Risk Factors And Prognostic Model Of Carbapenem-resistant Klebsiella Pneumoniae In Patients With Bloodstream Infection

Background:The infection caused by carbapenem-resistant Klebsiella pneumoniae(CRKP)is increasing year by year.Bloodstream infection(BSI)is one of the most important infections of CRKP,leading to high mortality.The study aims to investigate the risk factors of carbapenem-resistant Klebsiella pneumoniae bloodstream infection and the related factors of poor prognosis of patients and to establish a predictive model of infection and prognosis.Methods:1.The modeling group included patients with Klebsiella pneumoniae(KP)bloodstream infection admitted from September 2017 to December 2021.The patients were divided into the CRKP group and CSKP group according to whether they were resistant to carbapenem antibiotics.The validation group included patients admitted from January 2022 to June 2022.Univariate analysis and multivariate logistic regression analysis were used to screen out the independent influencing factors of CRKP-BSI in hospitalized patients.Based on the selected risk factors,a prediction model was constructed and visualized as a nomogram,and verified in the validation group.The area under the curve(AUC)of the receiver operating characteristic(ROC)curve was used to evaluate the discrimination of the prediction model,and the calibration plot was used to evaluate the calibration of the model.2.The CRKP-BSI patients admitted from September 2017 to December 2021 were included in the modeling group,and the patients were divided into the poor prognosis group and the good prognosis group according to the hospitalization outcome.Multivariate logistic regression analysis was used to explore the risk factors for poor prognosis of CRKP-BSI,and the model was constructed,evaluated,and verified.Results:1.A total of 628 KP-BSI patients were included in the study as the modeling group,including 433 patients in the CRKP group and 195 patients in the CSKP group.A total of 130 patients were included in the validation group,84 patients in the CRKP group,and 46 patients in the CSKP group.The influencing factors of CRKP bloodstream infection were nosocomial infection(OR = 4.166,95 % CI 2.397-7.294),ICU admission(OR = 4.102,95 % CI 2.518-6.683),shock(OR = 3.026,95 % CI1.665-5.501),hospitalization days before bloodstream infection(OR = 1.032,95 %CI 1.010-1.054),PTA(OR = 0.982,95 % CI 0.971-0.992),FBG(OR = 1.274,95 %CI 1.085-1.496).Based on the above influencing factors,a prediction model for CRKP-BSI was established :Logit(P)= In P/(1-P)= 1.427×[nosocomial infection(yes = 1,no = 0)] + 1.412×[ICU admission(yes = 1,no = 0)] + 0.031×[bloodstream infections before hospitalization days(days)] + 1.107×[shock(yes = 1,no = 0)] +0.242×[FBG(g/L)]-0.019×[PTA(%)]-0.446.The areas under ROC curves of the model group and the verification group were 0.872 and 0.902,indicating good differentiation of the model.The average absolute error between the prediction of CRKP-BSI occurrence risk and the actual CRKP-BSI occurrence risk in the calibration curve diagram of the model group and the verification group was 0.008 and 0.025,respectively,indicating that the prediction accuracy of this model was high.3.A total of 433 patients with CRKP-BSI were included in the study on the prognosis of CRKP-BSI patients as the model group,including 253 patients with a poor prognosis and 180 patients with a good prognosis.A total of 84 patients with CRKP-BSI were included in the verification group,with 47 patients in the poor prognosis group and 37 patients in the good prognosis group.Based on ICU admission(OR=8.172,95%CI 4.389-15.216),shock(OR=3.081,95%CI 1.836-5.172),blood purification(OR=3.788,95%CI 1.337-10.738),blood glucose(OR=1.133,95%CI 1.034-1.242),PTA(OR=0.977,95%CI 0.961-0.993)to establish a prediction model of poor prognosis in patients with CRKP-BSI:Logit(P)= In P/(1-P)= 2.101×[ICU admission(yes = 1,no = 0)] + 1.125 ×[shock(yes = 1,no = 0)] + 1.332×[blood purification(yes = 1,no = 0)]+0.125×[Blood glucose(mmol/L)]-0.023×[PTA(%)]+0.100.The area under the ROC curve of the model was 0.869,and the correction curve indicated good prediction accuracy.The area under the ROC curve of the prediction model in the external validation group was 0.826.The calibration curve shows that the predicted value curve and the actual value curve are the same Conclusion:1.The study found that nosocomial infection,length of hospital stay before bloodstream infection,ICU admission,shock,PTA,and FBG are important influencing factors for the occurrence of CRKP-BSI.Based on the above-influencing factors,the predictive model of CRKP-BSI has high efficiency and can identify patients with CRKP infection in the bloodstream in the early stage.2.ICU admission,shock,blood purification,blood glucose,and PTA are the factors affecting the prognosis of patients with CRKP-BSI.After the occurrence of CRKP-BSI,medical workers should effectively control the risk factors and identify high-risk patients with poor prognosis,and take appropriate intervention measures to improve the prognosis of patients.