Variables predicting trauma patient survival following massive transfusion.

INTRODUCTION The literature contains little information regarding demographic or transfusion-related factors associated with survival following massive blood transfusion in trauma patients. The objective of this study was to describe patient, transfusion, and laboratory variables contributing to survival in this population during the first and second days after arrival at the hospital. A secondary objective was to identify costs associated with massive blood transfusion. METHODS A 7-year, retrospective review of 13,005 consecutive trauma patient records yielded a sample of 46 who were transfused with > or =50 units of blood products in the first postinjury day. Descriptive statistics were computed to describe the sample, transfusion data, and laboratory values. Logistic regression was used to predict survival using selected patient characteristics, laboratory data, and transfusion characteristics for both the first and second days. RESULTS Overall survival among this group who received massive transfusion was 63%. No significant differences were found between survivors and nonsurvivors in age, sex, type of trauma, or amount of any of the blood components administered on Day 1. Nonsurvivors had higher Injury Severity Scores and shorter ICU and hospital lengths of stay. Controlling for other variables, only arterial base deficit levels made a significant unique contribution to predicting survival. The volume of blood transfused on Day 2 did not contribute to survival prediction. The average cost of blood transfusion was more than 49,000 US dollars per survivor and 51,000 US dollars per nonsurvivor. CONCLUSION Defining medical futility based solely on the volume of blood products transfused currently is unjustified. The search for other early indicators of survival in the trauma population must continue.