Twenty years into the epidemic, AIDS has become the world's leading infectious cause of adult death and the fourth leading cause of death worldwide. AIDS has killed more than 20 million people since 1981—an estimated 3.5 million in 2001 alone.1 In industrialized countries, mortality has decreased sharply with the advent of antiretroviral (ARV) therapy.2- 4 In the United States, for example, the introduction of highly active antiretroviral therapy (HAART) has decreased AIDS-related morbidity and mortality by up to 90%.5 In some treatment settings, including Brazil, there is also evidence of decreased HIV transmission.6,7 In 2002, however, the vast majority of people living with HIV do not have access to these medications: less than 4% of the 6 million people in the developing world in need of ARV therapy are being treated.8
The availability of ARV therapy in developing countries has become a contentious public health and human rights issue. Many experts have argued that introducing ARV therapy in settings of great poverty should not be a priority, claiming that it is neither cost-effective nor feasible. These conclusions are based not on assessments of established treatment programs, but rather on projected drug costs of a handful of nascent projects.9,10 In fact, very little operational research has been conducted to examine strategies to deliver ARVs in resource-poor nations, despite the fact that these countries have the largest burden of HIV. It is important to note that Brazil and other countries in which ARVs are used widely have robust health infrastructures, whereas many countries in Africa (and in parts of Latin America) have recently seen significant declines in already weak infrastructures.11 The reluctance to treat HIV in the developing world stems, in part, from the argument that the advanced medical technology, such as CD4 cell count and viral load measurements, available in the United States and other industrialized nations, is a prerequisite for a successful HIV treatment program. Although many experts believe this precludes the creation of quantifiably successful programs in resource-poor settings, we have established one in rural Haiti.
Haiti is the poorest country in the Western hemisphere and one of the poorest in the world.12 It is also the most HIV-affected nation in the hemisphere. More than 6% of the adult population is HIV positive13; the prevalence is twice as high in certain poor urban areas.14 More than 15 years ago, Partners In Health, a nonprofit dedicated to the remediation of inequalities in access to care, and its Haitian sister organization, Zanmi Lasante, began a modest AIDS-prevention effort in Haiti's Central Plateau, one of the country's poorest regions. Our initial efforts—from the late 1980s until 1995—were focused largely on prevention. We developed a voluntary counseling and testing (VCT) program, promoted and distributed condoms free of charge, and conducted information and education campaigns in the rural communities in which many Haitians live. Although there is reason to believe these prevention efforts were met with some measure of success—seroprevalence in the prenatal clinic has remained stable among women from the region15 —the general clinic continued to receive a growing number of patients with advanced HIV disease. More than 80% of these patients had lived in urban Haiti, where they had likely been infected.16 Many were returning to their home villages for care—or to die.
The severity of the situation prompted our initiative to integrate community-based prevention and care efforts once effective drugs became available. In 1995-1996, we introduced zidovudine to our prenatal-clinic formulary. Prior to this, a majority of young women who came to this clinic declined free VCT services; however, once this effective method of preventing mother-to-child transmission became available, more than 90% of women offered these services accepted them. In 1998, we launched the "HIV Equity Initiative" in order to complement prevention efforts with ARV treatment for infected individuals who would have died, in our opinion, without these drugs. Because measurement of CD4 cell counts and viral loads is not available in rural Haiti, a clinical algorithm—based on criteria that include the nature and frequency of opportunistic infections, weight, neurologic status, and severe hematologic abnormalities— is used to identify those patients in greatest need. Currently, some 200 of the more than 2100 HIV-positive patients followed in our clinic receive ARV therapy. To ensure adherence, use of ARVs is supervised by community-based health workers, called accompagnateurs, who visit patients daily (this strategy is termed directly observed therapy [DOT]-HAART).17 The program also provides a modicum of financial and social support to ensure adequate nutritional intake for both the patients and their families.
In assessing the efficacy of HIV therapy, measurement of viral loads and CD4 cell counts has become the standard of care in industrialized countries but is not possible in most developing countries. Our experience in Haiti allows us to suggest other indices that may permit a valid assessment of the efficacy of an integrated HIV prevention and care project, or even a nascent national AIDS program in a poor country. First, clinical response to therapy can be measured indirectly by following weight, symptoms, and ability to return to activities of daily living. In rural Haiti, these activities are focused around farming and caring for children. We estimate that more than 90% of DOT-HAART patients are able to resume such activities within 3 months of initiating DOT-HAART.
Second, a reduction in mortality can be measured as long as patients are being monitored daily by accompagnateurs in our project. In rural Haiti, mortality among ambulatory patients who receive DOT-HAART has declined dramatically—since the program began, we have not had a single death among patients receiving ambulatory DOT-HAART as described above. Case fatality rates among HIV patients not receiving antiretroviral therapy remain high, as they do elsewhere where these medications are not available.15
Third, changes in hospitalization rates can also be measured easily. By the early 1990s, more than 25% of all hospitalizations in the Zanmi Lasante clinic were HIV-related. In the past 2 years, this proportion has decreased by almost half. Patients receiving DOT-HAART are rarely hospitalized; most HIV-related hospitalizations are of patients who are not receiving DOT-HAART.15 Encouragingly, in our experience, these clinical indices match criterion standard laboratory findings. We have been able to obtain assays of viral load and CD4 counts at a Boston hospital for certain patients. In a subset of 60 DOT-HAART patients who were doing well clinically, over 80% had undetectable viral loads (less than 400 copies in peripheral blood).18
In addition to these clinical indices, simple serosurveys are already available to many community health centers in resource-poor settings. One way to gauge secular trends in burden of disease is to follow seroprevalence within prenatal clinics; as noted, at Zanmi Lasante it has remained unchanged—between 4% and 5%— over the past 4 years.18
Measuring rates of seroprevalence over time is an indirect way of measuring burden of disease in young adult populations. Introducing ARVs may also enhance prevention efforts in other ways. As we noted, adding zidovudine to the formulary of a prenatal clinic quickly led to an increase in the proportion of women requesting or accepting VCT. Similar trends have been registered in each of the Zanmi Lasante clinics since the introduction of ARVs. The medical staff has also been galvanized by the clinical improvement of their patients: the negative impact on the morale and performance of medical professionals, forced to stand by as patients with advanced HIV disease die, is only now being documented.19
Understanding the lived experience of patients who were near death upon initiating ARV therapy is an equally important means of assessing the impact of an integrated prevention and care project. We close with the case of a patient who was born in a rural village to a large family of peasant farmers, but later moved to Port-au-Prince to continue her primary education; she returned home several years later, HIV-positive and with 2 small children. For almost 10 years, her therapy was limited to treatment of opportunistic infections, which appeared with increasing regularity. By mid-1999 she was too weak to leave her bed, sick with life-threatening chronic enteropathy. In November 1999, she began therapy with zidovudine, lamivudine, and indinavir. Her diarrhea disappeared within 2 weeks, and she gained 26 pounds in the first 5 weeks of treatment. In 2002, she remains asymptomatic and works as an HIV outreach worker.
Like many of our patients in Haiti, she has strong feelings about access to ARVs: "What can I say? The medicines are eloquent enough. What they have done for me is amazing. Everyone was shocked when I went home . . . I was so sick before I started treatment. I was skinny, and the medicines made me big again. I was so weak I could not walk, and now look at me."
Singler J, Farmer P. Treating HIV in Resource-Poor Settings. JAMA. 2002;288(13):1652-1653. doi:10.1001/jama.288.13.1652-JMS1002-6-1