Risk factors for carbapenem-resistant Acinetobacter baumannii infection in children and A. baumannii-related mortality in the PICU: a retrospective study in Guangxi, China.

backgroundCarbapenem-resistant Acinetobacter baumannii (CRAB) is recognized as a critical-priority pathogen due to the considerable therapeutic difficulties it poses and its association with high mortality. Pediatric patients are particularly vulnerable to the adverse clinical outcomes of these infections.

methodsThis retrospective study analyzed the clinical and microbiological characteristics of patients at the Maternal and Child Health Hospital of Guangxi, China, between January 2019 and December 2025. It further assessed risk factors for both CRAB infection and A. baumannii-related mortality in the pediatric intensive care unit (PICU). Univariate and multivariate logistic regression analyses were performed to identify independent risk factors. A predictive model was then developed, and its performance was evaluated using the Hosmer-Lemeshow goodness-of-fit test and receiver operating characteristic (ROC) curve analysis.

resultsThe study included a total of 436 children, comprising 222 with CRAB infection and 214 with carbapenem-susceptible A. baumannii (CSAB) infection. In the univariate analysis, potential risk factors for CRAB infection (P < 0.05) included younger age, prior hospitalization, PICU admission, specific infections (enteritis, pneumonia, bloodstream infections), the use of invasive procedures or devices (mechanical ventilation, indwelling urethral catheters, central venous catheters, drainage tubes, parenteral nutrition, nasogastric feeding tubes), and prior exposure to glycopeptides, oxazolidinones, carbapenems, and corticosteroids. Potential risk factors for A. baumannii-related mortality in PICU (P < 0.05) included prior hospitalization, tumor, CRAB infection, septic shock, use of indwelling urethral or central venous catheters, parenteral nutrition, prior exposure to β-lactam-β-lactamase inhibitors, carbapenems, tetracyclines, trimethoprim-sulfamethoxazole, or corticosteroids, as well as levels of platelet (PLT) count, C-reactive protein (CRP), albumin (ALB), and alanine aminotransferase (ALT) levels. Subsequently, multivariate logistic regression model analysis identified admission to the PICU (Odds Ratio (OR): 5.136, 95% CI: 3.045-8.661, P < 0.001), pneumonia (OR: 2.618, 95% CI: 1.384-4.954, P = 0.003), carbapenems exposure (OR: 3.876, 95% CI: 2.277-6.598, P < 0.001), and prior corticosteroid exposure (OR: 1.860, 95% CI: 1.168-2.961, P = 0.009) as an independent risk factor for CRAB infection. In addition, the independent risk factors for A. baumannii-related mortality in PICU were prior hospitalization (OR: 3.656, 95% CI: 1.808-7.392; P < 0.001), CRAB infection (OR: 3.186, 95% CI: 1.406-7.219; P = 0.006), septic shock (OR: 3.417, 95% CI: 1.440-8.107; P = 0.005), parenteral nutrition (OR: 3.105, 95% CI: 1.248-7.725; P = 0.015), and decreased PLT count (OR: 0.997, 95% CI: 0.995-0.999; P = 0.001).

conclusionTwo logistic regression models were developed to predict CRAB infection in children and A. baumannii-related mortality among PICU patients, respectively. The findings highlight the importance of early identification and management of these risk factors. CLINICAL TRIAL NUMBER: Not applicable.