Carbapenem-Resistant Klebsiella pneumoniae Infections among ICU Admission Patients in Central China: Prevalence and Prediction Model

Objective. To investigate the prevalence of infections due to carbapenem-resistant Klebsiella pneumoniae (CRKP) among ICU admission patients in central China and develop a reliable prediction model. Methods. Five hundred and seven consecutive ICU admission patients with Klebsiella pneumoniae (KP) infection were enrolled in this retrospective multicenter case-control study from January 2014 to June 2018. The prevalence and antimicrobial susceptibility pattern were analyzed. Multivariate analysis was performed by logistic regression modeling to determine the risk factors. A prediction model was developed and verified using data from six hospitals in central China. Results. Of the total 507 isolates of KP, 244 (48.1%) strains were carbapenem resistant. The majority of these isolates were from sputum (30.9%) and blood (20.9%) samples. Tigecycline had good activity against CRKP (95.5%). The most common sequence type (ST) of CRKP was ST11 (84.4%), and 98.6% of them had the blaKPC-2 antimicrobial resistance gene. Thirteen variables were identified as independent risk factors for CRKP infection, including KP colonization or infection in the preceding year (OR=3.32, 95% CI 2.01-4.38), CD4/CD8 ratio <1 (OR=2.98, 95% CI 2.02-4.19), and parenteral nutrition >= 48 h (OR=1.88, 95% CI 1.22-3.04). The model developed to predict CRKP infection was effective, with an area under the receiver-operating characteristic curve of 0.854 (95% CI 0.821-0.884, p<0.001). Conclusions. ST11 carrying the blaKPC-2 antimicrobial resistance gene was the most common type of CRKP among the ICU admission patients in central China. The model demonstrated excellent predictive performance and exhibited good discrimination.