San Francisco serosorting may explain odd HIV data

STDs have risen, but not new HIV infections

About a year ago, it seemed to be a pretty safe assumption that HIV incidence was on the rise in San Francisco due to documented increases in sexually transmitted diseases (STDs) among high-risk populations, including men who have sex with men (MSM) who use the Internet to find anonymous sexual partners.

And when it comes to the HIV epidemic, where San Francisco goes, so does the rest of the nation, eventually, according to conventional wisdom.

But when the same public health officials and researchers who drew attention to the alarming increases in syphilis among MSM studied actual HIV incidence among this population, using the detuned assay test, they found exactly what they did not expect: HIV incidence was leveling off at the same time STDs were on the rise.

"The first thing we noticed was this rapid spike in syphilis among gay men in San Francisco, and we’d already seen this rise in gonorrhea and other indicators," says Willi McFarland, MD, PhD, director of HIV/AIDS Statistics and Epidemiology for the San Francisco Department of Public Health.

"So when this last indicator of syphilis went up, we thought this would make acceleration of the epidemic even worse, because STDs have a synergetic effect on HIV infection — people infected with both are more infectious to their partners," he says. Working with an investigator from the Centers for Disease Control and Prevention (CDC) in Atlanta, San Francisco public health officials measured the incidence of new infections among people seen at municipal STD clinics and elsewhere, and they found that the rate of incidence has been leveling off over the past several years, McFarland says.

Public health officials and investigators also found that safe sex behavior increased since 1998 among the population of HIV-negative MSM when they reported having sex with an HIV-positive partner.1 The most logical explanation for this counterintuitive result is that while risky sexual behavior is on the rise, the MSM having this sex are doing so with partners whom they know have the same serostatus, investigators concluded.

In other words, MSM were serosorting themselves. At least one hypothesis is that they are serosorting through the Internet — the very venue that makes it more likely they’ll increase their sexual risk behaviors.

"If persons who are the same HIV serostatus are in the same networks so that positives are having unprotected sex with positives and negatives with negatives, then you could see an increase in syphilis transmission without an increase in HIV transmission," McFarland notes. "So from there we looked for other data that would collaborate or refute this hypothesis."

One source of data investigators used was the behavioral data collected by the Stop AIDS Project in San Francisco. It recorded respondents’ serostatus and asked if they knew their partners’ serostatus. The results showed that from 2001 on, HIV- positive MSM were saying they did know their partner’s serostatus, McFarland says.

Likewise, HIV-negative MSM also reported that they knew their partner’s serostatus in the more recent years.

"Then we looked at counseling and testing data and the question of whether they knowingly had sex with someone who was HIV-positive," says McFarland. These data showed that people who were positive had fewer partners they knew to be negative, and so these additional data began to support the hypothesis that MSM were serosorting, he explains. "They could have high levels of unprotected sex so long as people were selecting partners of the same serostatus."

Although public health officials lack hard data regarding this outcome, it’s a possibility that the leveling off of HIV cases could be the result of the city’s prevention-for-positives campaign. This began in 2001 with a media saturation, featuring celebrity advertisements and physicians trained to provide prevention in HIV care clinics, he says.

"So we may have seen community-level evidence that it’s turning around, since 2001 was when it began to turn around," McFarland points out. Additional research is needed to further test the hypothesis and to discover how serosorting has been working in recent years, but one possibility seems to be the increased use of the Internet and chat rooms to find sexual partners, he says.

"It’s a semianonymous way of disclosing your status," McFarland adds. "People put in their serostatus and say they’re looking for someone with the same serostatus, and we know that partnering on the Internet is becoming more common."

This strategy of what some researchers call "negotiated safety" combined with HIV testing can work if it’s adhered to, but that’s one of the big concerns McFarland and other public health officials have.

"Part of me worries that this may be a little optimistic," he says. "It’s a lot to ask that people know their own serostatus and ask partners about them." Serosorting depends on people knowing their status if they’re positive and remaining negative if they’re not, and then being honest about their status when negotiating sexual encounters, McFarland notes.


1. Truong HM, McFarland W, Kellogg T, Dilley J. Increases in serosorting’ may prevent further expansion of the HIV epidemic among MSM in San Francisco. Presented at the 11th Conference on Retroviruses and Opportunistic Infections. San Francisco; February 2004. Abstract 843.