ICU-acquired infection in neutropenic patients

Among neutropenic patients, sepsis is a frequent cause of ICU admission, and its epidemiology is well described, but data on infection acquired during the ICU stay (ICU-AI) are lacking. We aim to describe, using a large nationwide cohort study, the incidence of ICU-AI for neutropenic patients in a large network of French intensive care units (ICUs). This study was conducted using the healthcare-associated infection surveillance cohort “REA-REZO” involving 265 participating ICUs. All patients admitted to a participating ICU from 1st January 2018 to 31th December 2023, for more than 48 h were eligible. The primary endpoint was the rate of ICU-AI. A matched analysis was conducted using non-parsimonious regression model to compare neutropenic and non-neutropenic patients. Matching process resulted in well-balanced 7545 patients’ pairs. ICU-AI incidence rate was 14.9 versus 13.0 AI per 1000 patients days in the neutropenic group vs non-neutropenic group (Incidence Rate Ratio [IRR] = 1.15 [1.05, 1.25] p = 0.002) with higher incidence rate for bloodstream infection (BSI) (p < 0.001) but comparable incidences rates of pneumonia (p = 0.503) and central-line associated BSI (p = 0.100). There were less enterobacterales among ICU-AI of the neutropenic group (33.8% vs 45.4%, p < 0.001), but higher rate of coagulase negative Staphylococcus (12.0% vs 5.2%, p < 0.001), Enterococcus spp, (8.9% vs 6.0%, p = 0.015) and Candida spp. (10.4% vs 6.2%, p = 0.001). Rates of ICU-AI involving multidrug resistant micro-organisms did not differed, neither did the rate of ICU-AI involving non fermenting Gram-negative bacilli. Neutropenic patients had higher ICU-AI incidence as compared with non-neutropenic patients, specifically through an increase of BSI incidence. Microbiology differs between neutropenic patients and non-neutropenic patients with no difference in multidrug resistant microorganism involvement.