0622 The Psychomotor Vigilance Task as a Diagnostic Tool for Hypersomnolence. (12th April 2019)

• Record Type: Journal Article
• Authors: Moron, Danielle; Mariano, Christianna; Bliwise, Donald L; Maness, Caroline; Olvera, Victoria; Saini, Prabhjyot; Rye, David; Trotti, Lynn Marie
• Is Part Of: Sleep. Volume 42(2019)Supplement 1
• Page Start: A247
• Page End: A248
• Publication Date: 2019-04-12
• Subjects: Sleep -- Physiological aspects -- Periodicals; Sleep disorders -- Periodicals; Sommeil -- Aspect physiologique -- Périodiques; Sommeil, Troubles du -- Périodiques; Sleep disorders; Sleep -- Physiological aspects; Sleep -- physiological aspects; Sleep Wake Disorders; Psychophysiology; Electronic journals; Periodicals; 616.8498
• DOI: 10.1093/sleep/zsz067.620
• Languages: English
• ISSNs: 0161-8105

Abstract

Abstract: Introduction: Current tools to evaluate daytime sleepiness lack sensitivity and specificity. Additions to the multiple sleep latency test (MSLT) have been proposed, including the psychomotor vigilance test (PVT). The PVT has been extensively examined in controls to characterize effects of sleep loss, but its use as a tool to diagnose and classify individual patients remains incompletely characterized. We assessed whether sleep clinicians/researchers could differentiate PVT performance between patients with hypersomnolence and non-sleepy controls tasked with either performing their best or simulating sleepiness. Methods: Twenty-eight patients (EDS group, mean age 35.2 (SD 15.8), 64.3% women) with excessive daytime sleepiness confirmed with clinical interview, Epworth>10, and MSLT mean sleep latency ≤ 8 minutes underwent 10-minute PVT, typically before MSLT nap 2. Thirty-seven age- and gender-matched, non-patient controls, all with Epworth≤ 10, also performed 10-minute PVT. Controls were instructed to perform their best (CON group) or to make their performance look as though they were sleepy, without specific instructions about how to do so (simulated sleepiness, CON-SS). Four sleep clinician-scientists (2 ABPN board-certified in Sleep Medicine, 1 ABSM board-certified, and 1 nurse practitioner) and two sleep researchers classified each participant's detailed PVT output, blinded to group allocation. Kappa was calculated using SAS MAGREE macro. Results: Individual raters performed similarly, with four reviewers correctly classifying 55% of cases, one 57%, and one 65%. Classification was significantly better than chance, with moderate agreement between raters (kappa 0.54, p < 0.0001). Considering each group separately, agreement remained higher than chance. Agreement was only fair for EDS (kappa 0.40), but was substantial for CON and CON-SS (kappa 0.62 for both). When EDS were misclassified, they were typically classified as CON (87.4% of misclassifications). When CON were misclassified, they were typically classified as EDS (89.2%). When CON-SS were misclassified, they were typically classified as EDS (74.3%). Conclusion: The PVT can provide supplemental information as part of hypersomnolence evaluation. However, used in isolation, it would result in substantial misclassification of hypersomnolent patients. Patients with pathologic sleepiness cannot always be distinguished from non-sleepy controls attempting to appear sleepy. Support (If Any): K23NS083748