What can the UK learn from international comparisons of routinely collected perinatal data? UK perspectives on the Euro-Peristat project

Abstract

Listen

Translate article

BackgroundClaims that perinatal and infant mortality rates in the UK are higher than those of all comparable countries are often made in a spirit of moral panic, to argue for policies or interventions that may or may not be effective at reducing the rates. Crude perinatal and infant mortality rates tend to be included routinely in successive national and international indicator sets, sometimes being interpreted as measures of population health and sometimes as measures of quality of health care. Such claims inevitably ignore the effect of differences in childbearing populations, data collection methods, and inclusion criteria. The Euro-Peristat project was established to move beyond presentation of unidimensional indicators in isolation and to present data about the outcome of pregnancy for mothers and babies stratified by birthweight and gestational age while presenting comparative data about maternity care and the characteristics of the childbearing population to set them in context. MethodsCollaborators from 26 European Union (EU) member states plus Iceland, Norway, and Switzerland took part in the most recent phase. The indicators were selected with a Delphi process for the first data collection for the year 2000 and revised to compile data for 2004 and then 2010. They were designed to be compiled from routine national vital registration, clinical, or administrative data collection systems in the form of cross-tabulations of aggregated anonymised data. Each country provided information about methods of data collection and inclusion criteria. No rankings were applied and the countries were listed alphabetically according to their official EU names. Data were published in the form of graphs of indicators and tables of data. Stillbirth and neonatal mortality rates were calculated with WHO definitions. Where possible, indicators of outcome were constructed with common cutoffs on the basis of gestational age or birthweight, to avoid bias towards higher reported rates in countries in which births and deaths are registered at lower gestational ages. FindingsAll indicators showed wide variation between countries, although some of this variation could be attributed to differences in data collection and ascertainment. Because of differences in data collection, data for the countries of the UK were compiled separately. Stillbirth rates at or after 28 weeks of gestation ranged from 1·5 per 1000 total births in the Czech Republic to 4·3 per 1000 total births in France, with rates of 3·8 per 1000 total births in England and Wales, 3·6 per 1000 total births in Scotland, and 3·4 per 1000 total births in Northern Ireland being among the higher rates. Neonatal mortality rates at 24 weeks or more of gestation ranged from 0·8 per 1000 livebirths in Iceland and 1·1 per 1000 livebirths in Finland to 4·3 per 1000 livebirths in Romania, with rates of 2·0 per 1000 livebirths in England and Wales, 2·1 per 1000 livebirths in Scotland, and 3·0 per 1000 livebirths in Northern Ireland. Rates for some countries were based on small numbers of events. There were also differences in risk factors. InterpretationThe interpretation of these and other indicators are discussed more fully in the first and second European Perinatal Health Report. Infant and perinatal mortality rates in the UK were not the highest in Europe but they were above the median. To be useful for identification of scope for improvements, international comparisons of the outcome of pregnancy should take account of random variation and differences in methods of data collection and population characteristics. Simple league tables based on crude rates are unhelpful. FundingEuro-Peristat is funded by the Health Programme of the European Commission's Directorate General for Health and Consumers. The funding source had no role in the writing of this abstract or the reports on which it is based.