This lecture will discuss some of the important issues associated with the theory and practice of clinical trials. The lecture will consist of seven parts, i.e. early history, background, strategy of experimentation, informed consent and its ramifications, compliance and its effect on bias and efficiency, double randomized consent design and ethical issues.

The early history will discuss the work of James Linde who carried out his classic studies on scurvy in 1747. However, the prophylactic effect of lemon juice was discovered by the East India Company almost 150 years before. In the 19th century, many American physicians journeyed to France to study with the famous French physician, P.C.A. Louis. He wrote that one must "count" in order to learn about the benefits of therapy. However, counting is not easy. He wrote that "unfortunately it requires more labor and time than the most distinguished members of our profession can dedicate to it."

One question which dominates the transfer of clinical investigation outcomes to the clinic is what proportion of therapies shown to be positive from a clinical trial are actually positive. It is shown that this proportion is not as high as one generally believes. Our conclusion is that there are large numbers of false positive therapies in current clinical use.

Clinical trials carried out in the U.S. require affirmative patient consent. As a result, the conclusions from a trial only apply to those patients who would give consent. Very often the consent process divides patients into prognostic groups where those who decline to give consent have poorer prognosis than those patients who do consent. As a result the conclusions of a trial may not strictly apply to the population of people with disease.

Compliance is a key factor in a well conducted clinical trial. However, lack of compliance may induce biases and result in very low efficiencies. This is especially important in prevention trials which attempt to change behavior, e.g. smoking, drinking, dieting.

Many physicians have been reluctant to participate in randomized clinical trials because they feel that the patient-physician relationship may be compromised. In an attempt to increase physician participation in randomized trials, a new way of designing randomized trials was introduced several years ago. These trials are called "randomized consent" or "pre-randomized" designs. These will be discussed from both practical and ethical points of view.

Finally, clinical trials have generated new ethical concerns. Several neglected ethical issues will be discussed which require more public discussion.

Back to Workshop Schedule