To construct a prediction model for social frailty in elderly patients undergoing maintenance hemodialysis.To explore the predictive value of modifiable factors such as dialysis symptom burden,family function,depression,and physical activity for the model.

By using the convenience sampling method,451 elderly patients undergoing maintenance hemodialysis in the hemodialysis room of a tertiary grade A hospital in Chengde City from November 2024 to April 2025 were selected as the research subjects for the construction group.Another 181 elderly patients undergoing maintenance hemodialysis in the hemodialysis rooms of two tertiary grade A hospitals from May to July 2025 were selected as the external validation group for the research.A cross-sectional survey was conducted using general information questionnaires,5-Item Geriatric Depression Scale(GDS-5),Dialysis Symptom Index(DSI),Brief Aging Perceptions Questionnaire(B-APQ),Family APGAR Index(APGAR),and the Score of the Chinese Version of the International Physical Activity Questionnaire-Short Form.Logistic regression analysis was used to screen the influencing factors of social frailty.The ROC curve was drawn using R language and internal and external validations were conducted.The distinguishability,calibration,and clinical practicability of the model were evaluated.Based on the predictive factors,the basic model and the full model were constructed.The predictive value of the modifiable factors was quantified.

The incidence rates of social frailty among elderly maintenance hemodialysis patients in the construction group and the validation group were 33.7% and 35.9% respectively.Logistic results showed that dialysis duration,burden of dialysis symptoms,family function,physical activity,and depression were the influencing factors of social frailty in elderly maintenance hemodialysis patients.Based on the above five factors,the area under the receiver operating characteristic curve of the Nomogram for internal and external validation was 0.888 and 0.876 respectively.The Hosmer-Lemeshow test results showed that both P>0.05.The calibration curves and decision curves of internal and external validation both showed that the model had good calibration and high clinical practicability.The modifiable factors significantly contributed to the predictive value of the model.

The constructed prediction model for social frailty in elderly patients undergoing maintenance hemodialysis has good predictive efficacy and clinical applicability.It could provide a tool for the early identification of social frailty in elderly maintenance hemodialysis patients.Medical staff could formulate corresponding intervention measures based on factors that can be intervened,such as the burden of dialysis symptoms,family function,depression,and physical activity,so as to delay the occurrence and development of social frailty in elderly patients undergoing maintenance hemodialysis.