Evaluation of clinical appropriateness of blood transfusion

Introduction

The appropriateness of health procedures has attracted considerable attention in recent years, particularly in the light of the need to control the growth of healthcare costs.

Clinical appropriateness can be defined as the use of interventions:

• - with documented clinical effectiveness

• - for conditions in which the effectiveness has been documented

• - with expected benefits exceeding the expected negative consequences.

The need to give the 'right' care, to the 'right' patient, in the 'right' way, should be kept clearly in mind independently of the awareness that health costs cannot be indefinitely expanded. In fact, economic considerations pertain to a different factor underlying clinical governance, the organizational appropriateness, which means applying clinically appropriate interventions on the condition that they use a reasonable amount of resources and providing that the consumption of these resources is as efficient as possible.

In Italy the need to assess the appropriateness of blood transfusion practice has been established by regulatory requirements and by good practice standards¹. Furthermore, the growing need to deliver sufficient and safe blood components of high quality make careful control of their appropriate use even more necessary.

Instruments needed to evaluate appropriateness

Two instruments are needed in order to evaluate the appropriateness of a health intervention:

1. a reference standard (the set of conditions, timing and settings for which an intervention has shown to be effective);

2. information about why and how the intervention has been requested/performed to verify its conformity in comparison to the reference standard.

The guidelines

The adoption of a reference standard is a requirement for every consideration regarding clinical appropriateness: clinical guidelines represent this standard. Guidelines developed by the GOBSAT method ('Good Old Boys Sitting Around a Table')², a term referring to an unstructured manner by which a group of self-selected experts discuss and write down their opinions, should no longer be valid or encouraged. Unfortunately, as argued by Reverberi³, not rarely 'panellists do not refrain from expressing their personal opinions assertively'.

Clinical guidelines should be evidence-based, 'systematically developed statements to assist in making decisions about care for specific clinical circumstances'⁴. Three attributes are needed for a guideline to be considered 'evidence-based'⁵: (i) the multidisciplinarity of the group assembled to produce it; (ii) the systematic evaluation of the best available scientific evidence to justify adopted statements; and (iii) the classification of the quality of evidence and of the strength of recommendations must be reported.

With regard to evaluation of appropriateness, there are no options other than adopting a reference standard. If the chosen model of care is 'practical evidence', that is, clinical decisions derived from personal intuition and judgment (which means, in fact, that every patient is managed according to each doctor's own experience) every intervention is, substantially, appropriate: in such a scenario the evaluation of appropriateness is, in fact, impossible.

Numerous clinical guidelines refer to the correct use of blood components^6–12. In 2008 a working group by the Italian Society of Transfusion Medicine and Immunohaematology (SIMTI) produced recommendations regarding the clinical use of blood components and plasma derivatives¹³. To be considered true guidelines, these recommendations need to be agreed with the scientific societies of other disciplines interested in transfusion therapy. Chapter D of the Standards of Transfusion Medicine, edited in 2007 by SIMTI¹⁴, recommends that guidelines are locally adopted by Hospital Transfusion Committees: some authors suggest that local ownership is a critical factor in encouraging the adoption of guidelines by clinicians¹⁵, although this is not confirmed by others¹⁶. Caution should be taken to avoid that local adaptations of national guidelines result in excessive variability in the definition of what is appropriate.

The blood request form

Suitable documentation is required to assess the justification for a transfusion within established clinical guidelines. Unfortunately, the clinical circumstances necessitating transfusion are often poorly documented on the blood request form and the patient's record, which is a major limitation of perspective and retrospective audits, the most common type of evaluation of appropriateness performed by transfusion services^17–19. Of course, if the reason for transfusion is not documented, no evaluation of appropriateness is possible, and the case can only be classified as 'non-complying request, unassessable appropriateness'. The judgment of the adequacy of request form compilation is not always uniform: in an analysis of 6,036 requests evaluated between January and April 2009 at Ravenna Blood Centre by ten Blood Centre physicians, the frequency of 'incomplete documentation' ranged between 0.8 and 24.6% (unpublished data). In an attempt to reduce this variability, a check-list for the evaluation of request conformity has recently been introduced. It sets out three questions to be answered: (i) from the information on the request form, is it possible to assign the transfusion justification to one of the conditions stated in the adopted guideline?; (ii) from the information on the request form, is it possible to evaluate the amount of blood components needed to correct that condition?; and (iii) is the request appropriate with respect to indication and to amount?

Questions 1 and 2 define the request conformity: only affirmative answers to both allow a subsequent judgment about appropriateness, while at least one negative answer implies that the request is considered not conforming and the appropriateness unassessable. An affirmative answer to all the questions allows the resquest to be judged appropriate. Not enough requests have been evaluated yet to establish the effectiveness of such an approach.

Blood request forms designed to incorporate the indications for transfusion could help to lower the rate of unassessable cases, as well as to reduce inappropriate use of blood products²⁰, although this has not been confirmed in the Ravenna experience. The problem could be definitively solved by the use of an electronic form (transfusion order-entry), including mandatory fields to be completed with choices from a structured list before the request is sent.

Application of clinical guidelines

Although clinical guidelines have been widely spread over the last years, this does not necessarily mean that the recommendations described are followed. Despite the efforts to improve the quality of patient care, research suggests that 30–40% of patients do not receive treatments whose effectiveness has been shown and that 20–25% of patients receive treatments that are unnecessary or potentially harmful^21,22.

International experience has shown that passive dissemination of recommendations and guidelines is ineffective, no matter how important the issue or how valid the methods are. Many different types of intervention have been described to 'close the gap between knowing and doing'^23–25.

Distribution of printed educational material is widely used but, according to a Cochrane analysis, appears to have limited beneficial effects on professional practice²⁶.

Educational meetings, conferences or workshops designed for healthcare providers, showed small, if any, efficacy on improving guideline application²⁷.

Local consensus processes, involvement of people in the decision-making process about issues that will affect them was of uncertain efficacy²⁸.

Educational outreach visits, use of trained persons who meet clinicians in their practice settings to give information intended to change their practice, showed consistent but small effects²⁹.

Local opinion leaders, individuals perceived by their colleagues as 'educationally influential', successfully reduce non-compliance with clinical guidelines, with an effectiveness comparable to other strategies³⁰: uncertainty remains about how to identify such leaders, and about the best methods they should apply to obtain desired results.

Audit and feedback is the most widely used strategy to improve the application of guidelines. Retrospective audits are referred to periodical review, summarized and presented to a group of individuals with the purpose that they modify their practice, if inconsistent with accepted guidelines.

In transfusion medicine, prospective auditing, that is the evaluation of requests performed before delivering blood components, has been considered the most effective, although the most labour-intensive intervention, because it takes place before the transfusion occurs, preventing inappropriateness rather than merely registering it. The effectiveness of audit and feedback, whether retrospective or prospective, has been shown to be small to moderate^31,32.

Benchmark studies could be useful for identifying variations in transfusion practice in specific diseases or surgical procedures¹: an interesting study, in which transfusion data from blood bank computer systems were linked with patients' data from hospital inpatients' databases based on the International Classification of Disease system, has recently been published³³.

Reminders, information for clinicians intended to recall the content of accepted guidelines, are usually associated, in transfusion medicine, with a new transfusion request form or computer order entry system. They generally achieve small to modest improvements in providers' behaviours³⁴.

On the whole, it could be stated that, with respect to the simple implementation of guidelines, interventions to enhance their application offer some advantages, although the observed advantages are modest to moderate, no single intervention has been shown to work better than the others, and the use of combined interventions has not shown clear advantages with respect to a single intervention. Studies performed in the field of transfusion medicine have shown analogous results. Furthermore, the effectiveness of interventions over time appears limited: a return to baseline conditions has been observed, both following completed interventions and during ongoing ones³².

Conclusions

Evidence-based medicine (EBM) is seen as a means to improve the scientific quality of clinical practice, thus ameliorating health care; there are, however, criticisms of the claims deriving from EBM³⁵. The main criticism, known as the 'EBM paradox', is that, as Sackett and colleagues themselves acknowledged, assessed by the rules of EBM itself, there is no evidence that EBM is a more effective means of pursuing health than medicine-as-usual. A second main criticism is that EBM denigrates clinical expertise, ignores patients' values and promotes "cookbook medicine"³⁶.

Many criticisms are founded on the lack of relevant evidence regarding the management of common clinical situations. There seems to be a misperception based on the idea, not rarely insufficiently opposed by EBM advocates, that only randomised trials or systematic reviews constitute 'evidence'. This conception leads to a binary, 'all or nothing', use of evidence, by which in the absence of strictly concordant, good quality studies, only each doctor's own experience and reasoning can be used to manage the 'grey zones' of practice. This contrasts with two concepts stressed by EBM: the concept of hierarchy of evidence and that of the need to deal with the best available evidence, when the best possible one is not available.

Prospective approval of blood requests is considered by many the best way to ensure transfusion appropriateness and is indicated as mandatory by SIMTI's Standards of Transfusion Medicine¹⁴. However, this represents a labour-intensive and often conflict-creating practice for transfusion medicine physicians, which has not yet demonstrated clear effectiveness in ameliorating clinical practice.

No evaluation of clinical appropriateness of blood transfusions can be made without convincing clinicians that there are rules, albeit imperfect, which establish what is, and what is not, appropriate. In the absence of this shared awareness the pursuit of appropriateness risks being a theoretical, although time-consuming, exercise. In the absence of efforts to convince clinicians about the need to follow evidence, transfusion medicine physicians will not be able to be, as wished by Grazzini¹, the 'gatekeepers of clinical appropriateness of transfusion practices'.