Early trajectories of antibiotic exposure and colonization pressure and risk of ICU-acquired carbapenem-resistant Gram-negative bacteria: A prospective cohort study

Background Carbapenem-resistant Gram-negative bacteria (CR-GNB) pose a critical threat in intensive care units (ICUs), with antibiotic exposure and colonization pressure identified as key associated factors. Prior studies have analyzed these two time-varying factors as static cumulative variables, obscuring heterogeneity in temporal patterns and their joint evolution. How distinct early trajectories of these factors jointly relate to CR-GNB acquisition risk remains unclear. Methods In this prospective cohort study conducted at four ICUs in a tertiary-care center in China from March 2024 to January 2025, we enrolled consecutive patients with systematic rectal surveillance cultures. We used group-based multi-trajectory modeling to identify distinct joint trajectories of daily antibiotic exposure (dose, duration, spectrum) and colonization pressure during the first five ICU days. Continuous-time Markov multi-state models with Day-5 landmark analysis were adopted to assess associations between trajectory groups and subsequent ICU-acquired CR-GNB, adjusting for baseline and cumulative covariates. Results Among 533 patients entering the Day-5 landmark analysis, three distinct trajectory groups were identified: Low exposure/low pressure (34.1%), escalating exposure/intermediate pressure (54.4%), and high exposure/high pressure (11.4%). CR-GNB acquisition occurred in 124 patients (23.3%), with rates of 9.3%, 26.2%, and 50.8% across trajectory groups, respectively. Compared with the low-exposure trajectory, adjusted hazard ratios were 1.68 (95% confidence interval [CI] 1.24–2.28) for escalating-exposure and 2.82 (95% CI, 1.53–5.19) for high-exposure trajectories (both P < 0.001). Conclusions This study identified distinct early trajectories of antibiotic exposure and colonization pressure that were associated with differential CR-GNB acquisition risk. This trajectory-based framework for early association-based risk stratification may inform targeted prevention strategies, but external validation is required before clinical implementation. Trial registration Chinese Clinical Trial Registry Identifier: ChiCTR2400081352. Registered 28 February 2024.