This post was prepared by Jason Lakin of the International Budget Partnership.
How should we allocate scarce resources to health sector priorities? Suppose we have a pot of $1000 available to spend on the treatment of two diseases. One disease kills 2000 people per year, and the other 1000 . If it costs $1 to treat 1 person, the pot of money could cover treatment for all of the people suffering from the less prevalent disease, or half of the people suffering from the more prevalent disease. (The cost of treatment is the same; assume, too, that people who are treated recover from both diseases with the same probability.) Of course, the money could be split up in different ways to cover some of the people suffering from each disease. What is the “right” way to divide the pot of money?
This is not an academic question. Governments confront this dilemma every day. Of course, they are often choosing between allocating funds to treatments that affect different populations, have different rates of success, and have very different costs. But that just renders the decisions even more complicated. A recent article in the New York Times provides a stark example of the trade-offs.
“In Africa’s two most populous nations, Nigeria and Ethiopia, the number of people who died of AIDS in 2007 — 237,000 — was less than half the 540,000 children under 5 who died of pneumonia and diarrhea. But this year, the $750 million the United States is spending on H.I.V. and AIDS in the two countries not only dwarfs the $35 million it is spending there on maternal and child health, but is also more than the $646 million it is spending on maternal and child health in all the world’s countries combined.”
Treating pneumonia and diarrhea can be done for a fraction of the cost of providing treatments for HIV, so it is natural that we spend more on treating HIV. But what is the cost of doing so in human lives? We can get a very rough idea by using some global estimates of the cost of providing treatment for people with these diseases.
First, we can use an estimated cost per case of treatment for HIV of US$485 (a 2007 estimate of the median price of a first-line ARV, lab costs and service delivery in Africa as reported by UNAIDS here). Let us ignore diarrhea for now, which is probably the cheapest of the three diseases to treat of those we are discussing. We can assume that if we treat all the cases of diarrhea and pneumonia as cases of pneumonia, that will be an upper bound on the cost of treating them. How much does it cost to treat pneumonia? This table provides some estimates for Africa, based on four different levels of severity. If we assume that the four types of respiratory infection are equally likely to occur, we get a weighted average cost of US$55 in 2001. Of course, these numbers are just estimates. I use them for illustrative purposes (and as long as it is cheaper to treat pneumonia and diarrhea than HIV, the point I am making stands).
Using these estimates, it would cost an additional US$29.7 million to treat all of the 540,000 kids who died from pneumonia/diarrhea in Nigeria and Ethiopia. Were this money to come out of the HIV budget, it would reduce the number of HIV patients that could be provided treatment by about 61,240. So, using these admittedly very rough estimates, our current allocation of resources from the pot of money for disease treatment suggests that we value the life of a person with HIV at 8.8 times the value of the life of a child with pneumonia.
In the same New York Times article cited above, Jeffrey Sachs questions this logic:
“Jeffrey D. Sachs, the Columbia University economist, countered that wealthy donors still spent far too little on global health and rejected what he called the wrong-headed idea that ‘we need to make a terrible and tragic choice between AIDS or pneumonia.’ The United States has invested heavily in the fight against AIDS, and other wealthy nations should pick up more of the cost of other global health priorities, he says.”
According to Sachs, the key problem is that the pot isn’t big enough. That may be true, but the needs are also unlimited. For example, in spite of all the money that has poured into HIV in recent years, half of those who need treatment still don’t get it. And things are only getting worse: the price of ARVs has fallen, but new guidelines from the WHO (based on new science) may advise that treatment begin earlier than in the past, in order to improve survival rates. This will almost certainly raise the overall cost of treatment.
Are the lives of people with HIV really more valuable than those of children with pneumonia? If not, are we willing to invest enough money in both diseases to cover treatment for everyone who needs it?
The debate on what diseases or health programs should get donor funds is a tricky one. There are various lobbies – HIV, vaccines, MDGs, maternal health, child health etc.. all competing for these funds. What these individual efforts do not see is that all these diseases and programmes are interrelated in some way. The issue cant be one against the other – HIV/AIDS v/s pneumonia or whatever else. It should be addressing all the needs, that is whoever has whatever ailment should get the requisite care and treatment. This is what comprehensive primary healthcare is about to which we had committed in 1978 at Alma Ata. This is now forgotten and segmented approaches of various lobbies/interests are being pushed globally and they distort national health policies taking away their comprehensive character and being replaced with selective programs and targeted approaches. These segmented interests on specific health issues lead to verticalisation of healthcare and make healthcare even more expensive. So the most effective way for use of donor funds would be to pool all such funds going to a country into a national health fund and let the health policy stakeholders in that country decide how to use them as per the needs within that country. Donee countries should not become victims of interests of donors.
Makes you think. Ravi is right that lobbies unhelpfully promote segmentation when an integrated primary health care approach (that rightly emphasizes prevention rather than waiting to cure) would be more sensible. But choices would still need to be made about how best to use limited resources, because primary health care doesn’t mean that everything can be done. The evidence is clear that comparatively way too much is spent on sexy ailments such as HIV/AIDS while millions die from basic, treatable stuff such as diarrhea. In Tanzania, I have seen any number of health centers which lack water and toilets, where women cannot deliver their babies safely, but which has a new building with 4 air conditioners and 2 Land Cruisers and weekly workshops on AIDS. The reason for this has less to do with the importance of some lives over other lives, and more to do, as Ravi notes, with the whims that sway the moral and political imagination of the donors, in contexts where ‘recipient’ governments haven’t got the desire or gumption to insist otherwise.
This is the wrong question to ask. Some questions deserving attention instead include:
1. If HIV patients are abandoned and left to die, why should governments trust donors not to abandon “cheaper” investments? What are the opportunity costs of treating whatever appears to be most cost-effective and “sustainable” in a given budget year?
2. Prioritizing diseases like pneumonia “over” HIV on cost-effectiveness grounds does not account for the “sunk cost” of the billions that have already been spent on HIV; if those patients are abandoned, almost all of them will die within a few years, greatly diminishing the value of previous investments. If HIV is abandoned as a donor priority, how many billions are we willing to waste and how many millions are we willing to leave to die?
3. Cost-effectiveness studies are usually misleading because of the short durations of study and their poor understanding of overlapping disease vulnerability. Cost-effectiveness models almost never account for scenarios in which patients benefiting from “high-impact” interventions get HIV and die early deaths a few years later due to the explosion in HIV transmission that would result from abandoning universal treatment access. The same people who get pneumonia are almost always the same people most at risk of dying from diarrhea, malaria, malnutrition, HIV, and a wide range of neglected tropical illnesses. Is there any evidence that the “cost-effectiveness” of treating pneumonia truly holds out over the lifespans of poor patients, as most of the models assume? I’ve never seen any.
The most cost-effective and “sustainable” health investments are to treat only the healthy wealthy and let the poorest of the poor die from the disproportionate disease burden they bear.
We’re better than that!
In response to the points raised by MRP:
The framework for these comments is interesting. Although my post did not mention cost-effectiveness once, it is mentioned several times here. I bring this up because there is a tendency, not only in MRP’s response, but more generally, to conflate cost-effectiveness analysis with an analysis of the implicit values that we have placed on certain objectives by dint of our actions.
I set out a simple choice at the beginning of the original blog. I argued in general terms that we find it difficult to choose between competing priorities. This problem is not limited to health or even to public finance.
Although I did not discuss it, we create tools like cost-effectiveness analysis (CEA) precisely in order to help us choose between priorities. But CEA is just one tool for helping us make decisions, and I did not (would not) put it forward as the only one. Serious promoters of CEA do not either; in this chapter outlining its usefulness (See http://www.ncbi.nlm.nih.gov/bookshelf/picrender.fcgi?book=dcp2&part=A1662&blobtype=pdf), CEA is considered to be “only one of nine criteria relevant for priority setting in health if the object is to decide how to spend public funds.”
CEA cannot, for example, answer questions that are best left to open debate over our core values. After all, CEA relies on a single philosophical tradition, utilitarianism. Assuming that, for a variety of reasons, we do not assign equal value to public interventions to improve the lives of everyone by an equal amount, CEA cannot give us the right answer.
For example, suppose that we are deciding whether to spend additional health funding on people suffering from a rare but acute disease versus people suffering from a common but less acute disease. CEA will aggregate over all individuals and individual benefits to tell us what kind of intervention is most beneficial and at what cost. But we might feel that people with a very acute disease that they cannot afford to treat on their own deserve more support than people with a very common disease that they can probably manage on their own. That decision will never be arrived at through CEA, because it is based on an alternative philosophical viewpoint to utilitarianism, and because fundamental values are at stake in questions of public versus private provision of welfare.
The fact that CEA cannot “answer” a question of this type suggests not that we ought to reject it, but that we ought to think carefully about using it along with a variety of other inputs into decision-making. MRP makes a similar point when arguing that CEA does not adequately deal with the fact that vulnerable populations are particularly likely to contract a number of diseases, that the occurrence of individual diseases are not independent events, and that (ergo) investing in pneumonia because it is cheap could still result in people dying later from HIV if we traded off financing HIV for pneumonia. These issues could be reframed as follows: Do we want to give extra weight to certain particularly vulnerable populations when determining resource allocations? How much weight should we give to protecting people from one disease if they are likely to die of another later on?
To my mind, this simply suggests that we need to think carefully about our goals. Does spending disproportionately on HIV treatment accomplish those goals? Is it true that the current distribution of funding on these disease priorities really goes to the most vulnerable populations? Do we want to fund disease priorities for the most vulnerable groups or for those where we can make the greatest difference?
CEA is only a single input into the process of answering these questions. But ignoring the trade-offs I have put forward does not make hard choices go away. The only fair condition for avoiding trade-offs is putting up all of the resources necessary to avoid them. Since all resources come from someplace, however, this could only be done in the health sector by impinging on other priorities, like education.
As long as there are limited resources, we will have to debate how to spend them. MRP is right that CEA is not the answer, and I never claimed that it was. I am not sure I see how trashing CEA gets us closer to that answer, however.
Just to add one more dimension to the debate. I reiterate what I said in my comment earlier that we should not be debating one disease against the other. That I think is not the issue. The issue is how do we spend resources we have in the best way as well as how do we generate more resources because for the health sector resources are never enough. Countries have achieved good comprehensive healthcare delivery with good outcomes (Sri Lanka, Costa Rica, Cuba etc..) with resources as less as 2% of GDP, including HIV care, and countries with even 15% of GDP (USA) think they are short of resources. The question is how we synergise all resources available and use them with both effectiveness and equity in our strategy. This would essentially require that we treat healthcare as a public good and once we do that then these debates do not arise, though constraints may still prevail leading to waiting lists for example.
Another linked issue is how specific diseases, medical procedures etc.. gain ascendancy in policies and agendas. And here we have to confront the pharmaceutical and allied medical services industries which push things which lead to increased profitability for them. With healthcare in most of the developing world not being considered as a public good we tend to fall prey to the Pharma industry’s machinations. Today even the WHO and other UN agencies are have become victims of the pharma industry.
It is very good information about some myths regarding the curing of HIV/AIDS infection. It was very useful for many, including me to explore and gather the information about HIV/AIDS. It is a must-read blog to explore the new information about HIV infection curable medicines. Keep it posting these kinds of informative blogs in the future!