue Za Zhi. 1989;12:13–20. 127. Sparic R, Malvasi A, Kadija S, Stefanović A, Radjenović SS, Popović J, Pavić A, Tinelli A. Safety of cesarean myomectomy in women with single anterior wall and lower uterine segment myomas. J Matern Fetal Neonatal Med. 2017;31:1–4. 128. Ramya T, Sabnis S, Schitra TV, Panicker S.  Cesarean myomectomy: an experience from a tertiary care teaching hospital. J Obstet Gynaecol India. 2019;69(5):426–30. 129. Arrieta-Blanco JJ, Oñate-Sánchez R, Martínez-López F, Oñate-­ Cabrerizo D, Cabrerizo-Merino MD.  Inherited, congenital and acquired disorders by hemostasis (vascular, platelet & plasmatic phases) with repercussions in the therapeutic oral sphere. Med Oral Patol Oral Cir Bucal. 2014;19(3):e280–8. 130. Zhao R, Wang X, Zou L, Zhang W.  Outcomes of myomectomy at the time of cesarean section among pregnant women with uterine fibroids: a retrospective cohort study. Biomed Res Int. 2019;2019:7576934. 131. Ehigiegba AE, Ande AB, Ojobo SI. Myomectomy during cesarean section. Int J Gynecol Obstet. 2001;75(1):21–5. 132. Kwon DH, Song JE, Yoon KR, Lee KY.  The safety of cesarean myomectomy in women with large myomas. Obstet Gynecol Sci. 2014;57(5):367–72. 133. Dedes I, Schäffer L, Zimmermann R, Burkhardt T, Haslinger C. Outcome and risk factors of cesarean delivery with and without

1025

cesarean myomectomy in women with uterine myomatas. Arch Gynecol Obstet. 2017;295(1):27–32. 134. Pergialiotis V, Sinanidis I, Louloudis IE, Vichos T, Perrea DN, Doumouchtsis SK.  Perioperative complications of cesarean delivery myomectomy: a meta-analysis. Obstet Gynecol. 2017;130(6):1295–303. 135. Hatırnaz Ş, Güler O, Başbuğ A, Çetinkaya MB, Kanat-Pektaş M, Bakay K, Çelik S, Şentürk Ş, Soyer-Çalışkan C, Gürçağlar A, Şahin B, Kalkan Ü, Çelik H, Kalyoncu Ş, Bıyık İ, Yassa M, Erol O, Akarsu S, Turhan U, Ulubaşoğlu H, Sparic R, Tinelli A. A Comparative multicentric study on serosal and endometrial myomectomy during cesarean section: surgical outcomes. J Investig Surg. 2020;34:1–8. 136. Sparić R, Kadija S, Stefanović A, Spremović Radjenović S, Likić Ladjević I, Popović J, Tinelli A. Cesarean myomectomy in modern obstetrics: more light and fewer shadows. J Obstet Gynaecol Res. 2017;43(5):798–804. 137. Huang SY, Shaw SW, Su SY, Li WF, Peng HH, Cheng PJ.  The impact of a novel transendometrial approach for caesarean myomectomy on obstetric outcomes of subsequent pregnancy: a longitudinal panel study. BJOG. 2018;125(4):495–500. 138. Brown D, Fletcher HM, Myrie MO, Reid M. Caesarean myomectomy-­a safe procedure. A retrospective case controlled study. J Obstet Gynaecol. 1999;19(2):139–41. 139. Rai A, Mishra MG.  A study on safety and feasibility of caesarean myomectomy: at a private institute. Int J Reprod Contracept Obstet Gynecol. 2017;6:2765–70. 140. Sapmaz E, Celik H, Altungül A. Bilateral ascending uterine artery ligation vs. tourniquet use for hemostasis in cesarean myomectomy. A comparison. J Reprod Med. 2003;48(12):950–4. 141. Cobellis L, Florio P, Stradella L, De Lucia E, Messalli EM, Petraglia F, Cobellis G. Electro-cautery of myomas during caesarean section--two case reports. Eur J Obstet Gynecol Reprod Biol. 2002;102(1):98–9. 142. Incebiyik A, Hilali NG, Camuzcuoglu A, Vural M, Camuzcuoglu H. Myomectomy during caesarean: a retrospective evaluation of 16 cases. Arch Gynecol Obstet. 2014;289(3):569–73.

Rupture of the Uterus: A Dramatic Condition in a Genital Organ

65

Leonardo Resta, Gerardo Cazzato, Eliano Cascardi, and Roberta Rossi

Rupture of the uterus is a dramatic condition characterized by the complete rupture of the wall from the mucosa to the serosa at various levels, with serious consequences for the woman and for the eventual product of conception. The sudden and unexpected onset can jeopardize frequently the woman’s life. Fortunately, the lesion is rare. Hofmeyr et al. [1] in a 2005 WHO review reported an incidence of 0.053% (range 0.016– 0.30) in community reviews and an incidence of 0.31% (range 0.012–2.9) in medical institution reviews. The incidence is higher in developing populations. The percentage rises in patients who have had a previous caesarean section and who attempt a subsequent vaginal birth up to 0.4–0.7%, and sees the patient’s age, gestational period and induction with oxytocin as an increase in risk. Uterine rupture almost always occurs due to an obstetric event: edema and the particular fragility of the uterine wall during pregnancy justify the danger. In particular, the destructive action of the extravillous trophoblast facilitates the dissociation of the muscle fibers of the uterus, especially when the nesting occurs in areas where the wall thickness is less than in the tubal horns or in the supra-isthmus. Normally these areas, as well as on the lateral margins of the uterine body, are covered by an endometrium less responsive to horL. Resta (*) Division Dipartimento dell’Emergenza e dei Trapianti, University d’Organo (DETO), Bari, Italy University of Bari “Aldo Moro”, Bari, Italy e-mail: [email protected] G. Cazzato · R. Rossi Dipartimento dell’Emergenza e dei Trapianti d’Organo (DETO), Sez. di Anatomia Patologica, Bari, Italy Università degli Studi “Aldo Moro”, Bari, Italy E. Cascardi Department of Medical Sciences, University of Turin, Turin, Italy Pathology Unit, FPO-IRCCS Candiolo Cancer Institute, Candiolo, Italy

monal cyclic stimulation and therefore less suitable for allowing the ligands present on the surface of the endometrial cells during the “implantation window” to engage the proteins present on the trophoblast surface of the blastocyst. This explains how the occurrence of nesting in these areas and the related risk of breakage is an exceptional occurrence. A certain portion of uterine ruptures is related to the presence of placental abnormalities, such as placenta accreta (Figs. 65.1, 65.2, 65.3, and 65.4). The grafting of the trophoblastic material between the myometrial bundles, sometimes up to the serous surface of the uterus, can cause a weakening of the wall even in the postpartum period, especially in cases of high uterotonic administration [2, 3]. Certainly in our years the most frequent occasion for uterus rupture is represented by the presence of a caesarean section scar (Figs. 65.5, 65.6, and 65.7). In these cases, the rupture can occur by implantation of the gestational sac on the scar, with similar aspects to the isthmic-cervical implant, but with the aggravating circumstance of the modifications present in the myometrium: the wall of the uterus is further thinned and the muscle bundles are reduced and distributed orthogonally or randomly or dissociated by abundant quantities of collagen fibers [4] (Figs. 65.8 and 65.9). As already reported, the caesarean section scar is a very dynamic event, in which the daily traction on the residual muscle fibers leads to peri-arterial regenerative phenomena, slow-healing granulomatous inflammatory processes, reactive proliferation of the overlying mesothelium. The nesting of the blastocyst, the infiltrative action of the extravillous trophoblast and the presence of chorionic villi often lacking for large tracts of the fibrinoid layer of Nitabuch justify the possibility of a “percreta” penetration of the placenta and the laceration of the uterus wall [4]. In addition to spontaneous rupture by nesting of the placental sac on the caesarean section scar, there is a real risk of rupture, regardless of the implant site, when you want to allow a trial vaginal birth in a woman who has previously

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 G. Cinnella et al. (eds.), Practical Guide to Simulation in Delivery Room Emergencies, https://doi.org/10.1007/978-3-031-10067-3_65

1027

1028

Fig. 65.1  Placenta accreta. On the right side the uterine decidua is extensively invaded by trophoblast. Necrosis foci and thrombosis are present on the left side, on the edge of the wall rupture

Fig. 65.2  The myometrium, occupied by trophoblast, presents necrosis (upper left side) and inflammation with consequent weakness of the uterine wall

L. Resta et al.

Fig. 65.4  Trichromic stain revealing massive invasion of arterial lumen by trophoblast. Physiological phenomenon in normal pregnancy contributes to the ischemic event in the myometrium in case of placenta accreta

Fig. 65.5  Insertion of the gestational sac in a scarring area od caesarean section. Fibrosis, myometrial dissociation, and arterial lumen invasion precede the uterine rupture

Fig. 65.3  In this case of uterine rupture because of placenta accreta the myometrium of the uterine wall is extensively invaded by trophoblastic cells. The dissociation of the muscular cells and the production of lytic proteins by the trophoblast ease the uterine rupture

Fig. 65.6  Same case of the previous figure. It is evident the aggression of the placental villi and of the extravillar trophoblastic cells

had one or more caesarean sections. This occurrence, well known in the past, has been increasing in recent years; however, it is judged to be contained in very low values, such that it is not recommended in most patients [5]. However, initiating a patient who has already been caesarized to have a vagi-

nal birth requires a careful preliminary study and the preparation of appropriate organizational measures to be ready for any emergency. The spread of uterus rupture in patients who have already undergone caesarean section is documented on the rise in various countries and even in

65  Rupture of the Uterus: A Dramatic Condition in a Genital Organ

1029

Fig. 65.7  Same case of the previous figure. Fibrinoid necrosis in the area of rupture

Fig. 65.10  A case of uterine rupture 24  h after extensive myomectomy. We observe surgical staples (on the upper left) and hemorrhagic spreading among the muscular fibers

Fig. 65.8  Decidual modification and trophoblastic infiltration of the tubal cornual. The myometrial wall is very thin

Fig. 65.9  The trichromic stain sows the aggression of the trophoblast among the myometrial fibers

China with the recent authorization of the third child [6]. There are still many perplexities and controversies regarding the real correlations with test vaginal birth, despite published meta-analyses [7, 8]. What has been said for patients who have already undergone a caesarean section also applies to patients who have previously undergone profound myomectomies (Figs. 65.10 and

Fig. 65.11  Same case of the previous figure. Necrosis, hemorrhage, and inflammation along the rupture fistula

65.11), multiple species, which have extensively modified the architecture of the muscles and the vascularity of the uterus. During the operation, the surgeon often finds a blood stratification coinciding with the breach caused in the wall of the uterus. By gently removing the clots, a break is detected composed of irregular, brittle, swollen edges, and a dark red color. Histologically, the evident fracture of the muscle fibers surrounded by hemorrhagic infiltrate prevails. The basic conditions that led to the breakup are also evident. In the case of implantation in poorly protected areas (supra-isthmic region or tubal horn) we can see the invasion of the trophoblast both in the interstitium and inside the arterial vessels. It is clear that the phenomenon of the massive entry of trophoblastic cells into the lumen of the vessel up to the complete obstruction of the same, a phenomenon usual in all pregnancies at the end of the first trimester for the transport of the cells and the modification of the arterial wall, produces a condition of wall ischemia that favors rupture. The areas of necrosis are the indicator of the process. In the case of the implant on a caesarean section scar, we see the fraying of the muscle fibers, already arranged in an irregular

1030

manner, with the massive infiltration of trophoblastic elements. Maternal complications include postpartum bleeding, more or less abundant, with possible hemoperitoneum and a state of surgical emergency [9–11]. The consequence will be in many cases not only a hysterectomy but also the possibility of urogenital lesions and in the most serious cases death. The fetus will suffer an inevitable ischemic insult aggravated by the condition of metahemorrhagic anemia of the mother with possible hypoxic-ischemic encephalopathy, motor paralysis, and death. The histologic examination of the uterus after the uterine rupture is very important to determine if there is a malpractice and this issue is fundamental during litigation liability and claims [12–20].

References 1. Hofmeyr GJ, Say L, Gulmezoglu AM. WHO systematic review of maternal mortality and morbidity: the prevalence of uterine rupture. BJOG. 2005;112(9):1221–8. 2. Jauniaux E, Dimitrova I, Kenyon N, Mhallem M, Kametas NA, Zosmer N, Hubinont C, Nicolaides KH, Collins SL.  Impact of placenta previa with placenta accreta spectrum disorder on fetal growth. Ultrasound Obstet Gynecol. 2019;54(5):643–9. https://doi.org/10.1002/uog.20244. PMID: 30779235; PMCID: PMC6699933. 3. Okaniwa J, Higeta D, Kameda T, Uchiyama Y, Inoue M, Iwase A. Postpartum unscarred uterine rupture caused by placenta accreta: a case report and literature review. Clin Case Rep. 2021;9:1587–90. 4. Pacheco LA, Resta L, Tinelli A, Malvasi A, Haimovich S, Carugno J.  The cesarean scar complication. In: Malvasi A, Tinelli A, Di Renzo GC, editors. Management and therapy of late pregnancy complications. Third trimester and puerperium. Cham: Springer; 2017. 5. Chaudhary V, Singh M, Nain S, et al. Incidence, management and outcomes in women undergoing peripartum hysterectomy in a tertiary care centre in India. Cureus. 2021;13(3):e14171. 6. Zhou Y, Mu Y, Chen P, Xie Y, Zhu J, Liang J. The incidence, risk factors and maternal and foetal outcomes of uterine rupture during different birth policy periods: an observational study in China. BMC Pregnancy Childbirth. 2021;21:360. 7. Motomura K, Ganchimeg T, Nagata C, Ota E, Vogel JP, Betran AP, et  al. Incidence and outcomes of uterine rupture among women with prior caesarean section: WHO multicountry survey on maternal and newborn health. Sci Rep. 2017;7(1):44093. 8. Baradaran K. Risk of uterine rupture with vaginal birth after cesarean in twin gestations. Obstet Gynecol Int. 2021;2021:6693142.

L. Resta et al. 9. Toijonen A, Hinnenberg P, Gissler M, Heinonen S, Macharey G. Maternal and neonatal outcomes in the following delivery after previous preterm caesarean breech birth: a national cohort study. J Obstet Gynaecol. 2021;42:1–6. 10. Liao H, Duan T.  Uterine rupture. In: Di Renzo GC, Bergella V, Malvasi A, editors. Good practice and malpractice in labor and delivery. Milan: Edra Publishing; 2019. p. 247–64. 11. Tinelli A, Kosmas IP, Carugno JT, Carp H, Malvasi A, Cohen SB, Simone Laganà A, Angelini M, Casadio P, Chayo J, Cicinelli E, Gerli S, Palacios Jaraquemada J, Magnarelli G, Medvediev MV, Jimenez Metello J, Nappi L, Okohue J, Sparic R, Stefanović R, Tzabari A, Vimercati A.  Uterine rupture during pregnancy: the URIDA (uterine rupture international data acquisition) study. Int J Gynaecol Obstet. 2021;157:76. https://doi.org/10.1002/ijgo.13810. PMID: 34197642. 12. Tinelli A, Vergara D, Ma Y, Malvasi A. Dystocia, uterine healing and uterine innervation: an unexplored intersection. Curr Protein Pept Sci. 2020;21(5):440–2. https://doi.org/10.2174/13892037206 66190717115744. 13. Trojano G, Damiani GR, Olivieri C, Villa M, Malvasi A, Alfonso R, Loverro M, Cicinelli E. VBAC: antenatal predictors of success. Acta Biomed. 2019;90(3):300–9. https://doi.org/10.23750/abm. v90i3.7623. PMID: 31580319; PMCID: PMC7233729. 14. Zaami S, Montanari Vergallo G, Malvasi A, Marinelli E.  Uterine rupture during induced labor after myomectomy and risk of lawsuits. Eur Rev Med Pharmacol Sci. 2019;23(4):1379–81. https:// doi.org/10.26355/eurrev_201902_17091. 15. Zaami S, Malvasi A, Marinelli E. Fundal pressure: risk factors in uterine rupture. The issue of liability: complication or malpractice? J Perinat Med. 2018;46(5):567–8. https://doi.org/10.1515/ jpm-­2018-­0070. 16. Vimercati A, Del Vecchio V, Chincoli A, Malvasi A, Cicinelli E.  Uterine rupture after laparoscopic myomectomy in two cases: real complication or malpractice? Case Rep Obstet Gynecol. 2017;2017:1404815. https://doi.org/10.1155/2017/1404815. 17. Malvasi A, Cavallotti C, Gustapane S, Giacci F, Di Tommaso S, Vergara D, Mynbaev OA, Tinelli A. Neurotransmitters and neuropeptides expression in the uterine scar after cesarean section. Curr Protein Pept Sci. 2017;18(2):175–80. https://doi.org/10.2174/1389 203717666160322150034. PMID: 27001063. 18. Sparic R, Dotlic J, Mirkovic L, Stamenkovic J, Kotlica BK, Nejkovic L, Babovic I, Malvasi A, Tinelli A.  Asymptomatic isthmico-­cervical uterine perforation with IUD  - our experience and literature review. Clin Exp Obstet Gynecol. 2016;43(6):896–8. PMID: 29944248. 19. Alkatout I, Honemeyer U, Strauss A, Tinelli A, Malvasi A, Jonat W, Mettler L, Schollmeyer T. Clinical diagnosis and treatment of ectopic pregnancy. Obstet Gynecol Surv. 2013;68(8):571–81. https:// doi.org/10.1097/OGX.0b013e31829cdbeb. PMID: 23921671. 20. Malvasi A, Zaami S, Tinelli A, Trojano G, Montanari Vergallo G, Marinelli E. Kristeller maneuvers or fundal pressure and maternal/ neonatal morbidity: obstetric and judicial literature review. J Matern Fetal Neonatal Med. 2019;32(15):2598–607. https://doi.org/10.108 0/14767058.2018.1441278. Epub 2018 Feb 21. PMID: 29466899.

Skills, Learning Curve and Simulation in an Italian University Clinic

66

Marica Falini, Simona Freddio, Antonio Malvasi, and Sandro Gerli

66.1 Simulation The approach to group training using simulators to improve preparation for obstetric emergency management was developed after aviation experts pointed out that 60–80% of air accidents were due to human error. Since the introduction of flight simulators more than 20 years ago, airlines have achieved a 50–80% reduction in accidents, particularly those due to human factors. The philosophy behind group training is based on learning best practice during times of lower stress to improve the chance that emergencies will be handled effectively. From data published in literature, there is evidence to support the use of simulation, team training and scenarios in midwifery to improve both clinical procedures and outcomes. The Confidential Enquiry into Maternal Deaths recommends the use of simulations to improve the effectiveness and efficiency of the obstetric emergency response [1].

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/978-­3-­031-­10067-­3_66.

M. Falini · S. Freddio · S. Gerli (*) Department of Medicine and Surgery, Section of Obstetrics and Gynecology, Perugia University, Perugia, Italy e-mail: [email protected] A. Malvasi Department of Biomedical and Human Oncological Science (DIMO), Unit of Obstetrics and Gynecology, University of Bari, Bari, Italy Santa Maria della Misericordia Hospital Perugia, University of Perugia, Perugia, Italy International Translational Medicine and Biomodelling Research Group, Department of Applied Mathematics, Moscow Institute of Physics and Technology (State University), Moscow, Russia The New European Surgical Academy (NESA), Berlin, Germany

Emerging evidence supports the development of simulations with multidisciplinary teams to improve care, adherence to guidelines and clinical outcomes. Simulations, predominantly used in obstetrics to improve technical skills and knowledge, include the following: breech delivery care [2–4], shoulder dystocia [5–9], perineal tear repair [10–12], operative vaginal delivery [13–16], emergency caesarean section [17, 18] and other conditions, such as eclampsia [19, 20] and postpartum haemorrhage (PPH) [21–23].

66.2 The “Martina Floridi” Obstetrical-­ Gynaecological Simulation Teaching Laboratory of the Degree Course in Midwifery at the University of Perugia Students can make use of the “Martina Floridi” medical, obstetrical-gynaecological simulation teaching laboratory, which was inaugurated on May 16, 2014. This laboratory is furnished with equipment that fully recreates a real labour and delivery room, which has been designed taking into account the latest developments in obstetrics, with innovative technical solutions combined with functional design (Fig. 66.1). There is also a neonatal care unit in which the care of newborns can be practised. In order to comply with the new “clinical competence” of a midwife, laboratory practices have been strengthened through specific courses and workshops for the study of clinical cases, equipping the laboratory with numerous high-­ fidelity mannequins for simulations of cephalic, breech, operative and shoulder dystocia: “Sophie and her Mum-­ Trainer for complete childbirth” (Fig. 66.2). “Head for use of suction cup and forceps Lucy” in addition to the already existing computerised Childbirth and Resuscitation Simulator for reproducing scenarios of obstetric emergencies and treatment simulations in multidisciplinary teams

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023 G. Cinnella et al. (eds.), Practical Guide to Simulation in Delivery Room Emergencies, https://doi.org/10.1007/978-3-031-10067-3_66

1031

1032

M. Falini et al.

Fig. 66.1  The “Martina Floridi” obstetrical-gynecological simulation teaching laboratory of the Degree Course in Midwifery at the University of Perugia. Students can make use of the “Martina Floridi” medical, obstetrical-gynecological simulation teaching laboratory, which was inaugurated on May 16, 2014. This laboratory is furnished with equipment that fully recreates a real labor and delivery room, which has been designed taking into account the latest developments in obstetrics, with innovative technical solutions combined with functional design

Fig. 66.3  Noelle Birthing Simulator S554.100 (SN M1502281)

Fig. 66.2  Sophie and her Mum-Trainer (Accurate) for complete childbirth mannequin

(Noelle Birthing Simulator S554.100 (SN M1502281) (Fig. 66.3) and Newborn Genuine Simulator). The “Noelle simulator” was, specifically, also equipped with a uterus that can be used for simulations of PPH management. The laboratory is also equipped with a Voluson ultrasound scanner, in order for midwives to learn the basics of the application of ECO OFFICE (Figs. 66.4 and 66.5).

Fig. 66.4  Noelle Birthing Simulator S554.100 (SN M1502281) and vaginal complicated birth simulation

A range of materials and teaching aids are available, adapted to the level of learning that the student needs to achieve.

66  Skills, Learning Curve and Simulation in an Italian University Clinic

1033

Fig. 66.6  Simulation with the Noelle Birthing Simulator S554.100 (SN M1502281) of vaginal birth delivery (tutors simulate Mac Robert’s maneuver)

Fig. 66.5  The “Martina Floridi” obstetrical-gynecological simulation teaching laboratory of the Degree Course in Midwifery at the University of Perugia. It is also equipped with a Voluson ultrasound scanner, in order for midwives to learn the basics of the application of Eco Office

66.3 Simulation in Midwifery Degree Training Simulation is widely used in the clinical training of students and health professionals. It is a valuable strategy for teaching, learning and assessing clinical skills at different levels of education: undergraduate, postgraduate and continuing. In the curriculum of our Degree Course, simulation in the workshop, setting in which a large number of possible clinical situations occur, is one of the most used teaching/learning modalities in the theoretical–practical courses, preparatory to the internship work that the student will carry out in the operating units with real patients. Students have the opportunity to train in the different techniques of obstetrical-gynaecological nursing and in scenarios of increasing care complexity under the guidance of a teaching midwife who assumes the role of team leader. Over the years we have found that simulation helps students to consolidate and enhance their knowledge and to increase their satisfaction and self-confidence; moreover, group training also gives them the opportunity to measure and improve their relational and communication skills both towards other team members and towards the patient. On the basis of our teaching activity in the laboratory, it seemed appropriate to produce videos in which the students’ simulation activities are filmed, in order to promote and disclose our experience and to produce teaching material for learning. There are numerous obstetric physiology and pathology scenarios for which students undergo intensive training with

Fig. 66.7  Video tutorial about simulation with Sophie and her Mum-­ Trainer (Accurate) of spontaneous vaginal birth delivery

role play and simulation on high-fidelity manikins in the teaching laboratory. Among these, the videos on offer cover assistance in spontaneous birth, assistance in breech birth, shoulder dystocia and PPH. Spontaneous birth: Laboratory simulation of delivery and afterbirth assistance and a positive assessment by the team leader precede the student’s entry into the delivery room (Figs. 66.6 and 66.7, Video 66.1). The most frequent obstetric emergencies that are simulated in skill laboratories are as follows: breech delivery (Figs.  66.8 and 66.9, Video 66.2), shoulder dystocia and PPH. Shoulder dystocia: Shoulder dystocia is simulated periodically to exercise the obstetric team in the manoeuvres that must be practised in this obstetric emergency (Figs.  66.10 and 66.11, Video 66.3).

1034

Fig. 66.8  Simulation with Sophie and her Mum-Trainer (Accurate) of vaginal breech delivery

M. Falini et al.

Fig. 66.11  Video tutorial about Simulation of shoulder dystocia with the with Sophie and her Mum-Trainer (Accurate) and with Noelle Birthing Simulator S554.100 (SN M1502281)

Fig. 66.9  Video tutorial about simulation with Sophie and her Mum-­ Trainer (Accurate) of vaginal breech delivery Fig. 66.12  Simulation of balloon application during PPH with the use of Noelle Birthing Simulator S554.100 (SN M1502281)

Fig. 66.10  Simulation of shoulder dystocia with the Noelle Birthing Simulator S554.100 (SN M1502281)

Postpartum haemorrhage: The simulation of PPH requires the training of a multidisciplinary team to face not only the manoeuvres and obstetric surgical techniques but also the anaesthesiology and resuscitation problems of PPH (Figs. 66.12, 66.13, and 66.14, Video 66.4). The simulation of operative delivery mainly concerns the application of the vacuum extractor (Fig.  66.15). The simulation on the manikin allows to learn the correct application of the cup on the fetal head avoiding the incorrect applications that can cause maternal and fetal lesions (Fig. 66.16).

Fig. 66.13  Simulation of the instruments and the balloon that must be prepared by the obstetric team to apply the balloon correctly

These are obstetrical pathology scenarios of low incidence in clinical practice and high complexity of care, which students will unlikely face during their training experience, but whose diagnosis and emergency treatment in autonomy and team are among the basic skills of a midwife, so it is essential that students practice periodically in dedicated settings with simulations in team working [24, 25].

66  Skills, Learning Curve and Simulation in an Italian University Clinic

1035

Fig. 66.14  Video tutorial about Simulation of postpartum hemorrhage with Noelle Birthing Simulator S554.100 (SN M1502281)

Fig. 66.16  Simulation of incorrect application of the vacuum, in fact the fetal head is in posterior occiput position and the cup is applied in the frontal region

References

Fig. 66.15  Simulation of vaginal operative delivery with application of the Kiwi vacuum on the mannequin

However, all students, at the end of the course, report having witnessed or taken part in the treatment of PPH at least once during their entire course of study [26]. In recent years, simulation has assumed ever greater importance in forensic medicine for liability and claim [27, 28]. We observed a significant reduction in malpractice claim rates after simulation training. Wider use of simulation training within obstetrics should be considered [29, 30].

1. Confidential Enquiry into Maternal Deaths in the United Kingdom. Why mothers die. London: Royal College of Obstetricians and Gynaecologist; 1999. 2. Hardy L, Garratt JL, Crossley B, Copson S, Nathan E, Calvert K, Epee-Bekima M.  A retrospective cohort study of the impact of In Time obstetric simulation training on management of vaginal breech deliveries. Aust N Z J Obstet Gynaecol. 2020;60(5):704–8. 3. Stone H, Crane J, Johnston K, Craig C.  Retention of vaginal breech delivery skills taught in simulation. J Obstet Gynaecol Can. 2018;40(2):205–10. 4. Deering S, Brown J, Hodor J, Satin AJ.  Simulation training and resident performance of singleton vaginal breech delivery. Obstet Gynecol. 2006;107(1):86–9. 5. Olson DN, Logan L, Gibson KS.  Evaluation of multidisciplinary shoulder dystocia simulation training on knowledge, performance, and documentation. Am J Obstet Gynecol MFM. 2021;3(5):100401. 6. Gurewitsch Allen ED.  Simulation of shoulder dystocia for skill acquisition and competency assessment: a systematic review and gap analysis. Simul Healthc. 2018;13(4):268–83. 7. Shaddeau AK, Deering S. Simulation and shoulder dystocia. Clin Obstet Gynecol. 2016;59(4):853–8. 8. Grobman WA. Shoulder dystocia: simulation and a team-centered protocol. Semin Perinatol. 2014;38(4):205–9. 9. Grimm MJ, Costello RE, Gonik B.  Effect of clinician-applied maneuvers on brachial plexus stretch during a shoulder dystocia

1036 event: investigation using a computer simulation model. Am J Obstet Gynecol. 2010;203(4):339.e1–5. 10. Sano Y, Hirai C, Makino S, Li X, Takeda J, Itakura A, Takeda S.  Incidence and risk factors of severe lacerations during forceps delivery in a single teaching hospital where simulation training is held annually. J Obstet Gynaecol Res. 2018;44(4):708–16. 11. Illston JD, Ballard AC, Ellington DR, Richter HE. Modified beef tongue model for fourth-degree laceration repair simulation. Obstet Gynecol. 2017;129(3):491–6. 12. Gossett DR, Gilchrist-Scott D, Wayne DB, Gerber SE. Simulation training for forceps-assisted vaginal delivery and rates of maternal perineal trauma. Obstet Gynecol. 2016;128(3):429–35. 13. Bligard KH, Lipsey KL, Young OM. Simulation training for operative vaginal delivery among obstetrics and gynecology residents: a systematic review. Obstet Gynecol. 2019;134(Suppl 1):16S–21S. 14. Rose K, Kwan L, Pluym ID, Zhang H, Han CS, Afshar Y.  Forceps-assisted vaginal delivery: the landscape of obstetrics and gynecology resident training. J Matern Fetal Neonatal Med. 2021;34(18):3039–45. 15. Mannella P, Giordano M, Guevara MMM, Giannini A, Russo E, Pancetti F, Caretto M, Simoncini T. Simulation training program for vacuum application to improve technical skills in vacuum-assisted vaginal delivery. BMC Pregnancy Childbirth. 2021;21(1):338. 16. Mhyre J, Ward N, Whited TM, Anders M. Randomized controlled simulation trial to compare transfer procedures for emergency cesarean. J Obstet Gynecol Neonatal Nurs. 2020;49(3):272–82. 17. Sultana N, Betran AP, Khan KS, Sobhy S. Simulation-based teaching and models for caesarean sections: a systematic review to evaluate the tools for the ‘See One, Practice Many, Do One’ slogan. Curr Opin Obstet Gynecol. 2020;32(5):305–15. 18. Foglia LM, Eubanks AA, Peterson LC, Hickey K, Hammons CB, Borgia LB, Light MR, Jackson A, Deering S. Creation and evaluation of a cesarean section simulator training program for novice obstetric surgeons. Cureus. 2020;12(9):e10324. 19. Akalin A, Sahin S.  The impact of high-fidelity simulation on knowledge, critical thinking, and clinical decision-making for the management of pre-eclampsia. Int J Gynaecol Obstet. 2020;150(3):354–60.

M. Falini et al. 20. Abraham C, Kusheleva N.  Management of pre-eclampsia and eclampsia: a simulation. MedEdPORTAL. 2019;15:10832. 21. Dillon SJ, Kleinmann W, Fomina Y, Werner B, Schultz S, Klucsarits S, Moreno W, Butsko A, McIntire DD, Nelson DB. Does simulation improve clinical performance in management of postpartum hemorrhage? Am J Obstet Gynecol. 2021;225(4):435.e1–8. 22. Pettersen G, Gauvin F, Robitaille N, Sansregret A, Lesage S, Levy A. Massive hemorrhage protocol application and teamwork skills. AEM Educ Train. 2020;5(3):e10513. 23. Cheloufi M, Picard J, Hoffmann P, Bosson JL, Allenet B, Berveiller P, Albaladejo P. How to agree on what is fundamental to optimal teamwork performance in a situation of postpartum hemorrhage? A multidisciplinary Delphi French study to develop the Obstetric Team Performance Assessment Scale (OTPA Scale). Eur J Obstet Gynecol Reprod Biol. 2021;256:6–16. 24. Active Studio. Utilizzo della simulazione nella formazione in ambito sanitario. n.d.. https://www.activestudio.it/ utilizzo-­della-­simulazione-­nella-­formazione-­in-­ambito-­sanitario/. 25. Draycott T, Sibanda T, Owen L, Akande V, Winter C, Reading S, Whitelaw A, Draycott T.  Does training in obstetric emergencies improve neonatal outcome? BJOG. 2006;113(2):177–82. 26. Morese A, Ragusa A. Organizzazione e formazione nelle urgenze ed emergenze in sala parto. In: Crescini A, Ragusa C, Urgenze, editors. Emergenze in sala parto. Padova: PICCIN Nuova Libraria; 2015. p. 3–10. 27. Hanscom R.  Medical simulation from an insurer’s perspective. Acad Emerg Med. 2008;15(11):984–7. 28. Iverson RE Jr, Heffner LJ.  Patient safety series: obstetric safety improvement and its reflection in reserved claims. Am J Obstet Gynecol. 2011;205(5):398–401. 29. Schaffer AC, Babayan A, Einbinder JS, Sato L, Gardner R.  Association of simulation training with rates of medical malpractice claims among obstetrician-gynecologists. Obstet Gynecol. 2021;138(2):246–52. 30. Weinschreider J, Dadiz R.  Back to basics: creating a simulation program for patient safety. J Healthc Qual. 2009;31(5):29–36.