The usefulness of general practice data to describe the mortality risk in children with congenital anomalies: a cohort study

Abstract

BackgroundMajor congenital anomalies occur in 2·5% of livebirths in the UK and are an important cause of infant death. However, little is known about postneonatal mortality risks for these children because of a lack of long-term follow-up data at a population level. Furthermore, how mortality compares with the general population is underinvestigated. We compared mortality rates of children with and without major congenital anomalies. MethodsA population-based cohort of 334 599 children born 1990–09 was obtained from the Health Improvement Network, a UK database of prospectively collected primary care data. Diagnoses of major congenital anomalies were classified with European Surveillance of Congenital Anomalies criteria. Mortality rates between age 1 month and 5 years were calculated per 1000 person-years for children with any major congenital anomaly and by 15 system-specific subgroups. Cox proportional hazards were used to compare the mortality risk in children with and without major congenital anomalies where the proportional hazards assumption held. FindingsMortality rate was 0·35 per 1000 person-years (95% CI 0·32–0·38) in the general population compared with 2·92/1000 (2·40–3·54) for all children with major congenital anomalies (n=6648). Hazard of death was nine times greater for children with major congenital anomalies than for those without (unadjusted hazard ratio 8·74, 95% CI 7·04–10·87) and 20 times higher for those with congenital heart defects (20·5, 15·1–27·8). In the 81% (269 802/334 559) of the population for whom linked maternal data were available, adjustment for sex, socioeconomic status, year of birth, and maternal age and smoking status did not change hazard estimates. The mortality rate was consistently higher for children with than for those without major congenital anomalies across all system-specific subgroups. InterpretationUse of routinely collected health-care data provides estimates of the relative size of mortality risks for children with major congenital anomalies in direct comparison with the general population. Results are consistent with a previously published study from the USA. Despite variation by system-specific subgroup, children with major congenital anomalies who survive the neonatal period remain at high risk of death. Clinical and social explanations for these differences need further investigation to identify potential opportunities to prevent deaths in such children. FundingRS is supported by a postgraduate scholarship from the Economic and Social Research Council.