She is far more likely to feel relief after taking the medicine than a patient who does not know if the pill she swallowed contains an active compound, a phenomenon known as the placebo effect. Similarly, knowing whether a patient received the drug or a placebo would likely influence a researcher's evaluation of that patient. For these reasons many clinical studies are designed to be double blind, that is, neither the patient or the researcher knows what treatment the patient has received.

In addition to well-matched treatment groups and double blinding, a good-quality clinical trial also incorporates controls for other factors unrelated to the drug being tested but that nonetheless may influence the treatment outcome. For example, THC reduces anxiety in some people to the extent that they mistakenly believe their symptoms have improved. Although anxiety reduction may be a valuable form of treatment for some patients, it also interferes with attempts to determine whether THC relieves specific symptoms. Successful clinical trials must therefore eliminate this influence—for example, by comparing the effect of THC on a particular symptom with that of a drug known to reduce anxiety but not the specific symptom being studied.

While double-blind, randomized, controlled clinical trials are the best way to evaluate a drug's effectiveness, such trials are not always feasible. For example, children, women of childbearing age, and the elderly are often excluded from experimental drug trials for safety reasons, yet patients in all of these groups take prescription medications. To get around this problem, medical scientists sometimes conduct single-patient trials in which individuals—including patients from vulnerable populations—are treated sequentially with several different medications or are given alternating doses of an experimental drug and a placebo. Although limited in scope, single-patient trials can permit objective comparisons between treatments.

The next six chapters describe scientists' initial attempts to test the safety and effectiveness of marijuana and cannabinoid drugs in the treatment of human patients. The discussion is limited to research on conditions that marijuana has been most often claimed to help, such as pain, AIDS, cancer, and muscular spasticity. Other sources, particularly recent books by Grinspoon and Bakalar and Mathre,1 discuss indications for marijuana beyond

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement