BACKGROUND: Ohio has the 11th highest infant mortality rate in the United States as of 2015. Two sources are used to ascertain how infants die in Ohio, vital statistics (VS) and child fatality review (CFR). CFR determines cause of death through a multidisciplinary board that reviews medical, law enforcement, death scene investigation, and social work records. We sought to use CFR findings to improve VS reporting.
METHODS: Infant mortality was defined as a death of an Ohio resident in Ohio before age 1 year. We linked CFR and VS datasets by infant during 2009–2013 using personal identifying information. We grouped causes of death into 11 categories on the basis of CFR classification. Kappa statistic was used to measure concordance.
RESULTS: Records were linked for 96% (4,985) of all VS infant deaths. Concordance was excellent for asphyxia (kappa 0.95), moderate for congenital anomalies (kappa 0.64), and weak (kappa <0.6) for 9 remaining categories. The largest discordance occurred where 1,100 (22%) infants were assigned prematurity by CFR and perinatal condition by VS. Cause of death was missing from 359 CFRs, compared with 0 from VS; 64% of these were coded as sudden infant death syndrome by VS.
CONCLUSIONS: Concordance between CFR and VS was low for the majority of categories. Although CFR is often considered the gold standard in Ohio, it is limited by missing data and insufficient detail for premature deaths. By using cause of death ICD-10 codes from VS additional detail on underlying causes of prematurity was added. Linking CFR and VS provides a more complete understanding of infant deaths in Ohio and may lead to improved VS reporting.