Multicriteria benefit–risk assessment using network meta-analysis

Abstract

The designed method enables quantitative BR analysis of alternative treatments using all available evidence from a network of clinical trials. The preference-free analysis can be useful in presenting the results of an MTC considering multiple outcomes.

Evidence-Based Prescribing

“…To enable a meaningful comparison between the outcomes, we calculated absolute risks by multiplying the odds ratios obtained from network meta-analysis with the average odds of events across the control arms of included trials, thereby placing all outcomes on the same scale. 30,34 We then applied a structured benefit-risk model that allows evidence on multiple outcomes to be combined using qualitative preference statements. 25,30,35 This benefit-risk model took into account the probability distributions of all outcomes of interest and quantified the uncertainty around a decision, while keeping outcome measurements and value judgments clearly separated.…”

“…30,34 We then applied a structured benefit-risk model that allows evidence on multiple outcomes to be combined using qualitative preference statements. 25,30,35 This benefit-risk model took into account the probability distributions of all outcomes of interest and quantified the uncertainty around a decision, while keeping outcome measurements and value judgments clearly separated. 25 Specifically, we sampled from the posteriors for the absolute risk on each outcome, which were translated to a partial use between 0 and 1 (where 1 was best possible and 0 the worst possible value) for all alternative treatments and for all outcomes.…”

“…30 To illustrate the promise of this combined approach, we revisited the recent systematic review of statins and combined information on multiple outcomes using qualitative preference statements in a proof-of-concept study (see below for the methods of combining network meta-analysis and multicriteria decision analysis). When combining the evidence on multiple outcomes, we adopted simple preference statements about the relative importance of different outcomes and considered the effect of statins on preventing mortality to be more important than either major coronary or cerebrovascular events, which were in turn more important than any one of the tolerability or harm outcomes.…”

Evidence Synthesis for Decision Making 1

“…The methods can clearly be extended further into comparisons that set benefits against side effects and analyses based on multicriteria decision analysis. 9 …”

JSMAA: open source software for SMAA computations

“…For normal and logit-normal distributed measurements also, a relative version of the distribution is available that contains an alternative-specific part and a baseline common for all alternatives on a single criterion (van Valkenhoef, Tervonen, Zhao, et al 2012). JSMAA supports exact, imprecise linear Downloaded by [University of Boras] at 16:54 05 October 2014 constrained (interval) or ordinal preference information (weights).…”