San Francisco is trying a new tactic to fight AIDS. Health workers are aggressively testing people for HIV and then immediately putting those who test positive on potent antiretroviral drugs.
Known as "test and treat," the approach relies on the fact that taking HIV drugs dramatically reduces a person's risk of transmitting the virus to others. As more and more people are put on medication, the epidemic theoretically should fizzle out.
Test and treat sounds good on paper, but some doctors and policymakers have doubts about its feasibility on a large scale.
An analysis just published in the journal PLOSOne questions whether test and treat is the best strategy in developing countries, such as South Africa, where getting drugs to many people is difficult and the rise of drug resistance is a potential problem.
The results bolster support for the current HIV guidelines set by the World Health Organization, which recommend giving antiretroviral medicines to people whose immune systems are already compromised by the virus.
In 2009, scientists at the WHO developed a mathematical model to look at how test and treat could alter the course of the HIV epidemic in South Africa. The model made a provocative prediction. If all 50 million South Africans could be screened for HIV and the estimated 6 million who are infected given antiretroviral drugs, the number of new HIV cases each year would plummet. The virus would be on its way to elimination.
Mathematical biologist Sally Blower, who co-authored the current study, says this model is too simplistic. It didn't take into account the potential need for expensive, second-line drugs if resistant HIV strains emerge, and it underestimates how long people really stay on the medicines.
When she adds these factors to the model, it predicts that test and treat will take 30 years longer and cost significantly more than originally thought. Plus, focusing treatment on only those who are sick appears cheaper in the long run — and almost as effective at stopping the epidemic in South Africa — as giving drugs to everyone infected.
"Getting 5 million more people treatment, that's kind of a fantasyland," Blower tells Shots. And, she says, it would be tough to sustain. "When you put people on medication, you need to have the resources to keep them on it," she says. Otherwise, resistance can emerge, and it becomes an ethical issue.
Epidemiologist Brian Williams, who helped develop the original WHO model, agrees that sustainability and commitment are key issues. "The worst possibility is to promise people the drugs and then have to take them away," Williams tells Shots.
But he doesn't think this is a reason to dismiss test and treat. "Eventually everybody [who is infected] is going to need the drugs. Why not start their treatment right away," he says. "When you get cancer, you don't wait until it metastasizes to get drugs."
Williams, who now works at South African Centre for Epidemiological Modelling and Analysis, also says it isn't clear whether drugs resistance will be a problem under test and treat. "Experimental evidence suggests that antiretroviral treatment causes drug resistance to go down."
Such questions can't be answered with theoretical models. Treatment strategies need to be evaluated experimentally with small projects, such as the one in San Francisco.
"Each year we are gaining experience and data from countries where treatment scale-up is progressing," Dr. Andrew Ball, a physician at WHO's HIV/AIDS Department, told Shots in an email.
WHO is using the results from these projects and many modeling experiments, to develop new guidelines for antiretroviral drugs. It plans to publish these guidelines in 2013.
In the end, Ball says, eliminating HIV will probably require many strategies in combination with more access to antiretroviral drugs, including more needle exchange programs, condoms and circumcision.
"The issues raised by the new paper are quite valid, and we need to look at them," Ball says. "But the most important information comes from the hard research."