Levels of Care
Health interventions need to reach people either by being provided at their homes, schools, and workplaces or by encouraging them to visit health facilities. Programs based in communities can reduce the costs and barriers that impede people's access to services, while general primary care can act as an interface between community health programs and individual clinical care, whether ambulatory or inpatient. District and referral hospitals are needed to provide more specialized or costly care to reinforce community and primary care services with interventions that are required when these levels cannot bring the specialized equipment or skills to bear.
Many countries have attempted to construct links between community-based health care resources and households for a range of health programs.1 These programs do not substitute for a health system, but they provide a channel for reaching families with information and resources. They also mobilize additional resources, such as volunteers' time, local knowledge, and community confidence and trust.
Community-level programs can include a range of interventions but generally focus on services related to safe motherhood, nutrition, and simple prevention and treatments. They commonly include the following services:
reproductive health and maternal nutrition
growth monitoring and promotion
supplementary feeding with either external supplies or local supplies
"Community-level programsgenerally focus onsafe motherhood, nutrition, and simple prevention and treatments."
These interventions collectively reduce such risk factors as malnutrition that account for as much as 40 percent of the disease burden in low- and middle-income countries.
Community programs are organized in a variety of ways, particularly with respect to the status and number of community workers. At one end of the spectrum are community-based programs that rely heavily on local resources and volunteer time. For example, Thailand recruited and trained 60,000 village health volunteers who were responsible for mobilizing and supervising 600,000 village health communicators, who in turn attended an average of 20 children each (box 6.1). At the other end of the spectrum, some countries recruit workers from target communities but then hire and supervise them as part of the public health workforce. In Costa Rica, for example, health workers were full-time employees, recruited and managed by the government's rural health program. Costa Rica hired two full-time health workers for every 350 children. DCP2 assesses the evidence and concludes that an effective ratio of health workers to children is in the range of 1 to 500 for full-time, paid community workers and 1 to 10 or 20 for local, part-time volunteers.
"an effective ratio of health workers to children is in the range of 1 to 500 for full-time, paid community workers and 1 to 10 or 20 for local, part-time volunteers."
In addition to developing staffing strategies, community programs must decide whether to include food supplementation to address malnutrition. Food supplementation can help a community health program achieve its goals when integrated with other services or it can become the program's primary focus to the exclusion of other health services. In some communities, food supplementation is part of a broader program that also addresses micronutrient supplements, pre-school education, growth monitoring, and sometimes improvements in local food supplies. At the other end of the continuum, the food distribution component has dominated some community programs, as occurred with India's Integrated Child Development Services Program.
Community programs also need to find a balance between health promotion activities and curative care. When primary facilities are distant or poorly supplied, community workers may be pressured into providing direct curative care, distracting them from other health promotion activities. When this happens, community systems may be coopted and "medicalized" with the addition of diagnostic and treatment modules. While such additions may succeed in giving people access to services that would otherwise be absent, they also divert community programs from the initial purpose of general health promotion. Community health workers not only can promote healthy behaviors and preventive action but can mobilize demand for appropriate services at other levels.
"Community health workers not only can promote healthy behaviors and preventive action but can mobilize demand for appropriate services at other levels."
Finally, community-based programs must seek to maintain a balance between extensive coverage and adequate supervision. A program that supervises tens of thousands of volunteers needs hundreds of supervisors. DCP2 finds that supervision ratios in effective programs are approximately 1 supervisor for every 20 community health workers. If this supervisory responsibility is laid on primary care facilities, it can become an excessive burden. If it is managed well, however, it can be the difference between ineffective and effective programs.
"Effective health programs build on established community practices."
Effective health programs build on established community practices. Hence in Thailand, health services are combined with religious organizations at the village level. In Indonesia, social organizations have played an important role. In Bangladesh, NGOs that had been successful in other areas, such as food security and education, became active in health. In Costa Rica, Honduras, and Jamaica, national heath services already had a sufficient local presence to initiate community-based health programs.
Organization at the local level can accomplish a good deal, but only with adequate guidance and resources from more central levels. Training, supervision, and supplies must be sustained to have a lasting effect. Generally this means linking community health workers with primary care facilities.
Thus the success of community health efforts depends critically on the context, including level of development of infrastructure, services, and socioeconomic resources. For example, in Indonesia, the weakness of the primary health care system encouraged community programs to shift toward providing individual clinical services. In Costa Rica, by contrast, community health programs could focus on prevention and health promotion as a complement to a stronger primary care system.
General Primary Care
The term primary care denotes several different, yet related, aspects of health systems.2 In some contexts it refers to certain activities, such as immunization and prenatal care. In others it denotes a level of care with relatively low technical and skill requirements. For some countries it is a strategy for structuring and managing health care. For others it is a perspective or philosophy. The broad concept of primary care includes a range of initiatives that are associated with the Alma Ata Declaration on Primary Health Care approved by WHO in 1978. More recently, the WHO Commission on Macroeconomics and Health described the need for developing services that are close to the client. Despite the variations in the specific use of the term primary care, the basic notion is a common one: recognition that a certain range of health care services must act as an interface between families and community programs on the one hand, and hospitals and national health policies on the other.
"primary caremust act as an interface between families and community programs on the one hand, and hospitals and national health policies on the other."
Since the 1978 Alma Ata Declaration, WHO, the World Bank, and specific countries have refined and constructed alternative packages of services under the umbrella of general primary care. DCP2 notes substantial convergence in the content of general primary care over time: maternity-related care (for instance, prenatal care, skilled birth attendance, and family planning), interventions to address childhood diseases (such as vaccine-preventable diseases, acute respiratory infections, diarrhea, and malnutrition), and prevention and treatment of major infectious diseases. The list is familiar from numerous studies of cost-effective packages and priorities for the global disease burden (table 6.1).
Nevertheless, local health facilities that are equipped exclusively to carry out these kinds of functions may not meet local demand for other kinds of curative care or may miss important local health threats altogether. Consequently, public health experts stress the importance of having a local management team to plan local care and support services for a defined population, ranging from 10,000 to 50,000 people. With a local management team responsible for a specific population, that team can set priorities and monitor progress, as well as ensure a good fit between national priorities and local needs and demand for health promotion and treatment.
"general primary care can potentially address up to 90 percent of health care demand in developing countries."
In practice, achieving this kind of planned, local effort is difficult in low- and middle-income countries for several reasons:
Primary care facilities frequently lack the resources they need to function.
Staff positions may remain unfilled or staff members may be absent.
Supplies may not be delivered or may have expired.
Facilities may not be properly maintained.
People often seek health care from a variety of traditional healers, pharmacists, and private medical professionals in addition to public services. This fragmentation can make proper surveillance and planning difficult to manage.
With effective use of adequate financial, institutional, and human resources, general primary care can potentially address up to 90 percent of health care demand in developing countries. The direct effect of general primary care is well documented. For example, local reductions of 5 to 32 percent in mortality among children in Liberia, Niger, and the Democratic Republic of the Congo are attributed to the provision of general primary care in these locations. A well-functioning general primary care system is also integral to the success of a health system overall, because it acts as the bridge between local care and care at the next levels, such as district and referral hospitals.
In most countries, district hospitals account for the largest share of inpatient services.3 They are generally designed to serve from 100,000 to as many as 1 million people with services that are more sophisticated, technically demanding, and specialized than those available at a primary care facility, but not as specialized as those provided by referral hospitals.
The range of services district hospitals offer includes diagnostics, treatment, care, counseling, and rehabilitation. The technical demands of these services require professionals with training and experience spanning the fields of family medicine and primary health care, obstetrics, mental illness, eye health, rehabilitation services, surgery, pediatrics, and geriatrics. They require substantial capital investment in facilities, equipment, and management. District hospitals may also provide health information, training, and administrative and logistical support to primary and community health care programs. When a district hospital's service area coincides with a local government administrative unit, it may be responsible for other public health functions throughout the district.
The strength of a district hospital is that it concentrates skills and resources in one place for conditions that are either uncommon or difficult to treat. It is also a repository of knowledge and diagnostic tools for assessing whether referral to an even more specialized facility is indicated. A district hospital can only realize these strengths if it is properly integrated with other levels of care that are also functioning well. If primary care facilities are not meeting local needs, for example, then people will bypass them and overwhelm hospitals with demands for services that could be provided more effectively and efficiently in other settings. Primary care facilities must also screen patients to identify those who do require hospital attention. Timeliness and adequate transportation are essential, as no amount of screening or referral can make a difference to a patient stranded in a distant village.
At the same time, district hospitals' concentration of resources give rise to their potential weakness. Too often, district hospitals benefit those who live nearby and are not readily accessible to the poor or to those dispersed in rural areas. District hospitals can serve populations most equitably when their concentration of resources is accessible to all, that is, when barriers created by poverty, low service quality, costly transportation, or remote geography are addressed.
"Too often, district hospitalsare not readily accessible to the poor or to those dispersed in rural areas."
Costs of care in district hospitals are sensitive to the salaries paid to their staff, the utilization rates, and the inclusion of additional health functions. Staff salaries generally account for the bulk of hospitals' recurrent costs even when salaries are low. When utilization rates are high, the average fixed cost per patient day will be lower. In some places, hospital utilization rates are below 50 percent and the unused capacity represents a substantial economic loss for the health system. In other cases, hospitals are overutilized, and even though this reduces average costs, it results in more rapid depreciation of facilities. Additional functions also raise hospitals' costs. These functions might include training new health professionals and providing continuing education for them; supervising, supporting, or managing primary-level services; and designing and implementing district-wide public health campaigns.
"a Kenyan district hospital in a rural communityserved 2,223 children, spent about US$10 per patient, and saved an estimated 215 lives at an average cost per life saved of US$104"
Cost studies indicate the potential range of hospital costs in low- and middle-income countries. A Tanzanian hospital spent approximately US$4 per patient per day, but the study argued that it was underfinanced and that a little more than US$12 per patient per day would be necessary to provide care according to the provider's own standards. Researchers estimated that inpatient care in Kenya cost about US$9 per patient per day and in Bangladesh cost almost US$16 per patient per day. Costs for South African hospitals in five districts ranged from US$40 to US$97 per patient per day.
DCP2 includes an exercise to estimate the cost-effectiveness of district hospital care. Despite the exercise's admitted roughness, it gives a sense of the possible order of magnitude. Using data from a study of a Kenyan district hospital in a rural community with reasonably good access to health care, it finds that the hospital served 2,223 children, spent about US$10 per patient, and saved an estimated 215 lives at an average cost per life saved of US$104, implying a cost per DALY averted of only about US$4 to US$5.
Strategies for improving district hospital care vary. In many countries, district hospitals have been turned over to local governments as part of a decentralization of public services. In others, hospitals are given greater decision-making, or even financial, autonomy. In parts of Central Asia and East Asia, particularly in the former Soviet republics and in China, public hospitals have become so dependent on fees paid by patients that they effectively function as private institutions. Most public hospitals receive budgets that are based on their staffing and size, but reforms in some places have introduced reimbursement based on the number and complexity of services provided, with mixed results. In other places, efforts focus on improving the quality of care in hospitals. One goal is to reduce hospital-acquired diseases, a serious problem in resource-constrained Sub-Saharan African countries, where blood transfusions and the re-use of needles can transmit HIV, hepatitis, and other infections. Improving blood safety would cost less than US$8 per DALY averted.
"in the former Soviet republics and in China, public hospitals have become so dependent on fees paid by patients that they effectively function as private institutions."
District hospitals are subject to various pressures that affect how well they carry out their role. Some of these pressures force district hospitals to intervene as if they were primary centers, while others push them toward functioning as public health management centers. The appropriate mix will result from balancing investments in district hospitals with investments in other levels of care.
The next level of care is the referral hospital, which provides complex clinical care to patients referred from the community, primary, or district levels.4 Referrals explicitly link the different levels of health care in both directions. Community, primary, or district facilities direct patients to a specialized hospital for care. The referral hospital, in turn, provides support and information to assist the other levels and refers patients back to them when appropriate. For the full benefits of linkages between levels of care to be realized, referral hospitals need to provide many forms of support, including advice on which patients to refer, proper postdischarge care, and long-term management of chronic conditions. Coordinated training in the use of shared protocols is also necessary, as is technology support by skilled technicians and scientists.
"For the full benefits of linkages between levels of care to be realized, referral hospitals need to provide many forms of support"
Referral hospitals can also provide important managerial and administrative support to other facilities, serving as gateways for drugs and medical supplies, laboratory testing services, general procurement, data collection from health information systems, and epidemiological surveillance. Sometimes they take on the role of managing transportation for medical supplies and staff, or even financial, payroll, and human resource management for lower-level facilities.
Other important functions of referral hospitals are research and training. In industrial countries, referral hospitals may be developing new technologies, but in developing countries they are more likely to be involved in research for piloting and introducing technologies that have been developed elsewhere, that is, assessing these technologies for their effectiveness and adaptability to a new context. Referral hospitals become the vehicle for disseminating such technologies by training new staff and providing continuing professional education for existing staff at different facilities. Research that is responsive to local disease burdens and local technology constraints fills a critical gap, because industrial countries and pharmaceutical companies do not generally undertake such research if they do not foresee sufficient returns to their investments. Research activities also help attract and retain the specialists needed to treat complex cases and train new specialists.
As with district hospitals, referral hospitals in low- and middle-income countries frequently end up offering a full range of services, from the most specialized to basic ambulatory treatment. The demand for basic care results from people bypassing poorly equipped or inadequately staffed lower-level facilities.
The costs and effectiveness of referral hospitals are highly sensitive to the range of services they provide, to staff wages, and to utilization rates. In general, they tend to be more expensive than district hospitals because they treat more complex cases, have more expensive inputs, and are also engaged in teaching and research. Studies indicate that per bed day, referral hospitals can be two to five times as costly as district hospitals.
Referral hospitals tend to be located in large urban areas, exacerbating unequal access to specialized care for rural and generally poorer citizens. Because referral hospitals are by definition specialized, redressing such inequities by constructing more facilities is not feasible. Rather, equitable access to referral hospital services requires improving referral from other levels of care and reducing transportation costs and other financial barriers for the poor.
Investments in and functions assigned to referral hospitals need to be properly balanced and coordinated with investments and functional assignments to district hospitals, primary care centers, and community health workers. Just as referral hospitals cannot function efficiently without the other levels fulfilling their roles, so too community, primary, and district hospital levels cannot function effectively without the ability to refer complex cases to specialized hospitals. Lower levels of care certainly require strengthening, but this is more likely to reflect inadequate financing of the entire public health system than a grossly excessive allocation to referral hospitals.