AbstractA marked increase or “epidemic” in ventricular septal defects (VSD) in recent years has been reported by the Center for Disease Control. Many pediatric cardiologists believe that this increase is simply a reflection of more intensive diagnosis and evaluation of infants throughout the country. Yet to our knowledge there has been no objective evidence for this explanation. We evaluated this possibility by considering records on live births occurring in 1970–1983 in the counties surrounding Albany, New York. In that period a single group of pediatric cardiologists has been evaluating all infants with suspected or confirmed cardiac defects in this area. We limited this analysis to ventricular septal defects unassociated with any cardiac syndrome complex. Thus, VSDs occurring as part of cyanotic heart disease or other complex cardiac “syndromes” were excluded. Consistent with the reported national trend, the estimated prevalence rate of ventricular septal defects diagnosed under 1 year of age in this period has increased from 1.0 per 1,000 live births in 1970 to 4.0 per 1,000 in 1983. Several factors indicate that this rise resulted from more intensive diagnoses at earlier ages in this period: (1) for all patients with VSD, the median age at diagnosis fell from 251 days in 1970 to 25 days in 1983; (2) the most severe VSDs (based on an index derived from pathophysiologic data) were constant from 1972 afterward at an estimated rate of about one per 1,000 (although the estimated rates were lower than this in 1970 and 1971) while the milder defects rose from an estimated rate under 0.5 per 1,000 to over 3.0 per 1,000 in this interval; (3) changes in diagnostic criteria for VSD occurred, so that a murmur thought to be innocent in 1970 might, if heard later in the interval, lead to a diagnosis of VSD; (4) echocardiography began in the latter part of the interval, and the sharpest increase in the estimated rate of VSD occurred after its introduction in 1977 in our area; (5) the prevalence of complex VSD remained relatively stable throughout the period, being 0.4 per 1,000 live births in the first 3 years of the study and 0.5 per 1,000 in the last 3 years; (6) there was an increase in diagnoses of spontaneous closure of VSDs from 7% in 1970 to 32% in 1983. All of these factors are consistent with the view that the increase of VSD resulted from changes in ascertainment due to changes in methods and criteria for diagnosis of th