10 Jul 06
Evidence-based red cell transfusion in the critically ill
By R Rana, B Afessa, MT Keegan, FX Whalen Jr, GA Nuttall, LK.Evenson et al
Critical Care Medicine. 2006;34:1892-1897
The implementation of evidence-based practice poses a significant challenge in the intensive care unit. In this quality improvement intervention we assessed the effect of an institutional protocol and computerized decision support for red cell transfusion in the critically ill.
Design
We compared processes of care and outcomes during the two 3-month periods before and after the introduction of a multidisciplinary quality improvement intervention.
Setting
Multidisciplinary intensive care units-medical, surgical, and mixed-in a tertiary academic center.
Patients
Consecutive critically ill patients with anemia (hemoglobin of <10 g/dL).
Intervention
Using the computerized provider order entry, we developed an evidence-based decision algorithm for red cell transfusion in adult intensive care units.
Measurements and Main Results
We collected information on demographics, diagnosis, severity of illness, transfusion complications, and laboratory values. The main outcome measures were number of transfusions, proportion of patients who were transfused outside evidence-based indications, transfusion complications, and adjusted hospital mortality. The mean number of red cell transfusions per intensive care unit admission decreased from 1.08 +/- 2.3 units before to 0.86 +/- 2.3 units after the protocol (p< .001). We observed a marked decrease in the percentage of patients receiving inappropriate transfusions (17.7% vs. 4.5%, p < .001). The rate of transfusion complications was also lower in the period after the protocol (6.1% vs. 2.7%, p = .015). In the multivariate analysis, protocol introduction was associated with decreased likelihood of red cell transfusion (odds ratio, 0.43; 95% confidence interval, 0.30 to 0.62). Adjusted hospital mortality did not differ before and after protocol implementation (odds ratio, 1.12; 95% confidence interval, 0.69 to 1.8).
Conclusions
The implementation of an institutional protocol and decision support through a computerized provider order entry effectively decreased inappropriate red cell transfusions.