Comparison of self-reported crashes, state crash records and an on-road driving assessment in a population-based sample of drivers aged 69-95 years.

OBJECTIVE The study evaluated associations between outcome measures used in driving research including self-reported crashes, state crash records, and an on-road driving test (ORT). METHOD A total of 750 community dwelling participants aged 69 to 95 were recruited via the electoral roll into a study on injury prevention. Of these, 509 were drivers, and data on self-reported crashes, and either state crash records, or an on-road assessment were available for 488. Crash history data were obtained from state records (five-year retrospective and twelve-month prospective), retrospective self-report (five-year) and prospective monthly injury diaries (twelve months). A subsample completed an on-road driving test. RESULTS During the last five years, 22.3% reported a crash, 10.0% reported a crash in the twelve-month follow-up period, 3.2% of the sample had state crash records during the previous five years, and 0.6% had state-recorded crashes during the twelve-month follow-up period. State crash records did not agree with any other outcome measure. Those who scored 5 or less on the ORT were more likely to report a crash in the past five years (55.4% vs. 36.8%; p = .009). Results did not differ when participants with probable dementia were excluded (n = 2). CONCLUSION The results suggest that caution should be applied when using state crash records as an outcome measure in driving research and suggest that in the Australian context, retrospective self-reported crashes over five years are preferable when objective measures of driving performance are unavailable.