Abstract

OBJECTIVE: To demonstrate the use of aggregated, locally collected birth notification data to examine trends in birth-weight specific survival for singleton and multiple births. DESIGN: Retrospective analysis of 171,527 notified births and subsequent infant survival data derived from computerised community child health records. Validation of data completeness and quality was undertaken by comparison with birth and death registration records for the same period. SETTING: Notifications of births in 1989-1991 to residents of the North Thames (East) Region (formerly North East Thames Regional Health Authority). OUTCOME MEASURES: Birthweight specific stillbirth, neonatal, and postneonatal death rates. RESULTS: There was close correspondence between the notification and registration data. For 96% of the registered deaths a birth notification record was identified and for the majority of these the death was already known to the Community Child Health Computer. Completeness of birth-weight data, particularly at the lower end of the range, was substantially better in birth notification data. Comparison with the most recent published national data relating to birthweight specific survival of very low birthweight singleton and multiple births suggests that the downward trend of mortality is continuing, at least in this Region. CONCLUSIONS: The use of routinely collected aggregated birth notification data provides a valuable adjunct to existing sources of information about perinatal and infant survival, as well as other information regarding process and outcome of maternity services. Such data are required for comparative audit and may be more complete than that obtained from registration or hospital generated data.