Abstract: Objective: To evaluate the incidence of falls in middle-aged and elderly (≥45 years old) patients with chronic kidney disease (CKD) in China and build a prediction model.
Methodology: Based on the database of China Health and Retirement Longitudinal Study (CHARLS), the end result is a fall within 2 years. The influence of CKD on falls was analyzed by propensity score matching (PSM) and logistic regression. According to the ratio of 7∶3, CKD patients were randomly divided into training sets or verification sets, and logistic regression was used to explore the independent influencing factors of their falls and build a prediction model. Receiver operating characteristic curve (ROC), area under curve (AUC), calibration curve and decision curve analysis (DCA) were used to evaluate the discrimination, accuracy, and clinical application value of the model.
Results: A total of 7352 people were included. Compared to non-CKD patients, CKD patients had a higher fall rate (24.47% vs 15.50%, P<0.05) and a higher risk of falling (OR=1.62, 95% CI 1.21~2.17, P=0.001). Activities of daily living and depression were independent influencing factors of falls. ROC curve shows that the AUC values of the training set and verification set are 0.67 (95% CI 0.61~0.73) and 0.68 (95% CI 0.60~0.77), respectively, indicating that the model has a certain degree of discrimination. The calibration curve indicates that the model has a good calibration degree. DCA curve display has potential for clinical application value.
Conclusion: Middle-aged and elderly patients with CKD are more likely to fall. The prediction model constructed in this study has certain prediction efficiency and clinical application value, which is helpful for doctors to formulate better treatment and prevention strategies, reducing patient mortality, and improving quality of life.

Key words: font-family:Inter, -apple-system, BlinkMacSystemFont, ", font-size:16px, background-color:#F9FAFB, ">China, middle-aged and elderly, chronic kidney disease, fall, predictive models