GLAA testifies on HIV surveillance
GAY AND LESBIAN ACTIVISTS ALLIANCE OF WASHINGTON
Fighting for Equal Rights Since 1971
P. O. Box 75265
Washington, D.C. 20013
Testimony on Bill 16-0116
"HIV Unique Identifier System Amendment Act of 2005"
Before the Committee on Health
May 26, 2005
Chairman Catania, Members of the Committee, and Fellow Citizens:
My name is Bob Summersgill. I am the Treasurer of the Gay & Lesbian Activists Alliance of Washington (GLAA), the nation's oldest continuously active gay and lesbian rights organization.
We have been working with the Mayor, Council, Department of Health and other community organizations for more than 7 years to develop and implement the best possible HIV epidemiological surveillance system that we can devise. Ideally, a surveillance system will be accurate, confidential, and minimize distrust from the public.
Bill 16-0116, "HIV Unique Identifier System Amendment Act of 2005," was introduced by Councilmembers Sharon Ambrose, David Catania, Jack Evans, Adrian Fenty, Jim Graham, and Phil Mendelson, and co-sponsored by Kwame Brown, Vincent Gray, and Vincent Orange to correct the problems with the implementation of the unique identifier system. The current requirement of a partial social security number (SSN) is a problem for a variety of reasons, most significantly discouraging people from getting tested for HIV.
The Centers for Disease Control and Prevention (CDC) has pressured all states and D.C. to track HIV infections in addition to AIDS, an advanced stage of HIV disease. The CDC suggested two systems as being acceptable for HIV tracking: names reporting, as we do with AIDS and many other diseases, and a unique identifier system. The Institute of Medicine issued a report in 2001, "No Time to Lose," [http://www4.nas.edu/onpi/webextra.nsf/web/web_extra_hiv], strongly supporting unique identifiers over names reporting to protect privacy and not discourage testing. Both systems are only as accurate as the number of people who get tested. If you do not get tested, you cannot be counted.
The CDC estimates that 25% of all people with HIV have not been tested and do not know their status. A study by Kaiser Permanente found 40-42% of people who test positive for HIV wait until they get sick, and miss years of life-extending treatment. In a survey released in April by the CDC of 5,600 men who have sex with men aged between 15 and 29, some 77 percent of those who tested positive for HIV did not know they were infected with the virus. Among African American men in the survey, 90 percent did not know that they were infected. Only 10% to12% of persons aged 18-64 years in the United States reported being tested for HIV during the preceding 12 months according to the CDC. [http://www.cdc.gov/mmwr/preview/mmwrhtml/mm5347a4.htm]
Knowing their HIV status helps infected people receive and maintain appropriate medical care and reduce their risk for transmitting infection. We must not further discourage people from getting tested.
History of the Unique Identifier System in the District
In 1999, The DC Council, at GLAA's urging, recommended to the Mayor that a unique identifier system be established to protect the confidentiality of people who test positive, and to maintain public trust in the health care system. Unlike most other diseases, HIV disease carries great stigma, and extra protections for privacy are warranted. Mr. Ronald Lewis, then the Director of the HIV/AIDS Administration (HAA), led the fight to oppose unique identifiers, supporting names reporting instead. Mayor Williams wisely chose unique identifiers, but unfortunately left Mr. Lewis in charge of implementation. Mr. Lewis has subsequently been removed from the Department of Health for gross incompetence. He was removed before the outright fraud and theft of funds at HAA came to light.
The original unique identifier plan was crafted without HAA input. However, HAA staff insisted, over objections from the working group, that the proposed identifier be modified to include the last four digits of the patient's social security number.
Scaring People Away from Testing
The current HIV unique identifier surveillance system raises safety and security concerns for anyone mistrustful of the government. This mistrust will be particularly acute among recent immigrants, especially undocumented immigrants who risk deportation even without being infected with HIV.
By asking for a partial Social Security Number (SSN) and country of origin, the current system makes it easy to identify immigrants. Considering that the United States bans immigration by anyone with HIV and the post-9/11 environment, non-citizens will feel uncomfortable with these questions and may be deterred from getting tested. Citizenship and legal or illegal status must not be barriers to health care. HIV is unconcerned with these issues and will spread between people regardless of their immigration status.
Catalina Sol of La Clinica del Pueblo described in the City Paper a client survey showing that 85 percent of Latino clients would avoid getting tested if they knew their names would be reported to the government.
HAA officials argued in 2001 that the last 4 digits of the SSN are not required for immigrants. They then stated that the code "9999" would be used in place of the last 4 digits of the SSN when there was no SSN. This is a distinction without a difference. It still highlights aliens in the code. Further, it is an admission that SSN is not needed. If it is optional then we can remove it altogether.
Washington, DC's immigrant community is large and extremely diverse. No educational system is likely to counteract the damage done by the questions in the surveillance system itself. An educational component is likely to miss those who are least literate and least proficient in English. The only way for the HIV surveillance system to not discourage testing is by removing the SSN altogether.
Use of SSN is Unreliable and Unneeded
Use of the partial SSN defeats the stated purpose of pursuing the policy in the first place, namely gathering of reliable information. Arguments to HAA noting the inherent problem of including the partial social security in terms of immigrants without a number, people misreporting the number, and general mistrust of the government collecting the information were ignored.
Information provided by HAA to the working group from the CDC explained that the failures of the unique identifier systems in Maryland and Texas were directly related to the inclusion of a partial social security number which were only 50% complete. Two evaluations of unique identifier systems conducted by the CDC in Los Angeles and New Jersey, found the social security numbers to be available in less than 20% of all cases reported. Massachusetts found SSNs on just 50% of their records. The CDC requires 85% reporting completeness for the system to meet minimum performance standards. First names, which HAA staff prohibited from being included, are 97.2 - 99.9% complete. Use of the first one or two letters of a person's first name — as is done with last name — would distinguish between two people born on the same day with similar last names.
The unique identifier code uses the date of birth; the first two letters of the last name; the number of letters in the last name; sex; race; and zip code. This provides an enormous number of possible unique identifiers.
I work with a membership database professionally. My database is not representative of, or as diverse as, the District's population, but it represents a size comparable to the predicted number of people with HIV. Comprised mainly of engineers across the country, only 16% are women and 86% are white. I generated unique identifiers from my database for comparison, leaving out zip codes, which would significantly skew the results. Of the 10,920 unique identifiers from my database, there were only 3 duplicate codes. That is only a .006 % duplication rate. Far below the 5% required by the CDC. Well within the CDC's requirements. The District's more diverse population should likewise find a very, very small duplication rate.
In the event that HAA identifies two codes as belonging to two different people, as would happen with the other collected data — transgender status, country of origin, mode of exposure, zip code, and occupation — a simple a, b, c, etc. can distinguish the codes without compromising privacy or confidence in the system.
Confidentiality of Data
While HAA reassures us that no AIDS surveillance information has ever been improperly disclosed, the InTowner newspaper has published photographs of medical records from a former D.C. health clinic blowing around the street. Improper release of medical records, including HIV surveillance information is routinely reported around the country. With the revelations about the widespread incompetence and fraud that has characterized HAA, there is an even greater concern about the proper care and confidentiality of those records. Again, the best way to insure confidentiality is to remove the more easily recognized data.
Additional Provisions of the Bill
The legislation also creates the tracking mechanism for transgender people. While this is not a requirement of the federal government, it is extremely important in the District. The only survey of transgender people to have been conducted by the Department of Health found that 25% of transgender people in the District are HIV+. A significant percentage of transgender people engage in sex-work in large part because of illegal discrimination based on their appearance. Transgender people also have much greater difficulty accessing basic health care. Consequently, the transgender community is believed to be a significant transmission vector for HIV. However, we are not currently collecting the epidemiological data that we need to assess the depth and scope of the problem. This bill requires that transgender status be collected with the other information.
The bill's prohibition on contact tracing reaffirms the current HAA policy. We agree with this policy because of the draconian nature and violations of medical confidentiality that has plagued other contact tracing programs. We do strongly support voluntary disclosure of HIV status to intimate partners.
Epidemiological surveys must do no harm. The bill will keep that harm from continuing. We urge you to pass this bill.
Thank you, I am available for any questions that you may have.
