The guidelines controversy: what do the experts think?

Following the recent controversy about Australian and NZ guidelines for preventing blood clots, Croakey asked some relevant parties for their views on:

• whether  health departments, hospitals,  safety and quality groups be reviewing their support for these particular  guidelines in response to the concerns raised in the MJA and elsewhere?

• whether such agencies also review their approach to endorsing  and disseminating guidelines generally?

• whether health and medical  groups producing guidelines should accept funding from interests with a commercial  stake in the guidelines’ recommendations?

Here are some of the responses (some of which are outdated by today’s NHMRC release of its own draft guidelines for public comment):

Professor Chris Baggoley Chief Executive, Australian Commission on Safety and Quality in Healthcare

1. It is important throughout this discussion to remember that concern about Venous Thrombo Embolism (VTE) is based on a very real clinical problem. The NHMRC: National Institute of Clinical Studies (NICS), whose role it is to help put evidence into practice, not only addressed VTE as a clinical priority area in its initial work program, it listed and published on the problem of VTE in its 2003 “Evidence-Practice gaps report Vol 1”. It provided a figure of 2000 deaths per year from pulmonary embolus annually. It also stated that PE causes or contributes to 10% of all deaths in hospital.

NICS followed this up with the 2008 “A review of developments 2004-2007” from the Vol 1 EPG Report and stated “While variations in practice are still occurring, there is reason to be optimistic that progress is being made in closing this evidence practice gap.”

2. NHMRC has not produced a VTE Guideline to date, although this is in process now and presumably will be released for consultation in the next 12 months. Alasdair Miller notes that NICS has selected two members of the Working Party to be on its own VTE guideline development working group, presumably for their expertise. He does point out that the ANZ Working Party on the Management and Prevention of VTE “comprises 14 eminent and appropriately qualified haematologists, physicians and surgeons” so it is not surprising that some would be selected to be involved in the production of the first NHMRC VTE Guideline.

3. In the absence of a local Guideline, what should be the response of those clinical leaders whose ethical and professional role it is to lead improved prevention and management  in VTE? They could do nothing, they could bring together evidence that is already published or they could produce their own evidence based guideline, using an approved NHMRC process. The worst outcome would seem to be the first alternative, yet paradoxically in doing nothing attention would not be drawn to any individual or group of individuals. Undertaking either of the latter two options could be applauded; the last option is very time and resource consuming.

4. As I understand it, the Working Party aims to produce a summary of evidence based published international guidelines and has not intended to produce a new set of guidelines.

5. Given all of the above, I will not be asking the Commission to take any specific action as a result of this article; we will remain focussed on the availablity of best guidance for clinicians in this important clinical condition. When the NHMRC Guidelines are produced, we will be encouraging these to be put in a format for easy use by clinicians, for these to be widely distributed to them and for them to take priority.

PS I did speak at the Action on VTE summit held in Sydney last year, as prevention of VTE is one of the elements of the Medication Safety Program of the Commission (for no remuneration!). We have funded NICS to role out its  successful public hospital VTE prevention program to the private sector, in 38 hospitals and the program, under the guidance of NICS, is going well.

***
Dr Jon Jureidini, Healthy Skepticism

Health departments, hospitals, safety and quality groups should be reviewing their support for these particular guidelines in response to the concerns raised in the MJA and elsewhere.  Ideally guidelines should be developed by non-conflicted panels who have access to raw data. This is a huge undertaking, but international co-operation would make it feasible, eg, Australia might take responsibility for some topics, knowing that the UK will cover others.

Health and medical groups producing guidelines should not accept funding from interests with a commercial stake in the guidelines’ recommendations. For obvious reasons. But this doesn’t solve the problem, unless those creating the guidelines have access to all the data. Otherwise, as David Healy points out, you end with biased recommendations even from non-conflicted committees, because what they rely on is biased (ie published data).

**

Dr Alasdair Millar, physician and author of current MJA article critiquing the guidelines

Health departments, hospitals, safety and quality groups should immediately reconsider their acceptance of the guidelines in respect of medical patients.  I have a paper in press in the Internal Medicine Journal (said to be published next month) which demonstrates that the risk of major bleeding with prophylactic low molecular weight anticoagulants is as high or perhaps higher than the reduction in the risk of symptomatic thrombosis, in medical patients.  This is because the risk of symptomatic thromboembolic disease in medical patients is so low.

The clinical trials used diagnostic techniques which revealed ALL thrombosis, including asymptomatic disease, and that is the source of the incidence figure for 17% in the Guidelines.  So the guidelines for medical patients are based on a measure of thrombosis which is not clinically relevant (this argument is in my earlier article).

It is interesting that Prof Fletcher in his rebuttal refers to the ACCP guidelines, hinting that anyone who doesn’t agree with them is automatically wrong.  However, even those guidelines do not take this asymptomatic/symptomatic question fully into account.

So in my view only a very small group of medical patients (those with cancer or a previous history of DVT or PE) should be considered for thromboprophylaxis, both to increase the prophylactic yield and to avoid causing a bleed in patients who wouldn’t have a clot in the first place.

Regarding whether agencies should review their guidelines processes generally, this is a much broader question than you think.  I tried to confront it in my paper.  Medical practice has become subject to the influence of a Quality Assurance industry that means well but whose approach is to ensure that for each activity there is a guideline, that there should be a process for adherence to these guidelines, that this process should be audited, and that a satisfactory audit means that patients are being given quality treatment.

Broadly speaking, this industry is satisfied when all the boxes are ticked and they can say “Ah, we can now report 90% compliance with this guideline” (orgasmic pleasure at 100%!).  This approach is unscientific, relies entirely on the guidelines being good in the first place, and can (and I believe, does) place patients at risk.

This argument can be developed further if you wish.  The thromboprophylaxis guidelines are a good example of how it happens in practice.  It is a very fundamental human behavioural trait to want to be seen to be doing something good, even when the appearance is illusory, and that is what is happening in the QA/clinical safety industry.  When you combine the approach of this industry with the marketing prowess of a global drug manufacturer, you can see that the dice is somewhat loaded.

Guidelines groups should not accept commercial funding, this is now widely considered to be unacceptable.  I suggest you read the JAMA article cited in Martyn van der Weyden’s editorial.

One interesting aspect of the Working Party Guidelines is that they could have been produced at no cost, and therefore with zero need for sponsorship.  E-mail is free and could have been used by the Working Group, as is publication in a medical journal.  There is no need to travel to Sydney for a meeting in an hotel, as Prof Fletcher described in his rebuttal.  Had the low-cost route been taken, none of the questions about the possible role of the drug company would have arisen.  That’s the point of avoiding the sponsorship.

***
Professor Ian Olver, Cancer Council Australia

Guidelines are best written by experts, but the ideal is that those experts cannot have a potential conflict of interest. Although evidenced based, it is the interpretation of evidence and the weighting of the clinical importance of various outcomes over a number of trials which allows opinion to form part of the recommendations.

Given that guidelines are designed to have widespread influence on clinical practice, simply having authors being transparent about the their relationships with industry is not enough, and we should insist that guideline authors have no potential conflicts to declare. Agencies should insist on this and insist that all relevant literature has been reviewed. Agencies such as the NH+MRC set independent standards for guidelines, for example which involve a clear audit trail of the papers upon which the guidelines are based and also insist on widespread stakeholder consultation  before guidelines are endorsed.

I don’t believe that guidelines are compromised if health and medical groups producing them receive funding from interests with a commercial stake in the recommendations.

***

Pharmacologist Professor Ric Day, University of NSW

Critical appraisal of any guidelines includes their source, motives for genesis, potential conflicts of interest.  These should always be considered. Then the content of the guidelines needs to be assessed.  Into the mix comes the ‘need’ for the guidelines and ‘alternatives’ i.e. pragmatic review of risks/benefits of adopting these particular guidelines under consideration.  The more unencumbered guidelines are by potential conflicts of interest and well resourced with best practice processes used in their production the more comfort there will be in accepting them. And the calibre of the individual contributors is of significance too.

Agencies should areview their approach to endorsing and disseminating guidelines generally. It is always good to be very critical of any guidelines as to their quality, processes etc including in particular possible conflicts of interest impinging on them in any way.

Health and medical groups producing guidelines should optimally not accept funding from interests with a commercial stake in the guidelines’ recommendations. If contemplated, then complete transparency regarding arrangements is necessary.  In some cases this might lead to more good than harm and it would be good to be open to that possibility.
**

Dr John Dowden, Editor, Australian Prescriber

Clinical trials help us to determine which patients will benefit from therapy to prevent deep vein thrombosis. The evidence from these trials should then be assessed in the production of guidelines. Implementation of these guidelines will then ensure that each patient receives the optimum treatment.

The problem with preventing venous thrombosis is that there no consensus on exactly who should be treated. The National Health and Medical Research Council is addressing this by developing a guideline through the National Institute of Clinical Studies.

Guidelines are expensive to produce and the money has to come from somewhere. The cost of production may be why some institutions are happy to adopt guidelines produced by somebody else. However, if there is sponsorship by a drug company, it inevitably raises questions of bias. Conversely, some would argue that a publically funded guideline may have a bias towards cost minimisation.

The success of a guideline probably depends on how much health professionals trust the source. They will also want to know how the guideline was developed and that it was based on evidence from the clinical trials.