Identifying clinically important difference on the Epworth Sleepiness Scale: results from a narcolepsy clinical trial of JZP-110

Background: While scores ≤10 on the Epworth Sleepiness Scale (ESS) are within the normal range, the reduction in elevated ESS score that is clinically meaningful in patients with narcolepsy has not been established.

Methods: This post hoc analysis of a clinical trial of patients with narcolepsy evaluated correlations between Patient Global Impression of Change (PGI-C) and ESS. Data of adult patients with narcolepsy from a double-blind, 12-week placebo-controlled study of JZP-110, a wake-promoting agent, were used in this analysis. Descriptive statistics and receiver operating characteristic (ROC) analysis compared PGI-C (anchor measure) to percent change from baseline in ESS to establish the responder criterion from patients taking either placebo or JZP-110 (treatments).

Results: At week 12, patients (n = 10) who reported being "very much improved" on the PGI-C had a mean 76.7% reduction in ESS score, and patients (n = 33) who reported being "much improved" on the PGI-C had a mean 49.1% reduction in ESS score. ROC analysis showed that patients who improved were almost exclusively from JZP-110 treatment group, with an area-under-the-curve of 0.9, and revealed that a 25% reduction in ESS (sensitivity, 81.4%; specificity, 80.9%) may be an appropriate threshold for defining a meaningful patient response to JZP-110 and placebo.

Conclusions: A ≥25% reduction in patients' subjective ESS score may be useful as a threshold to identify patients with narcolepsy who respond to JZP-110 treatment.