This review assessed the impact of psychotropic medications on simulated driving and concluded that impairments varied with the population studied, dose and time of testing, but were most commonly associated with benzodiazepines and tricyclic antidepressants. Overall, the reliability of these conclusions is limited by the poor reporting of the review methods and the apparent methodological limitations of the study data.

To investigate the effects of psychoactive or psychotropic medications on computer-simulated driving.

MEDLINE, EMBASE, PsycINFO and CanDRIVE were searched up to 31st March 2005; the search terms were reported.

Study designs of evaluations included in the review

Studies using a control group were eligible for inclusion. The included designs were placebo-controlled, double-blind and/or crossover randomised controlled trials (RCTs), non-randomised comparative studies and cross-sectional observational studies.

Specific interventions included in the review

Studies of benzodiazepines, newer (i.e. non-benzodiazepine) sedative-hypnotics, antidepressants, antipsychotics, lithium or newer mood stabilisers were eligible for inclusion. Studies of the effects of psychotropic medication in combination with alcohol were excluded. The majority of the included studies evaluated benzodiazepines or compared them with other medications including hypnotics, antidepressants/anxiolytics, antipsychotics and lithium. The remaining studies evaluated the effects of antidepressants/anxiolytics, antipsychotics, lithium or hypnotics. Further details of the specific drugs and dose regimens were provided.

Participants included in the review

The authors did not specify the types of participants that were eligible for inclusion. The participants in the included studies were mainly younger adults, with only one study investigating effects in older adults over the age of 65 years. Studies usually included more men than women and the majority of participants in the control groups were healthy. Only a small number of included studies examined the effects of drugs in patient populations such as those suffering from anxiety, insomnia, mood disorders, Meniere’s disease and schizophrenia.

Outcomes assessed in the review

Studies had to measure driving ability using a car simulator. Car simulators were categorised as complex, intermediate or simple. The included studies reported various outcomes: speed; steering; deviation from lateral position including tracking and lane drifting; reaction time or braking accuracy; driving errors including turning errors, coordination, gap acceptance, signalling and following distance; and vehicle collisions.

How were decisions on the relevance of primary studies made?

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

The authors did not state that they assessed validity. However, some aspects of study design, including random allocation and blinding, were reported.

The authors stated that the data were extracted in a non-blinded manner, but did not state how many reviewers were involved. Data relating to the time of testing and the type of driving simulator used were extracted and categorised as simple, intermediate or complex. Complex simulators used a realistic driver interface and simulated the road environment and driver tasks; simple simulators used a button response panel or steering wheel and foot pedal(s) with a computer panel. Only significant outcome effects were reported.

How were the studies combined?

The studies were grouped according to drug type and timing of effects after intervention (2 hours or less, 2 to 5 hours, and next morning) and combined in a narrative.

How were differences between studies investigated?

Some differences were described in the review text and were also apparent from the data tables.

Forty-four studies (n=1,336) were included: 33 placebo-controlled, double-blind crossover trials (29 of which were randomised), 6 parallel-design studies (4 of which were double-blind) and 5 cross-sectional studies. The sample sizes varied from 8 to 180 participants.

Benzodiazepines: 35 studies, 33 of which were carried out in healthy individuals, showed varying effects of different doses of lorazepam, diazepam, brotizolam, nitrazepam, lormetazepam, tiazolam and chlordiazepoxide. The effects varied with dose and time of testing relative to the dose, but were commonly associated with adverse outcomes for tracking and reaction time. Further details were reported.

Newer non-benzodiazepine hypnotics: 6 studies, including five of 7.5 mg zopiclone and five of different doses of zolpidem, found varying effects on numerous different outcomes. The effects varied according to the time of testing relative to the dose. Further details were reported.

Antidepressants and anxiolytics: 4 studies reported on 50 mg amitriptyline (3 studies), 50 mg trazodone (1 study), 40 mg fluoxetine (1 study), 30 mg paroxetine (1 study), 20 mg mirtazapine (1 study) and 20 mg buspirone (1 study). Few adverse effects on reaction time were reported, and those that were did not appear to be long lasting. Further details were reported.

Lithium: 4 studies reported data on lithium. One study was in healthy controls, two in patients with mood disorders, and one RCT in patients with Meniere’s disease. The effects varied and were generally adverse, with the exception of patients with Meniere’s disease, where reaction time and speed were not affected by lithium after 6 months of treatment. Further details were reported.

Antipsychotics: 5 studies in healthy individuals assessed the effects of 50 or 200 mg amisulpride, 50 mg chlorpromazine, 25 mg thioridazine, 0.5 mg haloperidol and 0.5 mg flupenthixole. In general, the studies reported no differences in reaction time, tracking errors and mistakes. One study assessed the effects of flupenthixole and fluphenazine (doses not stated) in patients with schizophrenia. This reported longer reaction times and worse driving accuracy in comparison with control groups. One study also suggested that adverse effects were worse with typical antipsychotics than with atypical antipsychotics. Further details were reported.

The level of impairment varied with the population studied, dose and time of testing, but effects were most commonly associated with benzodiazepines and tricyclic antidepressants. The most commonly reported drug-related impairments were in tracking and reaction time. However, the data on which these conclusions are based is limited by a lack of generalisability and standardisation, and small sample sizes.

This review answered a broad research question using poorly described methods, which makes it difficult to assess the level of reviewer bias and error. The risk of publication bias is also unclear, but is likely given the lack of any specific attempts to locate unpublished material. The authors’ synthesis was appropriate given the large heterogeneous group of studies which negates meta-analysis. It is also difficult to assess the reliability of the findings given the lack of any formal assessment of study validity, although the authors did refer to the overall methodological limitations of the studies. Overall, the reliability of the authors’ conclusions is limited by the poor reporting of the review methods and the apparent methodological limitations of the study data.

Practice: The authors stated that it is vital for clinicians to understand the effects of psychotropic medications on driving given their high and increasing prevalence.

Research: The authors stated that further studies are needed in patients with psychiatric illnesses across the age spectrum of drivers and, in particular, amongst older adults with and without cognitive disturbances. They also stated that as new, more sophisticated driving simulators are developed, protocols and external validations are required for standardisation.

Physican’s Services Inc. Foundation.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.