Predicting cognitive fitness to drive with touchscreen DriveSafe DriveAware

Abstract

Driving is a highly valued daily living activity easily disrupted by illness, injury, or age-related changes. General practitioners (GPs) are ultimately responsible for determining medical fitness to drive but lack valid and reliable tools. The desktop (original) version of DriveSafe DriveAware (DSDA) is a promising, valid and reliable test but it is not practical for medical practice. The concern of this thesis was the conversion of original DSDA into a touchscreen test of cognitive fitness to drive for GPs and occupational therapists to use in predicting patient driving performance without on-road testing. Because we were transitioning from a test administered and scored by a trained assessor to one where patient touch responses were scored in-app, we needed to develop an automatic data collection and scoring system that reflected the decision that would otherwise have been made by an expert-rater. We tested usability of the system with older adults then examined set scoring parameters to determine if these discriminated at-risk from comparison drivers. Results indicated the system we designed reflected the decisions that would have been made by a trained assessor. Next, we conducted a study to examine the internal validity, reliability, and predictive validity of data gathered with touchscreen DSDA. The criterion measure was outcome of a standardised occupational therapy on-road assessment. Rasch analysis provided evidence that touchscreen DSDA had retained the strong psychometric properties of original DSDA. However, results of a discontinued feasibility study indicated potential barriers to uptake of the test by physicians. Touchscreen DSDA may rather be a tool for occupational therapists to use in driver screening and addressing the community mobility need of their clients. Research indicates no there is no one best tool for screening fitness to drive. However, the thesis findings indicate touchscreen DSDA is one useful tool.