Drugs on the Market*
THE first essential in appraising a drug, personally or by proxy, is to judge the possibility of bias, which takes so many forms. The crudest form, financial interest, is almost necessarily present in the case of statements of drug firms. It may be crude indeed, in which case, it generally may be spotted by all but intellectual infants. But it can be assumed that even the best-intentioned firms are not blind to the interests of their stockholders, and are so hopeful of making a “strike” that they are likely to err on the optimistic side, once they are committed to marketing a drug. This does not necessarily disqualify their evidence, but it instinctively should enjoin extra caution in appraising their estimate, especially that of the salesmen who are quite apt to go somewhat further than the firm itself.
“Controls” are the one method by which an honest investigator can eliminate personal bias, conscious, unconscious or accidental. But when are controls “adequate” in kind and in degree? The answer is, “that depends.” It varies for each case. As to kind, they must be such that they cannot be modified at the whim of the investigator. That means usually a “blind” test, such as comparison with a placebo or an established drug with the conditions set down in advance, so that the investigator is left no discretion, once the series is started; for instance, alternate admissions receive the drug to be tested, the other the placebo. Of course, patients must be comparable,. . .