Date: Thu, 6 Jul 1995 10:48:57 -0400
From: Gabo3@aol.com

New York Newsday - Thursday, July 6, 1995
THE OMINOUS ODDS OF UNPROTECTED SEX
by Gabriel Rotello

	New York - Walt Odets, a California psychologist and well known critic of
AIDS prevention strategies in the gay community, often contends that "oral
sex is an extremely safe activity - about half as dangerous as driving a
car." In an article in last month's Harper's, he cited a study that seemed to
support his claim. That study, he wrote, "followed 6,704 [gay] men in the San
Francisco City Clinic Cohort and found that in five years there were
precisely two infections that might be attributed to oral sex. This
represents a .0003 percent risk of HIV transmission over five years." 
	A letter to the editor in the subsequent edition of Harper's pointed out
that, mathematically speaking, 2 out of 6,704 yields a .03 percent risk, not
.0003 percent, and Odets has accepted the correction. But that still leaves
the impression that in five years only two men out of a cohort of 6,704
became infected via oral sex. 
	That figure sounds quite low, and its publication in Harper's has had a
dramatic effect. Gay men nationwide have cited it as long-awaited proof that
oral sex is extremely safe. The problem is, the statistic is extremely
misleading. 
	For starters, the San Francisco City Clinic Cohort (SFCCC) only followed 675
gay men, not 6,704, for five years. It's true that over six thousand men
participated in an SFCCC hepatitis B study between 1978 and 1980. But when
those men were re-contacted in the mid 1980s and asked if they would
participate in an annual study of AIDS infection rates, the majority were
already HIV-positive. Only 675 men were ultimately studied at yearly
intervals beginning in the mid-1980s. 
	In 1991, epidemiologist Alan Lifson published a study of two of those
participants who, his research team determined, had become infected via oral
sex in 1988. But, Lifson says, "Our study should never be interpreted to say
that these are the only two men in the cohort in whom oral transmission
occurred. Our report was simply a case study of two men in whom oral
transmission occurred, to indicate that it does, in fact, occur." 
	Odets wrote to me that his estimates were not derived from Lifson's report,
and that additional information supported his "evaluation of the entire
subject of oral sex." 
	Nonetheless, a third oral infection in the SFCCC was confirmed after 1988,
and Paul O'Malley, the director of the SFCCC, emphasizes that there may have
been more. The problem, says O'Malley, is a familiar one in the
epidemiological assessment of oral risk - the so-called "masking effect."
Even if a man who becomes infected reports a single instance of anal
intercourse and hundreds of oral episodes, researchers attribute the
infection to the anal exposure, since anal sex is clearly the riskiest sexual
activity for gay men. This can "mask" the true extent of oral infection.
	One solution would be to study a population that engages only in oral sex,
but that was hardly the case with the SFCCC. Of the 675 men in that cohort,
very few practiced oral sex alone. "During the year those [first] two men
became orally infected," says O'Malley, "only 34 men reported receptive oral
sex with ejaculation and no anal sex. Of those 34 participants, two, or about
5 percent, seroconverted." 	
	That's not quite as alarming as it sounds, since over the entire five year
history of the cohort, only three oral infections were confirmed among the 49
men who became infected. Still, that means that approximately 6 percent of
the total seroconverters were orally infected. The best way to look at the
data, suggests O'Malley, is that at least three (and possibly more) out of
675 men in the cohort, or about .5 percent, were orally infected. You may
think that's high or low, but either way it's a far cry from 2 out of 6,704.

(Gabriel's column appears every Thursday in New York Newsday. His email
address is gabo3@aol.com)