A novel prognostic model for malignant patients with Gram-negative bacteremia based on real-world research

Abstract

Gram-negative bacteremia (GNB) is a common complication in malignant patients. Identifying risk factors and developing a prognostic model for GNB might improve the survival rate. In this observational and real-world study, we retrospectively analyzed the risk factors and outcomes of GNB in malignant patients. Multivariable regression was used to identify risk factors for the incidence of GNB, while Cox regression analysis was performed to identify significant prognostic factors. A prognostic model was constructed based on Cox regression analysis and presented on a nomogram. ROC curves, calibration plots, and Kaplan–Meier analysis were used to estimate the model. It comprised 1004 malignant patients with Bloodstream infection (BSI) in the study cohort, 65.7% (N = 660) acquired GNB. Multivariate analysis showed gynecologic cancer, hepatobiliary cancer, and genitourinary cancer were independent risk factors related to the incidence of GNB. Cox regression analysis raised that shock, admission to ICU before infection, pulmonary infection, higher lymphocyte counts, and lower platelet counts were independent risk factors for overall survival (OS). The OS was significantly different between the two groups classified by optimal cut-off value (log-rank, p < 0.001). Above all, a nomogram was created based on the prognostic model, which was presented on a website freely. This real-world study was concentrated on the malignant patients with GNB and proved that shock, admission to ICU before infection, pulmonary infection, higher lymphocyte counts, and lower platelet counts were related to the death of these patients. And a prognostic model was constructed to estimate the risk score of mortality, further to reduce the risk of death.