Petrelis Files

Arthur Sulzberger Jr.
Publisher, asulz@nytimes.com
The New York Times
New York, NY

Dear Mr. Sulzberger:

The New York Times reported on July 1, 2000, that San Francisco was in the midst of a dramatic rise of new HIV infections, with future new infections climbing from 600 up to 900 annually, possibly sparking a wave of thousands of new transmissions in other U.S. cities. [1]

The basis for this dire prediction, called a “sub-Saharan level of transmission” by Dr. Willi McFarland in a June 30, 2000, story in the San Francisco Chronicle, was use of STARHS, Serologic Testing Algorithm for Recent HIV Seroconversion, which was not approved at the time by the Food and Drug Administration, and still hasn't received the agency's approval. [2, 3]

The ominous alarms in the SF Chronicle article were echoed breathlessly in a gloomy July 1, 2000, Times article, written by Lawrence K. Altman, M.D., which reported on the investigational use of STARHS by the Department of Public Health in San Francisco, which “apparently is the only city that has applied the testing strategy in a series of years so as to detect trends” of new HIV infections.

With the fourth year of the anniversary of Altman’s pessimistic story soon upon us, I chose to spend part of my May Day this year pulling all the pertinent documents in my possession from the San Francisco health department pertaining to HIV antibody infection rates for the past several years, and examining them again with one question in mind: What is the current HIV rate of infection in San Francisco?

I have four sets of data to consider in answering the question. Some of the publicly available data from SF DPH is confusing and sometimes wildly inconsistent, but it is the best HIV epidemiology and surveillance that money from the Centers for Disease Control and Prevention can buy. All I’ve done is try to connect the many dots of conflicting HIV data from an astutely political local health agency, one that flexes considerable media muscle.

These are the four subsets of data from the SF DPH that I analyzed:

(A) Recent STARHS testing statistics from the city’s primary testing clinic.

Starting in January 2002, the health department’s monthly STD summary began to include the number of STARHS antibody tests administered at the municipal clinic.

“We have added a table showing HIV testing at City Clinic [the main HIV testing site in the city,] and the number of patients who test positive. Since December 2001, patients have been offered a second test that shows HIV infections within the past six months if their HIV test is positive. HIV-positive patients who are also positive for the second test [STARHS,] are listed in Table 2 as ‘recent positives,’” explained an editorial note in the January 2002 STD report.

The year-end figures for 2002 show 2432 HIV antibody tests performed, of which 103 were positive, and of those one hundred three positive results, 40 were classified as recent positives, using the STARHS method.

Numbers for 2003 reveal that 2776 antibody tests were conducted, with 109 positive for HIV, and of that number, 41 were found to be recent positives using STARHS.

So the number of tests rose, as did total number of positive test results, along with recent positives.

Still, the number of STARHS-identified recent positives is statistically stable, and in the middle double-digit range.

The STARHS data was omitted from the January 2004 monthly STD summary, without explanation, and is no longer a part of the summary, which is, of course, an unfortunate development for members of the public who wish to track such important data. [4, 5]

(B) The city’s annual HIV/AIDS epidemiology summaries.

1998

Since California law does not require names reporting of people testing HIV antibody positive, HIV infection rates for San Francisco are an educated guess by epidemiologists.

This means the annual HIV/AIDS summaries don’t provide precise HIV incidence rates.

The 1998 report says, based on many factors, that, “An estimated 15,249 San Francisco are living with HIV and 499 will acquire HIV each year.”

No numbers of antibody tests performed or HIV-positive test results are mentioned.

However, in order to arrive at this estimate, DPH’s report took into account the status of sodomy, anti-discrimination and domestic partnership laws, dating back to 1990. What effect, if any, these laws may have on determining San Francisco’s HIV transmission rate is unclear from the report.

1999

This year’s summary contained some HIV incidence data, gathered through blinded testing at the city’s primary site for STD testing.

“Between 1989 and 1998 the prevalence of HIV infection among MSM (including MSM IDU) declined from 55% in 1989 to 20% in 1998.

“However, during the same period the incidence of HIV infection among MSM fluctuated (range 4.3 – 10.4% per year) but remained statistically stable,” the summary noted.

“However, the relatively stable rate of new HIV infections suggests that current prevention strategies may be effective at holding HIV transmission rates stable but not effective enough to result in a further decrease in the rate of new infection,” researchers cautioned.

2000

The summary for this year documented that “the incidence and prevalence of HIV increased from 1997 to 1998 and again in 1999. This last year the incidence of HIV among MSM was 4.2% per year and the prevalence was 5.0%.”

A chart accompanying that statement illustrates that the incidence rate for 1999 was 4.2%, up from 2.2% the previous year.

Data was collected from anonymous HIV testing sites throughout the city.

2001

The report highlights the fact that this was the year DPH’s anonymous testing facilities began using STARHS.

“In general, there has been an increase in HIV prevalence and incidence among MSM seeking anonymous HIV testing in recent years. This increase is corroborated by recent rises in sexual risk behavior and sexually transmitted diseases in the city. Data from 2000 (available from January to September only) suggest that HIV incidence may have leveled off from that observed in 1999. However, the estimate, 3.9% per year, is still statistically significantly higher than 1996 to 1998. Findings remain to be confirmed with complete data for 2000 and 2001,” said the report.

Interesting that the researchers mention that the 3.9% figure for 2000 is indeed numerically higher than from 1996 to 1998, while they fail to note how it’s also a drop from 1999, when the figure was 4.2%.

It's almost as if the researchers cannot acknowledge any suggestion of a decrease, since attention is only called to cherry-picked data showing a rise.

2002

Comparable data from anonymous HIV testing sites is missing from the most current annual summary. [6, 7, 8, 9]


(C) HIV tests and positive results statistics, from the annual CDC-funded Counseling, Testing, Referral and Partner Counseling Referral Services report, which are culled from anonymous and confidential testing sites in the city.

Two caveats regarding the data are worth keeping in mind. The number of tests performed does not equal number of people of tested, and positive test results do not represent unduplicated or new HIV antibody infections.

1999

Number of tests = 25,871
No. HIV positive = 950
% HIV positive = 3.0%

2000

Number of tests = 27,267
No. HIV positive = 913
% HIV positive = 3.3%

2001

Number of tests = 25,006
No. HIV positive = 748
% HIV positive = 3.0%

2002

Number of tests = 19,091
No. HIV positive = 699
% HIV positive = 3.7%

[10, 11, 12]


(D) A 2004 syphilis and HIV study conducted by SF DPH, funded by CDC.

An abstract from the Trends in Primary and Secondary Syphilis and HIV Seroincidence among Men Who Have Sex With Men in San Francisco, 1998-2002, study was presented at the February 2004 Retroviruses and Opportunistic Infections Conference, by Dr. K. Buchacz, a CDC researcher.

These are the key points:

“Background: Syphilis facilitates the acquisition and transmission of HIV infection. To explore whether the current syphilis epidemic has been associated with increases in HIV incidence in San Francisco, we examined trends in HIV incidence in men who have sex with men (MSM) in 2 HIV testing populations and rates of primary and secondary syphilis among an estimated 50,782 MSM in the city from 1998 through 2002.

“Methods: We reviewed city sexually transmitted disease (STD) morbidity data. We determined HIV incidence by applying a sensitive/less sensitive testing algorithm to HIV-seropositive blood samples of MSM seeking HIV testing at the San Francisco City STD Clinic (SFCC) and at anonymous testing sites of the AIDS Health Project (AHP).

“Results: Primary and secondary syphilis rates increased more than 1000% from 1998 to 2002 (Table). HIV incidence was highest in 1999 and then tended to decline through 2002 at both sites (c2 test for trend, p = 0.13 for AHP, p = 0.36 for SFCC). At SFCC, however, HIV incidence was elevated in 2002 compared with 2001 (p = 0.12).

“Conclusions: To date, we detected no marked increases in HIV incidence among MSM in 2 HIV-testing populations, concurrent with the syphilis outbreak in San Francisco. However, changes over time in characteristics and volume of MSM testing in these 2 settings may distort possible inferences about HIV incidence trends among all MSM in San Francisco. Integrated HIV/STD prevention programs for MSM in San Francisco are needed to stop the spread of syphilis and potential increases in HIV infections.” [13]

What impresses me most about the conclusions is the remarkable detection of “no marked increases” among the gay community, over a four-year period, in a study that involved over 50,000 men. That the researchers say “potential increases in HIV infections” more than hints at how the alarming predictions reported in the New York Times almost four years ago have not been borne out.

I think it’s time for the New York Times to investigate and report on the current state of apparently leveling or declining HIV transmissions in San Francisco.

Regards,
Michael Petrelis

Sources:
1. New York Times story
2. San Francisco Chronicle article
3. CDC RFP using STARHS
4. S.F. DPH monthly STD summary
5. S.F. DPH monthly STD summary
6. S.F. DPH annual HIV/AIDS report, 1999
7. S.F. DPH annual HIV/AIDS report, 2000
8. S.F. DPH annual HIV/AIDS report, 2001
9. S.F. DPH annual HIV/AIDS report, 2002
10. CTR/PCRS report, 1999
11. CTR/PCRS, 2000
12. CTR/PCRS, 2002
13. CDC/S.F. DPH study on syphilis and HIV infection rates