Harnessing the potential of data-driven strategies to optimise transfusion practice

No one doubts the significant variation in the practice of transfusion medicine. Common examples are the variability in transfusion thresholds and the use of tranexamic acid for surgery with likely high blood loss despite evidence-based standards. There is a long history of applying different strategies to address this variation, including education, clinical guidelines, audit and feedback, but the effectiveness and cost-effectiveness of these initiatives remains unclear. Advances in computerised decision support systems and the application of novel electronic capabilities offer alternative approaches to improving transfusion practice. In England, the National Institute for Health and Care Research funded a Blood and Transplant Research Unit (BTRU) programme focussing on 'A data-enabled programme of research to improve transfusion practices'. The overarching aim of the BTRU is to accelerate the development of data-driven methods to optimise the use of blood and transfusion alternatives, and to integrate them within routine practice to improve patient outcomes. One particular area of focus is implementation science to address variation in practice.


I N TRODUC TION
Variation in clinical practice is a widely recognised phenomenon and to a degree, this is an inevitable feature of healthcare systems, including the National Health Service (NHS).Documentation of variation in practice does however serve a critical purpose in informing major national initiatives.In England, for example, the Getting it Right First Time (GIRFT) programme, is a NHS programme looking to improve the quality of care and reduce cost by reducing variation. 14][5] In all clinical specialities including transfusion medicine, there is unpredictability in how research evidence is adopted into clinical practice.
The general aim of this article is to provide a background to the work of the 'Blood and Transplant Research Unit' (BTRU) research programme funded by the NIHR on 'A data-enabled programme of research to improve transfusion practices'.Specific objectives of this article were to highlight the variation in transfusion practice and the use of transfusion alternatives commonly seen in medicine, and to provide an overview of strategies to improve uptake of evidence-based practice, including newer approaches based on electronic systems such as computerised decision support systems (CDSS).

VA R I ATION I N TR A NSFUSION CA R E
There are many examples of the complexity and variation in the practice of transfusion medicine: the many types of components that differ in their degree of matching requirements, the blood count and haemostasis test thresholds that trigger their use, the storage age, the volume that is prescribed and the additional specifications that may be applied.7][8][9] Slightly under 2.2 million blood components were issued in 2022 in the UK, 10 yet data from successive audits of practice suggest that as many as 20% of transfused blood components may have been given outside national standards and recommendations (Table 1). 2 A national audit in 2016 of the use of red cells showed that 16% of all red blood cell transfusions were considered inappropriate across 170 sites with 4328 participants. 11The national audit of blood transfusion programme also found inappropriate use of prophylactic platelets (37% of all transfusions) and preprocedure transfusions (19%).A further audit in 2019 (110 sites with 5155 participants) reported that 30% of patients with a pretransfusion Hb of >70 g/L were transfused without adequate clinical reason.This was even higher in patients with acute coronary syndrome or cardiorespiratory disease, where 61% had a pretransfusion Hb of >80 g/L.We have estimated that the number of unnecessary transfusions could be as many as 300-400 000 across the UK costing over £60 M per year.Of note, estimates of purchase costs of blood often do not fully consider the added costs of storage, testing and safe administration in hospitals. 12The importance of addressing inappropriate transfusions is highlighted by real concerns in the most recent Serious Hazards of Transfusion (SHOT) annual report describing cases of two preventable patient deaths and one major morbidity following ABO-incompatible red blood cell transfusion. 10hese national audits indicate limited compliance with guideline recommendations of strategies to promote alternatives or measures to minimise the use of blood.This is a serious concern given that blood is a limited resource. 13reoperative anaemia has been identified as a risk factor for the need for transfusion and for increased morbidity and mortality following surgery.Yet, audits continue to  1). 11Effective patient blood management (PBM) is recommended 14 and yet there is inconsistency in its implementation.A recent French cross-sectional study highlighted many deficiencies in perioperative anaemia management and correction of iron deficiency was poorly implemented. 15Yet data increasingly support the effectiveness of PBM.For example, a recent study in Germany used the records of 1.2 million patients to report a reduction in transfusion rate of 13.7% in the group who underwent active PBM. 16here is an accumulating body of evidence demonstrating the benefits of tranexamic acid (TXA) in surgical patients reducing the risk of major surgical bleeding and transfusion by 25% with no increased risk of thrombosis. 17,18Indeed, Roman et al. concluded that TXA is the single most effective PBM intervention. 19NICE recommended that adults having surgery where blood loss is expected to be moderate (>500 mL) should be offered TXA (NICE NG24) 20 and in 2016 this recommendation became a NICE Quality Standard (NICE QS138). 21Yet the 2021 National Comparative Audit found that only 67% of potentially eligible surgical patients were given TXA (Table 1). 11This means of the 1.5 million major surgeries each year, around 500 000 people do not benefit from TXA, leading to approximately 15 000 otherwise avoidable major bleeding events. 22Correct compliance with this quality standard alone could decrease the demand for blood by 33 000 units per year and save over £5 M in transfusion costs without considering the added benefits to patients.A recent national survey of anaesthetic trainees indicated that the use of TXA varies considerably between surgical specialities as does the availability of policies or the use of checklists to promote its use. 23qually concerning are lower levels of compliance with the NICE quality standard for patient information and education.A national audit found that only 26% of people who received a blood transfusion were given verbal and written information about blood transfusion. 11Another national study found this increases to only 50% in patients who may lack capacity, for example critically ill patients. 24Comparable data exist across other clinical settings; in maternity care, only 1.6% of women were offered full written information on the correct administration of iron and how to maximise absorption. 25There is also a lack of compliance to guidance for pretransfusion testing.A national audit demonstrated that only half of eligible patients with sickle cell disease had received the correct pretransfusion testing indicated in the guidance. 26he slow uptake of evidence-based recommendations is costly, both to patients and the healthcare system alike, with studies showing that PBM interventions can reduce the requirement for blood components translating into important potential cost and resource savings. 19As such, this represents a strategically important issue for service providers, policymakers, healthcare systems and funders.It is a major challenge for transfusion medicine because of its ubiquity throughout almost all areas of hospital practice.

R ECE N T BLOOD SHORTAGE S HIGH LIGHT TH E N E E D TO R E DUCE U N N ECE SSA RY BLOOD USAGE
The occurrence of serious transfusion-transmitted infections since the 1970s and the recent blood shortages in England highlight the importance of only using blood appropriately.Indeed, many countries are experiencing persistent challenges in providing an adequate and timely blood supply after the COVID-19 pandemic. 13The reasons are complex but include changing patterns of blood donation, and modified donor behaviour. 27,28mportantly, it is not only the availability of blood for transfusion that is important for patient care, but also access to the appropriate blood group type.Alloimmunisation to red cell antigens occurs commonly in patients with sickle cell diseaser and transfusion-dependent patients, and, at times, can be a real challenge for blood supply. 26The provision of better matched red blood cell units which might mitigate this risk is highly dependent on the size and diversity of the donor pool.At present, the range of blood groups seen in blood donors in England does not match those commonly present in ethnic minority groups, for example patients with sickle cell disease, thus limiting the ability of the blood service to provide well matched units to minimise the risk of red cell alloimmunisation in these patients.Whether new technologies based on matching genotyped blood donors and recipients, rather than the traditional labour-intensive serological approaches, have a role in addressing and minimising risks of alloimmunisation remains unknown and is the focus of ongoing research (Haem-Match). 29Another concern is the challenge in providing ABO compatible platelet transfusions. 30here are opportunities to improve patient outcomes by expanding both the diversity of the donor pool and the methods by which we match blood transfusions.

POOR LI N K AGE OF BLOOD SERV ICE A N D HOSPITA L IT SYSTE MS LI M ITS EFFORTS TO R E DUCE VA R I ATION
A major limitation in addressing overall blood supply problems is the inability to visualise in real time the full transfusion chain from blood donors to patients (vein-tovein mapping).As a result, changes in blood use are not immediately apparent to blood services, thus limiting their ability to respond to the changes in demand.One option in England has been a pilot 'live-link' between selected hospitals and NHS Blood & Transplant (NHSBT), the blood service in England, to provide near-real-time hospital blood stock and wastage information (vendor-managed inventory).It also faciltates automatic 'top-ups' of individual hospitals' blood stocks to predetermined levels.At present, it has limited current capability, being provided to only 20 out of over 150 hospitals supplied by NHSBT, and it does not allow information to flow from hospitals to NHSBT about the use of blood such as the number and type of patients receiving transfusions.

TH E STR E NGTHS A N D LI M ITATIONS OF I N TERV E N TIONS TO A DDR E SS POOR TR A NSFUSION PR AC TICE
There are a range of interventions to improve transfusion practice and promote adherence to national standards and recommendations.These include education, clinical guidelines, audit and feedback, and CDSS.Soril et al. reviewed the different quality improvement interventions in transfusion and found that the scale of any benefits achieved was often small. 31None of the included studies was graded as high quality; most were single site, often in tertiary specialist settings, and it was difficult to disentangle the effectiveness of different components within multimodal interventions.
The development of guidelines based on the best current evidence for good practice is a common starting point for addressing variation in practice.However, it has long been recognised that the publication of guidelines alone does not lead to change. 32Key barriers to evidence-based transfusion practice include clinicians' limited capacity to keep up to date with an evolving evidence base, resistant beliefs about transfusion benefits and the absence of strong drivers for change. 33Therefore, the dissemination of guidelines often needs to be supported by active implementation strategies.A recent overview using systematic review methodology identified 30 strategies targeting healthcare organisations, healthcare providers and patients to promote guideline implementation, including educational materials and meetings, and audit and feedback. 34][37] Traditional medical education meetings and workshops probably have only modest effects on clinical practice and, to a lesser extent, patient outcomes. 38Greater effects may be associated with a number of features, such as shorter meetings, better attendance, provision of additional take-home material, and targeting educational goals perceived as important.Research on educational approaches as a tool to change practice increasingly emphasises the importance of focussed topics and interactive methods, coupled with tools for self-learning and some form of competency assessment.Use of different teaching modalities to deliver education can be helpful 39,40 and this applies equally to transfusion medicine as to other aspects of healthcare. 41,42The COVID-19 pandemic saw an increased emphasis on distance or virtual learning, often termed e-learning. 43However, a recent survey of e-learning practices in transfusion highlighted continued uncertainty about its effectiveness. 44The main message appears to be one of caution that a single educational approach will be effective in delivering the desired change in practice and that it should form part of a wider implementation strategy, as described in a Cochrane review of audit and feedback. 45Other forms of educational initiative in transfusion such as Transfusion Camp offer a complementary means of education. 46

AU DIT A N D FE E DBACK
Audit and feedback can provide healthcare professionals with summaries of their clinical performance over a specified period of time with the intention of motivating improvement.It generally has modest effects on clinical practice; a meta-analysis of 140 trials of audit and feedback found a median 4.3% absolute effect. 45Effects varied considerably among trials, with a quarter finding large absolute effects on 16% or greater and a quarter finding no or even harmful effects.Larger effects were associated with lower baseline performance, feedback being delivered by a supervisor or colleague, provision of repeated feedback, providing feedback both verbally and in writing, and including clear targets and an action plan.The modest effects of feedback may translate into worthwhile population healthcare benefits when it is scaled up, such as through national clinical audit programmes.Furthermore, feedback based upon existing, routinely collected data offers efficiencies over relatively time-consuming patient case note reviews.However, there is still a gap between what audit and feedback can achieve and what is actually delivered.Analyses of national audit programmes in the UK found that they often did not fully utilise existing evidence on effective feedback methods, for example incorporating action plans. 47,48udit and feedback has a long history in transfusion medicine.However, limited progress with repeated national audits in the UK motivated the AFFINITIE programme (Audit and Feedback INterventions to Increase evidence-based Transfusion practice). 49The researchers first undertook a series of interviews and surveys to better understand the challenges of feedback delivery and effectiveness at hospital sites. 50The AFFINITIE research team then developed and evaluated two empirically and theoretically informed feedback interventions, 'enhanced content' to improve feedback clarity and usability and 'enhanced support' for hospital staff to act on feedback, on transfusion appropriateness.The effectiveness evaluation embedded two linked 2 × 2 cluster-randomised trials in national audits of transfusion for surgery (135 hospitals) and haematological disorders (134 hospitals) respectively.In this way, the trials evaluated the separate and combined effects of enhanced content and enhanced support against standard feedback.The enhanced feedback interventions were found to be no more effective than standard feedback. 51he likely reasons for the lack of effect included a lack of credibility of the audit standards, concerns about the data validity collected by the largely manual-based processes and evidence of variable (and often poor) enactment of the intended feedback at hospital sites.These lessons are directly relevant to the current design and delivery of ongoing or planned transfusion audit and feedback programmes.There remains considerable scope to enhance the effectiveness of audit and feedback programmes, 52 especially by drawing on evidence and expert-informed suggestions for optimising feedback. 53The AFFINITIE programme also demonstrated a methodology for embedding trials within existing largescale quality improvement programmes, thereby improving research efficiency and generalisability.It should be noted that the AFFINITIE researchers did not compare audit and feedback against no form of audit and feedback.Given audit and feedback likely has some effect, the key question remains how to enhance the effects, by testing different forms of audit and feedback. 54

COM PU TER ISE D DECISION SU PPORT SYSTE MS
There is growing interest in the development of CDSS within electronic health records (EHRs) to improve transfusion practice, although to date such systems are infrequently used and limited to specialist centres with high-quality EHRs.CDSS aim to improve safety by reminding clinicians to deliver recommended care and reducing errors in decisionmaking.They can include relatively sophisticated systems linked to patient-specific information allowing rapid implementation and scaling within EHR systems.CDSS have the potential to provide real-time feedback to requesters on appropriate blood ordering, as well as to clinicians on the use of blood.
The transfusion team in Oxford have described their experiences of CDSS, which has been reported to contribute alongside other blood management strategies to a reduction in total blood product costs of around 25% which may equate to a saving of around £1 million/annum without any negative impact on patient outcomes. 55Importantly while much of the focus has been on red blood cell transfusions, these tools may be as effective for transfusions of other blood components such as plasma and platelets. 56In another system, a recent best practice alert has been tested in a small, randomised trial and was reported to reduce platelet requirements. 57DSS are ubiquitous in primary and secondary care EHR systems.However, a systematic review of 108 studies (including 94 randomised trials) of CDSS found only a modest 5.8% improvement in the proportion of patients receiving evidence-based care and only a marginal 0.3% improvement in clinical endpoints. 58One striking finding was the significant variability in effect sizes; while some studies found large effects, others found none.System features and clinical context incompletely explained this variability, leading to the conclusion that the current literature 'provides little guidance for identifying the circumstances under which clinical decision support system interventions produce worthwhile improvements in care'.Furthermore, there are well-recognised problems of alert fatigue, distraction, and poor fits with clinician and patient needs during consultations. 59,60For example, clinicians are prone to ignore or discount multiple hazards highlighted in prescribing safety alerts, especially as alerts typically appear after they have made decisions to prescribe.
A focussed review on CDSS in transfusion described 20 separate studies, but nearly all studies were 'before and after' designs rather than randomised controlled trials or other more methodologically robust clinical trial designs. 61The review concluded that while implementation of a CDSS might improve red blood cell usage, there were many uncertainties regarding the optimal features of these systems, 62 as well as their impact on cost savings, effects on patient outcomes and the sustainability of any effects.The authors also made recommendations for standardised reporting of outcomes, not least the nature of the algorithm used.The integration of such systems into a diverse NHS landscape of electronic systems will undoubtedly be complex, but adherence to recommended standards will be of particular importance to ensure their widespread adoption and studies to determine their optimal configuration will be needed.

THE VALUE OF HOSPITAL DATASETS AND DATA REPOSITORIES TO EXPLORE UNWARRANTED VARIATION AND UNDERSTAND PRACTICE
This article started with a discussion on variation in transfusion practice.Accurate and timely data on the patterns of blood usage are key to understand and prioritise what needs to be done to improve practice.However, these data are often not readily available.Greater use of data within EHRs has huge potential for identifying variation in transfusion practice and for developing measures to promote appropriate use of blood, reduce blood wastage, improve blood stock management and reduce healthcare costs. 63Proof of principle work has established the feasibility of electronic collection of transfusion datasets from multiple hospitals in England, with value for more efficient and timely benchmarking of practices. 64Larger scale data collaborations for research are also underway (NIHR Health Informatics Collaborative, Table 2), but require significant bioinformatics support to standardise the transfusion data across multiple sites with their differing LIMS and EHRS.The initiative in England to create subnational secure data environments 65 should allow large-scale linkage of data between hospitals, primary care and NHSBT.This will provide a wealth of possibilities to transform patient care.For example, it may be possible for the first time to perform comprehensive studies of anaemia 13652141, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bjh.19158by University College London UCL Library Services, Wiley Online Library on [02/01/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License in surgery and pregnancy in primary and secondary care.Additionally, while current datasets available for research in transfusion medicine are retrospective, they could be combined with more powerful computational servers and cloud software to open up the potential for future iterations to be prospective, or even real-time.
The availability of large-scale datasets creates new opportunities for transfusion research (Table 2).One feature of the research environment in COVID-19 has been the creation of data repositories enabling many features of the social, behavioural, public health, management and economic impacts of the global pandemic to be explored, but which may have broader applicability to areas including transfusion medicine.A NIHR British Heart Foundation Cardiovascular Partnership links population healthcare datasets across the UK to study the relationships between COVID-19 and cardiovascular diseases.Another relevant dataset with over 50 000 patients is the Perioperative Quality Improvement Programme (PQIP), which supports the feedback of national variation in delivery of perioperative PBM interventions.It is being analysed to explore the effects of anaemia before surgery and how treatments such as iron affect patient outcomes. 66reoperative anaemia should be identified and treated at the earliest opportunity when surgery may be indicated, and ideally not in a preoperative assessment clinic in the days prior to surgery.Linkage between hospital and primary care data (e.g.Clinical Practice Research Datalink [CPRD] primary care database) could open up additional treatment opportunities which could positively impact patients' well-being and long-term quality of life.Linkages between a national haemoglobinopathy registry, hospital data and NHSBT could facilitate better and timely access to matched blood for patients with complex antibodies.Although it should be noted that many studies have audited the clinical information of patients who receive blood transfusions, the datasets used in these large-scale approaches should allow us to also consider patients who may have benefitted from blood components even when they were not administered or indicated.They may also provide a framework for building vein-to-vein linkages from the donor to the recipient.8][69] Over a million patients are now included in the German Patient Blood Management Network Registry, allowing a more comprehensive analysis of the relationships between anaemia, comorbidities and red blood cell transfusion. 70The more recent iteration of the Recipient Epidemiology and Donor Evaluation Study (REDS) programme is creating data repositories to include younger age groups. 71Using data at scale in large datasets should mean the study population is more representative of the entire population. 72Although evidence exists that inequality (e.g.ethnicity and deprivation) affects population health, this aspect is perhaps under-researched in transfusion medicine, yet data are emerging that practices need to be scrutinised. 73,746][77] ML may support the development of algorithms to predict transfusion requirements which could help reduce risk, improve patient outcomes and predict blood stock requirements.Directly embedding learning systems within EHRs has huge potential but their implementation remains a significant challenge. 78,799][90][91] How widely and rapidly these findings can be rolled out in hospital information systems remains to be seen.

N EW DEV E LOPM E N TS I N E NGL A N D: TH E BTRU I N DATA-DR I V E N TR A NSF USION PR AC TICE
In England, NIHR provided funding for 5 years for a BTRU programme focussing on 'A data-enabled programme of research to improve transfusion practices'.The overarching aim of the infrastructure grant is to accelerate the development of data-driven methods to optimise blood use and integrate them within routine practice to improve patient outcomes.The core structures of the BTRU are shown in Figure 1, with examples of activities shown in Table 2.This collaborative programme brings together haematologists, methodologists, surgeons, anaesthetists, data scientists and implementation experts to create a cross-cutting multidisciplinary team including both early career and established researchers, with embedded patient and public members.The work of the implementation group includes optimising feedback reports, in comparison with international F I G U R E 1 An overview of the BTRU.Cross-cutting research work-packages will address the following themes: (1) using hospital data to understand and address variations in blood use in hospitals; (2) using electronic systems to improve the sharing of information between hospitals and blood services for improving the blood supply chain; (3) using data from hospitals and GP practices to develop electronic tools to improve the outcomes for patients who might need transfusion; (4) investigating the costings of different pathways and processes for transfusion, given the need to understand how any electronic systems deliver value for money.recommendations, 53 for example, in a collaboration between the BTRU and the Blood Stocks Management Scheme (BSMS).The graphical abstract provides an overview of such strategies to address variation, in this case for stock inventory and rates of wastage between hospitals in England.
At the heart of the BTRU programme is planning for novel data linkages and large-scale data repositories including blood transfusion.Later ambitions will be explored by promoting and enabling an integrated process of monitoring and managing blood needs and use, including hospital-level databases, and data linkages between NHSBT as blood supplier to hospitals, to and from hospitals and also to primary care, but it is recognised that these developments will take time and additional resource to deliver.Understanding the cost-effectiveness and efficiency gains from electronic systems for better managing safe transfusions in hospitals and blood supply is an important need given the recent SHOT annual report which continues to highlight significant risks to patients and that errors in the transfusion processes still account for the majority of the reports. 10here is increasing evidence to support the impact of meaningful Patient Public Involvement and Engagement (PPIE). 92In the BTRU, a group of patient and public members play a key role bringing the views and needs of patients from a diverse range of backgrounds to our work, representing many communities commonly under-represented in healthcare research.They are involved in the research process from the design of research questions, data collection and analysis to the dissemination and implementation of our findings.PPIE views have driven a novel strand of transfusion research examining how differences in people's geographical location, ethnicity and socio-economic standing may affect the care they receive.From the perspective of donors, there is a recurring expectation that donated blood will be used to full effect and not be wasted. 93Patients are clear that unnecessary (inappropriate) transfusions should be avoided, and initiatives pursued to address variation in practice.Other relevant PPIE initiatives include reviewing research priorities identified through the activities of two James Lind Priority Setting Partnerships, 94,95 and the 'Choosing Wisely' for blood transfusion campaign. 96echnologies and human factor considerations to deliver improvements in the safety of blood administration include avoiding 'never events', such as ABO-incompatible red blood cell transfusions, remain top priorities for patients.Patients and public members are generally supportive of the use of routine clinical data but within clear boundaries and for specific purposes, and within a framework of national guidance (e.g.Goldacre report). 97

H E L PI NG TO SH A PE TH E BROA DER I N TE R NATIONA L EFFORT
The work of the BTRU should be viewed in the context of activities in many countries that are starting to explore the development of a fully linked electronic 'vein-to-vein' systems from the blood donor to the patient recipient.An exemplar system in Europe is the Swedish-Danish Scandinavian Donation and Transfusions (SCANDAT) database 98 and in North America, work through the REDS programmes.Such systems can be used to explore a range of donor and donation factors on patient outcomes. 99 longer term vision of the BTRU in England is to establish a so-called nationwide 'learning health system' to use rigorous, data-driven methods to continuously improve transfusion practice.Similar initiatives have been described in other international settings. 1003][104][105] These approaches are increasingly discussed in the literature and already used in public policy and in business to deliver cumulative improvements, for example companies 'randomising' potential customers to different presentations of online products to understand what drives purchases.Embedding randomisations efficiently into hospital transfusion systems will be challenging, and more clarity is needed on the optimal approaches for achieving this aim. 57,62These new strategies provide a real opportunity to improve transfusion practice, which is viewed by patient and public members of our PPIE panel as an imperative for researcher teams, given their custodianship of public health data and blood as a donated altruistic resource.Collaborations with colleagues across the globe promote the building of a strong shared learning approach, so that challenges to the use of data are collectively addressed, to ensure improvements benefit both patients and blood donors in all settings, including low-resource country settings.

CONCLUSION
There is a need to address the disparity between an expanding evidence base informing best transfusion practice and its uptake into routine clinical practice.For example, there are more than 48 randomised trials comparing different thresholds for red blood cell transfusion.In contrast, there have been very few randomised trials of approaches to implement their findings. 106,107Initiatives such as the BTRU in Data Driven Transfusion Practice aim to capitalise both on the increasing capacity for the collection of routine patient data and on the advent of interactive electronic systems to provide real-time machine-driven learning and thus effective feedback to individual clinicians and clinical teams to ensure optimal transfusion practice.By ensuring we take advantage of the data already being collected on a daily basis within the NHS, it is envisaged that preventative as well as therapeutic interventions can be optimised and improved across the wide range of specialities where transfusions are used.By utilising the emerging technological advances, we hope to develop an iterative flexible learning system which will have a long-lasting positive impact on patient outcomes 13652141, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bjh.19158by University College London UCL Library Services, Wiley Online Library on [02/01/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License including their quality of life by improving transfusion practice for all the patients that need blood.

AU T HOR C ON T R I BU T ION S
H. G. Evans, M. F. Murphy, R. Foy and S. J. Stanworth wrote the first draft of the manuscript.All authors contributed to the final version.AC K NO W L E D GE M E N T S This publication is supported by the National Institute for Health and Care Research (NIHR) Blood and Transplant Results of three recent national transfusion audits.For example, the 2021 National Audit of the National Institute for Health and Care Excellence (NICE) Quality Standard of Blood Transfusion found that only 58% of elective surgical patients were treated with iron preoperatively (Table T A B L E1 16% of all red blood cell transfusions given were considered inappropriate 37% of prophylactic platelet transfusions were inappropriate 19% of transfusions carried out preprocedure were inappropriate 6% of therapeutic platelet transfusions were inappropriate Note: National Comparative Audit of Blood Transfusion.https:// hospi tal.blood.co.uk/ audits/ natio nal-compa rativ e-audit/ .Accessed 31.7.2023. 1113652141, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bjh.19158by University College London UCL Library Services, Wiley Online Library on [02/01/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License demonstrate suboptimal identification and management of preoperative anaemia.
Examples of data repositories of potential relevance to patient blood management in England.Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bjh.19158by University College London UCL Library Services, Wiley Online Library on [02/01/2024].See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions)on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License T A B L E 2