Marjon Borgert

185 Implementation of a transfusion bundle Table 3. Logic regression model for appropriate transfusions Univariate analysis OR (95% CI) P -value Baseline (reference) 1 Baseline – intervention a 2.26 (1.65 to 3.10) < 0.001 Age 1 (0.99 to 1.02) 0.26 Gender 1.35 (0.85 to 2.13) 0.20 Apache IV score 0.99 (0.99 to 1) 0.11 Pre transfusion Hb 0.25 (0.19 to 0.32) < 0.001 Transfusion delay (hours) 0.99 (0.91 to 1.07) 0.79 Medical/surgical admission 0.82 (0.59 to 1.15) 0.26 Dependent variable: transfusion according to protocol = 1 (Hb reference value), protocol deviation = 0; a Intervention: use of the transfusion bundle during (post) implementation periods Table 4. Multivariable analysis OR (95% CI) p-value Baseline – interventions* 2.05 (1.47 to 2.86) <0.001 Pre transfusion Hb 0.26 (0.20 to 0.34) <0.001 *Intervention: use of the transfusion bundle during (post) implementation periods DISCUSSION The main nding of the current study is that the introduction of a transfusion bundle has resulted in a signi cant reduction of the percentage of inappropriate transfusions. E ectively, using the transfusion bundle helps to improve compliance to transfusion guidelines in daily practice. Since the landmark of Hébert et al. attempts have beenmade to reduce the number of transfusions given the concerns about the safety of transfusion as well as the rising costs and shortage of blood products. 1,19,20 Reducing the number of inappropriate transfusions is expected to result in improved clinical outcomes and reduced health care costs. 21 The transfusion care bundle may therefore diminish costs by reducing waste of scarce and expensive RBC products.