To the Editor.
—Dr Bok,1 discussing informed consent in tests of patient reliability, mentions two methodological problems that contribute to one horn of the dilemma of ethical research. If informed consent is required, then some patients will refuse to participate, thereby altering the population to which the results of the study can be generalized. In addition, some participating patients may give different answers to sensitive questions than they would have given had they been kept in partial ignorance of the study's goals. In so doing, they will change the meaning of their responses to something different from those they would have given had informed consent been avoided. Bok describes the other aspects of the dilemma very clearly so I will concentrate my comments on these two methodological problems.If informed consent is regarded as an essential part of the ethical design of a particular study, then some questions become
Holland PW. The Self-reporting of Cocaine Use. JAMA. 1992;268(17):2373-2374. doi:10.1001/jama.1992.03490170045013