- Mark J Rapoport: Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada. mark.rapoport@sunnybrook.ca
Driving a motor vehicle is central to the functional autonomy of patients with psychiatric illnesses. There have been many studies of the deleterious effects of psychotropic medications such as benzodiazepines, typical antipsychotics and tricyclic antidepressants (TCAs) on human motor skills; however, in the literature little attention has been paid to how such impairment affects driving ability. Computerised driving simulators offer a laboratory-based method of assessing the effects of specific psychotropic medications on driving abilities, in a standardised, controlled and safe manner. The purpose of the present article is to review research undertaken to-date on the effects of psychotropic medications on computer-simulated driving. A search of various databases, including MEDLINE, EMBASE and PsycInfo, was conducted. Forty-one articles assessing the impact of psychotropics on computer-simulated driving were identified. The pooled total number of subjects assessed in these simulator studies was 1336 (mean sample size 30.36 [SD 35.8]). The most common outcome measures in the various studies were speed, steering, deviation from lateral position (tracking, lane drifting), reaction time or braking accuracy, driving errors (e.g. errors in turning, coordination, gap acceptance, signalling, following distance) and vehicle collisions. The results of the studies were quite variable; however, the most common drug-related impairments included those of tracking and reaction time. Benzodiazepines and TCAs were most commonly associated with impairment, although the level of impairment was dependent on the population studied, the dose and the time of testing relative to drug administration. Computer-simulated driving provides a useful tool to research psychotropic-related impairment of driving abilities. Limitations of currently available data include the lack of generalisability, standardisation and small sample sizes.