Enhancing the Clinical Utility of DriveSafe DriveAware for People with Huntington’s Disease

INTRODUCTION

Huntington’s disease (HD) is a progressive, neurodegenerative disorder characterised by motor, cognitive and neuropsychiatric impairments [1–3]. Cognitive symptoms in HD can be detected many years prior to a clinical diagnosis [4–6] and affect many aspects of life, including driving [3, 8]. Impairments in cognitive domains such as sustained and divided attention [9, 10], processing speed [8, 11], planning ability [11, 12], and visual perception [8, 13] have demonstrated associations with reduced driving ability. Specifically, for people with HD, multiple cognitive impairments are a strong predictor of reduced driving ability [14–16]. Therefore, frequent examinations of driving-related abilities form an essential part of HD clinical care [8, 15].

Previously, our group assessed the clinical utility of DriveSafe DriveAware (DSDA) [17] for people with HD [18]. DSDA is a cognitive screening assessment designed to assess driving awareness and determine the need for intensive on-road testing [19, 20]. DSDA can be administered in 15 minutes on an iPad by a clinician, limiting both financial and time costs [20]. DSDA trichotomizes participants into three categories [20]. Participants who are categorised as ‘likely to fail an on-road assessment’ and ‘needs further testing’ are recommended to complete on-road assessments. Those in the third category, ‘likely to pass an on-road assessment’, are predicted to be safe drivers.

We showed that people with HD who ‘failed’ DSDA functioned more poorly in terms of cognitive and motor domains compared to those with HD who ‘passed’ [18]. We further categorised the ‘pass’ group into ‘low pass’ and ‘high pass’. The ‘low pass’ group were found to display significantly more cognitive deficits compared to the ‘high pass’ group. Because multiple cognitive impairments are a strong predictor of unsafe driving for people with HD [8, 15], we recommended the ‘low pass’ category is interpreted with caution.

This short communication examines whether DSDA could be improved for people with HD by incorporating variables from DSDA that are recorded, but not included in the scoring algorithm, so that driving appraisals are accurate. Specifically, we examined DriveSafe completion time, Intersection Rules completion time, and performance on Intersection Rules. Driving often involves rapid scanning of the environment to detect hazards and quick decision-making, and is highly reliant on processing speed [28, 29]. DSDA completion time is collected and necessary to consider as significant reductions in processing speed are present early in HD [4–6]. We examined whether people with HD completed DSDA subtests more slowly and had poorer performance on Intersection Rules compared to healthy controls. Furthermore, we explored whether completion time and Intersection Rules scores were associated with poorer performance on cognitive assessments previously shown to be predictive of driving ability.

METHODS

We recruited a control group age and gender matched to the HD sample. Our sample included 52 participants, comprised of 26 participants with genetically confirmed HD CAG expansion and a clinical diagnosis of HD (HD group) and 26 healthy controls (control group). The HD group was comprised of 19 symptomatic participants and seven premanifest participants. Total Functional Capacity (TFC) also varied; 13 participants were categorised as having no or mild functional impairment (TFC score = 11 to 13) and the remaining six categorised as having a moderate level of functional impairment (TFC score = 7 to 10), see Table 1.

In accordance with the manual [20], participants completed all three components of DSDA; DriveSafe, DriveAware, and Intersection Rules.

We then administered six neuropsychological tests assessing cognitive functions important for safe driving. We used standard testing and scoring procedures as indicated by manuals. The Trail Making Test A and B (TMT) [21] was used to assess sustained and divided attention. The Symbol Digit Modalities Test (SDMT) [22] was used to assess processing speed. The Mazes subtest from the Neuropsychological Assessment Battery (NAB) [23] was used to assess planning ability. The Line Orientation subtest from the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) [24] and the Rey Complex Figure Test (RCFT) [25] were used to measure visual perception.

We used independent-sample t tests to assess differences between the HD and control groups. Spearman’s rho was utilised for correlational analyses and Bonferroni adjusted p-values to control for inflated type-one error rate of multiple comparisons.

RESULTS

We found that the HD group were significantly slower on the DriveSafe subtest compared to controls (t(34.25) = 5.63, p < 0.001) (HD group: Msec = 798.46, SD = 236.46; control group: Msec =513.54, SD = 103.57). The difference in means was 4.75 minutes (95% CI 182.07 to 387.78), representing a large effect size (Cohen’s d = 1.56). For the Intersection Rules subtest, the HD group was significantly slower than the control group (t(50) = 3.25, p = 0.002) (HD group: Msec = 196.46, SD = 77.23; control group: Msec = 131.81, SD = 65.66). This difference of 1.08 minutes (95% CI 24.72, 104.59) represents a large effect size (Cohen’s d = 0.90). In contrast, accuracy on the Intersection Rules subtest did not differ between groups (p > 0.05). We found significant associations between slower times to complete subtests and poorer performance on neuropsychological assessments and lower scores on the DriveSafe subtest. In contrast, no significant associations were found between poorer performance on the Intersection Rules subtest and neuropsychological assessments or subtest scores, see Table 2.

Categorisations from DSDA for the HD group showed that 18 participants were classified as ‘likely to pass’, three participants were classified as ‘required further testing’, and five participants were classified as ‘likely to fail’. For the control group, all 26 participants were classified as ‘likely to pass’.

DISCUSSION

Overall, we found that the time taken to complete DSDA is associated with cognitive functions important for safe driving. While this information is not calculated in the DSDA scoring algorithm, impairments in processing speed occur early in HD. Therefore, considering how long is required to complete DSDA may tailor the assessment more specifically to those with HD and assist clinicians to make more detailed appraisals about driving ability, particularly in the absence of detailed cognitive assessments. Our results showed that the HD group were significantly slower at completing DriveSafe and Intersection Rules subtests, which likely reflects the group’s greater cognitive impairments, particularly processing speed. Slower processing speed has implications for safe driving [26, 27], and for people with HD who have impaired processing speed, they may be less efficient in scanning the environment and slower to respond to environmental changes. Because driving often involves rapid scanning of the environment to detect hazards [28, 29], completion time in DSDA may provide useful insight into processing speed and therefore, how people with HD are likely to perform on an on-road assessment.

Our finding that accuracy on the Intersection Rules subtest was similar for HD and controls is consistent with findings from the DSDA developers, who found the subtest was less sensitive at predicting driving performance compared to DriveSafe and DriveAware. This may be due to the restricted range of Intersection Rules, which only covers a limited range of rules testing judgement and knowledge and may explain why a difference was not found.

We found that slower completion times on DriveSafe and Intersection Rules subtests were associated with lower scores on DriveSafe and poorer performance on tests of attention, planning, visual perception and processing speed. Whilst the significant association between lower scores on DriveSafe and slower completion times indicates the assessment is identifying individuals with cognitive impairments likely to impact driving, our previous study raised concerns that some participants with HD were passing despite having many cognitive impairments. Given previous studies have demonstrated the importance of processing speed for safe driving in HD [8–13], our significant finding that people with HD completed DSDA more slowly and that slower times were associated with poorer cognitive performances may have implications for driving ability. Further research is needed to determine whether completion time in DSDA is correlated with a broader range of cognitive measures, particularly tests of psychomotor speed. Our initial results show that clinicians could consider completion time in DSDA when administering it to people with HD. Utilising completion time may tailor DSDA for people with HD so that it is further reflective of cognitive functioning and driving fitness.

The main limitation of the study was that the HD and control groups were small, which may limit generalisability of results. Importantly, we were unable to validate the utility of the DSDA plus timing strategy using on-road driving assessments, which are the gold standard for determining driving safety. Our results demonstrate initial support for the potential use of completion time to tailor DSDA for people with HD so that the assessment is more reflective of HD-related cognitive functioning, and therefore, safe driving in this population. This may improve the utility of DSDA and ensure people with HD are able to access a cost-effective and reliable measure of driving ability, particularly in multidisciplinary settings where access to detailed cognitive assessments is not available. Future research should assess whether DSDA completion time is associated with real-world driving performance to validate its use in DSDA for people with HD.

CONFLICT OF INTEREST

The authors have no conflict of interest to report.