M Ranchet1, A E Akinwuntan1,2, M Tant3, A Salch1, E Neal1, H Devos1. 1. Department of Physical Therapy, College of Allied Health Sciences, Augusta University, Augusta, GA, USA. 2. Dean's Office, School of Health Professions, The University of Kansas Medical Center, Kansas City, KS, USA. 3. CARA, Belgian Road Safety Institute, Brussels, Belgium.
Abstract
BACKGROUND AND PURPOSE: Physicians often struggle to screen out patients who are no longer fit to drive after stroke. The agreement between the recommendations of physicians and on-road assessors with regard to fitness to drive after stroke was investigated. METHODS: In this retrospective study, 735 patients with stroke underwent medical, visual and road tests at an official fitness-to-drive center of the Belgian Road Safety Institute. Physicians provided medical fitness-to-drive recommendations using one of three categories (favorable, reserved or unfavorable). On-road assessors used the same three categories to make practical fitness-to-drive recommendations. Agreement between the medical and practical fitness-to-drive recommendations was calculated using the percentage of agreement and prevalence and bias adjusted kappa (PABAK). Area under the curve (AUC) was used to predict the medical and practical recommendations after stepwise logistic regression analyses. RESULTS: The percentage of agreement was 73% and the PABAK was 0.60 (P < 0.0001). Physicians disagreed on 92% of patients classified as unfavorable and 80% of those classified as reserved by the on-road assessor. Previous visits to the driving center and number of comorbidities predicted medical fitness-to-drive recommendations (AUC = 0.68). Age, previous visits to the center, binocular acuity and driving experience constituted the best model to predict practical fitness-to-drive recommendations (AUC = 0.70). CONCLUSIONS: Although there was a moderate agreement between the medical and practical fitness-to-drive recommendations, physicians were less likely to screen out those patients who may pose an actual risk on the road. Demographic, clinical and driving factors differently affected the medical and practical fitness-to-drive recommendations.
BACKGROUND AND PURPOSE: Physicians often struggle to screen out patients who are no longer fit to drive after stroke. The agreement between the recommendations of physicians and on-road assessors with regard to fitness to drive after stroke was investigated. METHODS: In this retrospective study, 735 patients with stroke underwent medical, visual and road tests at an official fitness-to-drive center of the Belgian Road Safety Institute. Physicians provided medical fitness-to-drive recommendations using one of three categories (favorable, reserved or unfavorable). On-road assessors used the same three categories to make practical fitness-to-drive recommendations. Agreement between the medical and practical fitness-to-drive recommendations was calculated using the percentage of agreement and prevalence and bias adjusted kappa (PABAK). Area under the curve (AUC) was used to predict the medical and practical recommendations after stepwise logistic regression analyses. RESULTS: The percentage of agreement was 73% and the PABAK was 0.60 (P < 0.0001). Physicians disagreed on 92% of patients classified as unfavorable and 80% of those classified as reserved by the on-road assessor. Previous visits to the driving center and number of comorbidities predicted medical fitness-to-drive recommendations (AUC = 0.68). Age, previous visits to the center, binocular acuity and driving experience constituted the best model to predict practical fitness-to-drive recommendations (AUC = 0.70). CONCLUSIONS: Although there was a moderate agreement between the medical and practical fitness-to-drive recommendations, physicians were less likely to screen out those patients who may pose an actual risk on the road. Demographic, clinical and driving factors differently affected the medical and practical fitness-to-drive recommendations.
Authors: Megan A Hird; Kristin A Vesely; Tahira Tasneem; Gustavo Saposnik; R Loch Macdonald; Tom A Schweizer Journal: Front Neurol Date: 2018-02-13 Impact factor: 4.003