Sage Journals: Discover world-class research

Abstract

Objectives:

Monitoring antiretroviral treatment (ART) outcomes is essential for assessing the success of HIV care and treatment programs in resource-limited settings (RLS).

Methods:

Longitudinal analyses of clinical and immunologic parameters in HIV-infected adults initiated on ART between November 2004 and June 2008 at Management and Development for Health (MDH)-Presidents Emergency Plan For AIDS Relief PEPFAR supported HIV care and treatment clinics in Tanzania.

Results:

A total of 12 842 patients were analyzed (65.9% female, median baseline CD4 count, 106 cells/mm³). Significant improvements in immunologic status were observed with an increase in CD4 count to 298 (interquartile range [IQR] 199-416), 372 (256-490) and 427 (314-580) cells/mm³, at 1, 2, and 3 years, respectively. Overall mortality was 13.1% (1682 of 12 842). Male sex, World Health Organization (WHO) stage III/IV, CD4 <200 cells/mm³, hemoglobin (Hgb) <8.5 g/dL, and stavudine (d4T)-containing regimens were independently associated with early and overall mortality.

Conclusions:

Closer monitoring of males and patients with advanced HIV disease following ART initiation may improve clinical and immunologic outcomes in these individuals.

Keywords

Tanzania antiretroviral treatment outcomes

Introduction

In 2008, an estimated 33.4 million people were living with HIV. Sub-Saharan Africa remains the most heavily affected region, accounting for over two thirds of all new infections worldwide. In recent years, unprecedented efforts have been made by national governments and international organizations to scale-up HIV care and treatment services in low- and middle-income countries. By the end of 2008, as a result of these efforts, at least 4 million HIV-infected individuals in sub-Saharan Africa were receiving antiretroviral therapy (ART), an estimated 44% of those who are eligible.¹

Since ART rollout began, several studies examining ART outcomes in resource-limited settings (RLS) have been performed.² The majority have reported encouraging clinical, immunological, and virologic responses to therapy with sustained benefit over time. Mortality rates, the most important measure of ART effectiveness, have been estimated to range from 8% to 26% at 1 year after ART initiation in RLS, with the highest rates occurring among those with more advanced HIV at treatment initiation and shortly after ART initiation.³ Recently, high rates of mortality have been reported among patients lost to follow-up,^4,5 suggesting that mortality in programs without active patient tracking systems may be significantly underreported. Findings such as these have important implications for public health policy since determination of a program’s effectiveness is often based on survival.

In 2007, about 2 million persons were estimated to be living with HIV and AIDS in Tanzania, with approximately 600 000 persons with AIDS in need of ART.⁶ Over the last 5 years, a nationwide HIV care and treatment program aimed at providing HIV care and treatment services for all HIV-infected persons was implemented with support from international development partners, including Management and Development for Health (MDH)-Presidents Emergency Plan For AIDS Relief (PEPFAR). To date, very few studies have evaluated the treatment outcomes among HIV-infected individuals initiated on ART in Tanzania.⁷ In this article, we report mortality and immunological outcomes in response to ART and examine predictors of overall and early mortality among HIV-infected patients enrolled in a large urban HIV care and treatment program in Tanzania that provides ART within all levels of public health delivery system.

Methods

This prospective observational study was conducted at 12 MDH-PEPFAR supported HIV care and treatment clinics in Dar es Salaam, Tanzania, between November 2004 and June 2008. The MDH-PEPFAR HIV Care and Treatment Program (previously known as the Muhimbili University of Health and Allied Sciences-Dar City-Harvard School of Public Health-PEPFAR HIV Care and Treatment program) was established in 2004 and provides infrastructure, laboratory, and technical support to the 3 municipalities of Dar es Salaam (Temeke, Ilala, and Kinondoni). Inclusion criteria for this study were age >15 years, not yet initiated on ART, and not pregnant at the time of enrollment. At the MDH program, patients are evaluated monthly following enrollment and ART initiation. At each visit, they are examined by a physician and receive adherence and nutrition counseling and ART refills. Monitoring laboratory tests including hemoglobin (Hgb) and alanine aminotransferase (ALT) are performed 2 weeks after ART initiation and every 4 months thereafter. Immunonologic monitoring with CD4 counts is also performed at 4-month intervals. Viral load testing is not performed routinely. Pulmonary tuberculosis (PTB) screening with a chest X-ray and sputum smear for acid fast bacilli is performed at baseline on all patients and repeated in patients who have symptoms suggestive of PTB at follow-up visits. A comprehensive patient tracking system is in place to ensure all patients who are eligible for tracking (ie, missed visit, lost to follow-up, reported to be sick, and abnormal laboratory results) are contacted by phone or in person to encourage them to return to clinic. A network of community-based health workers and volunteer persons living with HIV and AIDS work with the MDH program to actively trace and ascertain the vital status of all patients who are lost to follow-up.

Patients are initiated on ART according to the National AIDS Control Program (NACP) ART initiation criteria (CD4 count <200 cells/mm³, clinical World Health Organization (WHO) stage IV and clinical WHO stage III with CD4 count of <350 cells/mm³).⁶ At the time of this study, standard first-line ART regimens included stavudine (d4T) or zidovudine (ZDV) plus lamivudine (3TC) and efavirenz (EFV) or nevirapine (NVP). To avoid drug interactions, EFV is substituted for NVP in patients receiving rifampin-containing antituberculous therapy. Recommended second-line regimens for patients failing to respond to first-line regimen include the combination of abacavir (ABC) + 3TC and lopinavir/ritonavir (LPV/r) or saquinavir/ritonavir (SQV/r). Antiretroviral drugs and co-trimoxazole are provided free of charge by the Tanzanian government; and since the inception of the MDH program, no patients have experienced treatment interruptions owing to drug shortages.

Patients were recruited for participation and enrolled in the MDH HIV Care and Treatment program following written informed consent, which was subject to ethical review by the Muhimbili University of Health and Allied Sciences and the Harvard School of Public Health.

Data Collection and Management

Patient demographic, clinical, laboratory, and therapeutic data are collected by physicians and nurses on standard case report forms and National Care and Treatment Center forms (CTC 2) which are completed at enrollment and at each follow-up visit. Data reviewers are stationed at each clinic to ensure adequacy and completeness of data recording by the health care workers. Data collected is then entered into a secure computerized database designed solely for the purpose of data collection and analysis. Unique patient identifiers are used. The database is updated daily by professional data entry clerks. Weekly quality assurance checks of the database are performed by the data management team to ensure data accuracy. Data collected for these analyses included baseline demographics, age, weight, height, body mass index (BMI), median upper arm circumference (MUAC), WHO clinical stage, history of prior or current PTB (the latter defined as receiving treatment for PTB within 1 month after ART initiation), ART regimen at initiation, enrollment and visit dates, date of death, and Hgb and CD4 count.

Treatment Outcomes

Primary outcomes of interest included death occurring in the first 90 days after ART initiation (early mortality), overall mortality and immunologic response to ART. Deaths were recorded after notification by family members, friends, patient tracking teams, or community-based health care workers. If the date of death was unknown but a patient was known to have died, the date of the last encounter with the patient was used as the date of death. Immunologic outcomes observed were the mean change in CD4 count over time and the proportion of patients with immunologic failure as defined by the following Tanzanian national criteria: (a) CD4 count less than pretherapy CD4 count after at least 6 months of therapy or, (b)³ 50% drop from the peak CD4 count value.

Statistical Methods

Proportions or median (interquartile range [IQR]) were calculated for baseline characteristics at the time of ART initiation. In this study, deaths were followed from date of ART initiation to date of death or date of last visit, whichever happened first. The assessment of immunological failure was also determined from the date of ART initiation to the date of first failure or date of last visit, whatever happened first. Kaplan-Meier methods were used to calculate the cumulative incidence of outcomes. To identify predictors of mortality, univariate and multivariate analyses were performed using Cox regression analyses. P values for the trend test were calculated using median scores test for ordinal variables and the likelihood ratio test was used to assess the statistical significance of nominal polytomous variables. The missing indicator method was used as needed.⁸ Generalized estimating equations (GEEs) were used to assess the change in CD4 count. Stepwise restricted cubic splines were used to assess nonlinearity of results and produce graphs of associations. The criterion for significance for all analyses was a P value of α = .05. All statistical analyses were performed with the statistical software package SAS, Release 9.1 (Cary, North Carolina).

Results

A total of 12 842 patients were included in this analysis. Baseline characteristics are shown in Table 1. The majority of patients were female (65.9%) and the median age was 36 (IQR 31-43) years. In all, 6855 (78.9%) patients had a baseline CD4 <200 cells/mm³ and 2619 (24.6%) of patients were WHO stage IV. The median BMI was 20.0 (IQR 17.7-22.7); 4052 (33.4%) adults patients had a BMI < 18.5. One third (31.8%) of the patients reported a past history of PTB and 13.4% were currently infected with PTB. All the patients were initiated on a nonnucleoside reverse transcriptase inhibitor (NNRTI)-containing ART regimen; 7451 (79.7%) patients were on regimens containing d4T.

Table 1

Basic Characteristics of Study Population (N = 12 842)

Characteristics	Median (IQR) or n (%)
Sociodemographics
Gender, female	4383 (65.9%)
Age (years), median (IQR)	36 (31-43)
MUAC, median (IQR)	24.0 (22.0-27.0)
BMI, kg/m², median (IQR)	20.0 (17.7-22.7)
BMI categories
<18.5	4052 (33.4%)
18.5-24.9	6425 (52.9%)
24.9-29.9	1252 (10.3%)
≥30	410 (3.4%)
Clinical
Days from MDH enrollment to ART initiation, median (IQR)	28 (18-42)
WHO stage
I	667 (6.3%)
II	1551 (14.6%)
III	5814 (54.6%)
IV	2619 (24.6%)
On anti-TB treatment within 1 month after ART initiation	1681 (13.4%)
Past TB history	4002 (31.8%)
ART regimen
d4T, 3TC, and NVP	5699 (61.0%)
d4T, 3TC, and EFV	1752 (18.7%)
ZDV, 3TC, and NVP	671 (7.2%)
ZDV, 3TC, and EFV	1225 (13.1%)
NVP-containing regimen	6370 (68.2%)
d4T-containing regimen	7451 (79.7%)
Months from ART initiation to last visit, median (IQR)	6.0 (1.9-12.5)
Laboratory
Baseline CD4 count, cells/mm³, median (IQR)	106 (40-187)
Baseline CD4 count <200 cells/mm³	6885 (78.9%)
Hemoglobin, g/dL, median(IQR)	10.1 (8.5-11.5)
Hemoglobin <8.5 g/dL	2560 (24.3%)

Abbreviations: ART, antiretroviral therapy; IQR, interquartile range; d4T, stavudine; 3TC, lamivudine; NVP, nevirapine; EFV, efavirenz; ZDV, zidovudine; ALT, alanine aminotransferase.

Significant improvements in immunologic outcomes were observed in patients on ART, with a median ART duration of 6.0 months (1.9-2.5). There was an increase in CD4 count throughout the 3 years following ART initiation, with no evidence of nonlinearity (P = .08; Figure 1). The median CD4 count 1, 2, and 3 years after ART initiation was 298 (199-416), 372 (256-490), and 427 (314-580) cells/mm³, respectively. The cumulative incidence of immunological failure defined as a CD4 count less than pretherapy CD4 count after at least 6 months on therapy at 1, 2, and 3 years was 10.5%, 17.5%, and 20.8%, respectively. When defined as a 50% drop from the peak CD4 count value, the cumulative incidence of immunologic failure at 1, 2, and 3 years was 5.7%, 16.8%, and 30.8%, respectively (Table 2).

Figure 1.

CD4 count by time in months since ART initiation. Overall slope 2.1 cells/mm³ per month (standard error = 0.24, P value <.001). ART indicates antiretroviral therapy.

Table 2

Cumulative Incidence of Mortality and Immunologic Failure

		Cumulative Incidence % (95% CI)
	Number at Risk	1 Year	2 Years	3 Years
Mortality^a	12 830	15.7 (15.3-16.1)	18.1 (17.6-18.6)	19.3 (18.7-19.9)
Immunological failure by different definitions
CD4 count less than pretherapy CD4 baseline after at least 6 months of initiating therapy^b	3109	10.5 (9.8-11.2)	17.5 (16.5-18.5)	20.8 (19.1-22.4)
50% drop from the peak CD4 count value^c	4820	5.7 (5.3-6.1)	16.8 (15.9-17.7)	30.8 (28.7-32.9)

Abbreviation: CI, confidence interval.

^aCumulative incidence of mortality at 90 days is 11.3% (95% CI 11.0%-11.6%).

^bDefined for patients with a baseline CD4 count and have been followed for at least 6 months from ART initiation.

^cDefined for patients with at least 2 CD4 counts taken after ART initiation. The maximum is the highest peak reached. Patients should have had a 50% drop from maximum peak at least 2 times.

The overall crude mortality rate was 13.1% (1682 of 12 842). The majority of deaths (76.4%) occurred in the first 3 months. The cumulative incidence of death at 1, 2, and 3 years was 15.7%, 18.1%, and 19.3%, respectively (Table 2). In multivariate analyses, predictors of early and overall mortality were very similar (Table 3). Males had a significantly higher risk of both early (hazard ratio [HR] 1.19; 95% confidence interval [CI] 1.06-1.34; P < .001) and overall (HR 1.19; 95% CI 1.07-1.32; P <.001) mortality. Patients with clinical and immunological parameters associated with advanced HIV at the time of enrollment were also at higher risk of early and overall mortality with a higher WHO clinical stage being the strongest predictor; CD4 <200 cells/mm³: (HR 1.62, 95% CI 1.33-1.97 [early mortality]; HR 1.42, 95% CI 1.20-1.67 [overall mortality]); Hgb <8.5g/dL: (HR 1.94, 95% CI 1.71-2.20 [early mortality]; HR 1.87, 95% CI 1.67-2.08 [overall mortality]); and WHO stage IV: (HR 4.99, 95% CI 3.81-6.53 [early mortality]; HR 4.61, 95% CI 3.66-5.82 [overall mortality]). The cumulative risk of mortality by CD4 count strata (<200 vs >200 cells/mm³) is shown in Figure 2. Patients with a BMI > 18.5 were also at significantly lower risk of mortality, although the relative risk of mortality was noted to increase again slightly at the upper extremes of BMI (Figure 3). Patients with a current or prior history of PTB had a significantly lower risk of both early and overall mortality. Patients initiated on a d4T-containing regimen were at higher risk of early and overall mortality compared to those on ZDV)-containing regimens (HR 1.41, 95% CI 1.15-1.73 [early mortality] and HR 1.43, 95% CI 1.20-1.70 [overall mortality]), respectively.

Figure 2.

Cumulative risk of mortality within strata of baseline CD4 count.

Figure 3.

Body mass index in relation to mortality risk.

Table 3

Results of Multiple Cox Regressions for Time to Overall and Early Mortality

	Early Mortality (<90 days)		Overall Mortality
	Hazard Ratio (95% CI)	P Value^a	Hazard Ratio (95% CI)	P Value^a
Sociodemographics
Gender	Reference		Reference
Female	1.19 (1.06-1.34)	<.001	1.19 (1.07-1.32)	<.001
Male
Age		.42^b		.28^b
30-39	Reference		Reference
15-19	1.04 (0.60-1.80)		1.15 (0.72-1.84)
20-29	0.91 (0.77-1.07)		0.92 (0.80-1.07)
40-49	0.97 (0.84-1.11)		0.98 (0.87-1.11)
50+	0.99 (0.82-1.18)		1.03 (0.88-1.20)
BMI		<.001^b		<.001^b
<18.5	Reference		Reference
18.5-24.9	0.55 (0.48-0.64)		0.60 (0.54-0.68)
25-29.9	0.42 (0.30-0.59)		0.47 (0.36-0.61)
30+	0.60 (0.37-0.97)		0.59 (0.40-0.89)
Clinical
WHO stage		<.001^c		<.001^c
I/II	Reference		Reference
III	2.29 (1.75-2.98)		2.26 (1.80-2.83)
IV	4.99 (3.81-6.53)		4.61 (3.66-5.82)
On anti-TB treatment
No	Reference	.011	Reference	.022
Yes	0.77 (0.62-0.94)		0.81 (0.68-0.97)
Past TB history
No	Reference	<.001	Reference	<.001
Yes	0.76 (0.66-0.87)		0.79 (0.70-0.89)
ART regimen
ZDV	Reference	<.001	Reference	<.001
d4T	1.41 (1.15-1.73)		1.43 (1.20-1.70)
Laboratory
CD4 count, cells/mm³
≥200	Reference	<.001	Reference	<.001
<200	1.62 (1.33-1.97)		1.42 (1.20-1.67)
Hemoglobin, g/dL
≥8.5	Reference	<.001	Reference	<.001
<8.5	1.94 (1.71-2.20)		1.87 (1.67-2.08)

Abbreviations: ART, antiretroviral therapy; BMI, body mass index; CI, confidence interval; d4T, stavudine; 3TC, lamivudine; NVP, nevirapine; EFV, efavirenz; ZDV, zidovudine; ALT, alanine aminotransferase; TB, tuberculosis; WHO, World Health Organization.

^aHazard ratios, 95% CI, and corresponding P were obtained from Cox proportional hazards models. Missing indicator method was used for missing variables. Only variables significant at P < .05 were included in the multivariate model. The multivariate model was adjusted for patient referral site, calendar month, and calendar year as indicators in addition to variables in the table.

^b P value for trend calculated using median score test.

^c P value for trend calculated using likelihood ratio test.

Discussion

In this study, we report ART outcomes from one of the largest HIV care and treatment programs in sub-Saharan Africa. Notable findings include robust improvements in CD4 count over time, and a relatively low risk of mortality, despite the advanced stage of HIV disease at ART initiation in most of our patients. Our findings provide yet more evidence that it is feasible to provide effective ART within all levels of public health delivery system in RLS.

Baseline demographics of our cohort were similar to other HIV cohorts in RLS; the majority of patients were female and most patients initiated ART at very advanced stage of HIV disease. Several clinical and laboratory markers associated with advanced HIV disease were observed in high proportions among study patients at enrollment, including a low CD4 count, BMI, Hgb, and higher WHO clinical stage. Advanced HIV disease at enrollment has been reported frequently among other HIV cohorts in Africa^9,10; a previous study from Tanzania reported up to 30% of patients starting ART with CD4 counts below 50 cells/mm³.¹¹ Factors that have been found to be associated with late disease presentation include lower educational level, male gender, stigma, inadequate referral systems, and poor access to treatment centers.¹² How to best promote early uptake of ART-eligible patients to treatment programs is a central challenge facing the scale-up of HIV care and treatment programs in RLS and a focus on reducing factors associated with late treatment access is therefore necessary.

According to Tanzania National guidelines, immunological failure is defined as a 50% drop from peak value in CD4 count or return to pre-ART baseline CD4 count or lower.⁶ In this study, a number of patients failed ART therapy according to these criteria. Of concern, an estimated one third of all patients met the definition of 1 of the criteria at 3 years. In prior studies where similar definitions of immunologic failure have been used, the rates of failure have varied. In a study from Ghana¹³ which used 3 criteria (2 of which are similar to the Tanzanian criteria), immunologic failure rates were as low as 12%. The insensitivity of immunologic failure criteria when compared to virologic failure criteria has been well described and may result in delays in antiretroviral switching as well as unnecessary treatment switches, both of which have been shown to negatively impact treatment outcomes.^14,15 It is therefore possible that, some of our patients may not have actually failed therapy using these criteria alone. On the other hand, these findings should not be completely dismissed; in a recent randomized trial of monitoring strategies (clinical findings, CD4 count, viral load) by Coutinho et al in Uganda, the addition of viral load measurements did not add any benefit over the use of the CD4 count plus clinical staging in predicting the risk of death.¹⁶

Mortality rates in our study were slightly lower than the rates in some other published studies from RLS.^3,17–19 We observed that the majority (1285 of 1,682, 76.4%) of deaths occurred within 90 days, a similar finding to other studies.^20,21 In a recent review by Lawn et al³ of 18 published HIV cohort studies from RLS, between 8% and 26% of patients who were not lost to follow-up died after 12 months, with the greatest burden of mortality occurring during the initial months of therapy. In this study, many of the same predictors of mortality were observed, especially those indicative of an advanced stage of disease at ART initiation including low CD4 counts and Hgb and high WHO clinical stage. The striking difference in mortality in patients with CD4 counts <200 cells/mm³ compared CD4 counts above 200 cells/mm³ has also been previously observed.

Males were also found to have a significantly higher risk of early and overall mortality in this study even after controlling for baseline clinical and immunologic status. A higher risk of mortality among men has been increasingly reported in other studies in RLS, although some of the data are conflicting.^22,23 Behavioral factors are likely to have contributed significantly to these outcomes. In recent studies, males have been noted to have higher rates of loss to follow-up and nonadherence than women²⁴ and report more HIV-associated stigma.²⁵

A notable finding in our study was the protective effect of current or prior history of PTB at ART initiation on mortality. This finding has also been reported in one other study and warrants further investigation.²⁶ The incidence of PTB in our patient cohort was much lower than expected, and although screening with chest X-rays (CXR) and sputum are performed routinely on every patient at the time of program enrollment, the low sensitivity and specificity of this type of screening is well known.^27,28 Potentially, more patients had undiagnosed PTB at ART initiation and were at higher risk of adverse outcomes than those who were actually diagnosed, perhaps because they were monitored less closely for other comorbidities such as the immune reconstitution inflammatory syndrome (IRIS) or were less likely to be receiving prophylactic co-trimoxazole.

Another unique finding in our study was the higher risk of overall and early mortality observed among patients on regimens containing d4T. An association between d4T-containing regimens and mortality was recently observed in a collaborative analysis of 12 prospective HIV outcome cohort studies from North America and Europe,²⁹ however there was no association of d4T and mortality in 2 other studies from RLS.^9,21 A further investigation into this relationship is needed, especially as d4T continues to be widely used in these settings. Stavudine has been found to be associated with numerous and potentially life-threatening side effects which could have contributed significantly to the higher rate of overall mortality seen in these patients.³⁰

There were several limitations to our study that may have influenced our findings. Several studies report adherence as one of the most important determinants of treatment outcomes, which we were not able to assess.^31,32 As this was an observational study, other potential confounders may have been present that may have affected clinical and immnunologic outcomes that were not accounted for (ie, opportunistic infections). In our study, viral loads were not measured and thus data on virologic outcomes were not presented, which would have been a more sensitive maker of treatment response in this cohort. Finally, the primary data were determined through physician reporting in clinical records and were therefore subject to potential inaccuracies or omissions in data. Despite these limitations, this study is one of the largest cohorts of HIV-infected patients on ART in a RLS and the program has one of the most comprehensive tracking systems ensuring reliable recording of deaths.

In conclusion, this study demonstrates feasibility and the effectiveness of large-scale HIV treatment initiatives in Tanzania and other RLS, where initiation of ART resulted in robust immunological responses and low overall mortality. However, greater efforts are needed to promote early HIV testing as well as prompt access to care and initiation of ART for those most in need of treatment. More intensive follow-up of patients in the first 3 months after ART initiation would help reduce the risk of early mortality as well as closer monitoring of high-risk individuals such as males and patients with advanced HIV disease at presentation. Nutritional interventions that address anemia and low BMI may be important as adjunct therapies to ART and need to be examined. Ongoing efforts to phaseout d4T from antiretroviral formularies must be reinforced, given the significant association between d4T use and mortality.

Footnotes

Acknowledgments

We thank Management and Development for Health (MDH), Dar es Salaam City Council, Muhimbili University of Health and Allied Sciences (MUHAS), Harvard School of Public Health (HSPH), and the Ministry of Health and Social Welfare for the guidance and collaboration in implementing a national HIV care and treatment program in Dar es Salaam, Tanzania. This program is supported by the US President’s Emergency Plan for AIDS Relief (PEPFAR) through the HSPH and by the Ministry of Health and Social Welfare, Tanzania. We thank all the patients and staff of the MDH supported care and treatment sites, who have contributed to these findings.

Authors’ Note

This paper was presented in part as a poster presentation #824 at the Conference on Retrovirus and Opportunistic infections (CROI) February 3-6, 2008, Boston, MA, USA.

Declaration of Conflicting Interests

The author(s) declared no conflicts of interest with respect to the authorship and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research and/or authorship of this article: Presidents Emergency Plan For AIDS Relief (PEPFAR).

References

UNAIDSAIDSEpidemicUpdate2009. http://www.unaids.org/en/KnowledgeCentre/HIVData/EpiUpdate/EpiUpdArchive/2009/default.asp. Accessed June 1, 2010.

Akileswaran

Lurie

Flanigan

. Lessons learned from use of highly active antiretroviral therapy in Africa. Clin Infect Dis. 2005;41(3):376–385.

Lawn

Harries

Anglaret

Myer

Wood

. Early mortality among adults accessing antiretroviral treatment programmes in Sub-Saharan Africa. AIDS. 2008;22 (15):18 97-1908.

Bisson

Gaolathe

Gross

Overestimates of survival after HAART: implications for global scale-up efforts. PLoS One. 2008;3(3):e1725.

Dalal

Macphail

Mqhayi

Characteristics and outcomes of adult patients lost to follow-up at an antiretroviral treatment clinic in Johannesburg, South Africa. J Acquir Immune Defic Synd. 2008;47(1):101–107.

The United Republic of Tanzania, Ministry of Health and Social Welfare. National Guidelines for the Management of HIV and AIDS, National AIDS Control Programme (NACP). 3rd ed. Dar Es Salaam, TZ: Government of Tanzania; 2008.

Johannessen

Naman

Ngowi

Predictors of mortality in HIV-infected patients starting antiretroviral therapy in a rural hospital in Tanzania. BMC Infect Dis. 2008;8:52.

Miettinen

. Theoretical Epidemiology. New York, NY: John Wiley & Sons; 1985.

Zachariah

Fitzgerald

Massaquoi

Risk Factors for high early mortality in patients on antiretroviral treatment in rural district of Malawi. AIDS. 2006;20(18):2355–2360.

10.

Laurent

Bourgeois

Mpoudi-Ngolé

Tolerability and effectiveness of first-line regimens combining nevirapine and lamivudine plus zidovudine or stavudine in Cameroon. AIDS Res Hum Retroviruses. 2008;24(3):393–399.

11.

Aboud

Bakari

Nyamtema

Immunological response to antiretroviral therapy in HIV-1 infected patients at Muhimbili national hospital in Dar es Salaam, Tanzania. Tanzania Dental J. 2007;14(3):30–33.

12.

Kigozi

Dobkin

Martin

Late-disease stage at presentation to an HIV clinic in the era of free antiretroviral therapy in Sub-Saharan Africa. J Acquir Immune Defic Syndr. 2009;52(2):280–289.

13.

Collini

Schwab

Sarfo

Sustained immunological responses to highly active antiretroviral therapy at 36 months in a Ghanaian HIV cohort. Clin Infect Dis. 2009;48(7):988–991.

14.

Reynolds

Nakigozi

Newell

Failure of immunologic criteria to appropriately identify antiretroviral treatment failure in Uganda. AIDS. 2009;23(6):697–700.

15.

Mee

Fielding

Charalambous

Churchyard

Grant

. Evaluation of the WHO criteria for antiretroviral treatment failure among adults in South Africa. AIDS. 2008;22(15):1971–1977.

16.

Coutinho

Mermin

Ekwaru

. Utility of routine viral load, CD4 count, and clinical monitoring among HIV infected adults in Uganda: a randomized trial [abstract 125]. In: Program and abstracts of the 15th Conference on Retroviruses and Opportunistic Infections (Boston) Alexandria, VA: Conference of Retroviruses and Opportunistic Infections; 2008:117.

17.

Castelnuovo

Manabe

Kiragga

Kamya

Easterbrook

Kambugu

Cause-specific mortality and the contribution of immune reconstitution inflammatory syndrome in the first 3 years after antiretroviral therapy initiation in an urban African cohort. Clin Infect Dis. 2009;49(6):965–972.

18.

Chang

Alamo

Guma

Two-year virologic outcomes of an alternative AIDS care model: evaluation of a peer health worker and nurse-staffed community-based program in Uganda. J Acquir Immune Defic Syndr. 2009;50(3):276–282.

19.

Bussmann

Wester

Ndwapi

Five-year outcomes of initial patients treated in Botswana's National Antiretroviral Treatment Program. AIDS. 2008;22(17):2303–2311.

20.

Marazzi

Liotta

Germano

Excessive early mortality in the first year of treatment in HIV type 1-infected patients initiating antiretroviral therapy in resource-limited settings. AIDS Res Hum Retroviruses. 2008;24(4):555–560.

21.

Ferradini

Arnoud Jean nnin

Pinoges

Scaling up of highly active antiretroviral therapy in a rural district of Malawi: an effectiveness assessment. Lancet. 2006;367(9519):1335–1342.

22.

Cornell

Myer

Kaplan

Bekker

Wood

The impact of gender and income on survival and retention in a South African antiretroviral therapy programme. Trop Med Int Health. 2009;14(7):722–731.

23.

Kamya

Mayanja-Kizza

Kambugu

Predictors of long-term viral failure among ugandan children and adults treated with antiretroviral therapy. J Acquir Immune Defic Syndr. 2007;46(2):187–193.

24.

Nachega

Hislop

Dowdy

Adherence to highly active antiretroviral therapy assessed by pharmacy claims predicts survival in HIV-infected South African adults. J Acquir Immune Defic Syndr. 2006;43(1):78–84.

25.

Bwambale

Ssali

Byaruhanga

Kalyango

Karamagi

. Voluntary HIV counselling and testing among men in rural western Uganda: implications for HIV prevention. BMC Public Health. 2008;8:263.

26.

Fairall

Bachmann

Louwagie

Effectiveness of antiretroviral treatment in a South African program: a cohort study. Arch Intern Med. 2008;168(1):86–93.

27.

Matee

Mtei

Lounasvaara

Sputum microscopy for the diagnosis of HIV-associated pulmonary tuberculosis in Tanzania. BMC Public Health. 2008;8:68.

28.

Van Cleeff

Kivihya-Ndugga

Meme

Odhiambo

Klatser

. The role and performance of chest X-ray for the diagnosis of tuberculosis: a cost-effectiveness analysis in Nairobi, Kenya. BMC Infect Dis. 2005;5:111.

29.

The antiretroviral therapy cohort collaboration: rates of disease progression according to initial highly active antiretroviral therapy regimen: a collaborative analysis of 12 prospective cohort studies. J Infect Dis. 2006;194(5):612–622.

30.

Stead

Osler

Boulle

Rebe

Meintjes

. Severe hyperlactataemia complicating stavudine first-line antiretroviral therapy in South Africa. Antivir Ther. 2008;13(7):937–943.

31.

Lanièce

Ciss

Desclaux

Adherence to HAART and its principal determinants in cohort of Senegalese adults. AIDS. 2003;17 (suppl 3):S103–S108.

32.

Palepu

Tyndall

Joy

Antiretroviral adherence and HIV treatment outcomes among HIV/HCV co-infected injection drug users: the role of methadone maintenance therapy. Drug Alcohol Depend. 2006;84(2):188–194.