Introduction and aims, objectives and definitions used in this Report

Gwyneth Lewis


A new title; a renewed purpose

The change of title of this, the seventh Report of the United Kingdom (UK) Enquiries into Maternal Deaths, from “Why Mothers Die” to “Saving Mothers’ Lives” more aptly reflects the purpose of this continuing, crucial, component of maternity service provision in the UK. To underline the proactive nature of this Enquiry, this Report, for the triennium 2003-2005, differs in other ways as well. It sets out ten overarching recommendations, which, where possible, are accompanied by suggested benchmarks and/or auditable standards to ensure more consistent implementation, monitoring and feedback. Whilst this does not take away the importance of the more specific recommendations and learning points made in each chapter, it will enable a more focused and strategic approach to implementing, and monitoring the key overarching recommendations which aim to provide every mother and her baby with high quality, safe and accessible maternity services.

Each Report’s recommendations are acted on by different people at many different levels. Examples include individual health practitioners, the Royal Colleges and other professional organisations, health authorities, commissioners, Trust risk and general managers, the Clinical Negligence Scheme for Trusts (CNST), the Welsh Risk Pool (WRP), the Health Care Commission (HCC) as well as central Government and its affiliated agencies. For example, in England, the National Service Framework for Maternity Services1 and its implementation strategy “Maternity Matters”2, both acknowledge the key part the findings of previous Reports have played in policy development. The Welsh National Service Framework for Maternity Services is similar3. Their recommendations help in protocol development, clinical audit and maternity service design and delivery.

In recent years, the Enquiry has expanded its remit to cover wider public health issues, and its findings and recommendations in this area have played a major part in helping in the development of other, broader, policies to help reduce health inequalities for the poorest families and for  socially excluded women. Without this current Report for example, it would not be known that, even in the UK in the twenty-first century, the most deprived  pregnant women have a risk of dying which is seven times higher than that of the broad majority of other pregnant women. And, by acting on similar findings in past Reports, this Enquiry has played a major part in re-defining the philosophy that now expects each individual women and her family to be at the heart of maternity services designed to meet their own particular needs, rather than vice versa.

Telling the story

The methodology used by the Enquiry goes beyond the scientific. Its philosophy, and that of those who participate in its process, also recognise and respect every maternal death as a young woman who died before her time, a mother, a member of a family and of her community. It does not demote women to numbers in statistical tables; it goes beyond counting numbers to listen and tell the stories of the women who died in order to learn lessons that may save the lives of other mothers and babies, as well as aiming to improve the standard of maternal health overall. Consequently, its methodology and philosophy now form part of a key strand in the World Health Organisations (WHO) overall global strategy to make pregnancy safer. A maternal mortality review tool kit, and programme, “Beyond the Numbers”4, has been introduced which includes advice and practical steps in choosing and implementing one or more of five possible approaches to maternal death reviews adaptable at any level and in any country. These are facility and community death reviews, Confidential Enquiries into Maternal Deaths, near miss reviews and clinical audit5.

Working together to save mothers’ lives and improve maternity care

It is because of the sustained commitment of all health professionals who provide maternity and other services for pregnant women in the United Kingdom that this Enquiry is able to continue as the highly respected and powerful force for change and improvement as it is today. Reading the Report or preparing a statement for an individual enquiry also forms part of individual, professional, self-reflective learning. As long ago as 1954 it was recognised that participating in a confidential enquiry had a “powerful secondary effect” in that “each participant in these enquiries, however experienced he or she may be, and whether his or her work is undertaken in a teaching hospital, a local hospital, in the community or the patient’s home must have benefited from their educative effect6”. Personal experience is therefore a valuable tool for harnessing beneficial changes in individual practice.

Learning lessons for continual improvement

This Enquiry is the first, and possibly best, example of use of the maternal mortality and morbidity surveillance cycle, now internationally adopted by the World Health Organisation’s programme “Beyond the Numbers” which promotes the use of  maternal deaths or morbidity reviews to make pregnancy safer4. The cycle, shown in Figure 1, is an ongoing process of deciding which deaths to review and identifying the cases, collecting and assessing the information, using it for recommendations, implementing these, evaluating their impact before refining and improving the next cycle. The ultimate purpose of the surveillance process is action, not to simply count cases and calculate rates. All these steps: identification, data collection, analysis, action and evaluation are crucial and need to be continued in order to justify the effort and to make a difference. The impact of previous findings of this Report continually demonstrate the contribution of such an observational study to both maternal and child health and the overall public health, and emphasise the need for it to continue in the future.

Figure 1.1 The maternal mortality or morbidity surveillance cycle

The Enquiry process is best described as an observational and self-reflective study which identifies patterns of practice, service provision, and public health issues that may be causally related to maternal deaths. This method of reviewing individual deaths has been described as “sentinel event reporting”. As Rutstein et al7, stated:

“Just as the investigation of an aeroplane accident goes beyond the immediate reasons for the crash to the implications of the design, method of manufacture, maintenance and operation of the plane, so should the study of unnecessary undesirable health events yield crucial information on the scientific, medical, social and personal factors that could lead to better health. Moreover, the evidence collected will not be limited to the factors that yield only to measures of medical control. If there is clear cut documented evidence that identifiable social, environmental, “life-style”, economic or genetic factors are responsible for special varieties of unnecessary disease, disability, or untimely death, these factors should be identified and eliminated whenever possible”.

It is this “Saving Mothers’ Lives” aims to achieve.

The evidence base

In the past some have questioned whether the Reports are 'evidence based'. The highest level of evidence of clinical effectiveness comes from systematic reviews of randomised controlled trials. The most comprehensive and up-to-date systematic reviews of relevance to these Enquiries are produced by the Cochrane Pregnancy & Childbirth Group, whose editorial structure is funded by the NHS Central Programme for Research & Development. The Co-ordinating Editor of the Group is a member of the editorial board of this Enquiry. 

Some Cochrane reviews are of direct relevance to topics highlighted by deaths described in recent Reports, and have been cited to support recommendations. These include treatments for eclampsia and pre-eclampsia, and antibiotic prophylaxis before caesarean section. However, many problems tackled in successive Reports have not been addressed by randomised trials, including prevention of thrombo-embolic disease and treatment of amniotic fluid embolism or massive obstetric haemorrhage.

An important limitation of randomised trials is that, unless they are very large, they may provide little information about rare, but important, complications of treatments. Safety issues are, therefore, sometimes better illuminated by observational studies than by controlled trials.

Many causes of maternal death are very rare and treatment options for these may never be subjected to formal scientific study. Inevitably, recommendations for care to avoid such deaths in the future rely on lesser levels of evidence, and frequently on 'expert opinion'. This does not mean that the Report is not evidence based, merely that, necessarily, the evidence cannot be in the form of a randomised control trial or case control study due to the relative rarity of the condition.

Severe maternal morbidity, “near misses”

The Enquiry has long wanted to extend its work to also review the cases of mothers who suffered severe obstetric morbidity and complications; so called “near misses”, but for lack of resources this has not yet been possible. However, this Report contains a chapter on the latest results of the Scottish Confidential Audit of Severe Maternal Morbidity8, and a summary of the first report of the United Kingdom Obstetric Surveillance System (UKOSS) for rare obstetric events, run by the National Perinatal Epidemiology Unit (NPEU) is contained in the relevant chapters of this Report7,9.

The United Kingdom Obstetric Surveillance System (UKOSS)

This new national system to study rare disorders of pregnancy was launched in February 2005. It enables the surveillance of a range of uncommon obstetric disorders, including conditions which may be classified as “near miss” events for maternal mortality. Descriptive, case-control and cohort studies are conducted.

Each month a reporting card including a simple tick box list of conditions under surveillance is sent to the nominated reporting clinicians (anaesthetists, midwives, obstetricians and perinatal risk managers) in each consultant-led maternity unit in the UK. The nominated clinicians are asked to return the card indicating whether there have been any women with any of the conditions seen in their hospital over the past month. The card also includes a box to indicate ‘nothing to report’. A data collection form is then sent back to any clinicians reporting cases to obtain details confirming the case definition, risk factors, management and outcomes. For some conditions, similar information is also collected about control or comparison women. All hospitals in the UK with consultant-led maternity units are participating in UKOSS, and hence these studies effectively survey the entire cohort of women giving birth in the UK. Several studies conducted through UKOSS provide incidence information which reveals some of the morbidity underlying the deaths detailed in this Report; this information is included in the relevant Chapters.

The aims and objectives of the Enquiry

The aim of the Enquiry is to save mothers and newborns lives by reviewing maternal deaths in order to learn lessons and formulate and disseminate recommendations which will lead to improvements in clinical care and beneficial health system changes for all women in the UK. Its objectives are:

The Enquiry’s role in setting clinical standards and contributing to clinical governance

The Enquiry is the longest running example of national professional self-evaluation in the world. Whilst much has changed since its inception in 1952, the lessons to be learnt remain as valid now as in the past. Whilst the Enquiry has always had the support of professionals involved in caring for pregnant or recently delivered women, it is also a requirement that all maternal deaths should be subject to this confidential enquiry and all health professionals have a duty to provide the information required.

In participating in this Enquiry, all health professionals are asked for two things:

At a local commissioning level maternity health care commissioners, such as Primary Care Trust (PCTs,) and Local Health Boards (LHBs) should commission services which meet the recommendations set out in this, and previous, Reports and ensure that all staff participate in the Enquiry if required, as part of their contract.

At service provider level the findings of the Enquiry should be used to:

At a national level

In every country, the findings of successive Reports have been used to develop national maternal and public health policies. For example, in England and Wales the findings of the Enquiry are used:

In Scotland, the findings of the Enquiry inform the work of equivalent bodies responsible for national quality initiatives. These include NHS Quality Improvement Scotland (NHS QIS), the Scottish Intercollegiate Guidelines Network (SIGN) and the Clinical Negligence and Other Risks Indemnity Scheme (CNORIS).

In Northern Ireland the findings of the Enquiry similarly inform policy development through the Department of Health, Social Services, and Public Safety for Northern Ireland (DHSSPS), inform quality and standards through the DHSSPS and other bodies including the Clinical Resource Efficiency Support Team (CREST) and the Regulation and Quality Improvement Authority (RQIA) and are included in postgraduate education.

Definitions of, and methods for, calculating maternal mortality

The ninth and tenth revisions of the International Classification of Diseases, Injuries and Causes of Death, (ICD9/10) define a maternal death as "the death of a woman while pregnant or within 42 days of termination of pregnancy, from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes". This means that there was both a temporal and a causal link between pregnancy and the death. When the woman died she could have been pregnant at the time, that is, she died before delivery, or within the previous six weeks have had a pregnancy that ended in a live or stillbirth, a spontaneous or induced abortion or an ectopic pregnancy. The pregnancy could have been of any gestational duration. In addition, this definition means the death was caused by the fact that the women was or had been pregnant. Either a complication of pregnancy or a condition aggravated by pregnancy or something that happened during the course of caring for the pregnant woman caused her death. In other words, if the woman had not been pregnant, she would not have died at that time.

Maternal deaths are subdivided into further groups as shown in Table 1. Direct maternal deaths are those resulting from conditions or complications or their management which are unique to pregnancy, occurring during the antenatal, intrapartum or postpartum period. Indirect maternal deaths are those resulting from previously existing disease or disease that develops during pregnancy, not due to direct obstetric causes, but which were aggravated by physiologic effects of pregnancy. Examples of causes of Indirect deaths include epilepsy, diabetes, cardiac disease and, in the UK only, hormone dependent malignancies. The Enquiry also classifies most deaths from suicide as Indirect deaths as they were usually due to puerperal mental illness although this is not recognised in the ICD coding of such deaths. The UK Enquiry assessors also classify some deaths from cancer in which the hormone dependant effects of the malignancy could have led to its progress being hastened or modified by pregnancy as Indirect although these also do not accord with international definitions. Only Direct and Indirect deaths are counted for statistical purposes as discussed later in the section on measuring maternal mortality rates.

ICD-10 also introduced two new terms related to maternal deaths. One of them is pregnancy-related death, defined as the death of a woman while pregnant or within 42 days of the end of her pregnancy, irrespective of cause. These deaths include deaths from all causes, including accidental and incidental causes. Although the latter deaths, which would have occurred even if the woman had not been pregnant, are not considered true maternal deaths, they often contain valuable lessons for this Enquiry. For example they provide messages and recommendations about domestic abuse or the correct use of seat belts. From the assessments of these cases it is often possible to make important recommendations. The ICD coding classifies these cases as fortuitous maternal deaths. However, in the opinion of the UK assessors, the use of the term fortuitous could imply a happier event and this Report, as did the last, names these deaths as Coincidental.

The other new term introduced in ICD-10 is Late maternal death, defined as the death of a woman from Direct or Indirect causes more than 42 days but less than one completed year after the end of the pregnancy. Identifying Late maternal deaths enables lessons to be learnt from those deaths in which a woman had problems that began with her pregnancy, even if she survived for more than 42 days after its end. However, although this category has only been recently recognised in the ICD 10 codes, and then only for deaths from Direct or Indirect causes, the previous three UK Enquiry Reports had already included all Late deaths notified to the assessors (including Coincidental deaths) occurring up to one year after delivery or abortion, as does this.

Table 1: Definitions of maternal deaths
Maternal deaths* Deaths of women while pregnant or within 42 days of the end of the pregnancy from any cause related to or aggravated by the pregnancy or its management, but not from accidental or incidental causes.

Direct* Deaths resulting from obstetric complications of the pregnant state (pregnancy, labour and puerperium), from interventions, omissions, incorrect treatment, or from a chain of events resulting from any of the above.
Indirect* Deaths resulting from previous existing disease, or disease that developed during pregnancy and which was not due to direct obstetric causes, but which was aggravated by the physiologic effects of pregnancy.
Late** Deaths occurring between 42 days and one year after abortion, miscarriage or delivery that are due to Direct or Indirect maternal causes.
Coincidental (Fortuitous)*** Deaths from unrelated causes which happen to occur in pregnancy or the puerperium.
Pregnancy-related deaths** Deaths occurring in women while pregnant or within 42 days of termination of pregnancy, irrespective of the cause of the death.
  • = This term includes delivery, ectopic pregnancy, miscarriage or termination of pregnancy.
  • * = ICD 9
  • ** = ICD 10
  • *** = ICD 9/10 classifies these deaths as Fortuitous but the Enquiry prefers to use the term Co-incidental as it a more  accurate description. The Enquiry also considers deaths from Late Coincidental causes.

Estimating maternal mortality ratios and rates

The international definition of the maternal mortality ratio (MMR) is the number of Direct and Indirect deaths per 100,000 live births. In many countries of the world this is difficult to measure due to the lack of death certificate data (should it exist at all) as well as a lack of basic denominator data, as baseline vital statistics are also not available or unreliable. The recent World Health Organisation publication “Beyond The Numbers; reviewing maternal deaths and disabilities to make pregnancy safer”4 contains a more detailed examination and evaluation of the problems in both determining a baseline MMR or interpreting what it actually means in helping to address the problems facing pregnant women in most developing countries.

Conversely the UK has the advantage of accurate denominator data, including both live and still births and has defined its maternal mortality rate as the number of Direct and Indirect deaths per 100,000 maternities. Maternities are defined as the number of pregnancies that result in a live birth at any gestation or stillbirths occurring at or after 24 weeks’ completed gestation and are required to be notified by law. This enables a more detailed picture of maternal death rates to be established and is used for the comparison of trends over time.

Furthermore, in the United Kingdom maternal mortality rates can be calculated in two ways:

ONS data are based on death certificates where the cause of death is directly or secondarily coded for a pregnancy-related condition such as postpartum haemorrhage, eclampsia etc.

For the past 50 years the Enquiry has calculated its own maternal mortality rate as the overall number of maternal deaths identified by the proactive case finding methodology used by this Enquiry has always exceeded those officially reported This is because not all maternal deaths are recorded as such on death certificates. For example, a large proportion of women known to the Enquiry who died of pre-existing medical conditions influenced by their pregnancy, for example cardiac disorders, epilepsy and some malignancies, were excluded from the official statistics. Other women excluded in official data are those who required long term intensive care and whose final cause of death was registered as a non pregnancy condition such as multiple organ failure even though the initiating cause was an obstetric event. Conversely, the maternal deaths known to the Registrars General may include Late deaths as it is not possible to identify from the death certificate when the delivery or termination occurred.

In order to aid the international comparison of the UK data with those from other countries calculated by using the ICD defined Maternal Mortality Ratio, this Report has also calculated the overall UK MMR as well as the more complete Enquiry maternal mortality rate. These are shown in Chapter 1. However, when making such comparisons, it is important to note two points:

Case ascertainment
The role of the Office for National Statistics

Since the introduction of a new Office for National Statistics (ONS) computer programme in 1993, all conditions given anywhere on the death certificate are now coded enabling a more extensive search of death entry information to identify all conditions listed which suggest a maternal death. In the past this has helped in improving case ascertainment, with a number of previously unreported deaths being identified. Fortunately for this Report the ONS record linkage study described below has identified very few additional cases of Direct or Indirect deaths. This is a reduction in the already small degree of under-ascertainment calculated for previous Reports.

For the past nine years, ONS have been able to match death records of women of fertile age living in England and Wales with birth registrations up to one year previously. The aim is to identify deaths of all women in England and Wales who died within one year of giving birth and to see how many additional cases can be found. The methodology, used in the past two triennia, was again applied for this Report and again shows that the majority of these deaths occurred Late, i.e. some months after delivery. The vast majority of these Late deaths were due to Coincidental Late causes and these are shown in Chapter 14.

Denominator data used for calculating mortality rates
Number of maternities

It is impossible to know the exact number of pregnancies which occurred during this or any preceding triennium since not all pregnancies result in a registered live or still birth. Because of the unreliability of these data, due to the lack of appropriate denominators, the most common denominator used throughout this and previous Reports is the number of maternities rather than to the total number of pregnancies. Maternities are the number of pregnancies that result in a live birth at any gestation or stillbirths occurring at or after 24 weeks’ completed gestation and are required to be notified by law. The total number of maternities for the United Kingdom for 2003-05 was 2,113, 831.

Estimated pregnancies

This denominator is used for calculating the rate of early pregnancy deaths. It is a combination of the number of maternities, together with legal terminations, hospital admissions for spontaneous miscarriages (at less than 24 weeks’ gestation) and ectopic pregnancies with an adjustment to allow for the period of gestation and maternal ages at conception. The estimate for the United Kingdom 2003-05 was 2,898,400. However, the resulting total is still an underestimate of the actual number of pregnancies since these figures do not include other pregnancies which miscarry early, those where the woman is not admitted to hospital, or indeed those where the woman herself may not even know she is pregnant.

Table 2: Maternal mortality definitions used in this Report
Maternal mortality definitions Reasons for use
UK Enquiry maternal mortality rates; Direct and Indirect deaths per 100,000 maternities. The most robust figures available for the UK and used for 50 years trend data in this Report.
The internationally defined Maternal Mortality Ratio (MMR); Direct and Indirect deaths per 100,000 live births. For international comparison although care needs to be taken in its interpretation due to the more accurate case ascertainment in the UK though the use of this Enquiry.
Deaths from obstetric causes per 100,000 estimated pregnancies. Because the data from spontaneous abortions and ectopic pregnancies are unreliable this denominator is only used when calculating rates of death in early pregnancy.


  1. Department of Health, England. Maternity Services. Standard 11 of the National Service Framework for Children, Young People and Maternity Services. London; Department of Health 2004. www.dh.gov.uk
  2. Department of Health. “Maternity matters: choice, access and continuity of care in a safe service”. Department of Health, London, April 2007. www.dh.gov.uk
  3. National Service Framework for Children, Young People and Maternity Services in Wales. Cardiff. 2005. www.wales.nhs.uk
  4. World Health Organisation. Beyond the Numbers - Reviewing maternal deaths and complications to make pregnancy safer. WHO; Geneva: 2004.: www.who.int/reproductive-health
  5. Lewis G, in British Medical Bulletin. Pregnancy: Reducing maternal death and disability. British Council. Oxford University Press.2003. www.bmb.oupjournals.org
  6. Ministry of Health. Report of the Confidential Enquiry into Maternal Deaths in England and Wales, 1952-1954. Reports on Public Health and Medical Subjects No.97: HMSO. 1954.
  7. Ruststein D, Berenberg W, Chalmers T, Child C, Fishman A et al. Measuring the quality of care; a clinical method. New England Journal of Medicine, March 11 1976, 582-588.
  8. Scottish Programme for Clinical Effectiveness in Reproductive Health. Confidential Audit of Severe Maternal Morbidity. 2nd Annual Report. SPCERCH December 2005. ISBN 1-902076
  9. Knight M, Kurinczuk JJ, Tuffnell D, Brocklehust P. The UK Obseteric Suveillance System for rare disorders of pregnancy. BJOG 2005: 112(3);263-5.
  10. Knight M, Kurinczuk JJ, Spark P and Brocklehurst P. United Kingdom Obstetric Surveillance System (UKOSS) Annual Report 2007. National Perinatal Epidemiology Unit, Oxford.