Transfusion insurgency: practice change through education and evidence-based recommendations

Abstract Background In 2000, we implemented an evidence-based guideline in the surgical intensive care unit (SICU) using a transfusion threshold of hemoglobin Methods We analyzed 2-month samples of admissions in even-numbered years from 1998 to 2006. Any infusion of packed red blood cells (PRBCs) was included. Results We analyzed data from 2,138 patients resulting in 5,130 transfusions. Thirty-six patients received >20 U of blood. The only difference between groups occurred in 2006 when renal failure increased. Transfusions decreased from 3.2 ± 0.34 (SE) to 1.7 ± 0.2. The number of patients who received blood also decreased. Mortality and length of stay (LOS) were not different among the groups. Every unit of blood transfused increased the mortality risk by 14%. Conclusions Implementation of an evidence-based transfusion guideline reduced the number of infused units and patients transfused without an increase in mortality.