Practical Transfusion Medicine 4th Ed.

47. Variation in transfusion practice and how to influence clinicians' use of blood in hospitals

Simon J. Stanworth1, J.J. Francis2, Michael F. Murphy1 & Alan T. Tinmouth3

1University of Oxford and NHS Blood and Transplant and Department of Haematology, John Radcliffe Hospital, Oxford, UK

2Health Services Research Unit, University of Aberdeen, Aberdeen, Scotland

3General Hematology and Transfusion Medicine, Division of Hematology, Ottawa Hospital and University of Ottawa Centre for Transfusion Research, Clinical Epidemiology Program, The Ottawa Health Research Institute, Ottawa, Ontario, Canada


Changing established patterns of transfusion practice is not easy. As the evidence base for transfusion medicine advances, there is an increasing need to ensure that important new research is rapidly implemented and that practice which is shown to be less effective (or cost-inefficient) is discontinued. Many of the methods used to facilitate change in clinical behaviour are familiar to hospital healthcare workers in the field of transfusion medicine. They include:

·        guideline adoption and dissemination;

·        educational materials or teaching sessions;

·        the use of audits with feedback; and

·        reminders.

But which is best or, put another way, which intervention delivers more effective and sustained change in transfusion prescribing behaviour in a cost-effective manner? Answers to these questions may help address comments frequently expressed at meetings, such as ‘What is the point of doing yet another audit?’ The wider volume of literature relevant to promoting clinical effectiveness and the uptake of effective practice is increasing substantially and this chapter will look at these issues with regard to reducing inappropriate transfusion practice.

Interventions to change practice

The different means of achieving changes in practice have been reviewed by many groups and cover a wide range of candidate ‘strategies’. One (empirically based) general taxonomy of interventions developed and used by the Cochrane Collaboration Effective Practice and Organization of Care (EPOC) Group is reproduced below, to illustrate the breadth of different interventions.

a. Professional

·        audit and feedback;

·        reminders;

·        educational – groups, meetings, outreach visits;

·        patient-mediated interventions; and

·        local opinion leaders.

b. Financial

·        fee-for-service and

·        provider incentives.

c. Organizational

·        clinical multidisciplinary teams and

·        changes to the setting/site of service delivery.

d. Regulatory

·        changes in medical liability and

·        peer review.

Professional interventions represent the main group of ‘strategies’ with which healthcare workers in transfusion are familiar. Audit and feedback is defined by EPOC as ‘any summary of clinical performance of health care over a specified period of time. The summary may also have included recommendations for clinical action.’ Audit and feedback is a very well-established tool for evaluation and improvement of the quality of clinical care and quality assurance in hospitals with the objective to improve patient care and outcomes through careful review of care provided against explicit standards. Where indicated, changes are then implemented at an individual, team or service level. Further monitoring is used to confirm improvement in healthcare delivery. Audit and feedback is widely used across the National Health Service (NHS) in the UK as a quality improvement method and is used by transfusion services worldwide to monitor parameters such as the number of units transfused/patient for a given procedure or condition and/or the proportion of patients receiving any blood. The proportion of transfusions deemed to be appropriate is sometimes assessed (also see Chapter 23). However, as currently used, audit and feedback does not work consistently and the effects are generally considered small to moderate. The method of feedback (e.g. at the individual or team level) is acknowledged as critical to delivering change, and will be discussed again later.

Educational outreach involves the ‘use of a trained person who meets with providers in their practice settings to give information with the intent of changing the provider's practice’. Educational meetings are defined by EPOC as ‘participation of healthcare providers in conferences, lectures, workshops or traineeships’. Other examples of interventions include the use of local opinion leaders who are ‘providers nominated by their colleagues as educationally influential’. Reminders are ‘patient- or encounter-specific information, provided verbally, on paper or on a computer screen, which is designed or intended to prompt a health professional to recall information’. Computer-aided decision supports would also be examples of reminders. In transfusion practice, reminders have been incorporated into transfusion request forms or computer order entry.

Changes in practice may also be instigated by combinations of these interventions and multifaceted interventions include ‘any intervention including two or more components’. Multifaceted interventions are likely to be more costly than single-component interventions. A recent example by Rothschild et al. of a combined approach to changing and improving transfusion decisions was a report of a study examining transfusion practices before and after a conventional educational intervention followed by a randomized controlled trial of a decision support intervention based on computerized physician order entry [1]. Education and computerized decision support both decreased the percentage of inappropriate transfusions, although, the residual amount of inappropriate transfusions remained high.

Organizational interventions are also widely used to deliver change in some areas of transfusion practice. In contrast to professional interventions looking at usage of blood, organizational interventions have been designed to deliver safer transfusion practice. Examples might include: the introduction of two-person pretransfusion bedside checking or patient identification using barcoding. Aspects of wider quality assurance systems for delivering safer transfusion practice might also be considered as organizational interventions for change, but these categories will not be considered further in this chapter.

Comparing and enhancing interventions to change practice

Different forms of behavioural interventions are undertaken by hospitals and blood transfusion services worldwide with the aim of changing transfusion practice, but there are few data on their relative or absolute effectiveness. The broad principles of clinical research appraisal should be applied to evaluate all studies of interventions, as for any pharmaceutical intervention in medical care. Randomized controlled trials would be expected to provide the highest level of evidence. For example, a controlled trial by Soumerai et al. of an intensive educational outreach programme (based around printed materials, presentations and face-to-face visits by transfusion medicine specialists) indicated good evidence for improved appropriateness of blood component use [2,3]. What this trial did not perhaps address were questions regarding the sustainability of changes and overall cost effectiveness.

An update of two previous systematic reviews [4,5] has identified and appraised all relevant publications evaluating different interventions in transfusion. The identification of the eligible literature in the field of practice improvement was challenging because of the wider numbers of potentially relevant journals and the lack of specific indexing to trials trying to change transfusion practice. In addition, many of the identified studies reported multiple and different endpoints or defined appropriateness of transfusion in different ways. In total, 43 studies (17 published since 2001) were identified and, taken as a whole, the reported results indicated variable effectiveness of guidelines, audit and feedback, and other interventions such as the implementation of a new transfusion request form or various education initiatives.

In general, interventions for the reduction of transfusion studied in clinical trials seemed to be effective. For red cell transfusions, the reported absolute reductions of inappropriate transfusions were –31% to +4%, an absolute reduction of 0.55 to –0.01 in the number of units transfused per patient and a relative reduction of –75% to +1% for the total number of units transfused (Table 47.1). Due to the lack of emphasis placed on detailing the interventions and use of a multifaceted intervention in most of the studies, determining the relative effectiveness of different types of interventions to change transfusion practice is difficult. However, no specific type of intervention appeared to be clearly more effective than others and there was no clear benefit of multifaceted interventions [4,5]. There were significant limitations to the quality of the evidence. Most of the studies were not controlled trials, but ‘before-and-after’ studies with no concurrent controls. This type of study design cannot account for changes that may occur over time for other reasons, and are also more prone to bias in favour of the intervention. Most were single-centre studies and many were performed more than 10 years ago. The reported successes in these published studies also raise the possibility of a ‘Hawthorne effect’, which describes an initial improvement in performance due to the simple act of observing the performance. These concerns about the true effectiveness and the durability of the effect of these interventions were raised in one study that reported a return to the baseline rate of transfusions three months after the completion of the intervention, and were also reported in a follow-up assessment from another study that also found a return to previous transfusion practice. None of the studies formally reported cost-effectiveness comparisons, and it is also likely that these publications will be susceptible to ‘publication bias’ in that studies with negative results may not have been submitted for publication.

Table 47.1 Changes in red blood cell utilization following introduction of behavioural interventions to change transfusion practice.


The results of the systematic review support the concept that interventions can be successful in changing physicians' transfusion practices. These findings are similar to other studies that have examined the effects of interventions to change physician practice in other settings. However, given the limitations of the studies, there is uncertainty as to the nature of the ‘active ingredients’ of interventions (or combinations of ingredients) that have the maximum effectiveness in delivering sustained (and cost-effective) changes in transfusion practice. For example, considerable scope exists to improve the impact of audit and feedback on patient care and outcomes; as described later, the effectiveness and consistency of audit and feedback may be considerably enhanced by ensuring that the feedback is behaviourally specific and accompanied by a behavioural target and a targeted action plan.

Variation in practice

The variability in transfusion practice should not be a surprise to any healthcare professional involved in this area of medicine. Indeed, in a wider context, all healthcare systems experience inappropriate variation in treatments and treatment rates that suggest there is considerable inappropriate practice. Alongside the public and political anxiety surrounding blood transfusion, this variation in practice has arguably been more comprehensively documented over a longer period of time for transfusion medicine than many other areas of healthcare, and is frequently reported as a key driver for the need for, and use of, interventions to change practice. One of the first and more commonly quoted studies on this topic was reported by the Sanguis Study Group [6]. They evaluated the use of blood components for elective surgery at multiple European hospitals and reported large differences between hospitals and clinical teams in the use of red cell transfusion for the same surgical procedures, with no clear explanation based on patient and clinical factors such as age, preoperative haemoglobin or perioperative blood loss.

Variation in usage has also been reported for other blood components. In a comparison of plasma use in a number of countries, the ratio of frozen plasma (FP) units to red blood cell units transfused varied from 1:3.6 in the USA to 1:8.5 in France [7]. Other studies have reported wide variation in the use of plasma and platelets among centres within the same country, including patients undergoing cardiac surgery and critical care patients. Other studies have suggested that FFP may be associated with the highest rate of inappropriate transfusion, with some studies indicating a rate of inappropriate transfusions as high as 50%.

However, the reasons for variation in transfusion practice are complex. Arguably some of this variation may be expected given the complicated nature of healthcare and wide differences between patients and their responses to intensive medical or surgical treatments, which are frequently the setting for transfusion. Although some variation is to be expected in any healthcare setting, it is variation against an explicit set of criteria or guidelines based on good evidence that is the key issue. When there is good scientific evidence that supports evidence-based transfusions guidelines, then practice variation that continues and deviates from these guidelines is a concern that requires scrutiny.

An increasing number of transfusion trials are being undertaken and reported using clinically relevant outcomes and are providing worthwhile additions to the evidence base for transfusion practice. Interestingly, these trials are beginning to challenge some of the preconceived notions of the clinical benefits of transfusion. For example, in the topic of red cell transfusions, a number of threshold studies are now published [8,9]. The transfusion requirements in critical care (TRICC) trial compared two transfusion thresholds for adult patients admitted to Canadian intensive care units; the results of a subgroup analysis showed a trend towards lower 30-day mortality in the restrictive transfusion group. In the paediatric intensive care unit setting, a similar restrictive transfusion strategy was found not to be inferior to a more liberal strategy. Comparable trials evaluating red cell transfusion strategies in neonates and recently for hip fracture surgery have also been published [9–11]. As a broad generalization, the combined weight of evidence from these trials of red cell transfusion does not support unrestricted use of red cell transfusions in many patient groups and therefore appears to argue against the intuitive desire to raise haemoglobin levels.

In the UK, national comparative audits of transfusion practice provide baseline information on compliance with standards and are undertaken through a collaboration between NHS Blood and Transplant and the Royal College of Physicians. These audits continue to demonstrate that approximately 20% of red cell transfusions, and an even higher proportion of transfusions of plasma and platelets, are not compliant with national guidelines, particularly in medical settings, and suggest that blood usage can be reduced [12,13].

Models of behaviour and change

It is important to study systematically the wider influences on decisions to transfuse in the context of the considerable variation in practice. At one level, inappropriate variation in blood usage may reflect poor knowledge by physicians. However, there are likely to be many other influences on clinical practice and a number of ways to improve and optimize transfusion practice that may not focus solely around knowledge (or the lack of it).

It has been suggested that the clinical actions of healthcare professionals are influenced by the same factors that influence human behaviour in general. Assuming that this is correct, models of change can then be used to understand the behaviour of health professionals better and to inform the development of interventions to change behaviour [14–16]. Some models propose that the ways in which people think about and manage changes in practice, and not just their knowledge, will influence their behaviour. For example, one synthesis of evidence proposes 11 ‘theoretical domains’ or factors that influence behaviour change in practice (Figure 47.1).

Fig 47.1 Depiction of 11 theoretical domains that may explain behaviour change in clinical practice. Reproduced from Michie et al. [16].


These domains could be explored using the following questions:

1. Do I (or my staff) have the necessary skills (e.g. communication skills, technical skills) to implement the recommended change? (Skills)

2. Is the recommended change in practice consistent with my professional training? Do I see it as my professional responsibility? (Professional role and identity)

3. Am I confident that we can do this effectively? (Beliefs about capabilities)

4. Do we have the necessary resources (e.g. time, staff, equipment, space)? (Environmental context and resources)

5. What protocols, procedures, monitoring or prompts will be needed to ensure that our policies are implemented? (Behavioural regulation)

6. Does the proposed new practice create unpleasantness, discomfort or work stress? (Emotion)

7. What are the views of other important or relevant people? Are my senior colleagues in favour of a change in practice? What will be the views of staff coming on to the ward on the next shift? Do the patients' relatives have strong views? (Social influence)

8. Is the new practice something that I may forget to do? Am I clear which patients should receive care differently? (Memory, attention and decision processes)

9. Where does the new practice fall on my list of clinical priorities? How important is it? (Motivation and goals)

10. Is it good practice? What is my understanding of the evidence about effectiveness versus risk? How do I weigh up the pros and cons? What might be the consequences of the action (for the patient, for other patients in the unit, for me or for the clinical team)? (Beliefs about consequences).

A behavioural perspective on enhancing the uptake of evidence-based transfusion practice may not only help unravel the complexities of decisions to transfuse but lead to new strategies for influencing clinicians' use of blood. For example, different domains influencing behaviour change can be elaborated further in ways that could identify the major drivers of decisions about whether or not to transfuse in a specific situation. These drivers focus on how individual clinicians might weigh up the relative importance of various factors to do with transfusion. Attitudes arise not only from the perceived advantages and disadvantages of transfusing, but also on how individuals balance them up against one another. It might be understandable if immediate adverse events were weighed more heavily than delayed adverse events, even though the likelihood of the delayed event may be greater. Similarly, the influences of other people with different views (say, the patient's family) is likely to be greater when those people are present in the ward than when they are distant. Further, factors influencing the clinician's control over the behaviour may exert different amounts of influence. For example, time constraints will have a more powerful influence when other critically ill patients are in greater need of attention. Published research has identified consistent patterns of influences on transfusion behaviour, encompassing theoretical domains of knowledge, social influences, beliefs about capabilities, beliefs about consequences and behavioural regulation. Of note, behavioural regulation (including techniques such as goal setting and providing feedback) was identified as a key domain influencing transfusion practice [17,18].

Implementation research

There is increasing recognition that the findings from clinical research will not change population health outcomes unless healthcare systems and professionals adopt them in practice. However, a consistent finding from many groups and organizations is that this transfer of research findings into practice is unpredictable and in many cases slow and inconsistent. This gap between evidence and practice is a strategically important problem for policy makers, healthcare systems and research funders because it limits the health, social and economic impacts of clinical research. There is no reason to believe that this is any different for transfusion medicine and the problem of slow uptake will become more important as new primary research is published. Recognition of this quality gap in practice has led to much more interest in active quality improvement strategies over the last 10 years, and a body of implementation research has developed in many different healthcare areas [19–21]. As a generalization, this again demonstrates that interventions can be effective, although providing less information to guide the choice or to optimize the interventions in actual practice.

The starting position for implementation research is the recognition that identifying factors predictive of clinicians' behaviour that are amenable to change may guide the design and choice of interventions with the highest chance of success. However, our understanding of potential barriers and enablers to quality improvement is largely limited in transfusion and hindered by a lack of a ‘basic science’ relating to determinants of professional and organizational behaviour and potential targets for intervention. A systematic investigation of the beliefs associated with inappropriate transfusion practice and barriers to change could therefore point to potentially effective ways to optimize practice.

As an example, audit and feedback is widely used and embedded within the NHS, providing an existing vehicle for quality improvement that can be optimized for transfusion. Systematic reviews indicate small to medium effects of 6–16% on clinical practice and patient outcomes, although effects are often variable [19,20]. Clearly, there is potential to improve the practice of audit and feedback. However, there is little rigorous evidence on how to optimize its content or delivery, target key health professionals and parts of the organization necessary for conducting it or reliably maximize its effectiveness. Research is beginning to provide data on how to enhance its effects and understand the key mechanisms for its effectiveness. Feedback can be enhanced by reviewing the written content to deliver direct responses better or by targeting its delivery to the relevant staff with discussion and agreement of action plans. Enhanced delivery might include practical guidance for clinical teams on how best to implement the process of delivering feedback, including the development of materials for clinical teams to facilitate discussion and agreement of contextually appropriate goals and action plans.


Many healthcare professionals are becoming more aware of the need to practise evidence-based transfusion medicine. However, there remains considerable variation in transfusion practice in spite of explicit criteria and guidelines. Some surgical teams carry out major procedures without blood transfusion by attention to patient care throughout the perioperative period. For example, a combination of educational support for algorithms for blood management and restrictive transfusion thresholds may offer a more effective approach to blood conservation than the implementation of more complex (or costly) single interventions, but better evidence on the relative effectiveness of these (or other) strategies is required. While interventions to change transfusion practice can be aimed at a number of levels (individual health care professionals, healthcare groups or teams, organizations providing healthcare, the larger healthcare system or environment), the majority of interventions have been aimed at individual practitioners, as ultimately it is the individual clinician who dictates much of the decision-making around patient care. More resources need to be devoted towards a better understanding of the promotion of clinical effectiveness and the uptake of clinically effective transfusion practice (evidence-based implementation). Evidence of ‘what to do’ is one thing, but learning ‘how to implement’ is another.

Key points

1. The evidence base for transfusion practice is poorly developed but is improving. As it increases, changes in transfusion behaviour will need to be enacted more frequently.

2. Transfusion practice is also characterized by variation in spite of explicit criteria and guidelines.

3. The decision to transfuse is a complex process and knowledge is unlikely to be the only influence; research into these factors is generally sparse.

4. A better understanding of the determinants of transfusion behaviour should guide the design and choice of interventions better in order to deliver changes and improvements in transfusion practice. These need to be evaluated in rigorous clinical studies.


1. Rothschild JM, McGurk S, Honour M et al. Assessment of education and computerised decision support interventions for improving transfusion practice. Transfusion 2007; 47: 228–239.

2. Salem-Schatz SR, Avorn J & Soumerai SB. Influence of clinical knowledge, organisational context and practice style on transfusion decision making: implications for practice change strategies. J Am Med Assoc 1990; 264: 476–483.

3. Soumerai SB, Salem-Schatz SR, Avorn J et al. A controlled trial of education outreach to improve blood transfusion practice. J Am Med Assoc 1993; 270: 961–966.

4. Tinmouth A, MacDougall L, Fergusson D et al. Reducing the amount of blood transfused. Arch Intern Med 2005; 165: 845–852.

5. Wilson K, MacDougall L, Fergusson D, Graham L, Tinmouth A & Hebert, PC. The effectiveness of interventions to reduce physician's levels of inappropriate transfusion: what can be learned from a systematic review of the literature. Transfusion 2002; 42(9): 1224– 1229.

6. Sanguis Study Group. Use of blood products for elective surgery in 43 European hospitals. Transfus Med 1994; 4: 251–268.

7. Stanworth S & Tinmouth A. Plasma transfusion and use of albumin. In: TL Simon, EL Snyder, CP Stowell, RG Strauss, BG Solheim & M Petrides (eds), Rossi's Textbook of Transfusion Medicine, 4th edn. Wiley-Blackwell andAABB Press; 2012 (in press).

8. Carless PA, Henry DA, Carson JL, Hebert PP, McClelland B & Ker K. Transfusion thresholds and other strategies for guiding allogeneic red blood cell transfusion. Cochrane Database Syst Rev 2010, Issue 10, Article No.: CD002042. DOI: 10.1002/14651858.CD002042.pub2.

9. Lacroix J, Hébert PC, Hutchison JS, Hume HA, Tucci M, Ducruet T, Gauvin F, Collet JP, Toledano BJ, Robillard P, Joffe A, Biarent D, Meert K & Peters MJ; TRIPICU Investigators; Canadian Critical Care Trials Group; Pediatric Acute Lung Injury and Sepsis Investigators Network. Transfusion strategies for patients in pediatric intensive care units N Engl J Med 2007, 19 April; 356(16):1609–1619.

10. Kirplani H, Whyte RK & Anderson C. The premature infants in need of transfusion (PINT) study: a randomised controlled trial of a restrictive (low) versus liberal (high) transfusion threshold for extremely low birthweight infants. J Pediatr 2006; 149: 301–307.

11. Carson JL, Terrin ML, Noveck H, Sanders DW, Chaitman BR, Rhoads GG, Nemo G, Dragert K, Beaupre L, Hildebrand K, Macaulay W, Lewis C, Cook DR, Dobbin G, Zakriya KJ, Apple FS, Horney RA & Magaziner J; FOCUS Investigators. Liberal or restrictive transfusion in high-risk patients after hip surgery. N Engl J Med 2011, 29 December; 365(26): 2453–2462. Epub 14 December 2011.

12. Stanworth SJ, Grant-Casey J, Lowe D, Laffan M, New H, Murphy MF & Allard S. The use of fresh frozen plasma in England: high levels of inappropriate use in adults and children. Transfusion 2010; 51(1): 62–70.

13. Estcourt L. National comparative audit of platelet transfusions, 2010. Key findings of the audit with regard to the inappropriate use of platelet transfusions.

14. Carver CS & Scheier MF. Control theory: a useful conceptual framework for personality-social, clinical and health psychology. Psycholog Bull 1998; 92, 111e135.

15. Ajzen I. The theory of planned behaviour. Organ Behav Hum Decis Process 1991; 50: 179–211.

16. Michie S, Johnson A, Abraham C, Lawton R, Parker D & Walker A. Making psychological theory useful for implementing evidence based practice: a consensus approach. Qual Safety Health Care 2005; 14: 26–33.

17. Francis JJ, Stockton C, Eccles MP, Johnston M, Cuthbertson BH, Grimshaw JM, Hyde C, Tinmouth A & Stanworth SJ. Evidence-based selection of theories for designing behaviour change interventions: using methods based on theoretical construct domains to understand clinicians' blood transfusion behaviour. Br J Health Psychol 2009; 14: 625–646.

18. Francis JJ, Tinmouth A, Stanworth SJ, Grimshaw JM, Johnston M, Hyde C, Stockton C, Brehaut JC, Fergusson D & Eccles MP. Using theories of behaviour to understand transfusion prescribing in three clinical contexts in two countries: Development work for an implementation trial. Implement Sci 2009, 24 October; 4: 70. PMID: 19852832.

19. Grimshaw JM, Thomas RE, MacLennan G, Fraser C, Ramsay CR, Vale L, Whitty P, Eccles MP, Matowe L, Shirran L et al. Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technol Assess 2004, Report, p. 8.

20. Jamtvedt G, Young JM, Kristoffersen DT, O'Brien MA & Oxman AD. Audit and feedback: effects on professional practice and health care outcomes. Cochrane Database Syst Rev 2006; Issue 2, Article No.: CD000259. DOI: 10.1002/14651858.CD000259.pub2.

21. Craig P, Dieppe P, Macintyre S, Michie S, Nazareth I & Petticrew M. Developing and evaluating complex interventions: the new Medical Research Council guidance. Br Med J 2008; 337: a1655.

Further reading

Cooksey R. A Review of UK Health Research Funding. Norwich: HMSO; 2006.

Effective Health Care. Bulletin on the effectiveness of health service interventions for decision makers. Centre for Reviews and Dissemination, The Royal Society of Medicine. Available at:

Eisenstaedt RS. Modifying physicians' transfusion practice. Transfus Med Rev 1997; 11: 27–37.

Glasziou P, Chalmers I, Altman D, Bastian H, Boutron I, Brice A, Jamtvedt G, Farmer A, Ghersi D, Groves T et al. Taking healthcare interventions from trial to practice. Br Med J 2010; 341: c3852.

Grol R. Implementation of changes in practice. In: R Grol, M Wensing & M Eccles (eds), Improving Patient Care: Implementing Change in Clinical Practice. Oxford: Elsevier; 2004.