RCTs that estimate the probability of a specific harm come in several flavors. Some compare a new therapy to placebo where the primary outcome is an efficacy measure but where a specific harm is also of interest. Although testing the safety outcome leads to large sample sizes, this type of trial is appealing because of the potential benefit of the therapy. Second are trials comparing active therapies when the new therapy, while potentially leading to a slightly higher rate of adverse events, is designed to be beneficial. Here, the sample size must be large enough to test both benefit and harm; the informed consent must clearly lay out both outcomes. The third type compares two agents of likely equal benefit in order to test whether the new one is more harmful than one already in use.Rational participants may be unwilling to enter such a trial if the informed consent form is honest.