Maternal mortality is a key indicator of the health status for women of reproductive age and is a marker of the performance of a country's health system, reflecting both access to and quality of prenatal and obstetric care.1,2 Remarkable progress has been made globally, with an important decrease in maternal deaths over the past 30 years.2
However, success in preventing maternal deaths never can be considered definitive. Recent reports describe an increase in the incidence of maternal mortality in the United States and the Netherlands.3,4 These findings may have various explanations, such as improvement in maternal mortality measurement through better case identification and characterization, or a real increase in the number of maternal deaths attributable to either a higher proportion of at-risk pregnant women or impaired access to and quality of obstetric care, or both.
France presents an interesting opportunity to examine this trend and to explore these hypotheses. As in most high-resource countries, the profile of pregnant women in France has changed over the past 15 years, with an increased prevalence of maternal age 35 years or older (from 15% to 19%), obesity (from 6% to 10%), and cesarean delivery (from 17.5% to 21%).5 Because these characteristics have been described as risk factors for maternal mortality,6–9 an increase in maternal mortality might be expected. However, France is characterized by a model of health care organization based on the principle of universal health coverage; factors related to access to care are expected to have little influence because prenatal and perinatal care is free and should be provided to all pregnant women. Because it is one of the largest European Union countries, and one with the most dynamic demographic indicators,10 the number of maternal deaths is large enough for useful analyses. The French Confidential Enquiry Into Maternal Deaths (Enquête Nationale Confidentielle sur les Morts Maternelles)11 provides reliable data regarding maternal mortality.
Our objective was to assess the trends over time in the ratio, causes, risks factors, quality of care, and avoidability of maternal deaths in France from 1998 through 2007.
MATERIALS AND METHODS
Since 1996, France has conducted confidential enquiries into maternal deaths, defined by French Confidential Enquiry Into Maternal Deaths as the death of a woman while pregnant or within 365 days of the end of pregnancy, irrespective of the duration and the site of the pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes, in accordance with the International Classification of Disease, 10th Revision (ICD-10), definition of maternal and late maternal deaths.12
The identification of possible maternal deaths begins with the analysis of death certificates of women of reproductive age conducted by the National Center for Cause-of-Death Statistics and Epidemiology. All death certificates were completed and signed by a certifying physician. Deaths of women with at least one cause-of-death code in the ICD-10, obstetric chapter (ICD-10, O00-O99), or any mention of pregnancy or puerperium in the text, or if the pregnancy check box was ticked (check box indicating a death during or within 1 year of pregnancy, introduced in 2000) were reported to the French Confidential Enquiry Into Maternal Deaths. Accidental deaths (road traffic accidents, burns, and murders) were excluded.
The certifying physician must formally consent before the enquiry can begin. Then, two assessors (an obstetrician and an anesthesiologist) not involved with the case before then visit the site to collect the relevant clinical information about the woman (her social and demographic characteristics, pre-existing conditions, and characteristics of her pregnancy, prenatal care, and delivery) and her death (circumstances and management of the morbid event and its fatal outcome). The assessors use a standardized anonymous questionnaire completed through interviews of clinicians and a review of the medical records and autopsy reports.
Each case is then anonymously reviewed in a plenary session of the National Experts Committee on Maternal Mortality, which is composed of 15 members including obstetricians, midwives, anesthesiologists, critical care and internal medicine specialists, and epidemiologists. The expert committee reaches a consensus (agreement by all the members) about the underlying cause of death, whether the death is a maternal death (as defined), the quality of care provided (suboptimal [yes or no]), and whether the death was avoidable. The quality of care is judged according to national clinical guidelines, published evidence, and professional experience. A maternal death is considered avoidable if one or more changes in the care provided or in the patient behavior might have averted the fatal outcome. The factors contributing to avoidable deaths are categorized as erroneous or missed diagnoses, delayed management, inadequate management, and poor maternal compliance with medical treatment or advice.
For this study, we included all maternal deaths identified through the French Confidential Enquiry Into Maternal Deaths between 1998 and 2007 for continental France (N=660). The overseas districts (Guyana, Martinique, Guadeloupe, and Reunion) were not included in this analysis, because death certificates from these regions were not collected centrally until 2001, and enquiries into maternal deaths there began only that year. A previous report analyzed maternal mortality in the overseas districts.13
The maternal mortality ratio, defined as the number of maternal deaths per 100,000 live births, was calculated for maternal deaths within 365 days of pregnancy end,14 within 42 days of pregnancy end,12 and between 43 days and 365 days of pregnancy end (late maternal deaths). Data regarding the number of live births come from the National Institute of Statistics and Economic Studies. The French Confidential Enquiry Into Maternal Deaths has been approved by the Commission Nationale de l’Informatique et des Libertés (National Data Protection Authority).
We analyzed time trends in maternal mortality over two 5-year periods (1998–2002 and 2003–2007) for ratios, both overall and by women's characteristics (age and nationality) and causes among all maternal deaths (N=660). Suboptimal care and avoidability were assessed among fully documented cases (n=392), excluding cases for which the certifying physician did not provide consent to conduct the enquiry (n=166) and cases for which the collected information was insufficient for such a judgment (n=102). Proportions were compared with χ2 tests or, when appropriate, Fisher exact tests. Statistical analyses were performed with Stata 10 software.
The French Confidential Enquiry Into Maternal Deaths identified 660 maternal deaths during the 10-year study period. Table 1 shows the mortality ratios by period; the overall maternal mortality ratio was 8.8 per 100,000 live births (95% CI 7.8–9.8) for 1998–2002 and 8.4 per 100,000 live births (95% CI 7.6–9.4) for 2003–2007 (no significant difference). Late maternal deaths (between 43 days and 365 days of pregnancy end) accounted for 8% (27/332) and 7% (24/328) of all maternal deaths identified by the enhanced system for each period.
Most maternal deaths, 78.9% (521/660), occurred in public hospitals; 9.9% occurred in private hospitals and 11.2% occurred at home or in an outdoor place. Again, these proportions did not change significantly over time. Information about pregnancy outcome was available for 628 of the maternal deaths; of these, 14.2% occurred during pregnancy (undelivered), 4% occurred after a spontaneous or induced abortion, 4.1% occurred after an ectopic pregnancy, and 77.7% occurred after a birth (live or stillbirth). None of these characteristics differed significantly between the two periods.
The mean age was 32.8 years (standard deviation 6.0) for 1998–2002 and 32.4 years (standard deviation 6.3) for 2003–2007. During both periods, maternal mortality increased with maternal age, with the lowest maternal mortality ratios observed in women aged 20–24 years and 25–29 years; significantly higher ratios were observed for women aged 30 years or older compared with those aged 25–29 years (Table 2). The distribution of maternal age among maternal deaths did not differ significantly between the two periods. The risk of maternal mortality was approximately three times higher among women of sub-Saharan African nationality than among French women (Table 2). Maternal mortality ratio by nationality groups did not change significantly over time. The change in the distribution of marital status among maternal deaths between the two periods mirrored the trend observed among all live births.15
Table 3 presents the distribution of causes of maternal deaths for 1998–2002 and 2003–2007, as assessed by the expert committee. There was no statistically significant change in either the distribution of causes or cause-specific maternal mortality ratios between the two periods. During both periods, approximately two thirds of the women died from direct obstetric causes. The percentages of hemorrhages, hypertensive disorders, and thromboembolisms decreased slightly over time, whereas the proportion of deaths attributable to amniotic fluid embolisms, cardiovascular conditions, and cerebrovascular accidents tended to increase, although those changes were not statistically significant. For the 10 years studied, hemorrhages remained the leading cause of maternal mortality (18% of deaths), accounting for 1.5 maternal deaths per 100,000 live births, followed by, among direct causes, amniotic fluid embolism (12%), thromboembolism (11%), and hypertensive disorders (10%), each with a maternal mortality ratio between 0.8 and 1.1. The maternal mortality ratio from indirect causes was similar during both periods; cardiovascular conditions and cerebrovascular accidents were the leading indirect causes and did not change significantly over time. The number of aortic dissections among the cardiovascular conditions causing death increased from 1 in 1998–2002 to 11 in 2003–2007. In the latter period, three of the 11 women who died from this condition were known to have genetic conditions placing them at risk, including Turner syndrome, Marfan syndrome, and Von Recklinghausen syndrome.
Of the 51 late maternal deaths, 18 were attributable to direct obstetric causes (two hemorrhages, two amniotic fluid embolisms, four hypertensive disorders, three thromboembolisms, one infection, five peripartum cardiomyopathies, and one other cause) and 33 were attributable to indirect causes (seven cardiovascular conditions, seven cerebrovascular accidents, four cases of cancer, and 15 other conditions).
The experts considered that they had sufficient information available to assess the quality of care for 67% (222/332) of the maternal deaths in 1998–2002 and 51% (168/328) of the maternal deaths in 2003–2007. The cases for which clinical information was sufficient were similar to the cases partially documented in terms of distribution of causes of death and of maternal age and nationality (data not shown). Table 4 shows the prevalence of suboptimal care among maternal deaths by specific causes of death. Overall, the committee found that suboptimal care decreased from 70% in 1998–2002 to 60% in 2003–2007 (P<.03). This difference was mainly attributable to the decrease in suboptimal care in the deaths attributable to indirect causes. Suboptimal care was most prevalent (approximately 80–90%) for specific leading causes of death, particularly hemorrhages and hypertensive disorders, as well as in less common subgroups, such as infection.
Adequate information was available to determine the avoidability of 224 maternal deaths for 1998–2002 and 168 maternal deaths for 2003–2007. Of these, 49.1% and 47.6% were considered avoidable. This difference was not statistically significant. Table 5 presents the proportion of avoidable deaths and their contributory factors by cause of death over the course of 10 years. Overall, almost half of the maternal deaths were determined to be avoidable, and this proportion was significantly higher for deaths attributable to direct causes (58%) compared with indirect causes (31.7%) (P<.001). The percentage of avoidable deaths varied substantially according to the cause of death: more than 90% of deaths were from hemorrhages and only 8% of those were from amniotic fluid embolisms.
Avoidable deaths usually had two or three contributory factors (Table 5). Among deaths from hemorrhage, the leading factors were inadequate management (eg, undertransfusion in cases of severe postpartum hemorrhage; 79.7%) and delay in management (eg, delayed haemostatic hysterectomy; 60.8%). Erroneous or missed diagnosis was more frequent among avoidable maternal deaths from thromboembolism (eg, lack of diagnostic examinations in presence of suggestive signs of pulmonary embolism in a postpartum context; 38.5%). In some cases, women's own decisions or actions, described as “poor maternal compliance with medical treatment or advice,” were considered to contribute to the death and its avoidability. This was reported more often for the indirect causes of avoidable deaths and for hypertensive disorders (26.6% and 26.1%, respectively; eg, patient with known pulmonary artery hypertension becoming pregnant despite medical contraindication).
The profile of maternal mortality in France, assessed through the enhanced surveillance system, remained unchanged from 1998 to 2007, despite the increase in the prevalence of maternal risk factors. Nonetheless, further improvement appears possible. Maternal deaths remain more frequent in some subgroups of immigrant women. In addition, approximately half of all maternal deaths and nearly all deaths attributable to hemorrhage are still considered avoidable; their most frequent contributory factor remains inadequate management.
The fact that maternal mortality did not increase is not likely a result of poorer measurement, because the French Confidential Enquiry Into Maternal Deaths enhanced method has not changed during the 10-year study period. The introduction of the check box for “during pregnancy or delivery” on death certificates in 2000 may have resulted in increasing the maternal mortality ratio by enabling us to identify maternal deaths not previously identified, but this did not occur. Nonetheless, to determine the existence of maternal deaths unidentified through the current French Confidential Enquiry Into Maternal Deaths method,16 the next step needs to be systematic legal authorization to link computerized birth certificates with the computerized death certificates of women of childbearing age and with the national hospital discharge database. This strategy has shown its usefulness in the United States and elsewhere.3,17–19
The increasing prevalence of risk factors among pregnant women in France5 might have been expected to lead to an increase in the maternal mortality ratio. We can hypothesize that the health care system was able to compensate, in part, for the increase in at-risk births. The higher maternal mortality rate among women from sub-Saharan Africa was explored in a previous study20 that indicated that the excess risk mainly concerned hypertensive disorders and infections, which are conditions that require good interaction between the health care system and patients for correct treatment. Moreover, suboptimal care was more frequent for non-French women compared with French women.20 However, the reduction (albeit not significant) in the gap in the maternal mortality ratio between women from sub-Saharan African and French women during the past 5-year period may be the beginning of a downward trend. This situation seems different from that in the United States, where the differences in maternal mortality ratios between ethnic groups have not stopped growing.3
Compared with other high-resource countries with enhanced surveillance systems (Table 6), the maternal mortality ratio in France is lower than it is in the United Kingdom,21 the Netherlands,4 or the United States,3 and it is comparable with the ratio in Canada19 or Australia.22 Moreover, maternal mortality in France has not increased as it has in other countries such as the United States or the Netherlands.3,4
However, the French maternal mortality profile shows that progress remains possible. Nearly half of all maternal deaths remain avoidable. The proportion of avoidability is highest for the most frequent causes, such as hemorrhages and hypertensive complications, which are also the principal components of severe maternal morbidity. Although avoidability findings are sometimes difficult to compare because the definition depends on a country's wealth and medical culture, the proportion of avoidable deaths seems higher than that in studies conducted in North Carolina (40% according to the Pregnancy-Related Mortality Review Committee)23 and in New Zealand (37% according to the Maternal Mortality Review Working Committee).24 Concepts of quality of care in obstetrics and intensive care are more universal. The latest results from the United Kingdom and the Netherlands showed proportions of suboptimal care (61% and 55%, respectively) that were, on the whole, similar to the 60% observed in France between 2003 and 2007.4,21
More than one third of the maternal deaths could not be fully documented. For these cases, the information is limited to social and demographic data and the cause of death extracted from the death certificate. Therefore, analysis of the quality of care and analysis of avoidability are not possible. Nonetheless, maternal characteristics are similar among included and excluded women, suggesting that women not included in this analysis did not differ from those for whom care and avoidability were assessed. If selection bias does exist, it is more likely to result in an underestimation of suboptimal care and avoidability than the opposite. Such bias is not likely to affect our conclusions.
Compared with other high-resource countries with enhanced surveillance systems (Table 6), the French profile of maternal mortality causes is characterized by hemorrhages (mainly atonic postpartum hemorrhages) and distinctly fewer deaths caused by hypertensive disorders than in the Netherlands or the United States. Cardiovascular diseases also were less frequent than in the United States, the United Kingdom, or Canada, where they are the leading cause of death. This might be explained ,in part, by the lower prevalence of cardiovascular risk factors and especially by the lower prevalence of obesity in France compared with these countries (United States United Kingdom, Canada),25,26 where there is a higher incidence of cardiovascular disease in the general population.27 Moreover, the free and systematic prenatal monitoring provided in France can affect the management of women with pre-existing diseases.
The report from the French Confidential Enquiry Into Maternal Deaths regarding the particular importance of maternal mortality attributable to postpartum hemorrhage in France has triggered the development of the first national clinical practice guidelines for postpartum hemorrhage management launched in 2004;28 these findings also stimulated the design of research programs to identify the factors, both individual and related to the organization of health care, that are associated with severe hemorrhages.29–33 The results also should contribute to changes in its management. The expected effect of this mobilization is a reduction in maternal mortality attributable to hemorrhages, but it is probably still too early for any measurable effect to be visible in the data analyzed here.
1. Atrash HK, Alexander S, Berg CJ. Maternal mortality in developed countries: not just a concern of the past. Obstet Gynecol 1995;86(4 Pt 2):700–5.
2. Hogan MC, Foreman KJ, Naghavi M, Ahn SY, Wang M, Makela SM, et al.. Maternal mortality for 181 countries, 1980-2008: a systematic analysis of progress towards Millennium Development Goal 5. Lancet 2010;375:1609–23.
3. Berg CJ, Callaghan WM, Syverson C, Henderson Z. Pregnancy-related mortality in the United States, 1998 to 2005. Obstet Gynecol 2010;116:1302–9.
4. Schutte JM, Steegers EA, Schuitemaker NW, Santema JG, de Boer K, Pel M, et al.. Rise in maternal mortality in the Netherlands. BJOG 2009;117:399–406.
5. Blondel B, Lelong N, Kermarrec M, Goffinet F. 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:e1–15.
6. Callaghan WM, Berg CJ. Pregnancy-related mortality among women aged 35 years and older, United States, 1991–1997. Obstet Gynecol 2003;102(5 Pt 1):1015–21.
7. Cevik B, Ilham C, Orskiran A, Colakoglu S. Morbid obesity: a risk factor for maternal mortality. Int J Obstet Anesth 2006;15:263–4.
8. Deneux-Tharaux C, Carmona E, Bouvier-Colle MH, Breart G. Postpartum maternal mortality and cesarean delivery. Obstet Gynecol 2006;108(3 Pt 1):541–8.
9. Clark SL, Belfort MA, Dildy GA, Herbst MA, Meyers JA, Hankins GD. Maternal death in the 21st century: causes, prevention, and relationship to cesarean delivery. Am J Obstet Gynecol 2008;199:36 e1–5.
10. Eurostat. Demography Report 2010—Older, more numerous and diverse Europeans. Luxembourg: European Commission; 2011.
12. World Health Organization. ICD 10: International statistical classification of diseases and related health problems. 10th revision. Geneva (Switzerland): World Health Organization; 1992.
13. Saucedo M, Deneux-Tharaux C, Bouvier-Colle MH. Understanding regional differences in maternal mortality: a national case-control study in France. BJOG 2011;119:573–81.
14. Atrash HK, Rowley D, Hogue CJ. Maternal and perinatal mortality. Curr Opin Obstet Gynecol 1992;4:61–71.
16. Gissler M, Berg C, Bouvier-Colle MH, Buekens P. Methods for identifying pregnancy-associated deaths: population-based data from Finland 1987-2000. Paediatr Perinat Epidemiol 2004;18:448–55.
17. MacKay AP, Berg CJ, Duran C, Chang J, Rosenberg H. An assessment of pregnancy-related mortality in the United States. Paediatr Perinat Epidemiol 2005;19:206–14.
18. Deneux-Tharaux C, Berg C, Bouvier-Colle M-H, Gissler M, Harper M, Nannini A, et al.. Underreporting of pregnancy-related mortality in the United States and Europe. Obstet Gynecol 2005;106:684–92.
19. Lisonkova S, Liu S, Bartholomew S, Liston RM, Joseph KS. Temporal trends in maternal mortality in Canada II: estimates based on hospitalization data. J Obstet Gynaecol Can 2011;33:1020–30.
20. Philibert M, Deneux-Tharaux C, Bouvier-Colle MH. Can excess maternal mortality among women of foreign nationality be explained by suboptimal obstetric care? BJOG 2008;115:1411–8.
21. Lewis G. Centre for maternal and child health enquiries. Saving mothers' lives: reviewing maternal deaths to make motherhood safer: 2006-2008. The eighth report of the confidential enquiries into maternal deaths in the United Kingdom. BJOG 2011;118(suppl 1):1–203.
22. Sullivan EA, Hall B, King JF. Maternal deaths in Australia 2003–2005. Maternal deaths series no. 3. Cat. no. PER 42. Sydney (Australia): AIHW National Perinatal Statistics Unit; 2007.
23. Berg CJ, Harper MA, Atkinson SM, Bell EA, Brown HL, Hage ML, et al.. Preventability of pregnancy-related deaths: results of a state-wide review. Obstet Gynecol 2005;106:1228–34.
24. Farquhar C, Sadler L, Masson V, Bohm G, Haslam A. Beyond the numbers: classifying contributory factors and potentially avoidable maternal deaths in New Zealand, 2006-2009. Am J Obstet Gynecol 2011;205:331 e1–8.
26. Knight M, Kurinczuk JJ, Spark P, Brocklehurst P. Extreme obesity in pregnancy in the United Kingdom. Obstet Gynecol 2010;115:989–97.
27. Bhopal RS, Rafnsson SB, Agyemang C, Fagot-Campagna A, Giampaoli S, Hammar N, et al.. Mortality from circulatory diseases by specific country of birth across six European countries: test of concept. Eur J Public Health 2012;22:353–9.
28. Goffinet F, Mercier F, Teyssier V, Pierre F, Dreyfus M, Mignon A, et al.. [Postpartum haemorrhage: recommendations for clinical practice by the CNGOF (December 2004)]. Gynecol Obstet Fertil 2005;33:268–74.
29. Audureau E, Deneux-Tharaux C, Lefevre P, Brucato S, Morello R, Dreyfus M, et al.. Practices for prevention, diagnosis and management of postpartum haemorrhage: impact of a regional multifaceted intervention. BJOG 2009;116:1325–33.
30. Deneux-Tharaux C, Dupont C, Colin C, Rabilloud M, Touzet S, Lansac J, et al.. Multifaceted intervention to decrease the rate of severe postpartum haemorrhage: the PITHAGORE6 cluster-randomised controlled trial. BJOG 2010;117:1278–87.
31. Driessen M, Bouvier-Colle MH, Dupont C, Khoshnood B, Rudigoz RC, Deneux-Tharaux C. Postpartum hemorrhage resulting from uterine atony after vaginal delivery: factors associated with severity. Obstet Gynecol 2011;117:21–31.
32. Belghiti J, Kayem G, Dupont C, Rudigoz RC, Bouvier-Colle MH, Deneux-Tharaux C. Oxytocin during labour and risk of severe postpartum haemorrhage: a population-based, cohort-nested case-control study. BMJ Open 2011;1:e000514.
33. Bonnet MP, Deneux-Tharaux C, Bouvier-Colle MH. Critical care and transfusion management in maternal deaths from postpartum haemorrhage. Eur J Obstet Gynecol Reprod Biol 2011;158:183–8.