Special Communication 

October 12, 2023

Red Blood Cell Transfusion: 2023 AABB International Guidelines

Jeffrey L. Carson, Simon J. Stanworth, Gordon Guyatt, et al

JAMA. Published online October 12, 2023. doi:10.1001/jama.2023.12914

Abstract

Importance  Red blood cell transfusion is a common medical intervention with benefits and harms.

Objective  To provide recommendations for use of red blood cell transfusion in adults and children.

Evidence Review  Standards for trustworthy guidelines were followed, including using Grading of Recommendations Assessment, Development and Evaluation methods, managing conflicts of interest, and making values and preferences explicit. Evidence from systematic reviews of randomized controlled trials was reviewed.

Findings  For adults, 45 randomized controlled trials with 20 599 participants compared restrictive hemoglobin-based transfusion thresholds, typically 7 to 8 g/dL, with liberal transfusion thresholds of 9 to 10 g/dL. For pediatric patients, 7 randomized controlled trials with 2730 participants compared a variety of restrictive and liberal transfusion thresholds. For most patient populations, results provided moderate quality evidence that restrictive transfusion thresholds did not adversely affect patient-important outcomes. Recommendation 1: for hospitalized adult patients who are hemodynamically stable, the international panel recommends a restrictive transfusion strategy considering transfusion when the hemoglobin concentration is less than 7 g/dL (strong recommendation, moderate certainty evidence). In accordance with the restrictive strategy threshold used in most trials, clinicians may choose a threshold of 7.5 g/dL for patients undergoing cardiac surgery and 8 g/dL for those undergoing orthopedic surgery or those with preexisting cardiovascular disease. Recommendation 2: for hospitalized adult patients with hematologic and oncologic disorders, the panel suggests a restrictive transfusion strategy considering transfusion when the hemoglobin concentration is less than 7 g/dL (conditional recommendations, low certainty evidence). Recommendation 3: for critically ill children and those at risk of critical illness who are hemodynamically stable and without a hemoglobinopathy, cyanotic cardiac condition, or severe hypoxemia, the international panel recommends a restrictive transfusion strategy considering transfusion when the hemoglobin concentration is less than 7 g/dL (strong recommendation, moderate certainty evidence). Recommendation 4: for hemodynamically stable children with congenital heart disease, the international panel suggests a transfusion threshold that is based on the cardiac abnormality and stage of surgical repair: 7 g/dL (biventricular repair), 9 g/dL (single-ventricle palliation), or 7 to 9 g/dL (uncorrected congenital heart disease) (conditional recommendation, low certainty evidence).

Conclusions and Relevance  It is good practice to consider overall clinical context and alternative therapies to transfusion when making transfusion decisions about an individual patient.

Introduction

Red blood cell (RBC) transfusion is a common and costly treatment; approximately 118 million units of blood are collected worldwide each year.^1,2 Clinicians should offer RBC transfusion to patients only when benefits outweigh harms. Harms include infectious and noninfectious complications; although serious reactions are infrequent, there remains potential for substantial harm (Table 1).^3,4 Patient advocacy groups support minimizing harms by avoiding transfusions without clear benefit.⁵

Although the average acquisition cost of a unit of RBCs is $215 in the United States,^6,7 it varies by country and region. Acquisition costs do not typically cover expenses of distribution, storage, processing, administration, and monitoring for complications.^7,8 Many blood transfusion providers face challenges, exacerbated by the COVID-19 pandemic, in maintaining adequate stocks of RBCs.⁹

Randomized controlled trials (RCTs) assessing outcomes of different transfusion thresholds typically compare higher hemoglobin thresholds (liberal transfusion strategy) with lower ones (restrictive transfusion strategy) for RBC transfusions. The numbers of these trials continue to increase. AABB guidelines in 2012 included 19 RCTs; in 2016, 31 RCTs.^10,11 In 2018, the Transfusion and Anemia Expertise Initiative published guidelines based on 5 RCTs for RBC transfusion in critically ill children.¹² In 2021, an updated Cochrane systematic review included 48 trials.¹³ Given the expanded evidence base and the prior absence of AABB guidelines specific to children, we reexamined the transfusion threshold evidence and provide updated guidance.

Guideline Development Process

The AABB commissioned and funded updated guidelines through the AABB Clinical Transfusion Medicine Committee. To encourage wide implementation of the recommendations, the board of directors supported recruiting experts in RBC transfusion from international professional organizations (eAppendix in the Supplement). These recommendations were developed in collaboration with and are endorsed by the International Society of Blood Transfusion, International Collaboration for Transfusion Medicine Guidelines, the Society of Critical Care Medicine, the European Blood Alliance, and the Society for the Advancement of Patient Blood Management.

These guidelines follow existing standards of trustworthiness,¹⁴ including use of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach for summarizing evidence and moving from evidence to recommendations¹⁵ to provide credible recommendations for clinicians caring for adults and children considered for RBC transfusions. These guidelines do not address transfusion in preterm neonates.

Perspective

The panel chose individual patients as the primary perspective but also considered public health considerations; for example, supply of blood.

Panel Composition and Conflicts

The international panel included members with expertise in transfusion medicine, supported by a GRADE methodologist (G.G.) and a patient partner (A.D.) (eAppendix in the Supplement). In accordance with AABB policy, individual members disclosed all potential financial, professional, or personal conflicts of interest; none had substantive conflicts.¹⁶Five members were authors of trials included in a systematic review on transfusion thresholds (J.L.C., S.J.S., Y.L., C.S.-O., and E.M.W.) and did not vote on corresponding recommendations.

Population, Intervention, Comparator, and Outcomes Questions

We provide recommendations for 2 questions:

1. For hospitalized, hemodynamically stable adult patients, should clinicians transfuse with a restrictive strategy (typical hemoglobin level <7-8 g/dL) vs a liberal strategy (typical hemoglobin level <9-10 g/dL)?
2. For hospitalized, hemodynamically stable pediatric patients (a) without congenital heart disease (infancy to 16 years), should clinicians transfuse with a restrictive strategy (hemoglobin level <7-8 g/dL) vs a liberal strategy (hemoglobin level <9-10 g/dL); and (b) with congenital heart disease, should clinicians transfuse with a restrictive vs liberal strategy based on the cardiac lesion?

We provide recommendations for patients with acute or prolonged need of transfusions, but not for those who are transfusion dependent (eg, hemoglobinopathies). For adults, we examined subgroups in which the harm and benefit of a particular transfusion threshold might differ from that of overall populations: preexisting coronary artery disease, cardiac surgery, orthopedic surgery, and oncologic or hematologic conditions.

We examined subgroups of children in whom the risk and benefit of transfusion threshold might differ from that of the overall populations of patients: those with heart disease (congenital or acquired) or surgery and hematologic or oncologic conditions. We excluded trials of preterm neonates, which have been reviewed elsewhere.¹⁷

Values and Preferences

Recommendations are based on the following values and preferences:

• Avoid the adverse effects after RBC transfusion (high value).
• Conserve resources related to RBC transfusions (high value) to ensure blood is available for individuals who need it most.
• Prefer the demonstrated benefits of a restrictive transfusion policy despite the remaining possibility of a small increase in mortality.

Comments and Modification

J.L.C., S.J.S., G.G., S.V., and M.B.P. prepared the draft guideline document that was modified and approved by all panel members and the AABB Clinical Transfusion Medicine Committee. Subsequently, the AABB board of directors and international partner organizations also reviewed the guidelines.

Evidence Review and Grading

Systematic Review

We developed recommendations based on recently published systematic reviews of transfusion thresholds in adults (Cochrane review conducted in 2021)¹³ and children (Transfusion and Anemia Expertise Initiative, 2018),¹² supported by literature searches up to February 2021. We reviewed evidence from 45 RCTs with 20 599 adults, 5 RCTs identified within the Transfusion and Anemia Expertise Initiative in 2018, and 2 additional pediatric trials (the 5 RCTs and 2 pediatric trials had a total of 2730 participants).^18-20 The systematic reviews included RCTs in which the transfusion groups were assigned based on a clear transfusion threshold, described as the hemoglobin concentration or hematocrit level required before RBC transfusion. Outcomes in adults included 30-day mortality, nonfatal myocardial infarction, pulmonary edema or congestive heart failure, stroke, thromboembolism, acute kidney injury, infection, hemorrhage, mental confusion, proportion of patients with an allogeneic or autologous RBC transfusion, hemoglobin concentration (postoperative or discharge), number of RBC units transfused, and quality of life. An updated search conducted in January 2023 identified 3 trials with 151 patients.^21-23 For children, outcomes included mortality, thromboembolism, infection, and transfusion requirements.

Analysis

We assessed risk of bias in each RCT as recommended by Cochrane,²⁴ assessed statistical heterogeneity by both I² and χ² tests,²⁵ and used the Instrument to Assess the Credibility of Effect Modification Analyses criteria for making inferences regarding subgroup effects.²⁶ All analyses were performed with Review Manager version 5.4 (Cochrane Collaboration).²⁷ Relative risks and the corresponding 95% CIs were calculated for each outcome with random-effects models²⁸ unless counterintuitive results mandated use of a fixed-effect model. We calculated absolute risks by applying the relative effect to the median of control group risks. When events were anticipated to be rare (eg, for thromboembolism), the Peto odds ratio informed relative effect estimates.

Rating Quality of Evidence and Making Recommendations

We used GRADE methodology to develop these guidelines (see the Supplement).^15,29 The panel came to consensus for quality of evidence ratings that were included in summary of findings tables that served as the bases for panel judgments.³⁰ In moving from evidence to recommendations, the panel considered criteria in GRADE’s evidence to decision framework.³¹ The panel came to consensus for all recommendations except for using different restrictive strategy thresholds by clinical subgroup in which a vote was required.

Good Practice Statement

In deciding when a particular patient should undergo transfusion, the panel considers it good clinical practice to consider not only the hemoglobin concentration but also symptoms, signs, other laboratory data, patients’ values and preferences, and the overall clinical context. Relevant variables include the rate of hemoglobin level decline, intravascular volume status, dyspnea, decreased exercise tolerance, lightheadedness, chest pain thought to be cardiac in origin, and hypotension or tachycardia unresponsive to fluid challenge. Clinicians should consider alternatives to transfusion, including medical treatment of anemia and blood conservation strategies.

Disclaimer

This practice guideline will not apply to all individual RBC transfusion decisions.

Recommendations for Adults

Recommendation 1

For hospitalized adult patients who are hemodynamically stable, the international panel recommends a restrictive RBC transfusion strategy in which the transfusion is considered when the hemoglobin concentration is less than 7 g/dL (strong recommendation, moderate certainty evidence).

Remark: in accordance with the restrictive strategy threshold used in most of the trials for subgroups of patients, clinicians may choose a threshold of 7.5 g/dL for patients undergoing cardiac surgery and 8 g/dL for patients undergoing orthopedic surgery or those with preexisting cardiovascular disease.

Recommendation 2

For hospitalized adult patients, the panel suggests a restrictive RBC transfusion strategy in which transfusion is considered when the hemoglobin concentration is less than 7 g/dL in those with hematologic and oncologic disorders (conditional recommendation, low certainty evidence).

Evidence Summary for Adults

The 45 RCTs with adult participants were conducted across a range of settings, including orthopedic surgery (n = 11), cardiac surgery (n = 8), hematologic and oncologic conditions (n = 7), critical care (n = 8), acute blood loss (n = 6), acute myocardial infarction (n = 3), and vascular surgery (n = 2). The most common liberal transfusion threshold was 9 to 10 g/dL and the most common restrictive threshold was 7 to 8 g/dL.

Table 2 presents the summary of findings comparing restrictive with liberal transfusion strategies for 30-day mortality, multiple morbidities, and transfusion requirements. Thirty trials including data from 16 092 participants evaluated 30-day mortality, with a pooled relative risk of 1.00 (95% CI, 0.86-1.16). The baseline mortality rate was 8.3%, and an absolute difference between transfusion strategies was 0% (95% CI, 1.2% fewer to 1.3% more deaths) (high certainty). The restrictive strategy resulted in a 32.4% absolute reduction (95% CI, 37.3%-27.5% fewer deaths) in receiving a transfusion.

Chance may explain differences in mortality estimates among the clinical conditions (test for subgroup differences, P = .34). Given limited trial data in hematologic malignancies (2 trials, N = 149 participants) and an upper CI limit consistent with substantial harm (6.2% rate of increased deaths in the restrictive transfusion strategy), certainty of the evidence for mortality in this population was rated low (Table 3). Given heterogeneity in results and an upper CI limit consistent with substantial harm (4.4% rate of increased deaths in the restrictive transfusion strategy), the certainty of the evidence was rated low for mortality in acute myocardial infarction (Table 3).

There were no apparent differences between transfusion strategies for the morbidity outcomes (Table 2). Data from 3 RCTs that enrolled 448 participants suggested the risk of bleeding in hematology and oncology patients was uninfluenced by transfusion strategy (relative risk, 1.03; 95% CI, 0.87 to 1.23; absolute difference, 0.6%; 2.7% fewer to 4.8% more bleeding events).^32-34

The most common restrictive transfusion strategy applied in the trials was 7 or 8 g/dL (Figure), although variations included critical care and cardiac surgery trials that used a transfusion strategy of 7 to 7.5 g/dL and orthopedic and acute myocardial infarction trials that used a restrictive strategy of 8 g/dL.^36-64

Rationale for Recommendations for Adults

The panel recommends that RBC transfusion be administered using a restrictive transfusion strategy of 7 g/dL for most hemodynamically stable adults (strong recommendation, high certainty evidence).

The panel was divided (by vote) on whether to recommend different restrictive transfusion strategy thresholds by clinical subgroup. The rationale for recommending a universal threshold of 7 g/dL is that many trials used this threshold, and there is no strong clinical or biological basis for expecting different effects between 7 and 8 g/dL (with the possible exception of cardiovascular disease and hematology or oncology; see later). Furthermore, the effects on mortality were consistent across all subgroups, and there were no apparent differences in outcomes between trials that used a threshold of 7 and 8 g/dL (see earlier) (Figure). Recommending a hemoglobin threshold of 7 g/dL would conserve more blood.

An alternative view is that the recommendations should closely follow the clinical trial evidence and avoid extrapolating trial results when a threshold of 7 g/dL has not been explicitly tested. Most of the trials in orthopedic surgery used a threshold of 8 g/dL, and the largest trial conducted in cardiac surgery used a threshold of 7.5 g/dL. Some members of the panel thought that higher hemoglobin thresholds might improve outcomes other than mortality, including improved function and recovery after surgery or acute illness.

For patients with acute and chronic ischemic cardiac disease, there remains substantial uncertainty regarding the safety of restrictive thresholds. As in the AABB’s previous guidelines,^10,11 the panel chose not to recommend for or against a liberal or restrictive transfusion threshold for patients with acute myocardial infarction. Although the pooled estimates of effects on mortality with acute myocardial infarction were almost identical to the overall effects, the absolute and relative risk estimates were imprecise, with wide CIs. The panel noted that the MINT trial (including 3500 participants with acute myocardial infarction) is nearing completion. MINT compares a liberal transfusion at 10 g/dL with a restrictive transfusion strategy of 7 to 8 g/dL.⁶⁵

In the setting of hematology and oncology inpatients, the panel suggests transfusion at 7 g/dL (conditional, low certainty evidence). Although the number of patients enrolled in these trials was smaller than that in many other clinical subgroups, because new RCTs have suggested neither harm nor increased bleeding when using a restrictive threshold, this recommendation differs from the 2016 guidelines.¹¹ There were insufficient trial data to inform recommendations in outpatient transfusion management.

Recommendations for Children

Recommendation 3

For critically ill children and hospitalized children at risk of critical illness who are hemodynamically stable and without a transfusion-dependent hemoglobinopathy, cyanotic cardiac condition, or severe hypoxemia, the international panel recommends a restrictive transfusion strategy in which a transfusion is considered when the hemoglobin level is less than 7 g/dL compared with one of less than 9.5 g/dL (strong recommendation, moderate certainty evidence).

Recommendation 4

The international panel suggests considering a transfusion threshold for hemodynamically stable children with congenital heart disease that is based on the cardiac abnormality and stage of surgical repair: 7 g/dL (biventricular repair), 9 g/dL (single-ventricle palliation), or 7 to 9 g/dL (uncorrected congenital heart disease) (conditional recommendation, low certainty evidence).

Evidence Summary for Children

The populations of children included in the RCTs were critically ill patients (n = 2),^20,35those with hematologic conditions (n = 1),⁶⁶ those with acquired and congenital heart disease (n = 3),^67-69 and those with severe (malarial) anemia (n = 1)^18,19 (Table 4). The largest single intensive care unit RCT reported a 51.8% absolute reduction in transfusions in the restrictive strategy group compared with the liberal strategy group,³⁵ with no significant difference reported for 30-day mortality within a meta-analysis of 5 RCTs (relative risk, 0.44; 95% CI, 0.04-4.45). In the latter analysis, the baseline mortality rate was 3.9%, with an absolute difference of 1.7% (95% CI, 0.2% fewer to 17.5% more deaths) (moderate certainty). There were no clear differences in the morbidity outcomes (Table 4). We evaluated the transfusion strategies on 30-day mortality in subgroups of heart disease (acquired and congenital) (eFigure 12 in the Supplement). Chance may explain differences in mortality among the clinical populations. The certainty of the evidence was rated as low because of small sample size and various surgical settings and clinical conditions.

Rationale for Recommendations for Children

It is likely that mortality is similar for restrictive strategies compared with liberal ones (moderate certainty, rated down because of inconsistency and the remaining possibility of an increase in 30-day mortality after application of a restrictive strategy of up to 3%).

Although the direct evidence was dominated by a single trial,³⁵ a large well-conducted RCT of transfusion volumes and timing in anemic children (hemoglobin level <6 g/dL) with malaria also supported the safety of a restrictive transfusion threshold. The panel concluded this evidence supported a strong recommendation.^18,19

Children with acquired or congenital heart disease form a subgroup in which there remains uncertainty regarding the pathophysiologic safety of restrictive thresholds, and the RCTs had recruited different populations of children undergoing surgery.

Discussion

The expanding number of RCTs of RBC transfusion thresholds informs best practice in adults and children. Many of the RCTs tested different protocols including thresholds for RBC transfusion that varied by clinical setting. The panel debated whether to recommend a threshold of 7 g/dL for all hemodynamically stable adults or adopt a higher threshold in select clinical subgroups (cardiac surgery, 7.5 g/dL; orthopedic surgery and chronic cardiovascular disease, 8 g/dL), ultimately concluding that each approach has its merits. Our guideline also now incorporates specific guidance for hemodynamically stable children, and the findings support recommendations for a restrictive strategy (threshold <7 g/dL for children, excluding those with congenital heart disease). Minimizing unnecessary complications of transfusion and responding to the ongoing global challenges of having a safe and secure blood supply will require effective strategies, including blood management programs, for implementation of these guidelines.

Good transfusion practice should rely not only on hemoglobin concentration thresholds but also incorporation of patients’ symptoms, signs, comorbid conditions, rate of bleeding, values, and preferences. This guidance is particularly important because clinicians commonly use only hemoglobin concentration to decide when to transfuse.⁷⁰ Blood management programs that audit blood should attend to these broader considerations in their policies and decisions. Given that RCTs demonstrated no effect on mortality,^71,72 the storage age of transfused RBCs need not be considered in transfusion decisions.

Similar to older guidelines,^73-78 this guideline and other guidelines published after 2016 continue to recommend restrictive transfusion strategies^79-83 (Box).

Box. 

Red Blood Cell Transfusion Guidelines Since 2016

Society and Recommendation

• UK National Clinical Guidelines Centre (2016)⁷⁹
  • Restrictive threshold (7 g/dL) for patients who do not have major hemorrhage or acute coronary syndrome or need long-term transfusion. In acute coronary syndrome, transfusion should be considered at a threshold of 8 g/dL. Clinicians should consider setting individual targets for patients with chronic anemia.
• European Society of Anaesthesiology (2017)⁸⁰
  • Target hemoglobin level of 7-9 g/dL in patients with active bleeding
• Frankfurt Germany Consensus conference (2018)⁸¹
  • Varied depending on clinical setting: 7 g/dL for critically ill patients, 7.5 g/dL in cardiac surgery, 8 g/dL in hip fracture and cardiovascular disease, and 7-8 g/dL in acute gastrointestinal bleeding
• Pediatric Critical Care Transfusion and Anemia Expertise Initiative (2018)¹²
  • Varied depending on clinical setting: 7 g/dL for hemodynamically stable critically ill children; for hemodynamically stable children with congenital heart disease, varied based on cardiac abnormality and stage of repair; 7 g/dL biventricular repair, 9 g/dL stage 1 and stage 2 palliation
• Society of Cardiovascular Anesthesiologists (2019)⁸²
  • Transfusion threshold of 7.5 g/dL is reasonable in cardiac surgery
• The Society of Thoracic Surgeons and affiliated groups (2021)⁸³
  • Restrictive transfusion strategy, although a specific hemoglobin level was not provided

Research Recommendations

Ongoing trials for patients with acute myocardial infarction, vascular disease, and neurologic disorders will inform transfusion practice.¹⁷ Further analyses of subgroups of trials using individual patient data from existing trials are needed by age, sex, preexisting cardiovascular disease, pregnancy status, and other clinical factors. There are gaps in the evidence regarding the needs of individuals with myelodysplastic syndromes who are transfusion dependent. To modify symptoms of anemia, such people may require higher thresholds for transfusions. Given the findings indicating the safety of restrictive thresholds, new trial designs should focus on the safety of lower transfusion thresholds (eg, 5-6 g/dL), incorporation of physiologic parameters, and the conduct of health economic analyses.

Conclusion

Our panel recommends restrictive transfusion strategies, typically with a threshold of 7 g/dL for both adult and pediatric patients. The panel recognizes important additional considerations, including signs, symptoms, comorbid conditions, and patient values and preferences, that will differ between patients. The recommendation is strong, based on moderate certainty evidence for most patients, but conditional, based on lower certainty evidence subgroups that include hematologic and oncologic disorders in adults and cyanotic cardiac condition in infants.