Document(s) 5 of 10

Women cradling babies sit along wooden benches outside the health clinic in Mvomero, a small rural community in Morogoro district. Inside, Mr Y.E. Kapito gently examines a baby to determine why he has a fever and to assess other health problems. The practitioner's quick reference guides are algorithmic, coloured wall charts that describe childhood illnesses, classification, and pathways to appropriate treatments.

Mr Kapito is seeing more patients these days. He estimates that the number has almost doubled in a year. But he hears of fewer child deaths. "It's been 6 to 8 months since I have heard about a child dying," he says. Samuel Hassain, who brought in his flu-stricken grandson, remarks, "Things have improved. People have faith in the services. They are treated well and get diagnosed properly."

Mvomero is one of many locations in Tanzania where the encouraging statistics showing lower mortality rates (particularly among children) take on a tangible form and a human face. It is here on the front lines -- that critical arena where theory meets practice -- where the reforms adopted by district health managers are proven either useful or ineffective.

One of the reasons why more children who come to the Mvomero clinic survive is that practitioners like Mr Kapito are now using a system called the Integrated Management of Childhood Illnesses (IMCI), one aspect of which is the use of those coloured charts. But such changes at the community level have also been accompanied -- and, in fact, made possible by -- some fundamental shifts in the way the health systems of Morogoro and Rufiji operate.

Essentially, improvements in the quality of health care have been the result of two factors. One is how district managers chose to use the planning tools; that is, what health interventions they decided to fund and prioritize in response to the portrait of burden of disease produced by the tools. DHMTs used these tools to overhaul the ways they planned, to revise their proportional allotment of funds, and to promote integrated solutions that offered multiple benefits from single health interventions.

The other key factor that allowed change to take place in the field was the availability of so-called "supportive interventions." Intended to restore some of the functional capacity that the health systems had lost because of years of under-funding and inertia, these interventions depended on districts' use of supplementary funding they had received from TEHIP. Both the use of the tools and the districts' investment in "supportive interventions" are examples of capacity building. However, they differ from each other in one key respect. The deployment of the tools is a form of passive capacity building: those tools didn't directly lead to change, but rather were one contribution that helped the districts enhance their own capacity to plan. The districts' provision of supportive interventions, on the other hand, is a more traditional form of capacity building that relies upon a direct transfer of skills through means such as training courses and targeted funding.

Supplementary funding

None of the changes to the ways that managers and clinicians go about their daily business would have happened without the provision of a modest financial top-up in the form of stakeholder funds to the districts. This new inflow of cash catalyzed an improvement of health care delivery by giving districts the financial means to act on their plans. These funds allowed districts to achieve new efficiencies in the daily operations of health systems and to increase spending where needed on interventions aimed at the most significant contributors to the local burden of disease.

The project first offered additional funding of up to US $2 per capita to the districts of Morogoro Rural and Rufiji in 1997. Surprisingly, however, both districts soon discovered that they were unable to absorb and spend this amount of extra funding. The $2 remained on offer for each of the first 3 years of the project. Still the districts did not have the capacity to absorb that amount of money. In 1997/98, 1998/99, and 1999/2000, the respective per capita consumption from the fund was US $0.57, $0.89, and $1.37. At this point, to spread the unspent money over a more reasonable period, the project was extended for 3 more years and the amount offered was reduced as the new national SWAp "basket" of US $0.50 per capita came into being (see box: "SWAp basket brings funding stability"). The average fund consumption over the first 4 years of the project was a US $0.92 annual per capita top-up contribution for the two districts. Expressed another way, the health advances experienced in Rufiji and Morogoro were achieved at a per capita annual cost less than what someone in North America would pay for a cup of coffee. 

SWAp basket brings funding stability The SWAp district council health basket funds — consisting of resources received from a range of international development agencies active in the health sector in Tanzania — is administered by the Ministry of Health and made available to all districts of the country. The phasing in of the SWAp basket between 2001 and 2003 has addressed the difficult issue of the sustainability of top-up funding in the two districts as TEHIP has been phased out. Its creation has helped to ensure that the benefits that came with TEHIP top-up funding will remain as long as this new basket stays in place.

These extra funds were used by district planners to support essential health interventions, the rehabilitation of health facilities, and training and capacity-building activities, as well as to assist with the purchase of essential drug supplies, transportation, communications equipment, and computers and computer software. Ceilings were placed on the amount of funds that could be spent for health facility rehabilitation. The only other caveat was that intervention spending had to be consistent with evidence about the burden of disease and had to support only interventions known to be cost effective.

The districts' initial lack of capacity to absorb the additional US $2 in funding was at first baffling to the DHMTs. Why were the new funds sitting unused in a bank account? The answer was that health systems in Rufiji and Morogoro lacked the administrative and management capacity to put the funds to work. To absorb the top-up funding would have required the skill base to transform financial resources into program spending. Staff would have needed the skills enabling them to draw up contracts, hold structured meetings, issue cheques, procure supplies, interact with accountants, etc. -- to do all the little things that are crucial to the functioning of a health system.

Capacity building in management and administration

The DHMTs saw in the mystery of the unspent funds a clue as to where they should invest a portion of the new money. It became clear that, for the system to be able to expand its budgets and service levels in the future, essential training and capacity building had to take place. Therefore, the DHMTs took stock of the missing skills and subsequently made use of a range of "supportive interventions" to address local deficits in management, administrative, and other skills.

For example, they acquired a modular training course called Strengthening Health Management in Districts and Provinces, developed by WHO (Cassels and Janovsky 1995), as a way of building the teams and increasing the confidence and skills of health management personnel. The basket funds also opened the space for the districts to purchase skills training for planners based on Ten Steps to a District Health Plan. Distributed by the Tanzanian Ministry of Health, the guide was developed by the Iringa Primary Health Care Institute (1997) in collaboration with the Nijmegen Institute for International Health.

This type of capacity building was crucial to the functioning of the DHMTs. Although the transfer of responsibility from high-level planners to locally based management teams was seen as central to Tanzania's health reforms, it was by no means assured that the team approach would succeed. Effective teams do not suddenly spring into existence -- they need cultivation to acquire essential tools and skills and to develop over time. Trust and cooperation between members must be fostered, specialties must combine in ways that complement each other, and teams must learn how to delegate responsibility. And as the team approach moved outward to encompass facility staff and, indeed, whole communities (which participated in health reforms by helping to renovate health centres) so too did the list of partners needing training and capacity building expand.

In addition to management and team-strengthening courses, the new funds allowed the districts to pay for other supplemental capacity building and training that responded to perceived needs in a number of areas, including

• Report writing (quarterly, technical, and financial);
• Computer applications and training;
• Administrative and financial procedures (including financial management, inventory control, payment, and ledger maintenance);
• Office management (including components of office management such as filing, communication, email, appropriate channeling of communications, and organization of meetings, including writing of minutes, determination of action items, and delegation of tasks); and
• Regular maintenance of vehicles, radios, computers, health equipment, capital items, solar power, etc. Maintenance was largely overlooked in Tanzania, where capital goods and infrastructure would typically break or wear out ahead of schedule. Ensuring that costly items would last longer -- because of regular maintenance -- was essential to increasing the cost-effectiveness of the health system.

The Integrated Management Cascade

A large part of the work leading to the improved routine functioning of the district health systems was bundled together into a strategy known as the Integrated Management Cascade (IMC). Focused largely on increasing capacity for the downward delegation of administrative duties to community-level workers (so that those workers became more engaged in the process of improving health care delivery), IMC was designed to ensure the smooth operation of the health system by improving the links between community health workers and supervisory personnel. These links led to greatly improved quality in the health system by reducing problems that had existed in numerous areas, such as delivery of drugs and supplies, distribution of staff pay, and the supervision of health workers dealing with expenditures and the referral of emergencies. All have major effects on the delivery of service.  

Putting supervisors and workers in closer contact required innovations in transportation and communication -- requirements made all the more urgent by the challenging terrain of both Rufiji and Morogoro. In Morogoro, the mountainous landscape makes it difficult to travel between the district headquarters and outlying health facilities. The same is true for the Rufiji River delta as well as the low-lying areas of the Rufiji flood plain, which are rendered impassible during the rainy season when the river can become a kilometres-wide torrent. Between March and May, the rains wash away sections of road, making many facilities inaccessible. Moving between locations therefore comes at a high price: finding alternative routes sometimes increases the distance threefold, leading to much greater use of fuel and staff time.

The solution? Districts used their top-up funding to equip each health centre with a solar-powered radio (a model that was robust and easy to operate) and to purchase a motorcycle, allowing supervisory personnel to travel quickly and cheaply between facilities. Rufiji also invested some of its funding in a high-speed boat to travel to villages in its vast tidal delta, since -- in addition to the problems of the rainy season -- such health facilities are normally accessible only by water at high tide.

These modest investments in appropriate-scale transportation and communications technologies have improved workers' abilities to perform their daily duties. "We have a motorcycle in each facility that enables us to make supervision tours to our satellite dispensaries," says J.R. Lifa, Clinical Officer-in-Charge of Mgeta Rural Health Centre in Morogoro. "We also have radio equipment in each health centre, which facilitates easier communication between zones and the DHMT office."

Mr Amadeus Mwananziche, in charge at the Mlali Dispensary, says that the enhanced communication has improved functions ranging from delivery of drugs to referral of cases. "We can now inform each other when there is an epidemic and it is a lot easier to learn of any problems anywhere within my area of operation."

The new transportation and communication capacity has also made it possible to establish a better organizational structure known as the cascade management system. Previously, a core of 4 or 5 managers in a central office directly supervised 50 to 100 facilities -- an almost impossible task, given the expectation that supervisors visit each dispensary 3 times a year. Now, supervisory responsibility "cascades" from staff in the central office to a secondary level of staff in the health centres. In turn they supervise a tertiary level of staff in the dispensaries within a reasonable traveling distance. This cascade structure, as depicted in Figure 2, has helped spread the benefits of new investments across the districts. Although labs and back-up drug supplies, for example, have been placed only in larger health centres, the use of radios and efficient means of transport have meant that clients who use smaller facilities can also share in the benefit of those investments.

Figure 2. The management cascade. Supervision is delegated to staff at selected health facilities. This map shows the cascade system for Rufiji district. Facilities of the same colour and symbol are grouped together.

Giving supervisors the physical means to oversee facilities within a more reasonable range has made it easier to solve routine problems. "We now have access to a working radio as well as email," explains Dr Harun Machibya, Morogoro Rural District Medical Officer. "A simple matter that can now be solved by radio communication would have previously meant an officer boarding a bus to Dar es Salaam." To date, the management cascade has produced many positive results:

• Actual supervision of peripheral facilities with time for supervisors to directly observe patient care;
• More coherent laboratory specimen collection and diagnostic laboratory reporting functions;
• Timely delivery of drugs, equipment, and supplies;
• Coordination of referrals of patients to the district hospital;
• Emergency epidemic support, such as during cholera outbreaks;
• Routine collection of health information and data;
• Notification of arrival of staff salaries (resulting in reduced closure of health facilities as health workers travel to collect salaries too soon and have to wait before returning);
• Improved maintenance of facilities and equipment, and replenishment of stationary, registers, etc.;
• Improved linkages and communication with communities;
• Locally conducted capacity-building workshops, technical training, and refresher courses; and
• Posting of replacement health staff when regular personnel are ill or have died.

Instantaneous communication by radio has also had an uplifting impact on morale: community health workers can now be assured that help is on the way, rather than remaining uncertain about when district supervisors will respond to their inquiries or requests.

Rehabilitation of health facilities

At a spectacular ceremony at the Hanga Dispensary in Rufiji, the village chairperson walks from the audience to the podium, greets the area's member of parliament and returns to her seat holding a certificate signifying that members of the community now own their health facility. This effectively makes community members partners with upper levels of government in the provision of health care.

All across Rufiji and Morogoro, communities have entered into the same unique arrangement. The government has agreed to pay the salaries of health workers in the community and for drugs and supplies. In exchange, communities plan and contribute to the renovation of the local clinic with their inputs (such as local building materials) and labour, and pledge to maintain and run the facility. This citizen participation -- part of a shift in emphasis away from central government toward local communities -- is described by Dr Harun Machibya as "a breakthrough." Says Dr Machibya, "Before, people asked for help [from the government] even when a lock was broken."

This new partnership answers a crying need in the districts. Over the years, communities and Ministry of Health officials have expressed a consistent desire to see health facilities renovated. Built in the 1960s and 70s, village dispensaries had been neglected for several decades and many of them were dilapidated. Early on, a small amount of money available through TEHIP was set aside to study renovation options. A Ministry of Health architect was contracted to conduct a needs assessment that would gauge the state of each facility and rank them in order of priority. Photographs were taken of each facility and the numerous deficiencies of each were catalogued -- a truly colossal task given that there were over 90 health facilities in Morogoro alone, demanding vast expenditures of time and energy.

It was soon clear that raising the necessary funds to renovate all these facilities would be impossible. But the Tanzanian "self-help" tradition from the postindependence period suggested a solution: what if officials asked communities to provide labour and some materials to offset the costs of renovation? A Tanzanian team familiar with community labour-based approaches was engaged to facilitate an initial demonstration exercise in 3 communities within each of the 2 districts. Community members were engaged in dialogue and work plans were produced. Rehabilitation of dispensaries in those communities took roughly 6 months. The communities' contribution of materials and labour meant that the renovation projects cost between 31% and 48% less than they would have cost under normal subcontracting practices.

Since then, this process has taken place in close to 40 communities. Some communities have gone well beyond what was in the standard dispensary blueprint by providing the health facilities with better latrines, water supply, and even extra maternity facilities. In some cases, the momentum from health facility rehabilitation has spurred other changes, with community members initiating new upgrading projects for mosques, schools, health workers' houses, and other local amenities.

The celebrations that invariably accompany completion of the health facility renovations provide an opportunity not just for revelry but also for government representatives to acknowledge the citizens' contribution. In Lusanga, for instance, as members of the community and the DHMT gathered in a carnival atmosphere, a high-ranking official presented a gift to the community. This gift took the form of more than 2 million Tanzanian shillings (US $1 800 in July 2004) worth of health equipment, including in-patient and delivery beds, suction equipment, and a blood pressure machine. Ownership of the building was then transferred to the community -- a gesture that stands as a reverse image of what occurred 30 to 40 years earlier, when communities transferred ownership of health facilities they had built to the central government.

What the districts did with budget planning tools

In addition to the "supportive interventions," other reforms -- aimed at changing how community health workers dealt with disease -- took place on a parallel track. Health managers at the district level, newly equipped with the planning tools, began restructuring health services to focus resources on the greatest need. Armed with new and current knowledge about the burden of disease in their district -- and with tools that allowed them to track spending in line with that burden of disease -- these managers tried to ensure that spending on individual health services would provide a significant pay-off in terms of lives saved and illness prevented. Districts also had new insight into the ways the health system could be improved since they had access to the results of research undertaken by the research consortia (see box: "What the research revealed").

What the research revealed Research into three distinct areas — health systems, health behaviour, and health impacts — provided critical insights that aided health care reforms in Rufiji and Morogoro. One of the most striking findings of the inquiries into health-seeking behaviours, for example, was that most deaths (close to 80%) occurred at home rather than at a health facility. This statistic underscored earlier doubts about the use of attendance and cause-of-death statistics — compiled by the government on the basis of health facility data only — as an aid for planning health budgets. Surely, this form of planning could not be reliable since it was based on only 20% of deaths. Since DSS information, by contrast, captures all deaths — those that occur in health facilities, in households, and elsewhere — it can be counted upon to give a more accurate and complete portrait of the burden of disease as experienced by the community. Another surprising revelation arising from research into health-seeking behaviour was that the people who had sought modern health care prior to their deaths greatly outnumbered those who had not. As illustrated in Figure 3, for malaria, 78.7% used modern care, only 9.4% used traditional care, and 11.9% used no care at all. These figures prove that the death rates in Rufiji and Morogoro were not primarily an outgrowth of a preference for traditional healers over modern health care (as some observers had speculated), but are more reasonably seen as related to problems of access, delay, or the apparent inability of modern health facilities to prevent these patients from dying. Formative research into the health systems planning process confirmed that planning was not being conducted as a response to the burden of disease, but instead was driven by a wide range of factors including donor agencies' agendas, bureaucratic inertia, and simple guesswork.

Figure 3. Initial care-seeking patterns. Care of first resort sought during the final illness by 320 fatal malaria cases in children less than 5 years of age in the Rufiji DSS sentinel area, 1999­2001. (Source: de Savigny et al. 2004)

What did burden of disease profiles and budget analysis tools tell district managers? Figure 4 provides some indications of how the actual local burden of disease corresponded to recent health care spending. In Morogoro district, for example, spending on malaria had been significantly lower than would have been expected for an illness that was the single biggest cause of mortality in the district. Following the introduction of the planning tools and additional funding, Morogoro's proportional expenditures came more into line with its burden of disease. Meanwhile, in districts where planning tools were not introduced, the imbalance remained striking.

Figure 4. Morogoro district health spending before (1996/97) and after (2000/01) the introduction of TEHIP budget mapping tool.

* Absolute funding levels for immunization and tuberculosis remained unchanged while funding for other interventions increased.

Evidence provided by the DSS indicated to health planners that two areas in particular accounted for a huge proportion of the burden of disease: acute febrile illness (including malaria) and a cluster of diseases affecting young children (acute febrile illness, pneumonia, diarrhoea, malnutrition, anemia, and measles). While a few individual districts did provide reasonable funding for these illnesses, aggregate numbers for all of Tanzania show, in general, insufficient funding to fight the biggest killers. Nationwide, childhood illnesses accounted for 37% of burden of disease: they received 17% of funding (Figure 5). A bigger problem is that the largest funding share (sometimes over 50%) went to a multitude of marginal problems with a cumulative burden of less than 15% (Figure 6).

Figure 5. Districts planning without tools: budget requests for IMCI in 30 district health plans, 2002.

Figure 6. Districts planning without tools: cumulative budget requests for all other interventions, each of which individually address less than 1% of the total burden of disease in 30 district health plans, 2002.

One way that district managers could more effectively face the challenges posed by the real burden of disease was by paying close attention to what went into the "minimum package of interventions" to be used by community health workers. They could, in other words, make sure that every remedy was useful in the local context. Table 2 details what was included in this minimum package in Rufiji and Morogoro. Every intervention in the package had to address a significant portion of the burden of disease, either a single or collective illnesses or conditions that contributed at least 2% or more to the burden of disease or an eradicable condition.

Table 2. Tanzania's national package of essential health interventions

Reproductive and child health

* Safe motherhood: maternal conditions

Intermittent presumptive treatment of malaria (pregnancy); antenatal care; obstetric care; postnatal care; gynaecology, STD, HIV/AIDS care; micronutrient supplementation for mothers

* Safe motherhood: perinatal conditions

STD screening; support for traditional birth attendants; safe delivery practices; newborn care; micronutrient supplementation for low birth weight babies; village birth registers

* Immunization

BCG (tuberculosis); diphtheria; pertussis; neonatal tetanus; measles; poliomyelitis; hepatitis B

* Integrated Management of Childhood Illnesses (IMCI)

Malaria; pneumonia; diarrhoea; measles; malnutrition; anemia

* Family planning

* Nutritional deficiencies

Nutrition information, education, and communication; breast-feeding support groups; growth monitoring and pupil health screening; micronutrient supplementation (iron, vitamin A); monitoring salt iodization; deworming; school feeding

Communicable disease control

* Malaria

IMCI (early care seeking and case management); insecticide-treated bed nets; intermittent presumptive treatment in pregnancy; home-based care; school health education about malaria prevention; epidemic preparedness; sustainable source reduction; information, education, and communication

* Tuberculosis and leprosy

Tuberculosis Directly Observed Treatment -- Short Course (DOTS); leprosy multidrug therapy; home-based care

* HIV/AIDS and STDs

STD prevention; information, education, and communication; condom promotion; blood screening; patient care, counseling, and social support; palliative care

* Epidemic preparedness

Cholera, measles, meningitis, plague, and malaria

Noncommunicable disease control

* Cardiovascular diseases

IEC on smoking, alcohol, diet, and exercise

* Diabetes

Preventive and promotive IEC; routine checking of blood pressure

* Neoplasms

Breast and cervical cancer screening

* Injuries and trauma care

* Mental disorders

* Anemia and nutritional deficiencies

Treatment for common diseases

* Helminths, skin, ocular, and oral conditions

Community health promotion and disease prevention

* School health

* Water hygiene and sanitation

* Information, education, and communication

Seven interventions

A new assault on disease

District managers also ensured they had well thought-out programs in place to combat the conditions that accounted for very large shares of the burden of disease. Since the most burdensome conditions were malaria and illnesses of young children, health managers put a heavy emphasis on malaria control and the Integrated Management of Childhood Illnesses (IMCI).

Malaria is a disease that already had been identified at the national level as a huge threat to public health. In the wake of daunting evidence that conventional drug therapies for malaria were failing, the Ministry of Health had embarked on a new antimalarial drug policy using a newer drug. The Ministry has also been involved in the promotion of insecticide-treated bed nets, which are deemed particularly vital in protecting pregnant women and babies from the devastating effects of malaria infection. TEHIP and its tools have fed into the antimalaria assault at both the district and national levels. Locally, the tools showed managers that insufficient resources were being applied to combating malaria and persuaded them to reverse this trend. Nationally, the project's advice was sought in the development of new mechanisms to increase the distribution of bed nets across Tanzania (see box: "The need for treated bed nets").

The need for treated bed nets Although the efficacy of insecticide-treated bed nets (ITNs) for preventing mortality was only established in 1996, the evidence in their favour is impressive. The Tanzanian Ifakara Health Research and Development Centre has determined that ITNs could prevent 30 000 deaths and more than 5 million clinical episodes of malaria annually in Tanzania. Research summarized in Net Gain, a copublication of IDRC and WHO, indicates that ITNs could reduce child mortality in Africa by at least 17% (Lengeler et al. 1996). Treated nets are also one of the most cost-effective ways of preventing death and illness. And their benefits appear to be community-wide rather than merely personal: recent evidence shows that an area with a high concentration of treated bed nets affords some protection for all people in that area, even those who do not own a bed net themselves. Treated bed nets were one of the selected essential health interventions in these evidence-driven districts.

It is important to note that there is a significant intersection between the two leading contributors to burden of disease: IMCI-addressable conditions and acute febrile illness (AFI). AFI (including malaria) represents over 60% of IMCI-addressable conditions in Tanzania. Indeed, malaria is more debilitating for children than it is for adults and is a contributor to many child deaths that are not recognized as having been caused by malaria. Districts have therefore focused particularly on combating malaria infection in the very young (see box: "'Malaria's threat to the young").

Malaria's threat to the young Many adults may not intuitively recognize malaria as a deadly disease for children because their own experience of malaria may be quite different. Adults who have survived malaria earlier in life are not likely to later die from it. Although they may suffer regular episodes, they have developed partial immunity and can be more readily treated. For children, however, malaria is debilitating in a number of ways. After the first few months of life, a baby will lose the immune protection it has acquired passively from its mother and may experience acute fever, anemia, convulsions, and other dangerous complications as a result of malarial infection. The child will then enter a downward cycle where lack of appetite and constant stress will push him or her off the growth curve, suppress the immune system, and leave the child more vulnerable to other diseases. When the child becomes thoroughly debilitated from fighting the malaria parasites, the next experience of illness — be it diarrhoea or another episode of malaria — is more likely to be fatal. Malaria is also directly responsible for many childhood deaths that masquerade as other illnesses. Since the malaria parasite multiplies in red blood cells — eventually causing the red blood cells to disintegrate, before moving on to inhabit other red blood cells — children with malaria exhibit acute anemia. Anemia in young African children is difficult for mothers and health workers to spot. A drop of blood from a child with severe anemia from malaria will often appear pale pink instead of red on a piece of blotter paper. This malaria-related anemia may set in motion a series of other tragic consequences. When the red blood cell count is low, it is difficult for oxygen to be transported through the blood stream. As a result, the child's heart is under enormous strain. Sometimes the result is sudden cardiac failure — an event that appears to parents and to many clinicians to be unrelated to malaria.

Integrated Management of Childhood Illnesses

The first adaptation and pretest (or trial) of the Integrated Management of Childhood Illnesses (IMCI) took place in Arusha, Tanzania, in February 1995. Since then, WHO and UNICEF have produced generic materials for this approach, and IMCI is now being implemented in many developing countries. 

IMCI is a care strategy that takes a "syndromic" approach, addressing the whole child by identifying and treating a range of possible common illnesses rather than simply focusing on one disease at a time. The fact that sick children will present with symptoms that may be caused by a number of possible diseases -- or that they may suffer several diseases at once, or have one condition masquerading as another -- provides a strong case for this integrated approach. 

When district managers in Rufiji and Morogoro recognized and sought to lower childhood mortality rates, adopting IMCI became an obvious choice. Implementing the program, however, required major efforts: retraining front-line health workers, reorganizing the use of clinic space, and promoting and encouraging a more active health-seeking role and better skills for parents -- all of which had to be addressed simultaneously.

Consider the fundamental shifts in practice that IMCI has required. Before IMCI, children brought by a parent to a clinic would be seen in a "factory line" process. The health worker might guess, for example, that the child's problem was diarrhoea and dispense a standard treatment. The process was one of quick evaluation and the rapid consumption of certain drugs. When districts sought to change this, they began with staff retraining. Groups of workers were taken out of the field and given a specific course of training and new bedside clinical instruction. When they returned to work, workers would be "validated"' to ensure that they had absorbed the new methods.

The differences in the way health workers are now expected to deal with patients start with the children's arrival at the clinic. Children are assessed to gain some sense of the seriousness of their condition. Patients exhibiting danger signs move to the front of the queue. To streamline the work, a clinic assistant takes care of simple processing such as taking temperatures and weights of patients in the queue. Today, a health worker examining a child is no longer expected to guess at the root cause of the child's ailment. Instead, the worker consults a set of algorithms: when certain signs and symptoms are present, the practitioner is guided to the possible causes of those symptoms and to the appropriate treatment and prescription. This is part of the "syndromic" approach -- the health worker is prepared to accept that the child may be suffering from a series of conditions working in concert, rather than from one particular disease.

Health workers believe that major gains have been achieved from adopting this approach. "I am impressed with the achievements attending the under-fives," says Tarsis Bwakila, clinical officer at the Ikwiriri Health Centre. "The way of managing diseases has improved. A good indicator of improved services is the increasing number of patients coming to our facility."

Another key feature of IMCI is that the parent, normally the mother, is integrally involved in the process. After the practitioner explains the main causes of the child's condition to the parent, the first dose of the appropriate treatment is given by the parent at the clinic, under the observation of the health worker. The health worker is therefore certain that the parent knows how to use the drug. (Diagrammatic instructions are also sent home with the parent as a reminder of how and when treatment is to be administered.) The parent is then told about the importance of proper and continued feeding (including the provision of more fluids), danger signs, and what to do if the child does not improve. To verify the parent's understanding, he or she is asked to recount the danger signs and the actions to take when these signs appear or when the child does not get better. This is part of an attempt to improve awareness and health practices in the home.

The selection of drugs also supports the practice of "syndromic" classification. Rather than prescribe drug A, which works very well against condition X, or drug B, which works very well against disease Y, the health worker has been taught that it is better to prescribe drug C, which does a reasonable job of combating both diseases X and Y. In other words, the idea is to choose drugs that have a broader spectrum of coverage. The IMCI methodology has proven highly adaptable to situations where adequately trained clinicians are in short supply. Personnel with limited clinical training can perform very well using the IMCI system. Indeed, because the process is guided by algorithms, lower skill workers may stick to the IMCI program more rigorously than better trained professionals and be less likely to revert to the use of "intuition."

The results of this process have been dramatic. The introduction of IMCI corresponded to a significant increase in the quality of child health services and a reduction of child mortality rates in Rufiji and Morogoro districts. In Morogoro, child mortality between the late 1990s and early 2000s has declined over 40% -- from about 35:1000 to around 20:1000 annual deaths in children under 5 years of age. An important share of this reduction has been shown to be the direct result of the IMCI system. 

In addition, there is reason to speculate that the corresponding lowering of adult mortality rates in those districts may also be related to improvements in skills brought about by the introduction of IMCI. Partly, this is likely the result of better health-seeking behaviour and patient attendance at facilities -- as adults see better results for their children, they are more likely to seek health care for themselves. In addition, adults may benefit from better treatment. As practitioners who deal with children also deal with adults, it is logical to expect that health workers would apply some of the organizational and clinical skills they learned in IMCI to the treatment of older patients.

Amadeus Mwananziche, medical assistant in charge at Mlali Dispensary, agrees with that assumption. "The training we received in IMCI and other areas has helped us a great deal in the management of other diseases," he says. Instituting an adult version of integrated management of illness is something being considered by WHO as a future health intervention, given the success of the "child version."

The two districts' positive introduction and experience with IMCI also underscores the critical need for continual retraining and upgrading of staff. This, in fact, is just one of many lessons arising from the TEHIP experience, which are explored in the
next chapter.

Conclusion

The multitude of program initiatives set in motion by TEHIP are wide-ranging but not scattered. They form a coherent whole that, we believe, has helped dramatically reduce infant, child, and adult mortalities in a population of approximately 741 000 (greater that the population of 66 of the world's countries).

In short, the relationship between these seemingly disparate program elements can be summarized as follows:

Document(s) 5 of 10