Differences Among International Pharyngitis Guidelines: Not Just Academic

The individual is inherently unpredictable. The mere fact that Epidemiology exists shows exactly this. Guidelines are built on the basis of population data and thus 'construct' a patient that is an average of a whole population or a subgroup thereof.

On the contrary the medical practitioner in her daily work relies on a different cognitive approach for problem solving. Confronted with an individual she relies on her immediate interaction with the patient where medical knowledge becomes inherent in the formation of the result of consultation.

The practitioner might feel that these two approaches are at odds in her day to day work. The contemplation of and adhering to formalized guidelines could be distracting in her building of a 'gestalt' in the form of diagnosis and therapy.

In my thinking there is no easy solution to that problem. We might come to fertile grounds if the two sides of the divide (guideline authors and medical practitioners) start to educate themselves about the way of thinking and reasoning of 'the other'. Then we might be able to use information that is beneficial and leave aside the misunderstandings.

Competing interests:   None declared

The excellent article of Matthys et al made it clear that the simple equation “best medical evidence = clinical practice recommendation” does not work in real life. It is only natural that the interpretation of evidence and the process of finding conclusions based on this interpretation depends a lot from the medico-cultural background in which the guideline was developed and is supposed to be applied. In my opinion this phenomenon is whether bad nor good but consequences should be drawn: I think the process of guideline development has to become much more transparent in the way that guidelines should provide information why a certain recommendation has been made and why this recommendation might differ from those of other guidelines. I am not as pessimistic as Del Mar (depending on my cultural background)on what most doctors expect from guidelines. A guideline may address different (evidence based) options within the context of its medico-cultural setting. Maybe doctors would accept guidelines that are not just giving recommendations but would more describe a “corridor” in which different approaches could peacefully co- exist, as long as evidence does support them. I am not sure if we really all have to agree on the same values, maybe a first step towards better guidelines could be to explicitly name the values that lie underneath their process of development.

Competing interests:   None declared

To begin, we would like to thank very much the authors of this article.

One can question the fact that, for a same disease, there is so many different guidelines (nearly as many as developed country) about the same subject. This could partly be explained by the bacteriological distribution. Larger antibiotic prescription could be explained by health system of course, and justified by streptococus strains which are more likely to generate acute rheumatic fever. The authors pointed at the main weaknesses of the French guidelines, and not only about sore throat. Most of the guidelines intend to be used in general practice. Whereas, they are all written by non GPs : ENT specialist, infectious disease physicians, microbiologists, pharmacists…mainly working in hospital with different recruitment. These physicians who treat the very seldom but serious complications of the sore throat speak about what they meet. It is not the same clinical expression of a disease for general practice.

The French guideline described in the article dates from 1999. In 2005, a new guideline (1) has been built which is not so different from the former. French physicians prescribed about 5 time more antibiotics than many other west European countries. This statement led in 2002, to a state advertising campaign held by the health authorities :“antibiotic is not automatic”. In the meantime, they suggested to the physicians the use of a diagnostic test. Since then, 50% of the GPs use this test, antibiotic prescriptions decreased of 17%, and 25% for the infants. A lot of informations still need to be given either to the people than to the physicians, not only GPs. EBM is a difficult art to practice. French guidelines and especially the one about sore throat is not really helpful.

1°) Agence Française de Sécurité Sanitaire des Produits de Santé ANTIBIOTHERAPIE PAR VOIE GENERALE EN PRATIQUE COURANTE DANS LES INFECTIONS RESPIRATOIRES HAUTESDE LADULTE ET LENFANTARGUMENTAIRE. 2005 ; http://recherche.sante.gouv.fr/search97cgi/s97_cgi

Competing interests:   None declared

It was with great interest I read the article by Matthys and colleagues about the differences between international pharyngitis guidelines. I appreciate their referencing our paper, “Evaluation and management of pharyngitis in primary care practice: the difference between guidelines is largely academic,” in their title. The main findings and conclusions of Matthys and colleagues, that there are surprising differences between international guidelines for pharyngitis despite easy availability of evidence, are important.

However, all of these differences are irrelevant if there is a large gap between what the guidelines say and actual practice. We found that the major problem in managing pharyngitis was not which guideline clinicians followed, but that clinicians routinely failed to follow any guideline. Early drafts of our article referred to the Belgian, Scottish, and Dutch guidelines. Our study was on primary care practice in the United States and focused on the differences between 2 American guidelines (US09 and US10 as labeled by Matthys and colleagues). Our findings would be all the more striking if held up against European guidelines that recommend avoiding testing and antibiotic treatment for sore throat. In our study, clinicians performed a rapid streptococcal test in 80% of visits and prescribed antibiotics to 47% of patients.

Matthys and colleagues write, “Differences among guidelines are not merely academic; they have important consequences for daily practice.” Maybe. But if clinicians fail to heed guidelines, all the well-founded guidelines in the world – even if they all agreed – will not affect clinical practice and are only an academic exercise.

Competing interests:   None declared

I read with much interest this well articulated international guideline comparison by Matthys and colleagues. To me the value of such research is that it objectively describes an array of clinical guidelines with reputedly similar aims and so may stimulate further curiosity about cross and sub-cultural management differences, and is not simply a look at "evidence variation". As a pragmatic family physician and researcher in the UK I am somewhat sceptical about the notion of homogeneous (or universal) guidelines that aim to represent a common external reality. Given the complexity and range of factors on the ground this goal seems to me presently "a bridge too far" for the evidence. But the attempt appears to be more than worth taking precisely because of the sort of variations illustrated. For example I am very surprised that excellent North Atlantic clinical scoring systems developed by specialists aren't used more widely by family physicians in Europe, and also that antibiotic resistance doesn't yet appear a big issue to all prescribers!(1) Notional differences in illness models tacitly used by differing groups of clinicians developing guidelines would support the author's conclusions and earlier comments made. 1. BMJ. Special Edition: Antimicrobial resistance.1998. 5th Sep, No 7159.

Competing interests:   None declared

In their elegant international comparison Matthys et al.[1] show that the main therapeutic advice in guidelines on sore throat varies 100% (between always and almost never antibiotics). Moreover tests for streptococci - as a proxy for possible serious complications that might be averted by antibiotics - seem to be used in an inconsistent way. They conclude that there is no evidence for the appropriateness of this regional variation.

We agree with that, but are less optimistic about their proposed cure: rigorous adherence to ‘guidelines’ for the development of guidelines. These ‘guidelines’ have become so extensive (amply illustrated in Matthys et al’s references) that this ‘medicine’ tends to throw away the baby with the bathwater.

As long as major parts of the central decision tree behind guidelines cannot be fully supported by unambiguous evidence guidelines will, in our opinion, always show variations. The reasons for this are the many subjective elements of human judgement[2], based on norms, values and cultural differences. There is nothing wrong with that as long as guideline developers clearly show in their document that all available evidence was rigorously searched, graded and used and that all subjective arguments used for branches (steps) that cannot be fully supported by evidence are made explicit.

In this way the remaining areas of uncertainty in guidelines are clearly visible and open to be bridged by supraregional consensus or, even better, by supplementary research.

1. Matthys J, De Meyere M, van Driel M, De Sutter A. Differences Among International Pharyngitis Guidelines: Not Just Academic. Ann Fam Med. 2007;436-443.

2. Thomas S. Standard setting in the Netherlands: impact of the human factor on guideline development [editorial]. Br J Gen Pract. 1994;44: 242-3.

Competing interests:   None declared

You have raised an important issue. Most guidelines have been validated for their feasibility in clinical practice. However, this does not mean that they will be implemented by physicians. If doctors are supported and regularly reminded they may change their prescribing habit, but more pragmatic trials in family medicine settings are less optimistic. Recently, we have demonstrated that discussing a new guideline on antibiotics for rhinosinusitis in small peer groups does not lead to significant changes in antibiotic prescribing (Qual Saf Health Care 2007;16:197-202). And of course, the key question is if following guidelines leads to better patient outcomes. To our knowledge this has not been tested in family medicine. It is an important area for research.

Competing interests:   None declared

This wonderful paper points a laughing finger at the emperor with no clothes. The range of different ‘recommendations’ arising from the different guidelines is startling. How can different attempts to get to the same point result in such different products? One might forgive different interpretations of the evidence, but it is hard to justify the selection of different sources of evidence. Is this the dreary cloak of parochialism? Perhaps it is a reflection on the fact that too many people still do not appreciate that the essence of evidence-based practice is empiricism. If we are burdened with too much patho-physiology (understanding of the mechanism of disease) we are likely to allow our prejudices to warp our objectivity.

But more than that, perhaps guidelines are worthless unless we adhere to the standard definition (usually paying tribute to“…systematised summary of the evidence to assist clinicians make clinical decisions…”). Implicit in such definitions is a sense that the doctor has to take responsibility for deciding what to do with the evidence. This means undertaking (at least) two jobs: deciding on the worth of the evidence (critical appraisal); and applying it to the individual patient. Unfortunately this may not be possible for quite easily anticipated reasons. For one, a study that represents good evidence for one question may be hopelessly biased for another. Good data may provide one decision for one patient, but quite different for another.

But, sadly, what most guideline-hungry doctors want is something else. “Just tell me what to do!” Effort-free solutions to information needs are probably impossible. In addition, abdicating these responsibilities to others may even be dangerous.

What we need is help in the difficult steps of evidence-based medicine, not just a pre-digested sop to eat in its stead. If guidelines can provide summarized up-to-date research in an evidence-based manner, and allow individualized decisions for the individuals we have to manage, using a standard easily-searched format, this would be ideal. We cannot use the evidence without risking doing so intelligently without effort.

Competing interests:   None declared

When most physicians consider guidelines, they envision standards for managing medical problems. The term – guidelines - gives the impression of authority. Those who write guidelines seem to write as if they represent authority. Researchers castigate physicians for not following guidelines. Payers reward physicians who follow guidelines.

But guidelines are not authoritative rather they are guides developed to inform care not to direct care.

This article on pharyngitis guidelines clearly provides a description of the problems of guidelines. Adult pharyngitis seems, at first consideration, a straight forward and simple clinical complaint. Yet, for several reasons, Dr. Matthys and colleagues found minimal agreement amongst 10 separate guidelines for diagnosing and treating adult sore throats.

This article should invalidate the myth that guidelines are merely an extension of evidence based medicine. First, the 10 guideline groups did not agree on which articles to include in developing their guideline. More important, these guidelines demonstrate that all guideline committees start with implicit biases.

Dr. Allan Detsky wrote an important article titled – “Sources of bias for authors of clinical practice guidelines” (CMAJ. 2006 Oct 24;175(9):1033, 1035.) In that article he enumerates many biases that members of guideline panels bring to their meetings.

The truth about guidelines is that they are molded though the value structure of the panel members. My colleagues and I recently published a discussion of competing US pharyngitis guidelines – “Pharyngitis Management: Defining the Controversy” (J Gen Intern Med. 2007 Jan;22(1):127-30). We concluded our analysis by stressing that the two guideline panels that we compared seemingly focused on different issues. The panel identified by Matthys and colleagues as US10 focused on symptom relief, while the panel identified as US09 seemed to focus on avoiding unnecessary antibiotic usage.

In considering antibiotic therapy for pharyngitis, one must weigh the benefits and risks of antibiotics. As I read the literature I see several benefits for treating group A streptococcal pharyngitis. First, prompt antibiotic treatment does decrease symptom duration by 1 – 2 days. Second, antibiotics likely decrease the probability of developing acute rheumatic fever. Third, antibiotics likely decrease the probability of suppurative complications. Fourth, antibiotics decrease person to person contagion. Antibiotics might also help patients with non-group A (mostly C and G) streptococcal pharyngitis. They also may prevent complications of Fusobacterium necrophorum associated pharyngitis.

The major risk of treatment is allergic reactions to the antibiotics. Many postulate that antibiotics for upper respiratory tract infections contribute to antibiotic resistance.

So this seemingly simple problem actually has multiple dimensions. As Dr. Matthys and colleagues nicely demonstrate each guideline group seems to have weighed the dimensions differently. Each reader could probably pick a guideline that fits their value structure. But the lack of consensus of these guidelines should make us strongly reconsider the guideline concept. As the authors appropriately state – “National guidelines on acute sore throat promote different clinical approaches, recommend different treatments, and cite different evidence. There is no evidence that regional variation is appropriate. Introduction of an explicit guideline development method for both European and North American guidelines may lead to more uniformity in the diagnosis and management of acute sore throat.”

Unless we can agree on values, we cannot achieve consensus on guidelines. This limitation hampers our ability to have standardized guidelines. This problem should make us rethink how guidelines are developed and implemented.

Competing interests:   None declared