MBRRACE-UK perinatal mortality surveillance

1. Background

MBRRACE-UK report on perinatal mortality for the whole of the UK and Crown Dependencies, as well as for each individual nation. The Office for National Statistics (ONS) also publishes annual data on perinatal mortality in England and Wales. Here we explain the reasons behind the differences in these statistics, and why we use slightly different definitions of stillbirth and neonatal death compared to ONS.

ONS figures on perinatal mortality are based on all births and deaths registered via the General Register Office. This includes all stillbirths registered at 24 weeks or more gestation, in line with the Stillbirth (Definition) Act 1992, and neonatal deaths of babies born at any gestational age.

Our goal is to publish data on perinatal deaths that are clinically comparable across different organisations, such as healthcare providers or the devolved nations within the UK. Since registered stillbirths and neonatal deaths are affected by some inconsistencies in reporting between organisations, we apply a rigorous definition of deaths to ensure comparability.

There are five key differences in the way MBRRACE-UK and ONS report stillbirths and neonatal deaths.

1. We report stillbirths where the birth occurred from 24 completed weeks’, irrespective of when the baby died in utero (including some unregistered deaths), while ONS statistics are based only on registered stillbirths.
2. We report neonatal deaths from 24 completed weeks’ gestational age. ONS has no gestational age limit.
3. We don’t include deaths following termination of pregnancy, which are included in ONS figures.
4. We report deaths by the year of the baby’s birth, while ONS report deaths by the year of the baby’s death.
5. We report some rates by place of baby’s birth, and others by place of residence. ONS report deaths by place of residence.

2. Key differences

In the UK, there are inconsistencies in how stillbirths are registered when the baby dies in utero before 24 weeks of pregnancy but is born after 24 weeks. For this reason, we report the number of stillbirths based on babies who died before or during birth at 24 completed weeks of pregnancy or later (or a minimum birthweight of 400g where the length of pregnancy is unknown). This definition includes all registered stillbirths as well as unregistered deaths where the baby died in utero before 24 weeks of pregnancy but was born after 24 weeks.

Information on both the date of birth and the date of confirmation of the death in-utero is reported to identify these deaths. For multiple pregnancies, where there is more likely to be a longer gap between an in-utero death and the subsequent birth, deaths are excluded where the death is confirmed before 20 completed weeks' gestational age.

This leads to the inclusion of approximately 40 additional unregistered stillbirths in England and Wales in MBRRACE-UK stillbirth rates that are not captured in ONS figures.

There is wide variation across the UK in whether signs of life are reported in babies born before 24 weeks of pregnancy. As a result, there are inconsistencies in whether these deaths are reported as a miscarriage or registered as a live birth and subsequent neonatal death. This variation in practice affects the accuracy of neonatal death data. To reduce this impact, while we collect information on neonatal deaths from 20 weeks gestation, we report the number of neonatal deaths based on how many babies were born at 24 completed weeks gestational age or later who die in the neonatal period. If an accurate estimate of gestation is not available a minimum birthweight of 400g is used to determine eligibility.

Many of the data tables published by ONS include all registered neonatal deaths, regardless of the gestational age at birth. This could include neonatal deaths occurring as early as 16 weeks gestation, or even earlier in some cases.

As a result, MBRRACE-UK neonatal mortality rates do not include around 650 to 750 neonatal deaths a year in England and Wales that are currently included in ONS figures, as the babies were born before 24 weeks' gestation.

Stillbirths and neonatal deaths following termination of pregnancy are outside of the scope of perinatal mortality surveillance by MBRRACE-UK, as they have different causes compared to other deaths and are monitored by other organisations such as the Department of Health and Social Care. Excluding terminations of pregnancy from perinatal mortality statistics also improves comparability between countries, where gestational limits for termination of pregnancy may vary or termination of pregnancy may not be allowed.

In many cases, it is not clear from the cause of death information in the stillbirth or neonatal death registration that the death followed a termination of pregnancy. We use rigorous methods to identify these cases, with healthcare professionals reporting when a stillbirth or neonatal death occurs following a termination of pregnancy.

This results in around 5 to 10 neonatal deaths (at 24 weeks’ gestation or later) and around 350 to 450 stillbirths (at 24 weeks’ gestation or later) each year in England and Wales that are not included in MBRRACE-UK rates but are reported in ONS figures.

Another, though smaller, reason for the difference is that we report deaths based on the year of the baby’s birth. For example, a baby born in December 2015 who died in January 2016 would be included in the 2015 statistics, reflecting a true birth cohort. In contrast, ONS figures are based on deaths within a single calendar year (a death cohort), leading to slight differences compared to MBRRACE-UK statistics.

In order to calculate mortality rates, we need a common denominator that includes all births and perinatal deaths. For organisations responsible for providing or commissioning maternity and neonatal services, such as trusts and health boards, we use the baby’s place of birth as the denominator. Reporting by place of birth provides valuable insight into the care pathways for babies, especially those who die during the neonatal period.

For organisations focused on public health, such as local authorities, and for calculating national rates, we report based on place of residence. This approach reflects the populations these organisations are responsible for serving.

3. Reviewing future reporting

Given the wide variation across the UK in reporting signs of life in babies born before 24 weeks of pregnancy, MBRRACE-UK has published guidance on assessing signs of life for babies born before 24 weeks gestation to improve consistency in death reporting. We collect information on both registered neonatal deaths and unregistered deaths of babies born with no signs of life from 22 weeks onwards. This allows us to monitor variations in reporting practices and regularly assess whether it is appropriate to publish mortality rates from 22 weeks gestation onwards. Aligning reporting from 22 weeks would be consistent with practices in many European and high-income countries, enabling better comparisons.

4. Version history