#294 Construction and evaluation of risk prediction model for peritoneal dialysis patients with volume overload

Abstract Background and Aims This study was designed to analyze the risk factors for the occurrence of volume overload in peritoneal dialysis (PD) patients and construct a risk prediction model. Method 97 peritoneal dialysis patients attending the General Hospital of Ningxia Medical University from January 1 2019 to May 31 2020 were collected and analyzed. The influencing factors of volume overload in PD patients were initially screened, the final independent influencing factors were determined, the risk prediction model was established, and the differentiation degree, true value and predicted value of the prediction model, as well as the effect of the prediction model were evaluated. Results The incidence of volume overload in PD patients was 52.58%, with ECW, E/T, and IPTH as independent risk factors for the occurrence of volume overload in PD patients. The AUC of the modeling set was 0.896, 95% CI: 0.8301–0.9627, sensitivity was 0.843, specificity was 0.870 with moderate differentiation, and the model predictive efficacy was favorable. Conclusion The risk prediction model established in this study showed that the influencing factors for the occurrence of volume overload in PD patients were ECW, E/T, and IPTH, which can help the clinical staff to recognize the high-risk group and intervene in time.