# Carbapenem-resistant Acinetobacter baumannii bloodstream infections in critically ill patients: prognostic factors and development of a nomogram

## TL;DR

This study identifies key risk factors for death in ICU patients with CRAB bloodstream infections and creates a predictive model to help guide treatment decisions.

## Contribution

A novel nomogram model is developed to predict 30-day mortality in ICU patients with CRAB bloodstream infections.

## Key findings

- 30-day mortality rate among ICU patients with CRAB-BSI was 60.0%.
- Age, impaired consciousness, corticosteroid use, SOFA score, and CRP level were independent risk factors for mortality.
- The nomogram achieved strong predictive accuracy with an AUROC of 0.863.

## Abstract

This retrospective study aimed to identify the clinical features and prognostic determinants in intensive care unit (ICU) patients with carbapenem-resistant Acinetobacter baumannii (CRAB) bloodstream infection (BSI) and to establish a personalized risk prediction model.

This retrospective cohort study included 185 ICU patients with CRAB-BSI at a tertiary care hospital between 2013 and 2023. Based on 30-day outcomes, patients were categorized into survival and non-survival groups. Independent risk factors for mortality were identified through univariate and multivariate logistic regression analyses. These factors were used to construct a nomogram prediction model. Model performance was evaluated by assessing discrimination using the area under the receiver operating characteristic curve with a calibration plot, and clinical utility via decision curve analysis.

The 30-day mortality rate among ICU patients with CRAB-BSI was 60.0%. Multivariate analysis revealed that age [odds ratio (OR) = 1.04, 95% confidence interval (CI): 1.02–1.07], impaired consciousness (OR = 5.10, 95% CI: 2.27–11.45), prior corticosteroid use (OR = 5.82, 95% CI: 2.10–16.12), Sequential Organ Failure Assessment (SOFA) score (OR = 1.26, 95% CI: 1.12–1.42), and C-reactive protein (CRP) level (OR = 1.01, 95% CI: 1.01–1.02) were independent risk factors for 30-day mortality. A nomogram incorporating these variables achieved an area under the curve (AUROC) of 0.863 for predicting 30-day mortality risk. The calibration curve indicated excellent concordance between predictions and observed outcomes, and decision curve analysis demonstrated significant clinical net benefit over a wide range of probability thresholds.

Mortality is high in ICU patients with CRAB-BSI and is significantly associated with age, impaired consciousness, corticosteroid use, SOFA score, and CRP level. The developed nomogram exhibits strong predictive accuracy and may function as a practical tool for quantitative management decisions.