维持性血液透析患者导管相关感染的病原菌情况、危险因素与预测模型构建

Pathogen status, risk factors and prediction model construction for catheter-related infection in patients with maintenance hemodialysis

  • 摘要:
    目的 探讨维持性血液透析(MHD)患者导管相关感染的病原菌情况、危险因素并构建预测模型。
    方法 采用病例对照研究设计,回顾性收集2022年6月—2024年3月在苏州大学附属苏州九院进行MHD治疗的208例患者的临床资料。分析患者导管相关感染情况,依据导管相关感染发生情况将MHD患者分为感染组和无感染组,并比较2组患者的临床资料。为避免过度拟合,采用最小绝对收缩和选择算子(LASSO)回归分析初步筛选变量,再采用多因素Logistic回归分析探讨MHD患者导管相关感染的危险因素。基于各风险因素构建预测模型,采用受试者工作特征(ROC)曲线的曲线下面积(AUC)评估模型的区分度; 应用K-fold交叉验证法对模型进行深度验证。
    结果 208例MHD患者导管相关感染率为24.52%(51/208), 其中导管局部感染22例,隧道感染17例,导管相关性血流感染12例; 共分离培养出革兰阳性菌36株,革兰阴性菌28株,真菌1株。感染组与无感染组在合并糖尿病、白蛋白、置管部位、导管留置时间以及导管维护频率方面比较,差异有统计学意义(P < 0.05)。LASSO回归分析和多因素Logistic回归分析表明,合并糖尿病(OR=3.651, 95%CI: 2.056~7.508)、白蛋白低(OR=1.782, 95%CI: 1.129~2.815)、置管部位为股内静脉(OR=2.298,95%CI: 1.269~4.162)、导管留置时间长(OR=2.959, 95%CI: 1.327~6.599)、导管维护频率≤2次/月(OR=2.373, 95%CI: 1.316~4.276)均为MHD患者导管相关感染的独立危险因素(P < 0.05)。基于独立危险因素构建的预测模型的AUC为0.885(95%CI: 0.838~0.931), 灵敏度为0.875, 特异度为0.847;100次K-fold交叉验证表明模型具有较好的泛化能力。
    结论 MHD患者导管相关感染风险较高,构建预测模型能有效评估MHD患者导管相关感染风险。

     

    Abstract:
    Objective To explore the pathogens status and risk factors for catheter-related infection in patients with maintenance hemodialysis (MHD), and to establish a prediction model.
    Methods A case-control study design was adopted, and clinical materials of 208 MHD patients treated in Suzhou Ninth People's Hospital of Suzhou University from June 2022 to March 2024 were retrospectively collected. The catheter-related infection status of the patients was analyzed, and they were divided into infection group and non-infection group based on the occurrence of catheter-related infections. The clinical materials were compared between the two groups. To avoid overfitting, the least absolute shrinkage and selection operator (LASSO) regression analysis was used for initial variable screening, followed by multivariate Logistic regression analysis to explore the risk factors for catheter-related infections in MHD patients. A predictive model was constructed based on the risk factors, and the area under the curve (AUC) of the receiver operating characteristic (ROC) curve was used to assess discrimination of the model. The K-fold cross-validation method was applied for in-depth validation of the model.
    Results The incidence rate of catheter-related infection in the 208 MHD patients was 24.52% (51/208), including 22 cases of local catheter infections, 17 cases of tunnel infections, and 12 cases of catheter-related bloodstream infections. A total of 36 Gram-positive bacteria, 28 Gram-negative bacteria, and 1 fungus were isolated and cultured. Significant differences were observed between the infection and non-infection groups in terms of complicating diabetes, albumin level, catheter insertion site, catheter indwelling time, and catheter maintenance frequency (P < 0.05). LASSO regression analysis and multivariate Logistic regression analysis revealed that complicating diabetes (OR=3.651, 95%CI, 2.056 to 7.508), low albumin level (OR=1.782, 95%CI, 1.129 to 2.815), catheter insertion in the internal femoral vein (OR=2.298, 95%CI, 1.269 to 4.162), long catheter indwelling time (OR=2.959, 95%CI, 1.327 to 6.599), and catheter maintenance frequency ≤2 times per month (OR=2.373, 95%CI, 1.316 to 4.276) were independent risk factors for catheter-related infections in MHD patients (P < 0.05). The AUC of the predictive model based on independent risk factors was 0.885 (95%CI, 0.838 to 0.931), with a sensitivity of 0.875 and a specificity of 0.847. A hundred rounds of K-fold cross-validation demonstrated good generalization ability of the model.
    Conclusion MHD patients have a high risk of catheter-related infections, and the constructed prediction model can effectively assess the risk of catheter-related infections in MHD patients.

     

/

返回文章
返回