Background: Strategies to restrict transfusions are gaining acceptance in critical care. We implemented an anemia management program (AMP) for trauma patients in the Surgical Intensive Care Unit. AMP was based on a transfusion trigger of 7 g/dL hemoglobin once hemodynamic sufficiency was achieved. We hypothesized that AMP would decrease the transfusion of packed red blood cells (PRBCs) and cost without detriment in clinical outcomes.
Methods: Transfusion data were retrospectively collected for all trauma patients treated in our Surgical Intensive Care Unit between July 2002 and December 2003. AMP was implemented in a step-wise fashion during a 6-month period (January to June 2003). Data were compared for the 6-month period before (Group I, July to December 2002) and after (Group II, July to December 2003) complete AMP implementation. Blood transfusion volumes were compared using negative binomial regression. Clinical outcomes (length of stay [LOS], death, myocardial infarction [MI], and ventilator-associated pneumonia [VAP]) were compared using risk ratios. Age, sex, and injury severity score (ISS) were examined as potential confounders.
Results: In all, 514 trauma patients were treated during the study period (n = 270 in Group I and n = 244 in Group II). Group I and Group II were similar in age (mean: 43.6 versus 42.9) and ISS (mean: 18.3 versus 17.0). Mean PRBCs per patient transfused decreased from 23.1 units to 17.1 units (p = 0.057), reflecting a 22.5% reduction adjusted for confounders (p = 0.097). Outcome data revealed no differences in LOS (mean: 6.4 versus 5.9, p = 0.920), risk of death (4.1% versus 6.1%, p = 0.158), or MI (0.7% versus 0.8%, p = 0.974), but a significant reduction in the incidence of VAP (8.1% versus 0.8%, p = 0.002). Total PRBC cost decreased during the study period from 503,000 dollars to 397,000 dollars.
Conclusions: An anemia management program appears to be safe when applied in the acute ICU phase of trauma care. Implementation of AMP in the ICU reduced the volume of PRBCs transfused with significant cost savings. No significant differences in length of stay, mortality rate, or MI rate were seen. The significant decrease in the rate of VAP requires further elucidation. Further long-term and larger studies are indicated.