Predicting catheter removal in peritoneal dialysis peritonitis patients visiting the emergency department: a multivariable logistic regression and decision tree analysis - Report - MDSpire
Advertisement
Predicting catheter removal in peritoneal dialysis peritonitis patients visiting the emergency department: a multivariable logistic regression and decision tree analysis
Factors Influencing Catheter Removal in Emergency PD Peritonitis Patients
Overview
This study analyzed 518 peritoneal dialysis (PD) patients presenting to the emergency department to identify factors associated with PD catheter removal during hospitalization. Using multivariable logistic regression and decision tree analysis, key clinical and laboratory predictors were identified to guide timely catheter removal decisions.
Background
Peritoneal dialysis is a common renal replacement therapy for end-stage renal disease but carries a risk of peritonitis, a serious infection that can lead to catheter failure and increased morbidity. Prompt antibiotic treatment and appropriate catheter management are critical to patient outcomes. Current guidelines recommend catheter removal in refractory or complicated peritonitis cases, but objective tools to determine optimal timing remain limited. Decision tree analysis offers a method to integrate clinical and laboratory data to support clinical decision-making in this context.
Data Highlights
Characteristic
Catheter Removal (n=31)
No Removal (n=487)
p-value
Age (years)
Mean ± SD
Mean ± SD
NS
Sex (Male %)
Data
Data
NS
CRP (mg/L)
Higher in removal group
Lower in no removal group
<0.05
Leukocyte count
Elevated
Lower
<0.05
Glasgow Coma Scale
Lower scores associated with removal
Higher scores
<0.05
Key Findings
Multivariable logistic regression identified elevated C-reactive protein (CRP), leukocytosis, and lower Glasgow Coma Scale (GCS) scores as significant predictors of catheter removal.
Decision tree analysis provided a visual and interactive model incorporating biochemical and clinical parameters to stratify risk for catheter removal.
Patients with refractory peritonitis, indicated by persistent infection markers despite antibiotics, were more likely to require catheter removal.
Early identification of high-risk patients can potentially improve outcomes by guiding timely catheter removal decisions.
Clinical Implications
Clinicians should monitor inflammatory markers such as CRP and leukocyte counts alongside neurological status (GCS) in PD patients presenting with peritonitis to identify those at increased risk for catheter removal. Utilizing decision tree models may enhance clinical judgment by integrating multiple variables, facilitating earlier intervention and potentially reducing morbidity. Objective tools can support adherence to ISPD guidelines and optimize patient management in emergency settings.
Conclusion
This study highlights key clinical and laboratory factors associated with PD catheter removal in emergency department patients with peritonitis. Integrating multivariable logistic regression and decision tree analysis offers a promising approach to improve decision-making and patient outcomes.
References
ISPD Guidelines 2016 -- Management of Peritoneal Dialysis-Related Infections
Chang Gung Memorial Hospital Study 2018 -- PD Catheter Removal Factors
