Development and Validation of a nomogram for Predicting Catheter Dysfunction in Patients Receiving Continuous Renal Replacement Therapy in Intensive Care Units: an observational Cohort Study

Abstract BackgroundContinuous renal replacement therapy (CRRT) has become an essential treatment modality for renal replacement in the intensive care unit (ICU). As the primary vascular access for CRRT, hemodialysis catheters benefit patients significantly, yet inherently pose risks of complications like catheter dysfunction. This study aimed to develop and validate a predictive model for catheter dysfunction that can provide clinicians with early signs.MethodsThe study was conducted using convenience sampling to select 300 patients from a tertiary Grade A hospital in China between October 2022 and November 2024. These patients, who were receiving CRRT in the ICU, were using a hemodialysis catheter. Patients were grouped by hemodialysis catheter dysfunction occurrence, with variables first screened via univariate analysis, followed by multivariate logistic regression to identify independent predictors of catheter dysfunction. With these predictors, a nomogram was constructed, and its predictive capability was assessed via the area under the curve (AUC), calibration curve, Hosmer-Lemeshow test, and decision curve analysis (DCA).ResultsSix factors of hemodialysis catheter dysfunction were identified: age, history of catheter placement, mechanical ventilation, red blood cell count, pre-pump pressure ratio > 0.66, and catheter retention time. The corresponding odds ratios (ORs) were 1.03, 2.89, 3.36, 1.60, 3.72, and 1.05 (all P < 0.05). Developed with six factors, the nomogram demonstrated in internal validation an AUC of 0.82 (95% CI: 0.757–0.884), a well-fitting calibration curve (Hosmer-Lemeshow χ² = 3.223, P = 0.92).ConclusionsThe developed nomogram can reliably predict the risk of hemodialysis catheter dysfunction in critically ill patients undergoing CRRT.Trial registration:Not applicable.