The Decision to Transfuse: One Size Might Not Fit All*

    loading  Checking for direct PDF access through Ovid


Although the transfusion of RBCs has been a mainstay of clinical practice for over 60 years, we still do not have clear evidence for which patients benefit from transfusion and under what circumstances. Since 1999, when the results of the landmark Transfusion Requirements in Critical Care study were published (1), more than 20 randomized controlled trials have been performed comparing outcomes among patients assigned to different transfusion strategies (2, 3) and citations therein. In most of these studies, the “liberal” transfusion strategy targeted a transfusion threshold at 9–10 g/dL, whereas the more “restrictive” transfusion strategy specified a transfusion threshold at 7–8 g/dL. In general, these studies found relatively few differences in the outcomes of the patients in the two transfusion arms. Some outcome differences were seen, but the specific endpoints which appeared to be affected by the transfusion strategy were not consistent between studies, possibly due to differences in the patient populations studied. Two meta-analyses have addressed this point by grouping these trials by patient category. One meta-analysis found no differences in mortality between subjects in the liberal and restrictive arms over several categories (2). A second meta-analysis observed differences in the composite outcome measure constructed for the meta-analysis, which included 30-day mortality and seven other ischemic and adverse events. Better outcomes were observed for the subjects in the liberal transfusion arm for both cardiovascular and orthopedic patients, but this was not the case for ICU patients. Thus, there continues to be an unmet need for more evidence to help guide clinicians on appropriate transfusion strategies.
Few studies of this nature have been conducted in patients with solid organ malignancies (4), or those with hematologic malignancies and myelodysplastic syndromes (5, 6). These patients often require extensive transfusion support (7). A study reported in this issue of Critical Care Medicine takes a much-needed step to address this gap (8). The Transfusion Requirements in Critically Ill Oncologic Patients (TRICOP) Study is a single center, randomized, double-blind, controlled trial comparing outcomes among subjects assigned to a liberal (< 9 g/dL) or restrictive (< 7 g/dL) RBC transfusion strategy. Adult patients with solid organ malignancies who developed septic shock within 6 hours of admission to the ICU were included. The number of units of RBC transfused to subjects in the restrictive arm was lower than in the liberal arm as was the proportion of patients receiving RBC. There was no statistically significant difference between the two arms of the study for the primary outcome measure (28-d mortality) although there was a trend toward increased mortality in the restrictive arm (45.0% vs 55.6%; p = 0.08). There were also no differences in most of the secondary outcomes including 60-day mortality, the need for various interventions, occurrence of several ischemic events or other severe adverse events, hospital and ICU length of stay, or hospital readmission rate. There was however a statistically significant difference in mortality at 90 days again favoring the liberal arm (59.1% vs 70.2%; p = 0.03).
This study was well-designed and achieved a high degree of internal validity. The patients in the two arms of the study were well-balanced for baseline characteristics, and compliance with the protocol was excellent with greater than or equal to 95% of subjects being transfused as assigned with no indication that protocol deviations were significantly more frequent in one of the arms. Good separation between the mean hemoglobin levels of the two arms was achieved over the 7-day observation period (approximately 0.5–2.0 g/dL among transfused patients).
    loading  Loading Related Articles