Some participants felt that PrEP used in isolation was ‘too much’ of a risk (if only approximately 70% effective). Respondent 1: That’s not enough, that’s not enough, exactly. Respondent 2: Its seventy-two

per cent effective but then there’s still that twenty-eight per cent. (HIV-negative MSM, FG) Participants also expressed a wider scepticism of PrEP, which emerged in two ways. GSI-IX Some thought that the advice to use condoms with PrEP indicated a continued scientific uncertainty: “I don’t know if this [will] work because since they said if you use it you can still use condom, that means they are not sure as well” (HIV-negative African woman). Second, some expressed scepticism in relation to the variability in reported efficacy rates in relation to adherence. As a result, there was a reluctance to trust these, unless the information came from a recognised, trusted and reliable source. R1: There’s so much out there right, on the internet an’ everything else right, if it came up on the news right? The logistic news on BBC one that this pill will prevent so…—like a hundred per cent—that’s when I would believe it would work. I [would not] trust any other pill… Question: So it has to be a hundred per cent? R2: ‘cause there’s

a lot a…no like—anyone can put something on the internet, …You’ve got to use like certain websites, you’ve got to look like .org, .gov, those type o’ websites. You’ve just done one that’s like www.medicine.co.uk it’s not a…it’s not like a legitimate website it could just be anyone postin’ stuff up there. Q: Well I suppose in the States it’s from the Food and Drug Administration so it’s

got that stamp on it. R1: Yes, but you [don't know] if they’re been a hundred per cent accurate on their job or just been lazy an’ the fact that we’re brave enough to take it to human trials right? (HIV-negative MSM, FG) In spite of being told that the FDA had approved the drug, these FG participants did not recognise the organisation or trust it as a reliable source, and were therefore sceptical about its statements and/or endorsement. Other participants were less concerned about PrEP not being 100% efficacious because they imagined AV-951 that it would be used in addition to condoms. However, participants expressed confusion or uncertainty in knowing how to interpret efficacy rates and their impact on risk reduction in this particular context. For example, in discussing the effectiveness of condoms in relation to other prevention strategies, one participant calculated the protection offered by PrEP, condoms and TasP with a hypothetical HIV-positive sexual partner. Assuming that PrEP was 90% effective at reducing HIV transmission, he explained: R: but there’s obviously still a 10% risk but, as you said, there’s the same risk with condoms.