While most conventional medications reach the public only after an extensive process of development and testing (see Chapter 10), medical knowledge concerning marijuana's potential benefits and risks has accumulated largely through its widespread use. As mentioned in the previous chapter, a recent poll indicates that approximately one in three Americans over age 12 have tried marijuana or hashish at least once, although only about one in 20 currently use these drugs.
Medical scientists know far more about marijuana's adverse effects than about its ability to relieve specific symptoms, mainly because of the difficulties of conducting clinical research on marijuana. In addition to securing financial support for their research, medical scientists who study marijuana must demonstrate their compliance with a multitude of federal and state regulations before carrying out their investigations (see Chapter 11). Thus, despite recent discoveries highlighted in Chapter 2, substantial clinical studies on the medicinal properties of marijuana remain scarce.
Yet clinical experiments must be undertaken to determine whether marijuana-based medicines live up to their promising performance in numerous basic science studies. Before any medication can be approved for sale by the U.S. Food and Drug Administration, it must pass a series of clinical trials to assure that it is both safe and effective. The trials, which are conducted on healthy volunteers and qualified patients, allow scientists to predict how drugs will perform in the general population.
In a well-designed clinical trial, patients are assigned to treatment groups in such a way that any possible biases in outcome are removed. For example, to compare two medications for nausea, the group of patients being treated with each drug should contain people of equivalent age, gender, and health status. Another approach to providing matched samples is the use of a “crossover” design in which all patients receive both the experimental drug and a placebo in random order.
Clinical trials should also be designed to eliminate the effects of both the patients' and the researchers' expectations concerning the results of the trial. Consider the patient who tries an experimental antinausea drug with the expectation that it will work.