Prediction model for 30-day mortality in cancer patients with carbapenem-resistant organism infections

BackgroundThis study aimed to develop and validate a novel model for predicting mortality risk in cancer patients with carbapenem-resistant organism (CRO) infections.MethodsCancer patients with CRO infections in Henan Cancer Hospital between January 2022 and March 2024 were included in this retrospective study. LASSO regression was used to construct a novel model for predicting mortality in cancer patients with CRO infections. Receiver operating characteristic (ROC), decision curve analysis (DCA), and clinical impact curves (CIC) were used to assess the predictive ability and clinical utility of the prediction model.ResultsA total of 417 cancer patients with CRO infections were included in the study. Fourteen factors were selected, including sample source, radiotherapy, blood culture, exposure to antibiotics after susceptibility testing, lymphocyte-to-monocyte ratio (LMR), lymphocyte-to-monocyte ratio (LMR), total protein (TP), blood urea nitrogen (BUN), calcium (CA), C-reactive protein (CRP), triglyceride (TG), procalcitonin (PCT), prothrombin time (PT), and thrombin time (TT), which were found to be associated with 30-day mortality in cancer patients with CRO infections. The areas under the ROC curve for the prediction model were 0.815 (95% CI: 0.767–0.857) and 0.801 (95% CI: 0.716–0.871) for the primary cohort and validation cohort, respectively. The models demonstrated good predictive accuracy, with p-values 0.479 and 0.786 on the Hosmer–Lemeshow test. DCA and CIC analyses confirmed the clinical utility of the prediction model.ConclusionOur model could be used as an effective individualized risk prediction tool for clinicians. It would provide personalized risk assessments for cancer patients with CRO infections.