Renal function as a cornerstone for risk stratification in acute heart failure

Introduction Renal dysfunction strongly influences prognosis in acute heart failure (AHF). While admission eGFR is a recognized predictor of mortality, the impact of in-hospital renal variation (ΔeGFR) remains uncertain. Understanding both static and dynamic renal parameters may refine risk stratification in AHF. Purpose This study aimed to assess the relationship between eGFR, ΔeGFR, and mortality in patients admitted with AHF. Methods We conducted a retrospective observational study including 121 consecutive patients admitted to a medical ward with AHF. Clinical, laboratory, and echocardiographic data were collected from electronic medical records. Renal function was assessed using estimated glomerular filtration rate (eGFR, CKD-EPI equation) at admission and its in-hospital variation (ΔeGFR). The primary outcome was 12-month all-cause mortality; secondary endpoints included 6-month mortality and length of hospital stay. Statistical analyses included logistic regression, receiver operating characteristic (ROC) curve analysis, Mann–Whitney U test, and Spearman correlation, as appropriate. Results Among 121 patients (median age 81 [75–87] years, 56.2% male), hypertension (78.5%), dyslipidemia (63.6%), and atrial fibrillation (62.8%) were the most prevalent comorbidities, while 33.9% had chronic kidney disease. Mean admission eGFR was 77.1 ± 13.4 mL/min/1.73m², and median ΔeGFR during hospitalization was −2.4 [−5.9 - +2.5] mL/min/1.73m². In multivariable logistic regression, both admission eGFR (OR 0.88, 95% CI 0.83–0.94, p <0.001) and ΔeGFR (OR 0.88, 95% CI 0.81–0.97, p =0.008) were independent predictors of 12-month mortality, with a model accuracy of 77.7% (Nagelkerke R²=0.31). ROC analysis demonstrated a significant inverse discriminative relationship between admission eGFR and mortality (AUC=0.233, 95% CI 0.142–0.325, p <0.001), whereas ΔeGFR showed no discriminative performance (AUC=0.513, 95% CI 0.398–0.629, p =0.820). At 6 months, non-survivors had lower baseline eGFR compared to survivors (67.5 ± 6.9 vs 78.6 ± 13.6 mL/min/1.73m², p <0.001), while ΔeGFR did not differ (p =0.196). No correlation was found between admission eGFR or ΔeGFR and length of hospital stay (ρ=−0.09, p=0.31; ρ=−0.01, p=0.90, respectively). Conclusion Both admission eGFR and ΔeGFR were independently associated with 12-month mortality, but only baseline eGFR demonstrated significant discriminative ability. These findings reinforce that renal function at presentation - rather than short-term variation - plays the predominant role in determining long-term prognosis in acute heart failure, emphasizing its importance in early risk stratification.