Impact of red blood cell transfusion and hemoglobin threshold on 1-year mortality among surgical sepsis survivors: A propensity score matching study

Abstract

BackgroundDespite the fact that red blood cell (RBC) transfusion is commonly applied in surgical intensive care unit (ICU), the effect of RBC transfusion on long-term outcomes remains undetermined. We aimed to explore the association between RBC transfusion and the long-term prognosis of surgical sepsis survivors. MethodsThis retrospective study was conducted on adult sepsis patients admitted to a tertiary surgical ICU center in China. Patients were divided into transfusion and non-transfusion groups based on the presence of RBC transfusion. Propensity score matching (PSM) and inverse probability of treatment weighting (IPTW)were performed to balance the potential confounders. ResultsA total of 1421 surgical sepsis survivors were enrolled, including 403 transfused patients and 1018 non-transfused patients. There was a significant difference in 1-year mortality between the two groups (23.1 ​% vs 12.7 ​%, HR: 1.539, 95 ​% confidence interval [CI]: 1.030–2.299, P ​< ​0.001). After PSM and IPTW, transfused patients still showed significantly increased 1-year mortality risks compared to non-transfused individuals (PSM: 23.6 ​% vs 15.9 ​%, HR 1.606, 95 ​% CI 1.036–2.488 ​P ​= ​0.034; IPTW: 20.1 ​% vs 12.9 ​%, HR 1.600, 95 ​% CI 1.040–2.462 ​P ​= ​0.032). Among patients with nadir hemoglobin below 70 ​g/L, 1-year mortality risks in both groups were similar (HR 1.461, 95 ​% CI 0.909–2.348, P ​= ​0.118). However, among patients with nadir hemoglobin above 70 ​g/L, RBC transfusion was correlated with increased 1-year mortality risk (HR 1.556, 95 ​% CI 1.020–2.374, P ​= ​0.040). ConclusionFor surgical sepsis survivors, RBC transfusion during ICU stay was associated with increased 1-year mortality, especially when patients show hemoglobin levels above 70 ​g/L.