Share this article on:

Early Clinical and Programmatic Outcomes with Tenofovir-Based Antiretroviral Therapy in Zambia

Chi, Benjamin H MD, MSc*†; Mwango, Albert MD‡; Giganti, Mark MS*†; Mulenga, Lloyd B MD, MSc*; Tambatamba-Chapula, Bushimbwa MD, MA‡; Reid, Stewart E MD, MPH*†; Bolton-Moore, Carolyn MD*†; Chintu, Namwinga MD, Mmed*†; Mulenga, Priscilla L MD*; Stringer, Elizabeth M MD, MSc*†; Sheneberger, Robert MD§; Mwaba, Peter MD, PhD¶; Stringer, Jeffrey S A MD*†

JAIDS Journal of Acquired Immune Deficiency Syndromes: 1 May 2010 - Volume 54 - Issue 1 - pp 63-70
doi: 10.1097/QAI.0b013e3181c6c65c
Epidemiology and Prevention

Background: In July 2007, amid some controversy over cost, Zambia was the first African country to introduce tenofovir (TDF) as a component of first-line antiretroviral therapy (ART) on a wide scale.

Methods: We compared drug substitutions, mortality, and “programmatic failure” among adults starting TDF-, zidovudine (ZDV)-, and stavudine (d4T)-containing ART. Programmatic failure was defined as death, withdrawal, or loss to follow-up.

Results: Between July 2007 and January 2009, 10,485 adults initiated ART (66% on TDF, 23% on ZDV, 11% on d4T), with a median follow-up time of 239 (interquartile range 98, 385) days. Those starting TDF were more likely to be male and more likely to have indicators of severe disease at baseline. In adjusted Cox proportional hazards models, ZDV- (adjusted hazard ratio [AHR] = 2.74, 95% confidence interval [CI] = 2.30-3.28) and d4T-based regimens (AHR = 1.92, 95% CI = 1.55-2.38) were associated with higher risk for drug substitution when compared with TDF-based regimens. Similar hazards were noted for overall mortality (ZDV: AHR = 0 .81, 95% CI = 0.62-1.06; d4T: AHR = 1.03, 95% CI = 0.74-1.43) and programmatic failure (ZDV: AHR = 0.99, 95% CI = 0.88-1.11; d4T: AHR = 1.11, 95% CI = 0.96-1.28) when compared with TDF.

Conclusions: TDF is associated with similar clinical and programmatic outcomes as ZDV and d4T but appears to be better tolerated. Although further evaluation is needed, these results are encouraging and support Zambia's policy decision.

From the *Centre for Infectious Disease Research in Zambia, Lusaka, Zambia; †University of Alabama at Birmingham Schools of Medicine, Birmingham, AL; ‡Zambian Ministry of Health, Lusaka, Zambia; §Institute of Human Virology, University of Maryland School of Medicine; Baltimore, MD; and ¶University Teaching Hospital, Lusaka, Zambia.

Received for publication June 9, 2009; accepted September 23, 2009.

This work was presented in part at the 16th Conference on Retroviruses and Opportunistic Infections (Abstract 142), held in Montreal Canada on February 8-11, 2009.

Correspondence to: Ben Chi, MD, MSc, Plot 1275 Lubutu Road, P.O. Box 34681, Lusaka, Zambia (e-mail:

Back to Top | Article Outline


Tenofovir disoproxil fumarate (TDF) is an important first-line agent for antiretroviral therapy (ART) in North America and Europe because of its proven effectiveness, favorable toxicity profile, and demonstrated regimen durability.1-3 In contrast, introduction of TDF into first-line therapy has been limited in resource-constrained settings, particularly in sub-Saharan Africa. Although experience with the agent has been favorable within clinical trials,4 the cost of TDF has been a major barrier to widespread use. A 1-year supply of the drug is estimated at $221 (U.S. dollars) when prescribed with lamivudine and nevirapine (NVP), a cost significantly higher than alternative first-line regimens incorporating zidovudine (ZDV; $153) or stavudine (d4T; $96).5 The toxicities associated with TDF may also be difficult to manage in settings where simple laboratory monitoring (e.g., serum creatinine) is poorly accessible. Recent work in Zambia, for example, demonstrated that approximately one third of individuals starting ART had some degree of renal insufficiency.6 Without appropriate monitoring of renal function, use of TDF could lead to iatrogenic renal failure in settings where options for management (e.g., dialysis) may be limited.

The Zambia government has implemented one of the most successful national programs for HIV care and treatment in sub-Saharan Africa. Services for HIV care and treatment have rapidly expanded, from 15,000 adults and children accessing ART in April 2004 to approximately 219,000 by December 2008 (unpublished data, A. Mwango). In July 2007, amid some donor controversy, the Zambian Ministry of Health introduced TDF as part of first-line therapy, making it the first African country to use the drug on a wide scale.7 By February 2009, more than 15,000 HIV-infected adults had initiated TDF-based therapy in Zambia's capital city of Lusaka, site of the country's largest ART program.8,9 In this report, we describe the early outcomes among patients initiating TDF-based ART in this urban African setting.

Back to Top | Article Outline


Clinical Care in Lusaka ART Program

We analyzed programmatic data from the national HIV care and treatment program in Lusaka, Zambia.8,9 This program is supported primarily by the Ministry of Health, with additional funding from the U.S. President's Emergency Plan for AIDS Relief and the Global Fund for AIDS, Tuberculosis, and Malaria. Clinical care provided across all sites has been standardized according to national protocols. Adults are enrolled and initially screened for ART eligibility. Individuals who have CD4 cell counts less than 200 cells/uL; World Health Organization (WHO) stage 4 clinical disease; or WHO stage 3 and CD4 cell counts less than 350 cells/uL are eligible for therapy. Since 2004, first-line ART has comprised a nucleoside reverse transcriptase inhibitor (NRTI) backbone, ZDV or d4T plus lamivudine, in combination with a non-NRTI (e.g. efavirenz [EFV] or NVP). In July 2007, TDF and emtricitabine were added as a nucleotide/NRTI backbone. Despite the higher cost associated with these drugs, the Ministry of Health based their decision on the TDF's favorable toxicity profile, its high genetic barrier to resistance, and its once-daily dosing schedule. Selection of a patient's first-line ART regimen is based on programmatic factors (e.g., drug availability) and medical contraindications. Because ZDV initiation is associated with clinically significant anemia,10,11 individuals with baseline hemoglobin concentration less than 10 g/dL are started on d4T- or TDF-based ART. The primary contraindication for TDF initiation is severe renal insufficiency, defined as creatinine clearance less than 50 mL/min. After an intensive lead-in period to monitor for adherence and early drug toxicities, clinical follow-up continues every 3 months. Treatment failure is diagnosed primarily by clinical or immunologic criteria; virologic testing is not widely available.12

Patient-level data are routinely entered into the program's electronic medical record,13,14 including detailed information regarding pharmacy refills at each visit. Deaths are recorded in the medical record after confirmation from family members, friends, or community health workers. “Inactive” patients have formally withdrawn from the HIV treatment program. “Late” patients are more than a month overdue for their last scheduled clinical or pharmacy visit. Through a clinic-based follow-up program, community health workers trace patients with missed visits.15

Back to Top | Article Outline

Analysis Cohort

We compared the clinical and programmatic outcomes among patients initiating TDF-, ZDV-, and d4T-based ART regimens. We analyzed programmatic data from treatment-naïve adults (≥16 yr) initiating ART across 18 primary care sites in Lusaka, Zambia from July 1, 2007 to January 31, 2009. Because the prescription of ART regimens may be based on specific medical conditions (e.g., anemia, renal insufficiency), we recognize the potential for “confounding by indication.”16 Although statistical methods can adjust for measured imbalances between the study groups, the possibility of residual confounding remains. For these reasons, we restricted our analysis to individuals with hemoglobin concentration 10 g/dL and greater and creatinine clearance 50 mL/min and greater at baseline, the two primary contraindications to specific ART regimens, in an attempt to minimize this potential bias. To compare demographic and clinical characteristics among individuals initiating different ART regimens, we analyzed categorical variables using Pearson's χ2 test. Wilcoxon rank-sum tests were used to compare continuous variables. Rates were calculated with 95% exact Poisson confidence intervals (CI).

Back to Top | Article Outline

Clinical and Programmatic Outcomes

We examined the relative impact of ART regimen on three outcomes: (1) drug substitution of the NRTI of interest (i.e. TDF, ZDV, d4T), a surrogate for drug toxicity; (2) mortality; and (3) “program failure,” a composite outcome comprising dead, inactive, and late patients. Owing to the high rates of early mortality noted in our population and others,8,17,18 we examined 90 day and less, greater than 90 day, and overall outcomes. We fit Kaplan-Meier curves to evaluate crude survival according to drug exposure category. We performed Cox proportional hazards regression to determine the hazard of each of the above-mentioned outcomes and the contribution of key covariates to those estimates. Inactive patients were censored as of the date they withdrew from care, and late patients were censored a month after their last scheduled visit.

In primary analysis, we categorized drug exposure for mortality and program failure based on an individual's prescribed ART regimen at the start of the follow-up period, as has been done in previous studies.19 For our 90 day and less and overall analyses, this was based on first prescription. For our post-90 day mortality, drug exposure was based on the regimen the patient was taking at day 90; however, we adjusted for drug substitution in the preceding interval. Baseline factors previously shown to be associated with mortality in our setting8 were included in the multivariable analysis: age, sex, body mass index, CD4 count, hemoglobin, WHO clinical stage, serum creatinine, and tuberculosis coinfection at enrollment. We included a measure of early adherence in the post-90 day analysis, using a variation of the commonly reported medication possession ratio.20,21 We divided the number of days late for pharmacy refills by total days on therapy over the first 90 days and then subtracted that percentage from 100%.12,22

Back to Top | Article Outline

Impact of ART Regimens on Renal Function

Because TDF is associated with renal insufficiency,23,24 we evaluated the impact of each drug regimen on creatinine clearance as calculated by the Cockcroft-Gault formula.25 We limited our analysis to those with available data at 6 or 12 months because routine creatinine screening has not been widely implemented. We calculated the mean change in creatinine clearance for each regimen and compared them with Student's t test. We also measured the proportion of individuals in each group whose creatinine clearance dropped below 50 mL/min at 6 or 12 months and compared them using Pearson's χ2 test. This threshold has clinical significance because individuals with creatinine clearance below it require TDF dose adjustment or substitution of the drug.26 All analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC, USA) and were approved by the institutional review boards of the University of Zambia (Lusaka, Zambia) and the University of Alabama at Birmingham (Birmingham, AL, USA).

Back to Top | Article Outline


Cohort Description

From July 1, 2007 to January 31, 2009, 10,485 ART-naïve adults meeting our eligibility criteria were initiated on HIV treatment. Of these, 6900 (66%) started a TDF-based regimen; 2446 (23%) started a ZDV-based regimen; and 1139 (11%) started a d4T-based regimen (Fig. 1). The median time of follow-up was 239 days (interquartile range [IQR]: 98, 385), which varied according to drug category: TDF (median 208 days; IQR: 89, 348), ZDV (median 306 days; IQR: 131, 479), and d4T (median 275 days; IQR: 122, 455). This difference in follow-up can be attributed to the gradual roll-out of TDF-based ART into Lusaka clinics after the Ministry of Health's incorporation of TDF into first-line therapy in July 2007 (Fig. 2). Comparison groups differed significantly according to various demographic and baseline medical characteristics. When compared with the ZDV- and d4T-based groups, those using TDF were more likely to be male and more likely to have indicators of severe disease at baseline (e.g., low BMI, low CD4 count, higher clinical stage, more frequent tuberculosis coinfection). Adults starting TDF were also more likely to be prescribed EFV rather than NVP, either to allow once-daily ART dosing or to accommodate the use of rifampin in patients requiring tuberculosis cotreatment (Table 1).

Back to Top | Article Outline

Drug Substitutions

TDF was associated with fewer drug substitutions (Fig. 3A). Overall, the rates were 9.0 per 100 person-years (95% CI = 8.1-9.9) for TDF, 27.0 per 100 person-years (95% CI = 24.6-29.5) for ZDV, and 21.5 per 100 person-years (95% CI = 18.4-24.8) for d4T. In the first 90 days, the switch rates were 8.4 per 100 person-years (95% CI = 7.0-9.9) for TDF compared with 57.7 per 100 person-years (95% CI = 51.3-64.4) for ZDV and 27.3 per 100 person-years (95% CI = 21.1-34.2) for d4T. After 90 days, the switch rates were 9.3 per 100 person-years (95% CI = 8.2-10.5) for TDF, 13.8 per 100 person-years (95% CI = 11.8-16.0) for ZDV, and 18.9 per 100 person-years (95% CI = 15.5-22.7) for d4T. In adjusted analysis, the overall hazard of switching from ZDV and d4T persisted when compared with TDF, with the greatest risk observed over the first 90 days (Table 2). Male sex and use of EFV were associated with reduced hazard for drug substitution. No associations were noted with age, CD4 count, hemoglobin, serum creatinine, clinical disease stage, and tuberculosis coinfection at time of ART initiation (Table 3).

Back to Top | Article Outline


Overall mortality for the cohort was 6.6 per 100 person-years (95% CI = 6.0-7.2), with considerable higher mortality noted in the first 90 days (13.5 per 100 person-years; 95% CI = 12.1-15.1) than after 90 days (3.3 per 100 person-years; 95% CI = 2.8-3.9) (Fig. 3B). In multivariable Cox proportional hazards models, overall mortality for those starting TDF were not statistically different from those initiating ZDV (adjusted hazard ratio 0.81, 95% CI = 0.62-1.06) or d4T (adjusted hazard ratio = 1.03, 95% CI = 0.74-1.43). These trends were consistent in subanalyses restricted to the first 90 days or after 90 days. Low BMI, low CD4 count, low hemoglobin, and advanced disease stage at baseline were all associated with elevated hazard of mortality. Women had higher risk of death, whereas individuals on tuberculosis treatment at time of ART initiation had lower risk of death (Table 3).

Back to Top | Article Outline

Programmatic Failure

We observed a high rate of programmatic failure within the cohort (31.0 per 100 person-years; 95% CI = 30.0-32.3). The overall rate was 32.2 per 100 person-years (95% CI = 30.5-33.9) for TDF; 28.1 per 100 person-years (95% CI = 25.9-30.5) for ZDV; and 32.2 per 100 person-years (95% CI = 28.5-36.0) for d4T (Fig. 3C). In adjusted models, hazards for programmatic failure among the ZDV and d4T groups were not statistically different from that of TDF (Table 2). Factors associated with programmatic failure were similar to those observed for the mortality outcome (Table 3).

Back to Top | Article Outline

Changes in Renal Function

Overall, 3539 of 6164 (57%) cohort members had creatinine clearance data available within the 6-month visit window. This was highest among those initiating TDF (2759 of 3776; 73%), followed by ZDV (523 of 1646; 32%) and d4T (257 of 742; 35%). The mean change between baseline (i.e., ART initiation) and 6 months was −14.7 mL/min for TDF, −12.7 mL/min for ZDV, and −13.8 mL/min for d4T (P > 0.05 for all comparisons). At 6 months, the proportion of individuals with creatinine clearance less than 50 mL/min was 73 of 2759 (2.6%; 95% CI = 2.1-3.3%) for TDF, 5 of 523 (1.0%; 95% CI = 0.3-2.2%) for ZDV, and 6 of 257 (2.3%; 95% CI = 0.9-5.0%) for d4T. The proportion in the TDF group was similar to that of d4T (P = 0.76) but higher than that of ZDV (P = 0.02).

At 12 months, 1402 of 3023 (46%) cohort members had creatinine clearance information available. Again, this was higher among individuals starting TDF (960 of 1560; 62%) compared with ZDV (294 of 1035; 28%) and d4T (148 of 428; 35%). Mean change between baseline and 12 months were -22.0 mL/min for TDF, -23.7 mL/min for ZDV, and -22.8 mL/min for d4T (P > 0.05 for all comparisons). At 12 months, the proportion of individuals with creatinine clearance less than 50 mL/min was 30 of 960 (3.1%; 95% CI = 2.1-4.4%) for TDF, 7 of 294 (2.4%; 95% CI = 1.0-4.8%) for ZDV, and 7 of 148 (4.7%; 95% CI = 1.9-9.5%) for d4T had creatinine clearance less than 50 mL/min. The proportions for d4T (P = 0.31) and ZDV (P = 0.51) were not statistically different compared with that of TDF.

Back to Top | Article Outline


Our early experience demonstrates that TDF is associated with fewer drug substitutions when compared with ZDV and d4T, a difference most pronounced in the first 90 days after ART initiation. When we compared the drug regimens according with other clinical and programmatic outcomes, no differences were noted. Although these results should be considered preliminary, the overall benefit associated with TDF is encouraging for programs that have incorporated the agent into first-line ART.

Rates for drug substitution for ZDV and d4T were consistent with our previous analysis of the Lusaka cohort.8 Most substitutions occurred within the first 90 days, particularly for ZDV. After 90 days, the rates for ZDV and d4T substitutions converged but remained higher than that of TDF. In multivariable Cox proportional hazards models, the risk for drug substitution for ZDV was nearly seven-fold higher than TDF during the 90 first days, but then approximated that of TDF in the interval that followed. Similar trends were observed for d4T, but significant differences were noted in both periods. These findings are consistent with data from clinical trials of TDF, in which the agent was associated with fewer adverse events when compared with either ZDV or d4T.1,2 They also demonstrate the positive impact of TDF in terms of regimen tolerability, a critical characteristic in settings where drug options may be limited.27

TDF has been shown to be an effective component of ART in a number of clinical trials. In separate studies, TDF-based regimens were found to be comparable with d4T-based regimens and superior to ZDV-based regimens in terms of virologic suppression.1,2 Because clinically evident disease progression is known to lag behind virologic treatment failure, we were not surprised to see the similar survival rates among the three comparison groups during this relatively short follow-up period. There appeared to be a trend toward improved survival associated with ZDV in our overall analysis, but it was not statistically significant. We believe this to be a limitation of our observational database, likely from the high degree of switching within the first 90 days among patients prescribed ZDV. It is also possible that, despite our efforts to control for disease status on our multivariable models, there was residual confounding from the considerably better health status among patients initiating ZDV.

Although many cohorts have focused solely on mortality as an outcome, the high rate of follow-up losses observed in many African programs suggests that ascertainment of vital status may be incomplete.15,28,29 Even when an “active” system for case finding is used, as has been done in our setting, reporting of deaths may be delayed or missed altogether.30-32 For this reason, we evaluated the outcome of program failure to provide a comprehensive, composite metric by accounting for death, program withdrawals, and follow-up losses. We found that the initial drug regimen was not related to differences in overall program failure. More targeted work is needed to determine whether TDF-based regimens, because of their reduced pill burden, can lead to improved adherence, retention, and virologic outcomes.

Although creatinine clearance data were not available for a significant proportion of cohort members, a noted limitation of this exploratory analysis, our results provide some reassurance regarding the impact of TDF-based ART on renal function at a population level. Unlike what has been observed in other African programs,33,34 mean creatinine clearance appeared to decrease over time across all patients initiating ART, but it is unclear whether these changes are clinically meaningful. Less than 3% of patients initiating TDF had a decline of creatinine clearance to below 50 mL/min, a percentage similar to that of ZDV and d4T at 12 months. This finding is consistent with clinical trials data.35,36 Future work should focus on identifying cost-effective and resource-appropriate algorithms for the monitoring of renal function among patients on long-term TDF. Although the renal failure appears to be a rare complication, it is difficult to detect clinically and may lead to severe long-term sequelae.

To our knowledge, this is the first report of clinical and programmatic outcomes associated with TDF-based ART in sub-Saharan Africa. We were able to reliably capture drug dispensations and link them to medical and outcomes information for a large population of adults on HIV treatment. However, we note several limitations. First, similar to many observational cohort analyses, there are inherent boundaries to the statistical methods used to draw associations between the drug regimens and clinical outcomes. Patients were allocated to the different comparison groups for reasons that were both selective (e.g., medical contraindications) and random (e.g., drug availability, provider preferences). We attempted to adjust for many of these potential effects through multivariable regression but acknowledge the possibility of residual confounding. Second, in an effort to minimize “confounding by indication,” we restricted our analysis cohort to individuals without anemia or severe renal insufficiency. Although we believe this approach was effective, it did limit the number of individuals who could contribute person-time in our analysis. Future work should consider alternative methodologies (e.g., use of propensity scores) to limit such biases without sacrificing sample size. Finally, our inability to reliably measure treatment failure was another limitation of this report. Numerous studies have highlighted the poor performance of immunologic algorithms,37-39 and the cost of routine virologic monitoring is prohibitive in our setting. For those reasons, we elected to focus on other clinical (e.g., mortality, toxicity) and programmatic (e.g., retention) outcomes for this analysis.

In summary, we observed favorable early outcomes with TDF-based ART. TDF was better tolerated, an important consideration in settings where drug options are limited. No differences were noted among regimens in terms of mortality or program failure. In a limited subset analysis, we found a low prevalence of severe renal insufficiency (i.e., <50 mL/min) at 6 and 12 months after ART initiation, with no increased risk associated with TDF use. Although longer-term data are clearly needed to fully validate Zambia's policy choice, these findings demonstrate the advantages of TDF-based regimens in African settings and support their continued use in the context of rapid service expansion. They also suggest that the up-front investments in first-line therapy such as with TDF may be a cost-effective strategy to increase regimen durability and improve overall patient outcomes.

Back to Top | Article Outline


The authors acknowledge the Zambian Ministry of Health for consistent and high-level support of operations research in the context of HIV program expansion. The findings and conclusions included herein are solely the responsibility of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.

Author contributions: B. Chi, A. Mwango, and J. Stringer designed the analysis plan, interpreted the data, and wrote the manuscript. M. Giganti provided data management, conducted statistical analyses, and edited the manuscript. L. Mulenga, B. Tambatamba-Chapula, S. Reid, C. Bolton-Moore, N. Chintu, P. Mulenga, E. Stringer, R. Sheneberger, and P. Mwaba contributed to the data interpretation and provided critical revision of the manuscript for intellectual content. All authors approved the final version for submission.

The work reported herein was supported in part by the President's Emergency Plan for AIDS Relief through a multicountry grant to the Elizabeth Glaser Pediatric AIDS Foundation from the U.S. Department of Health and Human Services and Centers for Disease Control and Prevention's Global AIDS Program (cooperative agreement U62/CCU12354) and a grant for Operations Research for AIDS Care and Treatment in Africa from the Doris Duke Charitable Foundation (2005047). Additional investigator salary or trainee support was provided by the National Institutes of Health (K01-TW06670; K01-TW05708; K23-AI01411; P30-AI027767) and a Clinical Scientist Development Award from the Doris Duke Charitable Foundation (2007061).

Back to Top | Article Outline


1. Gallant JE, Staszewski S, Pozniak AL, et al. Efficacy and safety of tenofovir DF vs stavudine in combination therapy in antiretroviral-naive patients: a 3-year randomized trial. JAMA. 2004;292:191-201.
2. Gallant JE, DeJesus E, Arribas JR, et al. Tenofovir DF, emtricitabine, and efavirenz vs. zidovudine, lamivudine, and efavirenz for HIV. N Engl J Med. 2006;354:251-260.
3. Willig JH, Abroms S, Westfall AO, et al. Increased regimen durability in the era of once-daily fixed-dose combination antiretroviral therapy. AIDS. 2008;22:1951-1960.
4. DART Virology Group and Trial Team. Virological response to a triple nucleoside/nucleotide analogue regimen over 48 weeks in HIV-1-infected adults in Africa. AIDS. 006;20:1391-1399.
5. Medecins Sans Frontieres. Untangling the web of antiretroviral price reductions: 11th edition. Available at: Accessed March 2, 2009.
6. Mulenga LB, Kruse G, Lakhi S, et al. Baseline renal insufficiency and risk of death among HIV-infected adults on antiretroviral therapy in Lusaka, Zambia. AIDS 2008;22:1821-1827.
7. Mwinga A. Inclusion of a tenofovir-based first-line regimen in Zambia: a bold step forward? Paper presented at 2007 HIV/AIDS Implementer's Meeting, 2007, Kigali, Rwanda.
8. Stringer JS, Zulu I, Levy J, et al. Rapid scale-up of antiretroviral therapy at primary care sites in Zambia: feasibility and early outcomes. JAMA. 2006;296:782-793.
9. Bolton-Moore C, Mubiana-Mbewe M, Cantrell RA, et al. Clinical outcomes and CD4 cell response in children receiving antiretroviral therapy at primary health care facilities in Zambia. JAMA. 2007;298:1888-1899.
10. Ssali F, Stohr W, Munderi P, et al. Prevalence, incidence and predictors of severe anaemia with zidovudine-containing regimens in African adults with HIV infection within the DART trial. Antivir Ther. 2006;11:741-749.
11. Moh R, Danel C, Sorho S, et al. Haematological changes in adults receiving a zidovudine-containing HAART regimen in combination with cotrimoxazole in Cote d'Ivoire. Antivir Ther. 2005;10:615-624.
12. Goldman JD, Cantrell RA, Mulenga LB, et al. Simple adherence assessments to predict virologic failure among HIV-infected adults with discordant immunologic and clinical responses to antiretroviral therapy. AIDS Res Hum Retroviruses. 2008;24:1031-1035.
13. Fusco H, Hubschman T, Mweeta V, et al. Electronic patient tracking supports rapid expansion of HIV care and treatment in resource-constrained settings [Abstract MoPe11.2C37]. Paper presented at the 3rd IAS Conference on HIV Pathogenesis and Treatment, 2005, Rio de Janiero, Brazil.
14. Morris MB, Chapula BT, Chi BH, et al. Use of task-shifting to rapidly scale-up HIV treatment services: experiences from Lusaka, Zambia. BMC Health Serv Res. 2009;9:5.
15. Krebs DW, Chi BH, Mulenga Y, et al. Community-based follow-up for late patients enrolled in a district-wide programme for antiretroviral therapy in Lusaka, Zambia. AIDS Care. 2008;20:311-317.
16. Walker AM. Confounding by indication. Epidemiology. 1996;7:335-336.
17. 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.
18. Zachariah R, Fitzgerald M, Massaquoi M, et al. Risk factors for high early mortality in patients on antiretroviral treatment in a rural district of Malawi. AIDS. 2006;20:2355-2360.
19. Mugavero MJ, May M, Harris R, et al. Does short-term virologic failure translate to clinical events in antiretroviral-naive patients initiating antiretroviral therapy in clinical practice? AIDS. 2008;22:2481-2492.
20. Sikka R, Xia F, Aubert RE. Estimating medication persistency using administrative claims data. Am J Manag Care. 2005;11:449-457.
21. Dezii CM. Persistence with drug therapy: a practical approach using administrative claims data. Manag Care. eb 2001;10:42-45.
22. Chi BH, Cantrell RA, Zulu I, et al. Adherence to first-line antiretroviral therapy affects non-virologic outcomes among patients on treatment for more than 12 months in Lusaka, Zambia. Int J Epidemiol. 2009;38:746-756.
23. Winston A, Amin J, Mallon P, et al. Minor changes in calculated creatinine clearance and anion-gap are associated with tenofovir disoproxil fumarate-containing highly active antiretroviral therapy. HIV Med. 2006;7:105-111.
24. Karras A, Lafaurie M, Furco A, et al. Tenofovir-related nephrotoxicity in human immunodeficiency virus-infected patients: three cases of renal failure, Fanconi syndrome, and nephrogenic diabetes insipidus. Clin Infect Dis. 2003;36:1070-1073.
25. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron. 1976;16:31-41.
26. Gilead Sciences. Full prescription information for Truvada, updated November 2008. Available at: Accessed March 1, 2009.
27. Hawkins C, Achenbach C, Fryda W, et al. Antiretroviral durability and tolerability in HIV-infected adults living in urban Kenya. J Acquir Immune Defic Syndr. 2007;45:304-310.
28. Rosen S, Fox MP, Gill CJ. Patient retention in antiretroviral therapy programs in sub-Saharan Africa: a systematic review. PLoS Med. 2007;4:e298.
29. 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.
30. Brinkhof MW, Pujades-Rodriguez M, Egger M. Mortality of patients lost to follow-up in antiretroviral treatment programmes in resource-limited settings: systematic review and meta-analysis. PLoS One. 2009;4:e5790.
31. Geng EH, Emenyonu N, Bwana MB, et al. Sampling-based approach to determining outcomes of patients lost to follow-up in antiretroviral therapy scale-up programs in Africa. JAMA. 2008;300:506-507.
32. Yu JK, Chen SC, Wang KY, et al. True outcomes for patients on antiretroviral therapy who are “lost to follow-up” in Malawi. Bull World Health Organ. 2007;85:550-554.
33. Peters PJ, Moore DM, Mermin J, et al. Antiretroviral therapy improves renal function among HIV-infected Ugandans. Kidney Int. 2008;74:925-929.
34. Reid A, Stohr W, Walker AS, et al. Severe renal dysfunction and risk factors associated with renal impairment in HIV-infected adults in Africa initiating antiretroviral therapy. Clin Infect Dis. A008;46:1271-1281.
35. Izzedine H, Hulot JS, Vittecoq D, et al. Long-term renal safety of tenofovir disoproxil fumarate in antiretroviral-naive HIV-1-infected patients. Data from a double-blind randomized active-controlled multicentre study. Nephrol Dial Transplant. 2005;20:743-746.
36. Gallant JE, Winston JA, DeJesus E, et al. The 3-year renal safety of a tenofovir disoproxil fumarate vs. a thymidine analogue-containing regimen in antiretroviral-naive patients. AIDS. 2008;22:2155-2163.
37. Reynolds SJ, Nakigozi G, Newell K, et al. Failure of immunologic criteria to appropriately identify antiretroviral treatment failure in Uganda. AIDS. 2009;23:697-700.
38. Badri M, Lawn SD, Wood R. Utility of CD4 cell counts for early prediction of virological failure during antiretroviral therapy in a resource-limited setting. BMC Infect Dis. 2008;8:89.
39. Bisson GP, Gross R, Bellamy S, et al. Pharmacy refill adherence compared with CD4 count changes for monitoring HIV-infected adults on antiretroviral therapy. PLoS Med. 2008;5:e109.

Africa; antiretroviral therapy; clinical outcomes; programmatic outcomes; tenofovir; Zambia

© 2010 Lippincott Williams & Wilkins, Inc.