The President’s Global Health Initiative (GHI) has outlined a new U.S. strategy for global health. Secretary Clinton recently described the GHI as aiming to bolster health system development and particularly primary care, tackle health problems that can be eliminated with relatively little investment (for “sustainability”), and focus on maternal and child health (MCH).
To an outsider, this all sounds pretty good; but to those who are familiar with the history of global health, the rhetoric around the GHI ominously harks back to past mistakes. In this blog entry, we discuss the history behind primary care in global health, to explain why previous attempts at “health systems development” perversely backfired in the 1970s, 80s, and 90s.
Why advocate for primary care?
What does a primary-care-based health system look like? Ideally, such a system facilitates access. Patients won’t need to travel vast distances and wait for hours (or days) to get care. They would get a range of medical conditions treated in one place, by a provider who is able to follow them over the long-term. The greatest burden of disease in the world now comes from chronic conditions (ranging from HIV to heart disease) requiring serial visits; even infectious diseases like TB require months if not years of regular check-ups to prevent drug-resistance and death. Visiting multiple providers or being “lost to follow-up” is known to worsen health outcomes and raise health costs as patients return later with complications. So we’d like to get patients preventive care, to avoid the devastating and costly consequences of end-stage disease—like the progression of undertreated diabetes to kidney failure, or the development of cirrhosis among the 2 billion people with untreated hepatitis B. The idea would be to stop focusing on tertiary hospitals filled with specialists who treat the urban wealthy, and train and support general practitioners to provide care through dispersed clinics in the communities where the sickest people actually live.
Most people can agree on these general goals.
So why hasn’t it happened?
In 1978, it seemed possible. That year, 134 countries gathered in Kazakhstan to sign the “Alma Ata Declaration”, claiming they would try to achieve “health for all by the year 2000”. Well, that obviously didn’t happen. Was it because of killer new epidemics like HIV? War? Famine? The answer was more mundane, and political: a group of governments ministers and organizations agreed that the Declaration would require too much long-term commitment. Building meaningful primary care systems in poor countries would require constructing clinics, training and adequately-paying skilled medical providers to staff the clinics, and developing reliable supply chains for pharmaceuticals and medical equipment.
Instead, they decided to limit their commitment to maternal and child health campaigns; these campaigns, like vaccination, are typically cheap and require little training. MCH programs were aligned with popular misconceptions that the majority of disease in poor countries could be prevented or treated with simple vaccines or a “basic minimum package” of pills. And because they focused on children and pregnant woman, the “innocents”, such campaigns avoided difficult issues behind the major causes of illness, which would require engagement with powerful lobbies (e.g., tobacco control) or uncouth topics (e.g., sexually-transmitted diseases).
Confusing selective with sustainable
One of the greatest legacies of Alma Ata was the confusion between selective healthcare and sustainable healthcare. The idea was that if a health program was cheap and required minimal infrastructure or personnel, it was “sustainable”.
The opposite turned out to be true. One of the clearest examples of this problem was in the training of “barefoot doctors” and “community health workers” to act as healthcare educations and minimalist care providers for the rural poor. Rony Brauman, the former president of Doctors Without Borders, explained:
“The basic assumption was that 90% of the world’s diseases could be easily prevented or treated…this assumption is absolutely false—it’s a demagogic lie that enables a number of NGOs [non-governmental organizations] and health and humanitarian institutions to establish their own special laws for the poor. We think that the world’s poor and destitute will have the right to benefit from our generosity through the promotion of health education. The West will have the doctors; the developing world will have ‘health educators’. We will have real practitioners; they will have community health workers…In just a few months after the 1978 conference, WHO trained teams of so-called community health workers—giving them pills, disinfectants, a few drugs, whatever… That was much easier and cheaper than getting involved in real healthcare policy for the Third World, which implies that you train people seriously, you pay them, and you organize a huge administrative and logistical system in order to supply health centers, health posts, and hospitals that allow them to function, that allow them to bring real medical care to deprived areas. Of course, the people who established and promoted this program would never have permitted themselves or their children to be treated by those community health workers.”
The cheapest care just produced substandard care, and was ironically the least sustainable care; the health workers felt as unsatisfied as the patients, and quit their jobs. The cheapest pills also failed to materialize in real health gains, since the majority of illness still came from diseases that they weren’t equipped to treat. Data from the programs reveal minimal to no progress in affecting mortality rates from these unequipped workers. And the emphasis on MCH failed to address social realities: as large-scale statistical analyses have shown, the major indicators of progress in child health are intimately related to the health of adults, as sick parents rarely make for healthy children. Age distinctions may be useful theoretically or for fund-raising, but when parents and children live in the same household, the chronic diseases that affect parents will prevent those parents from providing adequate nutrition and care for their children, both as the sickness itself takes hold on the parents, or when the financial costs do.
Turning primary care into a bad powerpoint presentation
In 1993, the World Bank published its annual World Development Report with a focus on health. This report heralded a new focus on developing health systems by giving-up the agenda posed by the Alma Ata Declaration, and instead focusing on markets to deliver healthcare services. Years earlier, the Nobel-prize-winning economist Kenneth Arrow demonstrated why this doesn’t work. But the World Bank approach emphasized that costs would be lower and benefits higher if patients paid for healthcare (so-called “user fees”, now found to be detrimental to health), and public-private partnerships would herald a new era of efficiency. The data now reveal the opposite; the poor were systematically excluded from privatized healthcare systems, which were ironically often less efficient and less effective that public systems as preventive health does not reap a profit, unlike fee-for-service programs that profit from excessive testing and reimbursements after people get sick. (For some perspective, take a look at p.101 of the Bank’s 1993 report to see just how much they underestimated the impact of the HIV epidemic).
To maintain the claim that health systems were still being developed, the Bank and others sponsored “Sector-Wide Approaches” (SWAps)—projects intended to improve health system management by redistributing funds from disease-specific programs to system-building exercises like leadership trainings for Ministry of Health workers. Forty-six years Gonzalez’s treatise on the tensions between “Mass Campaigns and General Health Services”, disease-specific (vertical) programs were again pitted against broad “health system” (horizontal) programs.
The problem with SWAps is that they sounded good, but lacked any real detailed plans. To improve the “health system” was so vague that funds were wasted. The introduction of SWAps in Zambia not only failed to produce noticeable health system management improvements, but diverted funding from the country’s tuberculosis treatment operation to useless consultant contracts and meetings for bureaucrats, resulting in critical medication shortages and ultimately precipitating the collapse of the country’s entire National Tuberculosis Program, which had previously been improving treatment rates dramatically. The funds appear to have been diverted to consultant reports, meetings and hotel room expenditures, while needed medicines were emptied from clinic shelves.
Leading papers on the topic of health systems development continue to use glittering generalities that are empty of what to actually do to improve the system, with statements like “reforms improving performance require information on explanatory factors” (no kidding?), or astoundingly-informative graphs like this one, which we could use to teach public health students what it means to produce a meaningless graphic:
“Health systems development” essentially turned epidemiology into a bad McKinsey consultant’s powerpoint slide.
Ironically, some of the health systems that are building up primary care in the most robust ways have cast off all this rhetoric and just invested in clear, specific plans for gradual healthcare development (look, for example, at Costa Rica or Sri Lanka). How did they do this?
They acknowledged that vertical programs are not the enemy. Why did SWAps fail? Because they were vague and had no specific targets, no process or outcomes measures to keep a program on track. The old guard of public health, in contrast, always cites the smallpox vaccination program as a glittering success—an example of disease eradication that took place because a specific target was the goal. The problem, of course, is that few diseases today are like smallpox; most of the world’s burden of disease consists of problems that cannot be solved with one pill or one vaccine, and that requires real medical care that is entwined with addressing social problems like poor sewage or food quality.
But the “new guard” of public health got the idea of how to adapt the benefits of vertical targets to broad primary care development. Partners in Health, for example, created model programs in the toughest conditions on earth by saying “hey, we have a specific protocol to treat tuberculosis or HIV, and we can do it by training people to learn how to use specific medications and look for specific side-effects or rashes or lab abnormalities”. They created a very targeted protocol, a way of keeping track of whether the protocol was working, and therefore were able to perform quality improvement programs to make sure the patients had good outcomes. But then they expanded: they said “hey, TB and HIV are not the only problems here. We have a lot of malnutrition in this community.” So they expanded into a malnutrition treatment protocol, and then a cervical cancer protocol, and so on…gradually building up a repertoire of quality vertical programs that, together, expanded service to provide a solid backbone of investment and resources and training that constitute primary care. By starting slow and going steadily through the major community problems, they did something remarkable: used local epidemiology to determine what their community needed first (for some communities, it’s HIV, for others, it’s diabetes), and used clear process and outcome indicators to improve quality and gradually expand. It’s why women in Sri Lanka now have nearly the same life expectancy as women in Germany, as a government-supported program started training providers in good birth practices, then expanded into other realms of women’s health, then broadly into community primary care.
This is what we do everyday in successful primary care clinics. There is nothing terribly unique about the approach of using vertical protocols to ensure good primary care systems. When expert bodies in endocrinology write diabetes treatment guidelines, they do not expect that the majority of American diabetics will be treated by the few specialized endocrine clinics available at academic teaching hospitals. The guidelines, rather, help provide specific measurable outcomes indicators for primary care clinics that serve the majority of American diabetics (e.g., allowing clinics to observe the track the hemoglobin A1c of their diabetic patients, or calculate what percentage of them are at the blood pressure goal of less than 130/80, and determine what the problems are if they’re not achieving these accepted goals).
The melding of vertical, disease-specific protocols with broad based horizontal infrastructure development has been generally foreign to global health institutions, who have tried to construct only the disease-specific programs without using the same personnel and resources to implement other treatments for the community (generating inefficient parallel infrastructures and political competition among disease advocates in different sectors—say HIV and chronic diseases); or, equally perversely, have tried to establish primary care without meaningfully-specific protocols, training, and outcomes measurements, resulting in disastrously-substandard “Third World care” (the “cheapest care”, in contrast to the Partners in Health use of community healthcare workers who are actively-equipped and undergo rigorous training and assessments, unlike the community workers of the 1980s).
So while many now decry HIV funding or funding for any specific disease-control program, or claim that “health system development” is more important than vertical funds, what matters more is how the programs are actually implemented, and whether they include clear targets and the goal of broadening their focus to community-based needs. We cannot ignore the voices of those with on-the-ground experience and medical knowledge, or we risk deploying a dangerous rhetoric that’s in some cases closing successful programs and taking medications away from patients, even when we intend to help those very same patients. The confusion of selective care with sustainable care through a focus on MCH, and the pitting of the vertical against the horizontal as a false dichotomy, must stop now if the GHI intends to avoid repeating the mistakes of the past.