Association between red blood cell transfusion and adverse clinical outcomes is Independent of cardiac history: a multicenter observational InPUT study analysis

PURPOSE: Red-blood-cell (RBC) transfusion is one of the most frequent interventions in critical care patients. While patients with acute cardiac conditions are more likely to receive transfusions at higher haemoglobin thresholds than other critically ill patients, data on RBC transfusion practice for critically ill patients with pre-existing cardiac conditions are scarce. METHODS: Using the International Point-Prevalence Study of Intensive-Care Unit Transfusion Practices cohort, weighted logistic regression investigated the association between the RBC units transfused and the primary composite outcome of 28-day mortality, new-onset acute kidney injury or ventilatory weaning failure. Interactions with cardiac history (acute coronary syndrome and/or heart failure) were tested. RESULTS: Cardiac history was present in 746 of 3643 patients (20%) and 894 of 3643 (25%) received at least one RBC unit. Transfusion rates were similar in patients with and without cardiac history (25% vs. 24%; p = 0.51). Among transfused patients, median nadir haemoglobin during ICU stay was slightly higher in those with cardiac history (7.6 g/dL vs. 7.4 g/dL respectively; p = 0.007), whereas stated haemoglobin transfusion threshold did not statistically differ (8.5 g/dL vs. 8.0 g/dL; p = 0.11). Each additional RBC unit increased the odds of the composite outcome in the whole cohort (2.18, 95% CI 1.85–2.56, p < 0.0001), without interaction with cardiac history (p = 0.44). CONCLUSIONS: RBC transfusion was commonly and similarly prescribed in critically ill patients with or without cardiac history. Each additional unit was associated with a worse outcome with no evidence of differential effect due to cardiac history. TRIAL REGISTRATION: NL9049 (Dutch Trial Register), registered on 16 November 2020.