Systematic review and meta‐analysis to predict mortality in heart failure with preserved ejection fraction: Development and validation of the HF‐DANAS score

Abstract

AbstractAimsThe morbidity and mortality of heart failure with preserved ejection fraction (HFpEF) continue to increase with the accelerating global aging process. During the past decade, the pathophysiology, diagnostic methods, and prognostic prediction of HFpEF have been revolutionized, resulting in new and effective management strategies. Dynamic prognostic assessment facilitates systematic clinical management of patients, and the aim of this study was to investigate the risk factors for mortality in patients with HFpEF and to develop a risk prediction assessment model.Methods and reultsData for the derivation cohort were obtained from three databases, PubMed, Embase, and Cochrane. The validation cohort was obtained from the Chinese Heart Failure Center database. The β‐coefficient was calculated based on the risk ratio (RR) and 95% confidence intervals (CI) corresponding to each risk factor to construct a mortality risk assessment model. A total of 30 studies were included in the meta‐analysis: 22 prospective cohort studies and 8 retrospective cohort studies, including 34 196 HFpEF patients. Seven predictors of all‐cause mortality in HFpEF patients were derived. Considering the need for feasibility in clinical practice, we performed subgroup and sensitivity analyses and determined the following cutoff values: age > 75 years (RR: 2.07, 95% CI: 1.83–2.35; P < 0.001), male sex (RR: 1.36, 95% CI: 1.17–1.59; P < 0.001), DM (RR: 1.23, 95% CI: 1.11–1.36; P < 0.001), anaemia (RR: 1.53, 95% CI: 1.41–1.67; P < 0.001), albumin concentration < 3.2 g/dL (RR: 1.29, 95% CI: 1.14–1.47; P < 0.001), AF (RR: 1.27, 95% CI: 1.12–1.43; P < 0.001), and NYHA class III/IV (RR: 1.63, 95% CI: 1.43–1.87; P < 0.001). The area under the receiver operating characteristic (ROC) curve (AUC) for this model was 71.3% (95% CI: 0.696–0.736), with an optimal cut‐off value of 10.75. The sensitivity and specificity were 0.778 and 0.566, respectively. According to this risk score, we divided patients into three risk classes (low, moderate, and high risk), the numbers of patients who died by the end of the 1‐year follow‐up were 23 (1.87%), 82 (5.62%), and 382 (15.52%) in these three groups, and the 5‐year mortality rates were 9.82%, 20.68%, and 43.28%, respectively.ConclusionsThis study developed an HF‐DANAS scoring system for the HFpEF mortality risk containing seven predictors, providing clinicians with a simple assessment tool that can help improve clinical management.