ISPOR 2020 PODIUM: Methodological Challenges with NMAs Assessing Long-term Efficacy in Psoriasis

Virtual ISPOR 2020 | TUESDAY, May 19th, 2020 | 5:30-5:45 PM ET

OBJECTIVES: In network meta-analysis (NMA) of psoriasis trials, cross-over after an initial placebo-controlled period limits the connectivity of long-term evidence networks. We illustrate the challenges with conducting NMAs assessing efficacy beyond cross-over in psoriasis by critically appraising the assumptions underpinning recent long-term NMAs.

METHODS: We compared three recent NMA studies (Armstrong et al. 2019, Sawyer et al. 2018, and Diels et al. 2017) assessing long-term (beyond 16 to 24 weeks) comparative efficacy in psoriasis and investigated how the three studies derived comparative efficacy estimates in the absence of a connected network. While each of the review methods address a slightly different research question, we assumed the underlying goal was to approximate the results from a standard connected network. The purpose of the study was therefore to assess the assumptions required for each approach that would provide unbiased estimates of the comparative effectiveness of included agents.

RESULTS: Armstrong et al. did not analyze comparative RCT data but instead removed placebo arms with cross-over, pooled individual active arms from 23 RCTs, and conducted naïve indirect comparisons between pooled arms (which assumes that all trials are equivalent in terms of prognostic factors, an unlikely assumption). Sawyer et al. analyzed a largely disconnected network with four comparative RCTs, as well as a larger network that was connected through 17 comparisons that assumed placebo response rates observed during induction period were maintained to week 52. Diels et al. only included comparative RCTs in their NMA and constructed a network of 11 RCTs by assuming a class-effect among TNFα inhibitors (i.e., all TNFα inhibitors have similar efficacy).

CONCLUSIONS: Recent analyses assessing long-term comparative efficacy in psoriasis have methodological limitations and should be interpreted with caution. Future studies should consider the assumptions underpinning NMAs and assess how alternative NMA methods can be leveraged to yield more rigorous long-term indirect treatment comparisons.

Fill out the form below and download the presentation. Schedule a meeting to talk with our experts.


Tim Disher, RN, PhD(c)
Associate Director, Data Analytics and Evidence Synthesis

Tim is an experienced analyst and emerging thought leader in the application of Bayesian methods to complex problems. His Vanier Canada Graduate Funded dissertation research focused on the  incorporation of multivariate evidence synthesis in…

Chris Cameron, PhD
Senior Vice President, Value and Evidence

Dr. Chris Cameron is a global thought leader in health economics and outcomes research with over a decade of experience. Prior to joining EVERSANA, Chris was a partner at Cornerstone Research Group Inc., and…