A nomogram for predicting the risk of peritoneal dialysis-associated peritonitis in patients with end-stage renal disease undergoing peritoneal dialysis: model development and validation study

Abstract

Objective: This study aimed to develop and validate a nomogram to predict the risk of peritoneal dialysis-associated peritonitis (PDAP) in patients undergoing peritopreneal dialysis.

Methods: A retrospective analysis was conducted on clinical data from 376 patients at Nanhai District People's Hospital in Foshan City, Guangdong Province, between December 2017 and December 2024. The dataset was randomly divided into a training set (n = 244) and a validation set (n = 132). Risk factors for PDAP were identified using Least Absolute Shrinkage and Selection Operator (LASSO) regression and logistic regression, and a predictive nomogram was developed and validated using R4.1.3. The model's performance was evaluated through receiver operating characteristic (ROC) curves, the Hosmer-Lemeshow goodness-of-fit test, decision curve analysis (DCA), and clinical impact curves (CICs).

Results: Eight potential predictors were selected by LASSO regression analysis. Multivariate logistic regression analysis confirmed that age, dialysis duration, albumin, hemoglobin, β₂-microglobulin, Potassium and lymphocyte count were independent risk factors for PDAP occurrence (P = 0.001). The nomogram's area under the curve (AUC) was 0.929 (95% CI: 0.896-0.962) in the training set and 0.905 (95% CI: 0.855-0.955) in the validation set. The Hosmer-Lemeshow goodness-of-fit test indicated a good model fit (training set χ² = 13.181, P = 0.106; validation set χ² = 8.264, P = 0.408). Both DCA and CIC revealed that the nomogram model had good clinical utility in predicting PDAP.

Conclusion: The proposed nomogram exhibited excellent predictive performance and clinical utility, providing a valuable tool for early identification and intervention in PDAP. Further external validation and prospective studies are recommended.

Keywords: Clinical prediction model; Dialysis-associated peritonitisc; End-stage renal disease; Nomogram; Peritoneal dialysis.

Fig. 1

Least absolute shrinkage and selection operator (LASSO) model feature selection. a: LASSO regression was used to analyze 19 general data and nutrition-related indicators of patients undergoing peritoneal dialysis. b: Bias of the LASSO cross validation

Fig. 2

Nomogram risk prediction model for PDAP in peritoneal dialysis patients. Abbreviations: PDAP:Peritoneal dialysis-associated peritonitis,HGB: Hemoglobin, ALB: Albumin. K: Serum potassium Na: Serum sodium, LYM: Lymphocyte count

Fig. 3

ROC curve was used to analyze the predictive value of the prediction model for the occurrence of PDAP. a: Training set. b: Validation set

Fig. 4

Calibration curve of PDAP prediction model. a: Training set. b: Validation set., X-axis represents the predicted risk of PDAP occurrence, while the Y-axis represents the actual occurrence. The diagonal dotted line represents perfect prediction

Fig. 5

Clinical decision curve analysis(DCA) of PDAP prediction model. a: Training set. b: Validation set. The X-axis represents the threshold probability and the Y-axis represents the net benefit expressed as a ratio. The red line indicates the net benefit of the therapeutic intervention in patients with PDAP; The gray line is the net benefit of treatment interventions for all, based on the statistical model; The black line is the net benefit of no treatment intervention for all

Fig. 6

Clinical impact curve (CIC) of PDAP prediction model. a: Training set. b: Validation set.The red curve in the CIC represents the number of people classified as positive (high risk) by the model under each threshold probability; the blue curve represents the number of true positives at each threshold probability