RR-PL-0412

Sunday 11:30, Palau de Congressos, Hall 2, Room A

RELIABILITY OF A ROAD TEST AFTER STROKE. Akinwuntan AE, Feys H, De Weerdt W, Baten G, Kiekens C, Verkammen K, Vanleeuwe T; Department of Rehabilitation Sciences, Faculty of Physical Education and Physiotherapy, Katholieke Universiteit Leuven, Belgium. PURPOSE: The road test is generally acclaimed as the most valid test of driving performance. In Belgium, the road test included in the predriving assessment for stroke patients has not been assessed for its reliability. The purpose of this study was therefore to determine the reliability of the road test performed by stroke patients in the Belgian Institute for Road Safety. RELEVANCE: One of the responsibilities of experts (physical therapists, neurologists, neuropsychologists and occupational therapists) who assess brain injured people including stroke patients who wish to resume driving, is to use reliable assessments of driving competence. SUBJECTS: Thirty patients with sequelae of stroke who had been driving before onset of stroke and wish to resume driving. METHODS AND MATERIALS: The road test consisted of a closed-course section performed within the premises of the institute and an in-traffic section which was performed on a standardized 20 km trajectory. Subjects were accompanied and evaluated during the test by an occupational therapist in an automatic opel kadet car fitted with a video camera for recording driving performance simultaneously. The video recordings were evaluated by another occupational therapist and an external assessor. Performance on the road test was evaluated on a 17-item checklist. Each item consisted of a number of sub-items (55 in total) which were scored on a four-point ordinal scale of poor, fair, sufficient and good performance. ANALYSES: Interrater reliability was assessed by comparing:(I) evaluations during the real road test and video recording as judged by the other occupational therapist and (II) between the two evaluations from the video evidence. Reliability at sub-item level was determined using weighted percentage agreements. All agreements with values >85% were considered as sufficient. Item per item interrater reliability and those of the overall performances were determined using Intraclass correlation coefficients (ICC). An ICC was considered low when <0.40, moderate between 0.40 and 0.59, moderately high when between 0.60 and 0.79 and very high when >0.80. RESULTS: Comparisons revealed that majority of the sub-items (41 out of 55) had >85% weighted agreement. Item per item reliability values varied from 0.42 to 1.0 for most of the items. Interrater reliability for the overall performance was 0.62 in real-life versus video and 0.80 in video versus video comparison. CONCLUSION: The reliability of assessing overall performance of stroke patients in the road test is moderately high and even better when assessed under the same condition. The outcome of this study provides a basis for comparison by future similar studies.