Unfortunately, medicine is like the fashion industry: It's trendy, filled with fads and the need to mindlessly espouse the current style.

To understand this, simply hark back to the rage of several decades ago: the problem-oriented medical record. The concept burst upon the scene like a Roman candle and made a similar exit, probably because it was mostly gimmick and not intellectually or philosophically based.

Our current fad is evidence-based medicine and again the intellectual and philosophical basis is scant. Even the inventors, Sackett et al, cannot define what they mean. In their book, Evidence-Based Medicine, they say, "Evidence-based medicine means integrating individual clinical expertise with the best available evidence from systematic research."

What's new about that? Certainly that is what I learned from my instructors when I went to medical school nearly 50 years ago and what my father told me he learned in medical school 80 years ago. In fact, Hippocrates understood the concept. It did not originate in Paris in the middle of the last century, as Sackett and his colleagues believe.

Hippocrates conducted a systemic project to determine the correlation of physical signs with imminent death to determine which were predictive. Never mind that the object was crassly commercial because hospitals (Aesculapions) in those days did not wish to admit dying patients because it ruined their reputations. It was a systematic research project and therefore was evidence-based medicine.

Having read Evidence-Based Medicine and come away not much the wiser, in sheer desperation I turned to the dictionary to see what the word "evidence" meant. Surprise. Evidence is simply "grounds for belief"-any grounds, not merely scientific. If a faith healer believes in what he is doing, by definition he is practicing evidence-based medicine!

Carrying the search further, I turned to a collection of articles from the British Medical Journal called "Logic in Medicine." There I learned the study of what knowledge is valid if covered by the branch of philosophy known as epistemology. It recognizes two forms of logic that produce knowledge: deductive and inductive.

Inductive logic is the basis of the scientific method and involves making observations to make conclusions. Deductive logic, on the other hand, allows conclusions to be made without observations. Certainly the former is preferable and in fact is what evidence-based medicine is all about. All the confusion could have been avoided if it simply had been accompanied by the appropriate adjective-"scientific." What Sackett et al really mean is scientific evidence-based medicine.

However, medicine cannot operate by the scientific method alone. Nobody has ever conducted a randomized controlled trial to prove that kidney transplants are beneficial. Nor will anybody ever do so. What, then, is the method of reasoning by which kidney transplants came to be considered an acceptable treatment? The answer is deductive logic, deduced from our knowledge of pathophysiology, not experimentally proven.

Unfortunately, there are times when a physician must act and cannot gather either inductive or deductive logic for support. This, then, causes him to go to the third level of evidence-experience or anecdote. Not highly valid, but in many instances better than inaction.

Finally, the evidence-that is, grounds for belief-may be faith.

As a stand-alone phrase without a descriptive adjective, "evidence-based medicine" is meaningless. Pure rhetoric. Or as was said many years ago, "Full of sound and fury, signifying nothing."

Clinical experience has been defined as "making the same mistake with increasing confidence for an impressive number of years," in contrast to "evidence-based medicine," which involves "perpetuating other people's mistakes instead of your own" (Anon.: Which humour for doctors? Lancet 351:1, 1998, quoting A Sceptic's Medical Dictionary, London, BMJ Publishing Company, 1997, cited in Sapira's Art and Science of Bedside Diagnosis by Jane M. Orient, Lippincott Williams & Wilkins, 2000).

Pamphlet No. 1071, April, 2000