A nomogram to predict bloodstream infections caused by third-generation cephalosporin-resistant Enterobacteriaceae.

OBJECTIVES

Third-generation cephalosporin resistance (3GCR) in Enterobacteriaceae bloodstream infections (BSIs) challenges clinical management in China. This study aimed to develop a predictive tool to guide antibiotic stewardship.

METHODS

A retrospective cohort of 459 patients with Enterobacteriaceae BSIs (185 3GCR, 274 third-generation cephalosporin-sensitive [3GCS]) from a tertiary hospital (2021-2023) was analysed. Multivariate logistic regression with backward selection identified predictors, integrated into a nomogram.

RESULTS

Independent 3GCR predictors included advanced age ( (Odds Ratio (OR) = 1.02, 95% Confidence Interval (CI): 1.01-1.04), liver cirrhosis (OR = 2.18, 1.26-3.77), solid tumours (OR = 2.04, 1.19-3.50), prior third-generation cephalosporin use (OR = 2.44 1.47-4.08), and elevated procalcitonin (OR = 1.01, 1.01-1.02). The nomogram showed moderate discrimination (Area Under Curve (AUC) = 0.751, 95% CI: 0.67-0.83) and good calibration (Hosmer-Lemeshow P = 0.757).

CONCLUSIONS

This nomogram, incorporating readily available clinical variables, facilitates rapid 3GCR risk stratification. Its implementation may optimize empirical antibiotic selection, improve outcomes, and mitigate resistance escalation in Enterobacteriaceae BSIs.