The association between frailty and in-hospital mortality in critically ill patients with congestive heart failure: results from MIMIC-IV database

Background Frailty correlates with adverse outcomes in many cardiovascular diseases and is prevalent in individuals with heart failure (HF). The Hospital Frailty Risk Score (HFRS) offers an integrated, validated solution for frailty assessment in acute care settings, but its application in critically ill patients with congestive HF lacks exploration. This study aimed to identify the association between frailty assessed by the HFRS and in-hospital mortality in critically ill patients with congestive HF. Methods This observational study retrospectively enrolled 12,179 critically ill patients with congestive HF. Data from the Medical Information Mart for Intensive Care IV database was used. The HFRS was calculated to assess frailty. Patients were categorized into three groups: non-frailty (HFRS 5, n = 7,961), pre-frailty (5 ≤ HFRS 15, n = 3,684), and frailty (HFRS ≥ 15, n = 534). Outcomes included in-hospital mortality, length of intensive care unit stay, and length of hospital stay. Multiple logistic regression and Locally Weighted Scatterplot Smoothing (LOWESS) smoother were used to investigate the association between frailty and outcomes. Subgroup analysis was employed to elucidate the correlation between frailty levels and in-hospital mortality across diverse subgroups. Results 12,179 patients were enrolled, 6,679 (54.8%) were male, and the average age was 71.05 ± 13.94 years. The overall in-hospital mortality was 11.7%. In-hospital mortality increased with the escalation of frailty levels (non-frailty vs. pre-frailty vs. frailty: 9.7% vs. 14.8% vs. 20.2%, P 0.001). The LOWESS curve demonstrated that the HFRS was monotonically positively correlated with in-hospital mortality. Upon controlling for potential confounders, both pre-frailty and frailty statuses were found to be independently linked to a heightened risk of mortality during hospitalization (odds ratio 95% confidence interval: pre-frailty vs. non-frailty: 1.27 1.10–1.47, P = 0.001; frailty vs. non-frailty: 1.40 1.07–1.83, P = 0.015; P for trend 0.001). Significant interactions between frailty levels and in-hospital mortality were observed in the following subgroups: race, heart rate, creatinine, antiplatelet drug, diabetes, cerebrovascular disease, chronic renal disease, and sepsis. Conclusion In critically ill patients with congestive HF, frailty as assessed by the HFRS emerged as an independent predictor for the risk of in-hospital mortality. Prospective, randomized studies are required to determine whether improvement of frailty levels could improve clinical prognosis.