Illustration by Bloomberg View
Illustration by Bloomberg View

Anyone trying to follow developments in HIV science has reason to feel bewildered of late. The news from HIV-prevention research has been fast-paced and confusing. What worked in one trial seems to fail in the next.

At the same time, the mixed results give policy makers clues about how to prioritize efforts to prevent AIDS. By highlighting the limitations of approaches that require human discipline, the findings show the value of methods that don’t.

On Nov. 25, researchers announced that they had discontinued a portion of a trial testing the ability of a vaginal gel to prevent HIV infection. Preliminary data showed that the gel, containing the AIDS treatment drug tenofovir, was ineffective. In a separate trial that concluded last year, the same gel had reduced the risk of HIV infection in female participants by 39 percent.

How could trials of the same product get conflicting results? A likely reason is that the two groups of women tested were given different instructions. In the first trial, the subjects were told to apply the gel before and after sex, and in the second they were told to apply it every day. Daily use may have proved too inconvenient, unpleasant or hard to remember. So the gel may work, as suggested by the first study, but may have been used too inconsistently to show an effect in the second one.

The adherence factor also may explain differing outcomes of studies that tested whether oral HIV treatments could prevent infection. A trial of established couples in which one partner was HIV-positive showed that the other partner’s risk of infection declined if he or she took a daily oral dose of tenofovir or tenofovir combined with emtricitabine. Subsequent studies of women at risk of HIV infection found no such benefit from either regimen.

Those women, however, weren’t necessarily in stable relationships. Arguably, the women with steady partners in the first trial led more predictable lives and may have taken their prescribed pills more regularly.

With any drug, adherence to a dosing regimen makes a difference in effectiveness. With chronic diseases, people on average take their prescribed medicine only about half the time. In the first study testing HIV pills as prevention, among homosexual men, blood tests indicated that only about half of participants actually took the drugs. Deploying these medicines for HIV prevention may be similar to deploying condoms: They’re effective, but not when people don’t use them.

What this tells us about combating HIV is that the less prevention methods rely on compliance, the better. New infections could be significantly reduced if efforts were focused on strategies that require minimal human adherence.

Two existing methods are of particular value. The first is male circumcision. About a million men have been circumcised for purposes of HIV prevention since 2007, after studies showed that the procedure reduced a man’s risk of HIV infection by a female partner by about 60 percent. To achieve 80 percent coverage in the 14 African countries with the highest HIV incidence would mean circumcising 20 million more men by 2015. The procedure, of course, requires the participant’s compliance, but once the cut has healed, male circumcision affords a lifelong benefit without further action.

The second strategy is preventing HIV transmission from a mother to her child in the womb or during delivery. Today, of the 2.6 million new HIV infections each year, 1 in 7 occurs this way. Yet this method of transmission is preventable by administering HIV drugs to the mother and newborn. The most effective regimens involve multiple doses, but even a single dose of nevirapine given to the mother at the onset of labor and to the baby after delivery can halve the risk of transmission.

Maximizing circumcision and mother-and-child programs would save lives. Sadly, neither of these strategies directly protects women, who account for 60 percent of HIV infections in sub-Saharan Africa. If a third study of tenofovir gel, applied vaginally before and after sex, shows efficacy in 2014, the product probably will be licensed, though it may have limited impact on the pandemic.

Given concerns about adherence, prevention methods for women that minimize the need for constant dosing are all the more necessary. Two promising strategies in development are a vaginal ring that would slowly release an HIV-killing drug and AIDS vaccines.

HIV is almost entirely preventable. That it continues to infect millions of people yearly speaks to its nature: It is spread mainly through sex, a basic, human act that tends to leave discipline behind. To defeat HIV, we need methods that don’t depend on our weak points.

To contact the Bloomberg View editorial board: view@bloomberg.net.