[Discrepancy between results of registration of perinatal cause of death by the CBS (Central Bureau of Statistics) and by personal studies in the Delft-Westland-Oostland region].

OBJECTIVE

To determine the reliability of the Dutch registration of causes of perinatal death by the Centraal Bureau voor de Statistiek (CBS, Central Statistics Office).

DESIGN

Prospective, descriptive.

SETTING

Region Delft-Westland-Oostland, the Netherlands.

MATERIAL AND METHODS

The registration was based on data concerning all deliveries of women domiciled in the region, irrespective of the ultimate place of delivery, during 1983-1992. By linking, in retrospect, a prospective regional registration system for perinatal mortality within the region anonymously to the CBS registration, the reliability of the latter registration with regard to the causes of death was determined. To establish the causes of death, all available data were judged by a team consisting of a gynaecologist, a paediatrician and a paediatric pathologist. The diagnoses were classified with the aid of the International Classification of Diseases 9 (ICD-9).

RESULTS

In 10 years, 28983 children were born in the region. Over this period, the CBS recorded 227 cases of perinatal mortality. The actual perinatal mortality was calculated as at least 247 cases. In 32% of the cases of stillbirth, the cause of death was not known at the CBS. Of the 82 cases in which the CBS had recorded a diagnosis, the causes of death were in agreement with those found in the regional study in 46%. Of the first-week mortality, the diagnosis was unknown at the CBS in one case and of the remaining 76 cases, the registration of the cause of death was the same in 68% of the cases.

CONCLUSIONS

Registration of the causes of death regarding perinatal mortality and particularly stillbirth by the CBS shows gaps, mostly due to incorrect reporting of the cause of death by the treating physician or autopsist, due to the fact that at the time of notification the morbid-anatomical diagnosis and/or laboratory data were not complete. For the study of the backgrounds of perinatal mortality the current CBS registration of causes of death appears unsuitable.