Abstract
It is important that physicians practise evidence-based medicine. Clinical experience is important, but there are a number of reasons why clinical experience can lead to the impression that ineffective treatments are effective. There are major educational and research challenges which must be met before clinicians can practice evidence-based migraine therapy more extensively than at present. Treating physicians will need to learn more about the principles of evidence-based medicine. Researchers will need to produce more and better clinical trials that address important clinical questions. The results of these trials will need to be reported clearly, and we need to improve the efficiency with which these results can be accessed. It is important that the pharmaceutical industry, clinicians, and academic health centres work together to meet these challenges.
Keywords
Many physicians believe that they practise evidence-based medicine (EBM), although most would agree that keeping up to date in today's world of rapid change in medicine is difficult. In fact, studies have shown that, as clinicians deal with patients, there is a frequent need for new clinical information, but this information is often not obtained by physicians for a variety of reasons (1).
Many physicians also express frustration at the lack of good evidence in many areas of medicine. However, although lack of good evidence for decision-making is still an issue in many areas of medicine, much good evidence does exist, but is not necessarily applied in the frontlines of patient care (2).
Other physicians feel that EBM may lead to a ‘cookbook’ approach to medicine, and that the art of medicine and ‘clinical expertise’ will be somehow lost, yet it is difficult to argue with the merits of EBM when it is defined as integrating individual clinical expertise with the best available external clinical evidence from systematic research (3).
Clinical experience
Some clinicians appear to depend more on clinical experience than on evidence from the medical literature. While clinical experience contributes greatly to the physician–patient encounter, it is important to maintain a healthy scepticism regarding treatment decisions based entirely on clinical experience. It is sobering to reflect that clinical experience led to the use of arsenic for migraine headache treatment for many decades, and this persisted well into the 20th century (4).
Sackett (5) succinctly outlines a number of reasons why the time honoured ‘trial of therapy’ can mislead clinicians, and indeed lead to the conclusion that a useless therapy is effective. These reasons are summarized in Table 1. Good knowledge of the scientific basis of medicine is therefore an important partner to clinical experience.
Reasons why clinical experience may suggest that ineffective therapies are effective
Modified from (5).
Migraine treatment and evidence-based medicine
Evidence-based migraine treatment presents formidable challenges. Not all of these challenges are unique to migraine, but some other areas of medicine may be more amenable to EBM than migraine. As in most areas of medicine, the foundation of EBM in migraine therapy is the results of well executed clinical trials.
Firstly, the aetiology and pathophysiology of migraine is still not completely understood. Although some treatments, notably the triptans, are based on our understanding of migraine pathophysiology, still too little is known. Prophylactic migraine treatment in particular suffers from a lack of an underlying pathophysiological framework. This makes it more difficult to develop potentially effective therapies that can then be tested in clinical trials.
Secondly, no specific diagnostic test is available for migraine. This necessitates the use of clinical criteria for migraine diagnosis. Although the International Headache Society (IHS) criteria (6) have been very helpful in migraine research trials, and also for better clinical diagnosis, they may still lack specificity and sensitivity. This introduces imprecision or unwanted ‘noise’ into the migraine diagnostic process and into clinical trials.
Thirdly, all outcome measures in migraine treatment are by necessity subjective. We have only the patient's subjective assessment and verbal response as outcome measures. Although in the end, it is perhaps only the patient's subjective assessment of a therapeutic response that really matters, it does make the scientific study of migraine more difficult.
Clinical trial design has come a long way over the last several decades, as has our understanding of what may influence the outcome of a clinical trial. Nevertheless, the high placebo response in clinical trials of symptomatic migraine treatment remains a challenge (7, 8). This high placebo-response rate is complex, and many factors may contribute to it. One factor may be the patient's perception of the chances of getting active treatment. There is evidence that the smaller the patient's chances of receiving the placebo vs. the active drug, the higher the placebo response rate (8). Migraine attacks by their nature are self-limited and by IHS definition last 4–72 h. However, some attacks experienced by patients with migraine are undoubtedly shorter and this could also contribute to the ‘placebo’ effect in clinical trials measuring outcome at, for example, 2 h, and especially at longer time points such as 4 hours. The overall result of these factors is that clinical trials of symptomatic migraine treatment need to be large, even for relatively effective drugs such as the triptans, and are therefore expensive.
The need for evidence-based medicine training
It is important that we rise to the challenge of providing evidence-based migraine therapy. It seems clear that, after the completion of training, the general physician's knowledge of a therapeutic area, e.g. hypertension, gradually becomes out of date. It has been shown that there is a statistically and clinically significant negative correlation between our knowledge of up-to-date care and the years that have elapsed since graduation from medical school (9, 10).
There is no reason to suspect that a physician's knowledge of migraine therapy is any different. It is gratifying to note that there have been many new developments in migraine therapy in the past decade. As a result, however, it is important that the knowledge of the practising physician be kept current, or as put so eloquently by Sackett ‘we become out of date and our patients pay the price for our obsolescence’ (11).
Perhaps almost as important, there is also evidence that training in EBM can help keep physicians up to date. It was found that graduates of a medical school with a problem-based EBM curriculum appeared to remain more abreast of important advances in the detection, evaluation, and management of hypertension, following graduation, than graduates of a more traditional medical school (12). Although there may be other explanations for these results, they do suggest that training in EBM can help keep physicians up to date in the currently rapidly changing medical environment. Happily, migraine therapeutics is now a part of this rapid change.
The need for more research
Educational challenges are, however, only some of the challenges to be met if we are to ground the contemporary practise of migraine therapeutics firmly in EBM. To practise EBM, there must be a solid evidence base from which to work. As in many areas of medicine, the evidence base for migraine therapeutics remains inadequate for a number of reasons.
One reason is that simply not enough good randomized clinical trials have been done in many areas of migraine therapeutics, even though much progress has been made. Jadad (13) has developed a wish list that, if fulfilled, would increase the ability of clinicians to practice EBM because a stronger evidence-base would be developed (Table 2). Evidence-based migraine therapy could be advanced significantly if we could meet the needs listed in Table 2.
Requirements of better evidence-based medicine
Modified from (13).
This supplement
How can we increase the inroads of EBM into migraine therapeutics? This supplement hopes to help in the process. In the next paper, Purdy (14) points out, on the basis of a needs assessment done for an EBM workshop held in Montréal in 1999, that many physicians, even specialist physicians, do not have as much knowledge of EBM as might be expected or desired, and that many are unsure which of the new symptomatic therapies to use in practice.
Wiebe (15) lays out for us the basics of EBM, and takes us through the five steps in the application of EBM to clinical problems.
Sheftell and Fox (16) raise many issues with regard to applying the evidence from clinical trials to clinical practice. They review a significant portion of the literature as it relates to symptomatic migraine treatment, and discuss in depth the various outcome measures that have been used, and discuss others which might be more desirable in certain situations.
Salonen (17) delves into the area of drug comparison trials, an important topic now that a number of good symptomatic migraine treatments are available in most countries. When to apply which treatment and in whom has become an important issue.
Finally, Gawel and Wiebe (18) carefully examine how to assess a therapeutic trial, using an important symptomatic migraine-comparison trial (19) as an example. As can be seen from their analysis, even a clinical trial designed by experts and carried out by competent researchers almost invariably raises a number of issues and unanswered questions.
The research challenge
This supplement tries to address some of the educational challenges around wider implementation of evidence-based migraine therapy. The research challenges are surely even more formidable. We need more and better data, and it is important that the medical community, including academic medicine and industry, work together to achieve this goal. It is interesting to note that a huge proportion of the current best evidence for symptomatic migraine therapy has come about through the efforts of industry, starting with the landmark clinical trials done during the sumatriptan development programme almost a decade ago (20–22). The pharmaceutical industry will no doubt continue to play a major role in expanding the migraine therapeutic evidence base. However, it is imperative that specialist physicians active in migraine therapy, along with their generalist colleagues and academic health centres, enlarge their role in migraine research and provide a balance to the industry-sponsored clinical trials. In this way, evidence can be obtained to guide migraine therapy into therapeutic areas in which there may be no incentive for commercial interests to fund research. In addition, any potential bias that might be introduced into migraine therapeutics, as a result of preferential funding by industry of selected aspects of migraine therapeutics, can be counterbalanced. Only in this way can the evidence base for migraine therapy be placed on a truly solid footing.
It is also important that all well executed clinical trials involving migraine treatment be published in full-length manuscripts, so that the trials can be assessed in detail by those involved in migraine therapy. One problem encountered by clinicians seeking to fully understand the nature of the various pharmacological treatments available for migraine is the unhappy tendency of some clinical trial sponsors, including corporate sponsors, not to fully publish the results. In this situation, abstracts are not good enough, as too few details are provided for full assessment of the trial. Many reviews of headache therapeutics, including some in this supplement, contain many references to abstracts, some dated 4 years or more previously. Surely the costs of publication are minimal compared to the cost of the trial itself. If limitation in personnel qualified to publish clinical trial papers is an issue, there probably are many academics at universities around the world who would welcome access to clinical trial data and the opportunity to publish the results of corporate-sponsored trials. New initiatives, such as that of the National Institute for Clinical Excellence (NICE) in Great Britain, which attempt to review all the evidence for a new intervention in a timely fashion and make it available to clinicians should also help to bring all clinical trials evidence into the public domain (see website at http://www.nice.org.uk/index.htm).
Efficacy and effectiveness
Efficacy can be defined as how well a treatment performs under the ideal and clearly specified conditions of a clinical trial (23). Effectiveness indicates how well a treatment performs in real-life clinical practice (23). We must not forget that, in addition to good clinical trials data, we need complementary data from the workplace to help us with the critical assessment of new headache therapies. More work in this area is needed, and large-scale well-designed clinical databases could help in this regard. From the participation of literally thousands of patients in the migraine clinical treatment trials of the last decade, it seems apparent that the migraine population is willing to help. There is every reason to believe that migraine sufferers will continue to contribute to clinical research. Because migraine is so common, data should not be difficult to obtain. However, a serious commitment from treating physicians is required. It is instructive to remember that literally hundreds of thousands of carotid endarterectomies done in clinical practice did not answer the question ‘does this intervention really make a positive difference to the patient, and if so to which patient?’ However, carefully done research trials involving only miniscule patient numbers in comparison with routine clinical volumes were able to provide conclusive answers to these questions (24).
Conclusion
I am confident that clinicians and patients together can use the currently available data and provide the additional data needed for the more complete practice of evidence-based migraine therapy. However, it will require hard work on the part of both, and a significant change in our attitudes both towards research and the practice of medicine to achieve this goal. The increasing recognition that medical school curricula need to change (25) with the provision of more EBM instruction will assist in this process.
Footnotes
Acknowledgements
I would like to acknowledge the support of Glaxo Wellcome who through an unrestricted educational grant to the Canadian Headache Society made the Evidence-Based Migraine Therapy Workshop in Montreal (August 1999) possible, and who also made possible publication of this supplement.