Document(s) 3 of 10

For several decades now, Tanzania -- like most other countries in sub-Saharan Africa -- has faced the dual burden of a crisis in public health and a grave shortage of resources with which to address this escalating, interlocking web of problems.

The health challenges confronting most of Africa have been well publicized: high on the list are the spread of deadly diseases and problems such as malaria, HIV/AIDS, tuberculosis, malnutrition, and anemia. Almost equally well known is that the effects of these devastating epidemics are exacerbated by conditions of poverty. Poor citizens are caught, in effect, in a vicious cycle: while their poverty makes them more vulnerable to the effects of illness and less able to afford proper treatment, succumbing to sickness in turn reduces their already-meager capacity to generate income.

Both aspects of this cycle can be seen as factors in the national context in Tanzania. While it has struggled with the onslaught of infectious diseases, high rates of child mortality, and widespread disability, Tanzania (one of the poorest countries in the world, with an annual per capita income of US $280) has until recently only been able to allot roughly US $6­8 per person annually to health care (Table 1). By comparison, according to Conference Board of Canada (2004) figures, Canada spends an annual US $2 809 per capita on health care, while annual per capita health care expenditures in the United States total US $4 819. 

Table 1. A profile of selected health-relevant indicators for Tanzania
(circa 2002)

Table 1. concluded

^a All statistics are from 2002 or represent the most recent estimate.
^b See http://www.tanzania.go.tz/census/.
^c As quoted in UNICEF (2004). 

The experience of the Tanzania Essential Health Interventions Project (TEHIP) suggests, however, that sudden injections of funding alone will not necessarily solve the health care crises facing the nations of Africa -- although certainly funding for health systems and health interventions must increase substantially over time. Nor is a panacea likely to be found in new and more effective drugs and vaccines -- although these innovations are also an important component in the battle to improve health for populations in the developing world. Instead, the project's experience makes clear that a critical missing link in reducing high levels of morbidity and mortality in developing countries is a system-level intervention: the ability to allocate health resources in strategic ways that target real and prevailing needs, and that enable health planners and individual health workers to operate more effectively at the level of the front line health facility. In other words, the efficiency of the health system (and the appropriateness of strategies in health care) is key to translating health care spending into an increase in health gains.

The policy implication we draw from this is that institutions and agencies concerned with improving the currently grim health outlook in Africa must take a more systemic approach -- turning at least some of their attention to apparently mundane matters within the health system, such as infrastructure, training, capacity building, human resources, and health planning, that form the foundation for future advances in the well-being of Africa's citizens. 

A history of hope and struggle

The story that unfolds in the pages ahead holds lessons that we believe can be applied widely throughout the developing world. At the same time, however, this is a narrative largely rooted within the specific geographical setting of Tanzania, a land of spectacular contrasts that range from the mountain forests of Kilimanjaro to arid plains, coastal deltas, and sun-soaked beaches. Indeed, these contrasts become immediately apparent by comparing the two districts within which TEHIP's work unfolded. Whereas Morogoro district is mountainous and lush, Rufiji is characterized both by a mostly dry, flat expanse in its interior and by a tidal delta on its coastline.

As enchanting as this landscape is, it has also served as the backdrop for a tragic and all too familiar story. Like their counterparts in countries across Africa, Tanzanians have suffered through a grave health crisis for the better part of a generation. While the full force of new or resurgent infectious diseases -- most notably malaria, HIV/AIDS, and tuberculosis -- has brutally rearranged the social landscape of an entire continent, national health systems have been wholly unable to meet this challenge, sometimes teetering on the brink of collapse.

In the specific case of Tanzania, it is especially and bitterly ironic that this health care crisis has occurred despite longstanding policies that had placed health care high on the national agenda. Since its launch as an independent unified country in 1964 (the product of an amalgamation of two former British protectorates, Zanzibar and Tanganyika), Tanzania, under the first postindependence government of Julius Nyerere, sought to ensure that Tanzania's citizens had access to education, health care, and clean water. Plans to provide these amenities and services centered on a new and unique social contract, whereby citizens who relocated to modern villages (typically taking the form of a collection of small hamlets) became the beneficiaries of government programs. Each village had pumped water, a school, and access to a clinic -- most often these facilities were built by volunteer labour from within those communities. The government kept up its end of the bargain by agreeing to maintain those structures and by dispatching teachers to the schools and health workers, drugs, and supplies to the new health facilities. This agreement led to new health care machinery being set in motion: medical training centres were built and large numbers of graduates poured out into the rural areas to provide the health care that the government had declared a public right.

While this momentum continued through the 1970s and into the 1980s, by the mid-80s the system was foundering. Arguably, part of the problem was that centrally planned health care management had been inefficient and unresponsive and was unable to maintain the health care infrastructure (such as the village dispensaries).

Yet another and indisputably massive factor in the erosion of Tanzania's health system was the international debt crisis that in the 1980s was creating similar disruption throughout the developing world. Economies dependant upon the export of natural resources were crippled by the twin scourges of dramatically falling commodity prices and rising interest rates that caused the debts of developing nations to increase exponentially virtually overnight. At the height of the debt crisis, Tanzania, like other developing countries, was faced with onerous -- perhaps impossible -- debt-repayment demands while its export income plummeted. With close to half of government revenues being channeled toward debt repayment -- at the expense of domestic social spending -- the negative impacts on Tanzania's health care system were dramatic and prolonged. Funds for training health staff and maintaining facilities were no longer available. Many clinics had no drugs or health supplies on the shelves. Wages were eroded by inflation, devaluation of the local currency, and by continuing austerity measures. Many health workers continued to perform their duties without receiving a salary. There were cases where facilities were operated by unqualified staff after the local clinician had died or simply left.

Early attempts to revive the system were unsuccessful. The introduction of user fees and other cost-recovery measures -- intended to infuse new funds into the system -- only served to drive more Tanzanians out of the orbit of the health care delivery system. People who had been unhappy with the low quality of services provided now became indignant at being asked to pay for those same low-quality services. Similarly, support from the international community began to wane as the outlook for health care in Africa became bleaker and "donor fatigue" began to set in.

Bold new initiatives

In Tanzania, as elsewhere in Africa, there is more hope today than existed during the darkest days of the mid-1980s to the early 1990s. Emblematic of the new ideas and new optimism that has been injected into the debate over health care in Africa is the 2001 report of the Commission on Macroeconomics and Health, which was formed by the World Health Organization (WHO) in 2000 to examine the relationships between health, development, and social equity, and to recommend measures to minimize poverty and maximize economic development. The world has also seen significant new funding and institutional muscle being applied to the problem of infectious disease, as exemplified by the Global Fund to Fight AIDS, Tuberculosis and Malaria. This major new funder was established in January 2002 as the outgrowth of work undertaken by the G-8 group of countries, leaders of African states, and UN Secretary General Kofi Annan. In addition, institutions such as the Bill and Melinda Gates Foundation, the Rockefeller Foundation, the United Nations Foundation, and the Roll Back Malaria Partnership have made health care in Africa a high priority at a time when international development programs have also redoubled their efforts on the continent. These are all positive and highly desirable initiatives reflecting a new political will to help deal with Africa's health reform challenges and representing international recognition of the need to commit the appropriate resources to that goal.

At the same time, however, echoes of the old era of structural adjustment programs -- designed to impose fiscal austerity on developing countries during the debt crisis -- continue to exert a restraining influence on national health care systems. For example, externally mandated hiring freezes still make it difficult for many countries to recruit new health practitioners necessary to staff their health facilities. Wages of health workers in many developing countries remain desperately inadequate, to the point where critically important public servants in many countries must consider other forms of work or working abroad to earn a living wage. The era of structural adjustment may be over, but the effects of earlier damage continue to cast a long shadow.

This presents a striking and disturbing paradox: at a time when major new funds are being promised for new therapies, technologies, and health interventions, the prospect is that those same funds will be channeled through weakened, fragile national health systems that remain inefficient, inadequate, and under-funded. A lack of ground-level capacity may well hamper the grand designs conceived of at the international level. Consider the potential obstacles, for instance, involved with bringing antiretroviral drugs to Africa to treat HIV/AIDS. Getting those drugs onto the dispensary shelves and into the hands of people who need them will require health systems that have information, communication, transportation, diagnostic, and human resource capacities sufficient to move the drugs to the right places, in the right numbers, at the right times, to the right people, and with the right counseling and follow-up.

TEHIP's piece of the puzzle

A central preoccupation of TEHIP has been to learn how a functioning and efficient health care system -- one with resources logically targeted at the most pressing health needs of the population -- could contribute to significant improvements in the health of the population. Essentially, the project's goals have been to help local authorities fix the gross technical and allocative inefficiencies that characterized health care delivery in two rural Tanzanian districts and, related to that, to help bring proportional spending into line with actual needs. The project has facilitated this process by encouraging an "evidence-based" approach -- that is, by promoting the use of evidence about the local-level burden of disease (as measured through mortality) and evidence on cost-effectiveness as the main determinants of how to establish priorities in health care budgeting.

This idea -- that increasing the overall effectiveness of public health services could have a major impact on the overall health of the population -- is not new. Before the effects of structural adjustment took hold in the 1980s, several influential reports emphasized the need for a focus on primary health care and strengthening of comprehensive health systems to reach the people in need and to improve health outcomes. Those documents included the 1978 Declaration of Alma-Ata (WHO and UNICEF 1978) and UNICEF's Child Survival Revolution initiative of 1982 (see UNICEF 1996) -- both of which stressed the need for equity, participation, and a multisectoral systemic approach to improving health.

After many years of crisis in the 1980s, a variation on that basic perspective re-entered the public eye with the publication of the World Bank's 1993 edition of its World Development Report (WDR93). In what appeared to be a reversal of previous Bank policy stressing fiscal restraint and drastic cuts to public sector programs, WDR93 proposed that increasing investments in health was key to economic development. It also prescribed that such investments should be based on evidence that would target and focus cost-effective interventions on the local "burden of disease" that exists in a particular ecosystem. For example, in an area where malaria accounts for 40% of the burden of disease, allocating only 5% of the budget would be inadequate for treatment and prevention. There is a natural tendency in human nature to put 80% of our efforts on 20% of the problem. In health systems, this leads to huge inefficiencies and low impacts.

Using this general principle of bringing efforts more in line with the weight of the problems, WDR93 proposed specific minimum packages of essential and primary health care interventions. The World Bank speculated that by taking this targeted "evidence-based" approach (wherein the burden of disease and cost-effectiveness become the determining factors of how budgets are spent, rather than administrative or political considerations or simple guess-work), relatively small funding increases could produce significant and tangible improvements, simply through correcting prevailing technical and allocative inefficiencies. Calculations included in WDR93 suggested that raising per capita public spending on health in low-income countries to US $12 annually -- a modest sum, but still higher than existing funding levels in Tanzania and neighbouring countries -- should lead to a 25% decline in the burden of disease. In short, WDR93 argued that, although increasing health funding is critical, the method by which those funds are allocated is also crucial to ensuring that new funding produces substantial improvements in health outcomes (Bobadilla et al. 1994).

Although the logic of WDR93 was generally well received, the report sparked very little discussion as to how this promising theoretical premise could be translated into practice. Notably absent from the report were answers to any of the "how" questions. How could local authorities get a true picture of the existing burden of disease? How could the new information be used to reengineer local health systems -- in other words, what mechanisms could be devised to allow decentralized planners to incorporate burden of disease into their work in a manageable and practical way? Essentially, TEHIP was created to answer the "how" questions. Conceived in October 1993, its aim has been to develop and test a set of simple, user-friendly tools to enable local-level health planners to plan on the basis of evidence.

The experiment has now run its course: the planning tools that evolved from several years of collaboration between local authorities and TEHIP have been put into use in two Tanzanian districts with populations totaling 741 000 -- a sample large enough to make it difficult to dismiss the results in these districts as the outcome of an "experiment" that would be difficult to replicate in "real life." Since 1997, District Health Management Teams (DHMTs) in the rural districts of Morogoro and Rufiji have been using an expanding tool kit to plan and implement health services to more precisely respond to local evidence. Concurrently, DHMTs have been assisted in their attempts to revamp their health systems by the provision of a financial top-up that brought their health budgets closer to WDR93's annual US $12 per capita target.

The results of these changes have been dramatic, with the two districts having witnessed marked improvement in health outcomes in the wake of the introduction of new planning methods and modest budget top-ups in the order of an additional US $1 per capita. In Rufiji and Morogoro, for example, child mortality fell by over 40% in the 5 years following the introduction of evidence-based planning (Figure 1). In the same period, the death rate for Rufiji adolescents and adults between 15 and 60 years old declined by 18%. Corresponding figures for districts that have not been using the planning tools -- and indeed all across most of Africa -- have been stagnant, at best, for children and increasing for adults. From neighbouring comparison districts, cross-references have been made with other contextual factors (such as differences in levels of rainfall, disease outbreaks, and health risks) that are known to affect mortality rates. Those factors do not appear responsible for the declining death rates in this case.

Figure 1. Reversing the trend in child mortality after district-level health system interventions in Tanzania.

Sources: National Sentinel Surveillance System, Ministry of Health Tanzania; Coastal Sentinel Demographic Surveillance System (TEHIP Rufiji DSS for 1999 to 2003); Central-East Sentinel Demographic Surveillance System (AMMP Morogoro DSS for 1992 to 2003); Tanzania Demographic and Health Survey 1992 for 1990 local baselines in Morogoro and Rufiji; Tanzania Demographic and Health Survey 1996 for 1995 baseline in Rufiji; UNICEF for Tanzania national under-five mortality trend ( http://www.childinfo.org/cmr/revis/db2.htm ).

The results from the two planning districts, therefore, support the earlier predictions in WDR93. Health system adjustments correlating health spending with burden of disease and cost-effectiveness do allow for significant improvement in health with modest increases in expenditures. Furthermore, we have no reason to believe that this result should be considered unique to the particular rural Tanzanian setting. The planning tools used by district health planners and managers are fully adaptable and can be used -- given the appropriate local statistical inputs -- by district health planners in other country settings.

Testing a potent idea

The process that ultimately led to the development and deployment of TEHIP's health planning tools began in October 1993, when Canada's International Development Research Centre (IDRC) convened an international conference in Aylmer, Quebec, Canada. Representatives of the World Bank, WHO, UNICEF, and other multilateral and bilateral organizations, nongovernmental organizations, universities, and ministries of health were asked to consider whether it would be possible to test the idea that evidence-based health planning could produce efficiencies that would lead to positive improvements in local health. They answered in the affirmative.

Subsequently, IDRC and the Canadian International Development Agency (CIDA) provided funding to launch what was then known as the Essential Health Interventions Project, or EHIP, (the lack of a "T" denoting that Tanzania had not yet been identified as the host country, and the Tanzanian Ministry of Health had not yet become a partner in the enterprise). The following 3 years saw a series of international meetings intended to collect as wide an array of expert input as possible and to refine the research design. The first meeting of EHIP's advisory committee, hosted by WHO, took place in Geneva in January 1994. There were 9 subsequent meetings in various cities culminating in the drafting of a "scope"' document in March 1996 (see TEHIP 1998), a call for proposals, and the final approval of research proposals in December of the same year.

This process of discussing and defining the parameters of research was much more exhaustive than is the norm -- most development initiatives would have much stricter timelines attached and less latitude to deal with theoretical concerns. However, there was a general feeling that since the program aimed to test an idea and approach that were potentially groundbreaking, and likely contentious, it was imperative to take time to "get it right" by exploring all the potential implications and pitfalls from the outset and by thoroughly thinking through the design model. This development and design phase of the project relied on high-level international and domestic expertise. When fieldwork began in Tanzania, however, direction came almost exclusively from local authorities and most funding went into the modest annual per capita incremental health budgets for the 741 000 residents of the two districts. The daunting challenge facing these small groups of district managers is underscored by that population figure. Expressed another way, 741 000 is a population greater than that of 66 (close to one third) of the world's countries (the population of Guyana, for example, is 705 000). Furthermore, this large population is spread over an area almost the size of Switzerland, with varied and often difficult terrain. Given the scope of their challenge, it is clear to us that the accomplishments of the district teams could indeed hold lessons for entire countries.

Understanding how this work led to actual changes in health care delivery in Rufiji and Morogoro requires recounting the particular stages in the project's life cycle. The first phase, noted above, was one of broad-based consultation that took place between 1993 and 1996. In 1994, the Tanzanian Ministry of Health responded positively to an invitation to participate -- triggering the selection of Tanzania as the research site, the engagement of the Ministry as a partner in the project, and the reorientation of EHIP to TEHIP. From this point, discussion began on how the research program could be made compatible with the specific circumstances of health care delivery in Tanzania.

Tanzania was interested in participating in the program largely because it had embarked on a program of health reforms receptive to decentralized, evidence-based planning and wanted to know how it could be implemented. A cornerstone of the Tanzanian health ministry's plans to reform the health system was the devolution to local authorities of health care delivery and management responsibilities, accomplished through the creation of District Health Management Teams (DHMTs) (see box: "More responsive planning through DHMTs"). TEHIP -- with its emphasis on decentralized control over health budgets -- would provide an opportunity to test whether the strategy being considered by the Tanzanian Ministry of Health was sound.

When the design phase concluded in December 1996, a second phase emerged, which involved setting up Demographic Surveillance Systems (DSSs) to collect comprehensive data on mortality to calculate the prevailing burden of disease in Rufiji and Morogoro. Although originally intended to monitor the impact of evidence-based planning, this data soon became part of the raw material TEHIP would use to create the "tool kit" intended for use by health managers.

Another aspect of the work involved putting the finished tools -- as they became available -- into the hands of the DHMTs. Once the project was designed and the tool kit progressively assembled, TEHIP scrupulously avoided interfering in how district health authorities used those tools or spent their money. Furthermore, additional funds flowing to Rufiji and Morogoro only consisted of top-up funds amounting to approximately US $1 per capita annually, which was needed to bring the districts closer to the minimum spending levels suggested in WDR93. (DHMTs could spend the additional funds not only on strategic interventions but also partly on initiatives that would make the health system more functional -- for instance, by upgrading management and administrative skills, increasing internal transport and communications capacity, or repairing substandard facilities). The only other advantage DHMTs enjoyed -- besides the additional funds -- was increasing access to the developing tool kit.

More responsive planning through DHMTs A key component of Tanzania's health sector reforms was the establishment of District Health Management Teams (DHMTs) in each of the country's 123 councils. Local health services had previously been planned centrally by administrators in the capital. However, the Ministry of Health felt that devolving planning and management authority to local teams -- composed of members with complementary skills and multiple areas of expertise -- would lead to policies and administrative practices that better suited local needs and conditions. Budgets, for example, could be allocated by DHMTs on the basis of prevailing local mortality rates rather than moving automatically in step with national health priorities. To perform these new functions, however, DHMTs would need training, tools, and the systemic support to help them determine the actual local prevalence of disease, to allocate funds appropriately, and to meet national standards of practice, reporting, and accountability.

Complexity anchored by fundamental questions

TEHIP evolved into a uniquely multifaceted and complex program integrating both research and development functions, employing the skills of Tanzanian researchers and practitioners, and examining a wide array of specific health care questions. During the course of this evolution, however, project workers would always come back to a series of questions and principles that defined the scope of the work and pointed toward its goals and potential outcomes.

One of the major results of those early years of research design and thorough consultation was the formulation of three interrelated questions that would serve to keep the project on course as its work proceeded:

• How and to what extent can decentralized district health plans be more evidence-based?
• How and to what extent can evidence-based plans be implemented by decentralized health systems?
• How, to what extent, and at what cost can such evidence-based plans have an impact on population health?

It is significant that, given the repeated use of the phrase "how and to what extent," these questions contain within them qualified rather than definitive statements. This reflects the project's assumption that health systems could never be managed entirely on the basis of evidence -- political factors, subjective judgments, and windfall opportunities will always intrude. Still, the intent was to ensure that evidence -- detailing the true burden of disease at the local level -- becomes as powerful a factor as possible in the allocation of health care resources. The aim was to move away from the norm of health care policy formation in developing countries, where spending is often dictated by numerous secondary factors that have little to do with the prevalence of disease or with a logical plan for maximizing health. Prominent among these factors is bureaucratic inertia -- the pressure to simply replicate the proportional allotments of last year's budget, added to or subtracted from the current year's financial picture. The paradigm advanced by donor agencies is another persuasive factor. Few health officials find it easy to turn down external funding aimed at, for example, a specific disease -- even if that particular disease is not a significant concern in the area and addressing it could divert attention and resources from more pressing population needs.

In addition to this key principle that evidence should guide health policy and resource allocation, there were other operational principles that emerged during the first consultative phase of the project. One such principle was that resources should be controlled by local authorities rather than by a central office. Another was that district management teams should be able to allocate those resources. These principles were reinforced by circumstances within Tanzania and by emerging policy trends in the field of international development. For instance, as we've already noted, Tanzania had committed itself to decentralization as part of its package of health care reforms. Donor agencies were also changing their approach: many had abandoned the practice of stipulating that funds should be spent in specific ways. The new practice -- adopted by some national development agencies -- of providing funding to be allotted by the recipient governments according to their reading of the most pressing need is known as the "sector-wide approach" (SWAp). The project's practice of providing DHMTs with a basket of funds -- which the teams would then divide to respond to the picture that emerged from the evidence -- was very much a precursor of the SWAp strategy for decentralized district health basket funding.

The need for an integrated approach

TEHIP was also guided by another principle that -- although it was not spelled out initially in the formal design of the project -- became clearer as time elapsed and as field experience accumulated. As the project's base of operations moved to Tanzania, it became increasingly obvious that any set of prescriptions for reforming health care delivery would have to keep the community-level health worker foremost in mind. The reality at the village level is that a dispensary or health centre will have one or perhaps two workers responsible for the care of thousands of people and for every aspect of the facility's operation. These one or two people shoulder a tremendous responsibility -- the success or failure of programs designed elsewhere comes down to how well those individual health workers can incorporate those imported plans into their daily routine. It therefore became obvious that any new initiatives should be seen as part of a complete package that makes organizational sense for the overburdened health worker. The more complicated and taxing the individual remedy becomes, the less likely that those remedies will succeed.

In response to this reality, our approach was to look for what we would call "integrated" solutions to problems. The integrated approach took many forms. It influenced methods of treating patients by suggesting that health care should address the overall condition of the patient and should not be directed in isolation toward diagnosing and dealing with specific individual diseases. It influenced the reorientation of health systems by stressing innovations and initiatives that could be integrated into the routines of community health workers and managers. It also had a major influence over how TEHIP as an organization was structured. 

In the next chapter, which describes the development and deployment of the tools, we will see numerous examples of this integration-oriented approach at work. For example, a guiding idea was that diagnostic tools should be kept simple. TEHIP was aware that it would be impractical for busy managers and health workers to use tools that presented too much information or burdened them with unnecessary complexity. The tools steered district priorities toward health interventions that formed "bundles" or "packages" -- which had the potential to deal more efficiently with overlapping or coexisting health problems -- as opposed to "stand alone" health interventions that dealt with single diseases on an individual basis. The rationale was that complementary and integrated sets of treatments would be easier for front-line workers to use and, therefore, were more likely to be successful.

This integrated approach contrasts with what is often termed the "vertical" model. We use this term to describe what we see as top-down efforts to control particular diseases, one at a time. Such programs often involve directives from high levels of authority, issued without adequate consideration for how the community-level health worker can integrate these new activities and responsibilities or how new initiatives will fit in with the prevailing local burden of disease or the structure and capacities of the local health system. We see the vertical approaches and campaigns as gaining momentum in an era when large amounts of money are flowing into international disease-control programs. This new funding creates tremendous pressure to be able to demonstrate results quickly and thereby creates a bias against taking a longer term, more systemic approach to health care.

Indeed, the integrated approach seems to run contrary to several established patterns that exert a powerful influence over the formation of health policy. In academia and within bureaucracies, there are few incentives to take an integrated approach that keeps the needs of the peripheral, front-line health worker and local manager in mind.

This bias against an integrated approach has a clear impact in the field. For example, the typical way of training health workers has been to take them out of the clinical setting to provide them with a few days of training on one particular disease, on an individual basis. TEHIP has taken the contrary stance, believing that it is better to train workers systemically -- training them to look at a range of signs and symptoms as indicators of a number of possible conditions. This approach makes particular sense in an environment where patients often suffer from more than one illness and where common symptoms may point to the presence of more than one disease.

Another example of the integrated mentality at work is internal: the connection between the project's "research" and "development" functions. Research to establish the burden of disease in the districts, to gauge the ways in which health budgets are allocated, and to determine how best the public can interact with the health system was undertaken simultaneously with the development and deployment of the tools. Both aspects of the program were meant to be complementary and mutually reinforcing. This is very different from the standard model in which research and development remain distinct entities with different staffing, separate budgets, and little communication between the two. Normally, after research has suggested a particular policy direction, additional funds must be raised to construct a pilot project and thereby translate the new ideas into actual change. Since TEHIP was a well-funded "research and development"' project operating within the context of a functioning, living health system, it could roll those normally separate functions into one package: providing DHMTs with access to relevant research and then enabling them to act upon those findings annually in subsequent planning cycles. Research was undertaken, tools for the DHMT managers were developed, and the tools were used to change the way the health system was managed and health care delivered at the community level -- all of this occurring within the same project cycle. In fact, the cycle continued after the tools had been put into practice. Ongoing research allowed for the creation of a kind of "feedback loop" that enabled practitioners to examine whether the tools and systemic changes were working, to refine and modify their developments, and to consider the development of new tools at any stage as the need for them became apparent.

Efficiency leads to equity

A final philosophical note to add, before we examine specific contributions, is that the project's work must be seen in the context of the sometimes-competing concepts of "efficiency" and "equity." The project's emphasis on efficiency -- its focus on maximizing the cost-effectiveness of health care spending -- should not be viewed as some cold-hearted accounting exercise. Rather, it was a means to an end, adopted specifically as a way of making the system more equitable.

There are two ways of assisting the poor through public spending. One way is to concentrate on the "distributional" issues -- for example, by creating programs that target "the poorest of the poor." Another means is to focus on the "production" issues -- in this case, that meant retooling the health system so it became capable of producing higher levels of health across all segments of society.

In other words, the reforms aimed to buttress the "universal" nature of this social benefit. Taking this approach made eminent sense in a situation -- such as existed in Rufiji and Morogoro -- where an inefficient health system produced consistently poor results and where most people were uniformly poor. Now, all citizens can go to the local clinic with a reasonable expectation of getting appropriate treatment for the most important population health problems. In reality, though, the biggest impact of increasing the efficiency of these universal health services is on the poor. This is because, in a country where the majority of people are poor, a minimum package of essential interventions addressing the diseases that account for the largest segments of the burden of disease will by definition benefit the poor. It is the poor who suffer the most from these diseases (such as malaria), so it is the poor who have the most to gain. Conversely, since the diseases that disproportionately affect the rich (such as cancer) did not appear as significant in burden of disease statistics, those diseases did not absorb significant resources within those intervention packages. Therefore, one effective consequence of such an approach is a transfer of public health care resources from the rich, who can afford private health care, to the poor, who rely exclusively on the public system.

Still, there is a small minority of rural poor that has not been positively affected by health care changes in Rufiji and Morogoro. Now that it has been demonstrated that "health production" issues can be effectively addressed, policymakers in Tanzania may well want to return to unfinished business and consider adopting a much more deliberate focus on equity -- that is, to pay "the equity premium" to widen the distribution of health care to those who still have not benefited from the ongoing health reforms. Our use of that phrase "the equity premium" acknowledges the reality that reaching the most marginal or remote subset of the population will not likely be accomplished in a "cost-effective" manner. Reaching this difficult-to-reach subset (let's assume it comprises 10% of the population) may cost as much as extending effective health services from 30% to 80% of the population. The extra cost does not change the fact, however, that achieving greater equity in the distribution of health care benefits is a desirable social goal that may be worth the additional expense. Pursuit of such a goal will be more effective when added to a system that has corrected its gross inefficiencies.  

Document(s) 3 of 10