Patient expectations of benefit from phase I clinical trials: linguistic considerations in diagnosing a therapeutic misconception.

The ethical treatment of cancer patients participating in clinical trials requires that patients are well-informed about the potential benefits and risks associated with participation. When patients enrolled in phase I clinical trials report that their chance of benefit is very high, this is often taken as evidence of a failure of the informed consent process. We argue, however, that some simple themes from the philosophy of language may make such a conclusion less certain. First, the patient may receive conflicting statements from multiple speakers about the expected outcome of the trial. Patients may be reporting the message they like best. Second, there is a potential problem of multivocality. Expressions of uncertainty of the frequency type (e.g., "On average, 5 out of every 100 patients will benefit") can be confused with expressions of uncertainty of the belief type (e.g., "The chance that I will benefit is about 80%"). Patients may be informed using frequency-type statements and respond using belief-type statements. Third, each speech episode involving the investigator and the patient regarding outcomes may subserve multiple speech acts, some of which may be indirect. For example, a patient reporting a high expected benefit may be reporting a belief about the future, reassuring family members, and/or attempting to improve his or her outcome by a public assertion of optimism. These sources of linguistic confusion should be considered in judging whether the patient's reported expectation is grounds for a bioethical concern that there has been a failure in the informed consent process.