iPhone app iPad app Android phone app Android tablet app More

Featuring fresh takes and real-time analysis from HuffPost's signature lineup of contributors
John R. Talbott

GET UPDATES FROM John R. Talbott
 

Monumental AIDS Breakthrough -- Or Is It?

Posted: 07/21/10 11:05 PM ET

Today, there are over 5,000 new stories listed on Google News from media sources around the world announcing a groundbreaking breakthrough in the prevention of AIDS in Africa. Here is a sample of some of the headlines.

"WHO, UN Hail 'Groundbreaking' Potential of AIDS Gel" ~ Voice of America

"Groundbreaking' gel halves HIV infection rates" ~ ABC Online

"Gel to cut HIV-infections an 'exciting discovery'" ~ BBC News

"Vaginal gel 'slashes HIV risks' in South African tests" ~ BBC News

"Vaginal Gel May Prevent HIV" ~ WebMD

"Vaginal gel cuts HIV infections as much as 54 percent, trial shows" ~ Washington Post

"Anti-HIV gel is declared breakthrough for women" ~ USA Today

"S.Africa eager to see speedy roll-out of AIDS gel" ~ Reuters Africa


As these headlines suggest, and the underlying stories report, a vaginal gel applied by women prior to sexual encounters has been found in a double blind field test conducted in South Africa to reduce AIDS infection rates by 39%. Upon first hearing this, it would be natural to conclude that use of this gel would reduce reported cases of AIDS in Africa by 39%; or, if you were a woman living in Africa you might conclude that use of this gel might dramatically help you from getting infected in the future. You would not be alone in thinking this was a groundbreaking discovery as the announcement of these test results brought a standing ovation from the audience of scientists and NGO'ers at the International AIDS Conference in Vienna yesterday.

But, do we really understand the significance of this test's actual reported results? As we all know, statistics can be used in very creative ways that sometimes misleads the reader as to the effectiveness of a new drug or procedure.

It turns out that the test did result in 39% fewer infections, per year. The annual incidence rate for the treated group was 5.6 % more infections per year, while the control group that was administered a placebo experienced an annual rate of new infections of 9.1%. 5.6% is indeed 39% lower than 9.1%, in percentage terms, but sometimes measuring things in percentage terms can be deceiving.

In fact, when one looks behind the percentages at the actual numbers of people tested in the study the story is much less convincing. This was a very small test with only 843 women completing the study and they were evenly split between a treated group and a placebo, or control group that received no treatment. The actual number of new infections then were 38 women in the treated group and 60 women in the control group in a little over a year. You can see that there was a difference of only 22 women infected between the two groups, a very small number, especially relative to the 22 million of Africans who are infected with HIV/AIDS. If as few as ten less women in the control group had not contracted the disease, the study would have reported a minor 24 % decline in incidence and the entire story would have lost its newsworthiness.

Second, even with these reported lower rates of incidence of 39%, this improvement is an annual figure. The study does not say that over your lifetime you will be 39% less likely to get AIDS with this treatment, only that at the end of one year, you will have a 39% better chance of having not contracted it.

If you look over a woman's entire sexually active lifetime of, say twenty years, say age 15 to age 35 in Africa, even the treated women in this study in this highly infected AIDS environment would end up with over 70% of their members becoming infected. The good news is that the treated women do better than the untreated, the bad news is that 5.6% of them are still getting infected every year and this compounds to infection rates of over 70% of the entire population over a sexually active lifetime. Yes, the non-treated will see infection rates of over 85% over their lifetimes, but is this really as big a difference as the news headlines and news stories suggest. It is sort of like saying that smoking cigarettes is 39% safer per year than smoking crack, but what difference does that make if both end up killing you in the end.

Finally, for those of you who have not understood what moral hazard is when discussed in the financial crisis literature, here is a beautiful example. South Africa has announced they do not want to wait for more and bigger studies or for FDA approval, they want to immediately pursue distributing the gel to their female population as soon as possible. But, have they thought this through? What if as a result of the women in South Africa using this gel before intercourse, their boyfriends or husbands decide it is not necessary to use a condom. Condoms have been found to be very strong preventative measures against AIDS transmission, so if condom usage declined by just 30% it would completely wipe out any reported benefit associated with this new gel.

UNAIDS and the WHO are anxious to report some success in their fight against AIDS given that their attempts at finding a vaccine so far have failed and their donors are getting impatient. To date, they are betting on this gel for women and circumcision for men as tools to prevent AIDS transmission. But, their circumcision studies suffer the same problem with statistics that this gel study has, a 50% reported reduction in annual transmission rates with circumcision does not mean you will be 50% less likely to get the illness over your lifetime, just 50% less likely to be ill at the end of the first year. Because people have many sexual encounters over many years, such reductions in annual infection rates will do little to reduce the incidence of AIDS in Africa over a lifetime. In addition, the circumcision field studies are suspect because for obvious reasons it was impossible to conduct double blind experiments when circumcision is the treatment being examined, men have a way of knowing if their genitals have been cut and so may change behavior accordingly.

I came to the conclusion that much of the AIDS in Africa was due to the numerous and informal sexual affairs common in Southern African nations, found both among married and unmarried men, especially with a much smaller highly infected group of very sexually active women who trade sexual favors for material goods and money. I write about this explanation in an academic article published here.

Because the infection rates in many Southern African nations have reached epidemic proportions with more than 20% to 25% of all adults becoming infected, I also concluded that these highly infected nations should institute universal, compulsory, but confidential testing of all of their citizens over the age of 12. While you might argue that patient rights are violated by such compulsory testing, this minor social cost is more than outweighed by the social benefit of stopping an indiscriminate killer in the community. Once people know their status, they can act accordingly, or anti-viral drugs can be taken to dramatically lower the transmission rates and hopefully, over time, completely eradicate this scourge of death on the continent of Africa. Action needs to be taken to eradicate this killer, not just treat its victims or attempt to control its spread.

John R. Talbott is the bestselling author of eight books on economics and politics that have accurately detailed and predicted the causes and devastating effects of the current financial crisis. In 2007, Talbott authored two articles on achieving AIDS prevention and the need for confidential and compulsory universal testing in the most infected countries after having spent three months living in AIDS ravaged Botswana.

 
 
 
 
 
  • Comments
  • 65
  • Pending Comments
  • 0
  • View FAQ
Comments are closed for this entry
View All
Favorites
Recency  | 
Popularity
Page: 1 2  Next ›  Last »  (2 total)
HUFFPOST SUPER USER
provgrays1
06:55 AM on 07/23/2010
Why do we never read of truly meaningful treatments or cures about any disease during the last
50 years?
HUFFPOST SUPER USER
Hugh7
04:43 AM on 07/23/2010
In three trials, they circumcised a total of 5,400 men and left a similar group intact. After less than two years, 64 of the circumcised men and 137 of the intact men had HIV. That difference, 73 circumcised men who didn't get HIV, is the whole basis of the claim that "More than 4 million new HIV infections could be prevented". It is known that studies cut short "for benefit" as these all were, are more likely to show benefit.

327 circumcised men, and a similar number of non-circumcised men, left the trials, their HIV status unknown. That is easily enough for the 73-man difference to be non-significant. Circumcised men who found they were HIV+ would leave because the trial had let them down. Non-circumcised men would leave because they had decided they didn't want to get circumcised - perhaps after talking to the circumcised men.

Not only were the trials not double blinded, they were not placebo controlled either. The true gold standard of clinical trials is both. Everybody involved badly wanted circumcision to be effective.

In at least six African countries, more of the circumcised men have HIV than the non-circumcised, according to the National Health and Demographic Surveys. In Malaysia, where 60% of the population is Muslim (and virtually all Muslim men are circumcised and hardly anyone else), 72% of new HIV infections are of Muslims. That certainly needs to be explained before pressing ahead with mass circumcision campaigns.
photo
HUFFPOST SUPER USER
raker
05:03 PM on 07/22/2010
Based on what I've read, the biggest obstacles to controlling the spread of HIV in Africa are cultural and religious practices. I've been told that HIV-positive men seek virgins for intercourse thinking this will rid them of the virus. Women are powerless in their sexual relationships, and it is strictly taboo to acknowledge AIDS in the family and the need to take sexual precautions.

The AIDS problem in Africa is a lot bigger than a lack of access to drug treatments. And who knows how much harm has been done by nutty religious Americans telling people not to use condoms.
photo
HUFFPOST SUPER USER
The African Man
05:58 PM on 07/22/2010
That's very racist and offensive. You've been told, you've been told...If you were told you the moon was made of blue cheese, I am sure you will not emphatically quote that with confidence.

(1) Intercourse with virgins has only been reported in certain areas of Africa. Try googling this up. There is no uniform African culture than can be generalizable

(2) There are areas in Africa with lower HIV rates than areas in the US e.g Senegal. Are those not Africans?

Again, this HIV issue especially when it comes to Africa is used to ventilate inner racial prejudices which you cannot easily express regarding cultures in the USA, since stigma against discrimination does not extend to Africans in Africa!
photo
HUFFPOST SUPER USER
raker
06:53 PM on 07/22/2010
You concede that some African men have sex with virgins to rid themselves of the virus while suggesting that this is different from what I said. It is exactly what I said.

Your accusations of racism are baseless. I would argue that people who assume that African life is exactly like American life, only it's maybe hotter and drier there, are the racially insensitive ones.

If the cultural and religious obstacles to stopping HIV aren't named honestly and we don't take steps to change them, the problem will never be controlled. If cultural practices come off badly for it, then so be it. For some people, political sensitivity trumps all, but where HIV is concerned, I vote for frank, blunt honesty and practicality.
03:36 PM on 07/22/2010
Another point not mentioned is that the gel seems to lose effectiveness after 18 months.
photo
HUFFPOST SUPER USER
Axlotl
02:17 PM on 07/22/2010
You have GOT to be kidding. Standard significance in this sort of test is a 5% difference, and you want to ask questions about 39%?
03:39 PM on 07/22/2010
Yes, because you're confusing raw numbers with relative numbers.

Now, I'm certainly for methods that can be shown to reduce risk, even if it is small. After all, even one saved life is better than none. However, we should not confuse the fact that something is working with the idea that it is working well. Again, the numbers are small so a shift of just a few people results in a large percentage shift. And given the high infection rate, a seemingly large change of 39% doesn't translate to that much of a reduction in total infected numbers.

It's a start. It's important.

It isn't the great saviour the hype is making it out to be.
photo
HUFFPOST SUPER USER
johntalbs
Author of Survival Investing
04:47 PM on 07/22/2010
I think it is time to go take that refresher course in statistics. The 5% you refer to means that just randomness could create these results 5% of the time. It has nothing to do with the reported improvement of 39%. As a matter of fact, there have now been 13 studies involving vaginal gels, so statistics says that you would expect that at least one of them would produce what look like good results even if there were no benefit to the treatment, just randomness over so many trials will eventually produce one that appears significant. Similar to monkeys typing the encyclopedia if you give them enough time.
07:56 PM on 07/22/2010
Are you saying there are 13 studies that showed no effect using standard statistical techniques while no studies have shown an effect until now?
photo
HUFFPOST SUPER USER
Axlotl
06:53 AM on 07/25/2010
I didn't make that direct connection, you did. I apologize for not doing the math for you - see, I estimated it in my head, because not to offend, but since I'm IN la(nother) stats course, I'm wagering you're the one who's rusty.

The control group has roughly 14.25% odds of contracting HIV, the other 9%. If you want to get snippy about it, the odds of 22 less infections happening is .8575^22 or P= .034 - less than 5%...which is A) why they went to press with this, B) significant, and C) why *you're wrong. I don't know what made you think you are smarter than the study authors in the first place, prof...

(Ah, right ; you're just a civil engineering B.S. and MBA. Nevermind..!)
02:10 PM on 07/22/2010
Pssst...

Google "House of Numbers" and "Gallo's egg"

Glad to be of service.
This user has chosen to opt out of the Badges program
11:39 AM on 07/22/2010
There will never again be a cure for any disease like there was for polio, because the pharmaceutical industry makes more money by treating rather than curing diseases.
02:18 PM on 07/22/2010
But don't they make money from polio vaccine? I have friends who have been cured of cancer.
03:42 PM on 07/22/2010
Right, because being able to cure a disease and thus having your researchers win the Nobel Prize for Medicine wouldn't attract investors.

This idea that medical professionals are only in it for the money and actually want people to be sick is ludicrous.

By the way: There is no cure for polio. There's a vaccine for it and it's so important that it's a standard one given to everybody in this country (assuming fools like McCarthy don't have their way). Are you seriously claiming having a customer base of every single person isn't worth going after?
photo
HUFFPOST SUPER USER
tinyrainbows
10:21 AM on 07/22/2010
Let's try this study...wear a condom.
03:44 PM on 07/22/2010
And when it fails?

Yes, wearing a condom is a very good way to stop the spread of HIV. You are assuming, however, that a woman can make a man wear one.

Hint: Don't be disingenuous and reply, "Well, she shouldn't have sex with him if he isn't going to wear one." It is not that simple. In a world that allows women to earn money and be reasonably independent of men, that may be a reasonable guiding point. Most women don't live in that world.
05:49 PM on 07/22/2010
And if the man thinks he doesn't need to wear one because he's circumcised, that just makes it easier for the woman to convince him to wear one, right?
photo
HUFFPOST SUPER USER
raker
09:38 AM on 07/22/2010
I learned long ago to be skeptical of headlines about groundbreaking advances in preventing or treating AIDS. There has been a steady flow of insubstantial stories since the beginning of the epidemic.

When I heard the vaginal gel story, I wondered why this would be specific to Africa, and why this gel is touted for vaginal use but not anal. Obviously, if the gel works in Africa it would work here too. So this looks like they are using Africa to test this new product that may or may not have therapeutic value. Activists have been fighting to get potential therapies into use as soon as possible, so maybe this is good or maybe not, but the publicity so far seems a little...off.
03:45 PM on 07/22/2010
It's touted for vaginal use because that's what was studied. You have to reduce the number of variables so that you can control for them.
photo
HUFFPOST SUPER USER
Morcat
09:25 AM on 07/22/2010
While I don't quarrel with the author's statical analysis, I suppose there is also a sense that if one life could be spared, it might be worth it. Donors to research organizations should get demanding. We need to demand the cure AND vaccine that are missing from this crisis -- which is far from being over. If only non-gay people could get AIDS, there would already have been a cure.
02:23 PM on 07/22/2010
A lot of straight people have aids. Demanding a cure will produce nothing. I appreciate the efforts of those who are working to produce a cure or vaccine. The progress in the field over the last 30 years has been astonishing.

You suppose that, in a sense, saving lives can be worth it !?!?!
03:47 PM on 07/22/2010
Um, you do realize that the most common method of HIV transmission is heterosexual sex, yes? That three-quarters of all cases of HIV infection were through heterosexual sex, yes?

The overwhelming majority of people with AIDS are straight. The reason why there isn't a cure is because this is a virus and viruses are exceedingly difficult to cure. Politically, well, most of the people with AIDS are Africans and we don't really care about them, now do we?
photo
HUFFPOST SUPER USER
Morcat
09:29 PM on 07/22/2010
Oops! I see what you mean. That last bit is badly said. It should have gone more like, "If the only people who could get AIDS were non-gays, we'd already have a cure."

I wonder what would have happened with polio if Salk and Sabin had said, "You know, we just don't know enough about these viruses, and it's probably too hard to find a vaccine." We don't have to settle for "It's just too hard," with AIDS, either.
09:01 AM on 07/22/2010
“It would be natural to conclude that use of this gel would reduce reported cases of AIDS in Africa by 39%” Is Africa only continent affected by Aids pandemic? We need to avoid creating a false impression that says only certain group of people gets Aids and are living in Africa! Unless we know something about Aids that other people doesn’t know. We need to salutes the University of Kwazulu Natal in South Africa and government of South Africa for making an attempt on Aids cure. We must ignore armchair critics who criticize for sake of criticism without any attempt of making a meaningful contribution while our people are dying like flies. We must also remember that this trial or research was implemented while former RSA Pres Dr Thabo Mbeki was still the Head of State and his late minister of health was still in her position so the notion that seeks to suggest that former president of South Africa was an Aids denialist is totally mischievous.Philani@Umlazi.Durban.South Africa
08:05 AM on 07/22/2010
While I understand what the article is trying to point out and I'm no fan of hyping up medical research, I'm wondering why no mention has been made of widespread rape as a form of transmission. Some of these women may not have the luxury of asking their "partners" to get a condom...especially as such rumors persist that having sex with a virgin can cure you of HIV in some of these ravaged countries.

With that said, if this gel is something that women can use to protect themselves in such instances of unwanted sexual contact, even if it is a slight improvement, it is more than what they had before. I'm not advocating not using condoms at all; but it's my understanding that many times, that's just not a choice these women have.
05:52 PM on 07/22/2010
If a man believes he doesn't have to wear a condom because he is circumcised, doesn't that make it harder for the woman he's trying to coerce to have sex with him to convince him to wear one?
photo
HUFFPOST SUPER USER
TheBurdicks
Whatever happened to my yellow bus?
07:49 AM on 07/22/2010
This "groundbreaking scientific breakthrough" is weakly, at best, supported by (inadequate) research. There is much better evidence supporting the distribution of relatively small amounts of cash among women at risk, i.e. those mired in poverty with no way forward, as an effective tool in the reduction of transmission. The well documented success of microcredit programs in developing countries may be the best model for this approach, enabling impoverished women at risk to become more self sufficient, through a self sustaining program. Even in the U.S., HIV/AIDS is disproportionately found among those living in poverty, and could be better addressed through an expansion of the social safety net.
In other developed countries exchange programs providing clean needles have been very successful, as has been decriminalization of drug use coupled with a public health approach focusing on prevention and treatment.

Although not the main focus of this blog, male circumcision as a preventive measure is even less supported by scientific evidence, and may have no real effect. Genital mutilation, female or male, can never be justified when forced on non-consenting minors. These practices should be banned internationally as a crime against innocent children.

http://www.doctorsopposingcircumcision.org/info/HIVStatement.html
HUFFPOST SUPER USER
DakkonA
www.DisentangledReality.com
07:44 AM on 07/22/2010
While most of the article is okay, there is one part I must dispute.

Your analysis of the underlying data betrays a misunderstanding of study design. "If as few as ten less women in the control group had not contracted the disease." You're not saying anything more than "if the results were different then the results would be different." They weren't different, and statistical tests can be done to make sure that its more than would be expected from random error. As long as the samples were randomly divided into the two groups, their rates should be identical.
photo
HUFFPOST SUPER USER
johntalbs
Author of Survival Investing
10:45 AM on 07/22/2010
Another point I failed to mention, this is the 13th study involving the application of a gel. How many studies do you think it would take such that random error might cause ten people not to contract the disease in the control group. This is a form of data mining. Lots of tests, and report the one that "succeeds".
01:31 PM on 07/22/2010
Arguing that 39% in the study period is not an impressive number is fine. To attack the methodology of the study (i.e. the "within the first" assumption) when this very methodology is used repeatedly in similar studies is weak however. Even for a layman, do you seriously expect these researchers to conduct a 20 year study on a new intervention for what is clearly an epidemic needing urgent solutions. Not to mention the complete waste of funds if it turned out the microbicide also held no promise.

You also fail to point out that this particular microbicide is different in that (1) it does not need to be taken immediately before intercourse, (2) it is designed as an HIV specific microbicide and not a general or hopeful microbicide; wild guess? this may actually improve chances of its effectiveness. And, this microbicide is not intended to be used as an alternative to condoms, but rather as an intervention for those women who find themselves unable to ensure condom usage because of men who refuse to use condoms (in a community where women are in no position to insist).

Even for someone like myself who is not very optimistic about this, your arguments sound mostly disingenuous. I sincerely hope you are not a scientist.
HUFFPOST SUPER USER
DakkonA
www.DisentangledReality.com
10:36 PM on 07/22/2010
Ah okay. Now I would have to ask whether this gel is substantially different from those others that could account for the apparent success of this trial? If they were all pretty much identical, I would have to agree with you.
03:53 PM on 07/22/2010
You're missing the point, betraying a misunderstanding of statistics.

His point was that the study group was so small that a very small change in the data has a large effect in the observed outcome. Given that no physical process is ever perfect every single time, there will always be variation in the result from one trial to another. But if what you're searching for is so very sensitive to fine variations, then you must be very careful when it seems like you have a positive result.

Here's an example: Suppose we have a test that is "98% accurate." You are given the test and it comes back positive.

Exactly how confident are you that your results are accurate?

Hint: I haven't given you enough information to make that determination. What might you need to know?
photo
HUFFPOST SUPER USER
johntalbs
Author of Survival Investing
04:59 PM on 07/22/2010
My point exactly, you are taking a small sample, and then reporting two small percentages of that small sample, and then expressing the difference in percentage differences of the percentages to get a big number like 39%. That is why I raised the issue of ten women. On this planet of close to 7 billion people, if just 10 women on earth in the control group had not contracted the disease we never would have heard of this study and 20,000 people in Vienna wouldn't be on their feet hooping and hollering.
HUFFPOST SUPER USER
DakkonA
www.DisentangledReality.com
11:20 AM on 07/23/2010
And my point was that it doesn't matter. The difference was large enough to be significant for that sample size. Statistics take a small sample size into account. That's why you use statistics and randomized group determination. Yes, more studies should be done, but that's no basis for discounting this result.
06:37 AM on 07/22/2010
This is an insane article.

How is South Africa supposed to pay for universal testing and anti-retroviral? What about compliance? How do you get the 75% of South African HIV patients to reliably take drugs so you avoid creating a new drug resistant mutation? And how long will this take? How many people will become infected and die before that happens?

The whole reason people are researching microbicides is that so many women in South Africa don't have the ability to negotiate for condom use. That is the victory in creating a working microbicide- that women who have no other way to protect themselves can find some protection from HIV. It isn’t meant to solve the HIV crisis.
photo
HUFFPOST SUPER USER
johntalbs
Author of Survival Investing
05:02 PM on 07/22/2010
Be careful who you call insane, I want to end this epidemic, once and for all. Certainly it would be cheaper to spend some money now to eradicate this illness than to continue to spend money every year in the future to try to control it.
06:31 PM on 07/22/2010
The only possible way to eradicate HIV would be an HIV vaccine, which we don't have yet. And even then it would take years like it did with smallpox.

You're suggesting that a country with a per capita income of $8,000 create an enormous, expensive HIV system, constantly monitoring the whole population. And if that was possible, compliance would still be an issue.

What you're suggesting is a pipe dream. Sure it sounds nice, but such a major change would take a decade and for the world to pour billions into South Africa. Meanwhile, more people are infected every day and microbicides are one way to prevent new infections.