"Did you fall asleep?" - Younger and older drivers' recollection of prior sleepiness while driving

Determining causality and driver culpability in fall asleep crashes requires extensive forensic examination, with post-crash interviews forming a critical part of the process. To date, there is no evidence regarding whether drivers' recollection of sleepiness and associated symptoms accurately reflects driving impairment during an earlier drive. To address this, our study examined whether established questionnaires assessing sleepiness (Karolinska Sleepiness Scale) and sleepiness symptoms (Sleepiness Symptoms Questionnaire) taken post-drive accurately reflected earlier driving impairment as characterised by near-crash events and lane deviations during a closed loop track drive. Sixteen younger (21-33 years) and seventeen older (50-65 years) drivers completed two 2 h afternoon track drives under supervision of a licensed driving instructor under two conditions: well rested (8 h sleep opportunity) and sleep deprived (0 h sleep). For younger drivers, all sleepiness symptoms (except mind wandering) significantly predicted prior severe (near-crashes AUC 0.78-1.00) and moderate (lane deviations AUC 0.78-0.94) driving impairment, with good to excellent accuracy. For older drivers, only severe impairment was accurately reflected in post-drive ratings (near crashes AUC 0.86-0.94), and for fewer symptoms (8/9 vs. 6/ 9). Optimal thresholds for the frequency of these symptoms, in addition to thresholds for minimum acceptable false positives are presented. Our data suggest that post-crash interviews should utilise questions around subjective sleepiness and associated symptoms, providing evidence regarding the presence and awareness of drowsiness while driving, taking into account the reliability of self-report evidence. This may aid in the investigation of fall-asleep crashes where drowsiness is suspected.