Utility of Electronic Case Reporting for Case Identification in Texas Birth Defects Surveillance

The utility of electronic case reporting (eCR) in birth defects surveillance is unknown. This evaluation assessed whether electronic initial case reports (eICRs) can serve as a potential case identification source and how eICRs compare to electronic health records (EHRs) in capturing demographic and diagnostic information.

Cases were identified from the Texas Birth Defects Registry's eCR data stream. Upon case confirmation, the EHR was requested and abstracted following routine abstraction processes. Next, the eICR html file was abstracted. The number, range, and mean of coded birth defects, as well as pre- and postnatal procedures were calculated for both data sources. Concordance between abstracted variables from the EHR and the eICR was evaluated for non-missing data using weighted kappa agreement statistics in SAS.

There were many missing data from the eICR. Fewer birth defects and pre- and postnatal procedures were reported in the eICR compared to the EHR. Most variables had low concordance, while a few variables had high concordance (e.g., infant sex [kappa = 0.95], infant birthdate [kappa = 0.99], primary defect code/diagnosis [kappa = 0.83 using 6-digit British Pediatric Association [BPA] codes; kappa = 0.92, 4-digit BPA codes]).

eCR may be a viable source for timely potential case identification and may also facilitate timelier referral to social services among eligible cases. eCR data files are not standardized across facilities, lack critical variables permitting examination of birth defect risk factors, and are generally not a comprehensive resource for all pertinent data related to a birth defect case.