Table 3 Summary of key findings and quality appraisal of the 16 included papers, presented by sub-group

Brunnauer (2016) [36] Germany; N

To explore the effect of mental illness on driving status in terms of driving licence possession and driving restrictions

N = 1859 recruited from five hospitals (N = 1546, psychiatric inpatients; N = 313, neurological inpatients as control group)

Quantitative:

Questionnaire

• Lower proportions of current drivers were found mainly in those with organic mental disorder and psychotic disorder.

• 67% of psychiatric patients reported having a valid licence and 77% of this group reported using their cars regularly.

• More than 30% of patients with psychotic and organic mental health conditions do not have driving licence despite suitable age and employment status.

• The best predictors for having driving licence were education and being partnered or in a relationship, which both reflect psychosocial status.

• Female, older age, pension holders, patients with organic mental disease or schizophrenia are associated with increased rates of driving cessation.

CEBM: level 4

AAN: Class IV

Bulmash (2006) [8] Canada; N

To examine the correlation between major depressive disorder (MDD) and driving performance using driving simulator

N = 47 (N = 18, outpatients diagnosed with MDD, required to be free of antidepressants); N = 29, control group)

Quantitative:

Naturalistic group comparison

Measures:

• Epworth Sleepiness Scale (ESS)

• Beck Depression Inventory (DBI)

• York driving simulator

• The depressed group experienced higher levels of general, on task sleepiness.

• Level of sleepiness was statistically associated with driving variables of speed and speed deviation.

• The depressed group showed a significantly slower steering reaction time (RT) and greater number of car crashes when compared to controls, which were characterized by a medium effect size.

• Slowed RT or increased crash risk were not significantly associated with severity of depressive symptoms.

• No statistically significant differences for road position, speed or speed deviation were found when compared to controls.

CEBM: Level 4

AAN: Class IV

De las Cuevas (2008) [37] Spain; N

To examine fitness-to-drive of people with mental health illness

N = 208 with psychiatric illnesses

Quantitative:

Naturalistic group comparison

• LNDETER 100 battery (cognitive functioning and psychomotor performance)

• Clinical Global Impressions- Severity of Illness Scale

• Only 33 out of 208 participants were compliant with the requirements of a driver’s licence, and 84% failed at least one of the tests.

• Driving patients showed better results than non-driving patients; however, 79.5% still scored too low to obtain or renew their driver’s licence

• Only 2 out of 10 professional drivers passed all the tests.

• None of the driving patients informed the authority of their mental health status and none stopped driving, although 10% recognized their inability to drive.

• In the 6 months prior to the research, only 3 of the 208 patients had been involved in a crash indicating a poor correlation between neuropsychological tests when predicting driving performance

CEBM: level 4

AAN: Class IV

McNamara (2015) [38]; Ireland; N

To explore the experience of driving of people diagnosed with bipolar disorder

N = 18, diagnosed with bipolar disorder; holding driving licence; recruited via selective sampling

Qualitative:

Thematic analysis of focus group data

• Occurrences of speeding, making poor decisions, losing feeling of control, decreased concentration and judgement, and impulsivity were identified by some participants when driving in manic or hypomanic states. They agreed “no driving” when unwell.

• Some participants reported bipolar was unlikely to impede their driving capabilities, compared to the general population.

• Every individual reported a unique experience on the effect of bipolar disorder on driving.

• Participants reported having an open communication with health professional is useful regarding fitness-to-drive.

• The fitness-to-drive guidelines were perceived as discriminatory by drivers with bipolar.

CEBM: level 5

AAN: Class IV

Niveau (2001) [39]; Switzerland; N

To examine if psychiatric patients that were reported to authorities have a higher risk than those were not reported to the police

N = 65 with psychiatric disorders (subjects: N = 31, reported to authorities / Control: N = 34) via purposive sampling

Quantitative:

Case control design

Measures:

• Reason for licence being revoked

• Violations

• Crashes

• Severity of psychiatric disease

• More males and those with lower level of schooling were reported to authorities, than the participants in the control group.

• A statistically higher rate of acute disorders was found among the reported cases than the controls. The reported subjects were found to show severe psychiatric antecedents; however, nearly half of them had a clean driving record.

• Less than 10% of clean driving records were found in the control group.

• Doctors basing their opinion about fitness-to-drive on their own professional criteria would underestimate other factors which could lead to bias in evaluating driving performance.

CEBM: Level 4

AAN: Class III

Segmiller (2015) [40]; Germany; N

To examine psychomotor driving skills of unmedicated first- and recurrent- episode people with schizophrenia, and skill decline with disease progression

N = 46 (N = 13 first episode and N = 13 recurrent episode schizophrenia, and N = 20 healthy controls)

Quantitative:

Naturalistic group comparison

Measures:

• Positive and Negative Syndrome Scale

• Computerised Weiner Test System (WTS) (visual perception, reactivity and stress tolerance, concentration and vigilance)

• 32% of untreated participants showed severe impairment in ability to complete the tests, which was interpreted as severe impairment in the ability to drive.

• A greater percentage of first episode patients (38%), compared to recurrent episode patients (25%), showed pronounced impairments on the WTS.

• The problems in psychomotor function seen in both groups could not be attributed to adverse effects of pharmacological treatments.

• Both groups of patients with schizophrenia had lower results on the WTS than healthy controls.

• Analyses did not reveal a chronological decline in psychomotor function over time (in the early stages of the disease).

CEBM: level 4

AAN: Class III

Zingg (2009) [41]; Switzerland; N

To investigate how performance of driving offenders with a psychiatric diagnosis relevant for driving, driving offenders without a psychiatric diagnosis, and control subjects differed on neuropsychological tests and self-performance on driving

N = 219 patients (N = 63 with diagnosis relevant to driving; N = 111 with diagnosis not believed to affect driving; N = 45 controls

Quantitative:

Naturalistic group comparison

Measures:

• Performance test system or Wechsler Intelligence Test

• Number connection test

• Modified card sorting test (computerized)

• Battery for attentional performance

• Participants who were offenders (with and without psychiatric diagnosis) scored significantly lower than the control group on tests of information processing, but only those with a psychiatric disorder also scored significantly lower on problem solving and cognitive flexibility, alertness, and visual scanning/reaction time.

• There were no significant differences between the groups on their self-assessment.

CEBM: level 4

AAN: Class IV

Langan (2009) [42]; Ireland; N

To investigate the level of documentation relating to fitness-to-drive in psychiatrists’ clinical notes

N = 44 outpatients record, N = 48 discharge records; diagnosed with psychiatric illnesses

N = 11 psychiatrists recruited from acute psychiatric unit

Quantitative:

• Secondary data analysis of medical records

• Questionnaire

• Minimal documentation on possible effect of illness on driving performance was found in outpatient records and discharged records, 0% and 2% respectively.

• More than half of both outpatient (57%) and discharged records (54%) contained documented advice on the importance of compliance with prescribed medication.

• 34% of outpatient records and 44% of discharged records contained documentation on the side effects of prescribed medication on driving ability.

• No advice regarding medication usage and operation of machinery or driving was found in both sets of records.

• Over 50% of psychiatrists indicated they advised patients on the adverse effect of illness or medication on driving performance but few documented this.

CEBM: not applicable

AAN: Not applicable

Menard (2006) [43]; Canada; Y

To examine Canadian psychiatrists’ attitudes, practices and knowledge on fitness-to-drive of people with mental health illness

N = 248 psychiatrists recruited via random sampling

Quantitative:

National cross sectional survey

• 64.1% of respondents strongly agreed or agreed on the importance of addressing fitness-to-drive.

• Only 18% of respondents recognised if their patients were active drivers.

• One-fourth of respondents strongly agreed or agreed they were capable in assessing fitness-to-drive.

• One-half of respondents believed that people with mental illness have a higher risk of a car collision.

• Psychiatrists were more likely to advise on the adverse effect of medication on driving, rather than the potential impact of the mental illness itself on driving.

CEBM: level 5

AAN: Class IV

Menard (2012) [44]; Canada; N

To understand the role of occupational therapists in determining fitness-to-drive of people with mental health illness, with a focus on screening, assessment, intervention, perceived barriers and facilitators to undertaking this role

N = 20 occupational therapists working in adult mental health recruited via purposive sampling

Qualitative:

Phenomenological analysis of focus group data

• An important role for occupational therapists was perceived, with regards to assessing driving ability and raising issues of driving with healthcare team and client.

• Limited evidence was found to exist to support practice.

• Some participants suggested discussing driving history and habit with their clients; understanding client’s perception of driving abilities and factors impacting on their driving capabilities; and identifying goals related to driving can ensure driving safety.

CEBM: Level 5

AAN: Class IV

Rouleau (2010) [45]; Canada; N

To explore factors that health professionals and clients considered in relation to their fitness-to-drive and recommendations psychiatrists provided after the assessment

N = 72 (N = 48, drivers, N = 24, non-drivers), with a mental health diagnosis recruited via purposive sampling

Quantitative:

Survey

• 27.7% of participants indicated illness and medications affect driving performance due to lower concentration (43.9%) and fatigue (29.4%).

• 28% of participants indicated illness was related to nervousness, 14.2% reported feeling unstable, 9.5% reported being more careful on the road, and 14.2% reported being more aggressive.

• Participants indicated that side effects of medication can cause dizziness (10%), nervousness (5%) and other general negative effect (15%).

• Psychiatrists focused on the impact of mood on driving and overall clinical assessment to determine a person’s fitness-to-drive, while occupational therapists focused more on physical status, impulsivity levels, and driving history. Both professionals also considered medication side effects, psychiatric symptoms, cognition and judgement when assessing fitness-to-drive

• Occupational therapists administered pen and paper test when assessing driving ability with people with mental health issues; however, the finding was not significant in predicting road performance due to small numbers.

CEBM: Level 5

AAN: Class IV

Vrkljan (2015) [46]; Canada; Y

To identify assessment tools and procedures used by assessors to determine fitness-to-drive, outcomes and recommendations using four case scenarios inclusive of one scenario of a 33 year old woman with schizophrenia (case#4)

N = 46 assessors in Canada

Quantitative: National Questionnaire

• 33 assessors (71.7%) would assess driving ability for a client with schizophrenia.

• Cognition was identified as a primary issue by 71% of the respondents for a client with schizophrenia.

• 17.6% of assessors stated that medication needed to be taken into consideration.

• 87.1% of assessors indicated using at least one observational test, followed by a physical (80.6%), perceptual motor (77.4%), or cognitive test (67.7%).

• 81.3% of respondents would conduct an on-road assessment.

• More than half of the respondents would proceed with licence reinstatement with ongoing monitoring if the client did well on both the off and on-road assessment.

• If the client passed the on-road assessment but failed the off road assessment, 19 respondents would recommend return to driving, with 8 would consider reassessment in 6 months.

CEBM: Level 5

AAN: Class IV

Crancer (1969) [47]; USA; N

To examine the possible link between particular mental health illnesses (psychotic disorders, personality disorders, and psychoneurotic disorders) and driving performance

N = 271 psychiatric patients with valid driver licence (N = 97 with psychotic disorders, N = 79 with personality disorders; N = 95 with psychoneurotic disorders) compared to a N = 687,228 control group

Quantitative:

Epidemiological analysis of secondary data

Measures:

• Crashes

• Violations

• Statistically higher crash rates were identified in both the personality (114% higher) and psychoneurotic groups (49% higher), than the control group.

• Crash rates in the schizophrenic group were similar to the control group.

• Each of the mental illness groups showed a statistically higher violation rates than the control group, particularly in reckless driving, negligent driving and driving with defective equipment.

CEBM: level 4

AAN: Class III

Eelkema (1970) [48]; USA; N

To examine the correlation between mental illness and crashes

N = 238, driving records from a hospital (1960); with matched comparison group (n = 290)

Quantitative:

Epidemiological analysis of secondary data

Measures:

• Crashes

• Violations pre and post hospitalisation

• Patients discharged from the hospital showed higher crash and violation rates per year.

• Greater crash ratio (>1) was found in people with psychosis and psychoneurosis; however, these groups had better records compared to the matched comparison group post discharge (crash ratio < 1).

• The highest crash rate was identified in personality disorders, with minimal improvement post discharge. With treatment, crash rates decreased in all categories except for personality disorders.

• Males with psychosis and psychoneurosis, and personality disorders showed a rising trend in violations.

CEBM: level 5

AAN: Class IV

 Kastrup (1977) [49]; Denmark; N

 Kastrup (1978) [50]; Denmark; N

Both studies used the same data set:

Part 1 (1977):

to examine characteristics of psychiatric patients that were involved in crashes

Part 2 (1978):

To investigate crashes psychiatric patients were involved in and its casualty severity

N = 2076 psychiatric patients involved in crashes, compared to N = 40,232 involved in all crashes

Quantitative:

Epidemiological analysis of secondary data

Measures for both:

Crashes

Part 1 (1977):

• As a group, people with psychiatric disorders accounted for 11% of the crashes in Denmark, and this included pedestrian and vehicle crashes.

• Approximately 6% of people who were killed or injured in 1973 in car or pedestrian crash had a psychiatric diagnosis.

Part 2 (1978):

• Drivers with a psychiatric disorder, who were involved in crashes, including pedestrian and vehicles:

- were over-represented by women; and

- were more frequently in the age group of 25–54 years.

• When compared to the normal population, greater proportions drove stolen vehicles without a valid licence and without seat belt at the time of crash, and were more frequently intoxicated.

CEBM: level 5

AAN: Class IV