Purpose

Cardiac surgery-associated acute kidney injury (AKI) is linked to poor outcomes. An observational study from Copenhagen, Denmark identified perioperative red blood cell (RBC) transfusion as a modifiable risk factor for AKI, with a dose-dependent relationship between the number of RBC units transfused and the occurrence and severity of AKI. We aimed to externally validate those findings in a larger population.

Methods

We conducted a retrospective observational study of adult patients undergoing nonemergent on-pump cardiac surgery at Toronto General Hospital (Toronto, ON, Canada) between 2016 and 2021. Acute kidney injury was classified using the Kidney Disease: Improving Global Outcomes (KDIGO) criteria. Data were analyzed using inverse probability weighted logistic regression.

Results

Among 5,204 patients, 798 developed AKI, with 77% classified as stage 1, 11% as stage 2, and 12% as stage 3. Patients with AKI were older, had lower preoperative hemoglobin levels and estimated glomerular filtration rates, longer cardiopulmonary bypass duration, and lower intraoperative hemoglobin levels. Red blood cells were administered to 37% of patients, with 14% receiving plasma and 32% platelets. Only RBC transfusion, alone or combined with other blood products, was significantly associated with AKI. The transfusion of 1–2 RBC units increased the probability of stage 1 AKI by 4% and stage 2–3 AKI by 2% compared with patients not receiving RBCs. The risk was especially pronounced with the transfusion of > 2 units of RBCs, which raised the probability of stage 1 AKI by 12% and stage 2–3 AKI by 9%.

Conclusions

This study confirms previous findings that RBC transfusion is associated with postoperative AKI in cardiac surgery patients. The association was strongest among patients who received > 2 units of RBCs. Prospective studies are needed to determine the optimal strategies for transfusion in these patients and evaluate potential alternatives.