Development and validation of a predictive model for treatment failure of peritoneal dialysis-related peritonitis in elderly patients
10.3760/cma.j.issn.0254-9026.2022.02.012
- VernacularTitle:老年腹膜透析相关性腹膜炎患者治疗失败的预测模型构建与验证
- Author:
Yu ZHANG
1
;
Liming YANG
;
Xueyan ZHU
;
Xiaoxuan ZHANG
;
Meijun LIU
;
Xiaohua ZHUANG
;
Ping LUO
;
Wenpeng CUI
Author Information
1. 吉林大学第二医院肾病内科，长春 130041
- Keywords:
Peritoneal dialysis;
Peritonitis;
Treatment failure;
Predictive model
- From:
Chinese Journal of Geriatrics
2022;41(2):185-190
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To establish and validate a predictive model for treatment failure of peritoneal dialysis-related peritonitis(PDAP)in elderly patients.Methods:Clinical data of peritoneal dialysis(PD)patients who were followed up from January 1, 2013 to December 31, 2019 at four Grade A tertiary hospitals in Jilin Province were collected.A total of 362 elderly patients with PDAP were eventually included as study subjects.Subjects recruited from 2013 to 2017 were used for model construction and the logistic regression model was used to screen risk factors for treatment failure of PDAP in elderly patients.A nomogarm was constructed to predict treatment failure of secondary PDAP using R language.The receiver operating curve(ROC)and calibration curve were used to evaluate discrimination accuracy of the model.Subjects from 2018 to 2019 were used as the cohort for validation of discrimination accuracy of the model.Results:Of 258 PDAP patients in the modeling cohort, 29 experienced treatment failure, including 15 PDAP-related deaths and 14 cases requiring catheter removal.The multivariate logistic regression model showed that types of pathogens( OR=8.849, 95% CI: 1.656-47.269, P=0.011), long dialysis age( OR=1.023, 95% CI: 1.005-1.042, P=0.013), pre-hospitalization antibiotic treatment( OR=5.123, 95% CI: 1.338-19.610, P=0.017), and dialysate white blood cell count on day 5>100×10 6/L( OR=7.085, 95% CI: 2.162-23.217, P=0.001)were independent risk factors for treatment failure of PDAP in elderly patients.For the nomogarm predictive model, the areas under the ROC curve(AUC)in the modeling cohort and the validation cohort were 0.818(95% CI: 0.735-0.902)and 0.762(95% CI: 0.656-0.889), respectively, and the calibration curves were close to a straight line with a slope of 1. Conclusions:Our nomogram predictive model based on types of pathogens, months of dialysis, pre-hospital admission antibiotic treatment, and dialysate white blood cell count on day 5 has demonstrated satisfactory discrimination accuracy for treatment failure of PDAP in elderly patients.
