Exercise training for chronic heart failure (ExTraMATCH II): Why all data are not equal

We, the ExTraMATCH II collaborative, recently completed an individual patient data (IPD) meta-analysis on the effect of exercise-based cardiac rehabilitation (ExCR) on mortality and hospitalisation in patients with heart failure. This analysis considered 18 randomised controlled trials, which included 3912 people with heart failure who were randomised to an ExCR intervention or control group (treatment as usual). Unlike the original ExTraMATCH IPD meta-analysis published in 2004 by Piepoli et al., this latest analysis did not demonstrate mortality and hospitalisation benefits from ExCR in patients with heart failure. We believe that some of the reasons for this difference in findings could be attributed to the data provided by the individual trials. We share here some of the data problems we encountered and provide our insights. In doing so, we also propose some recommendations for both future study designs and funding bodies, so that we may move towards a more efficacious approach to managing heart failure with ExCR in the future.