Abstract
In some therapeutic areas, a trial may have outcome measures categorized using one or more pre-defined levels. An example includes the Psoriasis Area Severity Index (PASI), which may report one or more of PASI-50, PASI-75, PASI-90, and PASI-100 end-points. Network meta-analyses (NMAs) can be conducted to consider all PASI information simultaneously. Given it is well established that trials within psoriasis are heterogeneous, we conducted and compared multi-PASI NMA models which adjusted for placebo response to account for heterogeneity among the trials with an NMA model with no adjustment. A Bayesian NMA was conducted using an ordered multinomial (simultaneous) model with a probit link. We investigated the results and model fit statistics of three models: no adjustment; a network meta-regression with a common beta (relationship between placebo response and treatment effect) across all PASI response levels; a network meta-regression with separate betas for each PASI response level. We also conducted NMAs on individual PASI outcomes. The NMA with no adjustment for heterogeneity fit the data poorly (high deviance information criterion [DIC] and heterogeneity) due to considerable variation in patient/study characteristics across studies. The NMA with a common beta fit the data better than the NMA with no adjustment, but produced results that were inconsistent with individual PASI outcome NMAs. The NMA with separate betas fit the data best (three statistically significant betas, lower DIC and heterogeneity) and aligned with results from individual PASI outcomes NMAs. NMAs considering outcomes with multiple levels should be flexible enough to adjust for heterogeneity. They should also consider separate betas for each response level because there may be different relationships between placebo response and treatment effect across various levels of response. NMAs should also assess whether results derived from simultaneous outcome NMAs align with findings from individual outcome NMAs.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
