Blood transfusion-associated sepsis during cardiogenic shock: a subanalysis from the SHARK-SHOCK registry
N Ghionzoli, A Stefanini, C Vandenbriele, H Soliman Aboumarie, C Sorini Dini, E Delcuratolo, C Stefanucci, M Barilli, R Nola, M Cameli, S ValenteAbstract
Background
Red blood cell transfusions (RBCTr) are commonly administered in cardiogenic shock (CS) to augment oxygen delivery, despite limited evidence supporting optimal transfusion thresholds in this setting. Current practice is largely extrapolated from heterogeneous populations and mostly guided by clinical judgement. Transfusion-related immunomodulation and infection risk may be particularly relevant in patients with severe haemodynamic compromise.
Purpose
To clarify whether a relationship exists between RBCTr administration and following development of sepsis in a heterogeneous population of overt CS.
Methods
We conducted a retrospective secondary analysis of a prospective cohort of 111 patients with Society for Cardiovascular Angiography and Intervention (SCAI) stages C-E CS referred to a single tertiary Italian Cardiac Intensive Care Units between 2021 and 2024. A fixed haemoglobin threshold of 8 g/dL for RBCTr was used. Sepsis was defined according to 2016 Surviving Sepsis Campaign criteria. Adverse events and outcomes related to sepsis and RBCTr were investigated, and a time-adjusted regression analysis was performed to establish their temporal relationship.
Results
Mean age of CS population was 62 ± 13 years, female-to-male ratio 4:1. Mean haemoglobin at admission was 12.4 ± 2.3 g/dL, 67% received at least 1 mechanical circulatory support (MCS) device over hospital course. Sepsis occurred in 32%, and 37% received at least one RBCTr. Sepsis was associated with longer median hospital stay (15 (6-35) vs 25 (12-57) days), higher continuous renal replacement therapy (26 vs 15%) and tracheostomy (9 vs 5%) requirements, and higher doses of noradrenaline (0.13 (0.07-0.25) vs 0.10 (0.05-0.20) µg/kg/min) (all P values < 0.05). A time-adjusted, significant relationship was found between RBCTr and further development of sepsis, even after adjusting for a comorbidity score, SCAI stage, the use of MCS, and invasive ventilation requirements (OR 4.44, 95% CI 1.98-9.92, P < 0.001).
Conclusions
In our analysis of patients with CS, RBCTr administration using a fixed cut-off was significantly associated with the risk of developing sepsis, even after adjustment for patient history and CS severity, and sepsis was associated with worse, non-mortality multiple clinical outcomes. Further randomised clinical trials are advocated to determine if stricter haemoglobin threshold is safe and can reduce RBCTr requirement and comorbidity, including sepsis.For image description, please refer to the figure legend and surrounding text.For image description, please refer to the figure legend and surrounding text.