Birth Settings and the Validation of Neonatal Seizures Recorded in Birth Certificates Compared to Medicaid Claims and Hospital Discharge Abstracts Among Live Births in South Carolina, 1996-2013.

Objective Neonatal seizures in the first 28 days of life often reflect underlying brain injury or abnormalities, and measure the quality of perinatal care in out-of-hospital births. Using the 2003 revision of birth certificates only, three studies reported more neonatal seizures recorded among home births ​or planned out-of-hospital births compared to hospital births. However, the validity of recording neonatal seizures or serious neurologic dysfunction across birth settings in birth certificates has not been evaluated. We aimed to validate seizure recording in birth certificates across birth settings using multiple datasets. Methods We examined checkbox items "seizures" and "seizure or serious neurologic dysfunction" in the 1989 and 2003 revisions of birth certificates in South Carolina from 1996 to 2013. Gold standards were ICD-9-CM codes 779.0, 345.X, and 780.3 in either hospital discharge abstracts or Medicaid encounters jointly. Results Sensitivity, positive predictive value, false positive rate, and the kappa statistic of neonatal seizures recording were 7%, 66%, 34%, and 0.12 for the 2003 revision of birth certificates in 547,177 hospital births from 2004 to 2013 and 5%, 33%, 67%, and 0.09 for the 1998 revision in 396,776 hospital births from 1996 to 2003, and 0, 0, 100%, -0.002 among 660 intended home births from 2004 to 2013 and 920 home births from 1996 to 2003, respectively. Conclusions for Practice Despite slight improvement across revisions, South Carolina birth certificates under-reported or falsely reported seizures among hospital births and especially home births. Birth certificates alone should not be used to measure neonatal seizures or serious neurologic dysfunction.