BACKGROUND: Bleeding into the chest is a life-threatening complication of cardiac surgery. Blood transfusion has been implicated as an important cause of harm associated with bleeding, based largely on studies demonstrating an independent association between transfusion and mortality. These studies did not, however, consider the possibility that bleeding may in itself be harmful, inasmuch as drains are inefficient at clearing blood from the chest and retained blood may compromise cardiac and lung function.
STUDY DESIGN AND METHODS: We undertook a multivariate logistic regression analysis of the risk factors associated with mortality in 2599 consecutive patients undergoing cardiac surgery. Unlike previous studies the risk factors examined included the volume of chest tube drainage at 24 hours. A stratified analysis was also undertaken that compared the adjusted risk of death for patients exposed or not exposed to a postoperative blood transfusion.
RESULTS: Blood transfusion was not an independent predictor of mortality (p = 0.4). Chest tube drainage was the strongest independent predictor of mortality (p < 0.001). In the stratified analysis, chest tube drainage remained an independent predictor of mortality for patients not exposed to a blood transfusion (p < 0.01). Furthermore, the risk of death of these patients was no different from patients exposed to a blood transfusion (p = 0.7 for interaction).
CONCLUSIONS: Our results argue that for patients undergoing cardiac surgery bleeding contributes to mortality through mechanisms unrelated to blood transfusion.