(790) Driving Patterns and Safety After Critical Illness: A Naturalistic Study With In-Vehicle Sensors

Introduction: Following critical illness, 50% of adults experience new or worsening physical and/or cognitive impairments affecting functional outcomes, including driving. The goal of this study was to use objective driving data collected by in-vehicle sensors to examine driving-related decisions and behavior patterns after ICU discharge.

Methods: A longitudinal naturalistic driving study was conducted with ICU survivor participants providing informed consent at hospital discharge (IRB #023-040). We installed in-vehicle sensors in participant’s personal vehicles for 6 months after discharge. Eligible participants were >40 years old, delirium positive, and were regular drivers in the 4 weeks prior to critical illness. We measured vehicle speed, acceleration, and location for each driving trip, enabling evaluations of hard braking events, swerves, and other measures (e.g., right/left turn ratios, trip distance). Validated acceleration thresholds were used to classify near-crash events.

Results: In 458 driving trips during the 6-months after hospitalization, 33% of participants did not resume driving during the first 6 months after hospital discharge. For participants that resumed driving, the mean driving resumption time was 38 days (SD=32 days). Returning drivers were classified into phenotypes of 1) immediate and full return; 2) delayed and full return; and 3) full return with subsequent reduction. Most trips occurred during weekday and daytime hours. Fifty percent of drivers that returned to driving experienced a near-crash event during the 6 months following ICU discharge.

Conclusions: Naturalistic in-vehicle driving sensors offer unique data in the evaluation of driving-related decisions and behavior patterns after critical illness. The rate of near-crash events detected in the first 458 driving trips suggest that full-scale investigation of near-crash and crash causation mechanisms are salient to critical care recovery. Patient-centered tools and care standards may be needed to manage the vulnerabilities specific to ICU survivors for driving resumption or driving cessation decision-making