The Guess Who: UNAIDS and its small problem with statistics -- and the truth

UNAIDS, the United Nations coordinating organization to combat AIDS, admitted it overestimated the size of the world-wide HIV-AIDS epidemic and will drastically slash the reported number of people suffering from the disease.

The Geneva-based organization calls these reductions a “revision of estimates” made after reviewing its research and monitoring methodologies. The adjustments, however, are substantial, bringing into question a decades worth of AIDS advocacy.

UNAIDS admits it overestimated the number of people with new infections at 2.5 million, a cut of more than 40 percent from last year’s calculation of 4.3 million. More startling is the reduction of last year’s assessment that 40 million people were infected with the disease. This statistic now reads 33 million.

Also, according to the group, the number of people dying each year from HIV-AIDS dropped from an estimated 2.9 million in 2006 to a revised 2.1 million.

UNAIDS officials claim their reassessment of the disease’s reach in India led to the greatest reduction in its prevalence and mortality rates. The advocacy group also admitted that “important revisions of estimates” took place in Angola, Kenya, Nigeria and Zimbabwe.

Alarming issues
UNAIDS statistics have long been questioned by researchers and health care advocates. Today’s Washington Post provides a takeaway message from the group’s announcement:

Having millions fewer people with a lethal contagious disease is good news. Some researchers, however, contend that persistent overestimates in the widely quoted U.N. reports have long skewed funding decisions and obscured potential lessons about how to slow the spread of HIV. Critics have also said that U.N. officials overstated the extent of the epidemic to help gather political and financial support for combating AIDS.

"There was a tendency toward alarmism, and that fit perhaps a certain fundraising agenda," said Helen Epstein, author of "The Invisible Cure: Africa, the West, and the Fight Against AIDS." "I hope these new numbers will help refocus the response in a more pragmatic way."

Over the past decade, global spending on AIDS has grown by a factor of 30, reaching as much as $10 billion a year.

But in its role in tracking the spread of the epidemic and recommending strategies to combat it, UNAIDS has drawn criticism in recent years from Epstein and others who have accused it of being politicized and not scientifically rigorous.

For years, UNAIDS reports have portrayed an epidemic that threatened to burst beyond its epicenter in southern Africa to generate widespread illness and death in other countries. In China alone, one report warned, there would be 10 million infections -- up from 1 million in 2002 -- by the end of the decade.

Peter Piot, the Belgian doctor who formed UNAIDS in 1995 recently lashed back at critics who, he says, worry too much about his groups’ estimates, which diverts attention from solving the AIDS riddle.

In a recent speech, he claimed:

"The challenge is ... complicated by the mixed messages circulating around the world," Piot said. "Denialist statements such as that 'UNAIDS overestimates the size of the epidemic,' and 'There's too much money for AIDS' don't help, not least because there's clearly a massive gap between what's needed and what's available."

Methodological questions persist
The scientific community has long held misgivings about the methodology UNAIDS uses to gather its prevalence estimates. Throughout much of the past decade, the group primarily relied on sentinel surveys, which monitor the prevalence of HIV in women who attend: 1) prenatal clinics; or, 2) clinics for the treatment of sexually transmitted diseases.

By only tabulating AIDS rates of sexually-active patients who visit health clinics, UNAIDS researchers admitted they wanted to track the people or groups most likely contribute to the expansion of the disease. This will provide them with a leg up when combating a concentrated malady like HIV-AIDS. Another advantage of this system is that these surveys could be taken anonymously because the clinics are settings where blood is often drawn.

Critics point out that UNAIDS long ignored the common practice of household surveys, which make use of large survey sizes in discrete geographic regions. (One of the problems with STD clinics is they most often reside in large cities, although much of Africa’s population still lives in rural areas.) Because household-based surveys are more thorough and time consuming, they are not susceptible to the wild variations of the sentinel surveys.

An example of inconsistent results recently took place in Burkina Faso. UNAIDS, using 2001 sentinel studies from various health clinics around Burkina Faso, reported the national AIDS rate stood at 6.5 percent, a much higher rate than anyone predicted. The next year, UN researchers conducted a similar study and saw the rate fall to 4.2 percent. After repeating the study once again in 2003, UNAIDS reported the average tumbled down to 2.3 percent.

About the same time, a different group of researchers published the results from a household-based survey which estimated the national prevalence rate to be 1.8 percent. Another benefit of household-based surveys is they provide researchers with important geographical information on the spread of AIDS. The researchers found that only 1.3 percent (on average) of rural people are infected, along with 3.6 percent of people living in urban areas. Yet, even these numbers can be deceiving: In the predominately Muslim rural north, for example, only .1 percent of the population is afflicted with HIV-AIDS, while up to 3.7 percent of people in the far southern part of the country. The study also found that prevalence rate for AIDS in the capital Ouagadougou to be as high as 4.2 percent.

Looking deeper
While UNAIDS should be applauded for acknowledging its shoddy science, I still have serious reservations with the group’s announcement.

Most importantly, UNAIDS kept with its sentinel surveys long after the knew these methods provided exaggerated claims of the spread of the disease. According to its own publication, the 2007 AIDS Epidemic Update, the group continually overestimated AIDS prevalence rates in 25 different African countries. (These rates include all adults between 15 and 49 years.)

Some of the “overestimates” were minimal, like the UNAIDS .8 percent rate of prevalence for 2004 in Senegal while a household survey found a .7 percent prevalence rate.

Some figures were staggering.

• In Botswana, researchers reported the AIDS prevalence rate in 2004 to be 25.2 percent, one of the highest in the world. That same year UNAIDS claimed the rate to be 38 percent. In 2005, the UNAIDS said the rate fell to 24.1 percent;
  

• In Central African Republic, UNAIDS announced the 2004 prevalence rate to be 13.5 percent. A 2006 household survey predicted the rate to be 6.2 percent. In 2006 UNAIDS reduced its estimate to 10.7 percent;

• In South Africa, researchers claim that in 2005, some 16.1 percent of the adult population was infected with the disease. UNAIDS put that number at 20.9 percent in 2004 and 18.8 percent in 2006.

Reviewing this data, one can see the group had plenty of time to revise and update its methodology. A cynic may even point out that as researchers began to lob criticism at UNAIDS’ findings, the group slowly began downsizing its estimates. If that's true, why did it take until 2006 for this to begin happening?

This is a group responsible for a $2.6 billion budget (in 2006-07), an organization that refers to itself as the UN’s coordinating body in the fight against AIDS. One would think it would want the most thorough, most geographic and gender specific information available regarding the disease it purports to fight.

The problem with these high estimates is that UNAIDS also views itself as a major portal in AIDS fundraising. Of course, the group never referred to its questionable statistics when it had its jar out for donations.

That didn't stop officials from making wild claims. In 2005, UNAIDS released a report claiming a “looming funding gap of $18 billion for H.I.V./AIDS in developing nations between 2005 and 2007.” This shortfall, the group said, could hurt access to prevention programs against HIV transmission and would jeopardize the antiretroviral treatment to 75 percent of individuals clinically qualified to receive these medicines in 2008 and slow down the progress made in providing support for AIDS orphans and other children affected by the disease.

The short version: If we don't get you're money, people are going to die.

Data-free research
Regular readers of this blog (my mother and some guy in Minneapolis, Minnesota) will no doubt understand my obligation to puncture what I feel is time-honored, but reality-free stereotypes of Africa. One of the most hallowed of those myths is that the continent of Africa doubles as a hospice for AIDS patients. This legend, perpetrated in part by the media of all stripes and fashionable personalities of all shades, provides for the rest of us a perfect symbol of Africa: A continent full of emergencies like AIDS, the slow, silent killer, producer of death, leaving only widows and orphans in its wake. Yet the subtext remains the same: Africans are too dumb, too foolish, too conservative and too horny to understand what the rest of the world already knows: AIDS kills! Stay away from it.

From Africa to the rest of world, the message clearly remains the same: We are constantly in trouble and we could use some more help.

Of course, UNAIDS, a long perpetrator of this myth for its private gain, has come out of the darkness and (somewhat) begged our forgiveness. In the coming weeks and months we will judge the contrition of this most high-paid and well oiled group. I guess they could start by merely telling the truth and quit being so damned arrogant.

Update: Added link to Washington Post story