Audrey McKinlay, John D. Corrigan, Jennifer A. Bogner, L. John Horwood

To investigate the concordance between medically documented childhood traumatic brain injury (TBI) and recall of same by adults aged 35 years. A total of 962 birth cohort members from the Christchurch Health and Development Study available at the 35-year follow-up. Childhood TBI information prospectively collected yearly over ages 0 to 15 years as part of the Christchurch Health and Development Study. At age 35 years, cohort members were administered the Ohio State University TBI Identification Method (OSU TBI-ID) to elicit recall of TBIs with loss of consciousness (LOC). Ninety-four individuals reported 116 TBI events. Twenty-five TBI events resulting in LOC, 17 (68%) were recalled (true positives) and 8 (32%) were not recalled (false negatives). LOC was incorrectly recalled for 56 events (false positives), but 868 individuals correctly recalled no TBI event (no LOC). A further 35 events were (correctly) recalled for which a TBI had been recorded but no LOC (true negatives; 91.8%). We evaluated the utility of the OSU TBI-ID to identify adult recall of childhood TBI with LOC occurring 19 to 35 years earlier. Most of the cohort accurately reported whether or not they had experienced a medically attended TBI with LOC, indicating that a positive result from the OSU TBI-ID provides useful screening information.