Confidential enquiries into maternal mortality


Prof. James Walker, Department of Obstetrics and Gynaecology, Leeds Institute of Molecular Medicine, Level 8, Clinical Science Building, St James University Hospital, Beckett Street, Leeds LS9 7TL, UK. Email


Please cite this paper as: Walker J. Confidential enquiries into maternal mortality. BJOG 2010;117:379–381.

Over the last 50+ years, obstetricians in the UK and beyond have looked on the UK Triennial Confidential Enquiry into Maternal Death1 as essential reading. These reports provide a collective experience about serious and often rare conditions, which have led to innovative research, guideline development2 and significant improvements in patient care.3 The vision that started the process in the early 1950s was inspiring, as knowledge and understanding are the key to enlightenment. The triennial report is eagerly awaited to determine whether there are new points to learn and whether changes in management have led to a decrease in mortality. It is part of a 360º audit cycle of service quality that started long before the concept was accepted by other areas in medicine, and was then forced on us by whose who believe they know better. It is a perfect example of the speciality assessing itself, highlighting concerns and developing improvement. The regular reports and reviews with their developing recommendations over the years and repeated audit can be credited with making the UK one of the safest places to have a baby in the world.4 This has been achieved by the development of a ‘no blame’ culture, by confidentiality of reporting and by review by a panel of independent assessors from both obstetrics and allied specialities, such as anaesthesiology, pathology and maternal medicine. Without this environment and structure, problems would not be exposed, causes would not be found and lessons would not be learned. Behind this detailed analysis is a network designed to maximise the ascertainment of data, the important starting point. Without a near-total collection of both maternal deaths and births within a country or area, an accurate maternal mortality ratio could not be calculated, change could not be assessed and comparisons could not be made.5

Many countries do not have accurate data collection and, even when data are collected, they are often facility (maternity unit) based and not geographically comprehensive. This may allow major problems to be identified and certain lessons to be learned,6 but provides little information about the true incidence of causes and the overall environmental reasons why women die.

It is estimated that, every year, approximately 500 000 women die world-wide in association with pregnancy and childbirth.5 That is one per minute. The Millennium Goal Number 5 is to ‘Reduce by three-quarters the maternal mortality ratio’. However, until the governing authorities in many countries believe that this appalling loss of life, with the resultant loss of mothers, wives and family structure, is unacceptable and start to count, analyse and learn, these deaths will continue. Maternal death should be made an officially notifiable event, forcing the collection of the data required to review the problems. Many countries are beginning to do this and some have demonstrated dramatic improvements,7 but there needs to be a wider will and commitment for these developments to succeed globally. Until then, many countries will remain dependent for leadership and guidance on countries with good reporting systems.

The UK Confidential Enquiries into Maternal Deaths presents the figures as the maternal mortality rate, which is the total number of direct and indirect deaths during pregnancy or in the first year after delivery per 100 000 maternities, defined as the number of pregnancies that result in a live birth at any gestation or stillbirths occurring at or after 24 weeks of completed gestation. This standardisation enables a more detailed picture of maternal death rates to be established and is used for the comparison of trends over time. However, because of the various reporting restrictions worldwide, the standardised World Health Organisation maternal mortality ratio is the number of direct and indirect maternal deaths per 100 000 live births up to 92 days after the termination of pregnancy. This results in different figures, and makes comparison with the UK difficult; the UK maternal mortality rate is always higher than the equivalent maternal mortality ratio, partly because of the longer surveillance period.1

The paper presented in this journal, the ‘Rise in maternal mortality in the Netherlands’, uses similar methods of reporting and assessment as in the UK, but presents the data using the World Health Organisation definition of the maternal mortality ratio. This means that the rates quoted are not directly comparable with those in the UK Saving Mothers’ Lives report.8 This is further complicated by the use of both nomenclatures within the paper. However, relative comparisons of the causes of death can be made and, as this is the second report from Holland using the same methodology, national comparisons over time are possible.

The strength of the review comes from the standardised methodology, the anonymisation of the cases and the assessment of the cases by an independent review committee that decided the cause of death and determined the standards of care. The ascertainment of the cases was by multiple methods with cross-referencing, maximising the collection of data, and not just by national birth certificate information which notoriously underestimates the mortality numbers.1 In the Dutch study, the national statistics office (Statistics Netherlands) under-reported maternal death by 33%. The quoted figure in the enquiry is 12.1 per 100 000 live births, but the national statistics data would only be 8.1 per 100 000 live births. The equivalent figures for the UK are approximately 13.8 per 100 000 live births quoted by the enquiry and 6.98 per 100 000 live births using death certificate data alone, an under-reporting of 49%.1 This calls into question the validity of national comparisons using ‘official’ national figures based on death certification alone.

The Dutch data are concerning as they show an increase in both direct and indirect causes of death in the years 1993–2005 compared with the previous report based on the period 1983–1992. A similar increase in indirect causes has also been seen in the UK, and it now constitutes 46% of total maternal mortality. In the Netherlands, it contributes only 23%. The reason for this difference could be better ascertainment in the UK of indirect deaths, but it also raises the question of why the rate of direct deaths in the Netherlands is significantly higher than in the UK and is still increasing, whereas it has been declining in the UK.

In the Netherlands, the main causes of direct death are pre-eclampsia, thromboembolism, haemorrhage and sepsis. Although mortality rates from thromboembolism, haemorrhage and sepsis are broadly similar between the UK and the Netherlands, deaths from pre-eclampsia are four times higher in the Netherlands than in the UK (maternal mortality ratio 3.5 versus 0.9). The deaths were associated with cerebral complications in 61% of cases and respiratory problems in 10%. Substandard care was present in 91% of cases, a problem observed previously.9 Substandard care included failure to control blood pressure, failure to stabilise prior to transportation, failure to deliver at the appropriate time and inappropriate fluid management. In the UK, pre-eclampsia used to be the leading cause of direct death, but care has improved significantly with the advent of clinical guidelines for its management,2,3 and maternal mortality has fallen significantly for this condition, with a reduction in deaths related to cerebral complications and no deaths from respiratory problems in the last triennium.1 This is clearly an area for urgent action.

Both countries have seen an increase in deaths associated with amniotic fluid embolism. The diagnosis of this condition is still controversial, leading to an estimated incidence ranging from one in 15 200 to one in 53 800 deliveries in North America and Europe.10 The increasing deaths from this condition may be a result of increased ascertainment, but this requires further investigation, as do intervention strategies.

Although there is a significant difference in the numbers reported between the UK and Holland, the causes of indirect death are similar, with cardiovascular disorders, cerebral vascular disorders and infection being the three main associations. Of the deaths from cardiovascular causes, death is mainly caused by dissection of the aorta or its branches, followed by cardiomyopathy and myocardial infarction, in the Dutch report, whereas, in the UK, myocardial infarction is the most common cause of death, followed by aortic dissection. The increase in acquired cardiovascular disease in both countries is associated with the increasing age of the pregnancy populations.

However, although, in the UK, there has been a linear increase in the rate of maternal death with increasing age, in the Netherlands, the maternal death ratio related to age appears to be a ‘U’-shaped curve, with the highest death ratios in the youngest and oldest age bands and the lowest between 30 and 35 years of age. This requires further investigation. The Netherlands has a lower teenage pregnancy rate than the UK,11 and the higher incidence of death may be related to a smaller but higher risk teenage pregnant group. Intuitively, one would expect an increasing mortality rate with age as a result of general health differences, but a higher rate in the young suggests problems with access to care that might place them at a disadvantage. The problem of access is also seen in some immigrant populations, which has been similarly highlighted in the UK reports.1

The report from the Netherlands demonstrates the value of detailed confidential enquiries at a national level. It highlights problems in care that should lead to increased investigation, development of interventional strategies and a reduction in mortality. This has been demonstrated in the UK in cases of pre-eclampsia, a particular cause for concern in the Netherlands, as well as thromboembolism. In the discussion, the authors highlight problems related to the transfer of women from low-risk care environments to teaching centres and the need to refer women with risk factors early. These factors highlight the differences in the care patterns between the UK and the Netherlands. The centralisation of care in the UK into large teaching centres lends itself to the provision of a high level of specialised care when required. However, this is seen as potentially detrimental to patient satisfaction and choice, as many women prefer to give birth locally. The fact that there are areas of concern in the Netherlands over rising maternal death ratios, despite their generally high socio-economic profile, as well as the previously documented high level of perinatal mortality,12 suggests that we should be cautious about moving our pattern of care towards theirs without careful consideration of a potentially adverse effect on maternal and perinatal mortality and morbidity. The UK has improved its safety for both mothers and babies by careful audit and guideline development. Care should be taken not to undo these changes by striving for political correctness.

Disclosure of interests

The author is Chairman of CMACE, the Centre for Maternal and Childhealth Enquiries, which produces the triennial report into maternal mortality in the UK for which he receives a small stipend.