Association between red blood cell distribution width-platelet ratio (RPR) and mortality in patients with heart failure from the MIMIC IV database: A retrospective cohort study

The association between the red blood cell distribution width-platelet ratio (RPR) and mortality in heart failure patients remains unclear. We aimed to investigate the potential non-linear relationship between RPR and 1-year mortality risk. A retrospective cohort study was conducted involving 6982 participants from the Medical Information Mart for Intensive Care IV (MIMIC-IV) database. Multivariable Cox regression and restricted cubic spline analyses were performed to evaluate the association between RPR and 1-year mortality, adjusting for potential confounders. We observed 1091 patients died in hospital and 2535 patients died during 1 year follow-up period. The prevalence or incidence of mortality did not show statistically significant differences among RPR groups in the overall study population. However, a positive association between RPR and the risk of mortality was noted after adjusting for multiple variables (HR = 1.38, 95 % CI = 1.06-1.81, P = 0.018). Analysis using restricted cubic splines indicated a U-shaped relationship between RPR levels and the risk of mortality (P nonlinearity <0.05), with the point of lowest risk at 0.104. Compared to this level, lower RPR (<0.104) was associated with increased mortality (HR = 0.046, 95 % CI: 0.004-0.546), as was higher RPR (>0.104) (HR = 2.656, 95 % CI: 1.692-4.170).This U-shaped association was consistent across subgroup analyses (all interaction P values > 0.05). RPR exhibits a U-shaped association with 1-year mortality in heart failure patients, suggesting both low and high RPR levels are linked to increased risk. RPR may serve as a relevant biomarker for risk stratification in this population. We incorporated RPR into the SOFA (AUC 0.731) and SAPS II (AUC 0.746) models, which significantly improved their predictive ability for in-hospital mortality. For 1-year mortality prediction, RPR + SAPS II (AUC 0.683) showed significantly improved accuracy, while RPR + SOFA (AUC 0.626) did not improve significantly.