Background We used a randomised controlled trial, with nested process evaluation, to measure the impact of the Behind the Wheel program on driving and community participation of drivers aged 75 years and older in northwest Sydney.

Methods Driving exposure/week was measured by continuous in-vehicle monitoring for 12 months. The Keele Assessment of Participation (KAP), self-regulation behaviour profile and depressive symptoms were assessed at 12 months. Using intention-to-treat, generalised estimating equations modelled driving exposure, adjusting for weekly measures and ordinal regression for behaviour profiles. A logic model was built to explain program inputs, outputs and outcomes, based on relationships between process measures (program fidelity, acceptability, dose delivered and received) and program outcomes.

Results 380 drivers enrolled (mean age: 80 ± 4 years), 366/380 completed the study. There was no between group difference in distance driven/week over 12-months (−5.5 km, 95% CI: −24.5,13.5 km), or KAP (−0.1, 95% CI: −0.6,0.3). The intervention group were more engaged in self-regulation (OR: 1.6, 95% CI: 1.1,2.3). Older drivers with low-function in the intervention group were three times more likely to report depressive symptoms (OR 3.1, 95% CI: 1.04,9.2). Intervention participants who developed a retirement from driving plan, on average, reduced their total distance driven per week (38 km, 95% CI: −7.5,−68.7 km) and kilometres driven outside of daylight hours per week (7 km, 95% CI: −3.5,−10.4 km). Both understanding of program content (β = 2.1, p = 0.03, 95% CI: 0.2–4.1) and achieving a safe mobility plan (β = 3.3, p = 0.003, 95% CI: 1.2–5.5) were important to engagement in self-regulation. Females were 2.7 times more likely to develop safe mobility plans than men (OR 2.7, 95% CI: 1.1–6.9).

Conclusions The program engaged older drivers in self-regulation but this did not translate to reduced mileage. The logic model informs decision making to channel resources to those who will benefit most.