Placebo or vehicle effects on dry eye signs and symptoms in randomized controlled trials: A systematic review and meta-analysis.

Abstract

Purpose: Dry eye disease (DED) is one of the most prevalent ophthalmic conditions. Placebo or vehicle administration is frequently used in DED trials, yet its effects remain poorly characterized. This study evaluates the presence, magnitude, and factors associated with DED vehicle or placebo effects to inform future trial design.Eligible vehicle- or placebo-controlled dry eye trials were identified using PubMed. Three authors independently extracted trial characteristics and outcome measures, including Schirmer test, tear breakup time (TBUT), conjunctival and corneal staining, and Ocular Surface Disease Index (OSDI). Random-effect models and meta-regression were used to evaluate placebo/vehicle effects and predictors. Forty-nine trials with 3529 participants in placebo/vehicle groups were included. Meta-analyses revealed significant placebo/vehicle effects on DED symptoms measured by OSDI (mean difference: -8.44; 95 % CI: -11.56 to -5.33, p < 0.01) and on signs including TBUT (0.50; 95 % CI: 0.13-0.87, p = 0.01), corneal staining (-0.55; 95 % CI: -0.90 to -0.20, p < 0.01), and conjunctival staining (-0.46; 95 % CI: -0.91 to -0.02, p = 0.04). Meta-regression revealed that a higher percentage of female participants and worse baseline OSDI scores were associated with a greater vehicle or placebo responses for OSDI (p = 0.04 and p = 0.001, respectively). No predictors were found for placebo/vehicle effects on DED signs. Placebo/vehicle effects in DED trials are substantial and should be considered in trial design. Female sex and worse baseline symptoms were associated with larger effects on OSDI. Future studies should explore mechanisms underlying these effects and approaches to mitigate their impact in DED trials.