0366 SUBJECTIVE-OBJECTIVE DISCREPANCIES IN TREATMENT OUTCOME WITH FOUR TYPES OF HYPNOTICS: A FIRST PASS ANALYSIS

• Published in: Sleep (New York, N.Y.) Vol. 40; no. suppl_1; p. A136
• Main Authors: Culnan, E, Khader, W, Morales, K, Kloss, JD, Baglioni, C, Gencarelli, A, Muench, A, Vargas, I, Riemann, D, Perlis, ML
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 28-04-2017
• Subjects: Sleep
• ISSN: 0161-8105; 1550-9109
• DOI: 10.1093/sleepj/zsx050.365

Abstract
Introduction:
Hypnotics may have different effects on subjective (sleep diary [SD]) and objective (PSG) measures of sleep continuity. Thus, the concurrent examination of both measurement strategies, while rarely done, is important because improvement in one domain may not be paralleled by improvement in the other domain. Meta-analytic data were used to assess the association of subjective and objective treatment outcomes by drug class.
Methods:
PubMed searches, from 1967-June 2016, yielded 327 possible articles of RCTs evaluating BZs, BZRAs, SADs, DORAs, or melatonin agonists (MELA). Weighted effect sizes (ES) were computed for the pre-post change data and were compared to the study placebo condition. Between measurement strategy differences were assessed using ES difference scores. Note: positive values favor sleep diary outcomes and negative values favor PSG outcomes.
Results:
Sufficient data were available for only BZRAs, SADs, DORAs, and MELA. When examining subjective-objective ES difference scores the following was observed: SL (BZRAs = 0.04; SADs = 0.02; DORAs = -0.03; MELA = -0.14); WASO (MELA = 0.03; BZRAs = -0.01; DORAs = -0.17; SADs = -0.26); and TST (BZRAs = 0.03; SADs = -0.03; DORAs = -0.10; MELA = -0.11).
Conclusion:
These preliminary results indicate that, overall, BZRAs produce the most concordant results on PSG and self-report measures (when taking into account placebo responding). The most discordant results, overall, were found with the DORAs and MELA, although the single most discordant value was for SADs on WASO. In each of these cases, PSG measures were found to show greater levels of efficacy. Given that subjective and objective treatment gains may reflect different benefits to the patient, discordance need not be viewed as better or worse efficacy. This said, medications that disproportionately affect objective measures may not be viewed favorably by patients and this may, in turn, adversely affect treatment adherence.
Support (If Any):
Partial support provided by R56AG050620 and R01AG041783.