Early observations and pilot data that first suggested a new direction
For most of the 20th century, the '10/30 rule' prevailed — transfuse to maintain haemoglobin above 10 g/dL and haematocrit above 30%. This threshold was based on physiological reasoning and expert consensus rather than clinical trial evidence. As awareness of transfusion risks grew (infections, immunomodulation, volume overload, iron overload), clinicians began questioning whether liberal transfusion practices were causing more harm than good. The stage was set for a paradigm shift when observational studies in the 1990s began associating red cell transfusion with worse outcomes in critically ill patients, though confounding by indication made causal inference difficult.
Landmark RCTs and pivotal trials that established the evidence base
The TRICC trial (1999) by Hebert and colleagues was the seminal study that challenged liberal transfusion. In 838 critically ill patients, a restrictive strategy (transfusion trigger 7 g/dL, target 7-9 g/dL) was at least as effective as a liberal strategy (trigger 10 g/dL, target 10-12 g/dL), with a strong trend toward reduced mortality. In patients under 55 and those with APACHE II scores ≤20, the restrictive strategy was significantly superior. The TRISS trial (2014) confirmed these findings in 998 patients with septic shock, demonstrating non-inferiority of a restrictive threshold (7 g/dL) compared to a liberal threshold (9 g/dL) with fewer transfusions. The FOCUS trial (2011) in 2,016 elderly hip fracture patients found that a restrictive strategy (trigger 8 g/dL) was non-inferior to a liberal strategy (trigger 10 g/dL) for death or inability to walk at 60 days.
Follow-up studies, subgroup analyses, and real-world validation
The TRICS-III trial (2017) extended restrictive transfusion evidence into cardiac surgery — a population where liberal practice had been particularly entrenched due to concerns about myocardial oxygen delivery. In 5,243 patients undergoing cardiac surgery, a restrictive strategy (trigger <7.5 g/dL) was non-inferior to a liberal strategy (trigger <9.5 g/dL) for the composite of death, stroke, MI, or new renal failure. The REALITY trial (2021) in acute upper GI bleeding and the MINT trial (2023) in acute MI provided context-specific guidance, with MINT suggesting potential harm from restrictive transfusion in myocardial infarction patients — highlighting that the optimal threshold is not universal but context-dependent.
Integration into clinical practice guidelines and recommendations
AABB (2012, updated 2016) provides a strong recommendation for a restrictive transfusion threshold of 7 g/dL for haemodynamically stable hospitalised patients, including critically ill patients. For postoperative patients and those with cardiovascular disease, a threshold of 8 g/dL is suggested. The NICE blood transfusion guidelines (2015) recommend a threshold of 7 g/dL for patients without major haemorrhage or ACS, and 8 g/dL for those with cardiovascular disease. The European Society of Anaesthesiology guidelines align with these thresholds while acknowledging ongoing uncertainty in acute coronary syndromes following the MINT trial results.
AABB Clinical Practice Guideline 2016
Strong recommendation for restrictive threshold of 7 g/dL in stable hospitalised patients; 8 g/dL threshold for cardiovascular disease or postoperative patients
NICE Blood Transfusion Guideline (NG24) 2015
Use 7 g/dL threshold and a target of 7-9 g/dL for patients who do not have major haemorrhage or ACS; 8 g/dL for cardiovascular disease
Now
Current standard of care and ongoing research directions
Restrictive transfusion (trigger 7 g/dL) is the standard of care for most hospitalised patients, including the critically ill, perioperative, and septic populations. The main area of genuine uncertainty is acute coronary syndromes, where the MINT trial raised concerns that restrictive thresholds might be harmful — a question not yet definitively resolved. Single-unit transfusion policies have become standard, replacing the historical practice of transfusing two units by default. Patient blood management programmes — encompassing preoperative anaemia correction, intraoperative blood conservation (cell salvage, antifibrinolytics), and restrictive transfusion triggers — represent the comprehensive approach. The field is moving toward personalised transfusion decisions based on physiological tolerance of anaemia (lactate, ScvO2, echocardiography) rather than a single haemoglobin number.
What is the current recommended transfusion threshold?+
For most stable hospitalised patients, including critically ill and perioperative patients, a haemoglobin trigger of 7 g/dL is recommended (AABB, NICE). For patients with cardiovascular disease (excluding acute MI) or postoperative patients with symptoms, 8 g/dL is suggested. In acute coronary syndromes, the threshold remains debated — MINT raised concerns about restrictive triggers, and many clinicians maintain a threshold of 8-9 g/dL in this population.
Should I transfuse one unit or two?+
One unit at a time. Single-unit transfusion with reassessment is now the recommended approach for non-bleeding patients. Each unit raises haemoglobin by approximately 1 g/dL. Transfusing two units when one suffices exposes patients to unnecessary donor exposure, cost, and transfusion risks. The 'two-unit default' was a historical practice with no evidence base.
What about transfusion in acute MI?+
This is the most uncertain area. The MINT trial (2023) compared restrictive (trigger <8 g/dL) versus liberal (trigger <10 g/dL) strategies in 3,504 patients with acute MI and found that non-inferiority of the restrictive strategy was not established, with a trend toward harm. Until further data emerge, many experts recommend a more liberal threshold of 8-9 g/dL in acute coronary syndromes, though this remains actively debated.
What role does preoperative anaemia correction play?+
Preoperative anaemia affects 30-40% of surgical patients and is an independent risk factor for transfusion, morbidity, and mortality. Patient blood management programmes advocate for identification and correction of anaemia (particularly iron deficiency) weeks before elective surgery using oral or IV iron, and occasionally erythropoietin. This reduces perioperative transfusion requirements and may improve outcomes independently of transfusion avoidance.
