Abstract: 【Abstract】Objective To analyze the risk factors for frailty in maintenance hemodialysis (MHD) patients, and to construct a prediction model for the risk of frailty so as to provide a reference for the prevention and alleviation of frailty in dialysis patients. Methods Convenient sampling was used to select a total of 145 MHD patients treated in the Blood Purification Clinic of our hospital. Logistic regression was used to construct a risk prediction model. The Hosmer-Lemeshow chi-square test was used to evaluate performance of the model. For the degree of fit, the area under the ROC curve was used to verify the predictive effect of the model. Results This study finally included 6 factors, namely gender (OR=7.385, 95% CI: 4.965～56.529, P=0.045), living style (OR=4.823, 95% CI: 1.138～20.446, P=0.033), nutritional score (OR=0.453, 95% CI: 0.255～0.807, P=0.007), hemoglobin (OR=0.146, 95% CI: 0.015～1.392, P=0.030), Charlson comorbidity index (OR=5.918, 95% CI: 0.465～75.240, P=0.012), and self-care ability score (OR=0.589, 95% CI: 0.551～1.142, P=0.032), to construct a risk prediction model. The results of Hosmer- Lemeshow chi- square test showed a better degree of fit (χ2=6.889, P=0.549) of the prediction model; the AUC under the ROC curve of the prediction model was 0.940 (P<0.001, 95% CI: 0.886～0.973), with the sensitivity of 86.4% and the specificity of 86.0%. Conclusion This prediction model has a better degree of fit and a better prediction effect. It is useful for medical staff to predict frailty in MHD patients earlier and to provide references in planning specific intervention measures.

Key words: Hemodialysis, Frailty, Risk prediction model