Sensational headlines circled the globe last Friday, June 30.
"S.F. HIV Rate Surges; alarming incidence of new infections
raises fears of scourge to come," exclaimed the San Francisco
"New HIV infections soar in San Francisco," declared the
Reuters new service.
"The percentage of new HIV cases in The City almost tripled
from 1997 to 1999," said the Examiner.
Epidemiology oftentimes seems to be more of an art form than a
science, at least in this town. And when it comes to the human
immunodeficiency virus, San Francisco can seem like ground zero
in the art of going out on a limb with disease forecasting.
San Francisco health officials were working overtime feeding
juicy quotes and leaking unpublished data to the media last
week � timed in advance of media frenzy surrounding next week's
International AIDS Conference in Durban, South Africa, and also
in the midst of heated negotiations in Washington, D.C. where
Congress is poised to cut Ryan White CARE Act funding to San
Francisco due to a dropped AIDS caseload.
The crux of the stories was that DPH had reviewed a dozen or so
indicators of the HIV infection rate in the city's gay
community � in particular looking at the number of men who took
the HIV antibody test at anonymous test sites � and found that
the rate of new infections had doubled, or even tripled, in the
last few years.
DPH officials were quoted saying that the estimated number of
new infections among gay and bisexual men in San Francisco had
gone up from 500 to almost 900.
Now, the same officials are singing a different tune.
"The 900 number is not an official DPH number � The comparison
to sub-Saharan Africa is unfortunate," Dr. Willi McFarland, an
epidemiologist with DPH told the Bay Area Reporter, referring
to a quote he gave to the Chronicle: "These are sub-Saharan
African levels of transmission."
While all the media reports quoted local officials stating that
a chief factor feeding the higher infection rate was the fact
that people with AIDS are living longer, creating a large pool
of infected people, who in turn are infecting others, McFarland
and DPH director Mitch Katz were later both instead emphasizing
other factors, including the size of the gay population itself.
"The reason why the  number of new infections is a less
reliable data point is that it's very sensitive to estimates in
the size of the population," Katz told the B.A.R. "And while I
have not yet reviewed the documentation that supports the 900
new infections a year, what I do know is that part of the
increase from 500 is due to a better estimate in the size of
the gay male population which increases its size, and also it
may be that the community itself is growing in size. Obviously
the more gay men or the more people at risk in general that you
have, the higher the number of new infections."
No vote was taken at a closed May 24 meeting of AIDS prevention
experts in San Francisco organized by DPH and the Center for
AIDS Prevention Studies of the University of California at San
Francisco. The meeting was convened in preparation for a
"community consensus" meeting to be held this fall.
The stated purpose of the meeting was to consider revising the
estimated number of new HIV infections taking place in San
Francisco each year, based on data collected from a variety of
It is no secret among prevention experts in San Francisco that
there has been an increase in new infections in the gay and
bisexual community, and a number of factors are feeding into
this, chiefly "barebacking," or unprotected anal sex, behavior
likely compounded by a concurrent epidemic of "party drugs"
including crystal methamphetamine in the gay community.
Prevention experts also point to a perception that new
combination antiviral therapies have made the disease less
Using the new detuned ELISA HIV antibody test, which helps
researchers determine whether an HIV infection has occurred
before or after 129 days prior to the drawing of the sample of
blood, researchers applied a complicated formula to come up
with their estimates of the number of new infections that were
leaked to the media sources and led to last week's headlines.
In an unreferenced abstract, among thousands to be published in
conjunction with the Durban conference next week, McFarland
reported a 2.1 percent infection rate at anonymous test sites
in 1996; a low of 1.3 percent in 1997; 2.3 percent in 1998; 3.4
percent in 1999.
None of the media reports provided the raw numbers used to come
up with these annualized percentages.
In 1996, McFarland told the B.A.R., of 3,505 gay and bisexual
men taking the test, there were 25 recent infections.
In 1997, of 3,114 gay and bisexual men, 14 were recently
In 1998�and here's the big jump leading to all the headlines�of
3,291 taking the test, 30 were recently infected.
In 1999, there was an unexplained drop in the number of gay and
bisexual men taking the test at anonymous test sites, but
according to McFarland, there were 1,995 who took the test, and
of these, 26 had been infected within the last 129 days.
Yet Jim Dilley, executive director of the UCSF AIDS Health
Project, which operates the anonymous test sites in San
Francisco under contract with DPH, noted about 9,000 people
took the HIV tests in 1999. Of those, 127 tested positive,
Dilley said, though some of these may have been older
infections, as the detuned ELISA test was not used.
At press time, DPH researchers had failed to explain why their
numbers are apparently a subset of the total number taking the
test at anonymous test sites.
"It's unfortunate that it got out before we were able to say as
a group that this is our best estimate of the number of new
infections," said Steven Tierney, director of HIV prevention at
DPH. "I do think that the number has clearly gone up."
"Obviously we feel very strongly that there is evidence that
there is increased HIV incidence in San Francisco which
parallels the reports of increased unsafe sex behavior which
we've made previously," agreed Katz.
Some of DPH's usual critics were quick to jump on the
apparently sloppy, incomplete data leaked from DPH to the major
media outlets last week.
"Instead of telling the public how many HIV tests have been
administered in San Francisco and the consequent number of
positive, negative and indeterminate results, UCSF and the DPH
have concocted an elaborate mathematical method using numerous
'indicators' and employing novel detuned ELISA tests unapproved
by the Food and Drug Administration to concoct HIV increases on
paper," said David Pasquarelli of ACT UP/San Francisco.
"These increases are not real and the study's methods,
indicators, and testing technology have never before been
employed for this purpose nor verified by researchers
unaffiliated with the AIDS industry," he added.