Treatment outcomes in Latin America, China and Botswana: successes and shortfalls

This article is more than 16 years old. Click here for more recent articles on this topic

Dramatic benefits have come from antiretroviral therapy rollout programmes in resource-limited settings and around three million people are now receiving antiretroviral therapy in low- and middle-income countries. However, such programmes are vulnerable and depend on many factors for their continued success. Significant disparities are being seen in results between and within different geographical areas and populations, the XVII International AIDS Conference heard on Monday.

At Monday morning's plenary session, Alex Coutinho, Executive Director of the Infectious Disease Institute of Kampala, Uganda, presented a high-level summary of the efforts, successes, and shortfalls of African programmes including antiretroviral therapy delivery, services for women and children, counselling and testing. More aggressive testing programmes have allowed many more HIV-positive people to be identified at early stages of disease, particularly in Africa. In many cases this allows people to access antiretroviral therapy before their disease progresses into serious clinical manifestations, allowing for greatly improved health outcomes and survival rates.

Much of the improvement in testing and treatment stems from prevention of mother-to-child transmission (PMTCT) programmes for pregnant women. According to data presented by Coutinho, 33% of HIV-positive pregnant women globally received PMTCT services, including antiretroviral treatment provision, in 2007 – up from 10% in 2004. And the number of children receiving antiretroviral therapy has more than doubled since 2005 – from 75,000 to almost 200,000 in 2007. However, only 12% of pregnant women are assessed for their own treatment needs, beyond the goals of PMTCT.


multivariate analysis

An extension of multivariable analysis that is used to model two or more outcomes at the same time.

adjusted odds ratio (AOR)

Comparing one group with another, expresses differences in the odds of something happening. An odds ratio above 1 means something is more likely to happen in the group of interest; an odds ratio below 1 means it is less likely to happen. Similar to ‘relative risk’. 


Injecting drug user.

middle income countries

The World Bank classifies countries according to their income: low, lower-middle, upper-middle and high. There are around 50 lower-middle income countries (mostly in Africa and Asia) and around 60 upper-middle income countries (in Africa, Eastern Europe, Asia, Latin America and the Caribbean).


Expresses the risk that, during one very short moment in time, a person will experience an event, given that they have not already done so.

At an oral abstract session on Monday morning, researchers presented data on large-scale antiretroviral treatment rollouts in the Caribbean and Latin America, Botswana, and China.

Caribbean and Latin America

Suely Tuboi, on behalf of the Caribbean, Central and South America Network (CCASAnet), reported on mortality rates during the first year of antiretroviral treatment for patients in seven Caribbean and Latin American countries. To date, no multi-cohort study has studied comparative mortality rates and risk factors after antiretroviral treatment initiation in this region. This study analysed the effects of the first year of antiretroviral treatment on mortality rates of adult, antiretroviral-naïve patients (aged 18 years or more) from sites in Argentina, Brazil, Chile, Haiti, Honduras, Mexico and Peru.

The study analysis included 5152 patients who initiated antiretroviral treatment between March 1996 and April 2007: 794 from Argentina, 522 from Brazil, 547 from Chile, 1672 from Haiti, 329 from Honduras, 416 from Mexico and 873 from Peru. The median age was 37 years, and 37% were female. The median baseline CD4 count was 109 cells/mm3 and varied between sites (163, 153, 116, 102, 105, 88 and 79 cells/mm3for Argentina, Brazil, Chile, Haiti, Honduras, Mexico and Peru, respectively). In all countries, the majority of antiretroviral treatment regimens (55% to 96%, depending on the country) were NNRTI-based.

Participants were analysed to the time of death, loss to follow-up, or 365 days since starting antiretroviral treatment. The primary outcome measurement was mortality (from all causes) within the first year of accessing treatment. The one-year mortality rate was 8.2% (95% confidence interval [CI] 7.4-9.1%) for the combined cohort. However, mortality varied widely across sites: from lowest to highest, rates were 2.1% for Argentina; 3.1% for Mexico, 3.6% for Brazil, 5.9% for Chile, 8.8% for Peru, 10.0% for Honduras, and 12.4% for Haiti.

Loss to follow-up rates differed even more widely between the seven countries: from a low of 0.6% in Honduras, to 3.1% in Peru, 3.7% in Chile, 3.8% in Haiti, 5.2% in Brazil, and 17% in Argentina.

Unsurprisingly, people who began treatment at higher CD4 cell counts had the best outcomes. In multivariate analysis adjusting for sex, age, clinical stage, calendar year, and type of regimen, hazard of death was associated with lower CD4 count at antiretroviral treatment initiation, with the poorest outcomes at baseline CD4 cell counts less than 50 cells/mm3. Compared to outcomes for this group, hazard ratios [HR] for one-year mortality: 0.79 (95% CI, 0.68 to 0.93) for CD4 cell counts between 50 and 100 cells/mm3, 0.59 (95% CI, 0.40 to 0.86) between 100 and 200 cells/mm3, and 0.44 (95% CI, 0.22 to 0.86) between 200 and 350 cells/mm3.

By multivariate analysis, an AIDS diagnosis at treatment initiation was also a significant risk (HR 2.93; 95% CI, 2.08 to 4.15), as was older age (HR=1.13 per 10 years; 95%CI, 1.02 to 1.25).

The investigators concluded that "the overall one-year mortality rate and risk factors for death observed in this region were similar to that reported for [other] lower income countries with active follow-up." Possible explanations for the high variability observed across countries included patient-specific factors such as background co-morbid conditions and different stages of disease at antiretroviral treatment initiation because of late presentation. Programme-specific factors were also proposed, including different criteria for treatment eligibility, differing rates of loss to follow-up (which could be a source of bias), and the age of the treatment provision programme itself.


Tendani Gaolathe of the Botswana Harvard Partnership presented findings on the results of six years of antiretroviral treatment in Botswana's public sector. An overall six-year survival rate of 88.6% has been observed among the over 75,000 people currently receiving antiretroviral treatment through Botswana's public health care system.

Botswana has had one of the highest HIV prevalence rates anywhere in the world – at an estimated 300,000 people living with HIV, prevalence is estimated at 17.1%. According to national estimates, roughly 113,000 HIV-positive Botswanians require treatment, out of an overall national population of only 1.7 million.

However, Botswana's national free antiretroviral treatment programme, which began in 2002, has been one of the world’s most successful in terms of outreach, reaching over 80% of those estimated to need treatment. At present, 32 hospitals and 128 satellite clinics offer screening and treatment services. As of December 2007, a total of 92,932 patients were on antiretroviral treatment in Botswana. (The projected number for the end of 2009 is 125,000.)

The great majority of these patients receive care through the public sector. However, those on antiretroviral treatment include 9514 private sector patients and a further 8336 patients who were out-sourced from the overloaded public sector to the private sector. (Such patients continue to receive free treatment under the public programme, but the delivery is implemented through private-sector health care sites.) These patients were not included in the reported analysis.

Earlier reports, such as the presentation at the Sixteenth International AIDS Conference in Toronto in 2006, have shown significant declines in mortality among Botswana's HIV-positive adults. In today's session, the overall six-year survival rate of the 75,082 public-sector antiretroviral treatment recipients was reported to be 88.6% (95% CI, 88.1% to 89.2%).

Of the 75,082 patients on antiretroviral treatment, 61% were female; 9.5% were children fifteen years of age or under, and 10.9% were adults 50 years of age or older. Just over a third (35.4%) of the adults initiated treatment with CD4 cell counts of 100 cells/mm3 or less; the great majority (all but 6.7%) had counts of 200 cells/mm3 or less.

The survival rate for women (90.8%; 95% CI, 90.3% to 91.3%) was better than that for men (85.1%; 95% CI, 84.0% vs. 86.2%). Unsurprisingly, early initiation of antiretroviral therapy was associated with improved treatment success, with six-year survival rates of 85.9%, 92.7%, and 83.9% for those who began treatment with CD4 cell counts above 200 cells/mm3, between 100 and 200 cells/mm3, and below 100 cells/mm3, respectively. The poorer outcome for those with the highest counts was not analysed, but Gaolathe suggested that it may have been because people with clinical disease would be most likely to start antiretroviral treatment at these CD4 cell levels.

Nearly half of the deaths occurred within the first three months after treatment initiation. The six-year survival rate for all patients who were still alive at three months was 94.3%. As Gaolathe stated, "if we can keep them alive for three months, we can probably keep them alive for years."

The researchers stated that "unwavering political commitment, translated into resources and collaboration of development partners, has enabled the successful implementation of Botswana's HAART program. The roll-out of treatment facilities to districts, training and authorization of nurses to dispense treatment to stable patients, and social mobilization have enabled high enrolment rates despite constraints on resources." They continued, "Botswana needs to ensure the sustainability of its treatment program, including its integration into the national response and health system. ART services should be further decentralized, and collaboration with the private sector and civil society further strengthened."


A presentation by Ye Ma of the Chinese Center for Disease Control and Prevention on the outcomes of adults receiving antiretroviral treatment in China focused on virologic outcomes.

Within the past five years, free antiretroviral therapy access has rapidly expanded in China. As of June 2007, a total of 35,557 patients nationwide have received antiretroviral treatment, with 29,184 (82.1%) remaining in treatment to the present. Resource limitations have prevented China from providing viral load testing for all people receiving treatment. However, a cross-sectional analysis of virologic response has been conducted to evaluate the virologic impact of the antiretroviral treatment programme, and to assess future requirements. Results of this analysis were presented at this session.

This was a cross-sectional survey in which 24 counties were selected from eight Chinese provinces: six in which HIV transmission was primarily through former blood plasma donation (FPD), and two in which injecting drug use (IDU) and sexual transmission were the primary sources of infection. (FPD accounted for half of the recipients overall.) Antiretroviral treatment recipients were stratified according to the length of time they had been receiving treatment – six to eleven months, twelve to twenty-three months, or 24 months or more. Study participants were then randomly sampled from each of these treatment-duration categories. Viral load tests were completed for each patient sampled; undetectable viral load was defined as less than 400 copies/ml.

Of the 5254 people who had been enrolled on antiretroviral treatment in these 24 counties, 4673 adults had been on treatment for over six months, and 1161 were selected as the final study sample. Mean age was 42.2 years, and 55.7% were male. Infection routes were: 55.6% FPD, 24.6% sexual contact, 19.9% IDU/other.

Rates of treatment response increased with the length of time on therapy. Of patients treated for six to eleven months, twelve to twenty-three months, or 24 months or more (N=218, 416, and 527, respectively), the proportions achieving viral load levels

No differences were found within duration groups according to age or gender. Other than treatment duration, the two other significant factors in treatment success were choice of initial antiretroviral regimen, and type of clinical site. The Chinese state-recommended first-line regimen was d4T or AZT, plus 3TC, plus nevirapine or efavirenz. Early on in the programme, however, limited supply of 3TC forced the alternate use of ddI. For people taking ddI-containing or otherwise nonstandard regimens, the adjusted odds ratio of virologic failure was 5.20 at six to eleven months (95% CI, 2.34 to 11.52), 4.25 at twelve to twenty-three months (95% CI, 2.38 to 7.59), and 2.46 at 24 months or more (95% CI, 1.29 to 4.67).

The site of care provision also had an enormous impact on outcomes. People receiving care at the level of a county hospital or larger site had by far the lowest failure rates. Failure rates after 24 months of treatment were significantly higher in smaller township hospitals (adjusted odds ratio [AOR], 8.15; 95% CI, 1.02 to 65.1) and highest of all in rural village clinics (AOR, 13.54, 95% CI, 1.64 to 111.92). This has major implications for the Chinese treatment model, which is based on community-level care. Presenter Ma described a critical shortage of experienced healthcare personnel at the rural level, emphasising the need for capacity-building initiatives to improve quality of care.


Tuboi S et al. Mortality during the first year of potent antiretroviral therapy in HIV-1-infected patients from 5 treatment centers in the Caribbean and Latin America. Seventeenth International AIDS Conference, Mexico City, abstract MOAB0203, 2008.

Puvimanasinghe JPA et al. Six years of HAART in the public sector of Botswana. Seventeenth International AIDS Conference, Mexico City, abstract MOAB0204, 2008.

Ma Y et al. Virologic outcomes of the adult AIDS patients receiving highly active antiretroviral therapy in China. Seventeenth International AIDS Conference, Mexico City, abstract MOAB0205, 2008.