NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
National Clinical Guideline Centre (UK). Blood Transfusion. London: National Institute for Health and Care Excellence (NICE); 2015 Nov. (NICE Guideline, No. 24.)
Red Blood Cells (RBCs) carry oxygen to cells. Anaemia reduces the oxygen carrying capacity of blood, but compensatory physiological responses (increased cardiac output; increased oxygen extraction by organs) mean healthy humans can tolerate anaemia to remarkably low haemoglobin concentrations when the circulating blood volume is maintained, for example with crystalloid fluid infusion. RBC transfusions are used to replace blood lost during haemorrhage or to increase a low haemoglobin concentration occurring for other reasons (for example, bone marrow failure or haemolysis).
Transfusion of stored RBCs carries risks and the timing of RBC transfusion, especially in relation to the ‘trigger’ haemoglobin concentration and target haemoglobin range, is an important clinical decision. The aim is to transfuse RBCs when the clinical benefits outweigh the risks. In addition, conserving RBC supplies for those patients in whom they are most clinically and cost-effective is vital, as RBCs are an increasingly scarce and costly treatment.
This chapter reviews the available evidence for the clinical and cost-effectiveness of RBC transfusions in different clinical situations. In addition, the evidence for particular sub-groups of patients in whom the tolerance of anaemia, and therefore the risk to benefit balance for RBC transfusion, may differ is reviewed. These include patients in whom the disease being treated could mean physiological compensation for anaemia may be compromised, for example, by heart disease or critical illness.
10.1. Review question: What is the clinical- and cost-effectiveness of red blood cell transfusion at different haemoglobin concentrations?
For full details see review protocol in Appendix C.
Table 69PICO characteristics of review question
Population |
|
---|---|
Intervention |
|
Comparison |
|
Outcomes |
|
Study designs |
|
10.2. Review question: What is the clinical- and cost-effectiveness of different target levels of post-transfusion haemoglobin concentrations for red blood cell transfusion?
For full details see review protocol in Appendix C.
Table 70PICO characteristics of review question
Population |
|
---|---|
Intervention |
|
Comparison |
|
Outcomes |
|
Study designs |
|
10.3. Methodology of clinical evidence review (threshold haemoglobin concentrations and target haemoglobin levels for blood transfusion)
The GDG was interested in establishing the most clinical and cost-effective haemoglobin thresholds at which red blood cell transfusion should be administered and the target haemoglobin levels to which red blood cell transfusion should be given. To this effect, two separate review questions were drafted as follows:
- What is the clinical and cost-effectiveness of red blood cell transfusion at different haemoglobin concentrations?
- What is the clinical and cost-effectiveness of different target levels of post-transfusion haemoglobin concentration for red blood cell transfusion?
The comparisons in each review included restrictive and liberal haemoglobin concentrations compared with one another. For full details see review protocols in Appendix C (C.1 and C.2).
On reviewing the evidence, it was acknowledged that the two reviews were very closely interlinked.
- The studies evaluated restrictive and liberal strategies such that patients transfused at restrictive haemoglobin thresholds received red blood cells until restrictive haemoglobin target levels were attained and vice versa. As a result, the data for the individual reviews are confounded and it is not possible to comment on threshold or target haemoglobin levels independently.
- The majority of the studies aimed to establish the effectiveness of different haemoglobin thresholds. In addition, some studies reported the target haemoglobin concentrations to which patients were transfused in each group. No evidence was identified independently on the effectiveness of different target haemoglobin levels irrespective of the thresholds of blood transfusion.
In light of the above, we have presented one combined evidence review with separate results for the sections on thresholds and target levels of haemoglobin concentrations. The evidence on the effectiveness of different target haemoglobin levels will be viewed in the context of the corresponding haemoglobin threshold levels at which the patients received blood transfusion.
10.4. Clinical evidence
RBC thresholds
We searched for systematic reviews and randomised controlled trials for addressing effectiveness of red blood transfusion at different haemoglobin concentrations.
Thirty four studies were included in the review;24,28,32,39,42,61,100,106,108,109,121,123,125,126,135,149,174,185,192,246,284,326,347 203 these are summarised in Table 71 below. Evidence from these studies is summarised in the GRADE clinical evidence profile below. See also the study selection flow chart in Appendix E, forest plots in Appendix K, GRADE tables in Appendix J, study evidence tables in Appendix H and excluded studies list in Appendix P.
Table 71
Summary of studies included in the review.
RBC targets
We searched for systematic reviews and randomised controlled trials for addressing effectiveness of different target levels of post-transfusion haemoglobin concentrations for red blood cell transfusion.
- Seven studies which compared haemoglobin threshold levels also reported the target levels of haemoglobin concentrations;63,123,125,126,174,192,318 these are highlighted and summarised in Table 71. As outlined earlier, the target levels of haemoglobin concentration are classified into restrictive and liberal and correspond with the restrictive and liberal thresholds of haemoglobin concentration for administration of blood transfusion.
- One study compared target levels of haemoglobin concentrations in patients with advanced gastric cancer who were receiving chemotherapy.241 No haemoglobin thresholds for transfusion have been reported in this study. However, the only outcome that could be analysed from this study was incidence of acute pulmonary oedema (new cardiac event) and this has been included in the meta-analysis. Other outcomes reported in this study include number of units transfused and quality of life (Karnofsky Performance Scores); these are not reported in an analysable format (mean and standard deviations not reported).
- Only one study was conducted in children.174
Evidence from these studies is summarised in the GRADE summary evidence profile (see sections 10.4.1 and 10.4.2). See also the study selection flow chart in Appendix E, GRADE evidence profiles in Appendix J, study evidence tables in Appendix H, forest plots in Appendix K, and excluded studies list in Appendix P.
10.4.1. Summary of the evidence (Summary GRADE profile)- haemoglobin thresholds for blood transfusion
10.4.1.1. Restrictive strategy compared with liberal strategy for blood transfusion (adults)
Table 72
Blood transfusions (adults).
Table 73
Length of hospital stay (adults).
Table 74
All cause-Mortality at 30 days (adults).
Table 75
New cardiac events (adults).
Table 76
Infection - adults.
Table 77
Adverse events (adults).
Table 78
Quality of life (adults).
10.4.2. Summary of the evidence - Target haemoglobin levels for blood transfusion
This section summarises the evidence on haemoglobin target levels for RBC transfusion from those studies that report haemoglobin thresholds for RBC transfusion and are included in the analysis and results presented in section 1.6.1 above. Hence the results are not independent of the results above.
10.4.2.1. Summary GRADE clinical evidence profiles
10.5. Economic evidence
10.5.1. RBC Thresholds
Published literature
One relevant economic evaluation was identified with the relevant comparison for RBC thresholds and targets and has been included in this review.320 One relevant economic evaluation was identified with the relevant comparison for RBC thresholds.203 These are summarised in the economic evidence profiles below (Table 95 and Table 96) and the economic evidence tables in Appendix I.
Table 95
Economic evidence profile: Restrictive versus liberal RBC threshold and target.
Table 96
Economic evidence profile: Restrictive versus liberal RBC threshold.
Two economic evaluations relating to this review question were identified but were excluded due to a combination of limited applicability and methodological limitations.239,343 These are summarised in Appendix Q, with reasons for exclusion given.
See also the economic article selection flow chart in Appendix F.
Unit costs
Relevant unit costs are provided in Appendix N to aid consideration of cost-effectiveness.
10.5.2. RBC Targets
Published literature
One relevant economic evaluation was identified with the relevant comparison for RBC thresholds and targets and has been included in this review.320 This is summarised in the economic evidence profile above (Table 95) and the economic evidence tables in Appendix I.
See also the economic article selection flow chart in Appendix F.
Unit costs
Relevant unit costs are provided in Appendix N to aid consideration of cost-effectiveness.
10.6. Evidence statements
Clinical
RBC thresholds
Restrictive strategy versus liberal strategy in adults
Thirty one RCTs compared restrictive strategies with liberal strategies for blood transfusion in adults. The evidence showed clinically important benefit with restrictive strategies with respect to the number of patients transfused and number of units transfused. The evidence suggested that there was no difference between the groups with respect to mortality, adverse events, new cardiac events, length of hospital stay, quality of life and infection, but there was some uncertainty. The evidence was of low and very low quality.
Restrictive strategy versus liberal strategy in children
Three RCTs compared restrictive strategy with liberal strategy in children. The evidence showed that there was a clinically important benefit with restrictive strategies in children with respect to the number of patients transfused. The evidence suggested that lower volumes and less number of units may be transfused in children with restrictive strategies, but there was some uncertainty. The evidence suggested that there may be no difference between the groups with respect to mortality at 30 days, ICU length of stay, adverse events and infections (nosocomial infection), however there was considerable uncertainty.
The evidence in children ranged from moderate to very low quality. No evidence was identified for the critical outcome quality of life.
RBC targets
Restrictive strategy versus liberal strategy in adults
Seven RCTs compared restrictive strategy with liberal strategy in adults. The evidence showed clinically important benefit with the use of restrictive targets for blood transfusion for the outcomes of number of patients transfused and number of units transfused. The evidence suggested that there may be a benefit with respect to all-cause mortality at 30 days, new cardiac events, length of hospital stay and adverse events in patients being transfuse to restrictive target levels, but there was some uncertainty. No difference was observed between the groups with respect to incidence of infections (pneumonia).
The evidence was of low and very low quality. No evidence was identified for outcome quality of life,.
Restrictive strategy versus liberal strategy in children
One RCT compared restrictive strategy with liberal strategy in children. The evidence showed clinically important benefit for restrictive strategy for the outcome number of patients transfused. The evidence suggested there may be a benefit with the use of restrictive targets with respect to infections, however there was considerable uncertainty. The evidence suggested that there may be no difference between groups with respect to the total volume of RBC transfused, mortality, ICU length of stay and pulmonary oedema but there was considerable uncertainty.
The evidence ranged from moderate to very low quality. No evidence was identified for the critical outcome quality of life.
Economic
One cost-effectiveness analysis found that, in anaemic, older, critically ill patients requiring prolonged mechanical ventilation, a restrictive red blood cell strategy was more costly and more effective than a liberal red blood cell transfusion strategy (ICER: £253,681 per life year gained). This analysis was assessed as partially applicable with potentially serious limitations.
One cost-consequence analysis found in a non-emergency cardiac surgery population (>16 years old) that a restrictive threshold was less costly (saving £178 per patient) had 2.1% more serious infection or ischaemic events, 0.7% more all-cause mortality at 30 days, 1.6% more all-cause mortality at 90 days, 1.2% more clinically significant pulmonary complications and fewer units of red blood cells transfused (saves 1 unit) compared to a liberal threshold. EQ-5D and hospital length of stay were the same in both groups. This analysis was assessed as partially applicable with potential serious limitations.
10.7. Recommendations and link to evidence
Recommendations |
|
---|---|
Relative values of different outcomes | The GDG considered all-cause mortality at 30 days, infections (including pneumonia, surgical site infection, UTI and septicaemia/bacteraemia), quality of life, acute and delayed serious adverse events and new cardiac events as the critical outcomes for decision making. Other important outcomes included the number of patients transfused, the number of units transfused and length of stay in hospital. |
Trade off between clinical benefits and harms | No evidence was identified to independently evaluate the effect of different thresholds or targets of haemoglobin concentration for blood transfusion because these interventions were linked in most studies. Evidence from the combined review of RBC thresholds and RBC targets comparing restrictive and liberal strategies suggested that the use of restrictive strategies (use of restrictive thresholds and targets) had a positive effect on important outcomes such as number of patients transfused, and number of units transfused. The evidence suggested that there was no difference between the groups with respect to mortality, adverse events, new cardiac events, length of hospital stay, quality of life and infection, but there was considerable uncertainty within the effect estimates to allow confident interpretation of clinical benefit or harm for these outcomes. There was evidence from three studies comparing restrictive thresholds with liberal thresholds in children. The evidence showed that there was a clinically important benefit with restrictive strategies in children with respect to the number of patients transfused. The evidence suggested that lower volumes and less number of units may be transfused in children with restrictive strategies, but there was some uncertainty. The evidence suggested that there may be no difference between the groups with respect to mortality at 30 days, ICU length of stay, adverse events and infections (nosocomial infection), however there was considerable uncertainty within the effect estimates to allow confident interpretation of clinical benefit or harm for these outcomes. No evidence was identified for the critical outcome quality of life in children. There was no clear evidence that using restrictive strategies (use of restrictive thresholds and targets) would harm the general population when considering RBC transfusions. Based on the evidence above, the GDG recommended the use of restrictive thresholds and targets for RBC transfusion for the majority of patients. The GDG specifically considered the importance of ischaemic heart disease and cardiovascular disease. No evidence was identified for patients with co-existing chronic ischaemic disease, including chronic cardiovascular disease. One study reported a sub-group analysis of the data in this group, and this was used to inform the wider consensus of the GDG on this topic135. The term ‘acute coronary syndromes’ encompasses a range of conditions including unstable angina, non-ST-segment-elevation myocardial infarction (NSTEMI) and ST-segment-elevation myocardial infarction (STEMI). All are due to a sudden reduction of blood flow to the heart, usually caused by the rupture of an atherosclerotic plaque within the wall of a coronary artery, and may cause the formation of a blood clot. Some evidence was identified for patients with acute coronary syndrome, but this was inconclusive. For patients with acute and chronic coronary artery disease the GDG agreed there are physiological reasons that anaemia may be less well tolerated, which include the fact that the heart relies more on increasing coronary blood flow to meet increasing oxygen demands than other organs (which can extract a greater proportion of oxygen from blood if required). Myocardial ischaemia occurs when coronary blood flow cannot meet the heart's oxygen requirement, either because a coronary blood vessel is occluded or when coronary blood flow is insufficient. This can occur in patients with a primary cardiac event, but might also affect patients with chronic coronary disease in whom other factors decrease coronary blood flow, such as shock or hypotension. The GDG noted that for many of the studies comparing restrictive with liberal strategies in patient groups with a high incidence of, or risk of, concurrent chronic cardiovascular disease the restrictive transfusion trigger was 80 g/L or greater. This was notably the case in the FOCUS trial (Carson 2011), which was undertaken in elderly patients with a prevalence of known cardiovascular disease undergoing emergency hip fracture repair. The GDG also noted that in a post-hoc analysis of the sub-group of patients in a large study of critically ill patients who had ischaemic heart disease (TRICC trial) (Hebert 1999), and in an a priori defined sub-group of patients with cardiovascular disease enrolled in a trial of restrictive versus liberal transfusion triggers for managing septic shock (Holst 2014), the mortality was lower with liberal transfusion. The restrictive transfusion trigger was 70 g/L in both of these trials. The GDG also took into account the findings of a recent large trial in patients undergoing cardiac surgery (TITRe2 trial) (Murphy 2015) in which mortality was reported to be lower in the liberal groups. Although the population studied in this trial had undergone cardiac surgery rather than presenting with an acute coronary syndrome, the GDG considered the evidence potentially relevant. The GDG therefore acknowledged the insufficient evidence for patients with coronary artery disease. For these patients the GDG agreed that clinical judgement was needed on an individual patient basis using information about disease severity and cardiovascular status (for example blood pressure and heart rate). Although restrictive strategies may be safe, the actual transfusion threshold and target haemoglobin may need to be higher than used for patients without coronary disease. . A research recommendation was made for red cell transfusion thresholds in patients with pre-existing coronary artery disease. The GDG discussed the specific needs of patients with major haemorrhage. It was agreed that in this clinical situation, status changes quickly and unpredictably and clinical judgement in relation to RBC transfusion is needed. In addition, the haemoglobin concentration may not reflect the circulating blood volume and the immediate risk to the patient is from hypovolaemia. |
Economic considerations | No economic analyses were identified to independently evaluate the effect of different targets of haemoglobin concentration for blood transfusion because this intervention was linked to thresholds. One cost-consequence analysis was identified which compared a restrictive to a liberal RBC transfusion threshold in a non-emergency cardiac surgery population (>16 years old). This analysis found that a restrictive threshold was less costly (saving £178 per patient) but had 2.1% more serious infection or ischaemic events, 0.7% more all-cause mortality at 30 days, 1.6% more all-cause mortality at 90 days, 1.2% more clinically significant pulmonary complications and fewer units of red blood cells transfused (saves 1 unit) compared to a liberal threshold. EQ-5D and hospital length of stay were the same in both groups. This analysis was assessed as partially applicable with potentially serious limitations. This analysis may not reflect full body of evidence for this comparison as it was based on 1 of 34 studies included in the clinical review for RBC thresholds. One cost-effectiveness analysis was identified which compared a restrictive RBC transfusion strategy to a liberal transfusion strategy. This analysis found that in anaemic, older, critically ill patients requiring prolonged mechanical ventilation, a restrictive RBC strategy was more costly and more effective than a liberal RBC transfusion strategy (ICER: £253,681 per life year gained). The increased cost of the restrictive threshold strategy compared to the liberal threshold strategy was largely due to an increased intensive care unit and hospital length of stay. Although, this study did not report cost-effectiveness in terms of cost per QALY it suggests that a restrictive threshold may not be cost-effective. The GDG noted that healthcare costs were high and mostly attributable to the ICU admission. The GDG also highlighted that the mortality difference reported could have occurred by chance due to the sample size in the trial. The GDG concluded that it is not possible to extrapolate the findings in this study to other transfusion recipients and therefore decided not to make a recommendation in this subgroup of patients based on this evidence. In this analysis, the higher cost of restrictive strategy was driven by an increased length of stay. However, the GDG noted that the clinical evidence in all other patients receiving transfusion showed a shorter length of stay when using a restrictive strategy compared with a liberal one. No relevant economic evaluations comparing different thresholds or targets of haemoglobin concentration for RBC transfusion in other populations were identified. The cost of RBC transfusion was considered by the GDG. Allogeneic RBC cost £122.09 per unit in England and North Wales.219 It was noted that this cost does not include all costs associated with a transfusion such as staff time, disposables, storage, wastage and laboratory tests. As part of the health economic model developed in this guideline, the additional cost associated with transfusion was estimated to be £70 per first unit transfused. Of note this estimate does not include costs associated with hospital stay or with the management of transfusion-related complications. Furthermore, these costs do not include consideration of the additional laboratory and clinical workload of taking or testing additional samples. Although not all costs and cost savings were quantifiable, the GDG were confident that restrictive strategy for RBC transfusion would be cheaper than a liberal one. Given that it was also concluded that the clinical evidence did not suggest harm from a restrictive strategy it was therefore judged likely to be cost-effective. The GDG considered that for specific subgroups such as in patients with major haemorrhage or acute coronary syndrome it may not be appropriate to use a restrictive threshold, and the economic savings from a restrictive transfusion strategy could be outweighed by the risk to patients of hypovolaemia and low haemoglobin concentration, respectively. Therefore, based on the clinical and economic evidence, the GDG agreed that a restrictive transfusion strategy should be offered to people in the absence of major haemorrhage or acute coronary syndrome. |
Quality of evidence | The quality of evidence for all outcomes was low or very low by GRADE criteria for the combined RBC thresholds and RBC targets review This was largely because of risk of bias arising from a lack of allocation concealment, inadequate blinding, inconsistency and serious or very serious imprecision. The GDG noted that for some important patient groups there is a physiological rationale for less restrictive haemoglobin transfusion thresholds such as those with brain injury, acute and chronic cardiovascular disease, but evidence from these groups was of poor quality. The GDG also recognised that some patients with inherited haemoglobin and/or red cell abnormalities, such as sickle cell anaemia or thalassaemia, may have a clinical indication for transfusion to a higher haemoglobin concentration. Similarly, some patients with chronic anaemia resulting from bone marrow failure, for example myelodysplastic syndromes, may require an individual higher haemoglobin target to minimise symptoms and improve quality of life. The economic evaluation was assessed as partially applicable with potentially serious limitations. |
Other considerations | The recommendation applies to adults and children requiring RBC transfusions. Although trauma patients were outside the scope of this guideline, the review included some studies with trauma patients. This was agreed because the population was not markedly different from this guideline's population, for example, some patients in critical care. The GDG discussed that the recommendation would, however, not be applicable to patients with major haemorrhage and it was important to highlight this in the above recommendation. For patients with cyanotic congenital heart disease, the GDG agreed that the reduced arterial oxygen content increased the risk from a low haemoglobin concentration. The GDG noted that some important patient groups in whom there is a physiological rationale, a higher haemoglobin threshold for transfusion may have clinical benefit such as those with brain injury, acute and chronic cardiovascular disease, but evidence from these groups was of poor quality. It was noted that there were two on-going trials comparing restrictive and liberal threshold strategies, the TRIGGER trial (transfusion in gastrointestinal bleeding) and the TRIST trial (transfusion of red cell in haematopoietic stem cell transplantation). The GDG also noted that recommending restrictive thresholds may be a sensible way of managing blood resource. The TRIGGER trial, an open label trial in patients with acute gastro-intestinal bleeding, was published following consultation of this guideline. The study was not included in the clinical review but the study results do not conflict with the recommendation.144 |
Recommendations |
|
---|---|
Relative values of different outcomes | The GDG considered all-cause mortality at 30 days, infections (including pneumonia, surgical site infection, UTI and septicaemia/bacteraemia), quality of life, acute and delayed serious adverse events and new cardiac events as the critical outcomes for decision making. Other important outcomes included the number of patients transfused, the number of units transfused and length of stay in hospital. |
Trade off between clinical benefits and harms | The evidence from the combined review comparing restrictive and liberal targets in adults suggested that there was clinically important benefit with the use of restrictive targets for the outcomes number of patients transfused and number of units transfused. The evidence suggested that there may be clinical benefit with respect to mortality (all -cause at 30 days), new cardiac events, length of hospital stay and adverse events in patients being transfuse to restrictive target levels, however there was some uncertainty within the effect estimates to allow confident interpretation of clinical benefit or harm for these outcomes. No difference was observed between the groups with respect to incidence of infections (pneumonia). No evidence was identified for outcome quality of life. There was evidence from one study comparing restrictive and liberal targets in children. The evidence showed clinically important benefit for restrictive strategy for the outcome number of patients transfused. The evidence suggested that there may be a benefit with the use of restrictive targets with respect to infections; however there was considerable uncertainty within the effect estimate. The evidence suggested that there may be no difference between groups with respect to the total volume of RBC transfused, mortality, ICU length of stay and pulmonary oedema, but there was considerable uncertainty within the effect estimates to allow confident interpretation of clinical benefit or harm for these outcomes. No evidence was identified for the critical outcome quality of life in children. The GDG discussed the specific thresholds and target levels for RBC transfusion and based the recommendation on the threshold and target ranges used in the studies. Many of the studies that used the most restrictive thresholds (Hb 70g/L) have been conducted in the sickest patient groups, for example critically ill patients in the ICU. Most other studies used a restrictive haemoglobin threshold of 80 g/L. The GDG considered that the safety of Hb triggers of 70g/L in these situations, especially for younger adult patients and those without cardiovascular co-morbidity, supported this threshold for most patients. |
Economic considerations | No economic evidence was identified to independently evaluate the effect of different thresholds or targets of haemoglobin concentration for blood transfusion. The specific threshold and targets outlined for a restrictive strategy are based on those reported in majority of the evidence identified in the clinical review. Economic considerations for a restrictive strategy are described in the LETR above. |
Quality of evidence | The quality of evidence for all outcomes was low or very low by GRADE criteria for both the combined RBC thresholds and RBC targets review This was largely because of risk of bias arising from a lack of allocation concealment, inadequate blinding and serious or very serious imprecision. There was also considerable variation within the studies in the levels of haemoglobin concentration used for haemoglobin targets. The economic evaluation was assessed as partially applicable with potentially serious limitations. The GDG agreed that the same recommendations should apply for children as for adults. |
Other considerations | The recommendation applies to adults and children requiring RBC transfusions. The GDG discussed that it was important to assess the patient's signs and symptoms for example the presence of chest pain, breathlessness, ECG changes, or other signs of myocardial ischaemia prior to setting transfusion thresholds. It was noted that there were two on-going trials comparing restrictive and liberal threshold strategies, the TRIGGER trial (transfusion in gastrointestinal bleeding and the TRIST trial (transfusion of red cell in haematopoietic stem cell transplantation). The TRIGGER trial, an open label trial in patients with acute gastro-intestinal bleeding, was published following consultation of this guideline. The study was not included in the clinical review but the study results do not conflict with the recommendation.144 The GDG noted that this recommendation, in the context of a cross cutting guideline focussing on the general principles of transfusion and the appropriate use of blood does not exclude any specific clinical condition, including oncology patients. However the detailed management of specific clinical conditions was not considered and the physician should follow appropriate guidance specific to clinical conditions, as appropriate. |
Recommendations |
|
---|---|
Relative values of different outcomes | The GDG considered all-cause mortality at 30 days, infections (including pneumonia, surgical site infection, UTI and septicaemia/bacteraemia), quality of life, acute and delayed serious adverse events and new cardiac events as the critical outcomes for decision making. Other important outcomes included the number of patients transfused, the number of units transfused and length of stay in hospital. |
Trade off between clinical benefits and harms | There was evidence from two studies comparing restrictive strategy with liberal strategy in patients with acute coronary syndrome. Evidence from these studies suggested benefit with the use of liberal thresholds compared with restrictive thresholds with respect to critical outcome mortality but suggested that restrictive thresholds may be better in this subgroup at reducing the number of units of allogeneic blood transfused and the length of stay in hospital. The quality of evidence was very low. The term ‘acute coronary syndromes’ encompasses a range of conditions including unstable angina, non-ST-segment-elevation myocardial infarction (NSTEMI) and ST-segment-elevation myocardial infarction (STEMI). All are due to a sudden reduction of blood flow to the heart, usually caused by the rupture of an atherosclerotic plaque within the wall of a coronary artery, and may cause the formation of a blood clot {NICE quality standard [QS68]} For patients with acute coronary artery disease the GDG agreed there are physiological reasons that anaemia may be less well tolerated, which include the fact that the heart relies more on increasing coronary blood flow to meet increasing oxygen demands than other organs (which can extract a greater proportion of oxygen from blood if required). Myocardial ischaemia occurs when coronary blood flow cannot meet the heart's oxygen requirement, either because a coronary blood vessel is occluded or when coronary blood flow is insufficient. The GDG also took into account the findings of a recent large trial in patients undergoing cardiac surgery (Murphy 2015) in which mortality was reported to be lower in the liberal group. Although the population studied in this trial had undergone cardiac surgery rather than presenting with an acute coronary syndrome, the GDG considered the evidence potentially relevant. The GDG agreed to recommend higher threshold levels for patients with acute coronary syndrome. Based on the thresholds used in the studies, the GDG recommended a level of 80-100 g/litre as the transfusion threshold for this group. The recommendation was based on the evidence and consensus expert opinion of the GDG members. |
Economic considerations | No relevant economic evaluations comparing different thresholds or targets of haemoglobin concentration for RBC transfusion in people with acute coronary syndrome were identified. The cost of RBC transfusion was considered by the GDG. Allogeneic RBC cost £122.09 per unit in England and North Wales.219 It was noted that this cost does not include all costs associated with a transfusion such as staff time, disposables, storage, wastage and laboratory tests. As part of the health economic model developed in this guideline, the additional cost associated with transfusion was estimated to be £70 per first unit transfused. Of note this estimate does not include costs associated with hospital stay or with the management of transfusion-related complications. Furthermore, these costs do not include consideration of the additional laboratory and clinical workload of taking or testing additional samples. The GDG considered that for specific subgroups such as in patients with acute coronary syndrome it may not be appropriate to use the same restrictive thresholds as for patients without acute coronary disease, and the economic savings from a restrictive transfusion strategy may be outweighed by the risk to patients of lower haemoglobin concentrations. The specific threshold and targets outlined for this subgroup of patients are based on those reported in the evidence identified in the clinical review and consensus expert opinion of GDG members. |
Quality of evidence | The quality of evidence for all outcomes was low or very low by GRADE criteria. This was largely because of risk of bias arising from a lack of allocation concealment, and serious or very serious imprecision. The GDG considered this evidence was not relevant to children, with the exception of those with cardiac disease that could compromise coronary blood flow or oxygen supply to heart muscle. |
Other considerations | The recommendation applies to adults with acute coronary syndrome requiring RBC transfusions. The GDG noted that for some important patient groups in whom there is a physiological rationale, a higher haemoglobin threshold for transfusion may have clinical benefit such as those with brain injury, acute and chronic cardiovascular disease, but evidence from these groups was limited and of poor quality. The literature suggests that there may be some evidence of harm with the use of restrictive red blood cell thresholds in populations with coronary ischaemia at baseline. In this guideline a level of 80–100 g/litre was used for patients with acute coronary syndrome, but further studies are needed to determine the optimal transfusion threshold for patients with chronic cardiovascular disease. The GDG agreed that further research was required in this area and drafted a research recommendation (see section 10.8). |
Recommendations |
|
---|---|
Relative values of different outcomes | The GDG considered all-cause mortality at 30 days, infections (including pneumonia, surgical site infection, UTI and septicaemia/bacteraemia), quality of life, acute and delayed serious adverse events and new cardiac events as the critical outcomes for decision making. Other important outcomes included the number of patients transfused, the number of units transfused and length of stay in hospital. |
Trade off between clinical benefits and harms | No evidence was identified for this recommendation. The GDG discussed the specific needs of patients who received transfusions for chronic anaemia that is not responsive to alternative interventions, including patients with inherited or acquired red cell or bone marrow disorders, including. It was agreed that these patients were a separate group who may not benefit from the thresholds and target levels set for the overall population. This was because the severity of anaemia may relate to an individual's level of symptoms such as fatigue and breathlessness, and their quality of life. These patients may need to have specific thresholds and targets set for them individually after clinical assessment. The recommendation was based on the consensus expert opinion of the GDG members. |
Economic considerations | No relevant economic evaluations comparing different thresholds or targets of haemoglobin concentration for RBC transfusion people with chronic anaemia were identified. The cost of RBC transfusion was considered by the GDG. Allogeneic RBC cost £122.09 per unit in England and North Wales.219 It was noted that this cost does not include all costs associated with a transfusion such as staff time, disposables, storage, wastage and laboratory tests. As part of the health economic model developed in this guideline, the additional cost associated with transfusion was estimated to be £70 per first unit transfused. Of note this estimate does not include costs associated with hospital stay or with the management of transfusion-related complications. Furthermore, these costs do not include consideration of the additional laboratory and clinical workload of taking or testing additional samples. The GDG considered that for this group of patients who require regular blood transfusion for chronic anaemia, setting individual thresholds and targets after clinical assessment would be current practice and would not have a significant economic impact. |
Quality of evidence | No evidence was identified for this recommendation and the recommendation was based on the consensus expert opinion of the GDG members. |
Other considerations | The recommendation applies to adults and children requiring RBC transfusions. |
10.8. Research Recommendations
- 3.
Red Blood Cell Transfusion: What is the clinical and cost effectiveness of restrictive compared with liberal red blood cell thresholds and targets for patients with chronic cardiovascular disease?
- Why this is important: The literature suggests that there may be some evidence of harm with the use of restrictive red blood cell thresholds in populations with coronary ischaemia at baseline. In this guideline a level of 80-100 g/litre was used for patients with acute coronary syndrome, but further studies are needed to determine the optimal transfusion threshold for patients with chronic cardiovascular disease.
- Review question: What is the clinical- and cost-effectiveness of red blood cell transfusion at different haemoglobin concentrations?
- Review question: What is the clinical- and cost-effectiveness of different target levels of post-transfusion haemoglobin concentrations for red blood cell transfusion?
- Methodology of clinical evidence review (threshold haemoglobin concentrations and target haemoglobin levels for blood transfusion)
- Clinical evidence
- Economic evidence
- Evidence statements
- Recommendations and link to evidence
- Research Recommendations
- Red blood cell: thresholds and targets - Blood TransfusionRed blood cell: thresholds and targets - Blood Transfusion
Your browsing activity is empty.
Activity recording is turned off.
See more...