RDW-to-Albumin Ratio as a Simple Biomarker for Early Mortality Risk After LVAD Implantation

Background and Objectives: Early risk stratification remains challenging in patients undergoing left ventricular assist device (LVAD) implantation. Red cell distribution width (RDW) and serum albumin reflect systemic stress and nutritional reserve; their ratio (RDW-to-albumin ratio, RAR) may provide a simple preoperative index. We evaluated whether preoperative RAR is associated with early mortality after LVAD implantation. Materials and Methods: We conducted a retrospective cohort study of LVAD recipients (2019–2025). RAR was calculated as RDW (%) divided by albumin (g/dL) from preoperative blood tests obtained 24–48 h before surgery. The primary endpoint was in-hospital mortality. The secondary endpoint was 90-day survival. In-hospital mortality was analyzed using logistic regression with parsimonious adjustment for INTERMACS high-risk status (profiles 1–2 vs. 3–7); penalized regression was used to reduce small-sample bias. Discrimination was assessed using receiver operating characteristic (ROC) analysis. Ninety-day survival was evaluated using Cox proportional hazards models. Results: Forty-seven patients were included (37 survivors; 10 in-hospital deaths). Higher RAR was associated with increased odds of in-hospital mortality and remained significant after adjustment for INTERMACS high-risk status (OR 1.68, 95% CI 1.04–2.90). INTERMACS high-risk status was strongly associated with in-hospital mortality (OR 17.89, 95% CI 3.19–138.07). RAR demonstrated good discrimination for in-hospital mortality (AUC 0.801, 95% CI 0.648–0.955). For 90-day survival, RAR showed a borderline association in unadjusted analysis (HR 1.28, 95% CI 0.98–1.68) and was not significant after adjustment (HR 1.20, 95% CI 0.89–1.63). Conclusions: In this small single-center cohort, preoperative RAR was independently associated with in-hospital mortality after LVAD implantation. These findings should be considered hypothesis-generating and require external validation.