Comparing Treatment Effect Measurements in Narcolepsy: The Sustained Attention to Response Task, Epworth Sleepiness Scale and Maintenance of Wakefulness Test

Abstract

Study objectives: To validate the Sustained Attention to Response Task (SART) as a treatment effect measure in narcolepsy, and to compare the SART with the Maintenance of Wakefulness Test (MWT) and the Epworth Sleepiness Scale (ESS).

Design: Validation of treatment effect measurements within a randomized controlled trial (RCT).

Patients: Ninety-five patients with narcolepsy with or without cataplexy.

Interventions: The RCT comprised a double-blind, parallel-group, multicenter trial comparing the effects of 8-w treatments with pitolisant (BF2.649), modafinil, or placebo (NCT01067222). MWT, ESS, and SART were administered at baseline and after an 8-w treatment period. The severity of excessive daytime sleepiness and cataplexy was also assessed using the Clinical Global Impression scale (CGI-C).

Measurements and results: The SART, MWT, and ESS all had good reliability, obtained for the SART and MWT using two to three sessions in 1 day. The ability to distinguish responders from nonresponders, classified using the CGI-C score, was high for all measures, with a high performance for the SART (r = 0.61) and the ESS (r = 0.54).

Conclusions: The Sustained Attention to Response Task is a valid and easy-to-administer measure to assess treatment effects in narcolepsy, enhanced by combining it with the Epworth Sleepiness Scale.

Keywords: ESS; MWT; SART; narcolepsy; sleepiness; sustained attention; treatment effects; validation; vigilance.

Figure 1

Factor analysis of CGI-C, SART, ESS, and MWT. (A) Factor analysis of the delta scores of the MWT (log-transformed), ESS, SART (log-transformed total error count), and CGI-C. (B) Factor analysis of the delta scores of SART outcome measures (all log-transformed), and CGI-C. The direction of the arrows represents the degree of correlation between the various measures. When arrows point in the exact same direction, they are perfectly positively correlated. Arrows pointing in opposite directions, with an angle between them of 180° indicate a perfect inverse, i.e. negative correlation. Arrows at right angles to one another reflect that the two measures are completely independent. The dashed line represents the 180° opposite of the CGI score. Panel A shows that the ESS and SART are more parallel to the CGI than the MWT, and the angle between them suggests that they capture different aspects. CGI-C, Clinical Global Impression of Change; ESS, Epworth Sleepiness Scale; log, log-transformed; MWT, Maintenance of Wakefulness Test; SART, Sustained Attention to Response Task.