Risk factors for bloodstream infections caused by carbapenem-resistant Acinetobacter baumannii in ICU patients with any site colonisation: A prospective observational study

Objectives The aim of this study was to develop and validate a risk score to predict the occurrence of bloodstream infections (BSI) caused by carbapenem-resistant Acinetobacter baumannii (CRAB) among colonised intensive care unit (ICU) patients, in order to support early clinical decision-making and antimicrobial stewardship. Methods Prospective observational study including adult patients admitted to ICU with CRAB colonisation at any site at the University Hospital of Pisa, Italy (June 2020-June 2023). The primary outcome measure was BSI caused by CRAB. A multivariable regression analysis was performed to identify factors independently associated with CRAB-BSI. Regression coefficients were used to develop a risk score for CRAB-BSI. Discrimination was evaluated using the area under the receiver operating characteristic curve (AUC-ROC). Results Among 283 colonised patients, 103/283 (36.4%) developed a BSI caused by CRAB. Median (IQR) time from colonisation to BSI was 4 days (1–8 days). On multivariable analysis, burns (OR: 8.219, 95% CI: 3.591–18.812, P < 0.001), number of colonised sites (per-site: OR: 2.197, 95% CI: 1.363–3.541, P = 0.001), respiratory tract colonisation (OR: 4.285, 95% CI: 2.179–8.426, P < 0.001), and cardiovascular disease (OR: 1.940, 95% CI: 1.068–3.524, P = 0.029) were independently associated with increased risk of BSI. The score ranged from 1 to 9 points. The AUC of the model was 0.817 (95% CI: 0.764–0.869, P < 0.001). The negative predictive value (NPV) was 97.8% in patients without septic shock and 72.2% in those who developed septic shock. Conclusions The rate of CRAB-BSI among colonised patients is considerable. The proposed score may be useful, after external validation, for a rational empirical use of new antibiotics.