Microsleep and sleepiness: a comparison of multiple sleep latency test and scoring of microsleep as a diagnostic test for excessive daytime sleepiness

Objective: To compare multiple sleep latency test (MSLT) and scoring of microsleep (presence of sleep electroencephalograph between 3 and 15 s in an epoch) as a diagnostic test for excessive daytime sleepiness (EDS). Design: A retrospective study. Setting: Sleep center at a tertiary care teaching hospital. Subjects: Patients referred to a sleep center who had an MSLT and one or more of the following symptoms; tiredness, sleepiness, memory loss, accidents/near accidents and gap driving. Interventions: Full night polysonmography (PSG) and next day MSLT were performed. Patients were classified as 'microsleep-positive' or 'microsleep-negative' according to presence or absence of microsleep. Results: Patients (n = 92) were divided into three groups according to their MSLT results; group A had an MSLT less than or equal to 5 min (n = 38), group B had an MSLT 6-10 min (n = 26), and group C had an MSLT > 10 min (n = 28). The number of patients with symptoms of tiredness and memory loss were statistically higher in group A compared with groups B and C (P = 0.036). The number of patients with symptoms of EDS, in groups B and C, was significantly higher in patients with microsleep than without microsleep (P < 0.05). By a paired McNemar's test, the better performance of adding microsleep to MSLT (sensitivity 42.9%; specificity 63.6%) to assess EDS was statistically significant (P = 0.0096). Conclusions: Microsleep determination during an MSLT is a more sensitive and specific test for EDS as compared to MSLT alone. (C) 2002 Elsevier Science B.V. All rights reserved.