1) IF a person takes their meds religiously
2) AND IF they are under medical observation
3) AND IF that observation shows their viral load has been undetectable (fewer than 40 viral copies per ml of blood) for six months continuously
4) AND IF they don’t have another sexually transmitted infection
THEN there’s no risk that they will pass on HIV in sex.

You can find the full paper in French here, and an unofficial English translation here. The key references: Quinn et al 2000, Castilla et al 2005, Melo et al 2006, Barreiro et al 2006.

Needless to say, there has been a lot of comment about this here, for example, and here, and here. For my money, the most sensible comment comes from Rex Wockner who points out, quite correctly, that the Swiss usually do their homework pretty thoroughly. So they’re probably right, but should that make anyone throw out the condoms with an untested partner? Of course not. Unless you have some preternatural way of telling that all the IFs are in order. Which none of us do.

The reaction from WHO and UNAIDS is predictably mealy-mouthed. They can’t quite bring themselves to say that the Swiss are right, any more than they can bring themselves to say that the risk of HIV transmission in oral sex for anyone who doesn’t have suppurating mouth wounds is basically zero. The fear is that people might interpret the Swiss statement to mean: if you’re on treatment, you’re not infectious. In other words. they’ll forget about the other IFs. It’s a patronising view: people cannot be trusted to digest the information and make informed choices. As in: “This guy tells me he’s on treatment and he’s good about taking his meds, but then I told him I was on the pill, so…. ”

We’ve been caught out by this many times before. Pretend everything is equally risky, and people will go ahead do dumb things across the board. Let people know exactly what the risks are, and they’ll often choose their dumb things more carefully. If we know that an infected person with good adherence to drugs and excellent medical care can have sex without condoms without endangering their partner, shouldn’t we tell them that? As the Swiss paper says: let the HIV negative partner decide if they want to use extra protection.

It is perhaps worth stressing, though, that all the data they reviewed came from studies in heterosexuals. One exception was a paper from San Fransisco suggesting that HIV incidence was falling among gay men since the initiation of treatment.(Porco st al, 2004) which has sadly since been overtaken by events. In other words, we can’t be quite so sure that a viral load of below 40 copies/ml protects against transmission in anal sex, let alone in injecting, both of which are more efficient than vaginal sex.

There’s something else in the Swiss paper that people haven’t been twittering about, but that I think is really interesting. More here.