Sleepiness assessment tools for predicting fitness to drive: a systematic review and meta-analysis

While the link between sleepiness and impaired driving performance is well-established, identifying high-risk sleepiness episodes is crucial for public safety. We assessed the predictive ability of sleepiness assessment tools in relation to impaired driving performance. A systematic review and meta-analysis searched four major databases, Medline (PubMed), Embase, Web of Science and the Cochrane Central Register of Controlled Trials (CENTRAL). Studies reporting associations between driving performance metrics (off-road events, Standard Deviation of Lateral Position, and reaction time) and at least one sleepiness measure (subjective: Epworth Sleepiness Score and Karolinska Sleepiness Scales [KSS]; objective: PERcentage of eye CLOsure [PERCLOS], Johns Drowsiness Scale, Karolinska Drowsiness Scale [KDS], EEG power alpha/theta ratio, and behavioural/physiological microsleep episodes) were included. Sixty-one studies were included (1441 participants, 353 observations). Both subjective (KSS [slope = 0.1, 95%CI 0.08; 0.12, p = 0.00]) and objective sleepiness measures (PERCLOS-80 [slope = 0.02, 95%CI 0.01; 0.04, p = 0.01]; KDS [slope = 0.00, 95%CI 0.00; 0.01, p = 0.02]; EEG power [slope = 0.01, 95%CI 0.07; 0.26, p = 0.00]; microsleep episodes [slope = 0.03, 95%CI 0.02; 0.03, p = 0.00]) were robustly associated with impaired driving performance. Certainty of evidence was low for KSS, PERCLOS-80, microsleep episodes, and ESS, and very low for KDS, EEG power, and JDS, owing to serious risk of publication bias and methodological limitations. These findings support associations between subjective and objective sleepiness measures and impaired driving performance but their predictive utility requires further investigation. Future research should assess combined tools under standardized real-world driving conditions in larger studies. PROSPERO: CRD42024427166.