Ratio of red blood cell distribution width to albumin: a predictive biomarker of In-hospital mortality in heart failure patients

The ratio of red blood cell distribution width (RDW) to albumin (ALB), known as RAR, functions as an innovative indicator related to prognosis. However, whether RAR can predict the in-hospital mortality (IHM) for heart failure (HF) patients remains ambiguous. This study included HF patients derived from the Medical Information Mart for Intensive Care III (MIMIC-III) and IV (MIMIC-IV) databases. To examine the association between RAR and IHM, multiple Logistic regression models were conducted, complemented by subgroup analyses. Additionally, to ascertain the optimal threshold for RAR, restricted cubic spline (RCS) regressions were applied. In the MIMIC-III (n = 9,413) and MIMIC-IV (n = 18,685) HF cohorts, the incidence of IHM was observed in 1,639 (17.41%) and 1,175 (6.29%) participants. Following adjustment for various covariates, RAR was shown to correlate with IHM (OR, 1.45 [95% CI, 1.08-1.39]). The areas under the curves for RAR were 0.683 (MIMIC-III) and 0.710 (MIMIC-IV), indicating superior predictive value than RDW and ALB. In subgroup analysis, younger HF patients with diabetes and without atrial fibrillation or anaemia showed higher ORs than older patients without diabetes or atrial fibrillation and those with anaemia, respectively. RCS indicated the OR for RAR was non-linear with IHM, and the optimal threshold for RAR prediction was between 4.5 and 5. An elevated RAR correlates with an increased risk of IHM in HF patients. Given that RAR can be readily derived from routine laboratory tests, it holds potential as a novel biomarker for identifying high-risk HF patients.