Placebo-controlled assessment of somnolence effect of cetirizine: a meta-analysis

Background: It has been found that the most common adverse reaction which occurs more frequently on cetirizine than on placebo is somnolence. However, the somnolence rate varied widely among different studies. The objective of this study was to assess the somnolence effect of cetirizine 10 mg daily compared to placebo in patients aged 6 years and older using meta-analysis and explore the sources of heterogeneity among different studies. Methods: Databases were searched for randomized controlled trials (RCTs) of cetirizine published before 2015. Overall risk differences (RDs) were determined by meta-analyses of 13 trials using the DerSimonian and Laird (D&L) method based on fixed-effects and random-effect models, respectively. The Q statistic, H statistic, and I² were calculated for heterogeneity analysis. Subgroup analysis, Galbraith plot, sensitivity analysis, and meta-regression were also performed to explore the sources of heterogeneity. Results: Various analyses showed that heterogeneity existed among the 13 trials and the placebo run-in period was the cause of heterogeneity. For RCTs without and with placebo run-in period, the overall RDs were 6.51% (95% CI, 4.47% to 8.56%) and 1.03% (95% CI, −0.13% to 2.19%), respectively, which indicated that the difference in somnolence rate between cetirizine 10 mg daily and placebo was not statistically significant for the subgroup with placebo run-in. Conclusion: The meta-analysis showed that the RDs between cetirizine 10 mg daily and placebo on somnolence rate were different for studies with and without a placebo run-in period. The results for studies with a placebo run-in period suggested that cetirizine 10 mg daily has no somnolence effect compared to placebo.