RESUMO
Sleepiness is one of the outcomes most used in randomized clinical trials (RCT) on the effect of treatments for obstructive sleep apnea (OSA). Furthermore, it is known that there is a placebo effect, especially in subjective measures. Therefore, given that sleepiness is a subjective measure, the objective of this systematic review with meta-analysis and three-level meta-regression was to assess the response to different placebos (pills and sham-CPAP) used in RCTs in OSA, both on subjective (Epworth Sleepiness Scale [ESS]) and objective (Multiple Sleep Latency Test [MSLT], Maintenance Wake Test [MWT], the Osler test and the Psychomotor Vigilance Task [PVT]). We observed a statistically significant placebo effect in both subjective and objective measures of hypersomnia, and in both sham-CPAP and pills. This placebo effect was greater, even clinically significant, in subjective measures (ESS: -2.84 points) and in those RCTs that used pills as a placebo. In the meta-regression, only a higher baseline value of the ESS and Osler test was related to the placebo effect.