Excuse the intrusion into the thread, but a couple of thoughts.
From the trial you cited, I would highlight the second clause:
"An HIV test is not required for entry on this protocol, but is required if the patient is perceived to be at risk. "
I know it's just one example, but HCV infection would be a MAJOR known risk factor for HIV, so the oncology analogy doesn't hold to the HVC trials.
I also see a serious flaw in BT's conclusion that “so it’s quite possible (indeed 40% likely) that many of the HCV patients will be co-infected [with HIV].
The part I have trouble with is the 40%. While it's true that there is a 40% co-infection rate with HIV in all HCV patients, it does not follow that that there would be a 40% infection rate in the people in this trial. Given the alarming increases in the rates of co-infection, the added complexity of managing co-infected patients, and the VERY high rate of co-infection (40%), it is unfathomable that very many patients could progress through the healthcare system with HCV all the way to a clinical trial without having an HIV test.
In the trial protocol you have to exclude known cases. What you have left are those who tested negative, and those who haven't been tested. This untested group would be the group that the 40% best applies to (but even this requires assumptions. It is probable that they weren't tested for HIV because they clinician perceived their risk to be lower for some reason, and their risk if less than 40%). This "untested for HIV" group could be as low as 10% of HCV patients (couldnt' find a number, just guessing).
40% of (pick your % that haven't been tested--10% say) is 4% of the enrollees. Add to that people who acquired HIV subsequent to being tested, and a few who might lie about their HIV status, and you have way under 40% in this trial.
Also, as has been pointed out, if the investigators don't do a pre-test of HIV, enrolling HIV positives yields very little scientifically. They'll get no pre-post data to learn anything about HIV infection and the co-infection will just complicate analysis of bavi's affect on HCV.
