Many epidemiological studies base their classification of congenital cardiovascular malformations in newborns upon a single, initial diagnosis. This study aimed to evaluate the effect of subsequent diagnostic investigations on the results of epidemiological studies. We used diagnostic codes from the Baltimore-Washington Infant Study from the time of birth and at ~1 year of age. Odds ratios and 95% confidence intervals were used to identify associations between changes in diagnoses and infant characteristics, time period, that is, before and after introduction of color flow Doppler imaging, and diagnostic variables. Of the 3054 patients with data at both time points, 400 (13.1%) had diagnostic changes. For congenital cardiovascular malformations of early cardiogenesis, such as laterality and looping defects, conotruncal malformations, and atrioventricular septal defects, significant associations were observed between diagnostic change and case infants large for gestational age (odds ratio=0.22, p=0.01), diagnosed initially by echocardiography only (odds ratio=2.05, p=0.001), or with non-cardiac malformations (odds ratio=0.60, p=0.03). For all other congenital cardiovascular malformations, significant associations were observed with echocardiography-only diagnosis (odds ratio=1.43, p=0.04) and non-cardiac malformations (odds ratio=0.57, p<0.001). We found no statistically significant differences between risk factor odds ratios calculated using initial diagnoses versus those calculated using 1-year update diagnoses. Changes in congenital cardiovascular malformation diagnoses from birth to year 1 interval were significantly associated with infant characteristics and diagnostic modality but did not materially affect the outcome of risk factor associations.
Keywords: CHD; cardiovascular malformations; classification; epidemiology.