Using a national register for the epidemiological study of congenital heart defects.

Data validity is a fundamental problem in epidemiology. An objective of birth defect registers is the collection of high quality data, often from a number of sources, for use in epidemiological research. This paper examines the use of data from a national register in New Zealand in the study of congenital heart defects. The prevalence rates of congenital heart defects are shown to depend on the definition of the terms employed, the methods and completeness of the case ascertainment, the correctness of the diagnoses, the nature and size of the population under study, and the duration of followup. Recommendations are made for increasing the utility of this register in the study of this major group of malformations.