A pre-drive ocular assessment predicts alertness and driving impairment: A naturalistic driving study in shift workers. (February 2020)

• Record Type: Journal Article
• Authors: Mulhall, Megan D.; Cori, Jennifer; Sletten, Tracey L.; Kuo, Jonny; Lenné, Michael G.; Magee, Michelle; Spina, Marie-Antoinette; Collins, Allison; Anderson, Clare; Rajaratnam, Shantha M.W.; Howard, Mark E.
• Is Part Of: Accident analysis and prevention. Volume 135(2020)
• Publication Date: 2020-02
• Subjects: Drowsy driving -- Ocular measures -- Driving impairment -- Predicting performance -- Fatigue; Accidents -- Prevention -- Periodicals; Accident Prevention -- Periodicals; Accidents -- Prévention -- Périodiques; 363.106
• DOI: 10.1016/j.aap.2019.105386
• Languages: English
• ISSNs: 0001-4575

Abstract

Highlights: Examined if a pre-drive ocular assessment can predict alertness and sleep-related driving impairment on the subsequent drive, in naturalistic driving in shift workers. Blink duration significantly predicted behavioural microsleeps and showed promise for predicting lane departures. A 4-minute pre-drive ocular assessment predicted lane departures and behavioural microsleeps with more accuracy than a 2-minute pre-drive ocular assessment. Abstract: Sleepiness is a major contributor to motor vehicle crashes and shift workers are particularly vulnerable. There is currently no validated objective field-based measure of sleep-related impairment prior to driving. Ocular parameters are promising markers of continuous driver alertness in laboratory and track studies, however their ability to determine fitness-to-drive in naturalistic driving is unknown. This study assessed the efficacy of a pre-drive ocular assessment for predicting sleep-related impairment in naturalistic driving, in rotating shift workers. Fifteen healthcare workers drove an instrumented vehicle for 2 weeks, while working a combination of day, evening and night shifts. The vehicle monitored lane departures and behavioural microsleeps (blinks >500 ms) during the drive. Immediately prior to driving, ocular parameters were assessed with a 4-min test. Lane departures and behavioural microsleeps occurred on 17.5 % and 10 % of drives that had pre-drive assessments, respectively. Pre-drive blink duration significantly predicted behavioural microsleeps and showed promise for predicting lane departures (AUC = 0.79 and 0.74). Pre-drive percentage of time with eyes closed had high accuracy for predicting lane departures and behavioural microsleeps (AUC = 0.73 and 0.96), although was not statistically significant. Pre-drive psychomotor vigilance task variables were not statistically significant predictors of lane departures. Self-reported sleep-related and hazardous driving events were significantly predicted by mean blink duration (AUC = 0.65 and 0.69). Measurement of ocular parameters pre-drive predict drowsy driving during naturalistic driving, demonstrating potential for fitness-to-drive assessment in operational environments.