randomly stumbled upon this :nkoWhaaaa: https://www.thelancet.com/journals/lanhiv/article/PIIS2352-3018(24)00217-0/fulltext
-
randomly stumbled upon this :nkoWhaaaa: https://www.thelancet.com/journals/lanhiv/article/PIIS2352-3018(24)00217-0/fulltext
-
(TL;DR: a new PrEP medication was trialled which can be administered as an injection twice a year, instead of a daily pill, and it had 100 percent efficacy)
-
Mx Amber Alex (she/it)replied to lis 🔜 MRMCD last edited by
@lis isn't that the one they'd like to sell for like $50,000 that would still turn a profit if sold at $40?
-
@lis Wow.
The 25th International AIDS Conference, held July 22–26 in Munich, Germany, was dominated by one topic: the results of the PURPOSE 1 trial of lenacapavir 6-monthly injections for the prevention of HIV. The novel approach to pre-exposure prophylaxis (PrEP) had remarkable results. Compared with the expected background HIV incidence, and in comparison to daily oral prevention with tenofovir-based drugs, lenacapavir given every 6 months was 100% effective in preventing HIV. There were no infections at all in 2134 participants. Equally remarkable was the study itself, which was done in a population of adolescent girls and young women in South Africa and Uganda. Confirmatory results from another trial, PURPOSE 2, which is ongoing in populations of men who have sex with men, are required for Gilead's product to be licensed for use, but by all accounts this is a mere formality, expected later this year. A twice-yearly injection to prevent HIV could be a game changer—a phrase that we do not throw around lightly in The Lancet HIV.
[...]
-
@lis Technically, only 'higher than 99.97%' efficacy, and that's before factoring in the possible sampling error.
-
Riley S. Faelanreplied to Mx Amber Alex (she/it) last edited by
-
@[email protected] hence why i said “had“, not “has”. as in “had 100% efficacy in the trial”
-
@lis Since you're measuring at a full-patient granularity, you have to set the boundaries of your measurement error at 'zero plus or minus half failures'. The study had 2134 participants, and of them, the measured failure rate was less than half out of every 2134, which is some 0.02343%.
You don't get to collapse your margin of error to zero merely because the event was in the past. This is not probability; it's uncertainty. These are different things.