Abstract 1114

Background:

Determination of the optimal hemoglobin concentration (Hb) for red cell (RBCs) transfusion is essential for patients undergoing cardiac surgery as mortality may be dependent on the severity of anemia and administration of RBCs. As there is considerable variability in transfusion practices, a pilot feasibility study is required to ensure that adherence to proposed strategies can be achieved before conducting a definitive study requiring thousands of patients. The primary objective of this study was to assess overall adherence to two transfusion strategies.

Methods:

We conducted a single centre parallel randomized controlled pilot trial to assess adherence to two transfusion strategies. The randomization sequence was created using block randomization by age and Cardiac Anesthesia Risk Score in groups of 4. High risk patients were allocated to a “restrictive” transfusion strategy (RBCs if their Hb was 70 g/L or less intraoperatively during cardiopulmonary bypass (CPB) and 75 g/L or less postoperatively following CPB); or a “liberal” transfusion strategy (RBCs if their Hb was 95 g/L or less during CPB and less than 100 g/L postoperatively). We defined adherence as adherence to the transfusion strategies for each patient for more than 90% of their days in hospital. We also assessed reasons for non-adherence and clinical outcomes. A total sample size of 50 patients was estimated to produce a 97% confidence interval equal to the sample adherence prevalence plus or minus 8% when the true prevalence of adherence was hypothesized to be 90%. We estimated our sample size using PASS 2002 software. The significance level was set at p<0.05 for all statistical analyses.

Results:

Fifty patients were randomized, 25 in the liberal group and 25 in the restrictive group. Overall, 99 units of RBCs were transfused in the liberal group compared 50 units in the restrictive group (p<0.01). Forty eight units of RBCs were administered intraoperatively for the liberal group compared to 14 RBC units for the restrictive group (p< 0.01). Ninety eight units were transfused in the liberal group according to the transfusion strategy compared to 42 units in the restrictive group (p=0.01). The median number of days of adherence to the transfusion strategy was 8 (interquartile range 12) in the restrictive arm and 6 days (interquartile range 3) in the liberal arm (Table, p<0.05). The proportion of patients with 100% adherence was higher in the liberal group i.e., more patients in the restrictive group received transfusions without meeting a transfusion criterion. This was also reflected in the proportion of red cells administered according to the transfusion strategy. The percent of adhered transfusion thresholds was 70% in the restrictive arm and 65% in the liberal arm. The adherence rate was highest in the operating room followed by the cardiovascular intensive care unit then the ward (Table). Reasons for non adherence included attending staff opinion that transfusion not indicated, patient refusal, the use of hemoconcentration to increase the hemoglobin concentration, hemorrhage, use of hematocrit instead Hb to transfuse the patient, confusion whether the transfusion strategy was to be applied on the ward, hyperkalemia, repetition of the complete blood count before transfusing the patient, and the staff was too busy.

Table:

Summary of Adherence Data

CharacteristicRestrictive Strategy (n=25)Liberal Strategy (n=25)
Proportion of patients with 100% adherence 80% 96% 
Proportion of RBCs administered according to the transfusion strategy:   
Operating room 100% 97.9 
Intensive care unit 78.8% 100% 
Ward 66.7% 100% 
Adherence to transfusion strategy, n (%) 35/50 (70%) 90/138 (65.2) 
Operating room 88.9% 80.4% 
Intensive care unit 67.6% 66.7% 
Ward 50% 11.5% 
CharacteristicRestrictive Strategy (n=25)Liberal Strategy (n=25)
Proportion of patients with 100% adherence 80% 96% 
Proportion of RBCs administered according to the transfusion strategy:   
Operating room 100% 97.9 
Intensive care unit 78.8% 100% 
Ward 66.7% 100% 
Adherence to transfusion strategy, n (%) 35/50 (70%) 90/138 (65.2) 
Operating room 88.9% 80.4% 
Intensive care unit 67.6% 66.7% 
Ward 50% 11.5% 
Conclusions:

This is the first randomized controlled trial that has assessed adherence rates to transfusion strategies in high-risk patients undergoing cardiac surgery. Determination of adherence rates to transfusion strategies is essential as adherence rates can impact outcomes if a considerable proportion of enrolled patients do not receive a transfusion according to their transfusion strategy. Poor adherence may impact outcomes despite the subsequent use of intention to treat analysis. Restrictive strategies can reduce transfusion, but further studies will be required to determine whether restrictive transfusions confer risk.

Disclosures:

Shehata:Canadian Blood Services: Employment, Research Funding.

This icon denotes an abstract that is clinically relevant.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution