Traditional meta-analyses synthesize aggregate data obtained from study publications or study authors, such as a treatment effect estimate and its associated uncertainty. An increasingly important approach is the meta-analysis of individual participant data (IPD) where the raw individual-level data are obtained for each study and used for synthesis. This study compares and discusses results from an IPD meta-analysis vs standard meta-analysis of randomized controlled trials of exercise cardiac rehabilitation in chronic heart failure (CHF).

Based on a previous systematic review, the Exercise Training Meta-Analysis of Trials for Chronic Heart Failure (ExTraMATCH II) identified and collected IPD from randomized controlled trials (RCTs) that compared exercise rehabilitation with a non-exercise control with a minimum follow-up of six months. For this abstract, the outcome of interest was all-cause mortality. Original IPD were checked for consistency and compiled in a master dataset. Standard meta-analytic models were used for aggregate data whilst two-stage and one-stage approaches, accounting for the clustering of participants within studies, were planned for statistical analyses of IPD.

Overall thirty-three RCTs were included in the original systematic review, whereas within the ExTraMatch II project, IPD on all-cause mortality were obtained from seventeen RCTs of approximately 3,700 patients. From aggregate data there was no significant difference in pooled mortality (relative risk 0.92, 95% confidence interval 0.67 to 1.26). IPD analysis revealed 701 events across exercise and control groups. Our ongoing IPD analyses will allow us to examine how patients’ characteristics (e.g. age, New York Heart Association functional class, ejection fraction) modify treatment benefit.

Given the limitations of current trial level meta-analysis evidence in CHF, access to individual data from several RCTs offers a timely and important opportunity to revisit the question of which CHF patient subgroups benefit most from exercise-based rehabilitation.