Risk factors of hemodialysis catheter-related bloodstream infection and prediction model / 中华肾脏病杂志

ABSTRACT

Objective:To investigate the risk factors for catheter-related bloodstream infection (CRBSI) in hemodialysis (HD) patients with tunnel-cuffed catheter (TCC) and construct a risk prediction model for the prevention and treatment of catheter infection.Methods:It was a retrospective study. Patients who had their TCC removed in Hemodialysis Access Center of the First Affiliated Hospital of Zhengzhou University from July to December 2020 were randomly divided into a training set (for model building) and a validation set (for model validation) in the ratio of 7∶3. The training set was divided into CRBSI group and non-CRBSI group with reference to the 2019 Kidney Disease Outcomes Quality Initiative clinical practice guidelines for vascular access, and the risk factors for the occurrence of CRBSI were analyzed. The odds ratio ( OR) values of the variables in the multivariate logistic regression analysis were applied to construct a risk prediction model, and the assessment ability of the model was validated in the validation set. Results:A total of 254 HD patients were included. The training set consisted of 179 patients with male-to-female ratio of 1.36∶1, age of (55.81±15.95) years old, median dialysis age of 18(8, 27) months, median TCC retention time of 15(5, 24) months, and 40 patients with confirmed CRBSI. Logistic regression analysis showed that, combined diabetes ( OR=2.711, 95% CI 1.174-6.258, P=0.019), history of catheter-related infection within 3 months ( OR=3.674, 95% CI 1.541-8.760, P=0.003), more than 4 times nursing interventions within 1 month ( OR=3.128, 95% CI 1.343-7.283, P=0.008), and central venous disease ( OR=2.572, 95% CI 1.130-5.854, P=0.024) were the independent influencing factors for CRBSI occurrence in HD patients with TCC. The OR values of the variables in the multivariate logistic regression were rounded to the assigned scores of the risk prediction model. The corresponding scores of each factor were summed in the training set to obtain the risk score. The receiver operating characteristic (ROC) curve was plotted, with area under the curve ( AUC) of 0.761(0.683-0.839) and maximum Youden index of 0.461, at which time the corresponding cut-off value was 6, with sensitivity of 90.0% and specificity of 56.1%. The model was validated in the validation set with AUC of 0.794(0.674-0.914) and cut-off value of 6, with sensitivity of 61.6% and specificity of 82.5%. Conclusions:Combined diabetes, history of catheter-related infection within 3 months, more than 4 times nursing interventions within 1 month, and central venous disease are the independent risk factors for CRBSI, and the prediction model based on the above factors has good efficacy in predicting the risk of CRBSI and can provide guidance for the prevention and treatment of CRBSI in HD patients.