Sj. Olsen et Ms. Durkin, VALIDITY OF HOSPITAL DISCHARGE DATA REGARDING INTENTIONALITY OF FATALPEDIATRIC INJURIES, Epidemiology, 7(6), 1996, pp. 644-647
This study of hospitalized fatal injuries to children in New York City
(1991-1992) compares the classification of injury intention on hospit
al discharge data with that on death certificates in 302 injuries for
which records could be linked. Using death certificates as the criteri
on, the sensitivity of hospital data for identifying intentional injur
ies ranged from 0.50 for children under age 5 years [95% confidence in
terval (CI) = 0.27-0.73] to 0.85 for adolescents (95% CI = 0.80-0.90),
and from 0.75 for white children (95% CI = 0.65-0.83) to 0.89 for non
white children (95% CI = 0.82-0.93). Hospital discharge data for this
population appear to understate both the incidence and the proportion
of fatal injuries due to intentional causes; particularly those occurr
ing to very young children and to white children.