Effect of a Restrictive vs Liberal Blood Transfusion Strategy on Major Cardiovascular Events Among Patients With Acute Myocardial Infarction and Anaemia

Ducrocq G. JAMA 2021:325(6):552-560 doi:10.1001/jama.2021.0135

Clinical Question

  • In patients with acute myocardial infarction and anaemia is a restrictive, compared with a liberal transfusion strategy, non-inferior with respect to major adverse cardiovascular outcomes at 30 days?


  • Anaemia is associated with major adverse cardiovascular events in patients with acute coronary syndrome
  • It is not known if transfusing patients with anaemia and AMI ameliorates this risk and often patients with acute coronary syndromes are excluded from trials comparing liberal and restrictive transfusion strategies
  • Previous trials have been very small with poor methodology
  • There is wide variation in practice
  • There is a growing body of evidence demonstrating harm from transfusion in several patient groups
  • TRICS-III, a trial in patients undergoing cardiac surgery with moderate-to-high risk of death showed a restrictive strategy of transfusion was non-inferior to a liberal strategy with respect to a composite outcome
  • The TRISS trial investigated liberal versus restrictive transfusion in patients with septic shock and demonstrated no significant difference in 90 day mortality
  • Patients with acute upper gastrointestinal bleeding had lower mortality with a restrictive strategy (Villanueva, NEJM 2013;368(1)11-21)
  • The TRICC trial assessed a liberal versus restrictive strategy in 838 ICU patients expected to stay >24 hours and showed no mortality difference. However, patients who were admitted after cardiac surgery were excluded from this study
  • Transfusion may cause platelet activation and increase thrombotic risk by causing platelet activation, aggregation and vasoconstriction
  • A storage defect in blood causing depletion in nitric oxide and 2,3 DPG may mean that the desired effect of improving O2 delivery may not be achieved with transfused blood


  • Randomised controlled trial
  • 1:1 randomisation to liberal or restrictive transfusion
  • Allocation concealment achieved by a web-based randomisation method with block randomisation and variable size blocks, stratified by centre
  • Initially the design had the restrictive group being transfused if Hb<7g/dL but this was changed to Hb<8g/dL to improve investigator adherence to the protocol
  • The patient could be randomised at any time from admission up to 30 days into their hospital stay
  • Patients consented to involvement in the trial prior to randomisation
  • Unblinded, although the outcomes were objective
  • The outcome was a composite MACE (major acute cardiovascular event) at 30 days. The outcomes of interest were all-cause death, recurrent myocardial infarction, emergency revascularisation prompted by ischaemia, stroke.
  • The analysis was as-randomised and as-treated (in non-inferiority trials this is done instead of intention to treat analysis)
  • The power calculation was based on observational data from the French registry of AMI which demonstrated MACE at 30 days to be 11% in a restrictive transfusion group and 15% in a liberal transfusion group
  • 630 patients were deemed necessary to achieve an 80% power to demonstrate non-inferiority in the restrictive group (assuming 5% would have protocol variation)


  • France (26 centres) and Spain (9 centres)
  • Enrolment occurred between March 2016 to September 2019


  • Inclusion: AMI (with or without ST elevation, symptoms in the last 48 hours and elevated biomarkers of myocardial injury) AND Haemoglobin between 7 and 10 g/dL during the admission. Age > 18 years
  • Exclusion:
    • Shock (SBP <90mmHg) with clinical signs of low cardiac output
    • AMI after CAGs or PCI
    • Life-threatening bleeding
    • Blood transfusion in last 30 days
    • Malignant haematologic disease
  • Participant numbers: 668 patients were randomised
  • Baseline characteristics: patients were well-matched at baseline
    • The median age was 77 years
    • 57.8% were male
    • 70% had NSTEMI, 30% had STEMI


  • Liberal transfusion (n=324)
    • transfusion for all patients with Hb < 10g/dL with a target post-transfusion Hb > 11g/d


  • Restrictive transfusion (n=342)
    • transfusion if Hb < 8g/dL with a target 8-10g/dL. Originally the threshold was 7g/dL but was changed to maximise investigator adherence to the protocol, before the first patient was included

Management common to both groups

  • Leukodepleted red cells were used
  • Transfusion could be administered at any time if there was a suspected massive bleeding event, before waiting for a haemoglobin result


  • Primary outcome: 30 day MACE (a composite score of major adverse cardiovascular events – all-cause death, stroke, recurrent MI, emergency revascularisation prompted by ischaemia)
    • At day 30, MACE had occurred in 14% in liberal group and 11% in restrictive group (45 vs 36 patients, between group difference -3% 95% CI -8.4% to 2.4%). This met the pre-specified non-inferiority criterion
    • Primary outcome analysis was of the as-treated population
    • As-randomised population had very similar results
  • Secondary outcomes: The components of the Day 30 MACE were reported as secondary outcomes. Due to high risk of Type 1 error, no formal statistical comparisons are made
  • Comparing Liberal Transfusion vs. Restrictive Transfusion
    • Death 25 vs 19 people (7.7% vs 5.6%)
    • Nonfatal recurrent myocardial infarction 10 vs 7 people (3.1% vs 2.1%)
    • Emergency revascularisation 6 vs 5 people (1.9% vs 1.5%)
    • Non-fatal ischemic stroke 2 vs 2 (0.6% vs 0.6%)
  • Patients who received >=1 unit blood transfusion
    • greater in liberal group (99.7% vs 35.7%)
  • Evidence of Harm: patients in the liberal group were more likely to
    • have acute lung injury
      • 7 vs 1 people (2.2% vs 0.3%)
    • develop infection (documented bacteremia or bacterial infection)
      • 5 vs 0 (1.5% vs 0%)

Authors’ Conclusions

  • In patients with AMI and anaemia, a restrictive transfusion strategy compared with a liberal transfusion strategy was non-inferior in regards to 30 day MACE outcomes


  • An important study in a key population where a paucity of evidence exists
  • Clinically important outcomes
  • Patients were analysed as randomised AND as treated
  • Follow-up was near complete (2 patients withdrew consent)
  • Allocation concealment


  • Physicians were unblinded to the therapy but the outcome measures were objective so bias to outcome likely low
  • The margin for non-inferiority was 25% so although this pre-specified margin was reached, this trial does not completely eliminate that a restrictive strategy is non-inferior. This minimal clinically important difference is arbitrarily agreed to in non-inferiority trials
  • Statistical significance for superiority was not met so a larger trial would need to be conducted to demonstrate superiority. This trial could be used to estimate sample size in a future larger trial
  • Outcome was assessed at Day 30. This is relatively short and could miss key events. A longer period for follow-up would add to the validity

The Bottom Line

  • Although not definitive this trial suggests that a restrictive transfusion strategy in patients with AMI and anaemia is safe with a non-statistically significant trend towards superiority
  • A larger trial would be needed to definitely guide management of anaemia in this patient population

External Links


Summary author:Celia Bradford @celiabradford
Summary date: 25/1/21
Peer-review editor: @davidslessor

Picture by: [author/site]


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.