Construction of a predictive model for death in patients with carbapenem-resistant Acinetobacter baumannii bloodstream infection
10.3969/j.issn.1671-8348.2025.04.016
- VernacularTitle:ICU耐碳青霉烯类鲍曼不动杆菌血流感染患者生存率的预测模型构建
- Author:
Siting YI
1
;
Pingjuan LIU
;
Mengmin YE
;
Jincheng ZENG
Author Information
1. 广东医科大学/广东省医学免疫与分子诊断重点实验室/东莞市医学活性分子开发与转化重点实验室,广东东莞 523808;中山大学附属第一医院医学检验科,广州 510080
- Keywords:
carbapenem-resistant Acinetobacter baumannii;
bloodstream infection;
risk factors;
nomo-gram;
prediction model
- From:
Chongqing Medicine
2025;54(4):884-888
- CountryChina
- Language:Chinese
-
Abstract:
Objective To investigate the risk factors for carbapenem-resistant Acinetobacter baumannii(CRAB)bloodstream infection in intensive care units(ICU)and construct a nomogram prediction model.Methods A retrospective analysis was conducted on the medical records of patients with CRAB bloodstream(BSI)infection in the ICU ward of the First Afliated Hospital,Sun Yat-sen University from January 1,2018 to December 31,2023.Cox regression analysis was usedo determine the risk factors for death in patients with CRAB bloodstream infection and to construct a nomogram.prediction model.The predictive ability of the no-mogramprediction model was evaluated by drawing calibration curves,receiver operating characteristic(ROC)curves,and consistency index(C-index).Results Septic shock(HR=7.770,95％CI:1.852-32.593)was an independent risk factor for mortality in patients with CRAB bloodstream infections,and hospitalisation day＞14 d(HR=0.331,95％CI:0.165-0.665)was an independent protective factor for mortality in patients with CRAB bloodstream infections.A nomogram was constructed based on the above factors,with a C-index of 0.725(95％CI:0.652-0.798).The predictive efficacy of patient survival rates at 20 d and 30 d was 0.831(95％CI:0.752-0.910)and 0.826(95％CI:0.715-0.937),respectively.The calibration curve was well fit-ted.Conclusion Septic shock was an independent risk factor for mortality in patients with CRAB bloodstream infections,and hospitalisation days＞14 d was an independent protective factor for mortality in patients CRAB BSI infections.The nomogram prediction model constructed accordingly has good predictive value for the sur-vival rate of patients in the intensive care unit who undergo CRAB bloodstream infections.
