The enemy in our gut? Infections due to extended-spectrum beta-lactamase producing Enterobacterales (ESBL-E) in hospital patients: a cohort study comparing 2971 ESBL-E-carriers and 2971 non-carriers.

PURPOSE: To quantify the risk of healthcare-associated infections (HAI) with extended spectrum beta-lactamase-producing Enterobacterales (ESBL-E) and other pathogens in carriers and non-carriers, and to identify risk factors for HAI. METHODS: In 2014 and 2015, a cohort study was conducted including all eligible ESBL‑E-colonized hospital patients and a randomly 1:1 selected sample of non‑colonized controls. Incidence densities per 1,000 patient‑days were calculated, and Fine-Gray sub-distribution hazard models were applied. RESULTS: In total, 2,971 ESBL-E carriers and 2,971 non-carriers were included. Among carriers, incidence densities were 3.29 (95% CI 2.76-3.93) per 1,000 patient days for HAI with ESBL-E, and 4.41 (95% CI 3.78-5.14) for HAI with other pathogens. In contrast, ESBL-E HAI among non-carriers was rare with an incidence density of 0.05 (95% CI 0.01-0.21) per 1,000 patient days. ESBL-E colonization was identified as most important risk factor specific for ESBL-E HAI followed by peptic ulcer, congestive heart failure and age. CONCLUSION: Incidence of HAI with ESBL-E was high among ESBL-E carriers, but HAI with other pathogens remained more common. ESBL‑E colonization was the dominant determinant of ESBL‑E HAI, with cases occurring almost exclusively in carriers. In conclusion, a known ESBL-E colonization status is an important predictor, but should not be the only factor to rationalize an empirical strategy targeting ESBL-E without considering additional factors.