Contents
A user’s guide
Alejandro R Jadad
Foreword
Murray W. Enkin, MD, FRCS(C)
Professor Emeritus, Departments of Obstetrics and Gynaecology, and Clinical Epidemiology and Biostatistics, McMaster University, Canada.
Around 600 BC, Daniel of Judah conducted what is probably the earliest recorded clinical trial. He compared the health effects of a vegetarian diet with those of a royal Babylonian diet over a ten day period.1 Despite the dramatic findings of the study, over four centuries elapsed before publication of the results. The trial had obvious deficiencies by contemporary methodologic standards (allocation bias, ascertainment bias, confounding by Divine intervention),2 but the publication has remained influential for over two millennia.
Other controlled clinical studies with methodologic weaknesses but important effects on practice have been undertaken during the ensuing centuries. Ambrose Paré (1514-1564), in an unplanned experiment, found that applying a soothing ‘digestive medicament’ to battle wounds produced better results than the traditional practice of cauterising wounds with boiling oil.3 Inoculation to prevent smallpox became popular after Maitland conducted a trial upon six Newgate convicts in 1721,3 although the numbers treated and the precision of the trial were not adequate to give a fair picture of the effects of the procedure. Jenner published his famous studies on vaccination at the end of the eighteenth century, based on 10 and 14 persons. Appalled by the ravages of scurvy among ships crews on long voyages, in 1747 James Lind conducted a comparative trial of the most promising scurvy cures, using as subjects 12 sick seamen on board the Salisbury at sea. ‘The most sudden and visible good effects were perceived from the use of the oranges and lemons.’ The British Navy did not supply lemon juice to its ships until 1795.3
The nineteenth century saw many major advances. Probably the most sophisticated trial of a preventive type was a before/after study conducted by Ignaz Semmelweis in 1847. He noted that maternal mortality was much higher among women delivered by physicians and medical students, who were in frequent contact with cadavers at autopsies, than among women delivered by pupil midwives. After considering various hypotheses he reasoned that ‘the cadaveric particles clinging to the hands are not entirely removed by the ordinary method of washing the hands’, and introduced the practice of more thorough washing and disinfectant.4 Maternal mortality among the doctor-delivered mothers dropped by 50 per cent in the subsequent six months, although still not to as low a level as that achieved by the midwives.
Credit for the modern randomised trial is usually given to Sir Austin Bradford Hill. The historic MRC trials on streptomycin for pulmonary tuberculosis5 are rightly regarded as a landmark that ushered in a new era of medicine. Their influence on the science of therapeutic evaluation was strengthened because the charismatic Hill followed up that work with lectures and articles6 reinforcing his message. Since Hill's pioneer achievement randomised trial methodology has been increasingly accepted, and the number of randomised controlled trials reported has grown exponentially. The Cochrane Library7 already lists more than 150,000 such trials, and they have become the underlying basis for what is currently called ‘evidence based medicine’. The concept has rightly been hailed as a paradigm shift in our approach to clinical decision making.8
It is not, however, the first such paradigm shift. A similar scientific revolution was hailed more than a century and a half ago, by the editor of the American Journal of Medical Sciences in 1836, in his introduction to an article which he considered to be ‘one of the most important medical works of the present century, marking the start of a new era in science’. It was ‘the first formal exposition of the results of the only true method of investigation (emphasis added) in regard to the therapeutic value of remedial agents’. The article that evoked such effusive praise was the French study on blood-letting in the treatment of pneumonia by PCA Louis.9,10
At that time blood-letting was the almost universally accepted ‘proper’ method of treating pneumonia. Louis used the quintessential Baconian approach of gathering vast amounts of data, which allowed him to make comparisons and systematically investigate the efficacy of treatments. His conclusion from that study was a bombshell; that the apparent efficacy of bleeding for pneumonia was a mere therapeutic illusion. His contribution to clinical epidemiology was to base recommendations for therapy on the results of collective experience, rather than on limited individual experience, tradition, or theory.
Louis' approach, and his evangelical zeal in promoting his methods created considerable controversy. He attracted many foreign disciples, including Oliver Wendell Holmes and William Osler who made their mentor's work available to American readers. He also attracted strong opposition, and his work was mired in controversy. His opponents were numerous and vociferous: ‘The physician called to treat a sick man is not an actuary advising a company to accept or deny risks, but someone who must deal with a specific individual at a vulnerable moment’ and ‘Averages could not help and might even confuse the practising physician as he struggles to apply general rules to a specific case.’ Practising physicians were unwilling to hold their decisions in abeyance till their therapies received numerical approbation, nor were they prepared to discard therapies validated by both tradition and their own experience on account of somebody else's numbers.10
Although doubtless they arose partly from an innate resistance to change, and partly from misguided self-interest, the arguments against a widespread application of the so-called numerical approach stemmed largely from a lack of understanding of its intent. When both practitioners and public finally became aware that collective experience enhanced, rather than replaced, the clinical skills of the individual physician, Louis' numerical approach became the basis of medical research and literature until the midpoint of this century. It was by no means a panacea, but was an enormous step on the way towards more effective health care.
The arguments heard against the numerical approach in the last century are remarkably similar to those used against evidence based medicine today. Worries are still being expressed that evidence based medicine confuses statistics with reality, results in a loss of clinical freedom, and ignores the importance of clinical experience and of individual values.11 These concerns stem from the mistaken belief that the proponents of evidence based medicine claim a multicentre double blind placebo controlled randomised trial to be the only way to answer a therapeutic question. This, despite the fact that Austin Bradford Hill himself said ‘Any belief that the controlled trial is the only way would mean not that the pendulum had swung too far, but that it had come right off its hook’.12 Evidence based medicine is simply the conscientious and judicious use of the current best evidence from clinical care research to guide health care decisions. It is another enormous step towards more effective health care. No more, and no less.
One reason for the sometimes expressed opposition to evidence based medicine is a lack of understanding of the meaning of a randomised trial. This failure of understanding is not due to a paucity of information; there is a vast literature about randomised trials, their purpose, their methodology, their limitations. Unfortunately, much of that literature has been incomplete, has been biased, or has been couched in impenetrable jargon. It is not surprising that it has often been misinterpreted.
That is why this book is so welcome. It is written in clear, explicit, and understandable language, for those who use, would like to use, or should use, the results of randomised trials. It provides an accurate and comprehensive description of the randomised trial, its importance, when (and when not to) do a trial, how to interpret the results, when (and when not to) translate the results into health care decisions. It is a book to read, reflect on, learn from, use, and enjoy.
References
1. Book of Daniel. In: The Holy Bible.
2. Grimes D. Clinical research in ancient Babylon: methodologic insights from the book of Daniel. Obstet Gynecol 1995;86:1031-4.
3. Bull BP. The historical development of clinical therapeutic trials. J Chron Dis 1959;10:218-243.
4. Semmelweis I. The etiology, the concept, and the prophylaxis of childbed fever (1861).Translated by Carter KC. University of Wisconsin Press. 1983.
5. Daniels M, Hill AB. Chemotherapy of pulmonary tuberculosis in young adults: an analysis of the combined results of three medical research council trials. Br Med J 1952; 1:1162-8.
6. Hill AB. The clinical trial. New Eng J Med 1952;247:113-19.
7. The Cochrane Library. Oxford: Update Software. 1998, issue 1.
8. Evidence-based Medicine Working Group. Evidence-based medicine: a new approach to teaching the practice of medicine. JAMA 1992;268:2420-25.
9. Louis PCA. ‘Researches into the effects of blood-letting in some inflammatory diseases and on the influence of tartarized antimony and vesication in pneumonitis. Am J Med Sci 1836;18:102-11 (cited in Rangachari 1997).
10. Rangachari PK. Evidence-based medicine: old French wine with a new Canadian label? J Royal Soc Med 1997;90:280-4.
11. Charlton BG. Restoring the balance: evidence-based medicine put in its place. J Eval Clin Pract 1997;3:87-98.
12. Hill AB. Heberden oration, 1965. Reflections on the controlled trial. Ann Rheum Dis 1966;25:107-13.
|
Home | Contents | Foreword | Introduction | Acknowledgments | How to order
© BMJ Books 1998. BMJ Books is an imprint of the BMJ Publishing Group. First published in 1998 by BMJ Books, BMA House, Tavistock Square, London WC1H 9JR. A catalogue record for this book is available from the British Library. ISBN 0-7279-1208-9