Because of the rarity and potential for extreme severity, amniotic fluid embolism (AFE) remains one of the most challenging conditions of pregnancy and childbirth to diagnose and treat.1 No pathognomonic signs can firmly support the diagnosis of AFE.2,3 Symptoms classically described include cardiovascular collapse, respiratory deficiency, clinical coagulopathy, coma and seizure, often preceded by premonitory signs.2,3 Similarly, some biological abnormalities are frequently observed: disseminated intravascular coagulation (DIC), hypoxemia, and acidosis. However, the presence of these signs and symptoms and their combination vary greatly and are not specific to AFE. Several biologic tests and histologic examinations have been proposed as diagnostic tools,4 but none can reliably confirm the diagnosis of AFE.
Consequently, there are no uniform diagnostic criteria for AFE, which remains both a clinical syndromic diagnosis and a diagnosis of exclusion. This lack of definition hampers the understanding of risk factors, diagnosis, treatment, and prognosis of this condition. Recently, a working group of the Society for Maternal-Fetal Medicine (SMFM) and the Amniotic Fluid Embolism Foundation proposed 4 uniform diagnostic criteria to identify cases of AFE for research purposes.5 This definition, based on the presence of hemodynamic and respiratory compromise, associated with strictly defined DIC, has never been tested on a specific population.
To better characterize the fatal cases of AFE and to study the relevance of the structured definition, our study had 2 objectives: first, to apply the structured definition proposed by the SMFM and the Amniotic Fluid Embolism Foundation to the cases identified by the National Experts Committee on Maternal Mortality (Enquête Nationale Confidentielle sur les Morts Maternelles) in France6 between 2007 and 2011; and second, to describe the epidemiology, clinical characteristics, and management by the presence or absence of all 4 diagnostic criteria for AFE. In addition, cases were reviewed for opportunities to improve the management of this severe condition.
This manuscript adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement. All maternal deaths due to AFE and identified through the French Confidential Enquiry into Maternal Deaths between 2007 and 2011 were included in the study population.7,8 The French Confidential Enquiry into Maternal Deaths was approved by the French Commission on Information Technology and Liberties. The use of the data is covered by this approval. The requirement for written informed consent was waived by the commission. The French Confidential Enquiry into Maternal Deaths has been conducted in France since 1996. In this permanent surveillance system, maternal death is defined as the death of a woman while pregnant or within 365 days of the end of pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not for accidental or incidental causes, in accordance with the International Classification of Diseases, 10th revision. The first step of this enquiry starts with the identification of possible maternal deaths by analyzing death certificates from the National Centre for Cause-of-Death Statistics and Epidemiology. All cases with a mention of at least 1 cause-of-death code from the International Classification of Diseases-10 obstetrics chapter, the words “pregnancy” or “puerperium” in the text, or the pregnancy check box ticked are reported to the French Confidential Enquiry into Maternal Deaths. In addition, possible maternal deaths are identified from computer-based linkage of the death register with the birth and fetal death register and from the national hospital discharge database. For each case, after consent is obtained from the certifying physician, a team of 2 assessors (an obstetrician and an anesthesiologist) interviews physicians in charge of the woman and reviews all medical charts, documents, and autopsy reports that are available on-site. The team then collects all relevant medical information about the woman and her death in a standardized anonymous questionnaire. Each report is then reviewed and collectively discussed in a plenary session of the National Experts Committee on Maternal Mortality, composed of 15 members including obstetricians, midwives, anesthesiologists, critical care and internal medicine specialists, and epidemiologists. The experts committee reaches a consensus (agreement by all the members) about the underlying cause of death, whether the death is a maternal death (as defined) and the quality of care provided (suboptimal [yes or no]). The quality of care is judged according to national clinical guidelines, published evidence, and professional experience. The experts committee established the diagnosis of AFE as the main cause of death by consensual decision based on the clinical, biological, and histological features reported.
A working group of the SMFM and the Amniotic Fluid Embolism Foundation recently proposed a structured definition of AFE for research purposes. This definition is based on the presence of 4 diagnostic criteria, all of which must be present5:
- “1. Sudden onset of cardiorespiratory arrest, or both hypotension (systolic blood pressure <90 mm Hg) and respiratory compromise (dyspnea, cyanosis, or peripheral capillary oxygen saturation <90%);
- 2. Documentation of overt DIC after appearance of these initial signs or symptoms, using the scoring system of the Scientific and Standardization Committee on DIC of the International Society on Thrombosis and Haemostasis ISTH, modified for pregnancy. Coagulopathy must be detected before loss of sufficient blood to itself account for dilutional or shock-related consumptive coagulopathy;
- 3. Clinical onset during labor or within 30 minutes of delivery of placenta; and
- 4. No fever (≥38°C) during labor.”
According to the modified ISTH scoring system for overt DIC in pregnancy, a score ≥3 is compatible with DIC in pregnancy, with platelet count >100,000/mL = 0, <100,000/mL = 1, <50,000 mL = 2; prolonged prothrombin time or international normalized ratio <25% increase = 0, 25%–50% increase = 1, >50% increase = 2; and fibrinogen level >2 g/L = 0, <2 g/L=1.5
To determine whether the deaths attributed to AFE by the national experts committee met the structured definition proposed by the SMFM and the Amniotic Fluid Embolism Foundation, we checked for the presence of the 4 proposed criteria in AFE-related maternal deaths identified by the national experts committee. Then, to better characterize cases that would be excluded by the structured definition of AFE for research studies, we described maternal, pregnancy, and delivery characteristics; clinical and biological patterns; and the use of specific laboratory tests for AFE separately for AFE-related maternal deaths with all 4 proposed criteria and those with at least 1 missing criterion.
The following baseline characteristics of the women and the pregnancy were extracted: maternal age (≤35 and >35 years), body mass index (<30 [normal weight] and ≥30 kg/m2 [obesity]), allergy, parity (0, 1–3, ≥4), and pregnancy achieved by assisted reproduction technology and multiple pregnancy (binary variables). Some characteristics of the delivery and the newborn were collected: noncephalic fetal presentation, labor, induction of labor and with prostaglandins or not, oxytocin during labor, artificial membrane rupture if spontaneous labor (binary variables), mode of delivery in 5 categories (spontaneous vaginal delivery, operative vaginal delivery, cesarean delivery before labor, cesarean delivery during labor, no delivery), newborn gender, macrosomia (≥4000 g), and neonatal death. The indications for cesarean delivery during labor were also divided into 3 categories: fetal indication only, maternal indication only, and both fetal and maternal indication.
To describe the clinical presentation of fatal AFE, premonitory signs and symptoms were recorded. Premonitory signs were classified as neurological signs (restlessness, confusion and/or agitation, numbness, tingling), fainting and/or sensation of doom, respiratory signs (dyspnea, cough, shortness of breath), and atypical signs (nausea and/or vomiting, arterial blood hypertension, skin rash, thoracic or abdominal pain). The following symptoms and their appearance as a first sign were searched: cardiovascular collapse including cardiac arrest, respiratory deficiency, seizure, and clinical coagulopathy. Clinical and biological features of the coagulopathy were described (initial with massive obstetric hemorrhage, delayed or only biological coagulopathy, platelet count, prothrombin time, fibrinogen level at the first assessment after the occurrence of the first signs or symptoms). The occurrence of abnormal fetal heart rate was also described in the context of AFE before delivery. The occurrence of these signs and symptoms was described according to the sequence of the labor, the membrane rupture, and the delivery (within 30 minutes of placental delivery or more). The time between the first symptoms to hemodynamic collapse and to maternal death was calculated.
The mention in medical files of the diagnosis of AFE by the clinician(s) in charge of the woman was collected, as was the time when this diagnosis was evoked: after the occurrence of premonitory signs or cardinal symptoms, at intensive care unit (ICU) admission, postmortem, or never mentioned. The use of specific laboratory diagnostic tests for AFE, such as the identification of amniotic fluid debris or fetal squamous cells in maternal blood and in maternal bronchoalveolar lavage fluid and measurement of plasma tryptase and insulin-like growth factor binding protein-1 levels (binary variables), and autopsy with lung anatomopathological examination were described. Results favoring AFE were collected for each specific biological test and autopsy.
Finally, components of clinical management were recorded when applicable. Among women with clinical coagulopathy, management of obstetric hemorrhage was characterized by use of sulprostone, intrauterine balloon, vascular ligation, uterine suture, hysterectomy, abdominal packing, blood transfusion of red blood cells, platelets, fresh frozen plasma, fibrinogen concentrate, and recombinant activated factor VII. Among women who survived more than 6 hours after the first signs, admission to the ICU and use of an extracorporeal life support system (ECLS) were described. Among women with cardiac arrest as a first sign during labor, the time between maternal cardiac arrest and fetal extraction was calculated.
Suboptimal care according to the national experts committee assessment was described overall and separately by obstetrics and intensive care.
The maternal mortality ratio for AFE, defined as the number of maternal deaths due to AFE per 100,000 live births, was calculated for the study period with exact Poisson 95% confidence interval (CI). Data regarding the number of live births were from the French National Institute of Statistics and Economic Studies and were used as a denominator. Maternal, pregnancy, and delivery characteristics; clinical and biological patterns of AFE; and specific biological tests were described separately for women with all proposed criteria and for those with at least 1 missing criterion according to the proposed diagnostic criteria for AFE by the SMFM and the Amniotic Fluid Embolism Foundation. Qualitative data are expressed as number (percentage) and quantitative variables as median (range). Statistical analyses were performed with Stata v13 (Stata Corporation, College Station, TX).
Between January 2007 and December 2011, among 3970, 669 live births in France, 429 maternal deaths were identified by the French Confidential Enquiry into Maternal Deaths; 39 (9.1% of all maternal deaths; 95% CI, 6.5–12.2) were attributed to AFE. The estimated maternal mortality ratio due to AFE was 0.95/100,000 live births (95% CI, 0.67–1.3/100,000). The confidential enquiry retained complete medical records for 36 of the 39 cases.
Among the 36 women who died from AFE according to the French national experts committee, 21 (58%) exhibited all 4 diagnostic criteria proposed by the SMFM and the Amniotic Fluid Embolism Foundation (Table 1). Among the 15 women with at least 1 missing criterion, laboratory documentation of overt DIC was lacking for 14: 7 women had clinical coagulopathy, but no blood test was performed in 5, and results were unavailable for 2; 6 other women presented cardiac arrest as a first sign without bleeding and never had a blood test performed; for these 11 women without a blood test performed, the period between the first symptoms of AFE and the death was ≤60 minutes. Finally, only 1 woman exhibited collapse followed by cardiac arrest without bleeding and had a blood test showing no DIC. Excluding the criterion of documented early DIC, 34 of 36 women presented all 3 of the remaining proposed criteria. Two women were missing multiple criteria: 1 presented cardiac arrest as a first sign 4 days after delivery, with autopsy favoring AFE; the other woman presented initial seizure and coagulopathy during pregnancy, with an autopsy and amniotic fluid debris or squamous cells in bronchoalveolar fluid favoring AFE.
Characteristics of the women who died from AFE, women with intra-/postpartum deaths from other causes and women from the National Perinatal Survey, a national representative sample of deliveries in France9 is shown in Supplemental Digital Content 2, Table S1, http://links.lww.com/AA/C37.
Among the 36 women who died from AFE, the maternal, pregnancy, and delivery characteristics were similar between women with all proposed diagnostic criteria and those with at least 1 missing criterion (Table 2).
Premonitory signs and symptoms of AFE occurred after membrane rupture in 34 of 36 women who died from AFE and before labor or more than 30 minutes after delivery in only 2. Premonitory signs of AFE were observed in more than three fourths of the women, regardless of whether all proposed criteria were met. If present, premonitory signs were multiple in two-thirds of the women (Table 3). The most frequent premonitory signs were neurological and fainting with or without a sensation of doom.
Collapse was observed in all cases except 1 and was a first sign in half of the cases (Table 3). The proportion of cardiac arrest as a first sign was similar in both groups. The proportion of clinical coagulopathy was lower, and if present, more often in the form of delayed hemorrhage in women with missing criteria than in women with all proposed criteria.
A complete standard blood test was performed in two-thirds of the women and at a median time of 65 minutes after the first signs (n = 17; range, 15 minutes to 23 hours 11 minutes) but in only 3 of the 15 women with missing criteria (Table 3). The most frequent initial coagulation abnormalities were fibrinogen level <2 g/L (n = 20) and prothrombin time >50% above normal (n = 19).
For 23 women, the first AFE sign occurred before delivery: abnormal fetal heart rate tracings were the first sign in 5 women and were associated with maternal signs in 9 others. In 8 other cases, the mother had a cardiac arrest and fetal heart rate was not measured. Fetal heart rate was considered normal in only 1 case.
Mention of AFE diagnosis by clinicians in charge of the women was available for 33 of the 36 women, including 4 first mentioned at ICU admission and 5 diagnosed postmortem. Among the 28 women with suspected AFE before death, specific biological tests for AFE were performed in 23, with similar proportions for women with and without all proposed criteria (Supplemental Digital Content 2, Table S2, http://links.lww.com/AA/C37). Tryptase activity was measured for 11 women: 9 presented low activity and 2 moderately increased activity, thereby excluding a diagnosis of anaphylactic shock.
Among the 15 women with at least 1 missing criterion, postpartum hemorrhage could have been selected as the primary diagnosis for 2 women: 1 presented with severe postpartum hemorrhage with coagulopathy but preceded by maternal fainting at the time of membrane rupture and cesarean delivery for abnormal fetal heart rate; the other had new-onset postpartum hemorrhage associated with convulsions. In a third woman, local anesthetic intoxication was recorded by clinicians, but local anesthetic concentration was not consistent with this diagnosis. Pulmonary thromboembolism could be an alternative diagnosis in 5 women presenting sudden onset cardiorespiratory arrest or collapse and respiratory compromise during labor, but biologic tests and autopsy favored AFE in 3 of the women, and for 2 others, symptoms were preceded by premonitory signs consistent with AFE. No case had evidence to support the alternative diagnosis of septic shock or anaphylactic shock.
Among the 30 women who developed obstetric hemorrhage, sulprostone was administered and hysterectomy performed in 21 and 18 women, respectively (Supplemental Digital Content 2, Table S3, http://links.lww.com/AA/C37). In total, 29 of 30 women with obstetric hemorrhage received a blood transfusion, with a median of 10 units of red blood cells (range, 0–22) and 11 units of fresh frozen plasma (range, 0–27). One woman with obstetric hemorrhage and coagulopathy in the context of fatal AFE did not receive any blood product.
The experts considered care as suboptimal for 20 of the 36 women (Table 4). Indicated hysterectomy was not performed or performed with delay in 13 women. Intensive care was considered suboptimal for 6 women, mainly because of delayed and/or insufficient blood transfusion. Overall, 30 of 36 women died within the first 24 hours of symptoms, including 18 within the first 6 hours. ECLS was performed in 6 of 18 women who survived for more than 6 hours and was not considered in the other cases. Among the 13 women with cardiac arrest occurring before the delivery, perimortem cesarean (n = 9) or instrumental vaginal delivery (n = 3) was performed in all but 1 case, including 5 performed within 5 minutes after maternal cardiac arrest, 1 between 5 and 10 minutes, and 6 delayed more than 10 minutes. Four neonatal deaths were reported: for all 4, fetal extraction was performed more than 5 minutes after maternal cardiac arrest.
This national study found that AFE accounted for 9.1% (95% CI, 6.5–12.2) of maternal mortality cases in France. Less than two-thirds of the women who died from AFE as identified by the national experts committee exhibited all 4 criteria proposed by the SMFM and Amniotic Fluid Embolism Foundation for AFE diagnosis in research studies. Documented DIC was the most commonly missing criterion, and was the only missing criteria for all but 2 women. According to the experts, the most frequent opportunities to improve future care include timely perimortem fetal extraction, timely hysterectomy, and adequate blood transfusion.
On assessing the level of agreement between AFE-related maternal deaths defined according to the national experts committee and the structured definition proposed by the SMFM and the Amniotic Fluid Embolism Foundation,5 we found that all the proposed criteria were present in all typical cases of AFE, except for documented overt DIC detected before severe blood loss, which was missing in 14 of the 36 cases. Rapid coagulation tests to permit rapid and early diagnosis of DIC are recommended in cases of suspected AFE to help establish the diagnosis. Nevertheless, in extremely severe presentations, clinicians focus on resuscitation. In the context of cardiac arrest, a blood sample could be impossible to obtain. Consequently, the performance of an initial blood test is often delayed. We agree that coagulopathy is a major criterion of AFE, but the requirement for early laboratory documentation of the coagulopathy may exclude typical cases of AFE from research studies. Consequently, we suggest expanding the criteria of early documented DIC proposed by the SMFM to early clinical coagulopathy with bleeding.
Indeed, as suggested by members of the working group who proposed this structured definition, AFE still remains a clinical diagnosis. Moreover, the use of a scoring system for DIC in pregnancy based on a modification of the DIC score of the ISTH for AFE diagnosis has been challenged.10 With an expanded definition that includes clinical coagulopathy, 31 of our 36 cases satisfy the definition for AFE-related maternal death. Of the remaining 5 cases, 3 had sudden cardiac arrest during labor without bleeding and autopsy favoring AFE. Another possibility is even to exclude the coagulopathy criterion in maternal death from AFE, as proposed in the definition based on experts’ consensus by the International Network of Obstetric Survey System.11 In this consensus, AFE is defined as “cardiorespiratory collapse within 6 hours after labor, delivery or ruptured membranes, with no other identifiable cause, followed by acute coagulopathy in those who survive the initial event.”
Neurological signs are not included in the definition proposed by the SMFM, but were present in 9 of 15 women who did not meet the definition, and in 8 of 21 who met the definition for AFE. Symptoms of restlessness, confusion, agitation, or numbness have traditionally been considered to support a diagnosis of AFE.
Not all clinical teams diagnosed AFE, even when the initial clinical presentation was typical and occurred in women with known risk factors for AFE. Management of severe AFE is largely supportive,3 but the early recognition of this diagnosis may allow the team to anticipate patient needs and to optimize care for this very severe complication of childbirth.12 This recognition could be crucial, because in the severe cases, we observed progression toward extreme severity is usually very rapid, and sometimes cardiac arrest occurs just a few minutes after premonitory signs. In particular, anticipating major obstetric hemorrhage observed in the context of AFE is certainly essential in optimizing patient care, as it would help to improve transfusion management by reducing delay in transfusion of blood products and probably in invasive surgical treatments such as hysterectomy.
In our study of AFE-related maternal deaths, the quality of care was considered suboptimal by the national experts committee in more than half of the cases. Delayed hysterectomy or the absence of hysterectomy although indicated was the most frequent suboptimal care, even though vascular ligation and/or uterine suture were performed in some women who died from AFE. In addition, the transfusion strategy was sometimes considered insufficient and/or delayed, consistent with previous reports of severe obstetric hemorrhage.13 The use of timely hysterectomy, a massive transfusion protocol, and coagulation monitoring should be encouraged in these contexts.14
Perimortem fetal extraction was performed in all but 1 gravid mother with cardiac arrest, but the 5-minute window between cardiac arrest and delivery was achieved in only one-third of cases. Even if the importance of timing with postmortem cesarean delivery to improve maternal hemodynamic status and minimize neurological damage is consistently mentioned in guidelines,15 delay in fetal extraction is classically reported in this context.16 Clinicians are very rarely confronted with this situation17; consequently, health care providers in maternity units should be proactively and regularly trained to manage maternal cardiac arrest, particularly to perform cesarean delivery timely. Recently, some authors have proposed to switch from “perimortem cesarean” to “resuscitative hysterotomy,” to insist in the crucial role of timely fetal extraction for both the mother and the fetus.18
Finally, in our study population, ECLS was used for one-third of women who died more than 6 hours after the first AFE symptoms and was not considered in the other cases. Even if ECLS is not routinely recommended for management of AFE,3 guidelines on adult advanced cardiovascular life support suggest that ECLS “may be considered for select cardiac arrest patients for whom the suspected etiology of the cardiac arrest is potentially reversible during a limited period of mechanical cardiorespiratory support.”19 Consequently, this treatment strategy should be discussed for every woman with suspected AFE and in-hospital cardiac arrest or cardiopulmonary collapse refractory to advanced life support medication and intervention.
Maternal death and AFE are both rare events, so while our study population is small, it is in the same order of magnitude of other contemporary cases series with detailed data of women with AFE-related maternal death.20–23 In our national population–based study, data were extracted from a national confidential enquiry system and not from an administrative hospital database using routine codes. AFE was diagnosed by consensus of the national experts committee, minimizing selection bias and overreporting of AFE-related maternal deaths, as opposed to data from population databases.24 Comparison with women presenting nonfatal AFE would have allowed for exploring characteristics and procedures associated with better prognosis, but similar detailed data on all women with AFE are not yet available in France. Still we were able to identify women who would be excluded from research studies with use of this proposed definition and to identify criteria that should be used to improve this definition. Another limitation of this study is the time that elapsed since the study period. Practices may have evolved since the data collection, but clinical patterns of severe AFE should not differ over time.3
In conclusion, although rare, AFE is responsible for 1 in 10 maternal deaths in France. Early blood testing is indicated to establish the diagnosis and to anticipate the need for massive blood transfusion and hemostatic support. Nevertheless, early testing may not be realistic in extremely severe cases, suggesting that the diagnostic criteria for AFE should not rely on early laboratory confirmation of coagulopathy. There are still opportunities to improve future care including timely perimortem fetal extraction, timely hysterectomy, and adequate blood transfusion.
We thank the medical assessors of the national confidential enquiry into maternal deaths who collected all the information on maternal deaths and the members of the National Expert Committee on Maternal Mortality. We thank Mrs Béatrice Blondel, PhD, from the INSERM UMR 1153, Obstetrical, Perinatal and Pediatric Epidemiology Research Team (EPOPé), Centre for Epidemiology and Statistics Sorbonne Paris Cité, Paris, France, for her comments on the manuscript. We are also grateful to Mrs Laura Smales from BioMedEditing, Toronto, Canada, for her help in editing the manuscript.
Name: Marie-Pierre Bonnet, MD, PhD.
Contribution: This author helped design the study and interpret the results, perform the statistical analysis, and write the first draft of the manuscript and the revised versions.
Conflicts of Interest: None.
Name: Diane Zlotnik, MD.
Contribution: This author helped collect the data, interpret the results, write the first draft of the manuscript, perform the statistical analysis, and approve the final manuscript.
Conflicts of Interest: None.
Name: Monica Saucedo, MD, PhD.
Contribution: This author helped collect and interpret the data, perform some statistical analysis and made significant revisions to the manuscript, and approved the final manuscript.
Conflicts of Interest: M. Saucedo is the statistical coordinator of the French Confidential Enquiry into Maternal Deaths.
Name: Dominique Chassard, MD, PhD.
Contribution: This author helped interpret the results and made significant revisions to the manuscript and approved the final manuscript.
Conflicts of Interest: Dominique Chassard is a member of the Experts committee of the French Confidential Enquiry into Maternal Death and is in charge of the specific analysis on maternal deaths due to amniotic fluid embolism.
Name: Marie-Hélène Bouvier-Colle, PhD.
Contribution: This author helped design the study and interpret the results. This author made significant revisions to the manuscript and approved the final manuscript.
Conflicts of Interest: M.-H. Bouvier-Colle used to be the scientific coordinator of the French Confidential Enquiry into Maternal Deaths and a member of the Experts committee.
Name: Catherine Deneux-Tharaux, MD, PhD.
Contribution: This author helped design the study, interpret the results, and made significant revisions to the manuscript.
Conflicts of Interest: C. Deneux-Tharaux is the present scientific coordinator of the French Confidential Enquiry into Maternal Deaths and a member of the Experts committee.
This manuscript was handled by: Jill M. Mhyre, MD.
1. Sultan P, Seligman K, Carvalho B. Amniotic fluid embolism: update and review. Curr Opin Anaesthesiol. 2016;29:288–296.
2. Clark SL. Amniotic fluid embolism. Obstet Gynecol. 2014;123:337–348.
3. Pacheco LD, Saade G, Hankins GD, Clark SL; Society for Maternal-Fetal Medicine (SMFM), with the assistance of Amniotic fluid embolism: diagnosis and management. Am J Obstet Gynecol. 2016;215:B16–B24.
4. Benson MD. Current concepts of immunology and diagnosis in amniotic fluid embolism. Clin Dev Immunol. 2012;2012:946576.
5. Clark SL, Romero R, Dildy GA, et al. Proposed diagnostic criteria for the case definition of amniotic fluid embolism in research studies. Am J Obstet Gynecol. 2016;215:408–412.
7. Saucedo M, Deneux-Tharaux C, Bouvier-Colle MH; French National Experts Committee on Maternal Mortality. Ten years of confidential inquiries into maternal deaths in France, 1998-2007. Obstet Gynecol. 2013;122:752–760.
8. Saucedo M, Deneux-Tharaux C, Bouvier-Colle MH; Le Comité national d’experts sur la mortalité maternelle. [Maternal mortality in France, 2007-2009]. J Gynecol Obstet Biol Reprod (Paris). 2013;42:613–627.
9. Blondel B, Lelong N, Kermarrec M, Goffinet F; National Coordination Group of the National Perinatal Surveys. Trends in perinatal health in France from 1995 to 2010. Results from the French National Perinatal Surveys. J Gynecol Obstet Biol Reprod (Paris). 2012;41:e1–e15.
10. Erez O. Proposed diagnostic criteria for the case definition of amniotic fluid embolism in research studies. Am J Obstet Gynecol. 2017;217:228–229.
11. Schaap T, Bloemenkamp K, Deneux-Tharaux C, Knight M, Langhoff-Roos J, Sullivan E, van den Akker T; on behalf of INOSS. Defining definitions: a Delphi study to develop a core outcome set for conditions of severe maternal morbidity. BJOG. 2017 Jul 29 [Epub ahead of print].
12. Gist RS, Stafford IP, Leibowitz AB, Beilin Y. Amniotic fluid embolism. Anesth Analg. 2009;108:1599–1602.
13. Bonnet MP, Deneux-Tharaux C, Dupont C, Rudigoz RC, Bouvier-Colle MH. Transfusion practices in postpartum hemorrhage: a population-based study. Acta Obstet Gynecol Scand. 2013;92:404–413.
14. Pacheco LD, Saade GR, Costantine MM, Clark SL, Hankins GD. An update on the use of massive transfusion protocols in obstetrics. Am J Obstet Gynecol. 2016;214:340–344.
15. Jeejeebhoy FM, Zelop CM, Lipman S, et al.; American Heart Association Emergency Cardiovascular Care Committee, Council on Cardiopulmonary, Critical Care, Perioperative and Resuscitation, Council on Cardiovascular Diseases in the Young, and Council on Clinical Cardiology. Cardiac arrest in pregnancy: a scientific statement from the American Heart Association. Circulation. 2015;132:1747–1773.
16. Jeejeebhoy FM, Zelop CM, Windrim R, Carvalho JC, Dorian P, Morrison LJ. Management of cardiac arrest in pregnancy: a systematic review. Resuscitation. 2011;82:801–809.
17. Mhyre JM, Tsen LC, Einav S, Kuklina EV, Leffert LR, Bateman BT. Cardiac arrest during hospitalization for delivery in the United States, 1998-2011. Anesthesiology. 2014;120:810–818.
18. Rose CH, Faksh A, Traynor KD, Cabrera D, Arendt KW, Brost BC. Challenging the 4- to 5-minute rule: from perimortem cesarean to resuscitative hysterotomy. Am J Obstet Gynecol. 2015;213:653–6, 653.e1.
19. Perkins GD, Jacobs IG, Nadkarni VM, et al.; Utstein Collaborators. Cardiac arrest and cardiopulmonary resuscitation outcome reports: update of the Utstein Resuscitation Registry Templates for Out-of-Hospital Cardiac Arrest: a statement for healthcare professionals from a task force of the International Liaison Committee on Resuscitation (American Heart Association, European Resuscitation Council, Australian and New Zealand Council on Resuscitation, Heart and Stroke Foundation of Canada, InterAmerican Heart Foundation, Resuscitation Council of Southern Africa, Resuscitation Council of Asia); and the American Heart Association Emergency Cardiovascular Care Committee and the Council on Cardiopulmonary, Critical Care, Perioperative and Resuscitation. Circulation. 2015;132:1286–1300.
20. Fitzpatrick KE, Tuffnell D, Kurinczuk JJ, Knight M. Incidence, risk factors, management and outcomes of amniotic-fluid embolism: a population-based cohort and nested case-control study. BJOG. 2016;123:100–109.
21. Main EK, McCain CL, Morton CH, Holtby S, Lawton ES. Pregnancy-related mortality in California: causes, characteristics, and improvement opportunities. Obstet Gynecol. 2015;125:938–947.
22. McDonnell N, Knight M, Peek MJ, et al.; the Australasian Maternity Outcomes Surveillance System (AMOSS). Amniotic fluid embolism: an Australian-New Zealand population-based study. BMC Pregnancy Childbirth. 2015;15:352.
23. Tanaka H, Katsuragi S, Osato K, et al. Value of fibrinogen in cases of maternal death related to amniotic fluid embolism. J Matern Fetal Neonatal Med.
24. Knight M, Berg C, Brocklehurst P, et al. Amniotic fluid embolism incidence, risk factors and outcomes: a review and recommendations. BMC Pregnancy Childbirth. 2012;12:7.