Evaluation of Data Sources for Critical Congenital Heart Defect Case Ascertainment in North Carolina

Background Surveillance of critical congenital heart defects (CCHDs) is important for research and public health practice. The recent implementation of universal pulse oximetry (POX) screening may provide an additional resource for improving the completeness of CCHD surveillance data. This analysis evaluates the utility of multiple data sources, including POX records, for CCHD case identification. Methods Using the cohort of CCHD cases confirmed by the North Carolina Birth Defects Monitoring Program born 2018–2021 ( N = 1035), sensitivity values were calculated for birth, infant death, and fetal death certificates; hospital discharge records; and POX records. Sensitivity was defined as the percentage of confirmed CCHD cases identifiable from a given data source. Sensitivity values were also stratified by demographic and clinical characteristics of cases. Results The sensitivity of the data sources for case identification varied widely: birth certificate (6.7%; 95% confidence interval [CI]: 5.1–8.3); fetal death certificate (29.8%; 95% CI: 18.0–41.7); POX record (59.5%; 95% CI: 56.2–62.8); infant death certificate (73.0%; 95% CI: 66.7–79.3); and discharge record (89.9%; 95% CI: 88.0–91.8). Sensitivity values for some data sources were significantly higher for cases delivered in larger hospitals, transferred after delivery, admitted to the neonatal intensive care unit, or prenatally diagnosed. Conclusions The POX record was only moderately sensitive, in part due to missing records, limiting its added value for case finding, though improvements in reporting may increase its utility. To help optimize staff time and resources, surveillance programs should consider the utility of supplemental data sources for CCHD case ascertainment.