A population-based study of neonatal seizures in Fayette County, Kentucky: comparison of ascertainment using different health data systems.

Based on a population-based, retrospective cohort study of neonatal seizures in Fayette County, Kentucky, from 1985 to 1989, estimates of neonatal seizure risk were calculated from computerized databases including hospital medical records, birth certificates, and death certificates. Computerized tabulations of hospital discharge diagnoses identified 97% of cases with a positive predictive value of 75%. Birth certificates had poor sensitivity for neonatal seizures overall (37%), but identified 67% of infants who seized on the first day of life. However, careless completion of the birth certificate produced a high number of false positives. Death certificates identified no infants with neonatal seizures.