Control Groups Without Best Standard Care

rough notes from a meeting with Michael Selgelid, 4 October 2007:

testing a new drug against best available treatment e.g. AZT versus placebo for mother-child transmission in poor countries

outcry because the control group got placebo rather than “best standard care”, which means what they would have got in a rich country

They were studying the question they actually wanted an answer to

but this conflicted with the Declaration of Helsinki which says that they had to get the best standard of care

but the standard of care in rich countries wasn’t available in poor countries so they weren’t deprived of anything

So some say that we should change the Declaration of Helsinki to say that control groups should get the standard of care available locally.

Michael’s ethical solution: make the treatment arm much bigger so that most people in the study are benefiting, even though still some are not.

There’s also the option of using historical controls instead of concurrent controls.

Does this issue arise in the case of TB? If so it should be flagged.

—-

interesting example: microbicide. If best standard of care is used in the control group, the study becomes enormously expensive, because the best standard of care is sex education; and if education is used then the infection rate plummets, so the trial has to become very large.