Intraoperative Transfusion of 1 U to 2 U Packed Red Blood Cells Is Associated with Increased 30-Day Mortality, Surgical-Site Infection, Pneumonia, and Sepsis in General Surgery Patients

Among surgical patients receiving transfusions of packed RBCs (PRBCs), the incidence and degree of transfusion-related immunomodulation are related to, among other factors, the type of product transfused and patient parameters. Autologous donation and routine leukocyte filtration have reduced some of the adverse effects, and today most RBC transfusions in the United States are leukoreduced. The amount of transfused RBCs has a direct impact on negative clinical effects, with PRBCs known to increase morbidity and mortality in select patients. The impact of low-volume leukoreduced PRBC transfusion on patients undergoing general surgery is not completely understood. Bernard and colleagues used data from the 2005-2006 American College of Surgeons National Surgical Quality Improvement Program. From 125,223 records, they obtained 30-d morbidity and mortality, and demographic, preoperative, and intraoperative risk variables. Occurrence rates for outcomes were calculated by the amount of intraoperative transfusion and by whether postoperative transfusions were given and then compared with data on patients who were not given transfusions. Infectious complications and composite morbidity and mortality were stratified across the number of intraoperative PRBC units received. Propensity for transfusion, procedure type, wound class, operative duration and >30 patient risk factors were used for the determination of risk of outcomes. A transfusion risk scoring system was created by assigning point values to the odds ratio associated with each transfusion risk factor. Risk factor points were added together to create the risk index for each patient. Values were divided into ranges of 0-5, 6-10, 11-15, 16-20 and >20. A total of 125,177 records were complete for analysis; 4788 patients (3.8%) received intraoperative PRBCs. Risk variables that most predicted intraoperative transfusion included procedure, procedure group, ASA physical status class, >38, preoperative transfusion >4 U, emergent procedure, esophageal varices, and age. Patients with a risk index ≤5 had a transfusion rate of 0.3% compared with 75.3% for patients with a risk index of ≥20. Patients given 1U PRBCs had higher rates of infection at the surgical site, urinary tract infections, pneumonia, sepsis/shock, composite morbidity, and 30-d mortality. After adjustment for various factors, a 1-U PRBC transfusion significantly increased the risk of mortality, composite morbidity, pneumonia, and sepsis/shock but not infection. When 2 U was transfused, risk for these outcomes remained increased, along with an increased risk for surgical site infection. Postoperatively, transfusion of >4 U RBCs increased the risk of pneumonia, urinary tract infection, sepsis, composite morbidity, and mortality. Patients given PRBCs intraoperatively should be highly selected, with mild anemia or hypovolemia being tolerated. Blood conservation measures and appropriate indicators for transfusion should be used to avoid the deleterious effects of transfusion of even small amounts (1-2 U) of PRBCs.