The problem is that none of us have a very good grasp of the concept of "odds." If we did, we wouldn't risk billions of dollars each year on state lottery tickets, we wouldn't bet on the 40-to-1 shot at the track, and we would never, ever try to fill an inside straight. We know, for example, that the risk of transmitting HIV during unprotected vaginal intercourse is 1 in 500, and we know the risk is even higher for unprotected anal intercourse -- yet some people are apparently willing to risk those odds, because new HIV infections attributable to unprotected sex occur every day in this country.
On the other hand, we know that the risk of transmitting HIV by kissing someone is so small it is almost unquantifiable -- maybe 1 in 2,000,000. Yet we can't quite shake the conviction that kissing might, just might, be risky. Because we don't fully understand the notion of relative risk, some of us assume we can continue to beat very short odds... and some of us worry unduly about very long odds.
The inadequacy of "odds" as a way of explaining relative risk was brought home to all of us last summer, when the Centers for Disease Control reported a case of apparent mouth-to-mouth transmission of HIV. Although this case was a highly unusual, one-of-a-kind event, it got picked up by the mass media -- and generated a kind of mass hysteria. What television commentators often failed to report was just how isolated and exceptional this case actually was. It occurred because both partners, the HIV-positive man and the HIV-negative woman, had undergone dental surgery. He had an exceedingly high viral load at the time; she had mouth lesions that were large enough to make her susceptible to infection by this route; and there was extensive mouth-to-mouth contact. The odds that this same set of circumstances would ever arise again are, well, astronomical. You've got a better chance of winning the lottery or being hit by lightning -- and we all know how long those odds are.
This isolated and unusual case does have two lessons to teach us, however. The first is that poor oral hygiene is a potential mode of infection -- and HIV-positive individuals with bleeding mouth sores shouldn't kiss anyone on the mouth. The second is that we should avoid talking about odds, and we shouldn't let one-of-a-kind cases worry us inordinately. The vast preponderance of all new HIV infections are transmitted through unprotected vaginal or anal intercourse, and this is where we need to focus our prevention efforts. What we need to remember is this: Kiss anyone you want, anywhere you want. But never, ever have unprotected sex.
Paul A. Volberding, M.D., is Editor-in-Chief of AIDS Care and AIDS Program Director at San Francisco General Hospital.