DEVELOPING countries could reduce their current burden of disease by 25 percent if governments were to make an essential national package of health services available to everyone. This is well within the reach of most developing coutries.
When it comes to health policy, one of the most difficult decisions for developing country governments to make is how best to put together a mix of health services that will be financed by public spending. Ideally, they would like to offer as much health care as possible to as many people as possible. But in practice, they end up concentrating resources in urban hospitals, which provide a wide range of services for a few, leaving other population groups, particularly in rural areas, with relatively little access. This allocation of public spending is inequitable and inefficient: costly treatments are prescribed that prolong life only slightly, while large populations are denied inexpensive services that extend life greatly, such as immunizations. As a way of resolving this problem, the World Development Report 1993 recommends that governments design and finance a national health package, embracing essential public health and clinical services that will substantially reduce the burden of disease (the present value of future streams of disability-free life lost as a result of death, disease, or injury) at affordable costs. This means that governments will need to review the value for money being obtained from the current mix of interventions they offer so that they can reallocate resources in the most cost-effective manner.
Value for money
No matter how health services are organized and paid for, what they actually provide are health interventions—specific activities meant to reduce disease risks, treat illness, rehabilitate, or palliate the consequences of disease and disability. Debates about whether health services should concentrate on “vulnerable groups” such as children, pregnant women, and the elderly, or about the relative roles of hospitals versus health centers, or about preventive versus curative activities, are at bottom debates concerning the proper mixture of interventions. In health, as in every other sector, customers want value for the money spent.
That is why the first step in designing a country’s essential health package is to determine the cost-effectiveness of a health intervention—the net gain in health compared with doing nothing, divided by the cost. Indeed, the developing countries that have been the most successful in improving health for a given level of spending have concentrated their public monies on highly cost-effective interventions. These success stories include China, Costa Rica, Chile, Sri Lanka, and Thailand—all countries with different income levels and political systems.
Cost-effectiveness of health interventions varies greatly
Source: Jamison and others Forthcoming; World Bank data.
Note: DALY, disability-adjusted life year. Interventions are specific activities intended to reduce disease risks, treat illness, or palliate the consequences of disease and disability; an intervention-year is an intervention repeated throughout the year rather than provided only once.
1 Includes some interventions that benefit all age groups.
When governments try to determine which interventions should be put in an essential national health package, they need to calculate the cost-effectiveness of the various possibilities. This chart shows the dollar costs and gains in DALYs for 47 key interventions.
Higher points represent interventions that are more effective in improving health as measured in DALYs saved; points farther to the right represent lower-cost interventions. Some interventions cost less than $1 per person benefitted, while others cost over $10,009. Some interventions add more than ten years of healthy life, while for others, the gain is equivalent to only a few hours or days of full health. Both axes are drawn on a logarithmic scale, so that the diagonal lines show equal cost-effectiveness ratios in dollars per DALY. These ratios vary widely, from as low as $1 per DALY to as much as $10,000. Higher lines represent more cost-effective interventions.
Four interventions illustrate the extremes of cost and health gain: vitamin A supplementation (low cost, high gain), chemotherapy for tuberculosis (high cost, high gain), environmental dengue control (low cost, low gain) and treatment of childhood leukemia (high cost, moderate gain).
How is cost-effectiveness determined? Given a common currency in which to measure cost and a unit in which to measure health effects, different interventions can be compared by what it costs to achieve one additional year of healthy life. Outcomes are measured in the same unit of disability-adjusted life years (DALYs) used to estimate the burden of disease. The ratio of cost and effect, or the unit cost of a DALY, is called the cost-effectiveness of the intervention; the lower that number, the greater the value for money offered by the intervention.
Despite the usefulness of this approach, to date, only a small share of the thousands of known medical procedures have been assessed for cost-effectiveness. Even so, the interventions that have been evaluated could deal with more than half the world’s disease burden. Just implementing the 20 most cost-effective ones could eliminate over 40 percent of the total burden and three quarters of the health loss among children.
The WDR’s cost-effectiveness estimates were largely drawn from “Disease Control Priorities in Developing Countries” (Jamison et al, forthcoming, World Bank), with the estimates based as far as possible on actual conditions in developing countries.
The WDR found huge differences in both the cost and the effectiveness of various health interventions, underscoring the importance of making such calculations (see chart). Indeed, because the value of interventions can differ so much, making allocative decisions badly costs lives. Spending $100,000 on chemotherapy for tuberculosis would achieve a gain of 35,000 DALYs, since about 500 patients would be saved, and they would be prevented from infecting others. But the same expenditure on diabetes management would save only 400 DALYs, because although 500 patients would benefit, each patient would gain less than one healthy year from a year of treatment, and there would be no significant benefits for others.
In addition, the results demonstrated that—confirming previous notions—many of the most cost-effective health interventions are preventive in character or fall into the category of primary care. For example, programs aimed at reducing childhood malnutrition and mortality, chiefly from infectious diseases, should continue to be supported by primary health care programs.
However, the findings also showed that not all preventive measures rank high in terms of cost-effectiveness (e.g., spraying to control the mosquitoes that carry dengue or malaria, and often water and sanitation investments), whereas several hitherto neglected interventions are very cost-effective: chemotherapy against tuberculosis; integrated prenatal and delivery care; mass programs to de-worm children; condoms plus information and education to combat AIDS; and measures against smoking, such as education and consumer taxes on tobacco products.
Content of the package
The WDR has come up with a suggested minimum package of health services, which is affordable by the majority of developing countries at current levels of health spending and would reduce the burden of disease by just over 30 percent in low-income countries. Eleven clusters of interventions or individual interventions are included in the package; apart from being cost-effective, these services address diseases responsible for a large share of the disease burden in developing countries (Table 1).
|Clusters of interventions||Main disease conditions addressed|
|Expanded program on immunizations||Measles, poliomyelitis, diphtheria, tetanus, whooping cough, yellow fever, and hepatitis B|
|School health program||Intestinal worms|
|AIDS prevention program||Sexually transmitted diseases and AIDS|
|Tobacco and alcohol control programs||Lung cancer, cardiovascular disease, cirrhosis, and injuries associated with alcohol abuse|
|Other public health programs (including family planning, health, and nutrition information)||These are not disease specific|
|Short course chemotherapy for tuberculosis||Tuberculosis in adults|
|Management of the sick child||Diarrheal diseases, pneumonia and other respiratory infections, malaria, measles, and severe malnutrition|
|Prenatal and delivery care||Perinatal mortality and morbidity, complications of pregnancy and delivery, low birth weight, unwanted pregnancies, and congenital syphilis and gonorrhea|
|Family planning||Perinatal and infant mortality and maternal mortality and morbidity|
|Treatment of sexually transmitted diseases||AIDS, syphilis, gonorrhea, chlamydia, and other sexually transmitted diseases|
|Limited care (mainly for adults)||Pain control, infection and minor trauma treatment, and advice to reduce chronic diseases|
However, the exact content of each country’s essential package will be largely determined by the epidemiological profile of the country (the distribution of disease burden across diseases) and the cost-effectiveness of the corresponding interventions. The size of the package (number of intervention clusters) will depend on the financial resources available for health care.
Clustering interventions improves cost-effectiveness through at least three mechanisms. Synergism between treatments or prevention activities is common, particularly in pediatric care. Joint production costs can substantially reduce the amount of resources needed were interventions to be provided separately. And the optimal use of specialized resources, such as hospital beds, requires a screening process to refer the most severe cases from the first level of care to other facilities.
An efficient health cluster should include interventions that can be given to the same individual, at the same time, and through the same mode of delivery (outreach community health worker, health center, or hospital). The expanded program on immunizations, for example, is a very efficient one, because it includes six vaccines provided through the same delivery system to the same individual, often at the same time.
How would the proposed minimum package affect the various populations group? For children under five, who contribute disproportionately to the burden of disease, interventions to control the major diseases in early childhood are among the most cost-effective. Five of the 11 clusters produce large benefits for this group: the expanded program on immunizations, the school health program, family planning, management of the sick child, and prenatal and delivery care. The WDR did not consider the synergism likely to occur when these interventions are provided together, so the estimates are on the conservative side.
As for adults, the minimum package addresses three major risk factors, which together account for as much as 20 percent of the burden of disease in this group: tuberculosis infection, dangerous sexual behavior, and tobacco smoking. Tuberculosis in India and in sub-Saharan Africa is responsible for about 8 and 11 percent, respectively, of the burden of disease in those aged 15–59. Sexually transmitted diseases are one of the most important causes of disease burden in young adult women, with AIDS now exacerbating the problem. And in many middle-income countries (e.g., the former socialist economies and China), tobacco consumption is responsible for a large proportion of premature deaths, mainly due to lung cancer and cardiovascular diseases. This burden is likely to rise rapidly.
Regarding mothers, the good news is that maternal mortality and morbidity, which are responsible for 18 percent of the disease burden in women aged 15–44, can be effectively prevented in developing countries. The package includes two powerful clusters of interventions to reduce substantially the burden of disease associated with reproduction. Family planning could reduce the number of maternal deaths by 25 percent, and universal access to prenatal and delivery care, another 40–50 percent.
Costs and payment
For low-income countries, the minimum package is estimated to cost $12 a person per year (Table 2). This rises to $22 a person per year in middle-income countries. The differences are due mainly to different demographic structures, epidemiological conditions, and labor costs in the two settings. About one third of this would go for public health and two thirds for essential clinical or individual services. Under these assumptions, total cost for all developing countries would be about $60 billion, which is less than total public spending on health in those countries now and only about 4 percent of world total health care spending.
|Country group and package component||Total current health spending per capita (dollars)||Cost of proposed package per capita (dollars)||Reduction of disease burden (percent)|
|All developing countries||41||15||25|
|Essential clinical services||—||10||19|
The cost of the essential package is best calculated at the country level in order to incorporate different epidemiological and demographic conditions and price-level differentials. It is clear, however, that in low-income countries, where governments typically spend about $6 per person for health and total health outlays are only about $14 per person, the affordability of the minimum package is problematic. This would require an increase in public spending for health, as well as a reorientation of current government health outlays away from discretionary, less cost-effective interventions to essential care. In middle-income countries, where public spending for health averages about $62 per person, the $22 minimum package is financially feasible. What is required is a political commitment to shift existing resources in the direction of public health and essential clinical care.
Middle-income countries with public health expenditures above $50 per person should expand the minimum package to cover interventions with slightly less favorable cost-effectiveness than those in the minimum package to address a wider spectrum of diseases and injuries. An essential package would logically address tropical diseases peculiar to the country (e.g., Chagas disease in Brazil) and include public health measures and clinical interventions to reduce injuries (mainly from motor vehicle collisions) and fight major noncommunicable diseases.
Who should pay for the essential package? There are strong efficiency arguments for directing government funding to public health interventions, because of the public goods nature of these services. And a number of the essential clinical services, including treatment of tuberculosis and sexually transmitted diseases, have large positive spillover effects. In addition, there are equity grounds for financing the package. The poor are disproportionately affected by the disease burden the package addresses, making public finance of the package an effective mechanism for reaching the poor. The main problem with universal government financing of an essential package is that it leads to public subsidies to the wealthy, who can afford to pay for their own services, with the result that fewer government resources go to serve the poor.
One way to solve this problem is by targeting public spending to the poor. In low-income countries, where current public spending for health is less than the cost of the minimum package, some targeting is almost inevitable. In countries where the wealthy do not use government-financed services because of the greater quality and convenience of privately financed services, targeting may be fairly easy. In middle-income countries, where a significant part of the population may be covered by private or social insurance, governments can mandate that the essential package be covered by all insurance policies.
The most sophisticated facility required to deliver the minimum elements of the package is a “district” hospital, which serves as the first level of referral from health centers. These hospitals offer basic surgery, emergency services, and some outpatient care; generally have 100–400 beds; and serve 50,000–200,000 inhabitants. The minimum package requires access to health centers and district hospitals throughout the country. On average, it requires about 1 district hospital bed, 0.1 to 0.2 physicians per 1,000 population, and 2 to 4 nurses per physician.
Governments can direct public spending to support the nationally defined essential package in several ways:
Where services are publicly financed and provided, they can reallocate public spending toward inputs—drugs, supplies, equipment, staff, and facilities—that support the package. In many countries, extending the basic infrastructure throughout the country and improving the quality of services at existing lower-level facilities are necessary steps to delivering the package. At the same time, governments can eliminate or greatly reduce financing of inputs for less cost-effective services. This might include closing wards or converting specialized hospitals to serve as district hospitals and curtailing the number of specialist physicians. In Canada, for example, provincial governments—not individual hospitals—control the decision to acquire technically advanced diagnostic tools.
At the same time, providers’ treatment decisions would not be micromanaged; they would be influenced, instead, by the nature of input availability. The specialized staff and equipment, for example, would not be available to treat certain highly fatal forms of cancer, but drugs would be available to provide pain relief. Similarly, staff, diagnostic services, equipment, and drugs would be available for treating malaria in young children. Budgetary and salary incentives could also be used to reward individual providers, facilities, or districts that achieved good coverage of the population with the services in the package.
Where services are publicly financed but privately provided, governments should reimburse only for those services in the essential package. This model of health care delivery is growing, but it is still uncommon in developing countries. At present the regulatory capacity to oversee such arrangements is poorly developed.
Adopting an essential national package of health interventions to be financed publicly, at least in part, and to be required of all privately financed insurance would help solve several problems in the allocation of health care resources.
• It would assure that public spending was concentrated on those interventions yielding the most health gain per dollar. This is not only good for health; it is likely to increase taxpayer satisfaction with the use of public money.
• By stating clearly what outputs the system is expected to provide, it would make it easier to estimate input requirements.
• It would clarify the boundary between public and private outlays, providing a basis for determining what to subsidize and what to leave to the private market.
• The definition of an essential package might also affect private spending by drawing attention to what is most cost-effective and regulating insurance accordingly.
• It would increase equity and help to either reduce poverty or alleviate its consequences by assuring that, at the least, the most justifiable health care is provided to the poor.