Abstract: Objective To identify the risk factors for carbapenem-resistant pseudomonas aeruginosa (CRPA) infection in the elderly patients and to develop a nomogram prediction model for early recognition and intervention. Methods A retrospective study was conducted on 558 elderly patients (≥75 years) with pseudomonas aeruginosa (PA) infection admitted to Geriatric Hospital of Nanjing Medical University from January 2022 to December 2023. The patients were randomly divided into a training cohort (n=391) and a validation cohort (n=167). Multivariate logistic regression analysis was performed to identify the independent risk factors for CRPA infection. A nomogram prediction model was constructed using R software. The model’s performance was evaluated using receiver operating characteristic (ROC) curve analysis, calibration plots, and the Hosmer-Lemeshow test. Results Multivariate logistic regression analysis in the training cohort revealed four independent risk factors for CRPA infection in the elderly patients, including septic shock (OR=7.054, 95%CI:1.708－29.133), tracheostomy (OR=2.931, 95%CI:1.247－6.889), hospitalization history within 3 months (OR=22.575, 95%CI:11.524－44.222), and carbapenem use (OR=3.203, 95%CI:1.420－7.226). The model demonstrated a good fit of observed data, with an area under the ROC curve (AUC) of 0.816 (95%CI: 0.778－0.850) and 0.798 (95%CI: 0.736－0.860) for the experimental and validation cohort, respectively. The calibration curves and Hosmer-Lemeshow test results further confirmed the model had a good predictive performance. Conclusions Our prediction model exhibits a good predictive value in identifying elderly patients at risk of CRPA infection and can be used to guide preventive and treatment measures.

Key words: aged, carbapenem-resistant pseudomonas aeruginosa, risk factors, predictive model, nomogram