Using hospital discharge data to track inflicted traumatic brain injury

Background: Inflicted traumatic brain injury (inflicted TBI) in infants has been the subject of recent preventive interventions, yet the burden of this problem is poorly defined. This article estimates the national incidence of hospitalizations due to inflicted TBI in infants aged <1 year using hospital discharge databases (HDDs) over 3 years, compares findings to previous infant inflicted TBI studies, and assesses differences in case composition and coding over time. Methods: The Kids' Inpatient Databases (KIDs) for 1997, 2000, and 2003 were used to estimate the incidence of inflicted TBI hospitalizations in infants. Cases were identified by the presence of ICD-9-CM codes indicating both TBI and abuse. Incidence estimates were generated by dividing the weighted number of inflicted TBI cases identified by the year-specific infant population. Changes over time were evaluated using Cochran-Armitage and ANOVA statistical tests. Results: The 1997, 2000, and 2003 KIDs yielded infant inflicted TBI incidence estimates of 27.5 (95% CI=20.8-34.0), 27.5 (95% 6=22.6-32.3), and 32.2 (95% CI=26.9-37.4) cases per 100,000 infants per year, respectively. KIDs estimates were within the range of previous inflicted TBI incidence studies that used active surveillance. Over time, the proportion of cases covered by Medicaid increased (p<0.0001). Conclusions: The KIDs call be used to generate incidence estimates of infant inflicted TBI hospitalizations that are consistent with previously published active surveillance estimates. HDDs call be used to assess the scope of infant inflicted TBI as well as trends in case composition and coding practices. Such assessments may be useful for the evaluation of prevention programs.