Evaluation of heart failure admission as a surrogate for mortality in randomized clinical trials: A meta‐analysis

Abstract Background Heart failure (HF) admission is used as a study endpoint in clinical trials. However, it remains unclear whether it can be a valid surrogate endpoint for mortality. Objectives To validate whether HF admission is a valid surrogate for mortality. Methods In PubMed and EMBASE, randomized controlled trials (RCTs) of interventions to treat patients with heart failure at the enrolment were searched on 13 April 2022. We extracted RCTs in which event numbers of both HF admission and all‐cause mortality were reported as either primary or secondary outcomes. Trial‐level correlations (R‐squared) between HF admission and mortality were assessed. We performed subgroup analyses by study year, follow‐up duration, baseline HF with reduced ejection fraction (HFrEF) or HF with preserved ejection fraction (HFpEF), and whether the intervention was pharmacological. We followed the Preferred Reporting Items for Systematic Reviews and Meta‐analyses guideline. Results A total of 117 RCTs met the criteria for inclusion. Overall, the trial‐level R‐squared between HF admission and all‐cause mortality was 0.39 (95% confidence interval (CI), 0.26 to 0.53). However, in the subgroup analyses, the trial‐level R‐squared was increased when the follow‐up duration was ≥24 months (0.70 [95% CI: 0.55, 0.85]), when intervention was pharmacological (0.51 [95% CI: 0.34, 0.68]) and when the baseline HF type was HFrEF (0.57 [95% CI: 0.42, 0.73]). Conclusions Our findings indicate that HF admission may not always be a valid surrogate for mortality in patients with HF. Rather, the surrogacy of HF admission may be dependent on clinical background and interventions.