Which is the Best Way to Assess and Follow-Up Fitness to Drive after Stroke?

Special Article - Stroke Rehabilitation

Phys Med Rehabil Int. 2015;2(6): 1054.

Which is the Best Way to Assess and Follow-Up Fitness to Drive after Stroke?

Björkdahl A1,2,3*, Nilsson L¹ and Jönsson U¹

1Sahlgrenska University hospital, Sweden

2Gothenburg University, Sahlgrenska Academy, Neuroscience and Physiology, Rehabilitation medicine, Sweden

3Ersta Sköndal University College, Campus Bräcke, Sweden

*Corresponding author: Ann Björkdahl, Ersta Sköndal University College, Campus Bräcke, Box 21062, 41804 Gothenburg, Swedan

Received: June 15, 2015; Accepted: July 23, 2015; Published: July 27, 2015


Aim: To explore the feasibility to make on-road assessments routinely at 3 months follow-up for all patients with a 3 months verbal prohibition of driving after stroke, to support the physicians decision of fitness to drive.

Methods: From September 2007 to December 2009 there were 151 stroke patients from the stroke units at the hospital eligible for inclusion in the study. Fifty agreed to be assessed by the Nordic Stroke Driver Screening Assessment (NorSDSA) and on-road assessment. As base for discussion about the consequences on resource use relative to accuracy, calculations were made to explore and compare an expected yearly cost for two different assessment conditions, the NorSDSA followed by on-road assessment in uncertain cases and on-road assessment for all cases.

Findings: The yearly need for driving assessments was estimated to 500 patients. With less accuracy than only on-road assessments the NorSDSA with the stipulated cut-off produced a cost benefit of 1,700 €. The NorSDSA resulted in 32% uncertain cases and the certain cases were in 17% in disagreement with the on-road assessment, the gold standard.

Conclusion: It is conceivable and could be recommended to make on-road assessments for all patients with stroke at 3 months follow-up as accuracy is of importance both for patients and society and may save resources in the long run.

Keywords: Driving; Cognitive impairment; Screening; Cost effective; NorSDSA; On-road


Many of the persons who survive a stroke are drivers [1]. Since a stroke can affect fitness to drive, physicians are obliged, by Swedish law to determine whether the patient can be allowed to drive after the injury. All patients not fit to drive should be reported to the Swedish Transport Agency that confiscate their driving license. However, as recovery may occur after stroke the Swedish National Guidelines for stroke recommend that the physician discuss driving with all patients before discharge from hospital and make a verbal agreement, with the patients that wish to resume driving, to refrain from driving during a period of 3-6 months. In too many cases there is a lack of compliance with the guidelines and no discussion and decision regarding driving is made. Similar problems seem to be present in many countries according to the report from Fisk et al. [2] indicating that nearly 90% of the stroke drivers do not receive any type of driving evaluation. A reason for the lack of decisions regarding driving after stroke may be that many physicians feel that they are poorly informed and trained in terms of assessing driving fitness [3-5]. In Sweden in some cases the patient gets a verbal prohibition of driving for 3 months at discharge from hospital but follow-up of the agreement with a new decision failure. With this background there is a need for practical and explicit routines on how to handle the issue of driving after stroke in order to facilitate the obligatory duty of the physician as well as make information and procedures clear and equal for all patients.

Driving is a complex activity that requires sufficient attention, perception, and information processing. All of these functions can be affected by a brain injury [1, 6]. To handle complex situations in traffic, driving requires rapid judgments, automatic reactions, and flexibility. It can be difficult for a physician to assess these abilities during a short clinical examination. In many countries there are special driving assessment centers. Sweden does not have that organisation and there are only few official centers for traffic medicine. Since there is a need for evaluations made by professionals with knowledge in traffic medicine, every region, hospital or care centre tries to organise their own routines for how to decide on fitness to drive.

After a stroke the physician, following the National guidelines, should make an agreement with all patients that are drivers to refrain from driving during 3 months, which is clear and easy. At time to decide if there should be a permanent prohibition or the patient should be allowed to resume driving after 3 months it is more problematic. Due to the complex situations that can be encountered during driving and the interacting abilities that are needed to deal with these situations, there is currently no universal, standardised way to assess fitness to drive [7, 8]. The available instruments seldom comprise items for the entire range of situations encountered during driving, which can range from simple to complex [9]. Different assessments also have different levels of ecological validity. In this case, the term ecological validity means the extent to which the assessment reflects a person’s true fitness to drive. There is evidence in the literature that suggests that assessments in real traffic with skilled evaluators have better ecological validity than neuropsychological assessments in a clinic [7, 10]. At present, the on-road assessment is the most accepted method for determining fitness to drive, although the methods have not been standardised [10, 11]. Consequently, the on-road assessment is often used as the gold standard.

However, other factors such as resource utilisation may also shape the choice of assessment method and on-road assessment is relatively time consuming and need collaboration with a driving school. Therefore, the ideal situation would be to have screening methods that allow the identification of cases that need a more thorough examination [12]. The Nordic Stroke Driving Screening Assessment (NorSDSA) is widely used in Sweden and specifically designed for assessing driving. It has been found to be one of the available tests that best reflect fitness to drive [13]. In spite of this, these kinds of tests are limited, and a study by Selander et al.[14] conclude that the NorSDSA should not be used as a stand-alone test to determine the fitness to drive of individuals, which is the most common way in Sweden [15].

From the perspective of the stroke driver, evaluations and information about driving after stroke is unclear and with unexpected results according to their self-appraisal and a decision to cease driving disrupt ability to participate in meaningful activities [16].

Given the above aspects of problems with routines, evaluations and information it is important to explore possible solutions to improve routines for the physicians, accuracy in evaluations and understanding of the patient for the decisions that are made. As onroad assessment has been found to be the most ecologically valid way to examine fitness to drive and yet not is the standard evaluation, it would be of interest to explore if it would be feasible and appropriate to use the on-road assessment routinely. Probably the main reason for not using the on-road assessments is that it is considered to be time-consuming and costly. The aim of the study was to explore the feasibility to make on-road assessments routinely at 3 months followup for all patients with a 3 months verbal prohibition of driving after stroke, to support the physicians’ decision of fitness to drive.

Research questions

1. Is it feasible and appropriate to make on-road assessments at a 3 months follow-up for all patients with stroke?

2. Is there an acceptable difference in accuracy between NorSDSA and on-road assessment as gold standard in order to choose NorSDSA instead of on-road assessment?


All patients admitted to a primary stroke unit at a university hospital between September 2007 and June 2009 was considered for inclusion in the study. Patients were excluded if they did not have a driving license or driving experience, or would be dependent on extensive care in the future. Patients with direct medical obstacles to driving, such as epilepsy, neglect or hemianopsia, were also excluded. Patients in need of an adapted vehicle were referred to a driving centre as this kind of resources was not available in the present structure. Adaptations would not be made before no further improvements would be expected, which is too early to determine at 3-months follow-up. The occupational therapists at the unit searched the medical records to determine if there were reasons for exclusion from the study according to the exclusion criteria’s.

Due to reorganisation of the stroke units at the hospital in June 2009, patients thereafter until the end of December 2009 were recruited from other hospitals belonging to the university hospital. From these hospitals patients in needed of a more thorough investigation were referred to the study by a physician.


The routine on the units was that all patients diagnosed with a stroke were prohibited from driving by verbal agreement until they had seen the physician for a new decision during a return visit 3 months after discharge. At discharge, the patients who were eligible for a driving assessment received a letter informing them that they would be contacted shortly before their return visit to the hospital. Just before the return visit, the patients were contacted by telephone and asked to participate in a driving fitness examination. Just before the return visit, the patients were contacted by telephone and asked to participate in a driving fitness examination. They were informed that the study was a possibility to have a more thorough assessment of fitness to drive and would supply the physician with material for rapid and correct decisions of driving.

If the patient consented to participate, the examination was performed in two sessions at the driving school. The first session, NorSDSA was administered. At the second session, an on-road assessment was performed assessed by a different OT. This assessment took place in a car with dual controls and the patient was accompanied by a traffic instructor responsible for safety and provision of instructions for the drive on a predetermined route that included a range of different traffic situations. Before finishing, the patient was asked to give his or her opinion of his/her performance and was also given feedback by the OT of her perception of the performance. The discussion after the driving produced information about the patient’s awareness of problems, which was also documented.

The OT documented the result in the medical record. In total, three OTs experienced in working with stroke patients and trained in traffic medicine, were involved in the assessments. As there was no available standardised observation method for the on-road assessment at the time, an observation template and guidelines were made to secure as equal assessments as possible. To assure acceptable inter-reliability, the OTs also simultaneously observed and independently assessed, the on-road driving of two different patients. These assessments were then compared and discussed. There were an agreement on the pass or fail result and good agreement regarding their notes on the observation template. The OT and the traffic instructor compared their experiences from the drive before the feed-back to the patient and a good agreement was found on the general question of pass or fail, 96% agreement on the 50 assessments in the study. However, the decision was to be made by the OT.


Nor SDSA: The Stroke Driver Screening Assessment (SDSA) is a cognitive instrument that was developed in Great Britain by Nouri and Lincoln to assess driving after stroke [17]. The SDSA has been translated and validated for Nordic traffic conditions, where it is known as NorSDSA [13, 18]. The adaptations made in the Nordic version included changing the illustrations of traffic situations so that the driver was on the right-hand side of the car. The traffic signs were also adapted to reflect Swedish conditions. The instrument has been shown to be valid and reliable [13, 17]. The instrument comprises four tasks: dot cancellation, directions, compass directions, and recognition of traffic signs. The time is recorded and the result is given in points. Dot cancellation examines precision and visual selective attention by a task to mark all groups of four dots on a sheet with several lines. In the Direction task, the patient is given a number of cards with a lorry and a car going in different directions and he/ she must find the right square in which to place the card according to the arrows in a co-ordinate system. Compass directions demands good visuospatial orientation, the ability to reason and understand the relationship between the different directions, the ability to divide attention, and good working memory, when placing cards with two vehicles leaving a traffic circulation in different directions in a coordinate system. Recognition of traffic signs examines the ability to use different traffic signs in different traffic situations, mental processing speed, and working memory.

The result of the test is divided into:

There is no clear instruction on how to handle the uncertain results. In this study we have chosen to see these results as uncertain and therefore in need of additional assessments.

On-road assessment: The assessment was based on the literature on how to perform a valid and reliable road test [10, 19]. The following traffic situations were included: traffic lights, circulation, crossroads, left turn, crossing bicycle path, prohibition to turn left, country road, highway, residential area, speed limits of 50, 70, and 90 km/h, and following traffic signs. The route was the same for all assessments and took about an hour to complete. The distance was 25Km.

During the drive, an OT observed and documented the patient’s performance, driving style, behavior and problems on an observation template according to the domains of assessment described by Patomella et al. [20, 21]. The instrument was under development and could therefore not yet be fully used. The domains that were assessed were maneuvering, orientation, conforming to traffic regulations, maintaining control, reaction, attention, and problem solving. After the drive, the observations were compiled and collated into a decision of pass or fail.

For a decision to fail the patient, the drive had to contain several insecure situations or "a risk" of serious accident or actions and situations that required the intervention of the traffic instructor for reasons of safety. The on-road assessment was documented in the medical records: the results in each observation template area were recorded along with a clear recommendation regarding pass or fail.

Administration: The median time for preparation before the assessment, performance, summation and documentation was 1,5 hours for the NorSDSA respectively 2,5 hours for the on-road assessment. As the study was exploring the need for assessments, it was convenient to use a driving school providing the resources just when needed. With this organisation there was a fee for the use of the car, and the instructor which the patient had to bear (1,000 SEK ≈ 106 €).


To examine the proportion in need of further examination with an on-road assessment after the NorSDSA, due to a not certain result of pass or fail, the four NorSDSA categories were calculated and the two middle categories representing uncertain cases were merged into one category. Second, the total time consumption and cost were computed and compared to the costs of a scenario where only onroad assessments were given in all cases.

As the accuracy of the results are of great importance further examinations were performed in order to explore if there were discrepancies in decisions between NorSDSA pass or fail and the gold standard, on-road assessment. For this purpose analysis of the agreement between NorSDSA and on-road assessments was performed with a Kappa analysis. This analysis yielded an agreement value and its significance. According to Peat [22], Kappa values of 0.5, >0.7, and >0.8 represent moderate, good, and very good agreement, respectively. Positive and negative predicted values were also calculated to determine the probability that Nor SDSA screening accurately detects the patients who are unfit and fit to drive.

Additionally the results of the two assessments were plotted together (see Figure 3) to explore if there were areas with total agreement. As base for discussion about the consequences on resource use relative to accuracy, calculations were made to explore how changed cut-offs on Nor SDSA would influence the costs.

Ethical aspects

The study was considered to be required for developing routines and an application for ethical approval was not acquired, but was approved by the director of department. However, ethical considerations were taken into account to ensure that the patients were not at risk. Written information was provided and the participants were asked for their consent to participate in the study.


Since there was a reorganisation in June 2009, only data collected before that was included in the analyses exploring the proportion of patients in need of a driving assessment after the stroke onset. In total, 441 patients had a stroke and were admitted to the stroke unit of the hospital during study period, September 2007 to June 2009. Of these, 110 (25% of the whole cohort) were eligible and included in the analyses of the number of patients that yearly would require decisions on driving. Every year, approximately 2000 patients are admitted to the stroke units of the university hospital. Extrapolation of the eligible rate of our patients to this number means that about 500 cases would have to be assessed for driving fitness by the physicians every year.

Thirty patients (27% of the eligible patients) agreed to be assessed for their fitness to drive and could be included in the comparisons between the two assessment methods (Figure 1). There were different reasons for the drop-out of the 80 eligible patients not included, such as they would stop driving, disagreed with making the assessment or had already got a clearance for driving from a physician not aware of the study. However, the sample was representative for patients were it may be questionable if the patient should drive or not three months after stroke. From the 41 patients referred from the other stroke units in the hospital complex after the reorganisation 20 agreed to participate in the assessments.