National rates of birth defects among hospitalized newborns

BACKGROUND: The Healthcare Cost and Utilization Project (HCUP) family of hospital discharge databases offer an unprecedented opportunity to generate national estimates of newborn infants with birth defects. This report estimates national hospital admissions for newborn infants diagnosed with birth defects computed from HCUP and compares them to pooled prevalence figures computed from state birth defect surveillance systems. METHODS: HCUP-derived rates of 36 birth defects from 1997 through 2001 were compared to rates derived from pooled data reported by 26 state-based surveillance systems stratified by inclusion of elective terminations in case definitions. Rate ratios (RRs) were calculated for each birth defect by dividing the rate derived from HCUP by the rate derived from the relevant surveillance systems. RESULTS: HCUP newborn hospitalization rates for birth defects closely approximate pooled birth defect rates for surveillance systems that do not include elective terminations. HCUP rates were not significantly different for 35 of 36 defects. Overall, 20 HCUP rates were within 10% of state rates, 11 more were within 20% of state rates, and only 1 differed by more than 50%. HCUP rates compared most closely to state rates for cardiovascular (VSD RR = 0.98, ASD = 0.96, pulmonary valve atresia and stenosis = 0.92), orofacial (cleft palate RR = 1.10, cleft lip = 1.06), and genitourinary defects (obstructive genitourinary RR = 1.01, bladder exstrophy = 0.97). HCUP rates compared less favorably to rates derived from surveillance systems that included elective terminations. CONCLUSIONS: HCUP data approximate state-based surveillance system data for defects that are easily recognized in the newborn period and infrequently a cause for elective termination. HCUP data can be used to examine the impact of public health efforts on the number of infants born with birth defects as well as the cost and consequences of variations in the hospital management of birth defects.