Mobility behaviour and driving status of patients with mental disorders - an exploratory study.

BACKGROUND
Driving is an important activity of daily life and an integral part of mobility. However, impact of mental illness on road mobility is widely unexplored.


METHOD
Driving status in 1497 psychiatric inpatients (PPs) and a clinical control group of 313 neurological inpatients (NPs) was investigated using a brief questionnaire.


RESULTS
67% of PPs (89% NPs) reported to have a valid driver's licence and 77% of them (92% NPs) reported to regularly use their cars. Within driver's license holders, patients with organic mental disorder (32%), substance dependence (37%) and psychotic disorder (40%) had the lowest proportion of current drivers. Higher educational qualification (odds ratio [OR] from 2.978 to 17.036) and being married/partnered (OR 3.049) or divorced (OR 4.840) significantly advanced the probability of possession of a driving license. Predictive factors for driving cessation were being female, an older age, drawing a pension and having an organic mental disease or schizophrenic disorder.


CONCLUSION
Mental disease has a negative impact on driving status and this is especially true for illnesses frequently being accompanied by distinct cognitive impairments. Factors predicting road mobility elucidate the strong relationship with psychosocial status indicating that recovery of driving competence should be an integral goal of treatment strategies.


Introduction
For most people driving is an important activity in daily life affecting physical, social and economic well-being. Driving mobility is also an important part of one's selfidentity that may influence health status (Edwards et al. 2010). Driver licensing has changed, indicating that especially in older cohorts an increased number of driving licenses are registered (Siren and Haustein 2013). Thus, medical factors may become increasingly important with respect to mobility behaviour and traffic safety.
The majority of research on mental health and driving has focussed on risk factors of driving safety. Albeit there is very little evidence from epidemiology, confirming that psychiatric illness results in dangerous driving, it seems that psychiatric patients have a higher than expected rate of involvement in road traffic accidents (Vaa 2003, Dobbs 2005, Marshall 2008). For example, schizophrenic patients seem to participate less frequently in road traffic actively, compared with healthy persons but, at the same time, they are at increased risk of being involved in a traffic accident (Edlund et al. 1989). With respect to psychomotor functioning, most patients with a schizophrenic or depressive disorder, under steady-state pharmacological conditions and prior discharge to outpatient treatment, show mild to moderate impairments with respect to driving skills (Brunnauer et al. 2004(Brunnauer et al. , 2006. A Germany-wide representative survey on mobility in 2008 revealed that about 88% of persons surveyed reported to have a valid driver's license and that most driving is for private reasons (INFAS and DLR 2010). Only few data are available concerning the effects of mental disorder on road mobility. A study on functional status in older outpatients with schizophrenia found that 52% of 83 middle-aged and elderly outpatients suffering from schizophrenia possessed a valid driving license, 43% of them were current drivers and driving status correlated with the degree of independence in their living situations (Palmer et al. 2002), Parmentier et al. (2005) reported a minimal decrease in road mobility among men due to mental disorders such as anxiety and stress, sleep disorders and depression. In people aged 65 years and older, predementia, Parkinson's disease, older age, and a high number of kilometres previously driven were common restriction factors in women and men (Marie Dit Asse et al. 2014). In patients with very mild to mild dementia it has been reported that 54%-78% of men ceased driving (Foley et al. 2000). Otherwise, driving cessation in older adults was associated with increased depressive symptoms (Ragland et al. 2005), decreased social and physical functioning (Edwards et al. 2009, Choi et al. 2012, and a decline relative to active drivers in cognitive functioning (Choi et al. 2014). Not least, it could be demonstrated that in patients suffering from organic mental disorder, training and pre-planning on giving up driving mitigated the negative consequences of post driving cessation (Musselwhite and Shergold 2013). Thus, mobility may be affected by medical factors but also has an influence on health status of patients.
The purpose of this exploratory study was to gain basic information about driving status of psychiatric inpatients and compare it with those of a clinical control group of neurological inpatients. Besides this, we were interested in the contribution of variables collected to the prediction of (a) possession of a driving license and (b) driving restriction in our clinical sample.

Materials and methods
In a multicentre study we examined driving status and road mobility behaviour in 1546 psychiatric inpatients (PPs) using a brief questionnaire that could be easily applied within clinical routine practice. All patients were at least partly remitted and prior discharge to outpatient treatment. Most patients were prescribed medications at the time of interview (PPs 88.0%; NPs 74%) and there was substantial variability in type and dosage of medications based on patient's needs. Patients with considerable comorbidity were excluded from the study.
Participating study centres were 5 hospitals for psychiatry, psychotherapy, psychosomatic medicine and/or neurology: kbo-Inn-Salzach-Klinikum Wasserburg/Inn, Psychiatric Hospital, Bezirkskrankenhaus Augsburg, Department of Psychiatry and Psychotherapy at the Ludwig-Maximilians University Munich, Pfalzklinikum Klingenmü nster, and Mediclin Klinik an der Lindenhö he Offenburg. A group of neurological inpatients (NPs, n ¼ 313) mainly suffering from peripheral neurological disorders (e.g. degenerative discopathy) was recruited from kbo-Inn-Salzach-Klinikum Wasserburg/Inn, Department of Neurology. Study procedures were identical at each site.
The short questionnaire comprised 10 questions with either a checklist of possible answers, or a yes/no choice (see Appendix A to be found online at http:// informahealthcare.com/doi/abs/10.3109/13651501.2015. 1089293). Basic demographic questions (age, gender, civil status, education and employment) and questions on driving status and mobility behaviour (possession of a driving licence, driving cessation, kilometres driven per year, frequency of utilisation and predominant purpose of the drive) were obtained from patients self reports. Clinical data (diagnoses according to ICD-10 classification and duration of illness) were surveyed by the treating psychiatrist/neurologist. Before starting the multicentre study the short questionnaire has been evaluated in a pilot study in the Department of Neuropsychology of the kbo Inn-Salzach-Klinikum.
The study was approved by the Ethics Committee of the Medical Faculty of the Ludwig-Maximilians-University Munich and was conducted in accordance with the Declaration of Helsinki. All participants gave written informed consent prior to inclusion in the study.

Statistical Analysis
Statistical analysis was performed using SPSS software (Statistical Package for Social Sciences, Version 11.5, SPSS Inc., Chicago III, 2002). Analyses were conducted in three steps. First, descriptive statistics were initially used to obtain means and frequencies of the variables collected. Data was then analysed with non-parametric tests (Chi-square and Mann-Whitney U test) or independentsamples t-tests with an alpha level of 0.05. In a third step we investigated the associations between driving status (driving license/no driving license and driving cessation/ no driving cessation) as the dependent variables, and sociodemographic (age, gender, education and employment status) and medical factors (ICD 10 F-diagnosis and duration of disease) as the independent variables. Odds ratios (ORs) from the binary logistic regression analysis were examined to determine the associations between the variables in the model and the likelihood of having a driving license and driving restriction controlling for the other variables.

Sociodemographic and clinical profile
A total of 1859 participants (1546 PPs, 313 NPs) answered the brief questionnaire; the mean response rate in the study centres was 87.8% (range: 79.5% -92.0%). Data from 19 patients with unclear diagnoses and 30 patients with intellectual disabilities (ICD10: F 70) were excluded from statistical analysis.

Driving status
Psychiatric patients and neurological patients significantly differed with respect to possession of a driving license (PPs: 67% vs. NPs: 89%, p50.01). Despite having a current driving license, 21% of the PPs refrain from driving compared with only 8% of the NPs (p50.01).
Among PPs, mainly those with organic mental disorder, substance dependence and psychotic disorder had the lowest proportions of current drivers. Figure 1 illustrates the proportion of patients with a driving licence/no driving license and the amount of patients with a driving license who regularly use their cars.
Concerning kilometres driven per year, most patients were infrequent drivers (510.000 km/year) with a significantly higher percentage in the PPs group (p50.01); significant differences in frequency of car utilisation between PPs and NPs could not be observed. Patients reported that they primarily use their car for private (52%) and less for business purposes (37%) or doctor visits (11%).

Logistic regression -possession of a driving license
The full model was statistically significant, 2 (df ¼ 17, n ¼ 1810) ¼ 209.84, p50.001, indicating that it was able to distinguish between patients having or not having a driving license. The model explained between 18.1% Table 1. Demographic and clinical characteristics and driving status of PPs (n ¼ 1497) and NPs (n ¼ 313). (Cox and Snell R2) and 25.7% (Nagelkerke's R2) of the variance and correctly classified 76.2% of cases. As shown in Table 2, four of the independent variables (civil status, education, diagnoses and duration of illness) made statistically significant contributions to the model. Patients being married/partnered or divorced were about 3-4 times more likely to report having a driving license compared with singles. Also patients having a graduation were 3-17 times more likely to report to have a driving license compared with patients having no graduation. Patients with substance dependence, psychotic disorder and longer duration of illness were less likely to have a driving license compared with NPs. Age, gender and employment status made no statistically significant contributions to the model.

Logistic regression -driving cessation
To analyse factors contributing to driving cessation, patients having a valid driver's license were separately examined. The full model containing all 16 variables was statistically significant, 2 (df ¼ 17; n ¼ 1249) ¼ 96.577, p5 0.001, and explained between 12.9% (Cox and Snell R2) and 21.3% (Nagelkerke's R2) of the variance. 82.6% of cases could be correctly classified.
Five of the 16 independent variables (age, gender, civil status, employment and diagnoses) made statistically significant contributions to the model (Table 3). Older age, being female, drawing a pension (compared with being unemployed), and having an organic mental disease or schizophrenic disorder (compared with NPs) was associated with a higher probability of driving cessation. Patients being married/partnered were less likely to restrict driving compared with singles.

Discussion
From our data we can conclude that driving a car is a relevant issue for patients with a mental disorder. About 67% of patients with a mental illness have a driving license and 77% of these patients report to drive regularly with their cars. The percentage of PPs possessing a driving licence is however significantly smaller when compared with our clinical control group of neurological patients, suggesting that psychiatric disease in particular has a negative impact on driving status. They also differ with respect to car utilisation from the clinical control group. PPs more often quit driving despite having a driving license, and while not differing with respect to frequency of car utilisation per week, they drive less kilometres annually. Since they use their cars primarily for non-business purposes, it can be assumed that driving restrictions largely affect the social functioning of PPs.
Patients with organic mental disorder, substance dependence and psychotic disorder had the lowest proportion of current drivers. Within variables studied, education and civil status were the best predictors for having a driving license, indicating the close relationship with psychosocial status. Predictive factors for driving cessation were being female, older age, drawing a pension and having an organic mental disease or schizophrenic disorder.
Before discussing our results, limitations of this investigation should be outlined. The purpose of the present exploratory study was to gather basic information about the driving status of psychiatric patients, and to formulate hypothesis for more definite studies on mobility behaviour. From our data we can only draw tentative inferences on reasons for not having a driving license or not driving despite having a driving license. In future studies this should be analysed systematically within a more comprehensive survey focussing in more detail on possible causes, e.g. withdrawn because of traffic violations, medical and/or economic reasons. Also issues of vocational/employment status and income on driving status are to be addressed more specifically. Furthermore, it would be interesting how patients compensate for not driving, as well as the impact of not driving on psychosocial and vocational activities. However, this would have gone beyond the scope of our exploratory investigation. Finally, since most of the data with respect to driving behaviour were obtained from   patients' self-reports potential biases cannot be ruled out and thus relatives or caregivers should also be interviewed to validate statements. Nevertheless, this is to the best of our knowledge the first multicentre study that surveyed data on automobile driving from a comprehensive sample of psychiatric inpatients. Driving while impaired by drugs and/or alcohol is prevalent in populations associated with substance dependence and drivers consuming illicit drugs and psychoactive medications have an elevated risk of being involved in a traffic accident (Vaa 2003;Kelly et al. 2004). Surprisingly, among those with a valid driving license only 22% actually refrained from current driving, whereas 78% reported that they drove in spite of suffering from substance dependence. Thus, health professionals should be concerned about drugs and driving. Given the high proportion of active drivers with substance dependence in our sample, these patients should be counselled with respect to traffic safety and assessments and driver rehabilitation programmes should be integrated within inpatient treatment programmes.
With respect to driving cessation, female gender and age were predictive factors in our sample. This is in line with investigations showing that female drivers were more likely to self-regulate their driving behaviour than male drivers (Vance et al. 2006;Marie Dit Asse et al. 2014) and that increasing age is consistently associated with increased rates of driving cessation (Edwards et al. 2010). Our data also indicate that particularly those patients frequently presenting distinct cognitive impairments such as organic mental disorders and psychotic disorders restrict or even abandon driving. Disability levels vary widely in these disorders and it is especially the level of cognitive deficits that appear to be a valid predictor with respect to everyday functioning in patients with cognitive impairments (Green et al. 2000;Pereira et al. 2008;Evans et al. 2014). This seems to become more evident in patients as they grow older. Longitudinal studies in older drivers have emphasised the importance of cognitive functioning in maintaining driving ability (Ackermann et al. 2008;Edwards et al. 2010;McLeod et al. 2014). Thus, the question arises as to whether patients are behaving spontaneously and are self-motivated due to the awareness of cognitive deficits, or whether external pressure, e.g. from medical or public authorities, family members, etc., causes them to adapt their driving behaviour, or both. Not least, as there is substantial heterogeneity in causes and neuropsychological deficit profiles in diagnostic groups investigated, these may have different implications with respect to driving status.
The maintenance and rehabilitation of driving ability should be an integral goal of the treatment of psychiatric patients considering the significance for mobility and independence and not to forget for social functioning. Closer inspection of factors contributing to driving cessation due to mental disorders may be crucial when conceptualising therapeutic approaches and counselling patients. Compensational factors, e.g. the use of medications or other aids to maintain driving competence must be systematically investigated in future studies. Further investigations in this field are warranted to differentially elucidate the relationship between psychiatric illness, driving status and compensational aspects to optimise rehabilitation efforts.

Key points
Mental illness has a negative impact on road mobility, which is especially true for patients with an organic mental disorder or a schizophrenic disorder, i.e. diagnostic groups frequently presenting distinct neuropsychological deficits Given the high proportion of active drivers with substance dependence, health professionals should be concerned about traffic safety for this patient group. Patients should be specifically counselled with respect to traffic safety within inpatient treatment Finally data indicate that road mobility is a relevant issue for patients with a psychiatric illness that is closely linked to psycho-functional status and thus should be an integral goal of treatment strategies