We agree that informed-consent forms should be improved (Nature 482, 16; 2012) beyond shortening and simplifying them. Participants need to know whether a trial is addressing an unmet clinical need, and that the test is necessary to answer a key question.

For example, a 'me-too' drug that seeks marketing authorization by comparison against a placebo, rather than against alternative products, will not solve a patient's unmet need. Neither is it in participants' best interests for a trial to establish the acceptability of a drug with inferior efficacy that provides no other advantage. If the informed-consent form were to explain these points, it is likely that far fewer patients would agree to participate.

Patients also need to understand why randomization is the only scientific and ethical way to provide meaningful results when a trial's outcome is unpredictable. Any other strategy, such as giving or withholding an untested drug, would be arbitrary and would not clarify the uncertainty.

Details of the therapeutic rationale, and of the pharmacological and toxicological properties of the experimental intervention, should be provided to trial participants on request.

Informed consent should reassure patients that trials aim to help them, and will allow onto the market only products that do so, rather than products that just meet commercial expectations. There is a long way to go in cleaning up clinical trials before the trust of patients is regained.