Restrictive vs. Liberal Blood Transfusion Strategies

Washington University Emergency Medicine Journal Club – September 16th, 2021


You’re working a shift on community medicine rotation in a medium-sized community-based ED one afternoon when you encounter Mr. S, a 62-year-old male here visiting his son from Florida, with a history of hypertension, hyperlipidemia, chronic iron-deficiency anemia, and coronary artery disease with a stent to his RCA ten years previously. He presents now with substernal chest heaviness that began while climbing the stairs at his son’s home and has persistent since. He took three sublingual nitroglycerin tablets with some relief, but is currently having 3/10 discomfort. He is not short of breath or diaphoretic. His physical exam is unremarkable. 

The patient’s ECG reveals sinus rhythm with very mild ST depression in the anterior leads and normal T-waves. You have no old records in your system for comparison. After one additional sublingual nitro he is now pain free. His labs begin coming back and reveal a high-sensitivity troponin T of 250 ng/L and a hemoglobin of 8.4 g/dL with a microcytic pattern. You go back and ask the patient about recent bloody stools or melena, both of which he denies. 

After giving the patient an appropriate dose of aspirin and starting him on a heparin drip for a non-ST elevation MI, you contact the on-call cardiologist with plans to admit to the hospitalist. The cardiologist agrees with the plan, but given the patient’s obvious myocardial infarction she suggests you transfuse the patient up to a hemoglobin of 10 g/dL. When you ask your attending, you are told this is not necessary and that there is prior evidence that transfusion to a level of 8 g/dL is sufficient, citing a study from 2013. You successfully admit the patient to the hospitalist without a transfusion, but wonder what that study actually showed and if there has been any additional evidence published in the interim. You also wonder what transfusion thresholds are appropriate for patients with active GI bleeding (a common cause of acute anemia) and high-demand states such as sepsis. After your shift, you decide to dive into the literature and see what the evidence shows… 

PICO Question

Population: Adult patients with anemia felt to require blood transfusion,
particularly those with gastrointestinal hemorrhage, sepsis/septic shock, and acute
coronary syndromes

Intervention: Transfusion based on a restrictive strategy with lower threshold
hemoglobin levels (≤ 7 g/dL, ≤ 8 g/dL)

Comparison: Transfusion based on a liberal strategy with higher threshold
hemoglobin levels (≤ 9 g/dL, ≤ 10 g/dL)

Outcome: Mortality, myocardial infarction, need for unscheduled revascularization,
stroke, number of units of blood transfused, adverse transfusion reactions, clinically
signiGicant volume overload, need for mechanical ventilation

Article 1: Carson JL, Brooks MM, Abbott JD, Chaitman B, Kelsey SF, Triulzi DJ, Srinivas V, Menegus MA, Marroquin OC, Rao SV, Noveck H, Passano E, Hardison RM, Smitherman T, Vagaonescu T, Wimmer NJ, Williams DO. Liberal versus restrictive transfusion thresholds for patients with symptomatic coronary artery disease. Am Heart J. 2013 Jun;165(6):964-971.e1. doi: 10.1016/j.ahj.2013.03.001. Epub 2013 Apr 8. PMID: 23708168; PMCID: PMC3664840. Answer Key.

Article 2: Ducrocq G, Gonzalez-Juanatey JR, Puymirat E, Lemesle G, Cachanado M, Durand-Zaleski I, Arnaiz JA, Martínez-Sellés M, Silvain J, Ariza-Solé A, Ferrari E, Calvo G, Danchin N, Avendaño-Solá C, Frenkiel J, Rousseau A, Vicaut E, Simon T, Steg PG; REALITY Investigators. Effect of a Restrictive vs Liberal Blood Transfusion Strategy on Major Cardiovascular Events Among Patients With Acute Myocardial Infarction and Anemia: The REALITY Randomized Clinical Trial. JAMA. 2021 Feb 9;325(6):552-560. doi: 10.1001/jama.2021.0135. PMID: 33560322; PMCID: PMC7873781. Answer Key.

Article 3: Odutayo A, Desborough MJ, Trivella M, Stanley AJ, Dorée C, Collins GS, Hopewell S, Brunskill SJ, Kahan BC, Logan RF, Barkun AN, Murphy MF, Jairath V. Restrictive versus liberal blood transfusion for gastrointestinal bleeding: a systematic review and meta-analysis of randomised controlled trials. Lancet Gastroenterol Hepatol. 2017 May;2(5):354-360. doi: 10.1016/S2468-1253(17)30054-7. Epub 2017 Mar 23. PMID: 28397699. Answer Key.

Article 4: Hirano Y, Miyoshi Y, Kondo Y, Okamoto K, Tanaka H. Liberal versus restrictive red blood cell transfusion strategy in sepsis or septic shock: a systematic review and meta-analysis of randomized trials. Crit Care. 2019 Jul 25;23(1):262. doi: 10.1186/s13054-019-2543-1. PMID: 31345236; PMCID: PMC6659290. Answer Key.

Bottom Line:

While current guidelines—including those from the American Association of Blood Banks, the American Association of Family Physicians, and the joint task force of EAST (Eastern Association for Surgery of Trauma) and the American College of Critical Care Medicine of the Society of Critical Care Medicine—recommend a restrictive blood transfusion strategy for most patients, there are specific patient populations whose physiology may affect such transfusion decisions. We chose to examine the literature on restrictive versus liberal blood transfusion strategies for patients with acute coronary syndrome, gastrointestinal hemorrhage, and sepsis/septic shock. 

An initial pilot randomized controlled trial conducted at 8 centers in the US compared liberal versus restrictive transfusion strategies for patients with acute coronary syndrome (ST-segment elevation myocardial infarction, non-ST segment elevation myocardial infarction, unstable angina, or stable coronary artery disease undergoing cardiac catheterization). While there was a strong trend toward fewer major adverse cardiac events (all-cause mortality, myocardial infarction,  or unscheduled revascularization within 30 days) with a risk difference (RD) of 15% (95% CI 0.7 to 29.3), this was a small (n = 110) pilot study whose results are not definitive. 

When this question was asked in a much larger RCT composed of 666 patients from 35 European centers, there was a trend toward decreased incidence of major adverse cardiac events in the restrictive versus the liberal transfusion group (relative risk [RR] 0.78, 1-sided 97.5% CI 0.00 to 1.17). As this was a non-inferioty study and the upper limits of the confidence interval did not cross the pre-specified non-inferiority threshold of 1.2, a restrictive strategy was deemed noninferior to a liberal strategy. Almost all of the patients in the liberal transfusion group received at least one unit of blood while only 35.7% of those in the restrictive group received any blood. There were also higher incidences of acute lung injury/ARDS, multiorgan system dysfunction, and infection in the liberal transfusion group, although these events were fairly rare. 

Patients with GI hemorrhage often present to the hospital with anemia or develop anemia while admitted. Given the risk of ongoing blood loss, transfusion thresholds in this patient population have also been questioned. A systematic review and meta-analysis from 2017 pooled results from five RCTs comprising 1965 total patients. Pooled risk of mortality was significantly lower in patients receiving restrictive transfusion strategies (RR 0.65, 95% CI 0.44 to 0.97). Risk of rebleeding was also lower in the restrictive patients, although this did not achieve statistical significance (RR 0.58, 95% CI 0.27 to 1.26). There were no differences in risk of acute MI, stroke, or acute kidney injury. Looking only at those patients with baseline ischemic heart disease, there was a statistically nonsignificant trend toward increased mortality with restrictive transfusion strategies (RR 4.28, 95% CI 0.86 to 22.31). Measured heterogeneity was low for overall risk of mortality (I2 = 0%) and for risk of rebleeding (I2 = 0%), although the pooled results are at high risk of clinical heterogeneity given differences in patient population and transfusion thresholds between studies. 

Another group in whom blood transfusion threshold requires special consideration is those with sepsis and septic shock. Given increased oxygen consumption and demand in these patients, and hence an increased need for oxygen delivery (Tuchschmidt 1991), higher transfusion thresholds could conceivably provide some benefit. In the original trial evaluating the efficacy of early goal-directed therapy  (EGDT) (Rivers 2011), patients with persistently low central venous oxygen saturation were transfused to a goal hematocrit of 30 percent. While EGDT has largely been abandoned in these patients, questions of transfusion threshold still need to be addressed. 

systematic review and meta-analysis on this topic from 2019 consisted of 3 RCTs with 749 total patients admitted to the ICU with sepsis or septic shock. Mortality at 28-30 days was similar in the restrictive and liberal transfusion groups (pooled OR 0.99, 95% CI 0.67-1.46), as was 60 and 90 day mortality (OR 0.91, 95% CI 0.55-1.51 and 0.85, 95% CI 0.49 to 1.47, respectively). There was no difference with regards to need for mechanical ventilation, vasopressor use, or need for renal replacement therapy. Again, there were issues with clinical heterogeneity as one study (TRICOP) included only cancer patients and the studies used different transfusion thresholds in the liberal transfusion arms. 

While far from perfect, the current literature supports restrictive blood transfusion strategies for patients with acute coronary syndrome, GI hemorrhage, and sepsis/septic shock. Given the nature of the intervention, patients and clinicians were not blinded in any of the studies (including those comprising the two meta-analyses) raising some risk of performance bias. Aside from this limitation and issues with heterogeneity in the meta-analyses, these studies appear to be at low overall risk of bias and should continue to inform clinical guidelines.