Measuring cognitive performance in space: validity of a 3-min PVT administered on a handheld computer

Kurzfassung

Objectives: Sleep disturbances due to irregular work schedules or circadian misalignment occur regularly during stays in space and are a major reason for medication. Reduced sleep quality and quantity may impair cognitive performance and are likely to result in human errors and accidents. Psychomotor performance degradation is a sensitive marker for sleep loss induced vigilance deficits. We implemented a Psychomotor Vigilance Task (PVT) of particularly short duration (3 min) on a portable handheld computer which can be administered in operational settings. The current study aimed at examining the validity of this test. Methods: In the DLR sleep laboratory 47 volunteers (21 female, mean age 27 ± 5 years) were examined during 12 consecutive days and nights each. Participants’ sleep was restricted to induce fatigue and different levels of performance impairment (38h awake, 4h sleep, and 4h sleep after moderate alcohol intake). Two nights of recovery were granted between sleep restriction conditions in a cross-over design. During wakefulness participants conducted a battery of performance tests in 3-hour intervals (in total 63 sessions) that reflect the operational demands in dynamic environments such as aviation/space (4 tasks) and road transport (lane change task). For reliability assessment, we calculated internal consistency (Cronbach’s α) and parallel test reliability with a 10-min PVT on a desktop computer (Pearson’s correlation). To assess validity, we correlated the 3-min PVT with the tasks of the test battery under the different sleep loss conditions. Results from the 9am test sessions are reported. Results: Internal consistency was very high with α = 0.98. Correlations with the 10-min PVT proved to be significant showing a good reliability under sleep restriction and extended wakefulness (r: 0.63 to 0.87). Three of the 4 tasks reflecting operational demands in aviation/space correlated significantly with the 3-min PVT (r: 0.45 and 0.7). The fourth task correlated poorly presumably reflecting additional different aspects of performance. The 3-min PVT correlated significantly with performance in the lane change task (r: 0.34 and 0.56). Conclusions: The 3-min PVT detected performance impairments after different degrees of sleep restriction reliably and with good validity. Impairments observed in the 3-min PVT paralleled those in the 10-min PVT and in performance tasks reflecting demands of operators in aviation/space and transport.