Document(s) 11 of 12

As TEHIP approached the end of its 8-year lifespan, the TEHIP team came to realize that the most difficult work was, in a sense, just beginning.

The discoveries of the past decade all pointed at the new question “what now?” The compelling evidence showing that targeted investments in health systems can dramatically lower mortality would mean little if international agencies and regional and district health planners remained unaware of these findings. And the mechanisms that had been developed to achieve practical gains on the ground — budgeting and planning tools, DSS systems to gather reliable health data, management cascades, etc. — would have minimal impact if they stayed on the shelf and out of the hands of the people responsible for local health planning and delivery. With TEHIP set to fade into history, it became increasingly clear that special efforts were required to ensure that advances made through the “special case” of TEHIP became commonplace features of day-to-day health care delivery over a wider geographic span.

Certainly, the political will to implement meaningful changes to health systems was evident both within Tanzania and beyond its borders. In 2003, delegations from Kenya, Ghana, and Uganda came to Tanzania to observe the progress that had been made in the two test districts and to investigate whether the TEHIP approach could be adapted to their countries. Within Tanzania, meanwhile, health officials were convinced that the TEHIP program should be scaled up nationally. If the results could be replicated outside Rufiji and Morogoro, they calculated, then the health benchmarks set down in the Millennium Development Goals would move within the country’s grasp.

Early scaling up

In fact, elements of the TEHIP package had been scaled up well before the end of the project, as a distinct entity, was in sight. Since TEHIP adopted an integrated research and development approach (with the “D” side of the “R&D” amalgam geared to quickly incorporating important research findings into practice) it had been possible to accommodate Ministry of Health (MoH) requests that aspects of a national rollout should proceed before all the long-term project data was in and analyzed. Rigorous scientific evaluation had conclusively demonstrated that certain treatments or interventions were medically safe and effective (although this verification, obviously, was not required before rolling out nonmedical interventions such as the TEHIP tools). Furthermore, TEHIP’s preliminary findings had made a convincing case that implementing particular interventions under real-life (and not experimental) conditions could bring down death rates in other districts.

And so — based on information produced in the demonstration districts as early as 1997 — the Ministry had decided to scale up the use of Integrated Management of Childhood Illnesses (IMCI) in 1999, to launch a national campaign promoting the use of insecticide-treated bednets (ITNs) in 2000, and to alter its antimalarial drug policy (in line with evidence from TEHIP and other data sources) in 2001.

The MoH decision to extend the sector-wide approach (SWAp) budgeting strategy beyond the two test districts was also crucial. This created decentralized District Health Baskets across the country, effectively doubling health spending and enabling districts to apply the appropriate funds to their most pressing health concerns.

The staggered pattern of mortality rate declines in the test districts and in the country as a whole show that moving forward on the rollout of proven interventions, while the TEHIP experiment was still in progress, had been a fruitful decision. While under-5 mortality began dropping in the TEHIP districts in 1998 (eventually reaching a reduction of more than 55% by 2003), statistics reveal falling under-5 deaths nationwide beginning in 2000 (reaching a 40% decline by 2005). The lag of only a few years shows how the Tanzanian government was able to realize broad-based gains by acting, in part, on TEHIP research findings while work in the test districts was still ongoing.

A prototype to inform a national scaling up of TEHIP products was also initiated in advance of TEHIP ’s closure as a project in mid-2004. In 2003, the United Nations Foundation (UNF) and IDRC funded a 3-year initiative that would move the use of TEHIP tools and strategies beyond the two test districts and into two entire regions — effectively scaling up from 2 to 11 districts. Staff from the Ministry of Health’s Zonal Training Centres (ZTCs) were engaged to produce manuals and other materials as the basis for training district health managers in the use of the tools and front-line health workers in the delivery of new packages of clinical interventions.

This exercise was to be a model for national-level rollout; one of its goals was to identify mechanisms that would aid in the countrywide scale up. It also represented the first time that TEHIP’s machinery would be deployed without any involvement by researchers. TEHIP had accomplished its research goals, and it was time to move the resulting products into daily use by health workers in ordinary health facilities across Tanzania — to make those products an integral, functioning part of the country’s health infrastructure.

Maintaining momentum

Still, despite this early partial rollout, the approaching closure of TEHIP in 2004 caused some trepidation. There was a realization that concerted and focused action would be required to sustain the momentum that TEHIP had brought to health system revitalization. If TEHIP wound down in the way that development projects typically do — with staff simply packing their bags and moving on to the next challenge — then much of that momentum would surely dissipate. Productive partnerships would be fractured as team members moved on to new jobs and new tasks.

The loss of TEHIP ’s budget also posed an obstacle to health officials seeking to institutionalize the lessons of the project, since it was evident that a fresh infusion of resources would be needed to roll out the TEHIP model countrywide, and to train a multitude of health workers who needed to acquire new skills.

There were also fears that TEHIP ’s winding-down would trigger a loss of “institutional memory” that would make it more difficult to communicate what had been learned in Rufiji and Morogoro. Researchers might be expected to champion the role of strengthened health systems by publishing papers in scholarly journals, but who would transfer the lessons of TEHIP ’s years of experimentation to the people responsible for actual service delivery throughout Tanzania, and who would advise officials from elsewhere in Africa seeking to adapt and replicate TEHIP ’s health gains in their own countries?

Enter the “exit strategy”

Facing these potential setbacks, IDRC and the Tanzanian MoH agreed on a new plan to build on the TEHIP legacy and propel health system renewal further onto the global health agenda. In early 2004, shortly before TEHIP closed its doors, IDRC approved funding for the development of an “exit strategy” that would ensure the orderly transfer of capacities and responsibilities from the project to local institutions, while advancing the training of personnel through the Zonal Training Centre system and encouraging cooperation and coordination among East African states.

Health officials in those states subsequently signaled their support for this policy thrust. In early 2005, the East African Community’s Health Ministers endorsed a regional approach linking research evidence with policy development — a statement that led to the ongoing planning for the Regional East African Community Health (REACH) Policy Initiative, designed to strengthen health systems and address cross-border health concerns such as avian flu and Ebola. Then, in mid-2005, Tanzania’s Ministry of Health confirmed its commitment to accelerate evidence-based health planning by accepting a plan drawn up by an independent consultancy (funded through the TEHIP Exit Strategy) for strengthening and decentralizing training to enable health workers to participate in TEHIP-style health delivery.

Viewing the full panorama of initiatives arising from the exit strategy, we see a wide span of activities, ranging from the visionary (e.g., planning for the REACH regional health policy initiative and an ambitious revamping of the ZTC system) to the apparently mundane (such as the transfer of TEHIP staff and TEHIP-developed infrastructure to national institutions). Yet each of these components — big and small — made possible the complex task of moving the gains realized in the two-district laboratory of Rufiji and Morogoro beyond those confines and into the wider world.

A well-defined and well-funded exit strategy is a rarity among development projects. It became part of the TEHIP story, however, in response to a realization that the widespread adoption of new methods, tools, and outlooks — even those whose efficacy had been proven convincingly — is not something that would happen by itself. Moving from test case to routine practice would require significant expenditures of time, resources, and thought. Careful planning would be needed to overcome institutional inertia (the reflex to keep doing things the way they have always been done), and to make sure that every part of the system was up to the larger task and that all those components could function together.

IDRC’s decision to provide significant funding (approximately CA $2 million) displays an acceptance of the idea that institutionalizing the TEHIP findings would not be an automatic process but something that had to be actively promoted. This realization was shared by other institutions, such as the Canadian International Development Agency (CIDA), which in November 2005 provided CA $7 million to extend the TEHIP benefits to the rest of the country. Happily, these and other agencies’ determination to not simply let the project end (with the risk that its innovations would fall out of use) paid tangible dividends. Building bridges with national, regional, and international institutions appears to have contributed to a new focus on health system strengthening at all those levels. The expansion of evidence-based planning within Tanzania, meanwhile, is moving hand-in-hand with continued improvements in health indicators across the country.

With hindsight, we can categorize the various initiatives undertaken after the TEHIP project closure under the following five thematic banners.

1. Dissemination

TEHIP placed significant emphasis, during its lifespan, on communicating its research findings to practitioners, politicians, communities, donors, and international health authorities through vehicles such as the TEHIP News and regular bulletins from the DSS systems. This approach continued during the exit strategy era. PowerPoint presentations, small brochures, newspapers articles — most of them centered on simple, easily understood charts and graphs — were aimed at crucial players in health care reform who likely had neither the time nor the academic background to digest scholarly papers in scientific journals.

After the TEHIP data was in and analyzed, it became essential to ensure that the project’s lessons were widely circulated and well understood. Staff focused on communicating clearly expressed health data, as the scaling up proceeded, so that politicians and bureaucrats would become more accustomed to setting policy informed by evidence. They also hoped that this impact on policymakers would move beyond Tanzania — creating a receptive climate across East Africa for implementing evidence-based health care geared to addressing the dominant burdens of disease.

This continued emphasis on dissemination took aim at a problem sometimes referred to as the “know–do gap.” In many cases, knowledge of the major causes of illness, of which interventions can most effectively deal with them, and what treatment packages should look like, does exist. Yet there is a chronic gap between this knowledge and what happens in communities. Often, crucial information does not flow to the people who set health policy and deliver health services. Partly, this is a systemic problem stemming from the lack of incentives for translating research findings into usable information that can solve problems: researchers will reap career rewards from publishing papers in scientific journals, but will not benefit from writing short articles that are accessible to health practitioners and policymakers. The “know–do gap” also arises from a discontinuity of roles, wherein no-one is given specific responsibility for ensuring researchers’ findings are delivered — in usable form — to the people who design new health policies and deliver health services.

While TEHIP researchers did indeed publish papers in international scientific journals (with the goal of making the case to the global health community that health systems deserve more attention), they also placed special emphasis on communicating TEHIP’s findings and recommendations to the often forgotten but important players at the local level. The first edition of Fixing Health Systems, published by IDRC in 2004, was also key to reaching policymakers at various levels, as well as agencies and donor partners that would become involved in funding later health system revitalization initiatives.

2. Replication

Since the goal of rolling out new techniques and methods is to strengthen an existing health system, scaling up and replicating a project’s results is not something that should be undertaken by the project itself, but rather by the health ministry in the host country. In other words, it is important to find existing and appropriate vehicles for delivering innovative health services that are not dependent on the project, but that will survive and flourish after the project has wound down.

The replication of TEHIP’s practices and results followed this model. The expanded use of the burden of disease tool and IMCI outside of Rufiji and Morogoro, for example, was instigated by the MoH and accomplished using its institutions. Later, another independent entity (the UNF) joined IDRC in funding the rollout of most of the other interventions to nine additional districts. In doing so, it created the design for a national scale-up process. One key aspect of this exercise was to identify an existing institution within the MoH structure (the Zonal Training Centre system) that could be used — given significant restructuring and revitalization — as the actual machinery for accomplishing a national scale up.

Later in the process — in 2007 — Tanzania’s finance ministry incorporated the TEHIP District Health Accounts tool into its PlanRep budgeting software, which all districts are now mandated to use. This is another example of a project tool becoming fully integrated into the national policy-making machinery.

One case where a TEHIP -initiated model is to be replicated outside of Tanzania is the Nigerian Evidence-Based Health System Initiative (NEHSI) — designed by Nigerian and international organizations under the coordination of IDRC, and funded by CIDA and IDRC. Nigeria is the most populated country in sub-Saharan Africa, and one where multiple vertical health programs have created inefficiencies and a lack of coherence in health care delivery. NEHSI will see the creation of DSS systems in two states to compile portraits of the local burden of disease, together with Multi-Stakeholder Information Systems that will ensure ongoing communication among communities, health facilities, planners, researchers, and development partners. This will allow for a more coordinated approach to primary health care planning, delivery, and evaluation, as well as more rational targeting of resources toward the most pressing local health challenges.

3. Planning for sustainability/institutional strengthening

As the sun set on TEHIP, it became essential to transfer knowledge, methods, and capacity — acquired over the project’s lifespan — to national, local, and regional institutions. These institutions would be responsible for maintaining the gains that had been achieved during the TEHIP years and for dealing with any new challenges that lay over the horizon.

One entity that took on a large share of that responsibility was the Zonal Training Centre (ZTC) system, which was a virtual institution annexed to an existing health training institution. Like many aspects of the Tanzanian health system, ZTCs were underfunded and in various states of disrepair. The staffing levels were often inadequate, and course content was geared mostly to delivering segregated vertical programs without any coherent, overall vision. Now, the ZTCs are being restructured and reinvigorated through funding allotments from the Ministry and health donor partner basket funds, which includes a CA $7 million grant from CIDA and other individual disbursements from agencies such as the World Bank, DANIDA, GTZ, USAID, JICA, and the Government of the Netherlands.

The importance of the ZTCs to the national scaling up stems from the obvious need to train front-line health workers in new services such as IMCI, which has made a huge contribution to improving children’s health outcomes in Tanzania. More broadly, however, the ZTCs offer an entry point in the search for solutions to the recognized and documented crisis in health-sector human resources that plagues most developing countries. Deploying and retaining effective health workers remains one of the most intractable challenges facing the continent’s health sector. With low levels of job satisfaction and a lack of incentives (e.g., poor wages and inadequate housing), health workers frequently leave their jobs, relocate to other countries, or take second jobs. Better training through improved ZTCs could represent a first step toward reinstating health workers as valued, respected professionals with a greater stake in their work.

Currently, ZTCs are gearing up for a more streamlined provision of training, wherein practitioners — rather than being pulled out of their workplaces for disconnected training sessions on vertical programs — will learn about key interventions as part of an organized and integrated curriculum. This will also pave the way for a more rational system for allocating and posting health workers in the field. Rather than simply justifying a list for a certain number of health worker cadres, with particular specialized training, it will increasingly be expected that clinical staff will have been trained and have all the required knowledge to deliver the entire national health package on a daily basis.

The ZTCs are also envisioned as a focal point for generating analysis and advice to the ministry. This new role is being defined by the regional centres, which have embarked on a process of creating their own business plans. The enhanced contribution of ZTCs to health management will include activities such as tracking which health workers need retraining or upgrading, providing input to determine what staffing levels are needed in particular areas, and performing follow-up supervision of all trained health workers to ensure high quality performance.

The design and construction of the REACH policy initiative provides another example, at the regional level, of a transfer of knowledge and capacity to a sustainable, ground-level institution. REACH will be the practical expression of the concept that TEHIP staff originally had envisioned as “the Duluti Institute.” Its primary function will be to synthesize, package, and distribute health policy options and information to policymakers, so that they are able to undertake more evidence-informed planning and policy-making. Its structure will see individual country nodes (in Tanzania, Kenya, Uganda, Rwanda, and Burundi) linked to a single regional hub at the East African Community level. Requests for information will filter upward from the country levels, so that the regional staff can prepare confidential briefings to inform politicians and bureaucrats on what policy directions the current evidence suggest.

REACH will also deal with cross-border health concerns such as epidemics and the spread of rare or emerging infectious diseases. At the time of this writing, REACH was involved in late-stage fundraising that would allow for a project launch.

Earlier analysis determined that a TEHIP-style approach to health delivery would be best suited to countries where

• there are accessible primary health care facilities within reach of most of the population;
• the health system is to some degree decentralized;
• there is a sector-wide approach (SWAp) funding model in place; and
• some form of health surveillance system exists to provide a foundation for gathering population health data, and there is some national health research capacity.

East African countries such as Uganda, Kenya, and Tanzania (which are participating in REACH) fit that description. Further afield, countries such as Burkina Faso and Zambia also share those characteristics, and thus have the potential to replicate TEHIP-style health delivery.

4. Continued data collection and population health research

The Rufiji DSS station was made financially sustainable and, subsequently, was turned over to the Ifakara Health Research and Development Centre. Another ongoing contribution in the area of data collection has been the establishment, with help from TEHIP, of the INDEPTH network. INDEPTH is comprised of 37 sites operating at the household level in 19 countries in Africa, Asia, Central America, and Oceania. It has produced a “DSS starter kit” to aid in setting up new DSS facilities, works to ensure compatibility of software and operations between different DSS sites, and is encouraging the pooling and comparison of health data across borders.

Gathering reliable health data is an essential cornerstone of evidence-based planning. It not only provides a comprehensive portrait of burden of disease (by capturing all at-home mortality and morbidity as part of the picture), but also provides continual monitoring of health conditions that can rapidly inform whether new types of health delivery or new policies are actually working. This allows health managers to adjust and improve upon the public health package design.

Throughout Africa, the norm for health data collection has been to tally information from cases diagnosed in hospitals and other health facilities, a method that yields incomplete results since many people who are ill stay at home or seek services somewhere else. Partly as a result of TEHIP’s example, there is increasing recognition of the crucial contribution of demographic surveillance that incorporates routine population surveys and verbal autopsies for people who have died at home.

This approach is being promoted by a new global health initiative for health information systems — the Health Metrics Network — housed at the WHO. Additionally, the fact that one of NEHSI’s main entry points into health system reform in Nigeria is information- gathering through DSS is another demonstration of this new understanding that DSS (or a similar system for gathering population-based health data) is a fundamental requirement for evidence-based health planning in settings where there are no vital events registration systems.

5. Fundraising

With the retirement of TEHIP as an organization, it was important to encourage new donors to step forward and address the multiple needs that had been identified over the course of the project, to build on the project’s achievements, and to address new stresses and bottlenecks that the health system will inevitably suffer.

New infusions of funds and expressions of commitment have indeed improved the prospects for health system renewal in sub-Saharan Africa. For example, Comic Relief, a UK charity whose previous concern had been largely with providing relief after events such as natural disasters, has ventured into the more complex matter of health system strengthening by awarding the Ifakara Centre a (UK) £5 million grant over 5 years. (This was the largest grant ever awarded by Comic Relief, and it makes Ifakara the first African institution that Comic Relief has funded directly.) With input from the original TEHIP team, Ifakara submitted a proposal to Comic Relief for a project seeking to replicate the kind of gains made by TEHIP in under-5 child health, by using the R&D approach to obtain similar improvements in maternal and neonatal care, both areas where the death rate remains unacceptably high. From the outset, Comic Relief had expressed a desire to back project designs that incorporated TEHIP methods, particularly the integration of research and development in actual field conditions.

In addition to linking research directly to development, this Comic Relief-funded work has many other characteristics that were hallmarks of TEHIP. It will address recognized needs and priorities set out by the Ministry of Health, and will engage researchers in partnership with district planners, front-line health staff, and communities. It will use DSS resources to monitor ongoing progress and will attempt to review and update existing TEHIP planning tools. It is not bound by short-term time restrictions and targets, but will allow for the pace of work to be dictated by conditions on the ground. It will feed results back to the Zonal Training Centres, so that advances in maternal and neonatal health will be incorporated into training and will inform policy and practice across the nation.

There are several other examples of organizations picking up where TEHIP left off — taking the growing interest in the links between improved health systems and better health to new frontiers, both geographically and conceptually. For example:

• The US-based Doris Duke Foundation is providing US $100 million in grants for health system research and development in Ghana, Kenya, Lesotho, Madagascar, Malawi, Mozambique, Rwanda, Tanzania, and Zambia.
• CIDA has launched a CA $450 million “African Health Systems Initiative” to help strengthen health systems across sub-Saharan Africa.
• The Wellcome Trust (UK) has joined with IDRC and the UK’s Department for International Development to provide support for health research capacity strengthening, including initiatives specifically focused on linking research to health policy, to strengthen health systems in Malawi and Kenya. Each country is due to receive approximately £11 million from the donor consortium over 5 years.

Those examples reflect a growing consensus that operational questions need to be addressed if Africa is to make significant strides toward achieving the MDGs. While 10 years ago, most health investment flowed to the development of new vaccines or other technological fixes, there is a growing focus on answering systemic questions such as how to remove the barriers that prevent health care from reaching the people who need it; how health packages can be better designed and delivered; and how improved personnel, transportation, and related logistical practices can help health systems do what they were intended to do.

These sorts of questions are now being asked at the highest levels. In 2004, for example, the WHO’s World Report on Knowledge for Better Health called for more innovative health system research to help bridge “the know-do gap.” In short, the world has changed since TEHIP set out to test the 1993 World Development Report hypothesis that modest investments in health systems could produce significant improvements in health.

Still, this new and promising climate does not justify complacency. Having mechanisms in place to improve health care delivery does not mean there will not be new challenges and setbacks. The factors that affect population health are dynamic and everchanging. Managing health systems to respond to shifting needs and variations in the relative burdens of disease requires continuous attention to a large number of factors, but even still, unexpected circumstances may arise that can derail even the most carefully laid plans.

It is also likely that a particular package of interventions, after producing dramatic results, may hit a “plateau” where improvements in health outcomes start to level out. This has been the story in Tanzania, where DSS data are suggesting that most of the improvements that can be expected from IMCI have already occurred. When the health system hits this kind of “hard floor ” — when the illness and death rates in a particular category seem to have become intransigent or fixed — this is the time when health planners must look to other areas where mortality levels can be brought down to similar levels. In other words, when the low-hanging fruit has been picked off the tree, it’s time to buy a ladder and start picking higher up.

This is what is happening in Tanzania, where health planners are now looking to improve outcomes in neonatal and maternal health. While under-5 deaths began to decline steeply in 1999, neonatal and maternal death rates remained constant (Figure 7). For further declines in under-5 mortality — needed to achieve the Millennium Development Goals — it is clear that greater efforts are needed on both neonatal mortality and maternal mortality. Making gains in these areas will be essential if Tanzania is to make the necessary strides toward achieving the MDGs related to child and maternal health.

Figure 7. Under-5, neonatal, and maternal mortality in Tanzania, 1992–2004.

In short, there are no “quick fixes” or “silver bullets” in the quest to reduce Africa’s unacceptably high death rates. Having an effective, functioning health system in place is merely one crucial stage in an ongoing process that requires continual vigilance, attention to a large range of details, and a willingness to tackle the new challenges. Having the proper tools at the ready will provide health planners with the means to improve population health. But making sure those tools are used to greatest effect will also require ingenuity, commitment, and sustained political will.

Document(s) 11 of 12