Michael D. Conway, Srishti Gupta, and Srividya Prakash
Web exclusive, December 2006
The deadly trio of HIV/AIDS, tuberculosis, and malaria kills nearly six million people a year, most of them in poor countries. Fighting and preventing such diseases is difficult anywhere, but particularly in developing countries, which often lack the money, institutions, infrastructure, or even political stability needed to cope. Indeed, as the global health community has come to recognize that no single entity can contend with such maladies effectively, more than 70 health alliances have been formed in the past decade. These global partnerships link the efforts of the public, private, and nonprofit sectors to focus much-needed attention, expertise, and money on a given problem—say, malaria or poor nutrition.
Evidence suggests that such collaboration works.1 Partnerships not only allow donors and governments to undertake larger-scale (and higher-risk) activities than they could by themselves but also bring visibility to the problems and thereby attract more funding. Furthermore, partnerships can help countries coordinate efforts and avoid the duplication of investments and activities that often occurs when well-intentioned nongovernmental organizations (NGOs) and other groups simultaneously address the same cause or tackle it in disparate ways.
However, the partnerships are far from perfect. As the development community knows, the introduction of a windfall of concentrated resources (whether money, pills, or advice) into an impoverished country’s resource-strapped health system can be overwhelming, counterproductive, and even harmful to the country’s existing health plans. To determine the extent to which such consequences actually affect the recipients, we interviewed some 350 people in 20 countries where global health partnerships are active (Exhibit 1).
2 The respondents included a mix of government ministers, multilateral and bilateral development agencies, NGOs, district health-management teams, and local health facility personnel. The study found that health partnerships often provide inadequate support for implementing their plans, burden countries with parallel and duplicative processes, and fail to communicate adequately with their partners—a factor that exacerbates other problems. However, the study also suggests ways for these organizations to alter their approaches in order to serve developing countries more successfully and thereby realize a greater share of their considerable promise.
The power of partnership
Global health partnerships are organizations that bring together groups—including governments, donors, NGOs, and a variety of private-sector representatives—into a formal, collaborative relationship dedicated to the pursuit of a shared health goal. Typically, the partnerships work directly with the governments of affected countries to develop and implement plans for aid. During the past decade, such partnerships have grown to become the predominant organizational model for addressing the complex health problems of low- and middle-income countries. Today the partnerships provide everything from technical assistance and R&D to advocacy and financing. Major funding partnerships, such as the Global Alliance for Vaccines and Immunization (GAVI) and the Global Fund to Fight AIDS, Tuberculosis, and Malaria, have attracted some $8.1 billion in the past five years and awarded grants totaling $3.6 billion (Exhibit 2).
These efforts can have dramatic results. In Niger, for instance, the Global Fund helped distribute more than two million insecticide-treated bed nets to all mothers with children under age five. The Global Fund estimates that this effort will protect some three and a half million children against malaria. Likewise, from 2000 to 2005, the vaccination efforts of GAVI and its partners are thought to have prevented nearly two million deaths from bacterial meningitis, caused by Haemophilus influenza type B (Hib disease).
Besides providing money and attention, the partnerships encourage countries to craft smarter policies. In Vietnam, for example, the combination of money and procurement expertise coordinated by GAVI allowed the country to revamp its patchwork program of infant vaccination. Today Vietnam vaccinates newborns for hepatitis B within 24 hours of delivery. Likewise, in Zambia, a program encourages local clinics to monitor vaccination rates and report data to GAVI and other donors. Better data collection has helped Zambia’s government strengthen its district-level vaccination programs and improve its own outreach.
Furthermore, global health partnerships raise the profiles of NGOs and help them scale up their efforts. This can mean the difference between life and death in countries such as Bangladesh and Zambia, where most health care isn’t delivered in public institutions.
First, do no harm
The benefits of health partnerships are not without costs, however. To be sure, some unintended consequences cannot be avoided. A country’s health system, for example, must serve the wide-ranging needs of an entire population in an integrated way; by contrast, global health partnerships focus on particular diseases or even facets of diseases (say, vaccine R&D or tuberculosis treatment). The infusion of large amounts of targeted resources into impoverished countries increases the potential for trouble, since their weak systems and institutions prevent them from absorbing help effectively. We identified two significant negative consequences that countries face in their interactions with health partnerships: inadequate support for implementation and the duplication of effort. A lack of effective communication on the part of health partnerships in their dealings with partners and recipient countries intensifies each of these consequences.
Inadequate support for implementation
For an impoverished country beset by disease, the expertise needed to craft and implement health programs can be as valuable as money. Global health partnerships therefore work closely with international agencies such as the World Health Organization (WHO) to determine the most effective courses of treatment and the best ways of using new technology. But we found that the recipients of aid often struggle to carry out the policy decisions that partnerships mandate as prerequisites for funding. What’s more, when partnerships shift their policies—for example, by adopting new, more expensive drug therapies—they inadvertently foster an environment of uncertainty and mixed signals. Recipients end up feeling, in the words of one interviewee, as if they were “being forced, without discussion.”
The problem goes beyond hurt feelings. When stakeholders in affected countries don’t know the rationale behind a given policy or don’t understand the trade-offs and logistics involved, transition planning and implementation become more difficult. Worse still, in some cases we found that government officials and local NGOs felt that the policies themselves were inappropriate given the financial constraints and limited health systems in these countries.
Consider Burkina Faso, which like much of Africa struggles in its fight against malaria with drug resistance and the cost of treatment. Domestic research showed that chloroquine (a relatively inexpensive therapy) was 90 percent effective in treating malaria in some districts there. Yet the Global Fund, on the recommendation of the WHO, mandated a change to the artemisinin combination therapy, which is far more effective but about 20 times more expensive. Burkina Faso, which had already applied for a chloroquine grant, duly changed its policy and applied for money to cover the difference. In the six months before funding was approved, the country was on its own. This gap, we found, damaged support among its policy makers and raised doubts about the new policy’s financial sustainability. We encountered similar sentiments in Angola, where Ministry of Health officials expressed frustration over requirements that they change their malaria program without any assurance of clear financial sustainability, planning, or support.
Worryingly, we found that such complaints extend to implementation assistance more broadly (Exhibit 3).
Countries invest heavily in applications for funding (efforts supported by health partnerships and their affiliates) but often can’t execute plans once the grant writers leave for the airport and the check arrives. Health ministers in several countries cited a broad lack of adequate support from partnerships as a significant problem. In Laos, we heard that health partnerships “tend to send people in for intense bursts of activity and leave reports with a lot of 'shoulds’ but not a lot of 'hows.’”
Without such hands-on support, we found, countries make unrealistic assumptions and generate overly ambitious targets or, worse, targets that are too modest. The results can be disheartening. In Tanzania, in-country partner organizations asked the government to scale back its plans to ask for HIV/AIDS funding—it had originally proposed $1 billion—because “they did not think the Global Fund had that kind of money and that the country could not absorb that much money.” Instead of asking that funds be disbursed over a longer time frame to give the country time to implement new programs, Tanzania asked for less money. The result was that only 10,000 people were treated before it ran out. Meanwhile, according to an estimate by the joint United Nations Program on HIV/AIDS (UNAIDS), at the end of 2003, 1.6 million people were living with HIV in Tanzania, where life expectancy is now an appalling 45 years.
Moreover, we found that country coordination forums—the mechanism that global health partnerships use to draw together the activities of donors, NGOs, and country-level health officials—are inadequately structured and too numerous to be effective. In Angola, for example, four coordinating bodies focus on HIV/AIDS, and none meets the country’s needs. What’s more, low-income countries lack senior managerial talent, so the same officials often serve on many committees, spreading already scarce talent even more thinly. (This proved to be true in Bangladesh, Burkina Faso, Tanzania, and Vietnam.) The costs of poor coordination extend to the frontline district health-management teams and NGOs. In Zambia, for instance, our interviews suggest that many of the field activities are carried out without the knowledge of either national or district-level planners.
Duplication, duplication, duplication
It’s no surprise that the countries where donors operate have deficient national health systems. In a place like the Democratic Republic of the Congo, still emerging from a five-year civil war that killed nearly four million people, it’s perhaps surprising to find any functioning system at all. Global health partnerships enter such countries prepared for problems with infrastructure, procurement, logistics, health information, financial systems, and human resources; poor governance and corruption are commonplace as well. Still, we found that a desire for expediency often tempts health partnerships to impose one-size-fits-all processes, which often duplicate—or undermine—a country’s existing procedures.
One problem area is financial planning. In Bangladesh, Ethiopia, and Vietnam, interviewees reported that scheduling clashes between health partnerships and governments have led to duplication, confusion, and misalignment. In Bangladesh, the Ministry of Health even felt compelled to redraw its annual operational health plan in the middle of the year to synchronize it with the grant calendars of global health partnerships. To be sure, the magnitude of the funding involved makes such choices palatable to many countries, but time and effort are wasted.
Further, the money that health partnerships spend in a given country often isn’t under the control of its government and is instead disbursed through duplicative mechanisms. This approach is wise in some cases, given the high levels of corruption and poor governance found in much of the developing world. But the practice also discourages transparency, making it difficult for countries to plan beyond short-term expenditures, to push donors to finance program gaps that countries have identified, or even to develop the ability to undertake such efforts. Indeed, in Angola, where most donors operate outside the scope of government ministries because of corruption, the national and provincial authorities have no way of knowing how much money flows through individual provinces. As a result, the provision of services is patchy, and successes are difficult to track, sustain, or ramp up.
Duplication is prevalent in monitoring and evaluation as well. Despite the importance of collecting surveillance data in the field to assess the efficacy and reach of individual health programs, some health partnerships inadvertently duplicate these efforts either by using several NGOs or country-sponsored programs to gather the same data in different formats or by using inconsistent indicators or time frames. Unnecessarily diverting a country’s scarce management time away from broader health system issues can undermine national planning and execution efforts.
Finally, some interviewees reported that health partnerships encourage countries to use procurement systems (say, for bulk purchases of vaccines) that duplicate those a country already has. Although this practice has succeeded in countries where national procurement systems are weak or nonexistent, such as Angola, the duplication can otherwise have harmful effects. In Burkina Faso, where leaders felt pressured to switch to the Stop TB Partnership’s Global Drug Facility for bulk purchases of tuberculosis drugs, this move actually reduced the country’s bulk-purchasing power.
A failure to communicate
Both problems—the inability of health partnerships to provide adequate support for the implementation of programs and the tendency to duplicate efforts—are amplified by poor communication. The resulting confusion and suspicion often lead to wasted time and lessen the effectiveness of all parties involved.
Communication among health partnerships, their partners, and the countries they help is deficient in two ways. First, we found that countries lack channels for delivering feedback to partnerships. Worse, many countries feel powerless even to try. One of the most common misunderstandings is the idea that countries can’t broach the subject of how partnerships might tailor their approaches to meet a country’s needs. As an official in Ghana told us, “We changed our [approach] to accommodate the Global Fund. We did not think about asking them to change—that would be impossible.” Moreover, when officials are willing to ask partnerships to be flexible, those officials often don’t know whether the partnership officials they meet are senior enough to discuss policy issues. Further, the time that elapses between a question and the answer (often several months) is hardly conducive to effective communication.
A second deficiency is that in-country development partners don’t feel sure of their roles and responsibilities in relation to the work of the health partnerships. To some extent, this problem is inevitable, since the partnerships tend to operate with small administrative staffs and keep a minimal presence in the countries themselves. When partnerships lack a dedicated “country face” or fail to structure clear agreements with partners, the countries and partners become frustrated. As a partner in Vietnam noted, “We are simply unpaid workers of [partnerships] like the Global Fund and GAVI. While there is more and more work, our staffing capacity has not been increased at all.” Poor communication also increases the likelihood that effort will be wasted. In Tanzania, for example, the partnerships’ inability to streamline the monitoring and evaluation requirements of their partners into one report left the job to district medical officers. Such officials, we found, spend a substantial amount of their time writing reports (Exhibit 4).
Perhaps the worst result of poor communication is that it leads to mistrust and to the propagation of myths about the partnerships’ intentions and policies. One Asian country’s officials told us that partnerships consider African countries to be a higher priority than Asian ones. In Ghana, we heard that the partnerships “probably already knew which countries would receive money before anyone applied” and that “they should have just told us that Ghana was not on the list.” Similar misunderstandings occurred with respect to which diseases would receive funding.
Making partnerships stronger
The countries where global health partnerships operate have some of the harshest health care environments imaginable. Local health systems are inadequate, opaque, and underfunded. Congo and Ethiopia, for instance, each spend about $5 a head on health care each year to cover a combined population of nearly 136 million. Despite such challenges, we believe that health partnerships have opportunities to reduce the costs they impose on recipient countries while enhancing the benefits they bring about.
Give support where it’s needed
To ensure that recipient countries have enough resources to use the partnerships’ money effectively, the partnerships must begin to let countries lead discussions on the timing, pace, and scale of a technology’s adoption. Countries should, for instance, be encouraged to phase in new treatments at the local level, so that the neediest people can be cared for without sacrificing gains made elsewhere. Health partnerships could support these efforts by drawing on their experience and connections to help local leaders determine how much money would be appropriate and by encouraging them to spend it effectively. The exercise would do more than make individual projects more effective; in fact, by stimulating demand for assistance, the process could help countries begin to build desperately needed management capacity, as well as physical and technical infrastructure.
As for improving the coordination between partners at the country level, health partnerships must take full responsibility—and be held accountable to countries when they fall short. Partnerships should work more closely with countries to leverage existing coordinating bodies and earmark some of the grant money for administrative purposes. Memorandums of understanding can spell out roles and responsibilities.
Complement, don’t compete
Partnerships must tailor their approaches, requirements, and processes so that they can better serve the needs of individual health systems. To avoid duplicating efforts—and thereby overburdening countries—global health partnerships should look to one another for support. With better collaboration, partnerships could work with countries to strengthen the unified multiyear health plans that many of them are striving to create, instead of the disease-specific plans seen today. A broader plan could cover priorities, programs, infrastructure requirements, and expected financial flows. Moreover, the arrangements could embrace all relevant stakeholders, from a country’s national health system to the private sector and the NGOs. Partnerships should also coordinate their reporting requirements so that local officials need only create one report for a given disease area rather than the multiple reports demanded of medical officers today.
Make the connection
Health partnerships should work quickly to boost the quality of communication among themselves, countries, and NGOs. Indeed, they must act soon in order to prevent the problems from diluting the quality of their programs and hardening negative perceptions held by the developing world’s leaders. First, partnerships should increase the size and quality of their administrative staffs to ensure that countries get prompt and appropriate attention. Second, they should designate lead partners within countries (for instance, an NGO or a multilateral agency, such as the WHO) and develop country-specific agreements with the lead partners so that responsibilities are clear. Countries could then look to these local experts, not faraway administrators, for information and hands-on support.
Just as poor communication exacerbates other problems, improved communication could have a positive, catalytic effect and enable health partnerships to serve countries better. By sharing information, for instance, partnerships could develop an easily accessible database of experts and technical providers to meet pressing—and recurring—needs, such as installing and upgrading IT systems for health care, opening new clinics, and designing advocacy campaigns for patients.
Global health partnerships are an innovative and successful way to combat the afflictions that most bedevil the developing world, but they’re not perfect. They must work more closely with the governments of affected countries—and with one another—to maximize their considerable potential.
About the Authors
Michael Conway is a principal in McKinsey’s New Jersey office, where Srishti Gupta is a consultant. Srividya Prakash is an alumnus of the San Francisco office.
The original research for this work was conducted by a larger team, which included Doan Hackley, Paul Jansen, Pooja Kumar, Melissa Lau, and Zeryn Mackwani, in addition to the authors. Melissa Lau also contributed heavily to drafting the final report. Contributions from the team were invaluable in the shaping of this work.
Notes
1 See, for example, Mark Pearson, Economic and Financial Aspects of the Global Health Partnerships, UK Department for International Development, 2004.
2 This assessment was undertaken in the summer of 2005, in partnership with the Bill & Melinda Gates Foundation. The study focused on the 20 countries shown in Exhibit 1. We chose them because they represent a variety of regions, population sizes, and levels of public-health spending. Moreover, at least two global health partnerships were active in each country, and all had received grant money from both the Global Alliance for Vaccines and Immunization (GAVI) and the Global Fund to Fight AIDS, Tuberculosis, and Malaria.
