Ascertainment of congenital malformations: a comparative study of two systems.

STUDY OBJECTIVE

The aim was to compare the Office of Population Censuses and Surveys (OPCS) notification system for congenital malformation surveillance and the Liverpool Congenital Malformations Registry (LCMR) with respect to efficiency and uniformity of ascertainment, diagnostic accuracy, and overreporting of minor malformations.

DESIGN

Manual matching of computer listings was done, using date of birth, sex, birthweight, and health district of residence. Maternal age was used to confirm the match.

SETTING

Data were collected in the English health districts of Liverpool, St Helens and Knowsley, Southport and Formby, South Sefton, and Wirral over the years 1980-1985.

SUBJECTS

1959 malformed children notified to OPCS and 2649 notified to LCMR were assessed.

MAIN RESULTS

35.5% of malformed children ascertained by LCMR within seven days of birth were not notified to OPCS; 35.7% of cases reported to OPCS were exclusions from the LCMR protocol as being trivial malformations according to EUROCAT (European Registry of Congenital Anomalies) guidelines. Misclassification was infrequent but in 11.2% of cases the additional malformations present were not notified to OPCS. Conditions readily diagnosed at birth, such as neural tube defects, exomphalos, and facial clefts, were well ascertained by OPCS but others such as oesophageal atresia and Down's syndrome were not.

CONCLUSIONS

It is important that national surveillance of congenital malformations should continue. However, several modifications to the present OPCS monitoring system are necessary, including greater standardisation of data collection, the exclusion of trivial and clinically non-significant malformations, and the inclusion of data on therapeutic abortions performed for fetal abnormality. These issues are being addressed by OPCS.