[Compare new therapies with old, not with a placebo: a plea for revision of the Declaration of Helsinki].

If no therapy is available for a disease and a new therapy may have beneficial effects, a well-designed placebo-controlled randomized trial will not immediately raise ethical questions. Pre-2008 versions of the Helsinki Declaration reflect this. However, the Declaration of 2008 allows placebo-controlled randomized trials even where an established effective therapy is available, providing this is methodologically inevitable and safe for patients. Placebo-controlled trials have important advantages for sponsors: they are easier to perform because fewer patients are required and small improvements are sufficient to show the efficacy of a new therapy. The authors consider both arguments open for interpretation and argue that the current revision of the Declaration of Helsinki should return to its pre-2008 version. They also suggest that, independently of this, IRBs should resume the policy of rejecting protocols that use placebo while withholding an effective treatment.