Publication

• Title: 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
• Acronym: REALITY
• Year: 2021
• Journal published in: JAMA
• Citation: Ducrocq G, Gonzalez-Juanatey JR, Puymirat E, Lemesle G, Cachanado M, Durand-Zaleski I, et al; 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;325(6):552-560.

Context & Rationale

• Background

  • Anaemia is common in acute myocardial infarction (AMI) and is associated with higher short-term and long-term adverse outcomes.
  • Transfusion could improve oxygen delivery during ischaemia, but has plausible harms (volume overload, inflammatory and prothrombotic effects, transfusion-related lung injury) and confounding by indication limits inference from observational data.
  • Pre-REALITY randomised evidence in AMI/active coronary disease was limited to small trials with imprecise estimates and uncertainty about the optimal haemoglobin trigger.¹²
• Research Question/Hypothesis

  • In patients with AMI and anaemia (haemoglobin 7–10 g/dL), is a restrictive red-cell transfusion strategy non-inferior to a liberal strategy for 30-day major adverse cardiovascular events (MACE)?
• Why This Matters

  • Transfusion practice in AMI varies widely, and both over-transfusion (exposure to harms, resource use) and under-transfusion (recurrent ischaemia) are credible risks.
  • Defining a safe trigger has immediate implications for cardiac critical care pathways, blood bank demand, and guideline recommendations.

Design & Methods

• Research Question: Whether a restrictive transfusion strategy is non-inferior to a liberal transfusion strategy for 30-day MACE in AMI patients with haemoglobin 7–10 g/dL.
• Study Type: Multicentre, open-label, randomised, non-inferiority trial; 35 hospitals (France and Spain); web-based central randomisation; stratified by centre with variable block sizes (2–6).
• Population:
  • Inclusion: Adults admitted with AMI (STEMI or NSTEMI) and haemoglobin between 7 and 10 g/dL at any time during hospitalisation.
  • Key exclusions: Cardiogenic shock (systolic blood pressure <90 mm Hg with signs of low cardiac output or requiring inotropes); MI related to PCI or CABG; transfusion within 30 days; known haematological disease; major or life-threatening bleeding at randomisation.
  • Setting: Acute cardiac care/ward; ICU admission occurred in a minority (secondary outcome).
• Intervention:
  • Restrictive strategy: Transfuse 1 unit of packed red cells when haemoglobin ≤8 g/dL.
  • Target after transfusion: Stop transfusion when haemoglobin ≥8 g/dL and ≤10 g/dL.
  • RBC product: Leucoreduced packed red cells (1 unit at a time), with haemoglobin reassessment after each unit.
• Comparison:
  • Liberal strategy: Transfuse 1 unit of packed red cells when haemoglobin ≤10 g/dL.
  • Target after transfusion: Stop transfusion when haemoglobin ≥11 g/dL.
  • Protocolised approach: The liberal arm resulted in near-universal transfusion exposure.
• Blinding: Open-label to clinicians and patients; all components of the primary end point (and acute heart failure) adjudicated by a critical events committee blinded to randomised group assignment and haemoglobin values.
• Statistics: A total of 300 patients per group were required to demonstrate non-inferiority using a 1-sided 97.5% CI with a non-inferiority margin of RR 1.25 and 80% power; allowing for 5% major protocol violations increased the required sample size to 630 (315 per group); primary inference required concordant non-inferiority in both as-randomised and as-treated analyses; superiority testing was prespecified if non-inferiority was demonstrated.
• Follow-Up Period: 30 days for the primary end point (with additional longer-term follow-up reported separately).

Key Results

This trial was not stopped early. Recruitment exceeded the prespecified minimum sample size (668 randomised; 666 included in the as-randomised analysis).

• The restrictive strategy substantially reduced transfusion exposure (35.7% vs 99.7%) and total red-cell units (342 vs 758), while maintaining similar 30-day MACE rates in both as-randomised and as-treated analyses.
• Biological separation was achieved: discharge haemoglobin 9.7 ± 1.0 g/dL vs 11.1 ± 1.4 g/dL (difference −1.4 g/dL; 95% CI −1.6 to −1.2).
• The primary non-inferiority result was driven by a one-sided CI framework; superiority was not demonstrated (log-rank P=0.21).

Internal Validity

• Randomisation and allocation: Centralised web-based randomisation; stratified by centre; variable block sizes (2–6) supports allocation concealment at enrolment.
• Dropout / exclusions: 668 randomised; 2 excluded from analysis (lost consent form; withdrawal of consent immediately after randomisation); 666 analysed as-randomised with complete 30-day follow-up for the primary end point.
• Performance / detection bias: Open-label transfusion strategy makes performance bias plausible (co-interventions and clinical thresholds could be influenced); objective clinical end points were adjudicated by a blinded critical event committee, reducing detection bias for primary components.
• Protocol adherence: High adherence with clear separation in transfusion exposure; nevertheless, deviations occurred (eg, 13/342 in restrictive group received transfusion when haemoglobin was >8 g/dL; 1/324 in liberal group did not receive transfusion).
• Baseline characteristics: Groups were broadly comparable (eg, median age 78 vs 76 years; NSTEMI 68.4% vs 71.3%; mean haemoglobin at randomisation 8.6 ± 0.8 vs 8.7 ± 0.8 g/dL), with modest imbalances in active bleeding at admission (10.5% vs 15.1%).
• Heterogeneity: A prespecified random-effects sensitivity analysis yielded the same point estimate for the primary end point (RR 0.78; 1-sided 97.5% CI 0.00 to 1.17), arguing against major centre-driven heterogeneity.
• Timing: Randomisation allowed at any time during hospitalisation once haemoglobin entered the 7–10 g/dL range; the distribution of time from admission to randomisation was not reported.
• Dose: Protocolised single-unit transfusion with reassessment; achieved haemoglobin separation at discharge (9.7 ± 1.0 vs 11.1 ± 1.4 g/dL) and at nadir (8.3 ± 0.8 vs 8.8 ± 0.9 g/dL).
• Separation of the variable of interest: Any transfusion 35.7% vs 99.7%; total units 342 vs 758; discharge haemoglobin difference −1.4 g/dL (95% CI −1.6 to −1.2).
• Adjunctive therapy use: Not protocolised; differential use of anti-ischaemic therapies, diuretics, and invasive strategies by group not reported (beyond standard-of-care AMI management).
• Outcome assessment: Composite clinical end point with central adjudication; safety outcomes partly investigator-reported (acute heart failure adjudicated).
• Statistical rigour: Non-inferiority required concordant findings in as-randomised and as-treated analyses and was met; superiority testing was prespecified but not supported by the observed data.

Conclusion on Internal Validity: Overall, internal validity appears moderate to strong: randomisation and follow-up were robust and biological separation was clear, but open-label delivery and a per-protocol (as-treated) co-primary framework introduce potential bias typical of transfusion-threshold trials.

External Validity

• Population representativeness: Older AMI population with substantial comorbidity (median age 77 years; creatinine clearance <60 mL/min in ~65%); predominantly NSTEMI; includes patients with (non-life-threatening) bleeding at presentation.
• Key exclusions: Cardiogenic shock, major/life-threatening bleeding, recent transfusion, post-procedural MI, and known haematological disease limit applicability to the sickest or most unstable AMI patients.
• Applicability: The liberal arm was highly protocolised (99.7% transfused), which may differ from contemporary pragmatic practice; nevertheless, the transfusion triggers (8 vs 10 g/dL) are globally relevant and operationally simple.
• Health-system context: Conducted in France and Spain; transfusion products (leucoreduced red cells) and AMI pathways may differ in other systems or resource-limited settings.

Conclusion on External Validity: Generalisability is moderate for haemodynamically stable AMI patients with haemoglobin 7–10 g/dL in high-income hospital systems; applicability is limited for shock, massive bleeding, and settings with different transfusion infrastructure or baseline transfusion practice.

Strengths & Limitations

• Strengths: Pragmatic inclusion of STEMI/NSTEMI with clinically relevant anaemia; multicentre randomised design; near-complete follow-up; clear protocol separation (transfusion exposure and haemoglobin levels); blinded adjudication of primary end point components.
• Limitations: Open-label design; non-inferiority margin choice and one-sided CI framework; variable timing of enrolment during hospitalisation; limited power for rare harms and for superiority; no screening log reported (selection processes cannot be fully evaluated); geographically limited to two European countries.

Interpretation & Why It Matters

• Clinical signal

  In stable AMI with haemoglobin 7–10 g/dL, a restrictive trigger (≤8 g/dL) reduced transfusion exposure substantially (35.7% vs 99.7%) with similar 30-day MACE rates in the trial’s non-inferiority framework.
• Resource and harm minimisation

  Restrictive transfusion achieved a lower discharge haemoglobin (9.7 ± 1.0 vs 11.1 ± 1.4 g/dL) and avoided >400 red-cell units compared with the liberal strategy, supporting stewardship where clinically appropriate.
• Decision-making nuance

  The trial’s interpretation hinges on a non-inferiority paradigm and the degree of residual uncertainty acceptable in AMI (where recurrent ischaemia is high-stakes); individualisation remains essential for ongoing ischaemia, heart failure, or borderline haemodynamics.

Controversies & Other Evidence

• Non-inferiority framing and clinical tolerance for uncertainty: The selected non-inferiority margin (RR 1.25) permits a potentially clinically important increase in events, even if non-inferiority is declared; this tension was explicitly highlighted in published correspondence and reply.³⁴
• Durability of effect beyond 30 days: A prespecified longer-term report from REALITY found numerically more 1-year MACE events with the restrictive strategy (111 vs 92), with hazard ratio 1.16 (95% CI 0.88 to 1.53), emphasising uncertainty about late hazards or competing risks not captured at 30 days.⁵
• Scale-up evidence (MINT): In the larger MINT trial (n=3504), the restrictive strategy did not significantly reduce death or recurrent MI compared with a liberal strategy; the primary composite occurred in 16.9% vs 14.5% (RR 1.15; 95% CI 0.99 to 1.34; P=0.07), and potential harms of restrictive transfusion could not be excluded.⁶
• Heart failure as an effect modifier: A prespecified subgroup analysis of REALITY reported no interaction between baseline heart failure status and randomised strategy for 30-day MACE, but observed higher all-cause mortality with a liberal strategy in patients with heart failure (Pinteraction=0.009 at 30 days), with more deaths due to heart failure at 30 days (4 vs 11).⁷
• Guidelines: AABB international clinical practice guidelines for red-cell transfusion in AMI recommend (conditional; low certainty evidence) a liberal approach aiming to maintain haemoglobin around 10 g/dL in adults with AMI and anaemia, reflecting ongoing concern about restrictive thresholds in an ischaemic myocardium context.⁸
• Evidence synthesis: A patient-level meta-analysis of transfusion-strategy trials in myocardial infarction has been published, reflecting the continued need for synthesis across heterogeneous trial designs and populations.⁹

Summary

• REALITY randomised 668 AMI patients with haemoglobin 7–10 g/dL to restrictive (≤8 g/dL) vs liberal (≤10 g/dL) transfusion strategies.
• Primary 30-day MACE was 11.1% vs 14.2% (as-randomised) with RR 0.78; 1-sided 97.5% CI 0.00 to 1.17, meeting the prespecified non-inferiority framework (margin RR 1.25).
• Restrictive transfusion markedly reduced exposure to transfusion (35.7% vs 99.7%) and achieved lower discharge haemoglobin (9.7 ± 1.0 vs 11.1 ± 1.4 g/dL).
• Safety outcomes were generally similar, though descriptive reporting noted more acute lung injury/ARDS in the liberal group (2.2% vs 0.3%).
• Subsequent evidence (including MINT and guideline recommendations) maintains equipoise and highlights that clinically important harms of restrictive transfusion in AMI cannot be fully excluded.

Overall Takeaway

REALITY is a landmark AMI transfusion-threshold trial because it demonstrated substantial transfusion reduction with a restrictive trigger while meeting a prespecified non-inferiority framework for 30-day MACE. However, its non-inferiority design and subsequent larger and longer-term evidence mean that restrictive transfusion should be applied thoughtfully, especially in patients with ongoing ischaemia or heart failure risk, and alongside evolving guideline recommendations.

Bibliography

• 1.Cooper HA, Rao SV, Greenberg MD, et al. Conservative versus liberal red cell transfusion in acute myocardial infarction (the CRIT randomized pilot study). Am J Cardiol. 2011;108:1108-1111.
• 2.Carson JL, Brooks MM, Abbott JD, et al. Liberal versus restrictive transfusion thresholds for patients with symptomatic coronary artery disease. Am Heart J. 2013;165:964-971.
• 3.Sato T, Terada R, Ikeda T. Effect of Restrictive or Liberal Blood Transfusion on Major Cardiovascular Events in Patients With Acute Myocardial Infarction and Anemia. JAMA. 2021;325(24):2505-2506.
• 4.Ducrocq G, Gonzalez-Juanatey JR, Puymirat E, et al. Effect of Restrictive or Liberal Blood Transfusion on Major Cardiovascular Events in Patients With Acute Myocardial Infarction and Anemia—Reply. JAMA. 2021;325(24):2506-2507.
• 5.Gonzalez-Juanatey JR, Lemesle G, Puymirat E, et al; REALITY Investigators. One-Year Major Cardiovascular Events After Restrictive Versus Liberal Blood Transfusion Strategy in Patients With Acute Myocardial Infarction and Anemia: The REALITY Randomized Trial. Circulation. 2022;145(6):486-488.
• 6.Carson JL, Brooks MM, Hébert PC, et al; MINT Investigators. Restrictive or Liberal Transfusion Strategy in Myocardial Infarction and Anemia. N Engl J Med. 2023;389(26):2446-2456.
• 7.Ducrocq G, Cachanado M, Simon T, et al. Restrictive vs Liberal Blood Transfusions for Patients With Acute Myocardial Infarction and Anemia by Heart Failure Status: An RCT Subgroup Analysis. Can J Cardiol. 2024 Sep;40(9):1705-1714.
• 8.Pagano MB, Stanworth SJ, Dennis J, et al. Red Cell Transfusion in Acute Myocardial Infarction: AABB International Clinical Practice Guidelines. Ann Intern Med. 2025;178(10):1469-1477.
• 9.Carson JL, Fergusson DA, Noveck H, et al. Restrictive versus Liberal Transfusion in Myocardial Infarction — A Patient-Level Meta-Analysis. NEJM Evid. 2025 Feb;4(2):EVIDoa2400223.