Topical Rosacea Therapies Compared in Network Meta-Analysis

Key Takeaways

• Ivermectin and encapsulated benzoyl peroxide were associated with greater lesion-count reduction and higher likelihood of Investigator Global Assessment success than metronidazole.
• Discontinuation due to adverse events or effects was broadly similar across treatments overall, although encapsulated benzoyl peroxide was associated with more discontinuations than metronidazole.
• Patient-reported outcomes and erythema had too little data for robust quantitative synthesis, and longer-term efficacy, tolerability, and standardized patient-reported outcomes remained areas for future trials.

In a network meta-analysis of 32 randomized trials involving 11,399 adults with moderate to severe rosacea, ivermectin and encapsulated benzoyl peroxide were associated with larger reductions in lesion count than metronidazole across 10 topical interventions. Investigator Global Assessment success was also higher with those two agents than with metronidazole. The analysis compared relative efficacy and tolerability across the topical treatments included in the network.

This systematic review and network meta-analysis evaluated randomized clinical trials in adults with moderate to severe rosacea across 10 topical interventions. Prespecified primary outcomes were mean change in absolute lesion count, Investigator Global Assessment success, and discontinuation due to adverse events or effects. Random-effects network meta-analyses were used alongside SUCRA rankings and GRADE certainty assessment, with risk of bias assessed using Cochrane RoB 2.0. Searches covered CENTRAL, MEDLINE via Ovid, Web of Science, and Embase from inception through August 6, 2025, and most studies followed participants for 8 to 16 weeks.

Within the network, ivermectin was associated with a greater reduction in lesion count than metronidazole, with an MD of 4.17 and a 95% CI of 1.85 to 6.48. Encapsulated benzoyl peroxide showed a similar association for lesion count, with an MD of 4.14 and a 95% CI of 0.62 to 7.66 versus metronidazole. For Investigator Global Assessment success, ivermectin was associated with an MD of 10.31 and a 95% CI of 2.85 to 17.77, while encapsulated benzoyl peroxide had an MD of 15.51 and a 95% CI of 2.35 to 28.68.

Discontinuation due to adverse events or effects was similar between treatments overall in the network meta-analysis. Encapsulated benzoyl peroxide, however, was associated with more discontinuations than metronidazole, with an MD of 8.33 and a 95% CI of 0.45 to 16.22. Data on patient-reported outcomes and erythema were too limited for robust quantitative synthesis. Longer-term efficacy, tolerability, and standardized patient-reported outcomes remained areas for future trials.