Key messages
• The evidence suggests that waiting to give blood transfusions to most adults and children until their blood counts are lower (haemoglobin levels of 7.0 g/dL to 8.0 g/dL) compared to higher blood counts (9.0 g/dL to 10.0 g/dL) does not lead to higher risks of death, heart attack, stroke, pneumonia, blood clots or infection. Giving less blood would reduce unnecessary transfusions, which have risks.
• Some studies, rather than simply measuring haemoglobin levels, are testing other ways of deciding when a transfusion is necessary, but it is not possible to say which tests are better.
• Adults with acute brain injury may have a better ability to walk or take care of themselves if given blood transfusion to maintain higher blood counts. More research is needed to improve our understanding of outcomes other than death, including quality of life, and to understand which people may benefit from transfusing at a higher haemoglobin count.
What happens in people who need blood transfusions?
Doctors and healthcare professionals often give blood transfusions to people with low blood counts (anaemia), for surgery, bleeding or illness, but transfusions should only be given when we know they help people. Blood is a limited resource and transfusion is not risk-free, especially for people in low-income countries where testing for viruses such as HIV or hepatitis may be less available.
What did we want to find out?
'Blood count' measures the amount of haemoglobin in the blood. Haemoglobin is a protein that carries oxygen around the body. A normal blood count is at or above 12 grams a decilitre (12 g/dL). We wanted to find out if it is safe to withhold transfusions until the blood count drops to low thresholds, rather than transfusing at higher blood counts.
What did we do?
We examined the results of studies that allocated people to one of two groups by chance (for example, by flipping a coin). In one group, the participants only received blood transfusions if their blood count fell below a higher threshold (typically, 9.0 g/dL to 10.0 g/dL). In the other group, the participants only received transfusions if their blood count fell below a lower threshold (typically, 7.0 g/dL to 8.0 g/dL). (The ‘threshold’ is the blood count level that would need to be met before a transfusion would be given.)
What did we find?
We found 79 studies including 31,457 adults and 2864 children. Participants were undergoing treatment for different reasons including: bone (orthopaedic), heart (cardiac) or blood vessel surgery; critical care; bleeding in the stomach or intestines; heart diseases; blood cancers; brain injury and childbirth. Studies compared higher or lower blood count thresholds for giving blood transfusion.
Transfusion
We found that adults who received transfusions only at lower blood count thresholds were 42% less likely to receive a blood transfusion than those who received them only at higher blood count thresholds. Findings were similar for children, but the number of studies including children was much smaller.
Death and harmful events
Overall, there was no clear difference in the risk of dying within 30 days of receiving, or not receiving, a transfusion for most participants in the two different threshold groups, with the possible exception of people with gastrointestinal bleeding, where the risk of dying was lower with transfusion given at a lower threshold.
There was also little to no difference between threshold groups for outcomes including the number of serious harmful events, such as infection, heart attacks, strokes and blood clots.
We found that critically ill adults with acute brain injury had better recovery in the longer term (6 to 12 months) with a liberal transfusion strategy.
What are the limitations of the evidence?
Most studies provided high-quality evidence; they were adequately conducted and used methods that minimised biases that could make the validity of the results uncertain. We are confident in the evidence regarding the likelihood of receiving a transfusion, death within 30 days of transfusion, heart attack, stroke and infection. We are moderately confident in the evidence for infections and for blood clots, but the number of events for blood clots was smaller. The number of studies varied between conditions, and more evidence is needed in some areas (e.g. blood cancer and vascular surgery).
Almost half of the studies reported on quality of life, but it was difficult to compare trials and determine whether one strategy was better than the other.
How up-to-date is this evidence?
This is an update of work last published in 2021; 31 new studies are included. The evidence is up-to-date to October 2024.
Read the full abstract
The optimal haemoglobin threshold for use of red blood cell (RBC) transfusions in anaemic patients remains an active area of research. Blood is a limited resource, and there are concerns about risks, including transmitted infections. If a liberal transfusion strategy does not improve clinical outcomes, or if it is equivalent, then adopting a more restrictive approach should be recognised as the standard of care.
Objectives
The aim of this review update was to compare 30-day mortality and other clinical outcomes for participants randomised to restrictive versus liberal red blood cell (RBC) transfusion thresholds (triggers) for all clinical conditions. The restrictive transfusion threshold uses a lower haemoglobin concentration as a threshold for transfusion (most commonly, 7.0 g/dL to 8.0 g/dL), and the liberal transfusion threshold uses a higher haemoglobin concentration to direct transfusion (most commonly, 9.0 g/dL to 10.0 g/dL). Increasingly, investigators are considering other strategies including physiological triggers (i.e. central venous oxygen saturation), alone or in combination with such thresholds, to determine when a transfusion is indicated, so it is important to assess this growing body of evidence in tandem.
Search strategy
We searched CENTRAL, MEDLINE, Embase, Transfusion Evidence Library, Web of Science Conference Proceedings Citation Index, trial registries and PubMed on 14 October 2024. We checked reference lists of published reviews and papers for additional trials.
Selection criteria
We included randomised trials of surgical or medical participants that recruited adults or children. We excluded studies that focused on preterm neonates. Eligible trials assigned intervention groups based on different transfusion strategies or thresholds, usually defined by a haemoglobin concentration below which a RBC transfusion would be administered. We included trials in which investigators had allocated participants to higher thresholds or more liberal transfusion strategies compared to more restrictive ones, which might include no transfusion.
Data collection and analysis
We used standard Cochrane methods. We pooled risk ratios across trials using a random-effects model. We assessed risk of bias using the Cochrane RoB 1 tool, and assessed the certainty of evidence using GRADE. We defined participants randomly allocated to the lower transfusion threshold as being in the 'restrictive transfusion' group and those randomly allocated to the higher transfusion threshold as being in the 'liberal transfusion' group.
Main results
Adult threshold comparison
We included 61 trials (27,639 participants), across a range of clinical contexts: orthopaedic, cardiac or vascular surgery; critical care; neurocritical care; gastrointestinal bleeding; trauma; acute myocardial infarction; haematological malignancies and postpartum haemorrhage. The haemoglobin concentration used to define the restrictive transfusion group in most trials was between 7.0 g/dL and 8.0 g/dL. The main outcomes were exposure to blood transfusion, 30-day mortality, neurologic function, myocardial infarction, congestive heart failure, cerebral vascular accident, infection and thromboembolism.
Studies were generally at low risk of bias.
Restrictive transfusion strategies reduced the risk of receiving at least one RBC transfusion by 42% when combining all clinical contexts (risk ratio (RR) 0.58, 95% confidence interval (CI) 0.52 to 0.65; high-certainty evidence), with a large amount of heterogeneity between trials (I² = 97%), reflecting diversity in the strength of estimates, not the efficacy of the policy. When combining all clinical contexts, restrictive transfusion strategies did not modify the risk of 30-day mortality compared with liberal transfusion strategies (RR 1.01, 95% CI 0.90 to 1.14; 44 studies, 22,575 participants; high-certainty evidence) or any of the other outcomes assessed including myocardial infarction, stroke, thromboembolism or infection (moderate to high-certainty evidence).
There were two exceptions in clinical populations. In gastrointestinal bleeding, 30-day mortality was lower with a restrictive transfusion strategy (RR 0.63, 95% CI 0.42 to 0.95; 4 studies, 1574 participants). In critically ill patients with brain injury, unfavourable neurological outcome at 6 to 12 months was lower with a liberal transfusion strategy (RR 1.14, 95% CI 1.05 to 1.22; 4 studies, 2297 participants) (moderate-certainty evidence).
Transfusion-specific reactions are uncommon, but occurred more frequently with the liberal strategy (Peto odds ratio 0.47, 95% CI 0.36 to 0.62; 18 studies, 11,505 participants).
Paediatric threshold comparison
We included eight trials (2764 participants), across a limited range of clinical contexts (critical care, cardiac surgery, haematological malignancies and severe malarial anaemia). The haemoglobin concentration used to define the restrictive group in critical care and haematological malignancies trials was between 6.5 g/dL and 8.0 g/dL, and between 7.0 g/dL and 9 g/dL in cardiac surgery trials, depending on the cardiac abnormality and stage of repair. Studies were generally at low risk of bias. There was no clear difference in 30-day mortality between restrictive and liberal transfusion strategies, although confidence intervals were wide (RR 1.22, 95% CI 0.72 to 2.08; 7 studies, 2571 participants; low-certainty evidence). There was moderate-certainty evidence of no clear difference between threshold strategies for infection; very low-certainty evidence of little to no difference for thrombosis and very low-certainty evidence of little to no difference for the outcomes multiple organ dysfunction and cerebrovascular accident.
Physiological triggers
Nine trials in adults (3818 participants) and one trial in children (100 participants) were identified. These tested different interventions and measures of physiological parameters in diverse clinical populations. The risk of bias was variable. Meta-analysis was not appropriate due to heterogeneity.
Authors' conclusions
A restrictive transfusion strategy significantly decreased the proportion of adults and children exposed to RBC transfusion. In most clinical contexts, there was no evidence of harm from a restrictive compared with a liberal transfusion strategy. Neurocritically ill patients, however, have better neurological outcomes at 6 to 12 months with a liberal transfusion strategy. Further work is needed to improve our understanding of outcomes beyond mortality, and to what degree the optimal strategies for transfusion should be modified in some patient populations, including different types of acute bleeding, cancers and subgroups of patients with myocardial infarction and other neurological injuries. New studies need to recognise the clinical context and the limitations of the adoption of a single threshold of haemoglobin, and to consider evaluating the use of physiological parameters to modify transfusion.
