Background: Preoperative anaemia and perioperative transfusion are associated with adverse outcomes after cardiac surgery with cardiopulmonary bypass (CPB). Coronavirus disease 2019 (COVID-19) may induce persistent haematological and endothelial alterations. Whether previous infection modifies the relationship between haemoglobin and outcomes is unclear. This study evaluated whether previous COVID-19 infection modified the association between preoperative haemoglobin concentration and perioperative transfusion burden after cardiac surgery with CPB. Early postoperative complications were analysed as predefined secondary outcomes.

Methods: This study represents a secondary analysis of a prospective observational cohort of adult patients undergoing elective on-pump cardiac surgery between 1 August 2022 and 30 October 2023. Patients were categorised according to previous COVID-19 infection. Surgery was performed at least seven weeks after infection. The primary outcome was perioperative transfusion burden, defined as the total number of blood products administered from the intraoperative period until hospital discharge. An interaction term between previous COVID-19 and preoperative haemoglobin was included in multivariable linear regression.

Results: A total of 280 patients were included, of whom 101 (36.1%) had a previous COVID-19 infection. Preoperative haemoglobin was comparable between patients with and without a previous COVID-19 infection (13.4 [12.4–14.8] vs. 13.8 [12.2–14.6] g/dL; p = 0.472). No significant differences in early postoperative complications, transfusion rates, or transfusion burden were observed according to previous COVID-19 infection. The COVID-19 × haemoglobin interaction was not statistically significant in the adjusted model (B = 0.002, 95% CI −0.081 to 0.085; p = 0.958). Lower preoperative haemoglobin, lower baseline platelet count, longer CPB duration, and CKD were independently associated with greater transfusion burden. In an exploratory multivariable analysis, preoperative anaemia remained associated with postoperative AKI after adjustment for relevant covariates (adjusted OR 5.80, 95% CI 2.23–15.06; p < 0.001).

Conclusions: No statistically significant interaction between previous COVID-19 and preoperative haemoglobin was demonstrated for perioperative transfusion burden. Lower preoperative haemoglobin was independently associated with greater transfusion burden, while preoperative anaemia remained associated with postoperative AKI after multivariable adjustment in an exploratory analysis. These findings support systematic preoperative anaemia screening and Patient Blood Management in cardiac surgery.