Infection-related ventilator-associated complications in ICU patients colonised with extended-spectrum β-lactamase-producing Enterobacteriaceae

Purpose To investigate the clinical significance of infection-related ventilator-associated complications (IVAC) and their impact on carbapenem consumption in mechanically ventilated (MV) patients colonised with extended-spectrum β-lactamase-producing Enterobacteriaceae (ESBLE). Methods Inception cohort study from the French prospective multicenter OUTCOMEREA database (17 ICUs, 1997–2015) including all ESBLE carriers (systematic rectal swabbing at admission then weekly and/or urinary or superficial surgical site colonisation) with MV duration > 48 h and ≥ 1 episode of IVAC after carriage documentation. All ICU-acquired infections were microbiologically documented. Results The 318 enrolled ESBLE carriers (median age 68 years; males 67%; medical admission 68%; imported carriage 53%) experienced a total of 576 IVAC comprising 361 episodes (63%) without documented infection, 124 (21%) related to infections other than ventilator-associated pneumonia (VAP), 73 (13%) related to non-ESBLE VAP and 18 (3%) related to ESBLE VAP. Overall, ESBLE infections accounted for only 43 episodes (7%). Carbapenem exposure within the preceding 3 days was the sole independent predictor of ESBLE infection as the causative event of IVAC, with a protective effect (adjusted odds ratio 0.2, 95% confidence interval 0.05–0.6; P < 0.01). Carbapenems were initiated in 9% of IVAC without infection, 15% of IVAC related to non-VAP infections, 42% of IVAC related to non-ESBLE VAP, and 56% of IVAC related to ESBLE VAP (ESBLE VAP versus non-ESBLE VAP: P = 0.43). Conclusions IVAC in ESBLE carriers mostly reflect noninfectious events but act as a strong driver of empirical carbapenem consumption. ESBLE infections are scarce yet hard to predict, strengthening the need for novel diagnostic approaches and carbapenem-sparing alternatives.