Driving Performance Under Treatment of Most Frequently Prescribed Drugs for Mental Disorders: A Systematic Review of Patient Studies

Abstract Background Mobility is important for daily life functioning, with particular challenges regarding road safety under pharmacological treatment in patients with a psychiatric disease. Methods According to PRISMA guidelines, a systematic literature search on PubMed database (January 1970 to December 2020) was performed. Primary endpoints were driving performance in on-road tests, driving simulator performance, or psychomotor and visual perception functions assessed to estimate fitness to drive according to legal regulations in patient studies. Results Forty studies were identified (1533 patients, 38% female, median age 45 years), of which more than 60% were cross-sectional and open-label trials. Under steady-state medication, 31% (range 27%–42.5%) of schizophrenic or schizoaffective patients under antipsychotics and 18% (range 16%–20%) of unipolar and bipolar patients under antidepressants showed severe impairment in skills relevant for driving. Data point to an advantage of second-generation antipsychotics compared with first-generation antipsychotics as well as modern antidepressants over tricyclic antidepressants with respect to driving. Most patients significantly improved or stabilized in driving skills within 2–4 weeks of treatment with non-sedative or sedative antidepressants. Diazepam significantly worsened driving the first 3 weeks after treatment initiation, whereas medazepam (low dose), temazepam, and zolpidem did not impair driving. In long-term users of sedating antidepressants or benzodiazepines, impairments in on-road tests were not evident. Conclusion The available evidence suggests that psychopharmacologic medicines improve or at least stabilize driving performance of patients under long-term treatment when given on clinical considerations. To enhance treatment compliance, existing classification systems of medicinal drugs concerning impact on driving performance should also incorporate information about effects of long-term-treatment.


Introduction
Modern societies demand a high grade of mobility, and there is evidence that driving cessation, for example, in cases of aging or chronic illness, affects social and economic well-being with impacts on health functioning (Edwards et al., 2010). Approximately 67% of patients with a psychiatric disease have a drivers' license; 77% of these patients report driving regularly with their cars and most of them (88%) use prescribed medication. Closer inspection of data also indicates that driving restrictions largely affect the social functioning of these patients (Brunnauer et al., 2016). Thus, road safety under pharmacological treatment is of great relevance for patients with a psychiatric disease and is therefore frequently discussed in clinical practice.
Besides their major mode of action, medicinal drugs can impair behavior most frequently because of sedating CNS adverse effects such as drowsiness or inability to concentrate. Behavioral correlates of drug-induced impairments may be a disruption of neuropsychological processes controlling behavior. Evidence supporting the concept of behavioral toxicity (e.g., DiMascio and Shader, 1968;Hindmarch I, 1994) comes from epidemiological data. Albeit causal relationships can hardly be drawn, there is considerable evidence that the use of psychoactive drugs is associated with an increased risk of traffic injuries, with a particular concern regarding benzodiazepines and tricyclic antidepressants (e.g., Ray et al., 1992;Leveille et al., 1994;Mura et al., 2003;Barbone et al., 1998;Bramness et al., 2008;Dassanayake et al., 2011;Elvik, 2013). Especially elderly users of sedating antidepressants have a more than twofold increased risk of being involved in road traffic accidents (Ray et al., 1992;Leveille et al., 1994;Dassanayake et al., 2011). Benzodiazepine use and the association with traffic accidents seems to be more frequent in younger drivers, and risk markedly increased by co-ingestion of alcohol (Dassanayake et al., 2011).
Empirical evidence from experimental studies also indicates that behavioral toxicity, particularly with sedating CNS effects in the acute phase of treatment, is of major concern with respect to driving performance (Ramaekers, 2003;Rapoport and Baniña, 2007;Berghaus et al., 2010;Strand et al., 2016;Brunnauer and Laux, 2017). However, cognitive dysfunction per se is also a core feature of many psychiatric disorders that is prominent in untreated patients and thus not necessarily linked to medication effects (Millan et al., 2012;Rock et al., 2014). Besides, there is evidence that long-term treatment with, for example, antidepressants or antipsychotics may help to improve psychomotor and cognitive dysfunction in patients with a psychiatric disease (Woodward et al., 2005;Keefe et al., 2007Keefe et al., , 2014Baune and Renger, 2014;Nielsen et al., 2015;Rosenblat et al., 2015;Prado et al., 2018).
Taken together stabilizing effects of pharmacological treatment have to be weighed against possible detrimental cognitive, vegetative-somatic, and psychomotor effects when validating drugs with respect to driving performance. Studies on this topic usually have investigated acute effects of medicines in healthy individuals that may be quite different from issues of long-term treatment of patients. Thus, the crucial question of which pharmacological treatment patients benefit most with respect to driving performance can only be validly answered by patient studies.
This systematic review gives an update of available evidence of patient studies on treatment effects on driving ability in psychiatric patients of the most frequently prescribed psychotropic drugs.

Methods
A systematic literature search in the PubMed database (January 1970 to December 2020) was performed using a combination of subject headings and keyword terms: "driving performance," "driving ability," "driving skills," "fitness to drive," "traffic safety," "on the road driving test," "driving simulation" AND "antidepressants," "antipsychotics," "benzodiazepines," "Z-drugs," "MAOinhibitors," and "mood-stabilizers." Furthermore, combinations of additional search terms were included to screen for further articles (search terms are listed in supplementary Materials). The titles and, if relevant, the abstracts of each citation were screened manually, and the full texts of relevant citations were then analyzed in detail. Besides, reference lists were manually searched for relevant studies. All identified articles were screened and, if necessary, discussed by 2 independent reviewers (Drs Brunnauer and Herpich).

Types of Outcome Measures
We integrated evidence of various research methodologies that focused on driving performance in psychiatric patients. Articles were selected if they (1) examined driving performance in an on-road test in real traffic or in a simulated on-road test, that is, on a closed circuit; (2) used a driving simulator with at least medium complexity, that is, a fixed hardware consisting of a steering wheel, brake pedals and simulated road environment; or (3) investigated psychomotor and visual perception functions to investigate driving skills according to legal regulations.
Articles in English or German language, in which study design, sample, and methods used had been described adequately, were considered for this review. The search results and study selection process are summarized in the following PRISMA flowchart (see Figure 1).

Study Selection
A total of 1417 hits were identified using the search terms mentioned above. The screening of titles, abstracts, and full-text articles (in a final step the analysis) of this literature search revealed 39 articles assessing driving ability with methodologies described above. Reference lists of the included studies and related reviews did not lead to an inclusion of further studies. Personal communication led to an inclusion of 1 further study. 22-120) with a total of 536 patients (37% females). The median age was 33 years (range 18-62 years). All investigations were designed as comparative clinical trials (cross-sectional and open label).
Antidepressants-The effects of antidepressants on driving performance were assessed in 14 clinical studies comprising 587 patients with primarily depressive disorders (44% female) with a median age of 45 years (range 18-78 years). Eight of these trials investigated driving performance within an open-label design, and in 6 trials patients were randomly assigned to treatment groups (4 under double-blind conditions).

Mood-Stabilizers-Three
publications were identified investigating driving performance in bipolar and unipolar patients and a group of patients with Meniere's disease. Sixty patients were investigated (48% female) with a median age of 45 years (range 18-63 years). Studies were designed as crosssectional or open-label trials.
Benzodiazepines and Z-Drugs-Eleven studies investigated the effects of benzodiazepines or Z-drugs on driving performance in patients with an anxiety disorder or patients with insomnia. The median sample size was 18 (range 12-44) with a total of 407 patients (22% females). The median age was 40 years (range 18-50 years) in the groups treated with tranquilizers and 55 years (range 25-75 years ) in studies investigating hypnotics. Most studies were designed as (randomized) double-blind cross-over trials. An overview of studies included is provided in Table 1.

Comparative Clinical Studies
Psychomotor functions according to legal regulations or driving simulator performance under long-term treatment with antipsychotic monotherapy were solely investigated within a naturalistic, non-randomized design, and patients were not investigated before and after treatment with respect to driving performance.
Under long-term treatment with antipsychotics and stabilized psychopathological conditions, schizophrenic and schizoaffective patients showed severe impairments in psychomotor functions related to driving skills in 31% (27%-42.5%) of cases. Second-generation antipsychotics (SGAs) seemed to be advantageous compared with first-generation antipsychotics (FGAs). Twenty-three percent (15%-27%) of patients treated with SGAs compared with 42% (34%-60%) of patients treated with FGAs showed severe impairments (Brunnauer et al., 2004(Brunnauer et al., , 2009(Brunnauer et al., , 2012. Differences could also be seen on driving simulator performance, with better results in patients treated with SGAs compared with FGAs in a risk simulation (Brunnauer et al., , 2009). Although in 1 study an advantage of clozapine on reactivity within SGAs could be shown (Brunnauer et al., 2004), altogether, no relevant differences with respect to driving performance between SGAs investigated was evident (Brunnauer et al., 2009;Brunnauer and Laux, 2012).

Treatment Combinations
While polydrug treatment is frequent in clinical practice, only a few studies could be identified that investigated driving performance under common psychopharmacologic treatment combinations in schizophrenic or schizoaffective patients. Irrespective of drug regimen, most patients do not seem to reach the level of psychomotor or driving simulator performance of healthy controls following subchronic or long-term polydrug treatment and psychopathologic stabilization (Hobi et al., 1981;Grübel-Mathyl, 1987;Wylie et al., 1993;Soyka et al., 2005;Fuermaier et al., 2019

Trizyclic Antidepressants (TCAs)/Amitriptyline
The acute and subchronic effects of low-dose nocturnally administered amitriptyline on driving performance have been investigated in chronic pain patients. Amitriptyline acutely impaired performance in an on-road driving test that diminished after subchronic treatment. It is important to note that patients' pain intensity ratings were unexpectedly low and moreover did not diminish under amitriptyline treatment; thus, there might have been a selection bias with respect to participants (Veldhuijzen et al., 2006).

Agomelatine
A novel approach to treat depression, focusing on circadian rhythms by melatonergic mechanisms, has been the development of agomelatine. After 4 weeks of treatment, patients significantly improved in both symptomatology and psychomotor skills relevant for driving, with most benefits occurring within the first 2 weeks of treatment. A functional level in psychomotor performance comparable with healthy controls could not be reached at the end of the 4-week study phase. In the on-road driving test in real traffic, patients were not significantly inferior to healthy controls and could in most cases be rated by a licensed driving instructor as sufficiently fit to drive after 4 weeks of treatment (Brunnauer et al., 2015).

Mirtazapine
Psychomotor and driving simulator performance in depressive patients during 14 days of treatment significantly improved under the noradrenergic and specific serotonergic antidepressant mirtazapine, indicating that partly remitted depressed patients show a better driving performance compared with untreated patients. Performance level in psychomotor functions of healthy controls was not reached within the subchronic or long-term treatment phase (Brunnauer et al., 2008). However, after 2 weeks of treatment, differences compared with healthy controls were not evident in driving simulator performance (Brunnauer et al., 2008;Shen et al., 2009).

Reboxetine
The effects of the selective noradrenergic reuptake inhibitor reboxetine on psychomotor function and driving simulator performance were investigated in a randomized comparative clinical study in patients with major depression. After 14 days of treatment, patients improved in driving skills, especially in tests measuring selective attention and reactivity. Furthermore, the frequency of accidents in the risk simulations with a driving simulator markedly decreased. Although patients were still inferior to healthy control subjects in psychomotor performance after 14 days of treatment, no differences could be shown in the driving-simulator tasks (Brunnauer et al., 2008).

Trazodone
Patients with primary insomnia treated with nocturnally administered trazodone performed slightly worse in a short-term Other ORT Long-term treatment (n. spec.) = impairment with hypnotics (treatment <3 y and >6 mo) Long-term treatment (n. spec.) = no impairment with hypnotics (treatment >3 y) Long-term treatment (n. spec.) = no impairment with anxiolytics Abbreviations: Aps, antipsychotics; BZDs, benzodiazepines; CPÄ, chlorpromazine-equivalents; DBC, double-blind crossover; DS, driving simulator; d, during the day (i.e., dosages were given at the day of driving assessment); FGAs, first-generation antipsychotics; n, nocturnal (i.e., dosages were given in the night before driving assessment); n. spec., not specified; ORT, on-road test; Other, cross-sectional or open-label study; PP, psychomotor performance; RCT, randomized clinical trial; RDB, randomized double blind; RDBC, randomized double-blind crossover; SGAs, second-generation antipsychotics; SORT, simulated on-road test. aOutcome: acute mean day 1-7; subchronic mean day 8-21; long-term mean >21 days. memory task, verbal learning, body sway, and arm muscle endurance. Negative effects on driving simulator performance were not found within 7 days of treatment. There was, however, a high dropout rate; 47 out of 63 individuals did not complete the entire study (Roth et al., 2011).

Venlafaxine
Four weeks of treatment with the serotonin norepinephrine reuptake inhibitor venlafaxine had positive effects on depressive symptoms and driving skills; most improvements were observable in the first 2 weeks. Although patients did not achieve healthy controls performance in psychomotor tasks within 4 weeks of treatment, significant differences could not be seen in an on-road test in real traffic compared with healthy participants' performance (Brunnauer et al., 2015).

Comparative Clinical Studies
The effects of long-term treatment with antidepressant monotherapy on psychomotor function in a clinical routine setting were investigated by Brunnauer et al. (2006). Sixteen percent of inpatients under pharmacologic steady-state conditions and prior to discharge showed severe impairments and thus had to be estimated as "unfit to drive." Controlling for confounding factors, a total of 28% of patients passed psychomotor performance tests without major impairments; 10% showed no major impairments in the TCA-group and 37.5% in the modern antidepressant group. Performance in on-road tests was assessed in 2 studies indicating that patients treated with selective serotonin reuptake inhibitors (SSRIs) or the serotonin norepinephrine reuptake inhibitor venlafaxine were inferior to healthy controls. Differences between treatment groups could not be shown, and driving impairment was significantly less in the treatment group compared with untreated depressive patients (Wingen et al., 2006;van der Sluiszen et al., 2017b). Grabe et al. (1998) did not identify differences between patients treated with TCAs, SSRIs, monoamine oxidase inhibitors, and common co-medications, whereas Brunnauer and Laux (2003) have shown less psychomotor performance impairment in patients treated with SSRIs and co-medication-approximately 37% were without major impairment-compared with treatment with TCAs (12%). No differences compared with healthy control patients could be shown in driving simulator performance (Hobi et al., 1981, Miyata et al., 2018 or in an on-road tests with patients treated long term (>3 years) with sedating antidepressants combined with TCAs, SGAs, or modern antidepressants (van der Sluiszen et al., 2020).

Treatment Combinations
The combined use of fluoxetine or moclobemide and benzodiazepines was differentially assessed in an on-road test by Ramaekers et al. (1997). In this study, 79% patients under fluoxetine and 73% under moclobemide additionally used benzodiazepines during treatment. Post-hoc analysis revealed that besides a linear decrease in driving performance throughout the treatment period in both groups, there was an apparent difference in patients taking competitive benzodiazepine co-medication metabolized by a P450 isozyme subject to inhibition by the antidepressant. Patients treated with fluoxetine decreased in driving performance until week 3 of treatment that diminished at week 6; under fluoxetine and competitive benzodiazepine treatment, a decline throughout the 6-week treatment period could be seen. Abbreviations: ---, no data available; (), inconsistent data; ↔, no impairment; ↓, impairment; ↑, improvement.

Comparative Clinical Studies
Seventeen percent of euthymic bipolar patients under long-term treatment with lamotrigine or lithium were severely impaired in psychomotor function relevant for driving, with a considerable advantage for patients treated with lamotrigine (Segmiller et al., 2013). Significant impairments compared with healthy controls could also be shown in driving simulator performance of bipolar and unipolar patients on stable lithium therapy (Hatcher et al., 1990;Jauhar et al., 1993).

Benzodiazepines and Z-Drugs
Tranquilizers Diazepam-Outpatients suffering from anxiety disorders showed a significant impairment in the first 3 weeks of treatment in an on-road driving test that diminished after week 3 (van Laar et al., 1992).
Medazepam-Subchronic treatment with medazepam in anxious patients showed that driving simulator performance was not adversely affected, while in a simulated on-road test, a minor increase of errors was discerned. As the author states, mean dosages were at a very low therapeutic level and thus general conclusions with respect to treatment effects should be drawn cautiously (Moore, 1977).

Hypnotics
Flunitrazepam-Patients with insomnia experienced a slight impairment of driving performance in an on-road test when treated for 7 days with flunitrazepam (Schmidt et al., 1986). On the contrary, a single dose administered on retiring to bed to patients with insomnia did not significantly affect driving performance the next morning (Vermeeren et al., 1995).
Flurazepam-Flurazepam had acute effects on driving performance in patients with insomnia. The ability to control the lateral position of the vehicle was significantly impaired, and the degree of impairment was more pronounced in females. Moreover, impairment was greater in the morning (10 to 11 hours) than in the afternoon (16 to 17 hours) after administration (Brookhuis et al., 1990).
Lormetazepam-Acute treatment with lormetazepam 1 mg did not impair driving performance in an on-road test of patients with insomnia when tested 10 to 11 hours and 15 to 16 hours post administration. However, 7 days of bedtime administration of 1 mg lormetazepam in primary insomnia patients induced significant impairment in driving simulator performance (Staner et al., 2005). Acute treatment with 2 mg impaired driving performance of insomniacs 10 to 11 hours post ingestion, but no residual effects after 15 to 16 hours were evident (Brookhuis et al., 1990).
Temazepam-There were no significant residual effects on performance in a driving simulation test 5.5 hours after intake of temazepam the next morning (Partinen et al., 2003), and no decrease of driving performance during an on-road test was found in patients with sleep disorder that received temazepam once a night for 7 days (Schmidt et al., 1986).
Zolpidem-Single-dose administration as well as repeated (7 days) treatment with zolpidem given at nighttime did not impair on-road or driving simulator performance of insomnia patients (Vermeeren et al. 1995;Partinen et al., 2003;Staner et al., 2005).
Zopiclone-A single, oral overnight dose of zopiclone significantly impaired on-road driving performance in insomnia patients who chronically used hypnotics as well as in patients who did not or only infrequently used hypnotics. However, the magnitude of impairment was less in the chronic users group (Leufkens et al., 2014a). Analogous results were obtained by Staner et al. (2005), who showed that single bedtime administration of zopiclone induced next-day driving impairments on a driving simulator.

Comparative Clinical Studies
Actual driving performance and driving-related skills of elderly chronic hypnotic users were investigated by Leufkens et al. (2014b). Patients were divided in frequent users (using hypnotics ≥4 nights/wk) and infrequent users (using hypnotics ≤3 nights/ wk). Driving, measured by a standardized highway driving and a car-following test, was not impaired in patients with insomnia irrespective of the frequency of use of hypnotics. Van der Sluiszen et al. (2019) suggested that impairment of driving performance was dependent on treatment duration in elderly patients. While the ability to drive was significantly impaired in patients treated with hypnotics and comedication for less than 3 years, no deterioration was found in patients treated for more than 3 years. In long-term users (>6 months) of anxiolytics and comedication, no clinically relevant differences in driving performance were evident compared with healthy controls. Table 2 gives an overview of controlled experimental patient studies on antipsychotics, antidepressants, mood stabilizers, benzodiazepines, Z-drugs, and driving performance.

Summary of Findings
Only 40 studies could be found according to our selection criteria, indicating a clear lack of patient studies on driving performance under psychopharmacologic treatment. No data were available that give information about causal relationships between antipsychotics, monoamine oxidase inhibitors, mood stabilizers, and driving performance.
Antipsychotics-Overall, with regard to antipsychotic medication, a great proportion of schizophrenic and schizoaffective patients does not reach the level of psychomotor or driving simulator performance of healthy controls following long-term treatment despite significant clinical improvement (Wylie et al., 1993;Soyka et al., 2005;Fuermaier et al., 2019). Keeping in mind the great heterogeneity between studies, on average 31% (27%-42.5%) of in-patients showed severe impairment of psychomotor function relevant for driving under steady-state pharmacological conditions prior to hospital discharge; that is, patients were treated in most cases 6 weeks and more in a stationary setting (Brunnauer et al., 2004(Brunnauer et al., , 2009(Brunnauer et al., , 2012Soyka et al., 2005). Considering that approximately 32% of young, unmedicated schizophrenic patients showed severe impairment of psychomotor skills relevant for driving (Segmiller et al., 2017), it can be assumed that impairments under steady-state pharmacologic conditions may be primarily due to the illness itself rather than to treatment effects.
Under common treatment combinations, a greater proportion of patients (32%-54%) showed severe impairment in functional domains relevant for driving (Grabe et al., 1999;Kagerer et al., 2003). As a polypharmacological approach is often administered in difficult-to-treat patients with poor treatment response, a selection bias towards a more severe clinical course might be discussed.
Summary-On average, one-third of schizophrenic or schizoaffective patients under treatment with antipsychotics showed severe impairment in skills relevant for driving compared with healthy controls, which may be primarily due to residual symptoms of the illness itself rather than to negative side effects of antipsychotic treatment. Studies point to an advantage of SGAs compared with FGAs with respect to driving skills, and there seems to be a disadvantage under antipsychotic polypharmacy. However, conclusions must be drawn cautiously with regard to the marked methodologic heterogeneity of the studies. Moreover, interpretation of data is also limited since adverse effects of pharmacological treatment and illnessassociated impairment cannot be disentangled within studies under review.

Antidepressants and Mood-Stabilizers
Approximately 16% of patients with major depression and 17% of bipolar patients showed severe impairment in psychomotor skills relevant for driving under long-term monotherapy with antidepressants or mood-stabilizers (Brunnauer et al., 2006;Segmiller et al., 2013). Data point to an advantage of treatment with modern antidepressants and SSRIs in particular over TCAs with regard to driving skills (Brunnauer and Laux, 2003;Brunnauer et al., 2006). Although Ramaekers et al. (1997) showed that specific combinations of antidepressants with benzodiazepines may have deleterious effects on driving performance, there is also evidence that under treatment combinations with lithium, carbamazepine, neuroleptics, and hypnotics-when given on clinical considerations-a similar distribution of patients showed severe impairment (18%-20%) compared with patients under antidepressive monotherapy (Grabe et al., 1998;Brunnauer and Laux, 2003).
Acute treatment with amitriptyline, but not subchronic treatment, had negative effects on the on-road performance of insomnia patients (Veldhuijzen et al., 2006). Under trazodone, no impairment in driving simulator performance could be seen in the acute treatment phase (Roth et al., 2011).
No patient data were available that give information about causal relationships of SSRI monotherapy on driving ability in patients.
Under long-term treatment with lithium, unipolar and bipolar patients showed a significantly lower performance than matched healthy volunteers in driving simulator performance, and there seems to be an advantage when patients were treated with lamotrigine compared with lithium (Hatcher et al., 1990;Jauhar et al., 1993;Segmiller et al., 2013).
Summary-Patients definitively profit within 2-4 weeks from treatment with antidepressants with respect to driving performance, and data point to an advantage when treated with modern antidepressants over TCAs and with lamotrigine over lithium. Although level of psychomotor performance of healthy controls was in at least a subgroup of (partly) remitted patients not reached, differences in driving simulator performance or on-road tests were less pronounced or even did not exist and thus may be an indication for compensational competencies with respect to driving performance in a "real world context." Not least, there is also evidence that the effects of sedating antidepressants on driving performance attenuate over time. Co-administration of other psychotropic medicines with antidepressants, when given on clinical considerations, does not seem inferior compared with trials on monotherapy.

Tranquilizers
Elderly patients treated long term (>6 months) under clinical conditions, that is, most were comedicated with antidepressants or antipsychotics, did not differ in driving performance compared with age-matched healthy controls (van der Sluiszen et al., 2019). Results of controlled experimental studies suggest that subchronic or long-term treatment with low-dose medazepam did not impair on-road driving performance of anxious patients (Moore, 1977). However, deleterious effects on driving performance in an on-road test could be shown the first 3 weeks of treatment with diazepam (van Laar et al., 1992).

Hypnotics
Most studies investigated effects of hypnotics either in a singledose regimen or after 7 days of treatment in patients with insomnia. Temazepam and zolpidem seem to be free of acute negative effects on driving performance when given to patients with a sleep disorder (Vermeeren et al., 1995;Partinen et al., 2003;Staner et al., 2005). Zopiclone, flurazepam, and dosedependent lormetazepam cause significant driving impairment in the acute treatment phase of patients with insomnia (Brookhuis et al., 1990;Staner et al., 2005;Leufkens et al., 2014a). When given as a single nocturnal dose, flunitrazepam seems to be free of deleterious effects on driving performance the next morning (Vermeeren et al., 1995); this does not appear to be true, however, when treated for 7 consecutive days (Schmidt et al., 1986).
There is some evidence that frequency of hypnotic intake does not play a major role in driving performance in elderly patients (Leufkens et al., 2014b) and that in chronic elderly users (>3 years), no driving impairments seem to become evident (van der Sluiszen et al., 2019). This is in line with studies indicating development of tolerance over time in long-term users (Pomara et al., 2015;van der Sluiszen et al., 2017a). However, there is also evidence that tolerance with respect to skills relevant for driving may be only partial (van der Sluiszen et al., 2017a).
Summary-Diazepam significantly worsened driving the first 3 weeks of treatment, whereas low-dose medazepam did not impair driving. Among hypnotics, flunitrazepam, flurazepam, lormetazepam (dose dependent), and zopiclone significantly impaired driving performance at treatment initiation, whereas temazepam and zolpidem were free of deleterious effects in patients with insomnia. There is evidence in long-term users that impairing effects on driving attenuate over time, which may be only partial.

Limitations
Major limitations of the current analysis are the small database and methodological constraints-60% of studies were cross-sectional or open-label trials-that do not allow to disentangle treatment effects from illness-associated impairment. Included studies show large variations with study designs, patient characteristics, and methodology used to measure driving performance. Besides, poor description of patient samples, with respect to clinical data (e.g., psychopathologic symptoms at time of testing, duration of illness, or driving characteristics) are missing in most cases. Thus, because of heterogeneity of studies, the summarized evidence was not suitable for a meta-analytic approach. It should also be noted that the literature review was restricted to PubMed database. Albeit screening of reference lists of included studies and related reviews did not lead to an inclusion of further studies in our review, it could be argued that the utilization of other databases might have provided additional information. Considering all these shortcomings, conclusions of this review should be drawn cautiously.

Conclusion
Much more controlled patient studies are needed to analyze the complex relationship between illness and medication to clarify a core issue: from which pharmacologic treatment do patients benefit, also with respect to driving performance. The current synopsis gives evidence that psychopharmacologic medicines under review improve or at least stabilize driving performance of patients when treated long-term under clinical considerations. To enhance treatment compliance and give information to professionals, patient information leaflets as well as classification systems of potential driving impairing medicines should also incorporate information about the stabilizing effects of long-term treatment on driving performance and recommendations on the passage of time that should be considered to minimize possible detrimental effects.
Due to the fact that particular drugs can place patients at risk-at least after treatment initiation-during everyday activities like driving a car, it is of crucial importance to ensure that physicians are aware of this and that they try to minimize potential disruptive effects on neuropsychological processes. Prescribers of drugs should be recommended to minimize the number of prescribed drugs, keeping in mind potential pharmacokinetic interactions, to adjust dosage regimens according to a patient's individual response, and to administer medicines in nocturnal doses whenever possiblefor example, in the case of sedating antidepressants-to mitigate adverse side effects on driving performance. Not least, patients should be educated on the potential risks of psychopharmacological treatments with respect to driving to attain a better awareness and self-monitoring of possible adverse drug effects.

Supplementary Materials
Supplementary data are available at International Journal of Neuropsychopharmacology (IJNPPY) online.