While some countries still have the opportunity to avert a full-scale AIDS epidemic by acting early to change the behavior of those at highest risk, others already have large numbers of infected people across many groups in the population. Chapter 1 presented evidence of the terrible impact of HIV/AIDS on individual welfare, in terms of human suffering and losses in life expectancy. What can be done to mitigate the impact of the AIDS epidemic on people and society? There are many impacts of the AIDS epidemic that cannot be quantified—for example, the emotional pain experienced by infected individuals and their families and the psychological damage wrought on surviving family members. These impacts are very important, but how to respond to them is beyond our expertise and best left to others. This chapter considers the economic aspects of three types of impacts—on infected individuals, on the health sector generally, and on surviving household members—and the ways in which government policies can help people to cope, given the many other pressing demands for scarce public resources.1

The first part of the chapter shows that there are affordable, effective, and humane ways for governments in low-income countries to help ease the suffering of individuals infected with HIV. However, both governments and individuals in the poorest countries should be wary of funding expensive treatments with uncertain benefits. The second part of the chapter suggests how governments can cope with the increased demand for and scarce supply of health care brought on by the AIDS epidemic in ways that are effective and compassionate, as well as fair and affordable. The third part proposes a strategy for developing countries to address the needs of poor families hit by the AIDS epidemic in the context of other poverty programs. The chapter concludes with a summary of the policy recommendations for governments attempting to cope with the impact of HIV/AIDS on health care and poverty.

What is the health impact of HIV/AIDS on an infected individual over the course of the disease? Are there effective, affordable treatments for people with AIDS in low-income countries? To answer these questions, this part of the chapter reviews the many illnesses that often afflict people with HIV/AIDS, the available treatments, and their cost. It distinguishes between three types of care: relief of symptoms, such as headache, pain, diarrhea, and shortness of breath, which is sometimes called palliative care; prevention and treatment of opportunistic illnesses (OIs); and antiretroviral (ARV) treatments, which attempt to combat HIV itself. Next it presents the amounts that developing countries are actually spending to care for people with HIV/AIDS. While this amount is often large relative to a country’s GNP per capita, it is usually too little to buy all the drugs needed to treat opportunistic illnesses, much less to pay for antiretroviral therapy. The section closes with a review of programs to assist with the home care of people with HIV/AIDS.

The discussion finds that although treatment of HIV itself is difficult and extremely expensive, some of the symptoms and opportunistic illnesses typically suffered by people with AIDS can be treated simply and at low cost. Some infectious diseases associated with HIV, especially tuberculosis, are somewhat more expensive to treat, but because they are infectious there are sound reasons for governments to subsidize treatment of any infected individual who would not otherwise get treated, regardless of the individual’s HIV status.

The pattern of opportunistic illnesses differs from country to country, depending on which diseases are prevalent, and the quality and amount of treatment available. The natural history of HIV illness and several of the most important opportunistic illnesses are defined in box 1.2. Figure 4.1 presents the proportion of AIDS patients who suffer from each of three OIs—tuberculosis, cryptococcosis, and Pneumocystis carinii pneumonia (PCP)—in six developing countries and the United States. Tuberculosis is most common in the three poorest countries, the Congo DR (formerly Zaire), India, and Côte d’Ivoire, becoming less common as per capita income rises. At the other end of the income gradient, PCP is most common in the United States, and is also common in the middle-income developing countries, Brazil, Mexico, and Thailand, but is rarely reported in the three lower-income countries. Cryptococcosis, a generic name for a group of fungal diseases that includes cryptococcal meningitis, shows no consistent pattern by income level, but infects at least 5 percent of people with HIV in all six countries. Among these three diseases, and indeed among all OIs, tuberculosis spreads most readily from people with HIV to others. As we discussed in chapter 1, tuberculosis greatly exacerbates the health impact of HIV in many developing countries, particularly in Africa and India, where it is the most common opportunistic infection.

Because of the variation in symptoms and in opportunistic illnesses, the cost and number of health care episodes for an HIV-infected person vary widely. Table 4.1 presents rough estimates for the average costs of pharmaceutical and inpatient palliative care of symptoms, prevention of tuberculosis and PCP, and curative care of the more common opportunistic illnesses. Estimated lifetime cost per patient for this care ranges from $300 to $1,000, depending on which drugs are used and the cost per day of inpatient care.

How effective are these treatments? In the early stages of HIV illness, palliative treatment can inexpensively relieve some of the pain, discomfort, and incontinence that otherwise rob people of the ability to enjoy life and contribute to their family and their community. Without symptomatic treatment, dehydration that results from diarrhea and nausea can kill in a few days. Fever and headache can be disabling for days or weeks. As shown in the top panel of table 4.1, drugs for palliative care are quite cheap. Hence, all but the poorest HIV-infected patients and their families are likely to be willing and able to buy these drugs, provided they are available. The sad truth is that these drugs are often not available, an issue we discuss below.

Moving down the table, we see that the opportunistic illnesses that commonly arise early in the course of AIDS can also be treated quite inexpensively. Treatment for thrush, toxoplasmosis, and pneumonia/septicemia can buy one to four years of life at an additional drug cost of $30 to $150—all but the very poor would probably be willing and able to pay for these treatments.

Rarer opportunistic illnesses like the fungal diseases tend to occur later in the course of the HIV infection and are more difficult and expensive to treat. For example, in the United States the average life expectancy after diagnosis with cryptococcal meningitis, the most common of the cryptococcosis diseases, is 320 days, while in the Congo DR, perhaps because of later diagnosis, this drops to 180 days, even with expensive state-of-the-art drugs (background paper, Perriëns 1996). Since a patient in the Congo DR might survive 30 days without treatment, such drugs would extend life by about 150 days for about $870. In Thailand earlier diagnosis would result in treatment extending life by perhaps 330 days for $1,740. Many patients in these two countries might decide against buying these drugs, even if they have the money to do so.

Table 4.1 Annual Cost per Patient of Palliative Care and Treatment of Opportunistic Illnesses, Sub-Saharan Africa and Thailand (1996 dollars)

In the final stage of AIDS, the immune system is so weak that a variety of infections spread throughout the body, leading to death. At this point, morphine to assuage extreme pain and the sensation of suffocation provides relief to the dying patient, and this in turn helps to ease the distress of the patient’s family. If purchased in bulk at international generic prices, enough morphine to ease the last two weeks of life would cost less than $4. But because of international controls on morphine distribution, this essential drug is rarely legally available in poor countries at any price.

The foregoing discussion has shown that many of the symptoms and opportunistic illnesses that occur in the early stages of AIDS can be effectively treated at low cost. Unfortunately, the low-cost generic forms of the needed drugs are often unavailable; even when they are available, people often lack information about their efficacy. Thus, many people pay much more than the $10 to $20 cited in the table for palliative treatments, while achieving no additional benefit. Governments can address these problems by facilitating the availability of generic drugs needed for palliative care and common opportunistic illnesses. For example, countries with a concentrated or generalized epidemic could add these medications to their list of "essential drugs," which are widely distributed. Governments can also help patients to make informed decisions by ensuring access to reliable information about the efficacy of various treatment options, both pharmaceuticals and traditional remedies. The degree of government subsidy for treatment will depend on the country’s overall health financing policy. We discuss this issue later in the chapter.

The treatments discussed above ease suffering and prolong life but ultimately fail to save the patient’s life because none attacks the underlying cause of illness—the continued spread of HIV within the body and the consequent decline of the immune system’s ability to recognize and repel biological threats. A few drugs have reduced the levels of HIV in the patient’s blood below the ability of laboratory tests to detect it. Unfortunately, these drugs are expensive and complex to administer, their long-term benefits are uncertain, and their efficacy varies greatly from one individual to another.

The first drug that showed evidence of inhibiting the spread of the virus in an infected patient was Zidovidine (AZT, or ZDV). When AZT was introduced in the late 1980s, the cost of a year’s dosage was about $10,000 in industrial countries. By 1997 the cost of a year’s dose had fallen to about $2,738 in industrial countries, while Thailand and a few other developing countries had negotiated bulk purchases for as little as $657 per patient per year. However, except for prevention of mother-to-child transmission, AZT rarely provided dramatic benefits, adding perhaps six months of healthy life for the average patient (Prescott 1997; Perriëns and others 1997)

A more effective therapy involving the use of three antiretrovirals was announced in June 1996. A year later, the U.S. government issued draft guidelines recommending early, aggressive treatment of HIV-infected individuals with triple-drug therapy (Brown 1997). However, it was clear that more time would be needed before the new therapies could be fully assessed. Some individuals taking the medicines in clinical trials have dramatically improved their health and no longer have detectable levels of viral RNA activity. Yet even among these patients the virus may only be hiding and could re-emerge. And other patients show little or no reduction in viral levels, while still others cannot tolerate the drugs. As of mid-1997 no studies had yet been completed estimating the average percentage of patients who could benefit from triple-drug therapy or the characteristics of patients most likely to respond favorably—or to relapse.

Does triple-drug therapy offer reasonable hope for treating the disease in developing countries? Even if the therapy is shown to be generally effective, three substantial problems will remain: the cost of the drugs themselves, the costs and difficulty of the monitoring needed for the therapy to be effective, and problems with patient compliance. Although all of these problems also exist in industrial countries, they are likely to be especially severe in developing country medical settings.

Table 4.2 shows the cost of the drugs and necessary monitoring in Thailand, one of the few developing countries where the therapy is available, and the United Kingdom or United States, and hints as well at the great complexity of regimen. Because most of the drug costs and all of the monitoring costs are lower in Thailand than in the two industrial countries, overall costs are a minimum of about $8,000 per year in Thailand, compared with a minimum of about $12,000 per year in the United Kingdom and United States. These costs are likely to decline over time, perhaps substantially. But even if costs fell to one-hundredth of current costs, or about $80 dollars per person per year, they would still be several times the total annual per capita expenditure on health in many low-income countries. Moreover, effective antiretroviral ther-apy requires a highly trained, specialized physician working in a well-equipped clinic with experience performing a wide range of sophisticated tests and procedures, all of which are in critically short supply in most developing countries.

Table 4.2 Annual Cost of Antiretroviral Therapy, Thailand, and the United Kingdom or the United States (dollars)

a. Triple-drug therapy consists of two of first group of drugs plus one of second group plus monitoring. Drugs are given daily. Which three drugs should be combined is a matter of current research and probably varies by patient.

Source: Background paper, Perriëns 1996; Prescott and others 1997; and Moore and Bartlett 1996.

In the event that cost and infrastructure problems could somehow be overcome, patient compliance would continue to pose serious difficulties. Patients undertaking triple-drug therapy must swallow up to 20 pills a day according to a complex schedule related to sleep and meal times. Failure to follow the schedule increases the chance that the virus will become resistant or that the patient will be too sickened by the drugs to continue treatment. Even well-educated patients with good clinical support have difficulty adhering to this demanding regimen; moreover, patients in the early stages of HIV infection are sometimes not willing to take drugs that make them nauseous when they otherwise feel healthy. In clinical trials in industrial countries, for example, as few as 26 percent of patients complied with the instructions (Stewart 1997). Problems with patient compliance are likely to be worse in low-income countries due to lower education levels and the many other problems that poor people in developing countries face.

Even with all these difficulties and uncertainties, many patients in developing countries will ask their physicians for triple-drug therapy, just as patients have attempted to obtain AZT. Governments will in turn face pressure to buy these drugs and to subsidize the necessary clinical services. When very few people have AIDS, total costs will also be small relative to other government expenditures. But as the epidemic progresses, the number of AIDS cases and the cost of the subsidy will escalate rapidly, drawing resources from other pressing social needs. At some point it will become evident that such a subsidy is unaffordable and also unfair to the many people who for a variety of reasons want government help but do not have HIV.

We have seen that medical responses to HIV/AIDS range from a few pennies to thousands of dollars. How much a country actually spends to treat a case of AIDS depends on many factors besides the differing cost of health care inputs. The most important of these is the amount of treatment that the HIV-infected person, his or her family, and any third party payers such as insurance companies or the government are willing and able to buy, and how much the government subsidizes health care and AIDS treatment. Figure 1-8 showed that across countries this amount is strongly correlated with per capita income. An in-depth study of AIDS expenditure in four countries and São Paulo State, Brazil, confirms this general pattern; the average total (public and private) AIDS expenditure varies from 0.6 times per capita GDP in Tanzania to 3.0 times per capita GDP in São Paulo; the average is a ratio of about 1.5 (background paper, Shepard and others 1996).

Where the AIDS epidemic is severe, health policymakers inside and outside government have sought ways to provide compassionate care at low cost. Three alternatives to expensive inpatient care are outpatient AIDS clinics, hospice care (residential low-technology care for the terminally ill), and home-based care.

One innovative program to deliver high-quality treatment of symptoms and opportunistic illnesses without the expense of hospitalization was an outpatient clinic started in 1989 in São Paulo, Brazil. Such clinics are especially well suited to serve urban HIV-positive and AIDS patients who are able to leave their homes. Later in the course of the disease, when the patient is less mobile, the hospice or nursing home provides a lower-cost substitute for inpatient care in a sophisticated referral hospital. However, since such facilities are rarely available in developing countries, the main alternative to the hospital is care at home.

What sort of home-based care is most effective? An analysis of the cost of eight home-based care programs in Zambia found that community-initiated programs were more effective and much less expensive than hospital-initiated programs (Chela and others 1994; Martin, Van Praag, and Msiska 1996). Assuming that the average patient with AIDS would survive six months with either type of care, the benefits of the care must be measured in reductions of hospitalization cost; reduced travel time to the hospital for the patient and the patient’s caretakers; increased patient satisfaction and comfort; and ancillary benefits to the community, such as improved understanding of the ways to prevent AIDS and decreased stigma toward HIV-positive people. Since the study found that patients who received home-based care reduced their hospitalization before death by only two days, the expenditure on the hospital-initiated home-based care programs of about $312 (6 months x 2 visits per month x $26 per visit) was much more than the $14.50 saving in hospital charges (2 days x $7.25 per day). On the other hand, the costs for six months of community-initiated home-based care averaged just $26, less than one-tenth the cost of the hospital-initiated program, and could almost be justified on the basis of reduced hospital use alone.

The tenfold cost difference between hospital- and community-initiated home care programs was due to the much larger expenditure on transport and staff time for the hospital-based programs. For example, on a typical day a team of trained hospital-based nurses could visit only four to eight patients, about a quarter of whom were away from home when the team arrived. As a result the hospital-initiated teams spent on average about two hours on the road in order to spend only fifteen minutes with the patient. In contrast, the community-initiated teams walked only a few minutes and spent an average of two hours with the patient.

If the low cost of the community-initiated home-based care program in Zambia can be generalized to other settings, it is possible that such care would be financed by the patients, their families, and their communities. Indeed, the community-initiated Zambian programs function well because of strong volunteer support from the local communities. Since the benefits of the program include the public ones of improved knowledge about HIV prevention and reduced stigma, there may be a government role in financing such programs, at least until their private benefits to patients’ families are sufficiently well understood for these families and communities to support such programs on their own. Where policies exist to facilitate access to health care for the poor, they should be extended to include community-based home care programs using the same eligibility criteria.