Validity of diagnoses of cardiac malformations in a Danish population-based hospital-discharge registry

Abstract

Studies of drug-induced birth defects require large sample sizes from high-quality datasets. The Danish discharge registries allow large sample sizes, but the quality of the diagnosis coding is largely unknown. We therefore examined the validity, expressed as the positive predictive value, of discharge diagnoses of cardiac malformations registered in the Danish population-based North Jutland County Hospital Discharge Registry. We reviewed the medical records of 418 (99%) of the 423 children with a first-time diagnosis of cardiac malformation between 1 January 1994 and 31 March 2002. Record review confirmed the diagnosis in 372 children, yielding an overall positive predictive value of 89% (95% confidence interval 86%, 92%) and a prevalence rate of 7.6 per 1000 livebirths. The positive predictive value was between 82% and 86% for three of the four commonest malformations, whereas that of the fourth, patent arterial duct, was only 41%. In conclusion, this registry is a valuable resource for studies of risk factors for cardiac malformations as a group and for studies of risk factors for several types of cardiac malformation. Prevalence rates based on data from this registry should be interpreted with caution.