www.andrewsullivan.com Daily Dish AN OPEN MIND: Update on the stats behind the HIV stats. Can't find them anywhere on the web. One of the researchers has invited me to go to San Francisco to sit down with him and go over them, but as yet, hasn't shown me the analysis or directed me to where it can be found. A respected AIDS journalist emails to say that the timing of the announcement is suspicious: "It's been my observation that alarming reports of HIV surges tend to pop up about a week or two before either a big AIDS Conference or a significant deadline in the federal budget process. Just so happens that the Retrovirus Conference opens in Chicago this weekend and, of course, one presumes Bush is putting his first budget together now." The last report of this kind from San Francisco calculated HIV-transmission rates in part by analysing whether there are employment discrimination protection for gays, domestic partnership provisions, and so on, in any given area. Quite how these would help one figure out HIV transmission rates is beyond me, and I look forward to seeing the logic. But the b.s. detector bell is ringing off the hook. - 1/29/2001 05:31:30 PM THE HIV NON-SURGE: The San Francisco Public Health Department study which alleges a sharp increase in HIV infection in that city among gay men has just been released. I'm grateful to the researchers for getting me a copy. What that means, of course, is that all those news stories written last week were written by reporters with no access to the actual data! Go figure. What's my read? There are six mini-studies of men seeking sex with men (gay and bisexual) that lead the researchers to their conclusion of soaring infection rates. I have to say that I'm unconvinced. It's possible, but nothing in the report proves it. Only one study has a reasonably representative sample of young gay men between the ages of 21 and 29, randomly selected from neighborhoods where AIDS infection is highest in the city. But the sample size varies considerably over time. In 1994 - 95, for example, there were 660 men in the sample. In 1998 - 99, there were 322. In 1993 - 94, there were a mere 261. Unsurprisingly, the smallest samples show the greatest variation from the mean. In 1993, at the height of the safe-sex era, HIV-transmission is calculated as 2.7 percent. The next year, the rate plummeted to 1.4 percent. The researchers provide no reason for this shift, and it's almost certainly a function of sample size. Last year's incidence rate of 1.8 percent is certainly within the realm of the average of the last several years which is around 1.4 percent. The "doubling" scare is more a function of the freak low figure for 1997 - 98, for which again there is no explanation. And the numbers themselves are tiny - a mere three HIV seroconversions in 1997; a mere 9 in 1998 - 99. You'd be statistically right if you declared a tripling in HIV rates from such numbers. But you'd also be really silly to draw such a conclusion from a total of 12 cases. - 1/31/2001 06:33:47 PM WAIT, THERE'S MORE: The other studies are just as weak. One looks at men who get anonymous HIV testing at public clinics. This, as the study's authors delicately put it, "may not be representative of the community" as a whole. No kidding. The sample sizes are also all over the place. The number seeking testing in 1996 was 3,488. In 1998 that number was 2,910. In 1999, the number fell again to 1,826 - a vast difference in sample size, and one that should set off alarm bells in drawing easy conclusions from one year to the next. The inference that transmission rates have increased is based on a new sort of biological analysis that can tell whether HIV infection is recent or not. Sure enough, more recently infected people are showing up at STD clinics. But what do these numbers tell us? They tell us that many fewer men are being tested than in the past and many more of them have been infected recently. One obvious conclusion from this is not an increase in HIV-transmission. It is that most gay men in San Francisco have already been tested for HIV and know their status. But those who may recently have had an unsafe incident or accident are the ones most likely to get tested. This is certainly just as plausible an explanation for the increase in incidence of HIV and a shrinking number of testers as an alleged outbreak of new HIV transmission. In the only study where the sample size seems to be increasing - for gay men showing up at STD clinics - the rate of HIV transmission is basically stable. In 1995, with a sample of 634 men, 5.8 percent showed up HIV-positive. In 1999, with a sample of 1,071 men, the rate had dropped to 4.7 percent. I'm sorry, but that isn't an explosion. To my mind, the only truly worrying sign in all the data is the number of rectal gonorrhea cases. This is a real number, since, unlike HIV, gonorrhea infections must be recorded by law. These are clearly rising, suggesting a declining use of condoms. The number has risen from 97 cases in 1995 to 160 cases in 1999 and 93 cases in the first six months of 2000. But this may not be evidence of HIV-transmission. Why? Because two HIV-positive men could transmit gonorrhea through unsafe sex, but not transmit HIV - because they both already have it. Another study shows that the rate of STDs among people with AIDS has doubled in five years, suggesting that unsafe sex may indeed be primarily an issue between HIV-positive people, not a means of HIV-transmission. The bottom line is: the case for a doubling in HIV-rates in San Francisco is simply not proven. If it's happening, these studies don't prove it. And journalists who trumpet the fact would do well to read the studies before broadcasting the executive summary. - 1/31/2001 06:30:56 PM THE GONORRHEA NON-SURGE: The sometimes crazy but often accurate AIDS activist Michael Petrelis emails to let me know that the one dark lining in my analysis of the San Francisco Health Dept HIV report may actually be a little lighter. I reported that the data shows a doubling of rectal gonorrhea among gay men in San Francisco in five years - a deeply worrying trend. The report didn't show the size of the testing samples so I took it as a stable and reliable indicator. It turns out it isn't. In their 1999 report, the San Francisco Department of Public Health explained the sampling background for gonorrhea testing: "In response to the city-wide increases seen in 1995, we began testing more MSM [men who have sex with men] seen at City Clinic for rectal gonorrhea (since infections may be asymptomatic): the number of tests increased from 542 in 1995 to 1285 in 1999 while the number of male visits per year remained approximately 11,000. This increase in screening could be expected to increase the number of reported cases. The number of cases with symptoms did not change between 1996 and 1999, which indicates that the increase in cases may be due to the increased number of tests." So the doubling of gonorrhea cases, by any reasonable assessment, is completely explained by the more than doubling of the number of tests. Quite why that wasn't made clear in the current report is beyond me. - 2/1/2001 04:20:07 PM
Feb. 1, 2001 Dear friends: Yippee! The SF DPH has finally produced some of its alleged HIV transmission data for Sullivan, and instead of buying their voodoo epidemiology, he has interpreted the data with deep skepticism. Overall, I believe this piece from Sullivan is excellent and will provide fodder for activists in SF like myself to demand even more HIV stats from DPH and, of course, more accountability from the researchers peddling their allegations of surges. However, I feel three of Sullivan's/DPH's points need addressing, so I have excerpted the points and responded to them.
No. 1 MP: During this period the SF DPH, and its private partners, were waging media campaigns about alleged increases in HIV and STDs among gay men. In 1999 the DPH claimed gay men were spreading syphilis when meeting sexual contacts from web chat rooms. Full pages ads appeared in local gay papers prominently featuring an enormous walking, ticking time-bomb symbolizing syphilis. It is my theory that media campaigns demonizing gays and gay sex, and using fear to scare us into more testing, actually may have driven men away from testing.
No .2 MP: I believe he is referring to the detuned ELISA test here. This detuned ELISA test has not been approved by the FDA for the purposes it is used by SF DPH. In the context of all the other questionable HIV data, I think skepticism is warranted when reading results from detuned ELISA tests conducted by DPH.
No. 3 MP: The most recent STD annual report for SF, for 1999, states the following. "The majority of these cases are reported from City Clinic: the proportion of cases seen at the clinic increased from 67 percent in 1996 to 74 percent in 1999. In response to the city-wide increases seen in 1995, we began testing more MSM seen at City Clinic for rectal gonorrhea (since infections may be asymptomatic): the number of tests increased from 542 in 1995 to 1285 in 1999 while the number of male visits per year remained approximately 1 1,000. This increase in screening could be expected to increase the number of reported cases. The number of cases with symptoms did not change between 1996 and 1999, which indicates that the increase in cases may be due to the increased number of tests. Since men with proctitis would most likely seek medical care, the trend in proctitis cases may be a better marker for trends in unprotected anal intercourse among MSM," reported the 1999 STD report. More screening seems to equal more cases. (Source: 1999 DPH STD report, DPH Home Page). Now that Sullivan has brought some much needed sanity and attention to the alleged HIV rate increases among gays in SF, it is time for Mike Shriver of Mayor Willie Brown's office to finally post his HIV data on the web so more of us can read and analyze it. In closing, I say to Andrew - thanks for getting the data and realizing the problems with SF's alleged HIV surges. --Michael Petrelis
Michael Petrelis
 TORONTO
tel/fax:(416) 406-HEAL
INDEX
Search our web site (400+ pages)
|
