MATCH-IT: A decision support tool for multiple comparisons presenting data from network meta-analysis to facilitate guideline development

Abstract

Introduction

Network meta-analysis (NMA) provides unprecedented opportunities to compare multiple treatment options (multiple comparisons) and are increasingly being used to inform clinical practice guidelines. However, the overwhelming amount of data generated from NMAs is challenging to present in comprehensible overviews for end-users, guideline panelists included. Acknowledging these challenges, we developed MATCH-IT—an interactive evidence summary displaying NMA results for multiple comparisons. In this study, we conducted user-testing and further developed MATCH-IT to support guideline panels in moving from NMA evidence to recommendations.

Methods

We user-tested the tool with guideline panelists and observed the use of the tool in panel meetings. User-testing sessions and guideline meetings were recorded, transcribed, and analyzed. The analysis informed the iterative development in the tool.

Results

We included four guideline panels and tested the tool with 15 panelists (four chairs, four methodologists, five clinical experts and two patient partners). User testing revealed both positive aspects and limitations of the tool. Interactivity allowed for dynamic display of the evidence during panels meetings and was highlighted as valuable. Further, participants felt that the tool provided overview of complex evidence, further facilitated by categorization of effects through colour coding. The inclusion of information on burden of treatment was highlighted as relevant and valuable. Regarding limitations, some users had issues discovering the interactive features. Earlier versions of MATCH-IT did not include sufficiently detailed information, such as the imprecision of effect estimates, which the users felt was needed for decision making. These findings led to refinements of the tool, including a new tutorial, inclusion of confidence intervals, and a new layer displaying more detailed information.

Discussion

Our study suggests that MATCH-IT may play a role in facilitating guideline development by easing understanding of NMA evidence and alleviating information overload in guideline panelists.