Does transfusion of blood products increase the risk of multidrug-resistant infections in critically ill patients?

Objectives : To investigate the association between transfusion of blood products during intensive care unit (ICU) stay and the incidence of ICU-acquired colonisation and infection with multidrug-resistant (MDR) bacteria (ICU-MDR-col, ICU-MDR-inf, respectively). Methods : Secondary analysis of two prospective multicentre cohort studies conducted in eight ICUs in France. Adult patients admitted to participating ICUs were enrolled consecutively from May 2019 to June 2021. All patients underwent systematic screening for colonisation with MDR bacteria at admission and weekly thereafter. Exposure to transfusion of blood products during ICU stay, including packed red blood cells (PRBCs), fresh frozen plasma (FFP) and platelet concentrates (PCs), was recorded. The primary outcome was the incidence of ICU-MDR-col and/or ICU-MDR-inf (ICU-MDR-col/inf) until ICU discharge, censored at day 28. Secondary outcomes included the incidence of ICU-MDR-col and ICU-MDR-inf, considered separately. Associations were assessed using mixed-effects Poisson regression models adjusted for prespecified confounders. Subgroup analyses were performed according to immunosuppression status at ICU admission. Results : Among 1103 patients, 358 (32.5%) received transfusion of any blood product. Overall, 292 (26.5%) patients experienced at least one ICU-MDR-col/inf event. Transfusion was associated with an increased incidence of ICU-MDR-col/inf (adjusted incidence rate ratio [aIRR] 1.31, 95% confidence interval [CI] 1.06–1.62; p = 0.013), mainly driven by ICU-MDR-inf (aIRR 1.62, 95% CI 1.16–2.24; p = 0.004), but not ICU-MDR-col considered individually. This association was restricted to PRBC transfusion. No significant heterogeneity was observed according to immunosuppression status. Conclusions : Transfusion of PBRCs was associated with an increased incidence of ICU-acquired MDR infections, with no significant heterogeneity observed according to immunosuppression status. Implications for Clinical Practice : These findings support a restrictive approach to transfusion and warrant further investigation into its immunological and ecological consequences in critically ill patients.