USING LINKED DATA TO EVALUATE MOTOR VEHICLE CRASHES INVOLVING ELDERLY DRIVERS IN CONNECTICUT

A deterministic algorithm was developed which allowed data from Department of Transportation motor vehicle crash records, state mortality registry records, and hospital admission and emergency department records to be linked for analysis of the impact of motor vehicle crashes on the elderly (65 years of age and over) population. Elderly drivers were involved in 8.4% of the motor vehicle crashes in Connecticut in 1995. Elderly drivers were associated with 5.2% of the linked medical records and 3.2% of the fatalities. Of the elderly drivers with linked hospital visits, 81% were treated in the emergency department and discharged; the rest were admitted to hospital, with median length of stay of 4 days. Geographically, crashes involving elderly drivers showed a bias towards more rural areas and away from the areas showing the highest overall motor vehicle crash rates. Logistic regression showed that, compared to the general population, crashes involving elderly drivers were more frequently correlated with driver illness (as reported by traffic enforcement personnel), a construction zone, violating traffic control, or failing to grant right of way, and less frequently with drinking or aggressive or dangerous driving. Conditions of diminished visibility were not identified as a significant factor, but elderly drivers were significantly more likely to be in a crash involving striking a deer.