Of the estimated 6 million people receiving antiretroviral therapy in 2009, 5.2 million were living in resource-limited settings.1 In resource-rich settings, long-term observational studies have reported increasing life expectancy2 accompanied by a rise in proportion of non–AIDS-related deaths.3,4 It remains debated as to whether these deaths are caused by HIV. In these settings, early response to treatment, measured by CD4 and viral load at 6 months,5,6 or latest updated values of CD4 count, viral load, anemia, and body mass index (BMI)7,8 were strong predictors of long-term mortality.
In resource-limited settings, where patients often initiate highly active antiretroviral therapy (HAART) at younger ages, at more advanced disease stages, with higher proportions of female and heterosexual mode of transmission,8,9 higher rates of early mortality during the first 6 months of therapy have been widely documented.9–16 However, a few studies have sufficient follow-up to assess long-term mortality rates and risk factors.11,17–19 As antiretroviral treatment programs mature, data on long-term mortality trends, causes, and associated factors are needed to inform clinical care and policy and to evaluate program effectiveness. In particular, it is important to assess whether characteristics at the start of HAART remain an important predictor of long-term mortality or if it is outweighed by initial response to therapy.
Thailand was one of the first lower-middle-income countries to launch a national treatment program in 2000.10 By 2009, >210,000 persons received antiretroviral treatment, an estimated 76% coverage of those in need of treatment.20 This study examines the short-term and long-term mortality trends and risk factors in a prospective observational cohort of adults who initiated HAART in a network of 43 public hospitals throughout the country.
Study Design and Population
Adults infected with HIV (age ≥18 years) received antiretroviral therapy in a prospective observational cohort study (Clinicaltrials.gov identifier: NCT00433030). The cohort study began in 1999, recruiting women who participated in trials on prevention of mother-to-child transmission of HIV21,22; from 2003, enrollment was extended to their partners and any adult with HIV infection presenting at participating sites; and from May 2005, a clinical trial comparing HAART monitoring strategies (Clinicaltrials.gov identifier: NCT00162682) was nested in the cohort. Participants provided written informed consent at entry, and the study was approved by the Thai Ministry of Public Health and local ethic committees. The criteria for initiation of therapy were based on national and World Health Organization (WHO) guidelines: Centers for Disease Control and Prevention (CDC) clinical stage B/C or CD4 count <250 cells per cubic millimeter.23,24 Adults, previously antiretroviral naive (except for previous prevention of mother-to-child transmission of HIV prophylaxis), who initiated HAART (defined as ≥3 antiretroviral drugs from at least 2 classes) before November 1, 2009, were included in the analysis.
Treatment and Follow-Up
Initial HAART regimens changed over time with increased availability of drugs: before 2003, protease inhibitor–based regimens were mostly used; then nevirapine-based fixed-dose combinations (with lamivudine and stavudine or zidovudine); and after 2005, tenofovir, emtricitabine, and efavirenz. Alternative drugs were provided in case of intolerance. Therapy was switched for clinical, immunologic, or virological failure by site physicians following Thai and international guidelines or, for patients in the clinical trial, based on viral load or CD4 count as per protocol (Clinicaltrials.gov identifier: NCT00162682). Second line was defined as a change of 1 or more drugs including a change of drug class. Patients received cotrimoxazole if CD4 count <200 cells per cubic millimeter and fluconazole if CD4 count <100 cells per cubic millimeter.
Patients attended the clinic monthly for a basic physical exam, drug refills, and adherence counseling conducted by a nurse. A physician saw them monthly in the first 3 months of treatment and 3 monthly thereafter or when referred by the nurse. Laboratory tests including CD4 count, viral load, and hemoglobin, liver enzymes, and amylase levels were conducted at the start of HAART, at 3 months, and at every 6 months thereafter.
Loss to follow-up (LTFU) was defined as patients who missed at least 1 scheduled visit and had no contact for >6 months. Hospital staff actively traced patients with missed visits through telephone calls and home visits. Patients who informed the staff of leaving the cohort, often because of relocation, were considered as voluntarily withdrawn. Data were collected on specially designed case report forms at site by trained nurses/physicians and sent to the study coordination center for double data entry and management.
Outcomes and Risk Factors
The outcome of interest was all-cause mortality. Deaths were defined as “early” if they occurred ≤6 months after starting HAART or “long-term” if they occurred >6 months and up to 5 years after HAART initiation. The immediate cause and underlying conditions were reported by site physicians. All death reports were reviewed and causes of death classified by 2 independent physicians based on the International Classification of Diseases-10 classification (http://www.who.int/classifications/icd/en/index.html). Deaths from unknown causes, or causes “possibly” or “probably” HIV related, among patients with latest CD4 count <200 cells per cubic millimeter at time of death were conservatively considered as “immunodeficiency related,” together with all AIDS-defining events.
Analysis on risk factors of mortality considered sex, age, CD4 cell count, viral load, CDC stage, BMI, anemia, history of tuberculosis, calendar year, antiretroviral regimen, cotrimoxazole and fluconazole prophylaxis at start of HAART (baseline), hospital size, and type of follow-up (monitoring strategy trial vs observational cohort); and viral load suppression, CD4 count change from baseline, BMI, and anemia at 6 months of therapy.
Age, CD4 count, and viral load at baseline were used as continuous variables after testing for nonlinearity using cubic spline method.25 Calendar year of HAART initiation was categorized guided by percentiles distribution. Anemia was defined as hemoglobin level ≤10 g/dL26,27; BMI was categorized as ≤18.5 or >18.5 kg/cm2.28 Antiretroviral regimens were categorized as nevirapine, efavirenz, or protease inhibitor based. At 6 months, viral suppression was defined as viral load ≤1000 copies per milliliter; CD4 change from baseline was categorized as ≤50 or >50 cells per cubic millimeter increase based on the literature.6 Anemia at 6 months was categorized as follows: persistent (anemic at baseline and at 6 months), new (anemic only at 6 months), recovered (anemic at baseline but resolved at 6 months), and never anemic.
Mortality rates were calculated per 100 person-years (PY). Patients were at risk from start of HAART until date of death, last visit, or at 5 years of treatment, whichever was earliest. Kaplan–Meier probability of survival was estimated up to 5 years of therapy. Cox proportional hazard models were used to assess factors associated with mortality. Because of evidence of interaction between baseline CD4, viral load, anemia, and follow-up time (P < 0.1), early and long-term mortalities were analyzed separately, at before and after 6 months based on the median time to death (Fig. 1).
Covariates with associated P value of ≤0.2 in univariate analysis were included in multivariate analysis, and covariates with a P value of ≤0.2 in multivariate analysis were included in the final model. Factors associated with long-term mortality were tested in a model including variables at baseline and at 6 months of therapy (model 1), and including only baseline risk factors (model 2). The proportional hazard assumption was assessed graphically. All analyses were based on intent-to-continue treatment, ignoring treatment changes, interruptions, or terminations.
To avoid loss of information and potential biased estimates because of missing data (15% of population had 1 or more missing data, all variables had ≤7% missing data) for the Cox univariate and multivariate analyses, we used the MICE (Multivariate Imputation by Chained Equations) method to impute missing values for covariates at baseline and 6 months, based on 20 cycles.29–31 Data were analyzed with STATA version 11 (Stata Corp, College Station, TX).
A total of 2048 patients started treatment in the cohort between 2002 and 2008; 241 were nonnaive at entry and 229 initiated a non-HAART regimen and were excluded, leaving 1578 (77%) included in the analysis. Seventy-four percent were women; at start of HAART, the median [interquartile range (IQR)] age was 33 years (28–38); CD4 count was 124 cells per cubic millimeter (57–196); 632 (40%) were in CDC stage B or C; 263 (17%) had anemia; and 1444 (92%) started a nonnucleoside reverse transcriptase inhibitor–based regimen (Table 1).
Follow-Up and Survival
The median duration of follow-up was 50 months (IQR 41–66), with a total follow-up time of 5935 PY. Overall, there were 89 (5.6%) deaths, 183 (11.6%) LTFU, and 142 patients (9.0%) voluntarily withdrew. The early mortality rate was 4.9 deaths per 100 PY [95% confidence interval (CI): 3.6 to 6.8] ≤6 months of HAART and declined to 1.0/100 PY (95% CI: 0.8 to 1.3) between 6 months and 5 years. Kaplan–Meier survival estimates (95% CI) were 97.6% (96.7% to 98.2%) at 6 months, 96.6% (95.6% to 97.4%) at 1 year, and 93.7% (92.3% to 94.8%) at 5 years of HAART. The probability of being alive and on follow-up was 94.4% (93.1% to 95.4%) at 6 months, 92.0% (90.6% to 93.3%) at 1 year, and 80.8% (78.5% to 82.8%) at 5 years.
Causes of Death
Of the 89 deaths, 37 (42%) occurred ≤6 months and 52 (58%) >6 months of HAART. Forty-six deaths (52%) occurred at the hospital and 43 (48%) at home. The cause of death was assigned in 72 cases (81%); the remaining 17 were unknown causes of which 14 occurred at home. The most common causes were non–AIDS-related infections (23 of 72) and AIDS-defining events (22 of 72). Other causes were as follows: 9 cardiovascular diseases, 6 violent deaths, 6 digestive system and liver diseases, 2 malignancies, and 4 others (Table 2). Using our conservative definition, immunodeficiency-related deaths accounted for 71% of all deaths [92% (34 of 37) of early and 56% (29 of 52) of late deaths].
The median time to early death was 1.9 months (IQR 1.1–3.9). Mortality rates were highest in patients with baseline anemia and CD4 count <50 cells per cubic millimeter (Table 3). In multivariate analyses, after adjusting for sex and age, early mortality was independently associated with anemia [adjusted hazard ratio (aHR) 3.6, 95% CI: 1.7 to 7.5] and low CD4 count (aHR 1.6, 95% CI: 1.1 to 2.2 per 50 cells decrease), whereas high viral load and CDC stage B or C had a weak association.
At 6 months, 1453 patients (92%) remained on follow-up. Median CD4 count at 6 months was 245 cells per cubic millimeter (IQR 168–334); median CD4 change from baseline was 114 cells per cubic millimeter (IQR 62–178); 1316 patients (92%) had a viral suppression ≤1000 copies per milliliter; and 92 (6%) had anemia (Table 1 and Fig. 1).
After 6 months, 52 patients (3.6%) died, with a median time to event of 19.0 months (IQR 10.4–31.4) after HAART initiation. Long-term mortality rates were highest in patients with new or persistent anemia, viral load >1000 copies per milliliter, and CD4 change from baseline ≤50 cells per cubic millimeter at 6 months (Table 4).
In analysis using model 1, including both baseline and time-updated variables at 6 months of therapy, after adjusting for sex, age, baseline viral load, and year of enrollment, factors independently associated with long-term mortality were as follows: persistent anemia (aHR 4.9, 95% CI: 2.1 to 11.6), CD4 count increase ≤50 cells per cubic millimeter (aHR 3.1, 95% CI: 1.6 to 5.7), and viral load >1000 copies per milliliter (aHR 2.8, 95%: CI: 1.3 to 6.1) at 6 months (Table 4). Similar results were obtained when using absolute CD4 cell count ≤100 cells per cubic millimeter at 6 months instead of CD4 change from baseline and when using viral suppression cutoff at 50 copies instead of 1000 copies per milliliter (data not shown).
In analyses restricted to baseline variables only (model 2), after adjusting for age, factors independently associated with long-term mortality were male sex (aHR 2.3, 95% CI: 1.2 to 4.2) and early year of enrollment (aHR 4.8, 95% CI: 2.2 to 10.7 for 2002–2003 vs after 2005), whereas baseline anemia (aHR 1.9, 95% CI: 1.0 to 3.6), low BMI (aHR 1.7, 95% CI: 1.0 to 3.0), and high viral load (aHR 1.4, 95%: CI: 0.9 to 2.2) had a weak association (data not shown).
In sensitivity analyses using the outcome of death and LTFU (failure), the risk factors were comparable, except for younger age at baseline, which became associated (aHR 1.2, 95% CI: 1.0 to 1.4, for early failure; aHR 1.4, 95% CI: 1.3 to 1.6 for long-term failure per 5-year decrease). Participation in a clinical trial was independently associated with a lower risk of long-term failure (data not shown).
In this cohort of antiretroviral therapy–naive patients starting HAART in public hospitals throughout Thailand, the probability of survival was high at 97% at 1 year and 94% at 5 years of treatment. Mortality rates declined from a peak of 4.9 per 100 PY in the first 6 months to 1.0 per 100 PY between 6 months and 5 years of treatment.
This is one of the highest survival estimates to date from a lower-middle-income setting. Previously reported survival at 1 year of HAART ranged from 74% to 92% in sub-Saharan Africa,14 87% to 97% in Latin America,15 and were 89% and 87% in Thailand and Cambodia,10,13 respectively. These higher survival rates are probably because of the comparatively fewer patients with advanced disease: 25% with CDC stage C versus ≥50% in sub-Saharan African12,14,32; median baseline CD4 count 124 versus 41 cells per cubic millimeter in an earlier Thai study.10 Of note, 37% of patients of our cohort were women enrolled after pregnancy, before advanced disease progression, as captured by the higher median CD4 count and better clinical CDC stage than other patients enrolled in the same calendar year. This healthy-women effect disappeared after adjusting on these factors.
There are a few studies in similar settings with sufficient follow-up time to compare survival at 5 years. The Development of Antiretroviral Therapy clinical trial in Uganda, comparing laboratory versus clinical HAART monitoring strategies, reported a similarly high rate of survival of 90% (95% CI: 88% to 91%) at 5 years in the laboratory monitoring arm.33 One treatment cohort study in Senegal enrolling from 1998 to 2002, with patients at more advanced disease stages at initiation as compared with our cohort, reported lower survival of 75% at 5 years (95% CI: 70.6 to 79.6).12
The mortality rate in the first 6 months in our cohort was twice as high as the rates reported in Europe and the United States (2.4/100 PY),9 although the long-term mortality rates were comparable (1.1–2.6/100 PY).5,34 In our cohort, AIDS-related and non–AIDS-related infections were the most common causes of death, and 71% of all deaths were considered as probably or definitely immunodeficiency related. However, over half of the long-term deaths were non-immunodeficiency related: cardiovascular or liver diseases, non-AIDS malignancies, and violent deaths, comparable with reports from European and North American cohorts.3
As observed elsewhere, predictors of early mortality were low CD4 count and anemia at baseline.9,12,18,35 Most early deaths occurred within 2 months of therapy and 92% were immunodeficiency related, highlighting the critical need for HAART initiation at higher CD4 thresholds as per revised WHO guidelines36,37 and for improving diagnosis and treatment of opportunistic infections.16
Predictors of long-term mortality were patients' early treatment outcome: patients with persistent anemia had 6 times higher risk of death as compared with nonanemic patients. Also, patients with low CD4 gain from baseline or unsuppressed viral load at 6 months had about 3 times higher risk of long-term mortality. After adjusting for these factors, baseline characteristics were no longer associated with mortality.
In European and North American cohorts, anemia at 6 months or updated hemoglobin values have been identified as independent prognostic factors for AIDS progression or death.7,8 In resource-limited settings, baseline anemia has been associated with early deaths,14,15,35,38–40 but a few studies have looked at anemia after starting HAART and subsequent deaths. A large study in Tanzania reported an association of hemoglobin decrease at 3 months of HAART with later deaths,39 although the median duration of follow-up was only 12 months. Similarly, in Zambia,41 anemia at 6 months was associated with subsequent mortality.
To our knowledge, this is the first large cohort study with a long enough follow-up period to show the effect of new or persistent anemia on long-term survival. Anemia in HIV-infected patients on HAART could be a sign of advanced disease or coinfections, or be caused by antiretroviral and antimicrobial therapy.42 Studies are needed to further investigate underlying causes of anemia and optimal interventions during the first months of therapy.
Our findings on the association between viral suppression and CD4 increase at 6 months and subsequent mortality are consistent with reports from European and US cohorts.5,6 In resource-limited settings, a few studies have included updated values of viral load and CD4 count: in a study in Senegal, high viral load at 6 months was predictive of 5-year mortality18; in Mozambique, low CD4 count increase at 3 months was associated with 5-year mortality,40 whereas in a cohort in South Africa, time spent within the low CD4 strata was associated with mortality.19 The role of viral suppression at 6 months highlights the importance of adherence to treatment to maximize its long-term benefits. As the lack of CD4 count increase possibly indicates the long-term effect of previous marked immunosuppression, initiation of HAART at higher CD4 level may in part avert this situation.
We examined predictors of long-term mortality using only baseline characteristics to inform areas where laboratory monitoring at 6 months is not routinely available. Baseline anemia, but not CD4 count, was associated with deaths after the first 6 months. Low baseline CD4 count was associated with long-term mortality only for patients with CD4 gain of <50 cells per cubic millimeter at 6 months (data not shown). Other studies have reported an effect of baseline CD4 count on long-term mortality but were often comparing low baseline CD4 counts (<25 cells/mm3) with ≥350 cells per cubic millimeter without considering anemia.43
Long-term mortality was 2 times higher in men as compared with women, after adjusting for other variables. This has been reported elsewhere10,44 and may be because of unmeasured factors, including coinfections and behavior or mode of transmission. As expected, long-term survival for patients enrolled in later years (after 2005) was higher than for the first patients enrolled in 2002–2003, most likely reflecting improvements in availability of treatment regimens and experience of health care providers.
There are several study limitations to consider. First, losses to follow-up may have led to an underestimation of mortality,45,46 although the rate of LTFU was relatively low at 12% at 5 years. Furthermore, patients LTFU would have access to free treatment elsewhere through the national antiretroviral program. At baseline, patients LTFU were younger, mostly women, with less advanced CDC stages, and enrolled in earlier years (data not shown); only calendar year was associated with mortality. In our sensitivity analysis considering LTFU and deaths as failure, risk factors were similar to those of mortality, except for younger age, which may have reflected higher mobility among younger populations. Also, patients enrolled in the monitoring trial were at lower risk of long-term failure, most likely reflecting the rigorous follow-up and commitment to participation in a trial.
Second, adherence was not included among potential risk factors, although viral load at 6 months was considered as a proxy of early adherence. Third, 45% of our cohort was followed-up in a clinical trial. These patients had a higher survival probability, although this was no longer significant after adjusting for calendar year of enrollment. We believe this reflects the wider benefits of accessing improved drugs and higher quality of care in the later years, rather than benefiting from being followed-up in the trial per se.
In conclusion, high long-term survival was achieved in a cohort treated in public hospitals in Thailand. Advanced disease progression at treatment initiation was associated with early mortality, which may be averted with implementation of updated WHO guidelines for earlier initiation of HAART. Resources should be devoted to optimization of early treatment adherence and to management of underlying conditions causing persistent anemia or slow immune recovery, which increase the risk of long-term mortality. Finally, monitoring and care of non–AIDS-related complications in patients surviving the first 6 months of treatment need to be integrated in the scale-up of treatment programs.
The authors thank all the patients and hospital staff at the participating sites.
2. Antiretroviral Therapy
Cohort Collaboration. Life expectancy of individuals on combination antiretroviral therapy
in high-income countries: a collaborative analysis of 14 cohort studies. Lancet. 2008;372:293–299.
3. Antiretroviral Therapy
Cohort Collaboration. Causes of death in HIV
-1-infected patients treated with antiretroviral therapy
, 1996-2006: collaborative analysis of 13 HIV
cohort studies. Clin Infect Dis. 2010;50:1387–1396.
4. Deeks SG, Phillips AN. HIV
infection, antiretroviral treatment, ageing, and non-AIDS related morbidity. BMJ. 2009;338:a3172.
5. Chene G, Sterne JA, May M, et al.. Prognostic importance of initial response in HIV
-1 infected patients starting potent antiretroviral therapy
: analysis of prospective studies. Lancet. 2003;362:679–686.
6. Grabar S, Le Moing V, Goujard C, et al.. Response to highly active antiretroviral therapy
at 6 months and long-term disease progression in HIV
-1 infection. J Acquir Immune Defic Syndr. 2005;39:284–292.
7. Harris RJ, Sterne JA, Abgrall S, et al.. Prognostic importance of anaemia in HIV
type-1-infected patients starting antiretroviral therapy
: collaborative analysis of prospective cohort studies. Antivir Ther. 2008;13:959–967.
8. Mocroft A, Ledergerber B, Zilmer K, et al.. Short-term clinical disease progression in HIV
-1-positive patients taking combination antiretroviral therapy
: the EuroSIDA risk-score. AIDS. 2007;21:1867–1875.
9. Braitstein P, Brinkhof MW, Dabis F, et al.. Mortality of HIV
-1-infected patients in the first year of antiretroviral therapy
: comparison between low-income and high-income countries. Lancet. 2006;367:817–824.
10. Chasombat S, McConnell MS, Siangphoe U, et al.. National expansion of antiretroviral treatment in Thailand
, 2000-2007: program scale-up and patient outcomes. J Acquir Immune Defic Syndr. 2009;50:506–512.
11. Egger M, May M, Chene G, et al.. Prognosis of HIV
-1-infected patients starting highly active antiretroviral therapy
: a collaborative analysis of prospective studies. Lancet. 2002;360:119–129.
12. Etard JF, Ndiaye I, Thierry-Mieg M, et al.. Mortality and causes of death in adults receiving highly active antiretroviral therapy
in Senegal: a 7-year cohort study. AIDS. 2006;20:1181–1189.
13. Ferradini L, Laureillard D, Prak N, et al.. Positive outcomes of HAART
at 24 months in HIV
-infected patients in Cambodia. AIDS. 2007;21:2293–2301.
14. Lawn SD, Harries AD, Anglaret X, et al.. Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa. AIDS. 2008;22:1897–1908.
15. Tuboi SH, Schechter M, McGowan CC, et al.. Mortality during the first year of potent antiretroviral therapy
-1-infected patients in 7 sites throughout Latin America and the Caribbean. J Acquir Immune Defic Syndr. 2009;51:615–623.
16. Lawn SD, Harries AD, Wood R. Strategies to reduce early morbidity and mortality in adults receiving antiretroviral therapy
in resource-limited settings. Curr Opin HIV
17. Bussmann H, Wester CW, Ndwapi N, et al.. Five-year outcomes of initial patients treated in Botswana's National Antiretroviral Treatment Program. AIDS. 2008;22:2303–2311.
18. De Beaudrap P, Etard JF, Ecochard R, et al.. Change over time of mortality predictors after HAART
initiation in a Senegalese cohort. Eur J Epidemiol. 2008;23:227–234.
19. Lawn SD, Little F, Bekker LG, et al.. Changing mortality risk associated with CD4 cell response to antiretroviral therapy
in South Africa. AIDS. 2009;23:335–342.
21. Lallemant M, Jourdain G, Le Coeur S, et al.. A trial of shortened zidovudine regimens to prevent mother-to-child transmission of human immunodeficiency virus type 1. Perinatal HIV
Prevention Trial (Thailand
) Investigators. N Engl J Med. 2000;343:982–991.
22. Lallemant M, Jourdain G, Le Coeur S, et al.. Single-dose perinatal nevirapine plus standard zidovudine to prevent mother-to-child transmission of HIV
-1 in Thailand
. N Engl J Med. 2004;351:217–228.
24. Sungkanuparph S, Anekthananon T, Hiransuthikul N, et al.. Guidelines for antiretroviral therapy
-1 infected adults and adolescents: the recommendations of the Thai AIDS Society (TAS) 2008. J Med Assoc Thai. 2008;91:1925–1935.
25. Hess KR. Assessing time-by-covariate interactions in proportional hazards regression models using cubic spline functions. Stat Med. 1994;13:1045–1062.
27. World Health Organization. Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. Vitamin and Mineral Nutrition Information System. Available at: http://www.who.int/vmnis/indicators/haemoglobin/en/
. Accessed December 21, 2011.
29. Royston P. Multiple imputation of missing values: update of ice. Stata J. 2005;5:527–536.
30. van Buuren S, Boshuizen HC, Knook DL. Multiple imputation of missing blood pressure covariates in survival
analysis. Stat Med. 1999;18:681–694.
31. White IR, Royston P, Wood AM. Multiple imputation using chained equations: issues and guidance for practice. Stat Med. 30:377–399.
32. Lawn SD, Myer L, Harling G, et al.. Determinants of mortality and nondeath losses from an antiretroviral treatment service in South Africa: implications for program evaluation. Clin Infect Dis. 2006;43:770–776.
33. Mugyenyi P, Walker AS, Hakim J, et al.. Routine versus clinically driven laboratory monitoring of HIV antiretroviral therapy
in Africa (DART): a randomised non-inferiority trial. Lancet. 375:123–131.
34. Lohse N, Hansen AB, Pedersen G, et al.. Survival
of persons with and without HIV
infection in Denmark, 1995-2005. Ann Intern Med. 2007;146:87–95.
35. May M, Boulle A, Phiri S, et al.. Prognosis of patients with HIV
-1 infection starting antiretroviral therapy
in sub-Saharan Africa: a collaborative analysis of scale-up programmes. Lancet. 376:449–457.
38. Srasuebkul P, Lim PL, Lee MP, et al.. Short-term clinical disease progression in HIV
-infected patients receiving combination antiretroviral therapy
: results from the TREAT Asia HIV
observational database. Clin Infect Dis. 2009;48:940–950.
39. Liu E, Spiegelman D, Semu H, et al.. Nutritional status and mortality among HIV
-infected patients receiving antiretroviral therapy
in Tanzania. J Infect Dis. 2011;204:282–290.
40. Palombi L, Dorrucci M, Zimba I, et al.. Immunologic response to highly active antiretroviral therapy
and mortality reduction in a cohort of human immunodeficiency virus-positive persons in Mozambique. Am J Trop Med Hyg. 2010;83:1128–1132.
41. Lambada M, Giganti M, Mango A, et al. Hb levels at 6 months and their Association with subsequent mortality among adults on ART in Lusaka, Zambia. Presented at: 18th Conference on Retroviruses and Opportunistic Infections; Poster #557; 2011; Boston, MA.
42. Volberding PA, Levine AM, Dieterich D, et al.. Anemia in HIV
infection: clinical impact and evidence-based management strategies. Clin Infect Dis. 2004;38:1454–1463.
43. Antiretroviral Therapy
Cohort Collaboration. Importance of baseline prognostic factors with increasing time since initiation of highly active antiretroviral therapy
: collaborative analysis of cohorts of HIV
-1-infected patients. J Acquir Immune Defic Syndr. 2007;46:607–615.
44. Maman D, Pujades-Rodriguez M, Subtil F, et al. Gender differences in long-term immune response to ART and mortality: a cohort analysis in 4 sub-Saharan HIV
programs. Presented at: 18th Conference on Retroviruses and Opportunistic Infections; Poster #557; 2011; Boston, MA.
45. Boulle A, Van Cutsem G, Hilderbrand K, et al.. Seven-year experience of a primary care antiretroviral treatment programme in Khayelitsha, South Africa. AIDS. 24:563–572.
46. Brinkhof MW, Dabis F, Myer L, et al.. Early loss of HIV
-infected patients on potent antiretroviral therapy
programmes in lower-income countries. Bull World Health Organ. 2008;86:559–567.
PROGRAM FOR HIV PREVENTION AND TREATMENT (PHPT) NETWORK
PHPT (Thailand)—Site and Principal Investigators (numbers of patients enrolled in each hospital are given in parentheses): Nakornping (138): P. Leenasirimakul; Rayong (129): S. Banchongkit; Prapokklao (116): M. Techapornroong; Samutsakorn (91): A. Chutanunta; Mae Chan (86): S. Buranabanjasatean; Chonburi (74): C. Bowonwatanuwong; Hat Yai (74): A. Nilmanat; Chiangrai Prachanukroh (71): P. Kantipong; Phayao Provincial Hospital (69): G. Halue; Lamphun (55): N. Luekamlung; Sanpatong (48): V. Klinbuayaem; Chacheongsao (46): P. Wittayapraparat; Ratchaburi (43): P. Sang-a-gad; Doi Saket (41): P. Sirijitraporn; Nong Khai (40): N. Yutthakasemsunt; Maharaj Nakornratchasrima (39): R. Lertkoonalak; Buddhachinaraj (38): S. Tansuphasawasdikul; Mahasarakam (38): S. Tongpan; Regional Health Promotion Centre 6, Khon Kaen (32): K. Vivatpatanakul; Bhumibol Adulyadej (27): S. Prommas; Mae on (27): N. Pattanapornpun; San Sai (27): W. Cowatcharagul; Samutprakarn (26): N. Eiamsirikit; Lampang (26): P. Pathipvanich; Nakhonpathom (23): S. Bunjongpak; Mae Sai (19): R. Paramee; Chiang Kham (15): Y. Buranawanitchakorn; Pranangklao (15): S. Pipatnakulchai; Phan (14): S. Jeungphichanwanit; Khon Kaen (13): W. Susaengrat; Kalasin (13): S. Srirojana; Somdej Pranangchao Sirikit (12): T. Hinjiranandana; Somdej Prapinklao (9): P. Maharom; Srinakarin (8): S. Anunnatsiri; Phaholpolphayuhasena (7): P. Chirawatthanaphan; Kranuan Crown Prince (7): A. Rattanaparinya; Roi-et (7): B. Jeerasuwannakul; Nopparat Rajathanee (6): J. Wongchinsri; Banglamung (4): J. Ithisuknanth; Klaeng (2): B. Chetanachan; Health Promotion Region 1 (1): S. Bounyasong; Prajaksilapakom Army Hospital (1): P. Nakchun; and Sankhampang (1): N. Pipustanawong.
PHPT Clinical Trial Unit—Sites monitoring: P. Sukrakanchana, S. Chalermpantmetagul, C. Kanabkaew, R. Peongjakta, J. Chaiwan, S. Thammajitsagul, R. Wongchai, N. Kruenual, N. Krapunpongsakul, W. Pongchaisit, T. Thimakam, R. Kaewsai, J. Wallapachai, J. Thonglo, S. Jinasa, J. Khanmali, P. Chart, B. Ratchanee, J. Chalasin, P. Krueduangkam, P. Thuraset, and W. Khamjakkaew; Laboratory: P. Tungyai, P. Punyati, W. Sripaoraya, W. Pilonpongsathorn, U. Tungchitrapituk, T. Thaiyanant, Y. Taworn, S. Surajinda, P. Mongkolwat, P. Sothanapaisan, J. Kamkon, D. Saeng-ai, A. Khanpanya, N. Boonpluem, A. Thongkum, N. Wangsaeng, A. Kaewbundit, R. Dusadeepong, C. Kasemrat, P. Khantarag, P. Pongpunyayuen, and L. Laomanit; PHPT Data Center: N. Naratee, S. Suekrasae, K. Yoddee, P. Chailert, T. Yaowarat, P. Chusut, R. Jitharidkul, R. Malasam, R. Seubmongkolchai, R. Suaysod, S. Chailert, N. Jaisieng, K. Seubmongkolchai, K. Chaokasem, A. Wongja, B. Amzal, T. Thanyaveeratham, W. Wongwai, J. Inkom, K. Saopang, A. Seubmongkolchai, S. Kreawsa, S. Tanasri, W. Chanthaweethip, P. Pongwaret, N. Homkham, T. Vorapongpisan, and S. Barbier; Administrative Support: N. Chaiboonruang, A. Lautissier, M. Honore, P. Pirom, D. Punyatiam, T. Sriwised, T. Intaboonmar, S. Phromsongsil, S. Jaisook, W. Champa, T. Tankool, S. Nupradit, N. Rawanchaikul, L. Summanuch, and S. Vorayutthanakarn; Tracking and Supplies: K. Than-in-at, M. Inta, and R. Wongsang; Drug Distribution Center: N. Mungkhala, D. Chinwong, and C. Sanjoom.