Print
Share
Still No Reason to Stall Male Circumcision, Forget the HIV Vaccine, or Throw Away Your Condoms
May 17, 2011
What if by taking a pill every day, all 33 million HIV-infected people in the world could not only fend off the deterioration of their own health, but also reduce their chances of infecting uninfected sex partners by 96 %? This is the prospect that is offered by newly announced results of the HPTN 052 trial. (See the Kaiser Foundation report here, the UNAIDS announcement here, the Global health Sushi report here and the trial registry info here. ) The trial has been cut short because only one among 877 HIV-infected people on anti-retroviral therapy (ART) infected his or her partner, while 27 among the 886 HIV-infected people did so. In view of the disadvantage apparently suffered by those taking the placebo instead of the real ART drugs, the researchers and their oversight board considered it unethical to continue to withhold the drugs from the couples in the control arm.
This dramatic evidence that ART lowers the infectivity of HIV supports the hypothesis of a beneficial biological external effect of ART. But is the evidence surprising and powerful enough to be a “game changer” as Michel Sidibe, the director of UNAIDS declared, or “to end, or at least diminish, a bitter feud within the AIDS world over how much funding should go to treatment versus prevention,” as the Wall Street Journal has suggested?
It has long been known that ART can have both biological and behavioral effects on prevention. A simple typology of these effects classifies them as either beneficial (because they slow transmission) or adverse (because they speed transmission). (See the table from our 2004 book reproduced at the bottom of this blog.)
So the new experimental finding adds support to the upper left quadrant of this table positing a beneficial biological effect and thus to the proposition that I blogged in 2008 here and here that these beneficial externalities of treatment might justify a greater allocation of resources towards treatment, as compared to a pure cost-effectiveness analysis that ignores such effects. In the majority of African countries that have not yet managed to treat a large proportion of those with more advanced AIDS disease, the push to use ART for prevention increases the tension between the ethical mandate to treat the sickest patients first and the stronger efficiency argument for allocating limited treatment resources to those who are not yet sick.
But it takes more than this confirmatory experimental trial to argue that prevention resources should be re-allocated towards treatment. Modeling shows that even if people start treatment when their CD4 count is 500, as they did in this trial, and then faithfully adhere to their daily dose for their entire lives, the future burden of the epidemic continues to rise until 2046. The two panels of the following figure project the future of the epidemic in Sub-Saharan Africa through 2050, incorporating the trial result that the infectiousness of treated persons is reduced by 96 %. These calculations also assume optimistically that financing will be found to enroll every HIV-infected person whose CD4 count is below 500, that circumcision roles out to 80 % of adult males by 2025 and that a 60% effective vaccine begins to roll out to all adults in 2025. Panel a of the figure shows that without enhanced behavioral prevention the number of Africans living with HIV/AIDS continues to increase until 2046, reaching a peak of 38 million before the AIDS transition occurs. And so does the cost of treatment, rising from its current level of around $6 billion to above $60 billion by 2050.
Figure. Simulations of the future African HIV/AIDS epidemic if treatment reduces infectivity by 96% and reaches everyone with a CD4 count below 500. Both scenarios assumed male circumcision reaches 80% by 2025 and a 60% effective vaccine begins to roll out in 2025. Panel a) assumes risk behavior is unchanged from current levels, while panel b) assumes a 50 % reduction in risk behavior.
(Source: Projections from the AIDSCost model. If you own a copy of Stata version 10 or later, you can produce the above projections by typing “findit AIDSCost” from inside Stata, installing the program, getting the data file and executing the following command for panel a:
“aidsproj using aidscgd2010.dta, uptake(.99) takeoff(2010) horizon(2050) cd4(500) maxep(0) gp(.96) weo regions(SSA) graph grcombine” For panel b: Same command replacing maxep(0) with maxep(0.5). )
These projections assume constant cost per patient-year of both 1st and 2nd-line treatment, which seems reasonable since current pressures on unit costs are in both directions. Perhaps the unit cost of treatment will fall as pharmaceutical firms recognize the profit potential of 35 million customers at a low price as compared to 5 million at a high price. But no matter how inexpensive the individual drugs, the number of people taking them daily would still rise by a factor of seven, creating an ever larger dependency of recipient countries on the donors and pharmaceutical firms providing the drugs.
In contrast panel b shows a scenario combining 99% uptake at a CD4 level of 500 with effective behavioral prevention which reduces infections from high risk behavior by 50 %. Adding only this behavioral change moves forward the date of the AIDS transition by 16 years, to 2030, so that by 2050 the number of Africans living with HIV/AIDS has declined to 24 million and annual treatment expenditures will be about 25 % less than without this behavioral change.
With these considerations in mind, behavioral HIV prevention still deserves the highest priority. The potential savings in AIDS treatment cost from even small behavioral changes accumulate to extraordinarily large benefits over the next 40 years and effective prevention shields the African countries and their citizens from an important part of the looming burden of this disease.
Source: Based on Table 3.3 on page 46 in Over et al., HIV/AIDS Treatment and Prevention in India: Modeling the Cost and Consequences (2004) available from Google Books here.
Possibly Related Posts
28 Responses to “Still No Reason to Stall Male Circumcision, Forget the HIV Vaccine, or Throw Away Your Condoms”
Post a Comment
We value frank and constructive exchanges and encourage you to use your real name in your comments.
May 17th, 2011 at 4:46 pm
Dear Mead–
Please show me the data on the behavioral interventions that you suggest should be scaled up, along with the evidence of their effects on incidence and associated statistical analyses.
Most people agree that we need a combination approach to prevention, but one that is evidence based, for instance, in the light of HPTN 052, combining HIV testing, ART, male circumcision, etc.
Let’s stop this inane debate pitting treatment vs. prevention. ART drastically reduces infectiousness and transmission and it is now proven in a randomized controlled trial.
We can make progress in prevention, if we start putting the thing we know work–with strong evidence to support them–into widespread use.
We can make progress, if instead of targeting now proven effective prevention methods (i.e. ART), we start putting greater scrutiny on interventions with little data to support their use (e.g. based on self-reported behavior change).
You once said to me that governments shouldn’t support ART provision as AIDS drugs were not a “public good,” as their benefits only accrued to those who use them. It is clear now that ART is actually a public good of considerable proportions, yet your resistance remains…
Gregg
May 17th, 2011 at 5:00 pm
Greg, I believe I said that at the time Martha Ainsworth and I wrote “Confronting AIDS,” we raised the question of whether ART might have benefits for people other than those treated, but considered the evidence too weak in 1997 to support economic policy. As I point out in the above blog, I had already shifted my view to accommodate some prevention effects in 2006 and then in 2008 and have done so again with my participation in the Institute of Medicine report “Preparing for the Future of HIV/AIDS in Africa: A Shared Responsibility,” for which we changed my projection model to include the biological prevention benefits of treatment. Now HPTN 052 the latest evidence to support ART’s benefits to the sex partners of ART patients.
The two simulations in the blog are both computed under the very optimistic assumption that the point estimate of incidence reduction among partners of 96 % would also apply when access becomes universal to all those with a CD4 count of 500 or less. The available media reports on HPTN 052 mention that in addition to the 28 new HIV infections shown to be from their partners, another 11 spouses were infected from outside the relationship. Given that recruitment into the trial was premised on both members of the partnership agreeing that they intended to remain sex partners and that all participants were presumably counseled about the dangers of unprotected sex in general, these 11 additional infections emphasize that ART will be insufficient to control this epidemic, a point that is also make by the simulations I present. This blog argues instead that behavioral prevention remains essential.
You are certainly correct that we have too little information about the cost-effectiveness of behavioral interventions, individually or combined. Cost-effective government-led behavioral interventions succeed if (A) using few resources, the government persuades individuals to reduce their risk behavior and (B) reduced risk behavior slows the incidence of HIV infections. No one doubts the second link in this causal chain: reduced risk behavior DOES reduce incidence. The problem is with the first link, from government action to behavior change. I think my prevention essay posted here (http://www.cgdev.org/content/p.....l/1424161/) presents reasons to hope that several promising avenues for improving link (A) are insufficiently explored.
May 17th, 2011 at 5:17 pm
I second Gregg’s question, but also pose another… Until the recent modeling–now supported by overwhelming RCT evidence–of treatment as prevention we couldn’t actually model the end of the crisis. Even your models here are about reaching everyone with CD4<500 with ART and then adding a group of "behavioral" interventions. Yet your suggestion is that somehow these are the "highest priority"? Mead, it's a logical jump I'm just not following. Of course, what's needed–as ever real programmatic thinker knows–is combination prevention with ART as one of several cornerstones to create overlapping prevention coverage across the community… But without the ART, it doesn't work.
So are you now saying that you DO support rapid scale up of ART to everyone below 500?
May 17th, 2011 at 5:42 pm
The adverse consequences in the last table are rather speculative, and in many respects what everyone supposes about all interventions involving sex, from contraceptive pills to circumcision. The benficial consequences, which are very powerfully associated with ARV treatment, need to be acted upon even if the effect is less than stellar. Making it possible is to my mind a political not economic question.
Treatment is prevention. We’ve known this from years ago in mother-to-child prevention programmes. Treatment and prevention are not an either/or. Finessing the treatment v prevention semantics now seems to me like Nelson prevaricating about whether to check his telescope with both eyes.
May 17th, 2011 at 6:11 pm
One key point is evident in this discussion but missed in the conclusions: The fundamental question is one of political will. We know the evidence. Now will the resources be made available to save the next generation?
The charts and the analysis presume treatment availability with behavioural prevention mechanisms as the variable. But they show nothing about what the results would be of a continuing stagnation of political will, and economic resources, for HIV/AIDS treatment. This ignores the reality.
HPTN 052 confirms that the political choice between prevention and treatment is a false one. Treatment is prevention. Yet in the face of now irrefutable evidence, this blog post still aims to strengthen this false dichotomy – concluding without justification that treatment should be given a lower priority than behavioural HIV prevention. Resources are needed for both biological and behavioural interventions, but this logic seems intended to undermine donor support for treatment.
It is not an acceptable argument that the cost of treating those who need treatment for HIV/AIDS (with benefits to the individual and the larger population) is more than governments want to bear, and we should therefore be excused as we watch an epic tragedy unfold.
May 17th, 2011 at 11:36 pm
Dear Professor Over,
It’s not your insistence that prevention is crucial that is worrisome to those of us who believe in healthcare as a human right. It’s the apparent unwillingness to engage with complexity and move past debates about why NOT to act. Cost-effectiveness analyses and mathematical modeling should be useful tools for advancing the conversation about how to obtain better health outcomes for ALL poor sick people (not just those yet to be infected), not for trying to prevent such conversations from occurring!
Anyone who takes the time to obtain the most fundamental understanding of HIV epidemiology can see that an incredibly complex biosocial epidemic requires comprehensive interventions including both prevention and treatment.
I would argue that the same is true for any disease (cancer, diabetes, heart disease, etc.) in pursuit of pragmatism and justice, but the case of HIV given recent evidence also mandates this dual approach based on sound science and economics.
One of your assumptions on the economics end strikes me as somewhat analytically flimsy. Global market prices for antiretroviral drugs as not static or unchangeable… they never have been in the past. In fact, two hours after you published this blog entry, UNITAID/CHAI/DFID announced a massive supply-side partnership that will yield savings of more than $600 million over the next three years.
http://unitaid.eu/en/resources.....tries.html
Of course the next step after last week’s galvanizing confirmation that treatment is indeed a pillar of prevention is not to advocate that we drop vaccine research efforts, cease scaling up access to circumcision, or divert funds away from condoms. Rather, we should see the HPTN 052 results as a clarion call to increase the size of the pie so that we can harness all of the tools of modern medicine to combat the first pandemic of modern globalization.
No one is arguing that prevention is not important – the critique is simply that the debates you revel in generate more heat than light. We simply wish that you would turn your considerable influence and intellect towards modeling things like the economic benefits accrued to a society by NOT allowing millions of children to be orphaned in need of support by the same programs that fund prevention and treatment. We simply wish you would use your mastery of cost-effectiveness metrics to estimate the true costs of doing nothing for those already infected and the true benefits of providing adherence support for things like food insecurity and transportation costs…
You have the power to be a champion for both rigorous public health economics and social justice. All it would take is opening your eyes to the fact that assumptions are sometimes just that; global health funding need not be a zero-sum game, and it can be used in a way that increases access to all primary care while increasing national accountability and capacity. There is more funding out there… perhaps you could also run a comparative cost-effectiveness analysis exploring the impact on the AIDS pandemic of reclaiming just a tiny fraction of the $82 billion in tax cuts handed to the top 2% of Americans last December.
As John Sloan Dickey, the President of your (and my) alma mater, always used to say during your years as a student: “The world’s troubles are your troubles, and there is nothing wrong with the world that better human beings cannot fix.” Please help us to fix them – and to do it without writing off the most vulnerable.
Sincerely,
Cameron Nutt
Dartmouth College Class of 2011
Health Policy
May 18th, 2011 at 2:46 am
The new research shows that adherent treatment with a good treatment regimen, when combined with strong prevention/safe-sex messages will reduce the risk of transmission by 96%.
Behavioral prevention messages have been promoted for 25 years, have historically been taken up by about 30% of the populations and have helped to reduce incidence rates by about 19% globally since the height of incidence in 1997-99. Behavioral interventions were never as broadly implemented as advertised. People and governments remain squeamish about comprehensive sex, gender, and sexual health education; condoms supplies have been historically inadequate; and socio-economic, gender, and cultural dynamics increase transactional, inter-generational, and concurrent partnership rates. Nonetheless, there are positive signs that well designed behavioral prevention policies can and do have an impact, often more so with young people.
The best behavioral intervention, besides abstinence and/or non-penetrative sex, which are never going to be majority approaches, is consistent and correct condom use, which has about the same prevention efficacy as treatment as prevention. The problem is that most heterosexual (and many MSM partners) do not want to use condoms in their most intimate relationships.
Of course, there are other biomedical prevention methods existing or in the pipeline which can have an important role in dampening the pandemic, for example, male circumcision and perhaps microbicides, PrEP, and even vaccines.
Instead of merely throwing cold water on the new treatment as/for prevention study, it would have been much more interesting if Mead Over could have modeled the difference between the current efficacy of behavioral and non-treatment biomedical prevention and then compared the AIDS Transition date from behavioral/bio-medical only to behavioral/bio-medical plus treatment for prevention. I’m fairly confident that the move in the Transition date would be even more dramatic than the one he models here.
One last point on costs of treatment, it would be much better if Mr. Over talked about cost dynamics in the real world instead of making assumptions about constant costs. On the plus side, the Clinton Foundations and others continue to work behind the scenes to improve market dynamics and costs in the generic industry and have just announced what may amount to $1 billion in ARV cost saving over the next three years (assuming treatment continues to scale-up). The Medicines Patent Pool is also trying to gather Big Pharma patents so it can out-licenses to generic producers who will make improved formulations and compete at efficient economies of scale in developing countries. Finally, other scientists are working on dose optimization and new medicines with longer half-lives and more bioavailability that will reduce the amount and thus the costs of APIs.
But, in the real world, the battle to lower prices is not uni-directional. Big Pharma continues to push for more stringent patent and data protections and stronger enforcement rights. It wants to ease patentability of minor changes in chemical entities and formulations and insists on patents for new uses; it wants to extend patent periods and to limit the use of IP flexibilities like compulsory licenses that allow generic competition; it wants to outlaw price controls and therapeutical formularies; it wants to create monopolies on test data; etc. etc. And, the US government backs Pharma up in its trade negotiations and behind the door pressure on developing countries.
An economist in the socio-economic realm would reference these real world factors and actually come out in favor of lower costs achievable through public policy. And he or she would point to the policy incoherence between the US seeking greater efficiencies in its global AIDS initiative on the one hand and its trade/IP policy that will have the predictable impact of ratcheting up the price of AIDS medicines.
As others have continually noted in criticizing Mr. Over’s economic/cost-effectiveness analyses, he looks at costs in very sterile terms, describing only the pay-out costs of treatment uptake as opposed the terrible social and economic costs of not treating people living with HIV, most of whom are in the most economically productive time of their lives and are parents of young children as well. I’m currently teaching in Durban, South Africa, where the antenatal infection rate is over 40%. Not treating is not an option, morally or economically. However, not addressing prevention is also not an option. Mead Over continues to make it sound like the two are in a zero-sum game.
May 18th, 2011 at 11:42 am
I agree with the above correspondents. The flaw in Mead’s argument is that some kind of \enhanced behavioural intervention\ will be able to cut incidence to a far greater degree than behavioural interventions have so far managed to do.
There is no doubt that behaviour change IS able drastically to reduce HIV incidence, but only in very special circumstances.
These include mass panic and public awareness in the early stages of an epidemic, as undoubtedly happened in the earliest stages of the HIV epidemic in gay men in some parts of America and the developed word: situations where behaviour change combined with such a saturation of HIV infection that the high-risk population either die or cease being high risk, as happened possibly in Uganda in the 1980s and Zimbabwe in the 2000s: and situations where condom use has historically been extremely low but awareness combines with increased availability to produce a sudden change in risk, as happened in Thailand in the early 1990s nad may possible – just possibly – be beginnig to happen in South Africa now.
Another example of sudden and effective behavioural change are the changes and consequent lower-than-expected HIV incidence that have happened in some of the biomedical prevention trials, due to the intensive and unusual support, monitoring and condom provision associated with the trial.
However a number of large meta-analyses of behaviour change programmes in a number of different populations, largely in the developed world, have found the figures Brook Baker cites above: about a 30% reduction in risk behaviour and at most a 20% reduction in STI or HIV incidence in the far smaller number of studies that used these as an endpoint.
It is possible that the hundreds of studies into behaviour change interventions were not in the main conducted as well as they might have been, and often measured outcomes that did not directly relate to HIV incidence, but I don’t think we are suddenly going to come up with a behaviuor change programme that will produce anything like the change in HIV incidence that Mead wants to see.
Why? Psychotherapy research shows that, while behaviour change programmes do work, their effect is not long lasting and they have to be renewed periodically. Behaviour change is slow, and even generational, and HIV infection is fast: we may have cut the smoking rate amongst US adults at least threefold since the 1950s, but it took 60 years to do so – and nort only does the amount of money and ingenuity poured into mass-media campaigns about smoking dwarf that poured into HIV, but it also required legal changes such as limiting cigarette availability that may not be applicable to sex. Small interventions such as individual and group work are impossible to standardise because of cultural and situational differences – look at the wildly different rates of success amngist the ‘popular opinion leader’ intevention amongst different populations – while mas-market interventions are impossible to subject to an RCT as by definition you cannot have a control group.
Does this mean I think we should give up on behavioural support and encouragement to have safer sex as part of HIV prevention? On the contrary, I think in the new world of biomedical prevention methods, we are ethically obliged to universalise it. After all, behavioural support is every bit as necessary to help people adhere to ARVs as it is to help them use condoms.
What we need to be doing is campaigning for a reasonably well-validated and well-conducted programme of behavioural support to be included in every imaginable HIV pevention programme. It shouldn’t be compulsory, because that assumes people haven’t already made a conscious decision about the degree of risk they wish to run, but it should certainly be available. It should become the ‘ground’ upon which we base all other interventions and if we neglect it, incidence may rise. Prevention programmes should also continue to contain an element of operational research and auditing to fine-tune all components to the local circumstances and population.
But we should stop imagining that by somehow intensifying behaviuor change programmes, or intorducing a magically ‘right’ one, we can cut incidnece and prevalence at a stroke. We can’t.
May 18th, 2011 at 2:50 pm
Dear All,
Thanks for this interesting discussion. However, I am troubled that so many of you disparage behavioral prevention. Please show ME the evidence that condom promotion, HIV testing, and mass distribution of ART have had any effect on HIV incidence at the population level, anywhere (except perhaps a very slight effect of mass ART in San Francisco, where the HIV rate has nevertheless been stable for years). Condom promotion and testing haven’t even succeeded in the artificial environment of a randomized trial, and the new drug based approach hasn’t been shown to be practicable on a population scale, especially in Africa, where the challenges of treating those who need the medicines to stay alive are formidable enough.
Behavior change, mainly partner reduction, has been shown to have saved millions of lives in Uganda, Thailand, Zimbabwe and the US gay community. Some of you may not like to hear this, but I urge you to read the many peer reviewed articles on the subject, some of which are referenced in my book “The Invisible Cure: Why We are Losing the Fight against AIDS in Africa” and others—particularly the Zimbabwe studies, have been published since.
It is true that many behavioral interventions have failed in RCTs, but none of these built on the lessons of those historical successes. Some of us have good ideas of what better behavioral interventions would look like, but getting funding to test them is all but impossible. The World Bank, USAID, NIH, CDC etc etc, barely even fund evaluations. Meanwhile, the most picayune drug trials get funded instantly. Now, why would that be, do you think?
Cheers, Helen
May 18th, 2011 at 3:20 pm
Dear Helen-
The issue for me is that we have little data on effectiveness of these kinds of approaches to HIV prevention, particularly those that offer up hard endpoints such as incidence and using rigorous experimental designs.
There are several studies now underway that will mix behavioral and biomedical approaches to HIV prevention and these need to be strongly supported by all of us (see Kurth et al, Curr HIV/AIDS Rep. 2011 Mar;8(1):62-72).
If we want to see a prevention revolution, we need to push for better and more prevention research and not simply be happy with what we have and imply that if we simply did more of these things, we’d see a bigger effect.
Frankly, AIDS prevention was colonized by psychologists early on and eschewed approaches that would address community-level and structural factors driving the epidemic.
It’s as if in preventing automobile related fatalities, we set out to counsel drivers across the planet one-by-one, instead of adopting seat belt laws and installing airbags.
Individual behavior change is difficult to accomplish, we need to move upstream in HIV prevention, if we want to see population level effects.
What worries me about Mead’s approach is that it is based on a “faith-based” notion of HIV prevention, in which he speaks mostly uncritically of the entire field. This is necessary from a rhetorical position for him in order to be able to juxtapose ART provision against the far superior, far more cost-effective prevention options.
HPTN 052 is not a trivial study–it has a 96% effect in an RCT in thousands of people. What worries me is that instead of saying: look ART is a key prevention tool, let’s figure out how to take what we know now about what works and get together and figure out how to combine prevention interventions for maximum local impact, we see Mead still railing against ART, which I think is counter-productive and no longer a tenable position.
So, let’s move forward together; we need more resources for global health (not just AIDS), and we need to marshall the evidence to help us craft the best local strategies for preventing new infections and preventing unnecessary deaths from AIDS, we need more research on behavioral and biomedical interventions, but also on interventions that look at ways to address structural drivers of the epidemic.
We have so much to do. Propping up the prevention vs. treatment false dichotomy is a diversion, a waste of time and effort.
I for one am ready to move on….
Gregg
May 18th, 2011 at 4:30 pm
I agree with much that my critics say about the weak evidence in favor of behavioral interventions and have expressed similar reservations to available evidence here. And I especially agree with Greg’s “let’s figure out how to take what we know now about what works and get together and figure out how to combine [treatment and behavioral] prevention interventions for maximum local impact.” I conducted the simulations for this blog, which show the benefit of combining prevention with treatment, in that spirit. However, all the caveats apply to these projections that I expressed about models in general here.
Brook is correct to point out that the current trends in the prices of ARVs are not auspicious. I support his view that the industrial organization of the international pharmaceutical industry, the political economy of their interaction with government and the economics of intellectual property for life-saving drugs for the poor all deserve deeper inquiry than I was able to provide in a short blog. Brook, are you conducting research in this area?
I would like to clarify: the simulations I have run here are not designed to pit treatment against prevention, but rather to suggest that placing all our bets on treatment as prevention will likely fail to meet our expectations. In my simulation, even with optimistic assumptions about male circumcision and an eventual vaccine, 99 % treatment uptake for all with CD4 less than 500 fails to control the epidemic before 2046. Furthermore, the fiscal burden for all of Sub-Saharan Africa climbs to $60 billion per year. The donors are now having trouble sustaining their constituency’s support for about $6 billion a year for treatment out of a total foreign assistance bill of about $100 billion per year, for all sectors. PEPFAR’s funding is holding steady while the Europeans are all cutting back.
I am concerned that donors will stop supporting treatment, as costs continue to increase dramatically. But, even if donors are able to increase their financing, will African governments agree to such levels of dependency? When I presented on the AIDS transition at Brandeis two months ago and in France at the University of the Auvergne last week, African students in the room expressed strong distaste for the increasing dependency of their governments on donor support for AIDS treatment.
It’s great to see Dartmouth students are reading CGD blogs. I hope you will read widely on our site, for there is much to learn here on many development topics other than AIDS. The question you raise, Cameron, of the impact of AIDS mortality on the overall economy is interesting and relevant, and you are correct that I have not focused on it in recent work. But, I have spent years considering this question. While important possible channels for deep impacts have not yet been sufficiently explored (such as the long term impact of orphanhood), the available evidence suggests that the most devastating impact of early adult mortality in Africa has been on the individuals whose lives have been cut short. Survivors cope, and they recover. It is important to remember that the surrounding economy is more affected by issues like banking reform, mobile-phone access, exchange rate policy, export vigor, business-stultifying corruption and, over the longer term, by schooling, than it is by even the highest AIDS mortality yet observed.
The continued vigorous economic growth of economies with severe AIDS epidemic, together with the devastating human suffering caused by AIDS convince me that the avoidance of this suffering and of the dependency of the African societies are sufficient reasons for governments to exert more effort controlling this epidemic—which, I maintain, must involve strengthening incentives for effective prevention, along with treatment. It will be self-defeating for advocates to attempt to pile on specious arguments about the impact of AIDS on macroeconomic growth, which will be quietly dismissed by the African ministers of finance who must make the budget decisions.
May 18th, 2011 at 4:49 pm
I work almost entirely on the generalized epidemics and will reserve my comment to those. What’s interesting to me is that this finding is really little new. The evidence from the Partners in Prevention Study (Donnell) and the Attia review as well as scads of biologic evidence were already rock solid that taking ARVs markedly reduce onward transmission. What is perhaps new, is that adherence among those folks with presumably little or no symptoms was good enough to prevent transmission. But even there, it is an open question as to how much that would be true outside the rarified world of the clinical trial and for the long term.
Concerns about “test and treat” for the generalized epidemics were never about the biologic efficacy. Rather to name some of the main ones: 1) Such an approach would miss the roughly 40% of transmissions that occur through sequential acute infections (something like a chain reaction) which occur too rapidly to be detected or prevented. 2) Whether a very large percentage of people would present for testing (still many in Africa have never been tested even once despite intensive efforts.) 3) Issues related to long term adherence 4) The huge burden of the cost both for the drugs and the service delivery for many millions. 5) Potential risk compensation.
One of the reservations about ARVs for people in the chronic stage of infection is that risk of transmission at that time is actually relatively low. Ron Gray’s recent article on new infections in Rakai is quite illuminating in my view.
What’s notable, but little appreciated is that prevention is actually working in sub-Saharan Africa per the 2010 UNAID report – albeit too slowly. Incidence overall is declining even more than that caused by the epidemic saturation phenomenon. Sexual behavior has changed markedly in several prominant countries and as Helen points out, associated with marked declines in incidence. RCTs are not the only kind of evidence and here we are talking about hard evidence based on incidence. We need to continue to build on that positive change in behavior.
All that said, of course we need a number of prevention tools. My own view is that strategically for the generalized epidemics the highest priority should be 1) Behavior change toward partner limitation (and some other things) 2) Male circumcision 3) Wide availability of condoms. I won’t list the other interventions except to say I do think there is a potential selective role for ARVs (beyond treatment) such as for IDENTIFIED HIV positive partner in a discordant partnership.
I agree with Gregg that we need to combine prevention strategies and move on to doing it.
May 22nd, 2011 at 3:34 am
Dear all,
Good to see this vigorous exchange.
While many specific points and opposing views are being aired, I am generally supportive of Mead’s post because it provides a counterbalance to the over-zealous, and in my view dangerous, reactions and interpretations the NIH study has provoked in many quarters. I offer two example of this:
Here is what the Lancet said about the NIH study announcement:
“Agencies such as President’s Emergency Plan For AIDS Relief and the Global Fund to Fight AIDS, Tuberculosis and Malaria need to reassess their prevention portfolios and consider diverting funds from programmes with poor evidence (such as behavioural change communication) to treatment for prevention. There is now an ethical imperative for guidelines to be revised to start treatment much earlier than recommended.”
And just look how Foreign Policy interpreted the results here: http://bit.ly/ltMmdb
Both of these reactions are overstated and as yet unfounded, and pretty scary when all we have at hand from the NIH is a press release. There are so many details about the NIH study that we just don’t know yet, and that’s part of the problem with publishing trial results by press release.
By way of one simple confounding example, we know that “at least seven of the 39 people in the study who did acquire infections had a different genetic strain of HIV to their positive partner” [thanks to Steven Fouch for quote].
I don’t think anyone here is seriously advocating for any ‘single strategy’ approach. We clearly need varied, overlapping and integrated approaches based on local realities.
This discussion is revealing in as much as it immediately divided into two clear ‘camps’, based largely on your perceptions of one another and respective positions vis-a-vis HIV policy, rather than on a debate about the specific NIH data (that have not even been released yet). So, good to see mention of the potential opportunities and obstacles that the putative ‘test and treat’ approach might face, such as those that Helen and Jim began to air most recently.
To me this exchange looks like entrenched, partisan defense of pre-established positions rather than real dialogue. Accusing Mead of pitting treatment vs. prevention is like the kettle calling the pot black. As far as I can tell, you are all doing it.
May 22nd, 2011 at 6:25 am
So, let’s move beyond entrenched positions, onto common ground. We need to expand access to treatment and drastically curtail new infections. Expanding access to treatment means more than pills into bodies, it means strengthening health systems, integrating HIV programs with TB, MCH and other key health initiatives. Curtailing new infections means pushing ahead with a new era of “combination prevention,” which includes male circumcision, testing, ART, condoms, syringe exchange, opiate substitution therapy, and behavioral interventions with strong evidence to support their effectiveness. In addition, we need to put all of these things to the test in an expansion of implementation science for treatment and prevention, while pushing for continuing investments in AIDS research both biomedical and behavioral. Finally, none of this is going to happen until we get the Obama Administration and other donor nations back on track to expand spending on AIDS and global health (even in these times of austerity), and get national governments in the developing world to take health seriously. We also need to boost activism–nothing gets done on the advice of experts: we need to support activism worldwide to keep our leaders feet to the fire. Think we can all agree on that.
May 22nd, 2011 at 10:45 am
Hi Gregg,
Certainly, I think we can agree on this, but I was particularly struck by this comment:
“We also need to boost activism–nothing gets done on the advice of experts: we need to support activism worldwide to keep our leaders feet to the fire.”
At the moment, the strongest activism in public health is being conducted by the pharmaceutical companies, if sometimes/often behind the scenes. These companies and their allies have all the politicians and all the big international health agencies by the cajones, and they have a strong influence over the medical and popular media as well. They are the ones in control of the policies, and the “group think”—particularly the promotion of the false belief that “behavioral prevention doesn’t work”. What do you propose to do about that?
Cheers, Helen
May 22nd, 2011 at 10:56 am
Helen-
I propose that you and I write something together. I am no fan of big pharma, but there are other institutions that push their own agendas as well (see the NY Times this morning on Gates Foundation education advocacy efforts), so the drug companies don’t have a monopoly on influencing public policy.
In any case, I wouldn’t agree with blanket statements such as behavioral interventions don’t work. Large drops in incidence early in the epidemic were the result of changes of behavior–how much of that was spontaneous and due to abject terror and how much of it was due to programs put in place by governments depends on where you look. In the USA, it was clearly the former; in Uganda and Thailand you could make a case for the later.
We need evidence. Both randomized controlled data and operational/implementation science data and we need more of it.
I think we would both agree that behavior change is difficult. As my friend Ted White here at Yale (and who worked at STOP AIDS in SF for a long time) says, if it was easy, he’d be thin. I think we veer towards biomedical interventions because in a sense they are simpler sometimes. But just because something is hard, doesn’t mean we don’t pursue it.
So, let’s call together for a renewed push on HIV prevention research of all kinds, with a focus on rigorous design of studies with biological endpoints when possible (sorry, the era of self-reported behavior change has to go), using “combination therapy,” that is a mix behavioral, structural and biomedical interventions, based on the specificity of local epidemics.
Gregg
May 22nd, 2011 at 11:39 am
Dear Gregg,
Sure, let’s keep talking, but you can’t just ask for everything, since a great deal of prevention research is already going on. Indeed, NIH and the other agencies spend a disproportionate amount of money on AIDS research, much of it frivolous, if you ask me.
What’s missing is serious attention to behavioral prevention, which is the only thing that has ever worked on a population-wide scale. Certainly many behavioral interventions have failed in RCTs, but that could be for two reasons, first that the interventions were not appropriate and second that behavioral RCTs are really hard.
For example, in one recent pilot study in Africa, the effect of the intervention was so powerful that everyone from the “control communities” kept rushing over to the “intervention communities” to see what was going on. This is great in some respects, but would obviously dilute the effect in an RCT.
An RCT in schools just might work, but the intervention would have to be carefully designed. So far, school-based interventions have failed, but I have some thoughts about why (See Epstein H. AIDS education programs miss target. AIDS. 2010 Aug 24;24(13):2140).
Also as you note, “reported behavior”/”reported behavior change” aren’t always great indicators. It depends on how the studies are done. There is a rich (but hopelessly underfunded) field of research trying to develop better interviewing techniques to help obtain responses that correlate with biomarkers.
Another possibility is triangulation: doing an intervention, and then conducting a before/after survey PLUS participant observation PLUS using coincident ANC or DHS+ prevalence changes to corroborate. This is cheaper and easier than setting up an RCT, and was essentially the technique used to demonstrate what happened in Uganda. In that case, all the stars lined up in one direction: partner reduction, motivated by community based and government led campaigns were the principle causes of the declines. There’s no disputing it—much as some in the international public health community might wish to obscure that finding and tell us that drugs are the only way out of this ongoing crisis.
Also, regarding the HIV declines among gays in the US in the 80s, it is certainly true that the Reagan Administration turned its back on the victims of the epidemic, but there were other groups, especially GMHC ?, that made clear recommendations early on to gay men about partner reduction, through organized campaigns, leaflets, outreach etc. The decline didn’t “just happen”—it was also organized. Plus, in those days, CDC did a great job of studying and disseminating the epidemiology of the disease and that helped the activists. Today, the international health agencies seemed to have lost their scientific bearings, on this subject at least.
Cheers, Helen
May 22nd, 2011 at 4:15 pm
On behavioural interventions, well I am waiting to see the results. A general trend in prevalence other than upwards, excepting one or two cold spots, is hard to discern. Witness the greater now than then HIV prevalence in South Africa, a country dear to my heart (see http://www.avert.org/safricastats.htm for a possibly accurate summary). And even the UK (gentler, arguable, but probably going up).
I am heartened by the reports from Lusikisiki, Eastern Cape, South Africa, where MSF first set down to give out ARVs and a locally run organisation continues the community support with very good treatment outcomes, there is an active Treatment Action Campaign and condoms are socially order of the day (in public discourse at least). But disheartened by my experience of villages north of the N2, a main dividing road in the province, where HIV is still called “this disease.”
Decent behavioural interventions, I would welcome the evidence and the results of their effective application. Decent treatment interventions, I would do likewise. They go together in my mind.
Prevention has as its end purpose important social outcomes like avoiding premature deaths, burden of funeral expenses and broken families that go beyond solely stopping additional cases of HIV. I really wish to achieve them all. It seems to me we need a more of a big picture focus, rather than discussing this or that intervention is better.
We have a long way to go and need all hands on deck, an integrated approach.
- matt
May 26th, 2011 at 7:45 am
A word to Helen:
Let me make it clear I don’t disparage behavioural interventions in the slightest. My non-AIDS job is psychotherapist and I am aware of the very rich research literature showing that support for behavioural change does work (in the case of 1-2-1 psychotherapy for symptom relief of anxiety and depression, it’s about 60% effective versus waiting-list control).
There are also occasions where behaviour-change campaigns have quite convingly led to, rather than just accompanied, risk behaviour reductions – the clearest to me being the Thai 100% Condoms capaign directed at sex workers and clients.
However I’m equally aware that the research literature also says that the exact intervention you make – what’s in the therapy ‘black box’ – is considerably less important that a) the willingness of the client to change b) the quality of relationship engendered (which in groups would imply peer relationship) and c) individual personal and learned skills of each therapist.
This rings true: a CBT programme conducted by a warm and caring professional feels very different from one conducted by an anxious rookie.
What this means is that I’m very pro behavioural support and very anti attempts to standardise or ‘manualise’ it or assumptions that it can be standardised like a prescription. This is a recipe for groups to parachute in stock behavioural programmes into cultures there the intervention is simply inappropriate – and do I have to utter the words ‘Uganda’ and ‘Abstinence’ to remind you that that can be counter productive – and expect them to work when delivered by anyone.
So behavioural support programmes in my opinion, while they can perhaps import ‘templates’ (an example I’ve personally trained in being the Positive Self Management Programme devised at Stanford), must be devised, managed and researched and audited locally, on behalf of, within and by the communities in which those programmes take place.
The Thai 100% condoms campaign was so successful precisely because it was invented in-country, promoted by an already respected politician, had a very precise taerget population, and deeply understood its own sexual culture. The same may or may not be true of Museveni and ‘zero grazing’, though the evidence is more muddy there.
Behavioural support programmes need to be included in any imaginable prevention scenario. But they must always be conducted accompanied by auditing and operational research and must always allow for flexibility in content and delvery to suit local conditions.
May 26th, 2011 at 8:09 am
Dear Gus,
Please read my book “The Invisible Cure: Why We are Losing the Fight against AIDS in Africa” and references therein, if you are still unconvinced about whether Zero Grazing worked. You are entitled to your own opinion, of course, by for those who have studied the campaign and the HIV decline in that country in detail, there is no question that it worked, and there are many surveys, peer-reviewed papers, etc to back this up.
Because of Museveni’s much later ties to the Bush adminstration and his embrace of abstinence during the 2000s, many people and even entire institutions have sought to discredit the Zero Grazing campaign. This has done enormous damage to the cause of HIV prevention.
You should also be aware that Thailand’s 100% condom program involved not just condom promotion, but also warnings against commercial sex, and triggered a 60% decline in male visits to sex workers, which had a powerful effect on HIV transmission. There was a good NEJM article on this in 1995.
Helen
May 27th, 2011 at 12:28 pm
I’ve appreciated these many comments, and Gregg’s call for some unity. In the spirit of unity, I do want to point out that I believe the lens of looking at these issues from the large scale/program/prevention lens in the generalized epidemics is rather different from the more clinical/small scale lens. That relates both to the nature of the epidemics, the nature of evidence and the practical potential for meaningful impact at scale.
On the nature of the epidemics, suffice it to say that while they vary in stage and maturity, there is something driving and actively propagating these epidemics. My strong (evidence-based) view is that both behavior and biology are at the root, and the “dangerous duo” of acute infection and multiple (notably concurrent) sexual partnering are what largely drives them.
Thus from the strategic perspective, the inability to address (for the most part) the UNC estimates as 38% of infections coming from antecedent acute infections – and beyond that largely drive the active propagation of these epidemics, – it is not just a trivial concern but rather a really PROFOUND one. The strategic model of “search and contain” has serious flaws for this epidemic. To use a military analogy – having great bombs was not enough to win the Viet Nam war. We had an elusive enemy. Thus my highest priorities are things that can span the population (male circumcision, partner reduction+, widespread condoms.)
Likewise, when folks rightly points out that an TasP approach depends on widespread testing, we must consider the serious programmatic limitations we have seen despite huge resources and emphasis on testing. For example the most recent DHS in the region I believe is Lesotho 2009. In this relatively advanced country that has attempted major outreach for testing, still only about HALF have EVER been tested. Recall that even the optimistic Granich model calls for annual testing of everyone who is sexually active. The other issues re TasP of cost etc are also obviously very important.
Then there is the issue of “evidence”. From the clinical perspective RCTs are the gold standard, though even in the clinical context there are serious issues of external validity and generalizability. From the large scale perspective, THE GOLD STANDARD SHIFTS. This is similar to Mike’s SINE QUA NON concept. The gold standard of evidence is evidence of what can work practically at scale. Based on that standard, the best evidence we have now supports behavior change (partner reduction,) since the improvements we have seen in declining incidence at country relate to safer behavior – particularly partner reduction. And though it is rightly argued that we haven’t seen that much impact of programs to influence behavior change, I would say the evidence from early Uganda and more recently Zimbabwe (albeit weakly) does provide evidence of impact even of weak efforts. But also there is no evidence of impact at the macro level in the generalized epidemics of virtually anything else – e.g. counseling and testing, STI Rx etc.
Now I don’t think my reaction is disappointment. Rather it is that I really don’t see that much new in these data and I’m concerned about declarations that this is a “game changer”. Or from the Lancet editorial to ” …consider diverting funds from programmes with poor evidence (such as behavioural change communication) to treatment for prevention…” These really worry me.
I want to return to Gregg’s call for unity. We all want to do the best to combat AIDS. We do need more unity, I believe based on better understanding. We probably all believe in the principle of combination prevention, including a role of TasP. But the real point is the strategy for prioritizing and implementing the various components coherently and practically, which calls for sufficient resources and execution.
May 27th, 2011 at 12:30 pm
I just wanted to say hear, hear to Jim’s fine post and reiterate that the TasP approach is great for truly monogamous discordant couples, but won’t prevent transmission during acute infection, and won’t prevent infections that come from outside the couple, which I gather were a significant fraction of the total in the HPTN trial (some 7-11/36 infections, right?).
For these infections, we need a different approach. Behavior change has indeed saved millions of lives in countries around the world, but when it has worked, it has not been achieved by changing the behavior of one person at a time either through VCT couseling or ABC messages aimed at individuals. It has always been achieved through collective change, preceded by the development of some sort of community consensus. People do what their peers do, and follow norms, etc. This is the case when you consider any sort of behavior change–not just sexual behavior, but also stopping smoking, losing weight, stopping FGM, etc. This is what happened in the three HIV prevention successes we know the most about: Uganda, Thailand and the US gay community. These campaigns worked because people found their own “prevention” language and promoted it–through programs as well as spontaneous interpersonal discussions.
International development agencies have yet to find the right way to stimulate this process through programs in the foreign cultures they are trying to help. This isn’t surprising, since it would be hard to imagine that a group of Ugandans coming here to help Americans stop smoking or lose weight would be very successful either, at least at first, nor would they be very persuasive in trying to get us to change our sexual behavior–unless they really listened to us, and collaborated deeply on developing such a program.
Rather than dispensing with behavior change programs (as the Lancet has indeed suggested) we need to think more deeply about what we could do to make them better. One thing that has not been done (enough) is to sit down with people in highly AIDS-affected communities, inform them about concurrency so they understand where their risks are coming from, and particularly how people can get HIV even if they are following their own custioms and not being typically “promiscuous”, and then ask them what they think they should do about it. [The growing emphasis on "special" populations (drug users MSMS prostitutes) in generalized epidemics is OK ,but most infections in E/S Africa are occurring outside of these groups, and people need to know that, and they need to know WHY.]
This approach of helping people understand their own risks and encouraging group action to change behavior collectively has worked a bit with FGM–through for example, the activities of groups like Tostan in West Africa–and it could work with HIV too, as Martina Morris has suggested. Unfortunately, the international agencies (PEPFAR, GFATM etc etc) have all but given up on behavior change and are increasingly putting all their funding into the clinic instead. This is great for those already infected and their (monogamous) partners, but will afford modest, if any protection to others who remain at risk.
May 27th, 2011 at 6:41 pm
Three points to add.
(1) Treatment as Prevention (TasP) is completely dependent on widespread testing, linkage to care, confronting stigma, behavioral reinforcement of safer sex and adherence, etc. It is a complex multicomponent intervention and is synergistic with MC, ABC, and other prevention stategies. Hence, when someone says, let’s divert resources from behavioral interventions (to increase testing, to increase care access, to decrease risk…) to increase ART coverage, I think they are building their castle on quicksand.
(2) The HPTN 052 trial ended but the study did not. This is misunderstood by many. We in the HPTN are following the couples for durability of the effect. HIV investigators over the age of 50 lived through the Concorde vs. ACTG 016/019 story for ZDV monotherapy, and we are concerned that durability of impact. [For some of you who are not intimate with HIV treatment research, the ACTG 016/019 studies demonstrated the benefits of zidovudine in treatment of HIV disease in slowing time to AIDS and reducting mortality; both trials were stopped early. The European Concorde Study was similarly looking at ZDV monotherarpy, but was not stopped early; it found the ZDV benefits to be very transient, suggesting that monotherapy was not as good as the American trials might have thought.]
(3) HPTN 052 was a superb study, in my view. No one knew that TasP (treatment as prevention) would achieve a 96% prevention benefit and this surely reinforces the public health utility of early therapy for the sake of the patient AND the community. Follow-up rates were high and community acceptance was remarkably good. Anyone who frames 052 results as “treatment instead of prevention” is poorly informed as the potential synergies of combination approaches and misses the point of 052 and the entire HPTN research portfolio testing a dozen strategies for HIV prevention (www.htpn.org).
May 27th, 2011 at 7:00 pm
(DH is in rural Africa and Sent this from his phone, with a request that it be posted.)
Hello from Zanzibar (where i have no computer & almost no internet access on this phone – which is also hard to write on…)
I appreciate the comments by Dave, Stan & others, including Helen’s apology/ explanation for her erroneous comment. Clearly this is a good/important study!
And, like Helen & a number of others, i too have been frustrated & discouraged by some of the hype & “anti-behav-chng” bias by some commentators (such as the Lancet editorial) The notion that this approach could make a massive dent in the overall epidemic seems ( considering acute infection,long term adherence, cost & other feasibility issues etc) questionable at best. Public health experts need to educate the public that “96 percent protection” in this trial does not even begin to translate into that level of impact at the population level…,
(& Dave, while I’m sympathetic to your basic point re treatment – prevention syngergies, what is the evidence for a prevention effect of hiv testing, in people who test negative (which BTW includes tmost of hose in acute infection phase)….
All that said, i have to say that, the more i think about it, i do believe there may be a non-insignificant potential for prevention. For example ,even tthough the study was conducted among heterosexual couples, what about the implications for high risk populations like msm & idu? Or, I always wonder/ worry what happens when a sex worker finds out she’s positive? Some will have “high ethical standards” &; will always use condoms or leave the business. But probably there are at least some (with hungry kids to feed etc) who might not always use a condom now – as a result? Putting such folks on early treatment could perhaps help protect the population at large in a substantial way, no??
May 28th, 2011 at 12:20 am
WSJ taking the line that “New evidence suggests the epidemic can finally be controlled”. More over-interpretation of the implications of NIH results.
Here: http://on.wsj.com/iAIJbW
May 31st, 2011 at 3:53 pm
1. I don’t think ART for prevention should be limited to stable discordant couples, nor should the HPTN 052 analysis be criticized for focusing on this aspect. The study appropriately focused on transmission to stable HIV-uninfected partners because transmission to outside/unknown HIN-uninfected partners could not feasibly be captured. Presumably ART also substantially reduced transmission to those partners. Given the strong interest in and epidemiologic impact of concurrency, I would think that, from an epidemiologic/prevention standpoint, HIV-infected persons with multiple partners should be prioritized for ART for prevention.
2. I agree that “treatment vs. prevention” is not a helpful discussion in the era of “treatment IS prevention.” However, I think it is also a disservice to forego vigorous, respectful, honest debate and further research about “proven-effective, high-impact prevention vs. prevention as we have always done it.” Resources are limited; there in fact are not sufficient resources, human or financial, to do everything.
3. As Jim Shelton and I have been discussing, there seems to a disconnect over the definitions of and criteria for “effective” and “impact.” As a way forward, it would be helpful to come to agreement about the terms of the discussion.
June 2nd, 2011 at 10:13 pm
As economists and public health specialists don’t seem to mix in good numbers, I’ll point the readers here to the Development Impact blog, where we also discussed this issue recently: http://blogs.worldbank.org/imp.....he-treated
June 6th, 2011 at 12:25 pm
[...] Mead Over at the Center for Global Development, notes in Still no reason to stall male circumcision, forget the HIV vaccine or throw away your condom that many are pointing to the new evidence that Treatment is Prevention as if it has already [...]