Print
Share
PEPFAR, Entitlements, and the Implications for U.S. Foreign Policy
July 7, 2010
When you’ve had enough light summer reading about the unintended consequences of relaxing regulations of the offshore drilling industry or of the U.S. financial sector, you can turn to an article in the July/August issue of Foreign Affairs on the unintended consequences of a program you thought was doing fairly well – the President’s Emergency Program for AIDS Relief or PEPFAR, which since 2003 has prolonged the lives of almost three million AIDS patients in the developing world. In the article, entitled “No Good Deed Goes Unpunished: The Unintended Consequences of Washington’s HIV/AIDS Programs,” (gated) Princeton Lyman and Stephen Wittels of the Council on Foreign Relations follow me in using the term “entitlement” to describe the implicit commitment the U.S. has made to each person to whom it provides a daily dose of antiretroviral therapy (ART) under PEPFAR. As I argue here, here and here, once a patient has been started on ART, withholding that person’s treatment will condemn an identified patient to death, an action which would expose the U.S. to more reputational risk than I believe it will be willing to accept. Lyman and Wittels say that “Because [ART] needs to be provided over a lifetime, treating HIV/AIDS patients is a serious long-term commitment. … [This kind of aid] will create a sense of entitlement in recipient countries and make patients directly dependent on the annual U.S. foreign aid appropriations process…” [Emphasis added.]
I argued that providing individuals with a life-sustaining drug on which they will thenceforth be dependent effectively constitutes a kind of “post-modern colonialism.” Similarly, Lyman and Wittels suspect that this kind of aid “could spawn as much resentment as gratitude.” They say that with continued expansion of the numbers of HIV/AIDS patients supported by the U.S., “[t]he effort might well produce gratitude among the patients benefiting from the aid, but as with other dependencies, it might also breed resentment, and African governments may be ambivalent, too.”
We also agree that a growing AIDS treatment entitlement will likely squeeze other health programs. Lyman and Wittels go even further and argue that AIDS treatment entitlements are likely to squeeze all of foreign aid. They concur that by channeling AIDS treatment assistance through the Global Fund for AIDS TB and Malaria, the U.S. could shift the entitlement burden to this international agency, which might then be able to share both the dependency and the fiscal burden among contributors.
Despite these important commonalities, Lyman and Wittels’ analysis differs from mine in two respects: (a) while I fear that the increasing dependency of PEPFAR countries on the U.S. government for the daily life-sustaining doses of ART would give the U.S. too much influence over recipient countries’ policies, they believe to the contrary that the humanitarian aspect of AIDS treatment actually ties the hands of the U.S. State Department, reducing their leverage over recipient policies; and (2) while I argue that donors should shift their objective from universal access to AIDS treatment to the achievement of an “AIDS transition,” they support the universal AIDS treatment objective. (See my recent Policy Forum piece in Science Magazine, which my colleague Bill Savedoff blogged here.)
On our first point of difference, Lyman and Wittels have changed my mind, persuading me with their tales from Ethiopia, Uganda and Zimbabwe that the U.S. will not threaten the withdrawal of PEPFAR and other humanitarian programs to influence the recipient’s policies, even when other democratic or humanitarian objectives are at issue. In fact, this commendable U.S. reticence is further evidence of the extent to which the patients on ART have an effective “entitlement”. Lyman and Wittels go on to say that “The Obama administration will need to recognize the paradox that in the absence of increases in other forms of aid, more humanitarian assistance will mean less leverage.” [Emphasis added.] And they point out that this fact casts doubt on “[t]he notion that development and diplomacy will always reinforce each other, one of the principles of Secretary of State Hillary Clinton’s plan to make them `twin pillars’ of U.S. foreign policy. … For one thing, [they say,] development efforts typically last much longer than the more immediate demands of diplomacy, a disconnect that is particularly acute in the case of PEPFAR.” This is pretty compelling stuff – with far-reaching implications for U.S. foreign assistance policy.
On the donor commitment to universal access, I believe that Lyman and Wittels are wrong to assert “None of these issues should be allowed to undermine the commitment to treat all HIV/AIDS patients.” As I argue in a podcast and this new essay, universal access to ART is a self-defeating goal which will perpetually recede as more patients are placed on treatment. Donors should instead commit to assisting recipient governments attain an “AIDS transition,” defined as enrolling enough additional ART patients to hold down AIDS mortality, while improving HIV prevention efforts until the verified number of new cases is pushed below the annual number of AIDS deaths. I am convinced that the secret to achieving an AIDS transition, either within a single severely affected country or globally, will be to enhance the incentives for HIV prevention at every level of HIV/AIDS policy. (See this essay.) When the AIDS transition has been achieved, and the total number of patients begins to shrink, universal treatment access will reemerge as an attainable goal.
3 Responses to “PEPFAR, Entitlements, and the Implications for U.S. Foreign Policy”
Post a Comment
We value frank and constructive exchanges and encourage you to use your real name in your comments.
July 8th, 2010 at 5:01 pm
Mead Over has studied the issue of AIDS policy as much or more than anyone and I am indebted to him for his writings on the subject. In his comment on our article in Foreign Affairs, he raises at the end an issue on which my colleague Steve Wittels and I pondered long and hard, i.e. for all the reasons described in our article, should the G8 rethink its commitment to enable eventually all who need ARV treatment for AIDS to be able to receive it?
In the end we decided not to challenge that commitment. It has such powerful human rights associations, and stands as such a great humanitrain commitment, it seemed almost churlish to suggest that it be abandoned. And yet, to be honest, the implications in our article and Over’s work, along with that of others, suggest that the commmitment is indeed unlikely to be achieved. Mead Over suggests a modified goal, and it is indeed a valuable proposal. It sets a significant target for treatment nevertheless, and yet puts more of a spotlight on prevention, and provides an end-state for eventual coverage. It does mean, however, that many will in the end be denied access to treatment and that will cause considerable outcry. Abandoning the universality of the original commtiment will also be hard for many dedicated advocates to accept. But by putting these propositions out on the table, ours on the unintended consiequences and Over’s alternative, we may be helping open a serious and broad-based debate on the issue that would lead to greater consensus on and acceptance of a more realistic and achievable goal. Steve and I hoped our article would help do that and this blog helps very much as well.
July 10th, 2010 at 5:48 am
Mead Over’s Global AIDS Transition Proposals -¬ How Weird Can You Get?
Brook K. Baker, Health GAP
July 10, 2010
Mead Over, in this blog commenting on a Lyman & Wittel’s article, which adopts Over’s “AIDS entitlement” rhetoric, references his recent essay: “The Global Health Transition.[1] In that earlier essay, and again in this blog, Over proposes a new paradigm for combating AIDS by preserving existing mortality reductions while reducing new infections through hard-headed prevention efforts. As explained in the longer essay, once the death rate exceeds incidence (new HIV infections), the “transition” will have been achieved. At the national level, according to Over, the AIDS transition paradigm forces donors and governments to limit treatment expansion to the achievement of actual reductions in new infections ¬ no proven prevention effect, no new treatment. At the sub-national level, treatment programs would be required to have a results-based prevention arm or a partnership with a results-based prevention partner. Successful programs could use a “cap and trade” system to sell treatment slots from well-performing to underperforming districts. Individual patients receiving subsidized antiretroviral therapy would be required to contribute a day a month to HIV prevention in their communities.
Over takes an axiom of epidemiology ¬ – it is important to reduce the reproduction rate of a disease ¬- and perverts it through a truly bizarre economic analysis and set of proposals.
• First, Over puts treatment and prevention in a hermetically sealed box and creates a bi-polar, win-win/lose-lose game. If prevention efforts work, additional people can be added to treatment. However, if prevention efforts fail, treatment must fail as well ¬ no new patients can start treatment. In this amoral game of life-or-death “incentives,” AIDS policymakers and service providers will be motivated to pursue better prevention outcomes not for the sake of prevention, but to preserve and extend treatment gains. However, if donors and implementers pursue bad or ineffective prevention policies ¬ for example, abstinence-only (a past favorite of PEPFAR), ¬ who pays? Not the policy maker; his or her life and job are secure ¬ it’s a poor person living with HIV in a poor country that is told to go home a die.
• Second, Over defined “transition achieved” as more death from AIDS than new HIV infection ¬, and he seems indifferent to how this crossover is achieved. So for example, if more PLWHA die – well that would be ok so long as the ultimate result was that deaths exceeded incidence. Accordingly, giving people bad medicines, or starting them late on treatment and thereby increasing the likelihood of mortality, would both be fine if the math worked out in the end.
• Third, Over misstates many “facts” and then uses these misstatements in his policy analysis. For example, he states that adding people to treatment has led to an explosion on the number of people living with HIV. However, his basic epidemiological model is wrong: only 4-5 million have been added to treatment, hardly an explosion to a population of 33 million. Moreover, despite access to medicines, many patients still die, especially in the first months of treatment if their CD4 cell counts are low, but even thereafter. Antiretroviral therapy has been shown to have significant life-extending benefits in rich countries, but survival in resource-poor settings is not as strong yet, primarily because many patients start treatment when their CD4 cell counts are extremely low. Even in rich countries, the patients who started treatment years earlier will eventually succumb to opportunistic infections unless a cure is discovered. Over also falsely claims that AIDS programming is focused more on prevention than treatment. However, all reviews of actual AIDS spending show that more is spent on prevention than on treatment. Likewise, he claims that Brazil and Thailand have achieved a lot on the treatment front, but little on the prevention front. However, the evidence is exactly to the contrary; Brazil’s bold prevention and treatment efforts have kept its prevalence rates at a fraction of those in Africa and Thailand’s 100% condom use campaign achieved dramatic results in reducing HIV rates among sex workers and their clients.
• Fourth, the main model that Over discusses is one in which only 15% of people needing treatment are added each year, presumably to keep down what he calls an international treatment entitlement.[2] This is roughly the loss to treatment during the first year of therapy in Africa, so this rate of “expansion” would actually result in a status quo treatment numbers. More to the point, leaving 85% untreated actually means that 85% of the people waiting in line for treatment will die because another 2.5-3 million PLWHA join the treatment-eligible queue each year.
• Fifth, Over suggests that treatment slots should be bought and sold like carbon-emission credits. Districts that can’t achieve provable prevention results could “buy” treatment from more successful districts. Since no HIV-positive person is actually a human, they are fungible.
• Sixth, because of their dependency debt, Over would require people receiving treatment, no matter what their circumstances, to devote one full day a month to prevention activities. Doesn’t matter what their job is or whether they are caring for sick relatives or children –¬ you’ve got to do
your community prevention work to “earn” treatment and to show your gratitude to benevolent donors who are footing the bill. There’s no comparable requirement anywhere in the world. Though I don’t know what chronic medical treatments Over or his family might require, but I assume he wouldn’t consider it appropriate to require them to volunteer a 1/30 of their time (per disease) to its prevention.
Over treats the most important scientific break-through ¬- that treatment might actually be the strongest form of prevention ¬- as an afterthought. A recent peer-reviewed study in The Lancet showed that placing people living with HIV on antiretroviral treatment created a 12-fold reduction in transmission to their HIV-negative partner.[3] This 92% decrease in the likelihood of transmission confirms the importance of scaling up widespread testing and reaching universal access (at least 80% coverage) to AIDS treatment. This finding suggests that the Over’s strategy of holding treatment hostage to prevention would undermine the world’s ability to eventually defeat HIV. Or more to the point, Over’s mind game is rendered meaningless by science ¬ AIDS treatment produces results-based prevention.
Notes:
[1] http://www.cgdev.org/content/p.....l/1424143/
[2] It¹s hard to determine what Over means by this 15%. Is it 15% of total treatment need, currently estimated at 15 million under the new WHO treatment guidelines? Or is it 15% of people who will die in two years if left untreated. That figure would be roughly 600,000 people a year based on the current AIDS death rate of approximately 2 million people a year (2 years x 2 million x .15 = 600,000).
[3] Deborah Donnell, et al. 2010. Heterosexual HIV-1 transmission after initiation of antiretroviral therapy: a prospective cohort analysis. Lancet Online.
July 14th, 2010 at 10:36 am
Thanks to Princeton Lyman and Brook Baker for their contrasting views on my essay proposing that the international community adopt the “AIDS Transition” as a new paradigm. Thanks also to Gregg Gonsalves for posting an opposing view as a comments to my colleague Bill Savedoff’s blog here.
Mr. Lyman and his co-author recognize that universal access to AIDS treatment is unlikely to be attainable, but as Mr. Lyman says, “[a]bandoning the universality of the original commitment will also be hard for many dedicated advocates to accept.”
Mr. Baker, as the co-chair of the Global Access Project, is just such a dedicated advocate. His critique of my AIDS transition paper is trenchant, but based partly on misunderstandings and in one case on an editing error. In this response, I will first clarify the misunderstandings and correct the error and then turn to differences of viewpoint that cannot be so easily explained away.
Mr. Baker’s correctly defines the AIDS transition in his own words as “preserving existing mortality reductions while reducing new infections through hard-headed prevention efforts.”
Here Baker has it exactly right. The transition would require both preserving mortality reductions and reducing new infections until annual infections are fewer than the number of deaths.
Given that he has correctly defined the AIDS transition, I don’t understand why Mr. Baker then turns the idea on its head as if I had said something completely different. He says:
“Over … seems indifferent to how this crossover is achieved. So for example, if more PLWHA die – well that would be ok so long as the ultimate result was that deaths exceeded incidence.”
But Mr. Baker has just finished characterizing the AIDS transition as “preserving existing mortality reductions.” So how can he think that could be reached by reversing those mortality reductions? What I really say in this essay is that allowing deaths to rise to exceed new infections would constitute a failure of the AIDS transition. I say this several times in the essay.
Furthermore, Mr. Baker seems to have missed the extended comparison I make between the demographic transition and the AIDS transition, including my expropriation of the iconic diagram of the declining death rate followed by the declining birth rate. I compare allowing the death rate to rise in order to slow the growth of the number of people living with HIV/AIDS to allowing free rein for the four horsemen of the Apocalypse in order to achieve the demographic transition. Both would constitute transition failure.
Perhaps Mr. Baker’s real concern is that comparison of the number of healthy life years saved through AIDS treatment with the greater numbers of healthy life years available through many other interventions might lead others to distort my definition of the AIDS transition in order to justify allowing AIDS mortality to rise. If so, I share his concern and will try to argue persuasively against such a distortion.
As I explain in the essay, my reasons for opposing such a reallocation of resources are first that donors which commit to continued patient recruitment are likely to have a greater stake in, and thus do a better job at, HIV prevention, and second that abandonment of the effort to reduce AIDS mortality would be a profound setback for the entire foreign assistance effort. After touting PEPFAR’s enormous success at prolonging the lives of AIDS patients, can donors really abandon the program and still hope to obtain taxpayer support for foreign assistance in the future?
I will use Mr. Baker’s helpful enumeration to respond to his other points. After addressing his second point, above, and his third and fourth points, below, I will end with a discussion of his first, fifth and sixth points.
Mr. Baker, in his third bullet, says my “basic epidemiological model is wrong”. Well of course my model is “wrong.” So are all the others. As the famous quotation from George Box has it, “All models are wrong, but some are useful.” A previous version of my model can be downloaded from our web site here. (We hope to have version 3.5 up and ready to download in a few weeks.) The model can be used to compute the future fiscal burden of AIDS treatment under a set of assumptions that are detailed in my third essay, available here. Since it incorporates all of the complexity that Mr. Baker suggests is essential, I hope the model is, in Einstein’s words, “As simple as possible, but no simpler.”
Mr. Baker says that prolonging the lives of 4 million does not create an explosion when the total number of people with HIV/AIDS is 33 million. The number on subsidized treatment is up to 5 million now and should reach 6 or 7 million if PEPFAR and the Global Fund reach their respective treatment 2014 goals. Whether this is an explosion is a matter of semantics, but a 20 % increase is large. If the rate of patient recruitment were close to 100%, the growth rate of people living with HIV/AIDS would be even larger, and if disinhibition accompanies treatment, as it has in many places, we will see an explosion which will rival that which was feared for the population.
Mr. Baker is right to call me out for a sentence that incorrectly criticizes the Thai and Brazilian HIV prevention efforts. In other places in these essays and in my other work I have effusively praised the Thai prevention effort and the Brazilian HIV prevention effort also deserves praise and emulation. Thanks to Mr. Baker I have corrected this typo in the draft that goes to the printer.
Mr. Baker says, in his fourth bullet, “the main model that Over discusses is one in which only 15% of people needing treatment are added each year… This is roughly the loss to treatment during the first year of therapy in Africa, so this rate of “expansion” would actually result in a status quo treatment numbers.” The appropriate comparison is between the annual number of people enrolled in treatment and the annual number of people who drop out or fail treatment. Since the number on treatment has grown from a few thousand in 2003 to more than 4 million, clearly the former number has been much greater than the latter all these years. Over this period, I estimate that the annual “uptake rate” in the average country has been about 15 % of those needing treatment in any year. According to my estimates, a 15 % uptake combined with a 3 % annual decline in incidence would succeed in suppressing AIDS mortality to its current level and get us to an AIDS transition by the year 2026. (See panel b of Figure 8.) However, if incidence does not fall, Baker is correct that 15 % uptake will not be enough to indefinitely suppress mortality. In fact, according to panel a of Figure 8, even an 80 % uptake would not achieve that goal.
Mr. Baker’s last paragraph references a recent Lancet article in support of the proposition that AIDS treatment “is the strongest form of prevention.” There is indeed evidence that antiretroviral therapy (ART) suppresses viral reproduction and thus reduces the infectivity of otherwise risky sexual contacts. But how much contribution can this beneficial spillover effect make to the AIDS transition? And how expensive would it be? In order to have a substantial effect on HIV incidence, Granich et al (2009) propose the so-called “test-and-treat” (TNT) approach, which would test every person in a country annually and immediately start people on lifelong treatment if they have become HIV infected. By making a set of assumptions that others have found implausible, the authors conclude that TNT would indeed slow the epidemic. In my second essay, I analyze this strategy and conclude that, over the next 40 years, it would cost at least $500 billion dollars and possibly as much as $800 billion. That would be between 2.5 and 4 times the cost of the 15 % uptake strategy that PEPFAR and other donors are now finding hard to sustain. Whatever the merits of the claims for the prevention benefits of TNT, I believe that it is simply not affordable.
In Mr. Baker’s words, the AIDS transition paradigm “puts treatment and prevention in a hermetically sealed box”. His first, fifth and sixth bullets take issue with this approach.
While Mr. Baker’s stark metaphor is invidious, it does capture a central idea of the AIDS transition. Think of the “sealed box” as the fiscal space that donors and domestic taxes make available to a recipient government to be spent on valuable social programs now and over the foreseeable future, like AIDS treatment, road construction and teachers salaries. The dimensions of this box might vary according to donor whim and tax revenues, but AIDS treatment entitlements and HIV prevention’s effects on future such entitlements are inescapably inside this box. While entitlements use up some fiscal space, effective HIV prevention, because it reduces the future demand for AIDS treatment, enlarges the fiscal space.
While Mr. Baker would apparently want the rate of treatment uptake to be completely independent of HIV prevention success, as it has in the past, I argue that donors and recipient governments should promise in advance that at least part of the gains from improved prevention go towards increased treatment uptake. Such an arrangement, I argue, would be “incentive compatible” because it would align the incentives of all interested parties towards improved HIV prevention.
One might think that a treatment advocate like Mr. Baker would like this idea. It’s almost the same as an earmarked tax dedicated to AIDS treatment. What logical reasons might lead advocates to oppose it?
First they might hold onto the hope that donors will commit to more than $500 billion on AIDS treatment in sub-Saharan Africa over the next 40 years. In that case, they would think they don’t need the extra fiscal space that improved prevention might buy. I believe this position is unrealistic and does a disservice to African governments and AIDS patients.
Or treatment advocates might be so pessimistic about HIV prevention that they are unwilling to link the rate of future treatment expansion to HIV prevention success. My second essay argues for optimism here. And Jim Shelton of USAID is also a prevention optimist in a comment posted here.
Or treatment advocates might simply oppose incentives of any kind. I hope not, since I think our best chance for eventually defeating this epidemic is to agree on a set of incentives for all of us – donors, recipient governments, susceptible populations and AIDS patients – which will efficiently engage our efforts towards holding down AIDS mortality while bringing the number of new infections even lower.