Correction to This Article
At the International AIDS Conference this week in Mexico City, the CDC came out with new estimates for the rate of HIV incidence in the United States. This story should have said that the rate plateaued with as many as 56,300 new cases annually, according to new CDC estimates--a striking indication that the strategy of targeting testing to only the highest-risk patients and those who directly request it hasn't worked.

HIV Screening: It's Not So Easy

Oral tests for HIV can be a trade-off: better speed and cost but lower accuracy than other tests.
Oral tests for HIV can be a trade-off: better speed and cost but lower accuracy than other tests. (By Tina Fineberg -- Associated Press)
  Enlarge Photo    
By Ishani Ganguli
Special to The Washington Post
Tuesday, August 5, 2008

In the controlled chaos of emergency departments across the country, doctors are working to curb the HIV epidemic, one swab of the gums at a time. The effort stems from guidelines published in 2006 by the Centers for Disease Control and Prevention to extend HIV testing to teens and adults in any health-care setting, regardless of their risk of having the virus. The D.C. Health Department launched its own campaign that year, distributing 80,000 free tests in ERs, community health centers and private doctors' offices.

But there are challenges to actually casting this wider net, as studies are starting to reveal.

Harvard-affiliated researchers were studying testing strategies at the Brigham and Women's Hospital in Boston when they realized that the rapid oral test they used in the ER produced many more false positives than expected.

Over eight months last year, Rochelle Walensky and colleagues administered the OraQuick Advance Rapid HIV-1/2 Antibody Test to 854 patients without known infection, they report this week in the Annals of Internal Medicine. Of the 31 patients with positive results who were given more rigorous follow-up tests, only five were found to be truly HIV-infected. Previous studies from urban centers including Washington have reported similar rates of false positives with OraQuick.

Rather than replacing the test with a more accurate one, the researchers learned to anticipate more false positives and act accordingly. Sacrificing accuracy for speed and cost-effectiveness is a familiar trade-off in screening for diseases such as breast cancer (as many as 90 percent of suspicious findings on mammograms turn out benign) but one that is new to HIV, with its socially charged history.

Although rapid blood-based tests tend to be more accurate, Walensky has found that patients are less comfortable with them. "Because of this [OraQuick] test, way more people accept HIV testing than otherwise might," she said. "But if in fact . . . there are more false positives than anticipated, we need to be prepared."

This means relaying these limitations to patients upfront and following up initial positive results with confirmatory tests, which take hours to weeks as opposed to minutes. (Such follow-up tests are already part of standard protocols.) The findings also point to the need to monitor what ends up in a patient's chart, said Walensky, an infectious disease doctor at Brigham and Women's and Massachusetts General Hospital. (At the Brigham, positive rapid results are not entered into a person's chart without confirmatory tests.)

Although only 13 percent of emergency departments offered the rapid tests in 2006, it's likely that these numbers are now higher. Each hospital has adopted its own combination of tests and testing strategy, said Bernard Branson, a CDC epidemiologist and lead author of the 2006 guidelines. "If we're going to . . . start widely testing in a population where there hasn't been wide testing before, we're going to have to be really careful to have accurate and robust testing," said Christopher Pilcher, an associate professor of medicine in the HIV/AIDS division of the University of California at San Francisco. "There are now a number of companies that are getting close to introducing a new generation of rapid tests," he added.

For decades, the complex social and political history of AIDS has set it apart. HIV activists working to protect their autonomy and privacy have long clashed with public health officials trying to curb a quickly spreading epidemic.

The result has been that HIV testing has been targeted to only the highest-risk patients and those who directly request it. HIV incidence in the United States, which had plateaued at 40,000 new cases annually, reached 56,300 new cases in 2006--a striking indication that this strategy hasn't worked.

But in the past decade, the tension between human rights and public health has been easing as the availability of effective antiretroviral therapies has changed HIV from a death sentence into a chronic disease. And so in 2006, the CDC revised its testing guidelines -- also loosening the requirements for consent and pre-test counseling -- in an effort to nudge HIV testing into the mainstream.

More than 25 percent of the more than 1 million HIV-positive Americans are unaware of their status, according to Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases. With routine testing, they can start treatment earlier, better preserving their health and drastically lowering their chances of passing on the disease.

Even though ERs have traditionally avoided unnecessary testing that might require extensive follow-up, knowing a patient's HIV status in the ER can provide critical insights into their diagnosis and treatment. "An internist may be the ideal person to initiate the HIV testing encounter, and handle any potential false-positive results. [But] many people under the current health-care system don't have that kind of provider," said Pilcher. And these are the Americans most at risk in the first place.

Has the strategy paid off in numbers diagnosed or lives saved? It's too early to tell, said Branson.

"If we screen people and don't get them into care, we've done little to help the patients," said Walensky.

Ishani Ganguli is a student at Harvard Medical School but was not involved in this research.

© 2008 The Washington Post Company