Email this article to a friend Respond to this article Read responses to this article Download to Citation Manager Search Medline for articles by: Dewhurst, J. Alert me when: New articles cite this article

reviews

The evidence base: rock of certainty or shifting sands?

I have a problem with evidence based medicine. This I know is a serious confession and could amount to professional misconduct, but in mitigation let me explain. The case I have is my own, so I will presume to tell you a little about myself.

After some years as a principal in general practice I joined the pharmaceutical industry. My medical history is unremarkable apart from two episodes of major depression, the first in my early thirties. For this I was given amitriptyline. The drug had no perceptible effect on the depression, but I became completely unable to concentrate, and my short term memory was seriously impaired. I had always done a lot of woodwork, but I found now that in the few seconds between taking a measurement and applying it to a piece of wood I had forgotten it.

However, the most disabling effect—one that I have never seen described—was a distancing from reality. I seemed to be living in a mental cocoon from which the outside world could be reached only with conscious effort. No one had warned me about driving a car, and on several occasions I avoided a serious crash only at the last second through a painful effort to get a proper awareness of my surroundings. The doctors' response was that this was because I was still depressed and that the dose must be increased.

After three or four months I simply stopped taking the drug, and after 24 hours of near manic euphoria I felt much better, having now to cope only with the depression. I was referred for psychotherapy. Whether this prolonged or shortened the depression there is no way of telling, but apart from the expense and time it didn't seem to do much harm.

When I joined the pharmaceutical industry, and later when running a clinical trials organisation, overseeing clinical trials was a large part of my job. I then realised for the first time how ineffective "effective" drugs actually are. A study may show that a treatment is effective, in that it shows a statistically significant positive response, even though less than half the patients receive any real benefit. The other patients may experience only the adverse effects or no effects. In one double blind, randomised study comparing a new antidepressant with amitriptyline and with placebo we were disconcerted to find that the best treatment was placebo. This study, in common with many other "negative" studies, was never published. Publication bias is probably greater than allowed for. But, more importantly, what is the meaning of "effective"?

Towards the end of my time with the clinical trials organisation I again developed full blown depression. Despite my account of my previous treatment with antidepressants I was put on tricyclics, the earlier history being considered irrelevant and misleading. I had depression and the evidence base was clear that antidepressants were effective.

All the old toxic effects returned, and some new ones. I was used to designing things in my head, and often I did this in bed before I went to sleep. I could design quite complex structures, turning them round and handling them in my mind in such detail that I could go straight to work on them the next day without need for drawings or other preparation. Now I was unable—without much confused thought and many blunders—to make even the simplest object.

I also have a benign familial tremor, inherited from my father. Before it had been mild and easily compensated for, but it now became so gross that I could not hold a cup or glass. I also became unsteady on my feet, partly because of trembling in my legs and partly because of loss of balance. At night I often staggered and occasionally fell.

After about four months, with the depression unimproved, I determined to stop treatment. This caused consternation. I was persuaded to take mianserin, fluoxetine, and a monoamine oxidase inhibitor, none of which had any benefit.

At this point I was referred to a consultant psychiatrist, who dismissed all previous treatments as misconceived. I was put on amitriptyline in escalating doses. I was now almost a zombie, and so far as I could distinguish the illness from the treatment there was no improvement in the depression. In desperation I stopped taking the tablets. After a day of euphoria I felt much better.

Why did I allow myself to be subjected to this treatment? With depression there is loss of self esteem and loss of will. Those treating me relied completely on the evidence base and were authoritative in the certainty of their advice.

So you see the problem I have with evidence based medicine. The evidence is not understood. Perhaps my problem is not so much with evidence based medicine as with its users, who find in it a useful but sometimes dangerous rock of certainty in an uncertain world. I'm sorry, colleagues, but you've still got to observe, and you've still got to think.

Wokingham, Berkshire John.Dewhurst{at}ukgateway.net

Rapid Responses:

Email this article to a friend Respond to this article Read responses to this article Download to Citation Manager Search Medline for articles by: Dewhurst, J. Alert me when: New articles cite this article