Taiwo, Babafemi O MBBS*; Idoko, John A MBBS†; Welty, Leah J PhD‡; Otoh, Ihedinachi MSc†; Job, Grace RN†; Iyaji, Paul G BPharm†; Agbaji, Oche MBBS†; Agaba, Patricia A†; Murphy, Robert L MD*
Adherence to potent antiretroviral therapy (ART) is a cornerstone of successful HIV/AIDS treatment; nonadherence predicts treatment failure1-4 and development of drug resistance.5,6 In resource-limited settings, the expectations of treatment partners, family members, and health care providers may motivate patients to adhere to preserve relationships and social support.7 Although it is possible to achieve high levels of adherence in resource-limited settings,8,9 it is desirable to develop additional strategies to improve adherence and patient outcomes.
Previous studies in resource-limited settings have documented the benefits that community-based health care workers or volunteers10-12 and clinic-based peers or nurses13,14 can have on adherence. An important attribute of these programs is that they ensure disclosure of HIV infection status, which is independently protective against virologic failure.15 However, many of these programs require specially trained personnel and may be unsustainable in the global scale up of ART. Further, models that rely on health care workers visiting patients' homes risk patient confidentiality. Interventions that leverage preexisting social networks, such as patient-selected treatment partners, are of particular interest, but little data exist on their virologic effects. Between 2003 and 2004, Idoko et al16 at the Jos University Teaching Hospital (JUTH), Jos, Nigeria, enrolled 175 treatment-naive patients in a single-arm pilot study to evaluate the impact of patient-selected treatment partners on virologic outcomes. Although the study found treatment partners beneficial, the small sample size and nonrandomization of the participants limited the generalizability of the findings. Therefore, we designed this large randomized study to determine the impact of treatment partners on attainment of undetectable viral load and adherence to pickup of prescribed antiretroviral (ARV) drugs.
The JUTH HIV clinic was one of 25 sites selected to administer ART when the Federal Government of Nigeria launched the National HIV treatment Program in 2002. Since 2004, the JUTH clinic has received additional funding for patient monitoring, drug procurement, and training through a United States President's Emergency Plan for AIDS Relief grant. At the time of study, the JUTH ART program was providing treatment to approximately 5000 HIV-infected adults. All patients received free ARV drugs and monitoring of CD4+ T-lymphocyte (CD4 cell) count and viral load as part of routine care. The Doris Duke Charitable Foundation funded research-specific activities and materials.
HIV-1-infected treatment-naive adults (>15 years of age) at the JUTH ART clinic who were eligible for treatment based on a clinical diagnosis of AIDS, CD4 cell count <350 cells per cubic millimeter with HIV-related symptoms, or CD4 cell count <200 cells per cubic millimeter regardless of symptoms were invited to participate. Willingness and ability to select a treatment partner were required. We excluded individuals with severe illness such as advanced malignancy or active opportunistic infection, including tuberculosis. The Institutional Review Boards of JUTH, Jos, Nigeria, and Northwestern University, Chicago, IL, approved the study. All participants gave written informed consent administered in English or the local language (Hausa), depending on patient preference.
Potential participants were screened between June 2006 and December 2007. Participants were stratified by residence to urban (Jos metropolis and Bukuru) or rural. Using a computer-generated allocation sequence, randomization was performed in a 1:1 ratio to treatment partner-assisted (TPA) ART (treatment arm) or patient-administered standard of care (SOC) ART (control arm).
Patients initiated ART containing a nonnucleoside reverse transcriptase inhibitor (NNRTI, nevirapine or efavirenz) and 2 nucleos(t)ide reverse transcriptase inhibitors (NRTIs). All ARV drugs were dispensed in the clinic pharmacy. At study entry, each patient participated in a 2-hour interactive adherence education session conducted in a dedicated room within the clinic. The session was led by an openly HIV-infected registered nurse, who was trained as an HIV adherence counselor. Sessions were conducted in the main local language (Hausa) and English. Content included information on ART, HIV drug resistance, symptom management, and emphasis on the importance of adherence and benefits of disclosure. Each patient was given a pillbox and educated on proper use.17 The study pharmacist, who was blinded to treatment arm, provided one-on-one reinforcement of the education provided by the adherence counselor plus information specific to each participant's regimen (dosing instructions, adverse effects, dietary issues, and adaptation of dose timing to daily routine). The pharmacist had formal training in adherence counseling and more than 3 years experience in the JUTH ART clinic. At each drug pickup visit (every 28 days), the pharmacist provided targeted counseling based on the participant's self report of adherence and adverse effects. Patients who had detectable viremia (viral load >400 copies/mL) at week 24 underwent intensive adherence retraining with the adherence counselor. This retraining was aimed at identifying individualized obstacles to adherence and practical tips for overcoming them.
In addition to standard care, patients randomized to the TPA arm chose a treatment partner who was aware of the patient's HIV infection and resided in the same house or in close proximity. Treatment partners attended one adherence education session similar to that for study participants. They were asked to observe the ingestion of HIV drugs at least once daily, assist with the reporting and management of adverse effects, and remind participants of drug pickup. No compensation was provided for participation, but treatment partners with financial constraints received travel stipends of US $2-14 (total).
Study data were collected in case report forms and entered into a FileMaker Pro Version 8 database. We collected the following baseline demographic data: age, gender, highest education (none, primary, secondary, or post secondary), length of time aware of HIV infection (<1 year or >1 year), HIV disclosure (yes = informed at least 1 person, who was not a health care worker, before study entry), and residence-to-clinic distance (<20 km, 20-100 km, >100 km). Once a month, each patient received a 28-day supply of ARV drugs from the pharmacy. We recorded each drug pickup (date, number of pills dispensed, and regimen). Patients were tested for plasma viral load (Amplicor HIV-1 Monitor v1.5, lower detection limit 400 copies/mL; Roche Diagnostics, Branchburg, NJ) and CD4 cell count (Partec CyFlow, Partec, Germany) at baseline and weeks 12, 24, and 48. “Mortality was determined from hospital records, clinic charts, social contacts, treatment partners, study staff, and social workers.”
Analyses were conducted using Stata, version 9.0 (StataCorp, 2005. Stata Statistical Software: College Station, TX: StataCorp LP). Findings are reported at significance level alpha = 0.05. Virologic outcomes were (1) the proportion of patients with >1 log10 decline in viral load at week 12, (2) early virologic success, defined as undetectable viral load at week 24, and (3) durable virologic success, defined as undetectable viral load at week 48. We treated missing virologic data in 2 ways. First, we analyzed all available data, with the stipulation that participants who were missing virologic indicators and were reported to have died were counted as failures. Second, we analyzed virologic outcomes using a noncompletion as failure rule, in which any missing value was treated as failure. We also examined drug pickup adherence (calculated from pharmacy records),18,19 changes in CD4 cell counts, and mortality. Drug pickup adherence was determined at weeks 24 and 48 and dichotomized as ≥95% and <95% adherent. Percent adherence was estimated by 100 minus percent of days alive but without medication (eg, a participant who had been 6 days late picking up prescriptions through week 24 was 96.4% adherent). For participants who died during the course of the study, adherence was computed through the day of death. For patients who were lost to follow-up, we assigned the maximum number of possible days without medication. Because the ARV drugs in each patient's regimen were dispensed together, we estimated adherence to the regimen rather than individual drugs. We assumed that patients did not obtain ARV drugs from other sources. This was plausible because there were limited sources of ARV drugs in Nigeria during the study and patients were instructed to obtain drugs only through the JUTH pharmacy. Adherence between the TPA and SOC groups in the first 24 weeks was compared using Wilcoxon rank-sum tests. The analysis was repeated at week 48 to examine the durability of the intervention.
We used logistic regression to compare adherence status, virologic success, and mortality between the TPA and SOC groups. We present unadjusted odds ratios and odds ratios adjusted for gender, age, residence-to-clinic distance, and HIV disclosure. We compared baseline characteristics between the TPA and SOC groups using χ2 tests for categorical variables and 2-sample t tests or Wilcoxon rank-sum tests for continuous variables. We examined changes in CD4 counts from baseline to weeks 24 and 48 using 2-sample t tests.
A total of 499 participants met inclusion criteria and were randomized to the intervention arm (TPA, n = 248) or control arm (SOC, n = 251). Three hundred patients were ineligible for participation because they were unable or unwilling to provide a treatment partner if randomized to the TPA arm; Figure 1 details participant disposition and retention. Baseline characteristics are summarized by treatment group in Table 1. Most of the participants (65%) were female with mean age of 34.2 years (SD = 8.9 years), reflecting the demographics of HIV-infected adults in Nigeria. There were no significant differences between treatment groups in baseline viral load, CD4 cell count, gender, marital status, length of time HIV status known, HIV disclosure status, residence-to-clinic distance, or educational attainment. Treatment partners selected in the TPA arm were a nonspouse relative (54%), a spouse (34%), a friend (10%), or a neighbor or other person (2%). The majority of patients initiated nevirapine-based ART (88.4%). There was no statistically significant difference in the distribution of NRTIs between the 2 study groups; however, more patients in the TPA arm received efavirenz (16.9% vs 5.6% for SOC). We accordingly controlled for type of NNRTI (efavirenz or nevirapine) when comparing virologic and adherence outcomes.
Effect of Treatment Partners on Viral Load
Virologic, adherence, and mortality outcomes are summarized in Table 2.
A similar proportion of patients in both arms had >1 log reduction in viral load in the first 12 weeks of treatment: 80% for both groups (available data); 78% of TPA versus 76% of SOC (noncompletion as failure). Considering noncompletion as failure, TPA participants were significantly more likely to achieve undetectable viral load at week 24 (61.7% versus 50.2%, OR = 1.58, 95% CI: 1.11 to 2.26, P < 0.05). When analysis was restricted to available data (with death counted as failure), there was no significant difference in attainment of undetectable viral load between the TPA and SOC groups at week 24, although there was a trend in favor of the TPA arm (64.3% versus 55.5%, OR = 1.44, 95% CI: 0.99 to 2.09, P > 0.05). At week 48, a greater percentage of TPA than SOC achieved undetectable viral load (65.3% versus 59.4%, noncompletion as failure), however, this comparison did not achieve statistical significance. Using available data, 69% of participants in each group achieved undetectable viral load at week 48.
Effect of Treatment Partners on Adherence
Both groups exhibited high levels of adherence, but the patients in the TPA arm consistently exhibited superior adherence at weeks 24 and 48 (Table 3). Through week 24, 89% of TPA patients were at least 95% adherent versus 72% of SOC patients (OR = 3.06, 95% CI: 1.89 to 4.94, P < 0.01), Table 2. Eighty percent of TPA participants were at least 95% adherent through week 48, compared with 67% of SOC participants (OR 1.95, 95% CI: 1.29 to 2.93, P < 0.01). When we examined adherence using the categories >90% adherent, 90-80% adherent, and <80% adherent, TPA participants were still significantly more adherent than SOC participants (data not shown).
At both 24 and 48 weeks, a greater fraction of SOC patients than TPA patients were late picking up drugs (χ2 tests, P < 0.05 at both 24 and 48 weeks). Figure 2 shows box plots comparing the numbers of days late for TPA versus SOC, among living participants who were late. It illustrates that among participants who were late, those in the SOC group tended to be late by more days than those in the TPA group, (Wilcoxon rank-sum tests, P < 0.05 at both 24 and 48 weeks).
At 24 weeks, patients in both arms had comparable increases in absolute CD4 cell counts from baseline: +164 cells per cubic millimeter (SD = 113.6 cells/mm3) among the 216 TPA patients with baseline and week 24 values and +158.5 cells per cubic millimeter (SD = 130.3 cells/mm3) among the 215 SOC patients with baseline and week 24 values. There was no statistically significant difference in the mean increase in CD4 count between the 2 groups (t = 0.47, df = 429, P > 0.05). The mean change in absolute CD4 cell count between baseline and week 48 was 181.2 cells per cubic millimeter (SD = 130.2 cells/mm3) for the 205 TPA patients with baseline and week 48 values and 182.5 cells per cubic millimeter (SD = 141.9 cells/mm3) for the 201 SOC patients with baseline and week 48 values. Again was no statistically significant difference in the mean increase in CD4 cell counts between the 2 groups (t = −0.09, df = 404, P > 0.05).
There were 41 (8%) reported deaths during the 48-week period of follow-up (Fig. 1). Twenty-six of these (63%) were in the TPA group, but there was no significant difference in mortality between the 2 groups (Table 2). A total of 21 deaths occurred within 12 weeks of initiating ART: 8 occurred between weeks 12 and 24 and 12 occurred after week 24. Causes of death or terminal symptoms were pneumonia (5), severe diarrheal illness (6), uncharacterized febrile illness/possible tuberculosis (10), meningitis presumed cryptococcal (5), confirmed pulmonary tuberculosis (1), Kaposi sarcoma (1), sudden cardiac death (2), hematemesis (1), motor vehicle accident (1), and unknown (9). It is unknown if these deaths occurred in the context of immune reconstitution illness. The most common causes of death in the first 12 weeks of commencing ART were cryptococcal meningitis, diarrhea, and uncharacterized febrile illness/tuberculosis. After 12 weeks, most common causes of death were uncharacterized febrile illness/tuberculosis and diarrheal illness.
Impact of Distance of Participants' Residence on Viral Load and Adherence
Living near the clinic was significantly associated with increased success in adherence and virologic outcomes, even after adjusting for treatment group (TPA versus SOC), gender, age, and HIV disclosure status. Compared with participants living >100 km from the clinic, participants living <20 km from the clinic were significantly more likely to be at least 95% adherent at week 24 [adjusted odds ratio (AOR) 2.31, 95% CI: 1.30 to 4.10, P < 0.01], at least 95% adherent at week 48 (AOR 2.35, 95% CI: 1.43 to 3.87, P < 0.01), have undetectable viral load at week 24 (available data AOR 1.85, 95% CI: 1.20 to 2.86, P < 0.01; noncompletion as failure AOR 2.05, 95% CI: 1.34 to 3.12, P < 0.01), have undetectable viral load at week 48 (noncompletion as failure AOR 1.64, 95% CI: 1.07 to 2.52, P < 0.05), and among those with early virologic success (>1 log10 reduction in viral load by week 12), have undetectable viral load at week 48 (noncompletion as failure AOR 1.69, 95% CI: 1.00 to 2.84, P < 0.05).
This is to date the largest randomized study of patient-selected treatment partners in a resource-limited setting. Using noncompletion as failure, more patients with treatment partners achieved viral suppression at week 24; there was no significant difference when analysis was limited to available data (with death considered as failure). At week 48, treatment partners had no significant effect on rates of viral suppression, mortality, or CD4 cell count increase. Nondurability of initial virologic benefit in treatment-partnered patients may be related to the fact that all patients who were viremic at week 24 received focused adherence retraining, regardless of study arm. In addition, it is possible that some treatment-partner fatigue occurred, and enthusiasm for intervening in another individual's ART waned over time. Time-dependent decline in treatment enthusiasm is well characterized among patients themselves in the form of pill fatigue, which is associated with waning adherence and treatment failure in some previously successful patients.20
Another recent randomized study found lack of durable virologic benefit with treatment partners (the study was designed to last 24 months but was discontinued after 12 because treatment partners had no apparent impact on virologic outcomes).21 The investigators in that study reported a mortality advantage in the arm with treatment partners, which we did not find in the current study. Although we tried to discriminate mortality from loss-to-follow up, our mortality data may be subject to reporting bias because outside-hospital deaths of TPA patients may have been reported by partners.
The high adherence in both arms of our study is consistent with previous studies in sub-Saharan Africa, as reviewed by Mills et al.22 More recently in Botswana, median adherence (as determined by pharmacy refill records-similar to the method used in this study) was 91%-97% after more than 6 months of ART.23 We did observe rates of viral suppression that were lower than drug pickup adherence, which may be partly explained by the limitations of drug pickup as a measure of adherence to therapy. Although validated in developed and resource-limited settings, and shown to correlate with virologic response and mortality,24-26 drug pickup adherence reflects maximum possible adherence because patients who pick up prescriptions from the pharmacy may not ingest the medications as prescribed. Other investigators have noted similar discordance between adherence determined using pharmacy records and virologic response.27
To our knowledge, this is the first randomized study demonstrating an association between TPA ART and improved drug pickup adherence. However, better drug pickup adherence, even at levels ≥95%, did not translate into durable virologic, immunologic, or mortality benefits, although virologic advantage was seen at week 24. Our findings emphasize the importance of assessing nonadherence endpoints when evaluating adherence interventions and the need to select interventions that improve actual drug ingestion and not drug pickup alone. “Failure to observe improved viral suppression in the more adherent TPA patients may in part be explained by viral suppression occurring in a substantial proportion of patients with modest adherence (as low as 54%) to an NNRTI-based regimen.23,28,29 This is in contrast to unboosted protease inhibitor-based therapy.29,30”
Unlike other investigators,31 we found that patients who lived close to the clinic achieved better adherence and virologic outcomes compared with those who lived far away. “This association could be interpreted as support for decentralizing HIV care services in resource-limited settings. Although we controlled for potential confounders such as education and income, we are unable to claim a causal relationship between residence-to-clinic-distance and adherence or virologic outcomes. This merits further investigation.”
“Other limitations may have biased our findings. First, although we controlled for drug regimen when comparing TPA to SOC, the arms did differ significantly in the proportion of patients receiving efavirenz-containing regimens (16.9% versus 5.6% respectively). Second, persons who were unable or unwilling to identify treatment partners were excluded; hence, more than 90% of study participants reported disclosure before enrollment. Our findings may not apply to patients who are unwilling to select a treatment partner or otherwise unlikely to disclose HIV status.” The high rate of screening failure in our study (approximately one third) suggests that the uptake of treatment-partner intervention may be limited in routine clinical practice. On the other hand, the high pre-enrollment disclosure rate among study participants could have attenuated potential benefits of treatment partners because disclosure alone is a major driver of adherence and virologic success.15,32
In conclusion, use of patient-selected treatment partners was associated with improved drug pickup adherence, but it had no durable impact on viral suppression, CD4 cell replenishment, or mortality. Strategies to convert enhanced drug pickup adherence to durable improvements in the other outcomes are needed. “The relationship between residence-to-clinic distance and patient outcomes requires further exploration because it may have implications for the ongoing scale-up of ART in resource-limited settings.”
We thank Isaac Abah, Chad Achenbach, MD, MPH, and Baiba Berzins, MPH, for assisting in the development and implementation of the study. We also thank the Patients and staff of the JUTH HIV Clinic.
1. Lima VD, Harrigan R, Murray M, et al. Differential impact of adherence on long-term treatment response among naïve HIV-1 infected individuals. AIDS. 2008;22:2371-2380.
2. Nachega JB, Hislop M, Dowdy DW, et al. Adherence to nonnucleoside reverse transcriptase inhibitor-based HIV therapy and virologic outcomes. Ann Intern Med. 2007;146;564-573.
3. Tuboi SH, Harrison LH, Sprinz E, et al. Predictors of virologic failure in HIV-1-infected patients starting highly active antiretroviral therapy in Porto Alegre, Brazil. J Acquir Immune Defic Syndr. 2005;40:324-328.
4. McNabb J, Ross JW, Abriola K, et al. Adherence to highly active antiretroviral therapy predicts virologic outcome at an inner-city human immunodeficiency virus clinic. Clin Infect Dis. 2001;33:700-705.
5. Maggiolo F, Airoldi M, Kleinloog HD, et al. Effect of adherence to HAART on virologic outcome and the selection of resistance-conferring mutations in NNRTI-or PI-treated patients. HIV Clin Trials. 2007;8:282-292.
6. Roge BT, Barfod TS, Kirk O, et al. Resistance profiles and adherence at primary virological failure in three different highly active antiretroviral therapy regimens: analysis of failure rates in a randomized study. HIV Med. 2004;5:344-351.
7. Ware NC, Idoko J, Kaaya S, et al. Explaining adherence successes in sub-Saharan Africa: an ethnographic study. PLoS Med. 2009;6:e1000011. Doi:10.1371/1/journal.pmed1000011.
8. Macalino GE, Mitty JA, Bazerman LB, et al. Modified directly observed therapy for the treatment of HIV-seropositive substance users: lessons learned from a pilot study. Clin Infect Dis. 2004;38(Suppl 5):S393-S397.
9. Koenig SP, Leandre F, Farmer PE. Scaling-up HIV treatment programmes in resource-limited settings: the rural Haiti experience. AIDS. 2004;18(Suppl 3):S21-S25.
10. Mukherjee JS, Ivers L, Leandre F, et al. Antiretroviral therapy in resource poor settings. Decreasing barriers to access and promoting adherence. J Acquir Immune Defic Syndr. 2006;43(Suppl 1):S123-S126.
11. Farmer P, Leandre F, Mukherjee J, et al. Community-based treatment of advanced HIV disease: introducing DOT-HAART (directly observed therapy with highly active antiretroviral therapy). Bull World Health Organ. 2001;79:1145-1151.
12. Weidle PJ, Wamai N, Solberg P, et al. Adherence to antiretroviral therapy in a home-based AIDS care programme in rural Uganda. Lancet. 2006;368:1587-1594.
13. Pearson CR, Micek MA, Simoni JM, et al. Randomized control trial of peer-delivered, modified directly observed therapy for HAART in Mozambique. J Acquir Immune Defic Syndr. 2007;46:238-244.
14. Sarna A, Luchters S, Geibel S, et al. Short- and long-term efficacy of modified directly observed antiretroviral treatment in Mombasa, Kenya: a randomized trial. J Acquir Immune Defic Syndr. 2008;48:611-619.
15. Ramadhani HO, Thielman NM, Landman KZ, et al. Predictors of incomplete adherence, virologic failure, and antiviral drug resistance among HIV-infected adults receiving antiretroviral therapy in Tanzania. Clin Infect Dis. 2007;45:1492-1498.
16. Idoko JA, Agbaji O, Agaba P, et al. Direct observation therapy-highly active antiretroviral therapy in a resource-limited setting: the use of community treatment support can be effective. Int J STD AIDS. 2007;18:760-763.
17. Petersen M, Wang Y, van der Laan M, et al. Pillbox organizers are associated with improved adherence to HIV antiretroviral therapy and viral suppression: a marginal structural model analysis. Clin Infect Dis. 2007;45:908-915.
18. Steiner JF, Prochazka AV. The assessment of refill compliance using pharmacy records methods, validity, and applications. J Clin Epidemiol. 1997;50:105-116.
19. Steiner JF, Koepsell TD, Fihn SD, et al. A general method compliance assessment using centralized pharmacy records: description and validation. Med Care. 1988;26:814-823.
20. Ostrop NJ, Hallett KA, Gill MJ. Long-term patient adherence to antiretroviral therapy. Ann Pharmacother. 2000;34:703-709.
21. Nachega J, Goliath R, Efron A, et al. Randomized trial of trained patient-nominated treatment supporters providing directly observed ART in South African adults initiating HIV therapy. Presented at: 16th Conference on Retroviruses and Opportunistic Infections; February 8-11, 2009; Montreal, Canada. Oral Abstract 143.
22. Mills EJ, Nachega JB, Buchan I, et al. Adherence to antiretroviral therapy in sub-Saharan Africa and North America: a meta-analysis. JAMA. 2006;296:679-690.
23. Bisson GP, Rowh A, Weinstein R, et al. Antiretroviral failure despite high levels of adherence: discordant adherence-response relationship in Botswana. J Acquir Immune Defic Syndr. 2008;49:107-110.
24. Grossberg R, Zhang W, Gross R. A time-to-prescription-refill measure of antiretroviral adherence predicted changes in viral load in HIV. J Clin Epidemiol. 2004;57:1107-1110.
25. Nachega JB, Hislop M, Dowdy DW, et al. Adherence to highly active antiretroviral therapy assessed by pharmacy claims predicts survival n HIV-1 infected South African adults. J Acquir Immune Defic Syndr. 2006;43:78-84.
26. 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(5):e109. Doi:10.1371/journal.pmed.0050109.
27. Nachega JB, Hislop M, Dowdy D, et al. Adherence to non-nucleoside reverse transcriptase-based HIV therapy and virologic outcomes. Ann Intern Med. 2007;146:564-573.
28. Bangsberg DR. Less than 95% adherence to NNRTI can lead to viral suppression. Clin Infect Dis. 2006:43;939-941.
29. Weiser SD, Guzman D, Riley ED, et al. Higher rates of viral suppression with nonnucleoside reverse transcriptase inhibitors compared to single protease inhibitors are not explained by better adherence. HIV Clin Trials. 2004;5:278-287.
30. Paterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med. 2000;133:21-30.
31. Carlucci JS, Kamanga A, Sheneberger R, et al. Predictors of adherence to antiretroviral therapy in rural Zambia. J Acquir Immune Defic Syndr. 2008;47:615-622.
32. Birbeck GL, Chomba E, Kvalsund M, et al. Antiretroviral adherence in rural Zambia: the first year of treatment availability. Am J Trop Med Hyg. 2009;80:669-674.
© 2010 Lippincott Williams & Wilkins, Inc.