Selection of Characteristics for Matching Adjusted Indirect Comparison in a Functional Mitral Regurgitation Population

OBJECTIVES: Indirect comparisons are a newer methodology approach within the medical device space. We compared baseline characteristics between patients with Functional Mitral Regurgitation (FMR) etiology from the CLASP and COAPT trials for weighting using the Matching Adjusted Indirect Comparison (MAIC) technique. The aim of this study was to understand the most appropriate covariates for weighting in future planned FMR analyses.

METHODS: Data from the FMR subset in the CLASP trial was used to compare to aggregated patient characteristics and outcomes from the COAPT trial. Clinicians with expertise in the disease area were consulted to determine which baseline characteristics were most appropriate for weighting in FMR patients. Clinicians pre-specified the weighting criteria for base case and sensitivity analysis scenarios. In the base case, weighting was conducted on prior myocardial infarction (MI), chronic obstructive pulmonary disease (COPD), stroke/transient ischemic attack (CVA/TIA), ischemic cardiomyopathy, and mean left ventricular ejection fraction (LVEF). In the sensitivity analysis, weighting was conducted on mitral regurgitation severity, prior MI, mean LVEF, and atrial fibrillation or flutter. We expected a larger Effect Sample Size (ESS) after weighting might indicate a higher sampling quality.

RESULTS: In the base case, the ESS for the CLASP FMR group decreased by approximately 40% (from n=85 to ESS=50.7). The sensitivity analysis led to a substantially smaller matched sample (ESS=38.6) and non-matched characteristics became more dissimilar to the COAPT population including variables with clinical importance for the population.

CONCLUSIONS: A higher ESS in the base case compared to the sensitivity analysis gave us a positive forward direction for future planned indirect comparisons in the FMR population.