ntmeds.com

PAIN TREATMENT: TESTING THERAPY FOR EFFECTIVENESS

The aim of these tests is to reveal the consistent action of the therapy liberated from the bias of patients and therapists who think it ought to work. The major bias is the placebo response, which is surprisingly powerful.The standard test for efficiency is the randomized double-blind placebo-controlled trial. ‘Randomized’ means that patients with a particular condition have been chosen at random in order to avoid choosing some special types who might tilt the result. ‘Double-blind’ means that neither the patient nor the people running the trial know if the real therapy or a mock therapy has been given. ‘Placebo-controlled’ means that the true therapy and a mock imitation of the true therapy have been given.To give an example, suppose that a pharmaceutical company wishes to market a new analgesic and that they have completed all the preliminary trials for safety and apparent beneficial effect. They are then required by law to submit their new tablet to this rigorous type of trial. A group of patients with some definite problems, a wisdom-tooth extraction for example, are asked to volunteer for a trial. They are told they will either receive the new tablet or a blank one that looks exactly the same. Then the patient, who does not know which tablet he received, tells an observer, who is also unaware of the nature of the tablet, whether the tablet reduced his pain. Finally, after all the data have been collected, the code is broken and it is calculated whether the new drug is superior to the placebo.This type of trial sounds simple, even if it is elaborate and costly, but there are severe problems. The first problem is that the tested group never includes all the types of people who might use the therapy: a test group in New York is unlikely to contain many Inuits. The group is usually deliberately restricted, to healthy young adult men, for example. This means that the results do not necessarily apply to Inuits, women, children, old people and so on.Much more serious is the famous difficulty in keeping a secret. A tested drug may have side effects such as drowsiness so it is obvious to patients and observers who has received the active therapy. There are ways around this particular problem by deliberately giving a sedative that is not believed to be an analgesic to compare with the tested analgesic that has sedative side effects. It will be seen that the crucial element of keeping patients and staff ignorant becomes more and more difficult as the therapy becomes more elaborate. Take the problem of subjecting acupuncture to a rigorous trial. What would be the placebo arm of the trial? The problems escalate to an extreme with surgical therapy, where it would be quite unethical to subject a patient to a general anaesthetic and a mock operation in order to test the true efficiency of the surgery.There are subtle ways around these problems, which we will discuss with particular therapies. However, no one should forget that the background for these trials is based on the powerful assumption that all people are the same and that individual psycho-social factors are irrelevant. This leads to the present vogue for ‘evidence-based medicine’, driven partly by the tradition of academic medicine and partly by the financiers’ need to identify proven therapy whose cost is justified by trial.The trials are designed to identify a class of pains, medicines and patients who respond reliably and to exclude a class where personal individuality is a factor. The separation of these two classes is itself an artefact because they interact.*52\219\2*

Share and Enjoy:
  • Digg
  • Sphinn
  • del.icio.us
  • Facebook
  • LinkedIn
  • Reddit
  • StumbleUpon
  • Twitter
  • Yahoo! Bookmarks

Random Posts

Comments are closed.