The Impact of Frailty on the Survival Prognosis of Maintenance Hemodialysis Patients and the Construction and Validation of a Survival Prediction Model

Objective: This study aims to examine the impact of frailty on the survival outcomes of patients undergoing maintenance hemodialysis (MHD) and to develop a predictive model for mortality risk. Methods: In this prospective cohort study, 400 MHD patients from Yidu Central Hospital of Weifang in March 2023 were enrolled.Patient data were collected through a questionnaire survey. Frailty status was determined according to the Fried phenotype frailty score, with patients categorized into the frailty group (≥3 points) and the non-frailty group (<3 points). Depression was assessed using the PHQ-9 scale, and anxiety was evaluated using the GAD-7 scale. Patients were randomly assigned to a training set (n = 280) and a validation set (n = 120) in a 7:3 ratio using R software. Kaplan–Meier survival curves were generated for frail and non-frail patients, and survival differences were compared with the log-rank test. Independent predictors of mortality were identified using the LASSO-Cox proportional hazards regression model, which was further applied to construct a mortality risk prediction model. The model’s performance was evaluated using the concordance index (C-index), calibration curve, and decision curve analysis (DCA). Results: The incidence of frailty among MHD patients was 45.75%. The mortality rate in the frailty group was 30.17%, significantly higher than 11.76% in the non-frailty group. Kaplan–Meier curves demonstrated a significant survival difference between groups ( P < 0.001). Multivariable analysis revealed frailty (HR = 1.854, 95% CI: 1.025–3.355), age (HR = 1.041, 95% CI: 1.013–1.070), depression (HR = 4.906, 95% CI: 2.000–12.035), anxiety (HR = 3.486, 95% CI: 1.778–6.831), cardiovascular disease (CVD, HR = 2.063, 95% CI: 1.126–3.781), serum creatinine (Cr, HR = 1.004, 95% CI: 1.003–1.005), and total cholesterol (TC, HR = 1.503, 95% CI: 1.133–1.995) as independent risk factors for mortality (all P < 0.05). The predictive model demonstrated a C-index of 0.903. In the validation cohort, the areas under the ROC curve at 6 months, 1 year, and 2 years were 0.889, 0.897, and 0.941, respectively. The calibration curve showed good agreement between predicted and observed outcomes, and DCA confirmed its clinical utility. Conclusion: Frailty is highly prevalent among MHD patients and represents an independent risk factor for all-cause mortality. A nomogram incorporating seven independent predictors provides accurate mortality risk estimates at 6 months, 1 year, and 2 years. This tool may facilitate the early identification of high-risk patients in clinical practice.