The Placebo Effect on Symptoms, Quality of Life, and Functional Outcomes in Patients With Angina Pectoris: A Meta-analysis of Randomized Placebo-Controlled Trials.

ABSTRACT

BACKGROUND: The placebo effect is a well described phenomenon in blinded studies evaluating antianginal therapeutics, although its effect on clinical research metrics remains unknown. We conducted a systematic review and meta-analysis to quantify the effect of placebo on end points of symptoms, life quality, and functional outcomes in randomized placebo-controlled trials (RCTs) of symptomatic stable coronary artery disease. METHODS: We systematically reviewed MEDLINE, EMBASE, and the Cochrane database for double-blind RCTs of antiangina therapeutics. Patients randomized to the placebo arm were the study population. Main outcomes were the changes in exercise performance (exercise treadmill test [ETT] parameters), quality of life (Seattle Angina Questionnaire domains), symptoms (Canadian Cardiovascular Society angina class) and drug usage (nitroglycerin tablets per week) between baseline and after placebo treatment. The primary outcome was ETT total duration time. Data were pooled with a random effect model. RESULTS: Seventy-eight RCTs (83% drug-controlled, 17% procedure-controlled) were included encompassing 4925 patients randomized to placebo. ETT total duration time was significantly improved after placebo treatment compared with baseline (mean, 29.2; 95% confidence interval, 20.6-37.8] seconds) with evidence of high heterogeneity (I2 = 98%) At subgroup analysis, crossover design was associated with a smaller placebo effect on ETT performance than parallel study design (P for interaction = 0.001). A significant placebo effect was observed for all secondary outcomes with overall high heterogeneity. CONCLUSIONS: A substantial placebo effect was present in angina RCTs across a variety of functional and life quality metrics. High variability in placebo effect size was present, mostly unexplained by differences in study and patient characteristics.