In many resource-limited settings, national antiretroviral (ARV) treatment programs are expanding after the World Health Organization (WHO) “3 × 5” initiative and with increased donor support for treatment programs.1,2 Thailand began a national ARV treatment program in 2000 as the Access to Care program and then expanded this program in 2004 as the National Access to ARVs for People Living with HIV/AIDS (NAPHA) program, with the goal of universal access to ARV treatment.3,4 As part of this HIV treatment program, all public sector regional, provincial, and district hospitals and some private and university hospitals in Thailand now provide ARV treatment to eligible HIV-infected patients.5 Previous literature has described successful implementation of HIV treatment programs in other developing countries, including Zambia, Malawi, and South Africa, with smaller numbers of patients and treatment sites.6-10 This article describes the coverage and outcomes of the national HIV treatment program in Thailand over the period of 2000-2007. To date, this report includes the largest observational treatment cohort of HIV-infected patients in a developing country.
ARV Treatment Program
Before 2000, ARV treatment in Thailand was only available through some research facilities and private hospitals. In 2000, the Thai government formed the Access to Care program, which included multisector cooperation between private and public hospitals and community and government organizations, involvement of persons living with HIV/AIDS, and free or reduced cost ARV drugs. In 2002, treatment access expanded when the Thai Government Pharmaceutical Organization (GPO) began production of a generic fixed-dose combination ARV named GPO-VIR, containing stavudine (d4T), lamivudine, and nevirapine (NVP). In 2004, the Royal Thai Government committed to providing ARV treatment for all eligible patients under NAPHA; and in 2006, ARV treatment was integrated into the universal health care coverage scheme.3
As part of the implementation of this program, the government conducted 2-day provider trainings on HIV medicine and treatment protocols, training more than 8000 physicians, nurses, counselors, laboratory technicians, and pharmacists in 1066 implementing hospitals between 2000 and 2006. Refresher training continues to be offered annually. Nongovernmental organization (NGO) staff and persons living with HIV/AIDS have also been trained to provide nonclinic-based care. Comprehensive care centers in 237 hospitals and health care centers provide a range of services including health education, self-care, home care, and psychosocial support and counseling.
In NAPHA and under the universal health care coverage scheme, patients self-refer or are referred by a provider from any hospital-based or NGO-based voluntary counseling and testing center or from a prevention of mother-to-child transmission site. HIV-infected patients are ARV treatment eligible if they have a CD4 count of 200 cells per cubic millimeter or less or have a symptomatic HIV diagnosis with a CD4 count of 200-250 cells per cubic millimeter. Definitions of asymptomatic HIV, symptomatic HIV, and AIDS in the NAPHA program correspond to the Centers for Disease Control and Prevention (CDC) clinical categories A, B, and C, respectively, with the addition of the following diagnoses which are also included in the Thai national guidelines definition of symptomatic HIV: central nervous system dysfunction; more than 10% weight loss; persistent dermatitis; anemia, lymphopenia, or thrombocytopenia; persistent cough or pneumonia; and persistent lymphadenopathy.11 Treatment failure in the NAPHA program is defined as either 2 consecutive CD4 counts, with a 30% or greater decrease from the treatment high, or a new AIDS diagnosis.
In Thailand, current first-line treatment regimens include GPO-VIR for most patients, with substitutions of zidovudine or efavirenz for patients unable to take d4T or NVP, respectively. Recommended second-line regimens include a ritonavir-boosted protease inhibitor (PI) regimen with indinavir, saquinavir, or lopinavir. Co-trimoxazole prophylaxis is recommended for all patients with CD4 counts less than 200 cells per cubic millimeter, and fluconazole is recommended for all patients with CD4 counts less than 100 cells per cubic millimeter.
After treatment initiation, CD4 monitoring is provided every 6 months. CD4 testing capacity has increased over time with testing only initially available in tertiary care hospitals but later also available in primary care centers. Beginning in 2006, viral load testing has been provided as part of the universal health care coverage scheme. However, due to the limited number of patients in this cohort with results, viral load was not included in the analysis. ARV toxicity monitoring, including alanine aminotransferase and hemoglobin testing, is recommended at routine scheduled visits after initiation of treatment. Chest x-rays are recommended for all patients before initiating ARV therapy. Most hospitals have capacity to diagnose common opportunistic infections, including Mycobacterium tuberculosis, Pneumocystis jiroveci, and AIDS-related skin conditions.
A clinical monitoring system for the national treatment program was implemented in late 2000. Hospitals report monthly on patient demographics, body weight, ARV regimen changes, CD4 counts, ARV drug adherence, opportunistic infections, deaths, and loss to follow-up. Data are collected on individual case report forms and entered into a Microsoft Access-based software program at the hospital site. Data are sent electronically to the Ministry of Public Health for national compilation, cleaning, analysis, and reporting.
All ARV-naive patients who initiated ARV therapy between January 2000 and December 2005 were included in the outcomes analysis, and patient follow-up data were included through March 2007. Data included in this analysis were demographics, baseline and follow-up CD4 counts, clinical stage, type of ARV regimen, and death. Data on opportunistic infections and prophylaxis, new AIDS diagnoses, and adherence were not consistently recorded and thus, were not included in this analysis. Lost to follow-up was defined as patients who were more than 3 months late for a scheduled visit or were coded by the provider as lost to follow-up. All patients identified as lost to follow-up were cross-checked by national identification number against the national death registry, and if found in this registry, patients were recoded as deaths. Immunological outcomes were determined for the subset of patients with available CD4 counts at 6, 12, and 24 months after treatment initiation.
Laboratory testing was performed according to routine procedures at primary or referral hospitals. CD4 testing is done with flow cytometry techniques either with dual platform (Becton-Dickinson FACSScan, FACSCalibur or Beckman Coulter EPICS) or with single platform (Becton-Dickinson FACSCount) methods.
Data Analysis and Statistical Methods
The outcomes analysis included all ARV treatment-naive patients enrolled before January 2006, thus allowing for at least 1 year of follow-up and for 3-6 months of potential reporting delays. Event rates and 95% confidence intervals (CIs) were calculated. Differences in CD4 counts at treatment initiation and at 6, 12, and 24 months were summarized using the median and tested using the paired Wilcoxon signed rank test. CD4 test results were counted as the 6-month, 12-month, or 24-month results if they were within a 3-month window of that follow-up visit, to account for patients who came early or late for a scheduled visit.
The Kaplan-Meier method was used to estimate survival probabilities. Cox proportional hazards models were used to determine hazard ratios for death at any time during the study follow-up period, adjusting for age, sex, hospital type, baseline CD4 count, AIDS clinical stage, and initial treatment regimen; enrollment cohort year was included as a strata in this model. Subsequent changes or discontinuations in regimens after treatment initiation were not included in the model. All patients were right censored at their last visit or in April 2007 if they were still in care at that time.
For the analysis of 12-month survival based on enrollment cohort year, patients with survival longer than 12 months were censored at 12 months after initiating therapy. To determine factors potentially associated with 1-year survival by enrollment cohort, multivariate Cox regression analysis was done. All analyses were done with SAS software version 9.1 (SAS Institute, Cary, NC).
Human Subjects Review
Data were collected as part of the routine ARV treatment monitoring program under the Ministry of Public Health, Thailand. Data elements that would have allowed identification of individuals were removed to create the analysis data set. Human subjects review was conducted by the CDC and determined this analysis to be a research not involving human subjects.
This treatment program was funded by the Royal Thai Government and the Global Fund for AIDS, Tuberculosis, and Malaria. The study design, data collection, analysis, interpretation of data, and writing of the report were jointly contributed by the CDC and Thailand Ministry of Public Health authors.
In the Access to Care program and its successor, NAPHA, 81,960 patients started ARV treatment between January 2000 and March 2007. The number of patients enrolled in NAPHA increased from 4327 in 2002 to 48,101 in 2004 and 81,385 in 2006 (Fig. 1A). By the end of 2006, 839 of 1066 hospitals in the NAPHA program were reporting data from all 76 provinces in Thailand (Fig. 1B).
For the outcomes analysis, 58,008 ARV treatment-naive patients who enrolled before January 2006 were included (Fig. 2). Patients were excluded if they were ARV treatment experienced or if they initiated treatment after December 2005. Among this patient population, 52.2% were male and the median age was 34 years [interquartile range (IQR) = 30-39 years]. The median CD4 count at treatment initiation was 41 cells per cubic millimeter (IQR = 13-113 cells/mm3), and 50.5% had clinical AIDS. The initial regimen for 53,466 of 58,008 patients (92.2%) was NVP and 2 nucleoside reverse transcriptase inhibitors, typically d4T and lamivudine as part of the fixed-dose combination pill (GPO-VIR). An efavirenz-based regimen was initiated at baseline in 3969 of 57,910 patients (6.9%), and a PI regimen was initiated in 395 of 57,910 patients (0.7%). Of the patients initiating a PI regimen, 223 (56%) initiated treatment in 2001, largely as a result of clinical trial participation and because early national treatment guidelines were not specific as to first-line or second-line regimens. Because most patients who initiated a PI regimen did so in 2001, baseline PI regimens were not included in the final Cox model assessing the hazard of death. Two percent of patients had previously received ARV prophylaxis as part of a prevention of mother-to-child transmission program (Table 1).
The median follow-up time was 1.6 years (IQR = 0.8-2.4 years). A majority (80.8%) of patients had more than 6 months of follow-up after the initiation of ARV treatment, and 67.8% of patients had at least 12 months of follow-up. At last follow-up, 43,648 (75.3%) patients were still on ARV treatment, 7637 (13.2%) had died, 1105 (1.9%) had temporarily stopped ARV treatment, 488 (0.8%) had permanently stopped treatment, and 5130 (8.8%) were lost to follow-up. For those patients who were lost to follow-up, the median time to loss was 5.9 months (IQR = 1.8-11.6 months). Among the 4018 patients in the 2000-2002 enrollment cohort, 750 (18.7%) had died and 525 (13.1%) were lost to follow-up.
Of all 7637 deaths, 5616 (85.0%) were known due to AIDS and 1025 (13.4%) had unknown cause of death. The crude AIDS death rate was 5.8 deaths per 100 person-years (95% CI = 5.7 to 6.0). Because the cause of death was not known in some cases where the report was from the national death registry and some deaths may have been misclassified, an alternative analysis in which all deaths were counted as AIDS deaths was conducted, and the crude AIDS death rate in this case was 7.9 deaths per 100 person-years. The median time to death after initiating ARV treatment was 3.2 months (IQR = 1.3-9.5 months). The risk of AIDS death was greatest in the first 6 months of treatment with 4034 (71.8%) of known AIDS deaths occurring during that time. The AIDS death rate in the first 90 days was 3.1 deaths per 100 person-years (95% CI = 3.0 to 3.3) and after 90 days was 2.7 deaths per 100 person-years (95% CI = 2.6 to 2.8), P < 0.001.
Among the 58,008 patients included in the analysis, 11,078 (19.1%) patients ever changed a regimen during the follow-up period. The switching rate was 13.3 per 100 person-years of follow-up (95% CI = 13.1 to 13.6). Among patients who changed a regimen, 10,238 (92.4%) patients switched to a different nonnucleoside reverse transcriptase-based regimen, and 523 (4.8%) changed to a PI regimen. Reasons for ARV changes included adverse events (66.9%), treatment failure (21.3%), tuberculosis treatment (8.1%), drug stock-outs (2.5%), and pregnancy (1.2%). The switching rate for toxicity reasons was 8.5 per 100 person-years of follow-up (95% CI = 8.3 to 8.7) and for treatment failure was 2.5 per 100 person-years (95% CI = 2.4 to 2.6).
CD4 count test results were available for 93.6% of patients at baseline and for 59.7% (27,474), 55.9% (21,620), and 55.2% (11,705) of patients in follow-up at 6, 12, and 24 months after treatment initiation. Median CD4 counts were 41 cells per cubic millimeter at baseline, 175 cells per cubic millimeter at 6 months, 226 cells per cubic millimeter at 12 months, and 312 cells per cubic millimeter at 24 months. Median increases in CD4 counts from baseline were 113 (IQR = 65-177, P < 0.001) cells per cubic millimeter at 6 months, 163 (IQR = 100-244, P < 0.001) cells per cubic millimeter at 12 months, and 246 (IQR = 155-360, P < 0.001) cells per cubic millimeter at 24 months. These results were similar when the analysis was restricted to patients with CD4 results available at all 3 time points (data not shown).
In a Cox proportional hazards model, the adjusted hazard of death was significantly associated with sex, age, weight, clinical HIV stage, baseline CD4 count, and type of hospital (P < 0.001) (Table 2). The Cox multivariate analysis was also conducted as an on-treatment analysis, excluding the 11,078 patients who changed or stopped a regimen, and the results were similar (data not shown).
In a Kaplan-Meier survival analysis, the overall probability of survival after 1 year was 0.89 (95% CI = 0.88 to 0.89) and after 5 years was 0.78 (95% CI = 0.77 to 0.79). For the subset of patients with baseline CD4 counts <50 cells per cubic millimeter, the probability of survival at 5 years was 0.73 (95% CI = 0.72 to 0.75) (Fig. 3A).
For patients enrolled in NAPHA in the 2000-2002, 2003-2004, and 2005 cohorts, the 2005 cohort had a significantly better survival rate 1 year after ARV treatment initiation than the other 2 cohorts [adjusted hazard ratio = 0.76 (95% CI = 0.68 to 0.85); P value for 2005 vs 2000-2002 and 2003-2004 enrollment cohorts <0.001] (Fig. 3B). After adjustment for sex, age, weight, HIV clinical stage, baseline CD4 count, baseline ARV regimen, and type of hospital, later enrollment was still found to be significant.
The national ARV treatment program in Thailand has rapidly scaled up ARV treatment with more than 80,000 patients placed on treatment since the year 2000. With the addition of data not included in this database but reported to the Ministry of Public Health from other public and private hospitals, approximately 115,994 patients had initiated ARV treatment in Thailand as of the end of 2006 (Thailand Ministry of Public Health, unpublished data, November 2007). Importantly, this scale-up has occurred at 1066 public health hospitals and been a multisectoral program at public, private, and university hospitals. This article reports the outcomes of a large national ARV treatment program, one of the largest cohorts of patients on treatment reported from a resource-limited country.12
Outcomes of this treatment cohort in Thailand were very good and comparable to those reported in other articles from resource-limited settings. The 1-year survival probability was 0.89, and 13.2% of patients died. Loss to follow-up in our analysis was 8.8%. These findings compare with observational treatment studies in Zambia, Malawi, and South Africa in which 7.0%, 18.6%, and 13.2% of patients died, respectively.6,9,10 In Malawi, survival probability was 0.81 at 12 months.9 As in our study, factors associated with death in Malawi included male sex and low baseline CD4 and in Zambia, low baseline CD4, body mass index, and WHO clinical stage.6,9 In Zambia, as in Thailand, most deaths occurred in the first 6 months of treatment, but the proportion of deaths in the first 90 days was lower in Thailand.6 In Brazil, in an analysis of all AIDS cases, irrespective of ARV treatment, the period of HIV diagnosis and the AIDS case definition were significantly associated with death.13 Similarly, we found later enrollment year to be significantly associated with a lower hazard of death, suggesting an impact of program changes and increasing provider experience with HIV treatment. However, although the median baseline CD4 count at treatment initiation in African countries was similar to that in Thailand, it was less than that seen in cohorts in the United States and Europe, suggesting an earlier entry to care in resource-rich settings.14,15
Limitations to this analysis include incomplete data on opportunistic infections, prophylaxis, body mass index, and ARV adherence. Furthermore, virologic testing was not routinely available until recently, and the infrastructure for CD4 testing has only gradually become more available in Thailand. Some selection bias may be present due to missing CD4 count results at follow-up visits, although results were similar in a more restricted analysis. In general, observational cohort studies have the limitation of possible bias in different treatment groups, for which statistical adjustments may not be able to fully adjust.16 Nevertheless, these findings demonstrate the outcomes of HIV treatment in a large government-sponsored program at multiple clinical sites and without the rigorous follow-up and incentives provided in a clinical trial.
Findings from this cohort highlight some of the realities of scaling up a large ARV treatment program in a resource-limited setting. For example, the absence of CD4 monitoring for asymptomatic patients early in the program may be a reason for the low median CD4 count at treatment initiation. Furthermore, patients with treatment failure have implications for ARV regimen availability and program costs. However, it is anticipated that the coverage and quality of ARV treatment and care will continue to improve in Thailand as CD4 testing to monitor asymptomatic patients and viral load monitoring to help detect early treatment failure and potential ARV drug resistance become available.17
The success of the Thailand national ARV treatment program is due, in part, to significant political commitment and a strong public health system. Factors cited in the success of the Thailand program include free access to treatment, public sector manufacture of ARVs, and a strong civil society.5 According to a 2005 WHO external review of the Thai national ARV treatment program, procurement and supply management have been well planned without any significant drug shortages, and training has been provided to a large number of health care providers.18 However, continuing efforts are needed to minimize loss to follow-up and to assess the use of and outcomes on second-line treatment regimens.
Finally, it is promising that as the national ARV treatment program has matured, the risk of death has decreased. These ARV treatment outcomes in Thailand are encouraging for other resource-limited countries that are also scaling up HIV treatment programs.
M.S.M. and S.C. led the data analysis and wrote the article; S.C. also contributed to study conception and design. P.Y., T.J., S.T., P.N., C.L., and S.P. contributed to study conception and design and assisted with interpretation of the data and writing of the article. U.S. and P.A.M. did the primary statistical analysis and contributed to the writing of the article. K.F. contributed to interpretation of the data and writing of the article.
The authors would like to express gratitude to Dr. Thawat Suntrajarn, Director General of the Department of Disease Control, for his encouragement and support in scaling up the NAPHA program in Thailand. The authors also thank the following individuals for their contributions to the NAPHA program: Dr. Manit Teeratantikanont (Department of Medical Sciences), Dr. Opart Karnkawinpong; Dr. Nopadol Piboonsin, Dr. Panumard Yarnwaidsakul, and Chertkiat Kleawkasikit. Furthermore, the authors would like to thank the following organizations for their contributions: Ministry of Public Health; Provincial Health Officers; Offices of Disease Prevention and Control 1st to 12th; Thai AIDS Society, The Thai Network of people living with HIV/AIDS (TNP+), Thailand Ministry of Public Health-US CDC Collaboration, and WHO.
1. World Health Organization. Scaling up antiretroviral therapy in resource-limited settings: treatment guidelines for a public health approach. Available at: http://www.who.int/3by5/publications/documents/arv_guidelines/en/
. Accessed April 20, 2007.
2. World Health Organization. The 3 by 5 initiative. Available at: http://www.who.int/3by5/en/
. Accessed April 20, 2007.
3. Chasombat S, Lertpiriyasuwat C, Thanprasertsuk S, et al. The national access to antiretroviral program for PHA (NAPHA) in Thailand. Southeast Asian J Trop Med Public Health
4. Phanuphak P. Antiretroviral treatment in resource-limited settings: what can we learn from the existing programmes in Thailand? AIDS
5. Ford N, Wilson D, Costa Chavez G, et al. Sustaining access to antiretroviral therapy in the less-developed world: lessons from Brazil and Thailand. AIDS
6. Stringer J, Zulu I, Levy J, et al. Rapid scale-up of antiretroviral therapy at primary care sites in Zambia. JAMA
7. El-Sadr W, Elul B, Hawken M, et al. Variability in populations enrolled and their outcomes in HIV care and treatment programs across countries in Sub-Saharan Africa. Paper presented at: 14th Conference on Retroviruses and Opportunistic Infections; February 19-23, 2007; Los Angeles, CA.
8. Van Cutsem G, Hilderbrand K, Coetzee D, et al. Clinical outcomes and emerging challenges after 5 years of ART in a South African township. Paper presented at: 14th Conference on Retroviruses and Opportunistic Infections; February 19-23, 2007; Los Angeles, CA.
9. Ferradini L, Jeannin A, Pinoges L, et al. Scaling up of highly active antiretroviral therapy in a rural district of Malawi: an effectiveness assessment. Lancet
10. Coetzee D, Hilderbrand K, Boulle A, et al. Outcomes after two years of providing antiretroviral treatment in Khayelitsha, South Africa. AIDS
11. CDC. 1993 revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep
12. Teixeira P, Vitoria M, Barcarolo J. Antiretroviral treatment in resource-poor settings: the Brazilian experience. AIDS
13. Kilsztajn S, Lopes E, Santos Nunes do Carmo M, et al. Improvement in survival among symptomatic AIDS patients by exposure category in Sao Paulo. J Acquir Immune Defic Syndr
14. The Antiretroviral Therapy (ART) Cohort Collaboration. HIV treatment response and prognosis in Europe and North America in the first decade of highly active antiretroviral therapy: a collaborative analysis. Lancet
15. ART-LINC. Mortality of HIV-1 infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet
16. Phillips A, Grabar S, Tassie J-M, et al. Use of observational databases to evaluate the effectiveness of antiretroviral therapy for HIV infection: comparison of cohort studies with randomized trials. AIDS
17. Kuritzkes D. Extending antiretroviral therapy to resource-poor settings: implications for drug resistance. AIDS
18. World Health Organization. Scaling up Antiretroviral Treatment: Lessons Learnt From Thailand. Report of an External Evaluation, 12-19 October 2004
. World Health Organization Regional Office for South-East Asia, New Delhi, India. 2007.