Meta-analysis and Indirect Comparisons: on Methods, Paradigms, and Biologic Treatments for Psoriasis

Meta-analysis offers a way to assess the clinical efficacy of a treatment by combining the results of randomized clinical trials while maintaining randomization; the combined effects, with their confidence intervals, can be represented with a forest plot. The efficacy of several different treatment options can be assessed with either direct or indirect comparisons. Indirect comparisons may be placebo-anchored as well as network meta-analyses (NMA) that use either a frequentist or Bayesian approach, depending on the statistical framework and the definition of probability selected. Indirect comparisons may also adjust for covariates or utilize individual participant data. Bayesian NMA are able to establish a rank order of efficacy based on probabilities or credibility intervals, which can be described by the surface under the cumulative ranking curve(SUCRA). Statistical superiority is demonstrated by pairwise comparisons, which are generally presented in league tables. This review provides clinical practitioners with detailed descriptions of these methods, drawing on examples from recently published NMA that rank the relative efficacy of biologic treatments for moderate to severe psoriasis. According to NMA findings, the four most effective treatments in both the short term (10-16 weeks) and the long term (approximately one year) are, in rank order, risankizumab (first in all studies that include it), brodalumab, guselkumab, and ixekizumab. However, the between-treatment differences are not always significant.