ICU患者并发导管相关血流感染预测模型的构建与验证

Construction of a prediction model for catheter-related bloodstream infection in intensive care unit patients and its validation

目的探讨重症监护病房(ICU)患者并发导管相关血流感染(CRBSI)的危险因素,并构建其预测模型。方法选取2021年2月至2023年2月广东省第二人民医院收治的402例ICU住院患者,其中2021年2月至2022年9月的280例患者作为训练集构建预测模型,2022年10月至2023年2月的122例患者作为验证集进行外部验证。单因素和多因素logistic回归分析患者发生CRBSI的危险因素,构建基于危险因素的联合预测模型并绘制列线图,验证其准确性。结果训练集280例患者中有15例发生CRBSI(5.4%),验证集122例患者中有8例发生CRBSI(6.6%),2组CRBSI发生率差异无统计学意义(χ^(2)=0.227,P=0.634)。训练集中多因素分析显示,导管留置时间、导管滑动/滑出、白细胞、白蛋白是ICU患者发生CRBSI的独立危险因素(P<0.05)。列线图模型验证结果显示,训练集与验证集的C-index分别为0.904和0.877,曲线下面积分别为0.845和0.822。决策曲线分析显示,列线图在训练集和验证集有较好的临床获益性。结论基于CRBSI危险因素构建的列线图可预测ICU患者发生CRBSI风险,能帮助医护人员及时采取有效的干预措施。

Objective To investigate the risk factors for catheter-related bloodstream infection(CRBSI)in intensive care unit(ICU)patients,and to construct a risk prediction model for the CRBSI.Methods A total of 402 ICU inpatients admitted to Guangdong Second Provincial General Hospital from February 2021 to February 2023 were collected in this study,of which 280 patients from February 2021 to September 2022 were allocated into the test group for construct a prediction model,and 122 patients from October 2022 to February 2023 in the validation group for external validation.Univariate and multivariate logistic regression were implemented to analyze the risk factors of CRBSI in ICU patients.A nomogram to predict CRBSI in ICU patients was constructed based on risk factors,and its accuracy was verified.Results CRBSI was detected in 15/280(5.4%)cases in the test group and 8/122(6.6%)cases in the validation group.There was no significant difference in the incidence of CRBSI between the two groups(χ^(2)=0.227,P=0.634).The results of multivariate analysis in the test group showed that catheter retention time,catheter sliding/sliding out,white blood cells,and albumin were independent risk factors for CRBSI in ICU patients(all P<0.05).The verification results from nomogram model showed that the C-index of test group and validation group was 0.904 and 0.877,respectively.The areas under the curve(AUC)of the test group and validation group was 0.845 and 0.822,respectively.Decision curve analysis showed that nomogram had a better clinical benefit in the test group and validation group.Conclusion A nomogram created based on the risk factors of CRBSI can predict the risk of CRBSI in ICU patients,which is helpful for medical staff take effective intervention measures.