The prognostic value of a laboratory-based frailty index (FI-Lab) in carbapenem-resistant Acinetobacter baumannii infections: a multicentre cohort study

BACKGROUND: Carbapenem-resistant Acinetobacter baumannii (CRAB) infections remain a major cause of mortality despite advances in antimicrobial therapy. Frailty, reflecting cumulative physiological decline, may strongly influence outcomes beyond age or comorbidity. This study evaluated whether a 30-item Frailty Index-Laboratory (FI-Lab), calculated within 48 h of admission, independently predicts early (7-day) and short-term (30-day) mortality among hospitalized patients with CRAB infections. METHODS: We conducted a multicentre retrospective cohort study including adults with microbiologically confirmed CRAB infection admitted to four Infectious Diseases Units in Apulia, Italy (Bari, Foggia, Taranto-Moscati, Bisceglie) between January 2020 and August 2025. The FI-Lab, derived from 30 biochemical parameters, quantified the proportion of abnormal laboratory values (range 0-1). Cox regression models assessed associations between FI-Lab and mortality, adjusting for age, comorbidity, SOFA score, septic shock, infection site, therapy type, and other covariates. RESULTS: Among 199 patients (mean age 70 ± 15 years; 38.2% male), the median FI-Lab was 0.56 (IQR 0.45-0.68). Seven-day, 30-day, and in-hospital mortality were 12.6%, 19.1%, and 20.6%, respectively. Each + 0.10 increase in FI-Lab was associated with higher 7-day (HR 1.52, 95% CI 1.22-1.89), 30-day (HR 1.68, 95% CI 1.36-2.08), and in-hospital (HR 1.63, 95% CI 1.32-2.01) mortality (all p < 0.01). Patients with respiratory infections had significantly higher 30-day mortality than those with bloodstream infections (HR 1.62, 95% CI 1.05-2.51, p = 0.029). No mortality difference was observed between cefiderocol- and colistin-based regimens. FI-Lab demonstrated excellent discrimination (AUC = 0.89 for 30-day mortality). CONCLUSIONS: The FI-Lab independently predicted early and short-term mortality in CRAB infections, outperforming traditional indices and highlighting the prognostic weight of biological frailty. The strong impact of respiratory infection and the absence of therapy-related differences suggest that host vulnerability may outweigh antimicrobial choice in determining outcomes. FI-Lab, easily derived from routine tests, could be integrated into clinical algorithms for early risk stratification in multidrug-resistant infection.