Jason Alexander, MD; Adam S. Cifu, MD

Summary of the Clinical Problem

Red blood cell transfusion is a common and potentially life-saving intervention, yet balancing the harms, benefits, scarcity of blood products, and cost remains complex. More than 13 million units of RBCs were transfused in the United States in 2013.1 Transfusions are safe, with only 0.24% accompanied by an adverse reaction, and life-threatening transfusion reactions occur at a rate of only 15.1 per 1 million blood components transfused.2 The average cost paid by hospitals to transfuse 1 RBC unit is $225.42, while the reimbursement rate from the Centers for Medicare & Medicaid Services is $194.86, resulting in a net loss to hospitals of 13.6%.

Characteristics of the Guideline Source

The guideline was developed by the UK NCGC on behalf of NICE, which funded and supported creation of the guideline.3 The NCGC assembled a guideline development group (GDG), a multidisciplinary team of health professionals, researchers, and lay members, to lead the process. All GDG members completed conflict-of-interest forms stating potential financial, business/professional, and intellectual conflicts at the start of guideline development and at all subsequent meetings. Members were required to partially or completely withdraw from the discussion if a declared conflict of interest was related to the review question being addressed. Following completion, the guideline underwent a peer review process for 6 weeks in which all comments from registered stakeholders were addressed and displayed on the NICE website (Table).

Evidence Base

The GDG identified 3 specific review questions related to RBC transfusions and completed a comprehensive literature review. Randomized clinical trials and systematic reviews were prioritized. Clinical questions were appraised using GRADE evidence profiles, and economic evidence profiles were generated to summarize cost and cost-effectiveness where available. All searches were updated prior to finalizing the guidelines to include the most current data. Final recommendations were reported in conjunction with standards set by the NICE guidelines manual, where strong recommendations were denoted by “offer” or “do not offer” and weaker recommendations were conveyed by “consider.”4

Thirty-four studies pertaining to RBC transfusions involving 17 553 patients were analyzed by the GDG. Risk ratios comparing restrictive (generally defined as a transfusion threshold of 7-9 g/dL) and liberal (generally defined as a transfusion threshold of 8-10 g/dL) transfusion strategies from the meta-analysis associated with this guideline demonstrated no significant difference in 30-day mortality (risk ratio, 0.95; 95% CI, 0.77-1.17), new cardiac events (risk ratio, 1.00; 95% CI, 0.54-1.83), or infections (risk ratio, 0.92; 95% CI, 0.83-1.01) and no significant difference in length of hospital stay (risk ratio, −0.52; 95% CI, −1.11 to 0.06, where 0 favors neither a restrictive nor a liberal transfusion strategy and negative numbers favor a restrictive strategy). Restrictive strategies resulted in fewer patients receiving an RBC transfusion (risk ratio, 0.65; 95% CI, 0.59-0.73), and when transfusions were required, fewer units were administered (mean difference, −1.13; 95% CI, −1.67 to −0.59).

Higher transfusion targets were found to confer some benefit in patients with coronary artery disease. One small trial randomized 110 patients with ACS or stable angina undergoing catheterization to either a liberal (transfusion for hemoglobin <10 g/dL) or restrictive (transfusion for hemoglobin <8 g/dL) transfusion strategy and demonstrated reduced 30-day mortality (1.8% vs 13%; P = .03) with the liberal strategy.5 A larger, more recent trial assessing postoperative management of anemia in 2003 patients following cardiac surgery also demonstrated higher mortality in patients receiving a restrictive (transfusion for hemoglobin <7.5 g/dL) vs a liberal (transfusion for hemoglobin <10 g/dL) transfusion strategy (4.2% vs 2.6%; hazard ratio, 1.64; 95% CI, 1.00-2.67; P = .045).6

Benefits and Harms

The guideline’s recommendations could help reduce RBC transfusions that are unlikely to improve patient outcomes. Expected benefits include limiting exposure to transfusion-associated risks, reduced health care costs, and greater availability of blood products for situations in which transfusions are indicated. Extrapolating these recommendations to patients not covered by this guideline, including pregnant patients or patients with sickle cell disease, may be harmful. Significant uncertainty remains regarding how best to manage RBC transfusions in these patients.

Discussion

Anemia is commonly encountered in clinical practice. Establishing evidence-based guidance regarding transfusion is warranted. The available evidence suggests that for anemic patients without ACS or major hemorrhage, a higher hemoglobin goal confers no significant clinical benefit and an increased risk of harm. This finding held true even among critically ill patients. It is reasonable to conclude that a hemoglobin threshold of 7 g/dL for patients meeting the criteria for restrictive transfusion does not lead to increased mortality or cardiac events and can be utilized in clinical practice. Patients with coronary artery disease should be managed with a higher hemoglobin goal.

The recommendations of the GDG differ somewhat from those released by the American Association of Blood Banks (AABB).7 Following a systematic review of 31 randomized clinical trials that included 12 587 patients, the AABB advises the same restrictive transfusion threshold of 7 g/dL for hemodynamically stable patients; however, this does not include patients undergoing cardiac or orthopedic surgery or who have preexisting cardiovascular disease, in whom a transfusion threshold of 8 g/dL is recommended. Furthermore, because of insufficient evidence, the AABB gives no guidance for transfusion thresholds in patients with ACS or chronic transfusion-dependent anemia or hematology/oncology patients with severe thrombocytopenia who are at risk of bleeding. These differences highlight the need for well-designed trials that include these patient populations.

Areas in Need of Future Study or Ongoing Research

There is limited evidence regarding optimal transfusion thresholds and targets in patients with chronic cardiovascular disease, although there is some suggestion liberal transfusion strategies may be of benefit for these patients. A post hoc analysis of patients with ischemic coronary disease in the TRICC trial8 as well as an a priori subgroup of patients with cardiovascular disease in critically ill patients with septic shock9 both showed benefits when hemoglobin thresholds for transfusion were higher, but well-designed studies will need to be performed before any significant conclusions can be made. These findings suggest that transfusions threshold and goals may vary according to patient population, and studies of these populations are warranted.

 

 

JAMA