Red blood cell transfusion and outcomes in acute pulmonary embolism.

Abstract

Blood transfusion has been associated with adverse outcomes in certain conditions. This study investigates the prevalence and outcomes of red blood cell (RBC) transfusion in patients with acute pulmonary embolism (PE). Retrospective study of consecutive patients from 2000 to 2012 admitted to a tertiary hospital with a primary diagnosis of acute PE. Transfusion status during the hospital admission was ascertained. Mortality was tracked from a state-wide death database and analysed using multivariable modelling. A total of 73 patients (5% of all patients admitted with PE) received RBC transfusion during their admission. These patients were significantly older, had more co-morbidities, worse haemodynamics, higher simplified pulmonary embolism severity index scores, and lower plasma sodium and haemoglobin (Hb) levels at admission. Unadjusted mortality for the transfused group was significantly higher at 30-day (19% vs 4%, P < 0.001) and 6-month (40% vs 10%, P < 0.001) follow-up. Multivariable modelling showed RBC transfusion to be a significant independent predictor of mortality at 30-day (odds ratio 3.06, 95% CI: 1.17-8.01, P = 0.02) and 6-month (hazard ratio (HR) 1.97, 95% CI: 1.12-3.46, P = 0.02). Sensitivity analysis confirmed that transfused patients had higher mortality than non-transfused patients in the subgroup of patients with Hb <100 g/L. RBC transfusion in patients hospitalized with acute PE is rare and appears to be associated with increased risk of short- and long-term mortality, independent of Hb level on admission. This finding underscores the need for future randomized controlled studies on the impact of RBC transfusion in the management of patients admitted with acute PE. [Correction added on 4 May 2018, after first online publication: the word 'serum' was changed to 'plasma' throughout the article where appropriate.].