Ivinson replies—I am grateful for Jon Cohen's clarification. It is indeed an important issue and as such clarity is important. Cohen agrees that a consensus was reached but disagrees with respect to what that consensus applied to. Ruth Macklin has told Nature Medicine that despite the absence of formal minutes, immediately following the meeting "an official procedural agreement [was] recorded". This read: "... Differences of opinion on whether the standards of counselling (sic) and treatment for trial participants who contract HIV should be those of the sponsoring country or those of the host country were resolved by consensus in favour (sic) of 'the highest practically attainable standard' (which could be arrived at through a fully collaborative decision of the sponsoring and host partners)."

However, in the weeks following the meeting, Macklin tells us, discussions between a few of the meeting's participants (likely including some of the "dissenters" to whom Cohen refers) resulted in this paragraph being amended to read:

"... However, participants in the Geneva meeting did not succeed in reaching consensus on what level of treatment should be provided to participants in a vaccine trial who become infected during the trial. Some people at the meeting argued that the 'best proven therapy' should be provided, in compliance with the current CIOMS international guidelines. Others at the meeting argued in favor of what they believe to be a more realistic standard—'the highest practically attainable' level of treatment. No amount of further discussion brought these two different opinions closer together. Rather than leave the matter entirely unresolved, participants agreed to set aside for the present any attempt to set a substantive standard, and agreed instead on a 'procedural' solution. That solution is to leave decisions about the level of care to the host country in a vaccine trial, those decisions to be made in full collaboration with sponsors of the trial. The result is that different countries will almost certainly decide upon different standards to apply to their own situation. However, in no case may the host country and the sponsor decide on a level of care that is lower than the 'highest practically attainable' level."

Wordsmithing aside, what this tells us is that whereas a change in wording, to date discussed by only a minority of the attendees, has seemingly facilitated broader agreement, the intention remains that trial participants be guaranteed access to the "highest practically attainable standard of care" in the host country while leaving open the issue of whether they should be offered "the best proven treatment". It is this consensus (procedural or otherwise) that I reported.

See “Clarifying AIDS vaccine trial guidelines” by Jon Cohen