Longitudinal assessment of post-concussion driving reaction time

Concussed patients present multiple neurocognitive and motor impairments including slowed reaction time (RT), a function essential to driving. We compared driving RT between concussed and non-concussed individuals across their concussion recovery (aim 1) and explored whether clinical concussion outcomes were correlated with driving RT uniquely in the concussion group (aim 2).

We recruited collegiate athletes (26 concussed and 23 age- and sex-matched controls) to complete the sport concussion assessment tool (SCAT5), a computerized neurocognitive test (CNS Vital Signs), and a driving simulation across 3 timepoints: ≤72 h, asymptomatic, and unrestricted medical clearance. RTs were recorded in response to 4 unanticipated driving events. CNSVS included 10 measures of cognitive function. General linear mixed models assessed interaction between group and time for aim 1 and group and concussion assessment outcome for aim 2 (α = 0.05). Pairwise comparisons with Cohen's d values were used following significant interactions and main effects.

There was a significant main effect for timepoint, such that pedestrian RT was slower at the ≤72-h timepoint relative to both the asymptomatic (p value = 0.023) and unrestricted medical clearance (p- value = 0.022). There were no other significant group-by-timepoint interaction or timepoint main effects for yellow stoplight RT (p-value range = 0.334-0.798), vehicle incursion RT (p-value range = 0.234-0.925) or vehicle cross RT (p-value range = 0.177-0.364). There was no significant group main effect (p-value range = 0.077-0.955), assessment outcome main effect (p-value range = 0.099-0.999) or interaction (p-value range = 0.103-0.998) for predicting any of the RTs, except for executive function (p = 0.046), motor speed (p = 0.006), and psychomotor speed (p = 0.027) predicting vehicle cross RT regardless of group.

This study demonstrates that driving RT may not differ between acutely concussed and healthy individuals or may not be detected on a short, simulated drive. Current clinical concussion outcomes poorly relate to driving RT. More research is needed to determine when it is safe to return to driving post-concussion.