Rise in maternal mortality in the Netherlands

Abstract

Please cite this paper as: Schutte J, Steegers E, Schuitemaker N, Santema J, de Boer K, Pel M, Vermeulen G, Visser W, van Roosmalen J, the Netherlands Maternal Mortality Committee. Rise in maternal mortality in the Netherlands. BJOG 2009;117:399–406.

Objective  To assess causes, trends and substandard care factors in maternal mortality in the Netherlands.

Design  Confidential enquiry into the causes of maternal mortality.

Setting  Nationwide in the Netherlands.

Population  2,557,208 live births.

Methods  Data analysis of all maternal deaths in the period 1993–2005.

Main outcome measures  Maternal mortality.

Results  The overall maternal mortality ratio was 12.1 per 100 000 live births, which was a statistically significant rise compared with the maternal mortality ratio of 9.7 in the period 1983–1992 (OR 1.2, 95% CI 1.0–1.5). The most frequent direct causes were (pre-)eclampsia, thromboembolism, sudden death in pregnancy, sepsis, obstetric haemorrhage and amniotic fluid embolism. The number of indirect deaths also increased, mainly caused by an increase in cardiovascular disorders (OR 2.5, 95% CI 1.4–4.6). Women younger than 20 years and older than 45 years, those with high parity or from nonwestern immigrant populations were at higher risk. Most substandard care was found in women with pre-eclampsia (91%) and in immigrant populations (62%).

Conclusions  Maternal mortality in the Netherlands has increased since 1983–1992. Pre-eclampsia remains the number one cause. Groups at higher risk for complications during pregnancy should be better identified early in pregnancy or before conception, in order to receive preconception advice and more frequent antenatal visits. There is an urgent need for the better education of women and professionals concerning the danger signs, and for the training of professionals in order to improve maternal health care.

Introduction

In high-income countries, relatively few women die as a result of pregnancy, childbirth or puerperium. However, next to perinatal mortality, the maternal mortality ratio (maternal mortality per 100 000 live-born children, MMR) reflects the quality of obstetric care, and can be used for international comparisons. Although the absolute numbers are small, maternal death is just the tip of the iceberg of serious maternal morbidity.

The Dutch Maternal Mortality Committee (MMC), instituted by the Netherlands Society of Obstetrics and Gynaecology in 1981, published its first report in 1998, covering the years 1983–1992.¹ The only two other countries performing periodical nationwide confidential enquiries are the UK² and South Africa,³ although other countries publish reports on maternal mortality using vital statistics or regional data, which often results in underreporting.⁴

The MMR in the period 1983–1992 in the Netherlands was 9.7 per 100 000 live births. The level of substandard care in women with pre-eclampsia (93%) and the 20% underreporting to the enquiry indicated goals for improvement in the following decade. Recommendations were made to improve care, especially for immigrant populations, and to report cases to the MMC to lower the percentage of underreporting. This paper contains a nationwide analysis of maternal deaths during the period 1993–2005, examining trends in maternal mortality and evaluating the level of implementation of the recommendations of the previous report.

Methods

All maternal deaths reported to the MMC during pregnancy or within 1 year after pregnancy between January 1993 and December 2005 in the Netherlands were included in the study. The MMC is made up of eight obstetricians and one internal medicine specialist working in the field of maternal medicine, appointed by the Dutch Society of Obstetrics and Gynaecology. The members are from both university and peripheral hospitals. Maternal mortality cases were voluntarily reported to the MMC by obstetricians and, in some cases, by midwives and GPs. Additional cases were collected after a cross-check with the data collected by Statistics Netherlands, which collects all vital data from the Netherlands. Finally, a request to report every death during or within 1 year after pregnancy in the study period was submitted to all 98 obstetric departments in the Netherlands.

Maternal death was defined and classified according to the World Health Organisation’s International Classification of Diseases, 10th revision (ICD-10).⁵ Deaths were classified as direct, indirect or fortuitous. The MMR is defined as the number of direct and indirect maternal deaths per 100 000 live births up to 42 days after the termination of pregnancy. A single underlying cause or mode of death was assigned to each case by the members of the MMC. The underlying cause of death is the disease or injury which results directly in death or initiates the chain of events leading directly to death. The mode of death is the disease or injury that ends life directly. Late maternal death was defined as the death of a woman from direct or indirect obstetric causes more than 42 days but <1 year after the termination of pregnancy.

Substandard care was defined as all care factors which may have resulted in low standards of care and which had a probable negative influence on the chain of events leading directly to death. It could be assigned to any person involved in the care of pregnant women and to the pregnant woman herself. Avoidance of such factors did not necessarily mean that death would have been prevented. The standard of care was the care as stated in national guidelines.^6–11 If there was no (appropriate) guideline, the best available evidence was used. The anonymised cases were individually assessed for substandard care factors by the members of the MMC. After that, all cases were discussed at a group meeting for a final decision on classification and substandard care factors. When consensus could not be reached, the decision was based on the assessment of the majority of the group.

Every 3 years, an update of all maternal deaths with trends and recommendations is presented at the National Congress of the Netherlands Society of Obstetrics and Gynaecology. The national guidelines are updated every 5 years, with specific recommendations and updated references. Furthermore, lectures are given on specific maternal mortality topics around the country, and sometimes published in case reports in the Netherlands Journal of Obstetrics and Gynaecology. The implementation of the recommendations and guidelines should lead to a decline in maternal mortality, measured in the subsequent report on maternal mortality.

A confidential enquiry was completed on each case reported to the MMC. For each maternal death, data were collected by the MMC on a standard questionnaire including information concerning general and obstetric histories, as well as the index pregnancy. Other sources of information, such as antenatal charts, laboratory and bacteriological results, pathology and autopsy reports and professional correspondence, were also analysed.

Results

In the study period, 414 deaths were identified. The classification of cases is shown in Table 1. A total of 367 cases was reported directly to the MMC and, after the cross-check with Statistics Netherlands, another 47 cases were identified (11%, 47/414). From these cases, only information on the cause of death and age was available. Of all direct and indirect cases, 24 were classified as late maternal deaths. From one of the late cases, only information from Statistics Netherlands was available. The MMR was 12.1 per 100 000 live births (309/2 557 208 live births in this period). This represents a statistically significant rise compared with the period 1983–1992 with an MMR of 9.7 (OR 1.2, 95% CI 1.0–1.5).¹

Table 1.   Numbers of identified and available cases for confidential enquiry, the Netherlands, 1993–2005

Classification	All identified deaths (n)	Late deaths (n)	Available for enquiry (n, %)
Direct	236	12	196 (83)
Indirect	97	12	93 (96)
Fortuitous	81	27	78 (96)
Total	414	51	367 (89)
Fortuitous/late excluded	309		266 (86)

Statistics Netherlands reported 208 cases of maternal death in the years 1993–2005. The underreporting in the official statistics was thus 33% (101/309).

The causes of direct and indirect maternal death are listed in Tables 2 and 3. MMRs for direct and indirect causes were higher in this study period than in the previous period 1983–1992. This rise was almost statistically significant for direct causes (OR 1.2, 95% CI 0.99–1.5) and statistically significant for indirect causes (OR 1.5, 95% CI 1.0–2.1).

Table 2.   Underlying causes of direct maternal deaths* in the Netherlands, 1983–1992 versus 1993–2005

Live births Underlying cause	1983–1992		1993–2005		OR (95% CI)
	1 860 807		2 557 208
	n*	MMR**	n*	MMR**
*Number of deaths, late deaths included. **Maternal mortality ratio, late deaths excluded. ***Miscellaneous: obstetric medication (2), pneumonia treated as embolism (1), abortion (2).
(Pre-)eclampsia	51	2.7	93	3.5	1.3 (0.9–1.9)
Thromboembolism	21	1.1	44	1.6	1.5 (0.9–2.6)
Obstetric haemorrhage	11	0.6	18	0.7	1.2 (0.6–2.5)
Genital tract sepsis	10	0.5	20	0.7	1.5 (0.7–3.1)
Genital tract trauma	8	0.4	3	0.1	0.3 (0.1–0.9)
Complication of caesarean section	8	0.4	4	0.2	0.4 (0.1–1.1)
Sudden death in pregnancy	12	0.6	22	0.8	1.3 (0.7–2.7)
Choriocarcinoma/hydatidiform mole	7	0.1	3	0.0	0.3 (0.9–1.1)
Complication of anaesthesia	4	0.2	2	0.1	0.4 (0.1–1.7)
Ectopic pregnancy	4	0.2	5	0.2	0.9 (0.3–3.1)
Acute fatty liver of pregnancy	0	0.0	6	0.2	P = 0.04
Amniotic fluid embolism	2	0.1	11	0.4	4.0 (1.0–16.1)
Miscellaneous	6	0.3	5***	0.2	0.6 (0.2–1.9)
Total	144	7.3	236	8.8	1.2 (0.99–1.5)

Table 3.   Underlying causes of indirect maternal deaths* in the Netherlands, 1983–1992 versus 1993–2005

Live births Underlying cause	1983–1992		1993–2005		OR (95% CI)
	1 860 807		2 557 208
	n*	MMR**	n*	MMR**
*Number of deaths, late deaths included. **Maternal mortality ratio, late deaths excluded. ***Miscellaneous: carcinoma (3), Steinert disease (1), systemic lupus erythematosus (2), liver cirrhosis (1), renal disorders (2), HIV (1), neurological disorders (2).
Cardiovascular disorders	13	0.6	45	1.6	2.5 (1.4–4.6)
Cerebrovascular disorders	19	1.0	15	0.6	0.6 (0.3–1.1)
Infectious diseases	3	0.1	9	0.4	2.2 (0.6–7.5)
Mental disorders	2	0.1	7	0.3	2.5 (0.6–10.8)
Diseases of blood(-forming organs)	6	0.3	4	0.2	0.5 (0.1–1.6)
Endocrine disorders	3	0.2	3	0.1	0.7 (0.2–3.2)
Pulmonary disorders	1	0.1	2	0.1	1.5 (0.2–11.1)
Miscellaneous	1	0.1	12***	0.3	8.7 (1.5–52.3)
Total	48	2.4	97	3.3	1.5 (1.0–2.1)

Pre-eclampsia was the leading cause of maternal mortality, being twice as frequent as the next most common cause. Death as a result of pre-eclampsia, thromboembolism, sepsis and amniotic fluid embolism occurred more often than in the period 1983–1992, but this was not statistically significant. Acute fatty liver of pregnancy was statistically significantly more frequently reported (P = 0.04). In one of these six cases, the woman was pregnant with twins (17%). In addition, the number of indirect deaths caused by cardiovascular diseases was significantly higher (OR 2.5, 95% CI 1.4–4.6), being the second most frequent cause of maternal death, followed by death as a result of thromboembolism. Maternal mortality caused by genital tract trauma was statistically significantly lower in the period 1993–2005 (OR 0.3, 95% 0.1–0.9).

The overall mean age was 30.9 years (range, 17–50 years). The mean age of all direct deaths was 30.7 years (range, 17–50 years), and 31.3 years (range, 17–45 years) for all indirect deaths. MMR (late deaths excluded) was lowest in women between 30 and 35 years of age, and the highest risks were found in teenagers and women aged 45 years and above (Table 4).

Table 4.   Maternal deaths and live births in each age group with corresponding maternal mortality ratio (MMR), Statistics Netherlands data included; the Netherlands 1993–2005

Age (years)	No. deaths	Live births*	MMR	OR (95% CI)
*Statistics Netherlands.
15–19	10	29364	34.1	3.2 (1.7–6.0)
20–24	33	230 812	14.3	1.3 (0.9–1.9)
25–29	86	749 981	11.5	1.1 (0.8–1.4)
30–34	113	1 045 676	10.8	1.0
35–39	76	436 089	17.4	1.6 (1.2–2.2)
40–44	12	62 689	19.1	1.8 (0.99–3.2)
≥45	3	2597	115.5	10.7 (3.6–31.9)

Women pregnant with their second child showed the lowest risk, whereas women pregnant with their first child and with parity three and above were at higher risk of maternal mortality (Table 5).

Table 5.   Maternal deaths and parity in each age group with corresponding maternal mortality ratio (MMR); the Netherlands 1993–2005

Parity	n	Live births*	MMR	OR (95% CI)
*Statistics Netherlands.
0	146	1 168 068	12.5	1.4 (1.1–1.9)
1	81	924 351	8.8	1.0
2	39	326 278	12.0	1.4 (0.9–2.0)
3+	23	138 511	16.6	1.9 (1.2–3.0)

The majority (178, 62%) of women who died had been in good general health before pregnancy, whereas 111 (38%) had chronic conditions, such as hypertension (n = 26), cardiovascular disease (n = 11), diabetes (n = 6), asthma (n = 5) and sickle cell disease (n = 4).

Of the 143 women of parity one or above, 66 (46%) had undergone uneventful previous pregnancies. Twenty-six women had pregnancy-induced hypertension (n = 8), pre-eclampsia (n = 12) or HELLP syndrome (haemolysis, elevated liver enzymes and low platelet count) (n = 6) in a previous pregnancy. Other previous complications were caesarean section (n = 22), postpartum haemorrhage (n = 9), perinatal death or a child small for gestational age (n = 12).

From the 289 direct and indirect deaths, including late maternal deaths, 83 (29%) were immigrant women. The women originated from sub-Saharan Africa (n = 31), Surinam and the Dutch Antilles (n = 20), Turkey (n = 11), Morocco (n = 8), Asia (n = 7) and Libya (n = 1). Five women came from western countries other than the Netherlands. The number of live-born children was known for almost all ethnicities, but only for the period 1996–2005. The MMR could thus only be calculated over that time period (Table 6). The MMR was substantially higher for all nonwestern women (20.7, OR 2.1, 95% CI 1.6–2.7). The highest MMR was found in women from sub-Saharan Africa and Asia. In ethnic minorities, especially in women from sub-Saharan Africa (55%), the most frequent cause of death was pre-eclampsia.

Table 6.   Maternal deaths and ethnicity; the Netherlands 1996–2005

Origin	n*	Live births**	MMR***	OR (95% CI)
*Number of deaths, late deaths included. **Statistics Netherlands. ***Maternal mortality ratio calculated with number of deaths and live births between 1996 and 2005, Statistics Netherlands; late deaths excluded. ****Maternal deaths originating from sub-Saharan Africa, Asia; no subgroups defined by Statistics Netherlands.
Dutch native	165	1 516 332	10.0	1.0
Western immigrants	5	164 834	2.4	0.2 (0.2–0.7)
Nonwestern immigrants	66	294 170	20.7	2.1 (1.6–2.7)
Surinam/Dutch Antilles	19	64 786	26.2	2.7 (1.7–4.3)
Turkey	8	66 925	10.5	1.1 (0.5–2.2)
Morocco	6	71 556	8.4	0.8 (0.3–1.7)
Other immigrants****	33	90 903	34.1	3.3 (2.3–4.8)

Of all 93 cases of maternal death from pre-eclampsia, 80 (86%) were available for further analysis. Cerebral haemorrhage (36 cases, 45%) was the leading mode of death and, in most cases, was associated with high systolic blood pressure (n = 32, 89%) and low platelet counts (n = 26, 72%). The diastolic blood pressure was above 110 mmHg in 19 cases (53%). All modes of death are shown in Table 7. Eclampsia complicated 35 cases (44%). Adult respiratory distress syndrome complicated eight cases: in four of these, the woman received an overload of intravenous fluids. In 74 women (93%), risk factors were present: chronic hypertension (n = 12, 15%), pre-eclampsia in a previous pregnancy (n = 10, 13%), twin pregnancy (n = 9, 11%), obesity (n = 7, 9%), in vitro fertilisation (n = 4, 5%), maternal age above 35 years (n = 18, 23%), vascular or kidney disease (n = 2, 2%), sub-Saharan origin (n = 16, 20%) and nulliparity (n = 48, 60%). Nulliparity was the only risk factor in 26 women (33%).

Table 7.   Modes of death in 80 cases of maternal mortality caused by pre-eclampsia, the Netherlands, 1993–2005

Mode of death	n (%)
Cerebral complication	49 (61)
Cerebral haemorrhage	36 (45)
Cerebral oedema	8 (10)
Encephalopathy	5 (6)
Adult respiratory distress syndrome	8 (10)
Disseminated intravascular coagulation	6 (8)
Multiple organ failure	3 (4)
Liver rupture	7 (9)
Miscellaneous	7 (9)

Of all cases of mortality caused by thromboembolism, pulmonary embolism was the most frequent mode of death (33 cases, 75%), often within 14 days after delivery (19 cases, 70% of this group). Thirty-seven of the 44 deaths (84%) from thromboembolism were available for complete analysis. Eleven women died in the antepartum period (30%). Risk factors for thromboembolism were present in 26 women (70%), caesarean section (n = 10), maternal age above 35 years (n = 10) and obesity (n = 9) being the most frequent. Fourteen women were eligible for thromboprophylaxis according to the Dutch thromboprophylaxis guideline,¹⁰ and all received this, either during puerperium or hospitalisation. Ten women (27%) were already on thromboprophylaxis at the time of thromboembolism. Three other women suffered from hyperemesis gravidarum complicated by dehydration and immobility. They were admitted to hospital and received no thromboprophylaxis.

Of the 45 women dying from cardiovascular diseases, 39 (87%) had acquired heart diseases. Death was mainly caused by aortic dissection or its branches (n = 19), cardiomyopathy (n = 7), myocardial infarction (n = 5) and peri/myocarditis (n = 4). Congenital heart disease caused six deaths (13%). For complete analysis, 42 cases (93%) were available. Twenty women died antepartum (13 dissections) and five died more than 42 days postpartum. In 28 women (65%), risk factors were present: maternal age above 35 years (n = 17), chronic hypertension (n = 10) and obesity (n = 7).

In the Netherlands, nine cases of suicide were reported to the MMC. Seven were classified as indirect deaths and three of these were late deaths. Two other cases were classified as fortuitous deaths.

A total of nine women delivered at home. Three women died within 48 hours postpartum. One woman showed delay in care as a result of delivery at home. Five women died between 1 and 6 weeks after home birth. In one woman, there was a delay in diagnosis.

Sixty-eight women (24%) died undelivered. Of these deaths, 29 (43%) occurred between 28 and 37 weeks of gestation. Nine (13%) deaths occurred after 37 weeks. Caesarean section had been performed in 49% of the women who died. Caesarean section in this study was often performed in haemodynamically nonstabilised women, because of maternal or fetal compromise. In four women, a complication of caesarean section was the only cause of maternal death. All of these caesarean sections were urgent procedures during labour. Three women had severe haemorrhage after the procedure, two caused by haemorrhage from an artery or vein. In one woman, haemorrhage during the procedure resulted in disseminated intravascular coagulopathy.

There were 22 maternal deaths in twin pregnancies, seven of which occurred after in vitro fertilisation and one after ovulation induction. Ten pregnancies in total occurred after in vitro fertilisation.

Substandard care was identified in 158 of all 289 cases (55%): the classification is given in Table 8. Most substandard care was identified in direct cases (63%), especially in those with mortality caused by pre-eclampsia (91%). Other characteristics are given in Table 9.

Table 8.   Substandard care in 289 cases of maternal mortality, the Netherlands, 1993–2005

	Dutch natives	Immigrants	Total
Values are denoted as n (%).
Direct maternal mortality	78 (58)	46 (74)	124 (63)
Indirect maternal mortality	23 (32)	11 (52)	34 (37)
Pre-eclampsia	46 (94)	27 (87)	73 (91)
Patient
Delay consulting doctor	4 (2)	4 (5)	8 (3)
Refusing medical advice/treatment	13 (5)	5 (6)	18 (6)
Communication difficulties	0 (0)	8 (10)	8 (3)
Midwife/GP
Inadequate antenatal visits	3 (1)	1 (1)	4 (1)
Delay in diagnosis	15 (7)	8 (10)	23 (8)
Delay in referral to hospital care	13 (6)	5 (6)	18 (6)
Obstetrician
Inadequate antenatal visits	2 (1)	3 (4)	5 (2)
Delay in diagnosis	44 (21)	22 (27)	66 (22)
Despite diagnosis, no or inadequate therapy	45 (22)	34 (41)	79 (27)
Antihypertensive medication	27 (13)	18 (22)	45 (16)
Prophylaxis of eclampsia	8 (4)	5 (6)	13 (4)
Inadequate referral to perinatal centre	5 (2)	0 (0)	5 (2)
Unstabilised delivery	11 (5)	9 (11)	20 (7)
Unstabilised transport	5 (2)	3 (4)	8 (3)

Table 9.   Some characteristics of 289 maternal deaths in the confidential enquiry, the Netherlands, 1993–2005

Factor	n (%)	General population* (%)
*Data from Perinatal Registration, the Netherlands.^25 **Deliveries in nonwhite immigrants.
Initial antenatal care by
Midwife/GP	151 (52)	79.5
Obstetrician	136 (47)	20.5
Concealed pregnancy	2 (1)
Referral by primary care giver	173 (60)
During pregnancy	131 (76)
In labour	22 (13)
After birth	20 (12)
Death during pregnancy	68 (24)
Death after pregnancy	221 (76)
Within 2 days	96 (43)
Within 2–7 days	35 (16)
Between 7 and 42 days	67 (30)
Late deaths (>42 days)	23 (10)
Duration of pregnancy (week)
<24	37 (13)	0.4 (20–24 weeks)
24–28	18 (6)	0.4
28–34	62 (21)	1.8
34–37	51 (18)	5.2
>37	121 (42)	91.1
Deliveries	212 (73)
Mode of delivery
Vaginal birth	108 (51)	84.8
Spontaneous	84 (40)	74.9
Caesarean section	104 (49)	15.1
Home birth	9 (4)	23.2
Death at home	33 (11)
Antepartum	16 (48)
Within 6 weeks postpartum	13 (39)
After 6 weeks postpartum	4 (12)
Death in nonwhite immigrants	82 (28)	14.6**
Preterm delivery	113 (39)	9.5
Autopsy performed	117 (40)
Admission to intensive care unit	183 (63)
Perinatal mortality	121 (42)
Fetal death <24	36 (12)
weeks of gestation Perinatal mortality >24	86 (30)	1.1
weeks of gestation

Discussion

Maternal mortality in the Netherlands has shown an increase compared with the period 1983–1992. This is a matter for concern.

The rise in maternal mortality may be caused partly by better reporting to the MMC. The percentage of cases not reported to the MMC is still significant, although it has improved compared with the period 1983–1992, in which 80% of cases were available for complete analysis. There is a remarkable difference in underreporting between direct (17%) and indirect (3%) cases. The underreporting to Statistics Netherlands did not improve over time: it was 26% in 1983–1992 and 33% in 1993–2005. Vital statistics reported by Statistics Netherlands are used for international comparison, but are not sufficiently accurate.¹² In the Confidential Enquiry into Maternal Deaths in the UK in 2000–2002, all but two cases were completely available for coding according to type, from a total of 391 deaths. Comparison of the death register with the birth register from Statistics Netherlands has now become possible in the Netherlands and, hopefully, we will have a higher percentage available for analysis in the next period.

The methods used in the enquiry, however, remain unchanged, and it seems unlikely that the increase is caused by registration or classification bias. Furthermore, the same trend was seen in the UK.² Both enquiries are audits on a nationwide level using the same definitions. The pregnancy-related mortality surveillance in the USA has also shown a rise in maternal mortality since 1982, with a mortality ratio of 13.2 in 1999.¹³

Demographic changes could partly explain this increase. The risk of death is influenced by both age and parity. In the general population, the percentage of mothers aged ≥35 years at delivery increased from 5.2% in 1980 to 20% in 2003 in the Netherlands.¹⁴ The mean total number of children (1.8) for each woman remained virtually unchanged in this period, and thus parity is less influential. Maternal age at birth was shown to be an important risk factor for all causes. In the first enquiry in the period 1983–1992, no maternal mortality was seen in women above 45 years, whereas three women in this age group died in the subsequent study period (MMR 115.5, OR 10.7, 95% CI 3.6–31.9).

A higher maternal age may be associated with an increase in the need for assisted reproductive techniques. In vitro fertilisation is associated with a higher maternal mortality in Australia (MMR 25.7 versus MMR 10.9 in the general population).¹⁵

The increase in maternal deaths from cardiovascular disease is similar to the findings in the UK and may be associated with the increase in obesity, chronic hypertension and maternal age in the Netherlands.

The statistically significant decrease in maternal death caused by genital tract trauma, such as uterine rupture, seems unusual in the light of the increase in the percentage of caesarean sections. It may be that more primary caesarean sections were performed after a previous caesarean section, but it may also have resulted from earlier anticipation by obstetricians of impending uterine rupture. The rise in maternal death caused by acute fatty liver of pregnancy and amniotic fluid embolism may be attributable to improved awareness of these diseases by obstetricians. A high index of suspicion of these diseases leads to more diagnostic tests to confirm the diagnosis. In a cohort in the UK Obstetric Surveillance System, a relationship was found between twin pregnancy and acute fatty liver of pregnancy (18%).¹⁶ We found a comparable percentage (one in six cases, 17%), but our numbers were too small to draw firm conclusions.

The MMR of sub-Saharan African and Asian women was more than three times that of the native Dutch group. In contrast with our first enquiry, MMRs of Moroccan and Turkish women were similar to the MMRs of indigenous women. Twenty percent of all births in the Netherlands are in immigrant populations today, whose fertility rate is 40% higher than that of the native Dutch population.¹⁴ In our study, 29% of the women who died were immigrants, and substandard care was found to be more frequent in immigrant women. In one of ten cases, communication difficulties were mentioned in medical files, and this probably is an underestimation. The fact that MMRs in women from Morocco and Turkey were similar to those of native Dutch women could be interpreted as a result of increased health literacy within a multicultural context, as these groups have been in the Netherlands for a relatively long time.

Changes in care could have contributed to the rise in maternal mortality. The percentage of caesarean sections in the Netherlands increased in the study period from 8.1% to 13.6%.¹⁷ In only four cases did caesarean section contribute directly to death but, in unstable women, it can influence the chain of events leading to death.

Differences in the quality of care could explain some of the changes in MMR. Substandard care, however, decreased from 78% of all direct cases in the period 1983–1992 to 63% in the period 1993–2005. In contrast, substandard care in pre-eclampsia did not show such a decline (93% versus 91%).

Pre-eclampsia is the leading cause of maternal mortality in the Netherlands, showing a rising trend in the period 1983–2005 (OR 1.3, 95% CI 0.9–1.9). The MMR for gestational hypertensive disorders is markedly higher than in the UK (MMR 3.5 versus 0.9).² Substandard care in the UK for hypertensive diseases was 72% in the period 2003–2005, in which 44% major substandard care was involved.² In previous reports of the Confidential Enquiry into Maternal and Child Health, hypertensive diseases were also ranked higher in the UK. Specific recommendations were subsequently given, especially for better monitoring of blood pressure and control of fluid balance. This advice possibly contributed to the decline in maternal deaths caused by hypertensive diseases. Most substandard care in maternal mortality caused by hypertensive disease in pregnancy in the Netherlands concerned insufficient diagnostic testing where indicated, inadequate management of hypertension by obstetricians, no use or inadequate use of magnesium sulphate, inadequate stabilisation before transport to tertiary care centres and/or delivery and failure to consider timely delivery.¹⁸ In addition, fluid overload in women with pre-eclampsia dying as a result of adult respiratory distress syndrome is of serious concern. In the UK, the advice to limit the administration of intravenous fluids in women with pre-eclampsia led to a decline of mortality caused by adult respiratory distress syndrome in pre-eclampsia.²

All women who were eligible for thromboprophylaxis according to the Dutch guideline received this.¹⁰ In the UK, most direct deaths are caused by thromboembolism (MMR 1.94). This is slightly higher than the Dutch MMR (1.6), whereas the UK has a higher caesarean section rate (23% versus 15.1%). The number of deaths caused by postpartum embolism after caesarean section decreased over the years in the UK, probably as a result of increasing vigilance and better application of thromboprophylaxis protocols. Current guidelines in the UK advise that doses of low-molecular-weight heparin should be adjusted to the woman’s weight.² This is also advised in the Dutch guideline, but not strongly. With the increasing number of morbid obese women, this may also be a factor in the smaller number of deaths caused by thromboembolism in the Netherlands. Obese women should be aware of the risks of obesity, preferably before conception, and should be educated about the danger signs. Risk assessment for thromboembolism should be undertaken for every pregnant woman.

In contrast with the UK, where suicide was the leading cause of maternal deaths when late deaths were included in the period 2000–2002, only nine cases of maternal death from suicide were reported in the Netherlands.^19–21 Cases of suicide are classified as indirect death if the pregnancy is thought to have influenced the psychiatric state of the woman. If such a connection cannot be found, the case is classified as fortuitous. There may be underreporting in the Netherlands, because often these women die more than 42 days after delivery and are no longer under the care of an obstetrician. In the period 2003–2005, the number of maternal deaths as a result of suicide declined in the UK, which may indicate that the recommendations to identify women most at risk and the application of appropriate management are having a preventative effect.

We recommend that women at increased risk for complications during pregnancy, such as pre-eclampsia, thromboembolism and heart disease, should be recognised better through appropriate medical, obstetric and family history taking. When pre-existing risk factors are present, women should receive appropriate preconceptional advice and, when pregnant, should be seen more frequently than low-risk pregnant women. If necessary, the woman should be referred to (university) hospital care. All pregnant women need to be educated about the danger signs associated with serious complications in pregnancy. Hopefully, this will lead to a decrease in the high percentage of delay in diagnosis.

Special efforts should be directed towards nonwestern immigrant populations, especially sub-Saharan African women, because they face the highest risks.²²

However, life-threatening complications are often unanticipated and occur suddenly. The rise in maternal deaths from hypertensive diseases and amniotic fluid embolism suggests that there is room for improvement in handling such emergencies in the Netherlands. There is evidence that simulation training programmes improve skills and thus reduce complications in emergency situations.^23,24 Such programmes should also be incorporated in the education of obstetricians and midwives.

In conclusion, the significant rise in maternal mortality and, especially, mortality from hypertensive diseases, as well as the high prevalence of substandard care involved in these cases, should lead to serious concern. Changing this situation must be a top priority for Dutch obstetrics.

Disclosure of interest

There are no conflicts of interest.

Contribution to authorship

All authors contributed to the conception and design and/or acquisition of data and the analysis of data. All authors contributed to the drafting or revising of the article and all approved of the final version to be published.

Details of ethics approval

For this study, no ethics approval was needed, as it is a retrospective study of maternal mortality.

Funding

No funding.