The tale of two sciences: Medical knowledge in the 21st century

Peter Schattner, Monash University

Jeanne Daly Evidence-based Medicine and the Search for a Science of Clinical Care, California, University of California Press, Berkeley and Los Angeles, 2005 (275 pp). ISBN 0-52024-316-1 (hardcover) RRP $178.95.

When you go to the doctor, you want them to listen to you and to be caring, but you also want them to know what they are doing. Doctors also want to do the right thing by their patients, even in the days when that meant using leeches to treat various ailments. There has long been a scientific aspect to medicine, and the recent advent of what has come to be known as ‘evidence-based medicine’ (EBM) is in that tradition.

EBM has been defined as ‘the conscientious, explicit and judicious use of current best evidence in making decisions about the care of individual patients’ (Sackett et al. 2000).Sackett and his colleagues go on to explain that the practice of EBM means integrating clinical expertise with the best available external clinical evidence from systematic research. Given that this approach to the practice of medicine seems almost self-evident, why has it aroused such controversy within the profession?

Jeanne Daly, an Australian academic and author of several books about medical research, has written a history of the development of EBM in a new book titled Evidence-based Medicine and the Search for a Science of Clinical Care. She explains the basis of the controversy over EBM through a series of interviews she conducted with its leading exponents. The book provides readers with a fascinating insight into the development of EBM and gives interesting, brief synopses of the lives of its leaders. For anyone interested in the intellectual evolution of modern medicine, this is a valuable and accessible book, free of technical jargon and full of illustrative anecdotes.

EBM protagonists successfully promoted the discipline into a world-wide phenomenon.

Daly explains that the 20th century saw great advances in knowledge about the pathophysiology of disease. This was particularly so for infectious diseases, but not limited to them. As a result, there was increasing faith in the authority of experienced physicians who believed that the key to clinical practice lay in understanding the pathophysiological basis of disease, which became the dominant scientific paradigm in medicine.

One of the first to challenge this paradigm was Alvan Feinstein, an American known as the father of clinical epidemiology. This discipline relies heavily on the use of statistics to interpret clinical phenomena. For example, one can calculate the risk of an individual smoker developing lung cancer by examining population data on the incidence of lung cancer in smokers compared to non-smokers. Feinstein thought that clinicians should make decisions about diagnosis, prognosis, and therapy for individual patients on the basis of such probabilities rather than some underlying biomedical theory of the disease process. He believed that this kind of empirical approach to patient care was likely to lead to better clinical outcomes (Chapter 2).

David Sackett and his colleagues at McMaster University in Toronto, Canada, took clinical epidemiology one step further. Clinical epidemiology had made very little headway among clinicians, and so the McMaster group developed an empirical method that could be used in day-to-day practice. This method is known as EBM and it was named as such in 1990 by one of the McMaster disciples, Gordan Guyatt. The process is: ask a clearly delineated clinical question; conduct a thorough search of the literature which pertains to that question; critically appraise the validity and strength of the evidence; and apply this to the original clinical problem.

Through numerous workshops, lectures, publications and seminal textbooks, the EBM protagonists successfully promoted the discipline into a world-wide phenomenon. As Daly describes it, charismatic leadership helped create a scientific movement with few if any rivals in the realm of modern medicine (Chapter 4).

EBM has also had a significant impact on the practice of medicine in Australia. It is taught at undergraduate level and its language has widely penetrated professional discourse. Former authority figures are ubiquitously challenged by the question ‘Where is the evidence?’, and the medical literature is now replete with EBM terms such as ‘the number needed to treat’ (to prevent one adverse outcome).

The Cochrane database has not proved to be directly useful to many clinicians.

One example is in the management of hypertension. In the bad old days (prior to 1990), patients whose blood pressures were consistently higher than ‘normal’ were told that they required drug treatment to prevent complications. However, neither the patient nor the clinician really understood what the level of risk for that individual was. Nowadays, the profession is able to base its treatment on better information, namely the ‘number needed to treat’ (NNT). In this case, the NNT might be 50, that is, 50 patients with hypertension have to be treated for five years in order to prevent one stroke. This quantification of risk can be very helpful in guiding clinical decisions.

Almost contemporaneously with the McMaster development, the British developed their own version of EBM (Chapter 6). Archie Cochrane, a doctor in a mining community in Wales, argued that each clinical speciality should collect and synthesise all trials conducted in its field and then issue regular updates and critical summaries. Out of this vision grew the Cochrane Collaboration (Chapter 7), an international fraternity of academic volunteers that produces systematic reviews of clinical trials (or, when trials are not available, ‘weaker’ forms of study design such as observational studies).

Although the McMaster and the British schools developed different approaches, they agreed on two basic concepts: clinical trials are the most valid means of testing ‘what works’; and that this evidence should be actively considered when making decisions about individual patients.

The main difference between the two schools is that the McMaster group initially suggested that clinicians themselves should search for and appraise original research papers, whereas the Cochrane group were ‘knowledge synthesisers’. This meant that the Collaboration did the hard work for clinicians by sifting through the literature and compiling it into a database comprising the best studies in the field. When studies on a topic were similar enough in their aims, methods, and patient characteristics, they would combine the results of individual trials to produce a ‘meta-analysis’. Such analysis is considered to be at the top of the evidence ‘hierarchy’, with observational studies and individual case reports lower down the pyramid.

The difference between searching for and appraising the literature, and looking up a database of pre-appraised reviews (in the Cochrane Library), highlights what came to be one of the principal objections to EBM—that is was not practical for clinicians. Over time, proponents of EBM conceded that it was not realistic to expect clinicians to practice the four steps of EBM (questioning, searching, appraising, and applying). Summaries of evidence, including that produced by the Cochrane Collaboration, would be more accessible for the majority who are ‘users’ of EBM rather than ‘doers’ of EBM. These summaries are freely available on the Collaboration’s website.

Randomised controlled clinical trials are almost impossible to conduct in some fields.

Unfortunately, even the Cochrane database has not proved to be directly useful to many clinicians. Systematic reviews are long and technical, and the database does not cover a great number of clinical areas of interest (no more than 10 per cent of published trials have been summarised by the Cochrane Collaboration to date). These difficulties have led to two further developments: implementation research and clinical practice guidelines (Chapter 9).

Implementation researchers study what makes clinicians change their professional behaviour. They examine which incentives (carrots and sticks) and educational strategies are most effective in encouraging doctors to take account of the best available evidence. This has proved to be a very complex and difficult area, because clinicians are more resistant to change than many would like or expect.

Clinical practice guidelines are recommendations based on evidence, with knowledge gaps filled in by expert advice. Ideally the guidelines are tailored to take into account regional traditions, and are promoted by ‘opinion leaders’. However, even these have not been widely successful in bringing about changes in clinician behaviour. Again, this seems likely to be related to resistance to change per se rather than any fault in the design of guidelines.

The most successful approach to changing clinician behaviour appears to be a combination of strategies. For example, repeated information about the limited benefits of the use of antibiotics for mild respiratory tract infection, the risks associated with these medications, clinical audits, peer-based educational discussions, clinical guidelines based on systematic reviews, public education, and the rising cost of drugs have led to a decline in the use of antibiotics for this purpose in recent years.

One of the ironies of clinicians relying on experts producing evidence summaries is that the genuflection before authority has come full circle, although the new elite are those with academic rather than biomedical expertise. David Sackett came to understand this paradox and therefore decided in 2000 that he would ‘never again lecture, write or reference anything to do with evidence-based clinical practice … [because] … my conclusions came to be given too much credence and my opinion too much weight’ (p. 240). He realised that he had become an ‘expert’, and people were now listening to him out of respect rather than because his utterances were necessarily well founded.

What are EBM’s other shortcomings for the non-academic clinician? Daly describes several, including the enormous weight that EBM gives to clinical trials. While clinical trials might be useful for medical interventions (testing what works), they are not particularly relevant for many other clinical encounters. For example, trials do not help if a patient is after reassurance or an explanation rather than effective treatment.

What is required is a science broader than that provided by EBM—a science of clinical care.

Another problem with clinical trials is that their application to individual patients requires a leap of faith. For example, trials have very strict patient entry criteria so that people with serious co-existing illnesses are usually excluded. That means that the probabilities associated with subjects in a trial might not necessarily apply to very different patients who attend doctors. A common example involves the treatment of high cholesterol in the elderly; do the results of trials on younger subjects apply to them? We cannot be sure because there is no exact science which bridges the gap between studies on groups of people and individual patients.

Another shortcoming is that trials are almost impossible to conduct in some fields. That is why health service research is so contentious. It is simply impractical to randomise large-scale community programs and it is rarely possible to ‘double blind’ them (that is, to prevent those who involved from knowing whether they are in the treatment or non-treatment group). Daly points out that clinical epidemiology, which is the basis of EBM, is difficult to apply to populations, even though epidemiology itself was originally a population-based science. In fact, tension has existed between the disciplines of clinical epidemiology and public health, with many pointing out that EBM has failed at the community level to address such important social issues as health inequalities.

So, what is the place for EBM some fifteen years after its advent? Daly concludes that EBM has been very useful, but in a more limited way than its originators had hoped. It has improved the science of testing ‘what works’ and has led to very useful knowledge syntheses in, for example, systematic reviews. It has relied on quantification to produce objective knowledge and reduce uncertainty. This reductionist science, embodied in the randomised controlled trial, has been highly successful but is narrowly focused, and does not cover all that happens in clinical encounters. In particular, it does not provide a solution for other forms of medical knowledge, such as that which is derived qualitatively. Its original degrading of the status of experts has also ironically led to a new kind of authority, leaving the clinician as reliant on others for guidance as was previously the case.

Daly concludes that what is required is a science broader than that provided by EBM—a science of clinical care (Chapter 10). She asserts that this must be an interdisciplinary science, and should draw upon fields traditionally beyond the scope of medicine, such as sociology. She observes that the Danish academic, Henrik Wulff, argued that clinicians need to integrate theoretical knowledge from biomedicine, empirical knowledge of outcomes in patients (preferably in the form of randomised controlled trials), and an interpretative understanding of patients that includes an ethical perspective. Obviously, EBM has a well-earned and rightful place within this panoply, but the struggle for pre-eminence among the various disciplines is still being played out today.

EBM is of major interest and direct relevance to consumers of health care. Its impact has been quite profound in several clinical areas, including that of ‘complementary medicine’ (also known as alternative or integrative medicine). Should you take shark cartilage to prevent cancer, or Echinacea to treat colds? The answer does not lie in their claims to be ‘natural’, or that your friend swears by them. When it comes to medical therapeutics, the public needs to be guided by the evidence.

Reference

Sackett, D., Strauss, S., Richardson, W.S., Rosenberg, W., & Haynes, R.B. 2000, Evidence-based Medicine: How to Practice and Teach EBM, 2nd edn, Churchill Livingstone, London.

The Cochrane Collaboration 2005 [Online], Available: http://www.cochrane.org/index1.htm [2005, Oct 14].

Peter Schattner is a general practitioner and part-time clinical associate professor in the department of general practice at Monash University. He has a particular interest in evidence-based medicine and has taught this subject at undergraduate and post-graduate levels. He has also researched the role of computers in promoting a more evidence-based general practice.