Syndicate content

Add new comment

Submitted by Ryan Cooper on

I agree. Although I would alter the order of reactions above and give a slightly different angle to the arguments. 1) 'clinical equipoise' argument should be applied equally to any prospective study that does not treat 100% of eligible population, regardless of the assignment rules (random, other sort of algorithm). 2) If there is excess demand of eligible subjects to receive a particular treatment at a given point in time, the ethical question linked to 'clinical equipoise' has to do with the decision of leaving eligible subjects without treatment and not about the mechanism or criteria chosen to select Treatment and Control groups. RCT vs. non RCT has to do with the later. So any design (RCT or not) that leaves subjects without treatment in presente of 'clinical equipoise' would be un ethical. 3) But, in most situations there is excess demand, and thus regardless of the desirability of leaving people without treatment, it simply is not technically or financially possible. In this situation, as David reminds us, RCT can be the most fair way of assigning treatment, given that after applying all reasonable targeting criteria, a lottery gives everyone the same chance. 4) There is the first issue David mentions about costs and opportunity costs. I have 2 questions though: a) Say I have 3 equally costly competing treatments that aim to meet a common social objective. If I can apply 'clinical equipoise' to one of them and not to the other 2 would this not imply that the equipoise one is better than the others? 2) Assuming a situation with 3 alternatives where I can apply 'clinical equipoise', does the concept apply to certainty of the direction of impact or also to the magnitud of impact. If it was only the former, I think your cost effectiveness and alternative cost argument is stronger. If it was also de later, I we could think of cost effectiveness comparisons and ranking without any experiment given that I would know the magnitud of the cost, the direction of impact, and also the magnitude of impact.

Having said this, a last question: In reality is it common to face 'clinical equipoise' in development economics? I agree its fair to consider this concept while prioritizing questions to answer, but does this not happen implicitly in academia?