Perinatal mortality surveillance 2004



  • The stillbirth rate in 2004 remained high (5.7 per 1000 births) after a significant increase between 2001 and 2002.
  • Three-quarters of stillbirths delivered after 28 completed weeks of gestation.
  • Multiple births had a higher stillbirth rate and neonatal mortality rate than singletons, 3.2 and 7.0 times higher, respectively.
  • A high proportion of stillbirths and neonatal deaths were born to women living in socially deprived areas.
  • Both the stillbirth rate and neonatal mortality rate were higher in women of Black, Asian or Other ethnicity.
  • Postmortem examination rates have plateaued at 42% after a significant fall throughout the 1990s and early 2000s.



Total death notifications

In 2004, there were 3791 stillbirths and 2257 neonatal deaths in England, Wales and Northern Ireland reported to CEMACH (Table 2).

The stillbirth rate for 2004 was 5.7 per 1000 total births equating to more than one in every 200 babies born. Following a significant increase in the stillbirth rate in 2002 (from 5.4 in 2001 to 5.7 in 2002), the rate remained high in 2003 (5.8 per 1000 total births) and in 2004 is still higher than at any point seen since the change in registration of stillbirths from 28 weeks of gestation onwards to 24 weeks of gestation onwards in 1992.5

The neonatal mortality rate in 2004 was 3.4 per 1000 live births. This represents a decrease from 2003 of 0.3 per 1000 live births.

Also presented in Table 2 are the total number of late fetal losses and perinatal deaths. Perinatal deaths are presented in two ways. Firstly, according to the definition in use in the UK, that is all stillbirths and early neonatal deaths. Secondly, the perinatal mortality rate is presented according to the World Health Organization definition, which includes all babies delivering from 22 weeks of gestation showing no signs & life (i.e. late fetal losses and stillbirths) and all early neonatal deaths.6




Mortality in singleton and multiple births

Multiple births are at a greater risk than singleton births of adverse perinatal outcome, including stillbirth and neonatal death. Table 3 shows the stillbirth and neonatal mortality rates according by the plurality of the birth.

The stillbirth rate of multiple births was 3.2 times that for singletons. An even greater disparity is seen for neonatal death with the neonatal mortality rate for multiple births 7.0 times that for singletons.

In 2004, there was a decrease in both the stillbirth rate and neonatal mortality rate for multiple births compared with the rates seen in 2003 of 3.1 per 1000 total births and 0.8 per 1000 live births, respectively.




Cause of death





The cause of death for stillbirths is currently classified using the Extended Wigglesworth classification supplemented by the Obstetric (Aberdeen) classification. Details of these classification systems can be found at www.cemach.org.uk/pdn_classifications.htm.

Figure 2 shows the cause of death of all stillbirths. The largest identifiable group is deaths due to severe/lethal congenital anomalies accounting for 15.1% of all stillbirths. This is followed by antepartum haemorrhage (10.0%) and death from intrapartum causes (7.3%). Just over 50% of stillbirths are antepartum fetal deaths that remain unexplained using the current classification systems. This situation is clearly unsatisfactory and CEMACH will be reviewing the classification systems currently in use in order to improve the reporting of the cause of death for stillbirths.

From 2005, CEMACH will report on fetuses that were small for gestational age and will also include information on the presence of fetal growth restriction.

Cause-specific stillbirth rates according to timing of death (antepartum versus intrapartum) are shown in Appendix B.




Neonatal deaths

The cause of death for neonatal deaths is classified according to the extended Wigglesworth classification. For 2004, the largest proportion of neonatal deaths was classified as death due to immaturity, 48% (Figure 3). This was followed by lethal/severe congenital anomalies (22%) and death due to intrapartum causes (11%).

Cause-specific early and late neonatal mortality rates can be found in Appendix B.




Maternal risk factors




Maternal age

The association of maternal age with stillbirth rate and neonatal mortality rate is well documented with higher rates of both being experienced by women of very young ages and those of older ages.

Age-specific stillbirth and neonatal mortality rates are shown in Figure 4. Data used to create this figure can be found in Appendix B.





CEMACH collects self-reported maternal ethnicity in order to explore the association between ethnicity and perinatal death.

The distribution of maternal ethnicity for all reported late fetal losses, stillbirths and neonatal deaths is shown in Table 4.

The calculation of ethnicity-specific mortality rates is hindered by the fact that neither registration statistics for England and Wales nor those for Northern Ireland collect information on maternal ethnicity. We have, however, attempted to calculate them using the information on maternal ethnicity collected in England as part of the delivery record of the Hospital Episodes Statistics (HES).4 There are a number of problems with doing this, not least that the data collection of maternal ethnicity on HES remains incomplete with 75% of all birth records having a stated ethnicity for the period 2003–04. The resulting rates should therefore be considered as an approximation only. The methodology used to calculate these estimated rates is further described in Appendix C.

Estimated maternal ethnic-specific mortality rates (Table 5) show significantly higher stillbirth rates and neonatal mortality rates for women of Black ethnicity (2.8 and 2.7 times higher, respectively), Asian ethnicity (2.0 and 1.6 times higher, respectively) and Chinese and other ethnicity (1.9 times higher in both cases) when compared with rates for women of White ethnicity.





The relationship of stillbirths and neonatal deaths with social deprivation was explored by the application of an Index of Multiple Deprivation (IMD 2004)7 score, a measure of deprivation at the small area level. The relative distribution of stillbirths and neonatal deaths according to quintile of deprivation is shown in Table 6 and the methodology further described in Appendix C.

Over one-third of all stillbirths and neonatal deaths were born to mothers resident in the most deprived quintile (compared with expected 20%). While this would appear to substantiate previous work that shows that deprivation is associated with adverse perinatal outcome it is not possible to draw any firm conclusions from this data in the absence of published information about deprivation in the general maternity population.

These deprivation scores are derived from area of residence-based statistics. For future reports, CEMACH hopes to look at individual level occupation and social class data by linkage with registration data collected by the ONS and NI GRO.




Characteristics of the baby




Birth weight

Table 7 shows the distribution of all live births, stillbirths and neonatal deaths according to birth weight. At the time of writing, birth weight data for live births were not available split by multiplicity and these figures therefore include singleton and multiple births. Two-thirds of all stillbirths and over 70% of all neonatal deaths had a birth weight of less than 2500 g compared with only 7.6% of all live births in England, Wales and elsewhere. Birth weight specific mortality rates can be seen in Table 8.

Figures for birth weight displayed in this section are for cases resident in England, Wales and elsewhere as denominator data for all live births in Northern Ireland were not available at the time of writing.




Gestational age

Figures for gestational age displayed in this section are for cases resident in England only as denominator data for all live births in Wales and Northern Ireland were not available at the time of writing.

Table 9 shows the distribution of all live births (singletons and multiples) according to gestational age at delivery. Nearly three-quarters of stillbirths were delivered after 28 completed weeks of gestation. Over 70% of all neonatal deaths were born preterm, i.e. before 37 completed weeks of gestation.

Gestational age-specific stillbirth rates and neonatal mortality rates are shown in Table 10.

In future reports, CEMACH intends to present figures showing distribution of stillbirths and neonatal deaths by birth weight adjusted for gestational age. This will enable a greater exploration of the association of being small for gestational age with stillbirth and neonatal death.




Commentary on findings

Perinatal mortality rates have been falling since the 1950s but throughout this period stillbirths have been and remain the largest contributor accounting for nearly 70% of perinatal deaths in 2004. By the 1990s, the stillbirth rate had stabilised and it was therefore of particular concern when an increase was observed in 2002. This report finds that these higher rates continued to be observed in 2004.

Nearly three-quarters of stillbirths deliver after 28 completed weeks of gestation, a time when a baby could be expected to survive if delivered alive in a good condition. Confidential enquiry assessments of maternity care in over 400 stillbirths in 1996–1997 noted that care had been suboptimal in 45% of cases.8 Concerns regarding the quality of risk assessment antenatally and the management of fetal wellbeing were recurring themes in previous confidential enquiry work. The current classification system of perinatal deaths is also limited as half of stillbirths are categorised as `unexplained'. In particular the role of identification of intrauterine growth restriction (IUGR) in stillbirth and neonatal deaths needs further study, especially with respect to the accurate diagnosis antenatally and at postmortem examination. To further our understanding of the causes of perinatal deaths two areas need improvement: the classification of stillbirths and neonatal deaths and the standardisation of the observations and measurements made at perinatal postmortem (see Section 5).

Data gathered about perinatal deaths are used to identify risk factors relevant to clinical practice and the planning of future services to improve the health and wellbeing of mothers and their babies. This report finds that mothers living in socially disadvantaged areas had a higher proportion of stillbirths and neonatal deaths than might be expected. Babies born to women of Black, Asian or Other ethnicity had higher stillbirth rates and neonatal mortality rates than women of White ethnic origin. Risk factors are not uniform across the various ethnic minority groups and further research is required to increase understanding of the socio-demographic profiles and determinants of health and wellbeing in pregnancy among Britain's ethnic minorities.

High-quality relevant information is fundamental to the identification of risk factors. It is a tribute to maternity professionals in England, Wales and Northern Ireland that CEMACH has continued to achieve such high ascertainment levels of perinatal deaths despite the challenges of reorganisation. However, assessing risk is also dependent on the provision of denominator data on all births. Information relating to maternity care based on all births is widely recognised to be inadequate and maternity information has been identified as a priority for the implementation of the National Service Framework for Maternity Services. CEMACH looks forward to working with the relevant national initiatives to improve the quality of maternity data for audit and surveillance activities.

From 2005 onwards, CEMACH plans to develop the reporting of perinatal mortality data in order to provide a more comprehensive picture of the factors associated with perinatal deaths in England, Wales and Northern Ireland (see Section 7). This, combined with improvements in maternity data on all births, will provide a more extensive and relevant picture of maternity and neonatal care in the future.