A few weeks ago, someone asked about drug use and PrEP, asking if it was more common for people on PrEP to be using drugs. At the time, Website A had just started offering PrEP as an HIV status, and it seemed a little early to start building any models of PrEP and drug use. Now, after three weeks, I can start to make some reasonable assertions.
The top-line is no. Looking at the six US cities dataset, there is no significant difference between men on PrEP and the site as a whole. (x^2(1, n=23594)=.86, p=.35) Now, that being said, there's a bit of a caveat here. A lot of people (27.5%) simply do not report a drug use status, and that also is true for HIV status (18.1%). In this analysis, I lumped non-responders into the "open to drug use" category. Moving them into "no drugs," the difference becomes even smaller (x^2(1, n=23594)=.09, p=.76).
However, this is one of the challenges of this particular approach to the data. Since non-response is an option (and, as I remember, the default) option, there is no clear interpretation of what that means. In this case and at this time, either way I slice it, it doesn't matter, but looking at some of the trend lines, missing data may become a real issue.
There's also some useful discussion about how PrEP is being positioned right now. My perception is that the people I know on PrEP are mostly middle/upper-middle class men with insurance and doing it as a form of personal responsibility. PrEP is probably a really good idea as a risk-reduction measure for "partiers," but that runs the risk of re-enforcing an image of PrEP as a party drug. ("Good guys use condoms, bad boys use PrEP.") Right now, there just isn't enough data. I'll re-visit this in a few weeks and see how the numbers have changed.