Back to Publications
Prioritizing Healthcare Interventions: A Comparison of Multicriteria Decision Analysis and Cost-Effectiveness Analysis
Value in Health
PUBLISHED
29 January 2022
CITATION
Wilson R, Chua J, Pryymachenko Y, Pathak A, Sharma S, Abbott JH. Prioritizing Healthcare Interventions: A Comparison of Multicriteria Decision Analysis and Cost-Effectiveness Analysis. Value in Health 2022;25(2):268-275. doi:10.1016/j.jval.2021.08.008
Abstract
Objectives To investigate the extent to which stated preferences for treatment criteria elicited using multicriteria decision analysis (MCDA) methods are consistent with the trade-offs (implicitly) applied in cost-effectiveness analysis (CEA), and the impact of any differences on the prioritization of treatments.
Methods We used existing MCDA and CEA models developed to evaluate interventions for knee osteoarthritis in the New Zealand population. We established equivalent input parameters for each model, for the criteria “treatment effectiveness,” “cost,” “risk of serious harms,” and “risk of mild-to-moderate harms” across a comprehensive range of (hypothetical) interventions to produce a complete ranking of interventions from each model. We evaluated the consistency of these rankings between the 2 models and investigated any systematic differences between the (implied) weight placed on each criterion in determining rankings.
Results There was an overall moderate-to-strong correlation in intervention rankings between the MCDA and CEA models (Spearman correlation coefficient = 0.51). Nevertheless, there were systematic differences in the evaluation of trade-offs between intervention attributes and the resulting weights placed on each criterion. The CEA model placed lower weights on risks of harm and much greater weight on cost (at all accepted levels of willingness-to-pay per quality-adjusted life-year than did respondents to the MCDA survey.
Conclusions MCDA and CEA approaches to inform intervention prioritization may give systematically different results, even when considering the same criteria and input data. These differences should be considered when designing and interpreting such studies to inform treatment prioritization decisions.