Objective: We conducted a randomized, multi-site, controlled trial of a cognitive-behavioral adherence intervention for patients initiating or changing an antiretroviral (ART) regimen.
Design: A 3 × 2 factorial design was used with the primary randomization assigning patients (1: 1: 1) to one of two adherence interventions or usual care.
Methods: The five-session adherence interventions consisted of cognitive–behavioral and motivational components, with or without a 2-week pre-treatment placebo practice trial. Intent-to-treat analysis used probability weights and regression tree analysis to account for missing data.
Results: A total of 230 patients were randomized; 199 started ART, of whom 74% completed the 48-week study. Electronic monitored adherence outcomes between the two intervention groups did not differ significantly and were thus pooled in analyses. At week 4, 82% of intervention patients had taken at least 90% of their prescribed ART doses, compared with 65% of controls (P < 0.01); this group difference dropped to 12% at week 12 (72 versus 60%; P = 0.15) and 11% at week 24 (66 versus 55%; P = 0.28). Mean adherence in the intervention group was significantly higher than the control group at week 24 (89 versus 81%; P < 0.05) only. There were no group differences with respect to HIV-1 RNA throughout the study.
Conclusions: The effects of the cognitive–behavioral intervention on adherence were modest and transient, and no effects were observed on viral load or CD4 cell count. More robust effects may require a more intense intervention that combines ongoing adherence monitoring and individualized intervention ‘dosage’ that matches the need and performance of each patient.
From the aRAND Corporation, Santa Monica
bLos Angeles Biomedical Research Institute at Harbor-UCLA Medical Center and the David Geffen School of Medicine at UCLA, Torrance
cUniversity of California Irvine, Irvine
dSanta Clara Valley Medical Center, San Jose
eUniversity of Southern California- Los Angeles County Medical Center, Los Angeles
fUniversity of California San Diego, San Diego, California, USA.
Received 14 December, 2005
Revised 20 February, 2006
Accepted 23 March, 2006
Correspondence to Glenn Wagner PhD, RAND, 1776 Main St., Santa Monica, CA 90407, USA. E-mail: email@example.com
Although antiretroviral therapy (ART) is a potent weapon in the clinical management of HIV, adherence is its ‘Achilles heel’ . Many studies document that adherence to ART has a critical bearing on virologic, immunologic, and clinical outcomes [2–10]. Although valuable clinical benefit can be achieved with moderate levels of adherence , the likelihood of lasting and complete virologic suppression is maximized when adherence is greater than 90 or 95% [2,3,5,12]. Unfortunately, mean adherence rates to ART generally range from 50 to 80%, and only a minority (20–40%) of patients typically achieve 90% or greater adherence when measured by electronic monitoring [2–5,13]. This report documents the findings from a randomized, controlled trial of a multicomponent cognitive–behavioral intervention designed to train patients to adhere well from the outset of ART.
In a recent review of HIV adherence intervention research, Simoni et al. identified 14 randomized, controlled trials of antiretroviral adherence interventions that had follow-up assessments . They found that even the most effective interventions produced effects that were modest and short-lived, and numerous methodological problems were identified (small sample sizes, reliance on self-reports, absence of intention-to-treat analysis). The field has yet to identify an intervention with robust and enduring effects on adherence and clinical outcomes.
We developed an intervention based on the information–motivation–behavioral skills (IMB) model of behavior change [15,16]. The intervention components include providing education about HIV, ART and the importance of adherence, tailoring the regimen to the person's daily routine, using problem-solving skills to overcome identified adherence barriers, reframing beliefs and attitudes about treatment to increase adherence self-efficacy, and facilitating positive social support for adherence. These components are commonly present in other state-of-the-art interventions . Our intervention is distinctive in including a 2-week, pre-ART placebo practice trial that simulates the challenges of ART adherence, with the exception of treatment side-effects.
We report findings from California Collaborative Treatment Group (CCTG) 578, a three-arm, randomized, controlled trial to evaluate the intervention described above. The primary goals of the study were to: (1) determine whether the intervention is superior to usual clinical care in improving adherence to ART at the primary test point (week 4) and the months thereafter; (2) assess the additive effects of supplementing the cognitive–behavioral components with the pre-treatment practice trial; and (3) determine the effects of the intervention on virologic and immunologic outcomes of treatment.
The study was conducted between September 2001 and February 2005 at the five HIV primary care clinics that comprise the CCTG, a multi-institutional, HIV clinical research organization.
Eligible patients were adults (age ≥ 18 years) in stable health (no active opportunistic infection) and planning to begin, restart, or switch to a new ART regimen containing a protease inhibitor (PI) or non-nucleotide reverse transcriptise inhibitor (NNRTI). ART-experienced patients had to report either having had problems with adherence or a belief that they could benefit from the intervention. Other eligibility criteria included HIV-1 RNA ≥ 3000 copies/ml, no active substance abuse, and English or Spanish speaking. All study participants gave informed consent, and study procedures were approved by the institutional review board at each institution.
Study design and procedures
The study consisted of a randomized 3 × 2 factorial design with three levels of adherence intervention crossed with pharmacologic-assisted therapy or standard care. Adherence was the primary outcome and week 4 the primary test point; virologic response was the secondary outcome. The primary randomization was to one of three adherence intervention groups: (1) a five-session training intervention that combines cognitive-behavioral components and a 2-week practice trial (enhanced intervention); (2) the same intervention as above but without the practice trial (cognitive–behavioral intervention); or (3) no intervention, but usual clinical care. Randomization followed a 1: 1: 1 ratio and was stratified by clinic site and whether the patient was ART-naive or experienced. The secondary randomization was to pharmacologic-assisted therapy with therapeutic drug monitoring (TDM) or standard care on a 2: 1 ratio within each of the three adherence intervention groups. Analysis from the TDM component is not presented here . Plasma concentrations of PIs and NNRTIs were measured at week 2 and reported to clinicians in the TDM arm at week 6; hence the TDM intervention occurred after the primary adherence end point.
Interviewer- and self-administered questionnaires were administered at screening (week –4) and weeks 4, 12, 24 and 48; blood draws were done at weeks –4, –2, 0 (treatment baseline), 1, 2, 4, 6, 12, 18, 24, 32, 40, and 48. Participants received modest compensation for time and travel at most study visits.
Description of adherence interventions
Both the enhanced and the cognitive–behavioral interventions were administered to participants individually by a research nurse. The first three sessions of each intervention were conducted before the patient started ART, and sessions 4 and 5 were scheduled during the first 2 weeks of treatment. The exercise content of each session was structured, but adapted to fit the needs of the individual patient; each session lasted approximately 30–45 min. A detailed outline of the content of each session can be provided upon request from the lead author (G.W.). The content of the cognitive–behavioral intervention was identical to this, except that mental rehearsal replaced the practice trial. Adherence data from the electronic monitoring caps (see below) were used to facilitate the adherence training and thus were shared with intervention patients during intervention sessions (not at follow-up assessments).
Patients in all three groups received usual care practices for enhancing adherence throughout the study. A pre-enrollment evaluation of usual care revealed that each site used the following core strategies at the start of treatment: education about the importance of adherence and the regimen's dosing instructions; tailoring the regimen to the patient's daily routine; and offering patients a pill box. After the first month of treatment, follow-up visits were generally scheduled every 3 months (or more frequently as clinically indicated), and procedures related to adherence typically consisted of inquiries about side effects and whether the patient was taking all prescribed doses.
Intervention fidelity was maintained through extensive training and supervision conducted by Dr. Wagner. All intervention sessions were audio-taped. The first four enhanced and cognitive–behavioral interventions conducted by each research nurse were closely monitored and supervised, as was every fifth intervention of each type thereafter. To limit contamination, the training of the nurses stressed the need not to deviate from the clinic's normal procedures for enhancing adherence, while still providing treatment as warranted. Usual care included some of the intervention strategies, but in a less structured, systematic way and did not include practice trials.
Electronic monitoring caps [electronic drug exposure monitor (eDEM) caps; AARDEX Ltd., Zug, Switzerland] were the primary method of adherence measurement, and were used during the 2-week practice trial, the first 4 weeks of ART, and the 2 weeks preceding assessments at weeks 12, 24 and 48. Data from the caps were downloaded into computer software to calculate the percentage of the prescribed doses taken and percentage of prescribed doses taken within specified time windows (e.g., for twice-a-day regimens, patient-reported intended dosage times were set 12 h apart, with ± 2 h windows around each intended dosage time). Adherence scores were adjusted for self-reported instances in which multiple doses were removed at once, doses were dispensed from another container, or the cap was removed without taking a dose. At weeks 4, 12, 24, and 48, participants were asked to report the number of doses taken and missed for each antiretroviral over the previous 3 days.
CD4 cell count, HIV-1 RNA level (Roche Amplicore ultrasensitive version 1.5, limit of detection 50 copies/ml; Roche Molecular Systems, Branchburg, New Jersey, USA) and all standard laboratories were done centrally (Quest Diagnostics Laboratories, San Clemente, California, USA).
The initial power analysis called for a sample size of 270 to detect a 20% difference in adherence between the enhanced intervention and the control group. Bivariate correlations were computed between continuous variables and adherence summary scores. Univariate and bivariate analyses (t-tests, analyses of variance) were used to assess differences between groups on continuous variables, and chi-squared tests were used for categorical variables. Patients with missing adherence data at specific assessment points included study dropouts and patients who remained in the study; many patients for whom data were missing at one assessment had data available at a subsequent assessment. To correct for missing data, our intent-to-treat analysis used a post-hoc weighting approach . At each assessment, patients who were most like those with missing data were assigned greater weights, thus providing representation of the patients with missing data in the analyses. The weights were calculated as the inverse of the probability of being a patient with adherence data present at the specific assessment. This probability was estimated via a logistic regression in which the outcome variable was presence or absence of adherence data and the covariates were baseline variables that predicted this outcome of non-missingness: age, sex, ethnicity, employment, education, medical insurance, parenthood, history of intravenous drug use, depression, substance use, length of time receiving care at the clinic, intervention group and clinic site.
Of the 241 patients screened for study eligibility, 230 were randomized (enhanced, 75; cognitive–behavioural, 79; control, 76), and 199 started ART. The flow of the patient sample through each time point in the study is shown in Fig. 1. Rates of treatment initiation did not differ across the three groups. Demographic, background and medical characteristics of the sample are listed in Table 1. In comparison with the remaining sample, the 31 that did not initiate treatment differed only in that they were more likely to perceive themselves as being infected by injection drug use (22 versus 6%; P < 0.01) and less likely to have a college degree (4 versus 18%; P < 0.05).
Among the 135 on ART at week 48, 112 (56% of the sample who started ART) were on ART continuously throughout the study. During study participation, 118 (59%) of the 199 patients who started ART had at least one change in ART regimen. The proportion of patients taking PI-based ART ranged from 60 to 66% over the course of the study; 72 to 76% of patients were on twice-a-day regimens, depending on the assessment, versus once-a-day (24–27%).
Intervention effects on adherence
Adherence measures were essentially identical (within one percentage point with regard to both mean percentage of prescribed doses taken and percentage of patients with at least 90% adherence) between the enhanced and cognitive–behavioral groups up to week 24 (data not shown). At week 48, adherence increased in the enhanced group, but declined in the cognitive–behavioral group; however, attrition also reached its peak at week 48. Given the lack of statistical differences in adherence and attrition rates between the two groups, these data were pooled and compared with the usual care control group. Figures 2 and 3 illustrate the electronic monitored adherence levels throughout the study for those receiving either the enhanced or cognitive–behavioral interventions versus the usual care control group. The analyses in these figures are based on the intent-to-treat analysis with all 230 randomized patients. Equivalent results were obtained when the analysis was repeated with the 199 patients who started ART. Furthermore, patients randomized to receive TDM feedback did not differ from non-TDM patients on any adherence variable at any assessment.
Percentage of prescribed doses taken
In group comparisons regarding the percentage of prescribed doses taken as measured by electronic monitoring, and 90% adherence as the marker of ‘good’ adherence, intervention patients were more likely to demonstrate good adherence in comparison with those in the control group (82 versus 65%; P = 0.01) at week 4; at weeks 12 and 24 this group difference declined to 12 and 11%, respectively (see Fig. 2). Up to week 24, the mean percentage of doses taken by patients of the intervention group remained at 90% or above, compared with nearly 80% in the control group (see Fig. 3).
Analyses involving only patients with actual data present at the specific assessment resulted in equivalent findings to those described above. For example, in comparison with control patients, intervention patients were more likely to have taken ≥ 90% of prescribed doses at week 4 (83 versus 66%; P = 0.01), and have higher mean adherence at week 24 (90.2 versus 82.5%; P = 0.02).
Self-reported adherence data showed no group differences at any assessment point (data not presented). The mean percentage of prescribed doses taken ranged from 95 to 98% within each group over the course of the study.
Percentage of prescribed doses taken within specified window of time
The proportion of intervention patients who took at least 90% of their doses within the specified time window was 56%, compared with 40% in the control group at week 4 (P = 0.05); at week 12, the group difference narrowed to 10% (P = 0.26), and at weeks 24 and 48 the groups were almost identical. There was an approximate 5% group difference in mean percentage of prescribed doses taken on time throughout the study, but no statistical significance was found. Similar findings were observed in unweighted analyses involving only patients with actual data at each assessment, and which did not attempt to represent patients with missing data.
Relationship between complete intervention attendance and adherence
Among the 132 patients who started ART and were assigned to either the enhanced or cognitive–behavioral intervention, 114 (86%) attended all five sessions. A greater proportion of patients who completed all sessions took at least 90% of their prescribed doses at week 4 (85 versus 50%; Fisher's exact test = 0.03) and week 12 (75 versus 43%; Fisher's exact test = 0.09) as measured by electronic monitoring, compared witho those who missed at least one session. Those who attended each session were also more likely to take at least 90% of their doses within the specified time window at week 4 (59 versus 25%) and week 12 (54 versus 29%); however, these differences were not significant.
Intervention effects on virologic and immunologic outcomes
The three groups (enhanced, cognitive–behavioral, control) did not differ with regard to the amount of change in HIV-1 RNA and CD4 cell count from baseline, nor the proportion of patients with HIV-1 RNA < 400 copies/ml at each time point (see Table 2). Similar results were obtained when analyses were repeated controlling for whether the patient was antiretroviral naive or experienced at baseline.
Relationship between adherence and virologic and immunologic outcomes
At week 24, a greater amount of reduction in HIV-1 RNA from baseline was significantly correlated with a higher percentage of doses taken as measured by electronic monitoring (r = −0.20; P = 0.03) and self-report (r = −0.28; P = 0.001). At week 48, the reduction in HIV-1 RNA from baseline was significantly correlated with the percentage of doses taken as measured by electronic monitoring (r = −0.30; P = 0.004) and self-report (r = −0.28; P = 0.003), and the mean percentage of doses taken in the specified time window (r = −0.23; P = 0.03). Patients who took at least 90% of their prescribed doses (measured by electronic monitoring) at week 48 experienced a greater reduction in log10 HIV-1 RNA from baseline than did the remaining sample (mean change of −2.7 versus −2.2; P = 0.05); those who took at least 90% of their prescribed doses within the specified window of time at week 48 also experienced a greater reduction in HIV-1 RNA (mean change of −2.8 versus −2.3; P = 0.04). At week 48, 59 of 137 (43%) patients had HIV-1 RNA levels below 400 copies/ml, but this was unrelated to any measure of adherence. The change in CD4 cell count from baseline to week 48 was positively correlated with self-reported adherence at week 48 (r = 0.21; P = 0.03), but no other adherence variable.
The cognitive–behavioral intervention evaluated in this study helped patients take at least 90% of prescribed doses in the initial weeks following the completion of the intervention, and to take doses on time. Intervention patients sustained a mean adherence level of 90% or more for 24 weeks, and a large majority maintained this adherence level throughout the study. The nearly 10% difference in mean adherence between the groups at week 24 was statistically significant; however, although such a difference has been found to be clinically meaningful with regard to viral suppression and disease progression in other studies [6,8,10], no such associations were found in this trial. Adherence rates generally remained high across all groups, with few group differences beyond week 4. Therefore, as in other published controlled trials of ART adherence interventions, the effects on adherence observed in this study were modest and relatively short-term [19,20], and no effects were found with regard to virologic and immunologic outcomes [19,21–23].
Study data showing that most intervention effects on adherence dissipated after the discontinuation of the training sessions, together with research demonstrating that adherence declines over time [24,25], suggest there is a need for ongoing adherence monitoring and maintenance training. Some patients may need extensive ongoing training whereas others may need minimal or no training to achieve and sustain optimal adherence, as indicated by the high adherence levels sustained by many of our control patients. Adherence interventions to date have been implemented uniformly to all study participants, rather than adjusting the need for intervention and maintenance training based on individual adherence performance. If a simple and accurate method can be identified for ongoing adherence monitoring for all patients, then adjusting the amount of training to match the needs of the patient may result in more efficient use of clinic resources and better acceptance of the program to patients, providers, and administrators. Although the study data failed to demonstrate any benefit of practice trials as a supplement to cognitive–behavioral training, these trials could be a valuable tool for evaluating adherence readiness and adjusting the amount of pre-treatment training, as supported by our previous research [26,27].
The study data did not reveal any intervention effect on HIV-1 RNA levels or CD4 cell counts. The ability to detect an intervention effect on these outcomes was hampered by several factors. A majority of the sample had previously been on ART; thus, it is likely that many had antiretroviral resistance at screening. Drug resistance attenuates response to ART  and the relationship between adherence and resistance is more complex than was once thought ; nonetheless, it remains clear that higher levels of adherence are strongly associated with superior clinical and virologic outcomes [2–8]. This is supported by our study findings, which serve to validate the study's adherence data.
Levels of adherence across all three groups of patients were remarkably high. Mean electronically monitored adherence levels generally ranged from 80 to 95% over the course of the study, which is much higher than most studies using electronic monitoring [2–5,13] and calls into question whether the patients in the study are representative of patients in general care. One possible explanation for the overall high adherence is the extensive study procedures and high frequency of study visits, resulting in a potential effect of attention across all groups. However, since the visit schedule was nearly identical for all study participants, it would not explain any observed group differences. The high adherence rates may reflect a high level of attention to adherence as part of usual care at the study sites, or that a sizeable proportion of the sample may have already had good adherence skills at the outset of the study, thus contributing to a ceiling effect that limited the study's ability to assess the full potential of the intervention. Variations in usual care across clinical settings and in adherence skills across patient populations need to be considered in interpreting the effectiveness of adherence interventions in clinical trials .
In summary, results from this controlled trial revealed that the front-loaded cognitive–behavioral intervention had modest, transient effects in improving antiretroviral adherence, and no effects on viral load and CD4 cell count. For effects to be more robust and durable, interventions may need to vary the amount of training, and perhaps the nature of the training strategy as well, utilizing the full armament of adherence enhancing strategies (e.g., cognitive–behavioral counseling, beeper/alarm reminders, directly observed therapy) depending on the needs of the patient. Accordingly, a practical and accurate method to assess the patients' need for adherence intervention would be of great value.
We would like to acknowledge the CCTG research group (Drs Richard Haubrich, J. Allen McCutchan, Miguel Goicoechea, Jeremiah Tilles, Catherine Diamond, Eric Daar, Mallory Witt, Loren Miller, Mario Guerrero, Carol Kemper, and Robert Larsen), as well as the study research nurses at each site (Beth Michel, Rebecca Sutton, Edward Seefried, Tomasa Maldonado, Robert Sanchez), the CCTG Data Center (Andrew Rigby, Pete Suggett and Shelly Sun), and the clinicians and patients who participated in the trial. The CCTG-affiliated clinics that participated in the study are located in San Diego (UCSD), Irvine (UCI Medical Center, General Clinical Research Center, College of Medicine, with funds provided by the National Center for Research Resources, 5M01RR 00827-29), Santa Clara (Santa Clara Valley Medical Center), and Los Angeles County (USC County Medical Center; Harbor-UCLA Medical Center).
Sponsorship: This research was supported by funds from the National Institute of Mental Health, grant number RO1MH61695, and the Universitywide AIDS Research Program of the University of California, center grant numbers CC99-SD003 and CC02-SD-003. Quest Laboratories and Monogram Biosciences provided all laboratory assays.
1. Rabkin JG, Chesney MA. Treatment adherence to AIDS medications: the Achilles heel of the new therapeutics. In: Ostrow D, Kalichman S, editors. Psychosocial and Public Health Impacts of New HIV Therapies. New York: Kluwer/Plenum; 1999. pp. 61–68.
2. Paterson DL, Swindells S, Mohr J, Brester M, Vergis EN, Squier C, et al
. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med 2000; 133:21–30.
3. Arnsten JH, Demas PA, Farzadegan H, Grant RW, Gourevitch MN, Chang CJ, et al
. Antiretroviral therapy adherence and viral suppression in HIV-infected drug users: comparison of self-report and electronic monitoring. Clin Infect Dis 2001; 33:1417–1423.
4. Howard AA, Arnsten JH, Lo Y, Vlahov D, Rich JD, Schuman P, et al
. A prospective study of adherence and viral load in a large multi-center cohort of HIV-infected women. AIDS 2002; 16:2175–2182.
5. McNabb J, Ross JW, Abriola K, Turley C, Nightingale CH, Nicolau DP. Adherence to highly active antiretroviral therapy predicts outcome at an inner-city human immunodeficiency virus clinic. Clin Infec Dis 2001; 33:700–705.
6. Bangsberg DR, Perry S, Charlebois E, Clark RA, Roberston M, Zolopa AR, et al
. Non-adherence to highly active antiretroviral therapy predicts progression to AIDS. AIDS 2001; 15:1181–1183.
7. Haubrich R, Little SJ, Currier JS, Forthal DN, Kemper CA, Beall GN, et al
. The value of patient-reported adherence to antiretroviral therapy in predicting virologic and immunologic response. AIDS 1999; 13:1099–1107.
8. Hogg RS, Heath K, Bangsberg D, Yip B, Press N, O'Shaughnessy MV, et al
. Intermittent use of triple-combination therapy is predictive of mortality at baseline and after 1 year of follow-up. AIDS 2002; 16:1051–1058.
9. Wood E, Hogg RS, Yip B, Harrigan PR, O'Shaughnessy MV, Montaner JS. Effect of medication adherence on survival of HIV-infected adults who start highly active antiretroviral therapy when the CD4+ cell count is 0.200 to 0.350 × 10(9) cells/L. Ann Intern Med 2003; 139:810–816.
10. Bangsberg DR, Hecht FM, Charlebois E, Zolopa AR, Holodniy M, Sheiner L, et al
. Adherence to protease inhibitors, HIV-1 viral load, and development of drug resistance in an indigent population. AIDS 2000; 14:357–366.
12. Matthews WC, Mar-Tang M, Ballard C, Colwell B, Abulhosn K, Noonan C, et al
. Prevalence, predictors, and outcomes of early adherence after starting or changing antiretroviral therapy. AIDS Patient Care & STDs 2002; 16:255–270.
13. Wagner GJ, Ghosh-Dastidar B. Electronic monitoring: Adherence assessment or intervention? HIV Clin Trials 2002; 3:45–51.
14. Simoni J, Pantalone D, Frick PA, Turner B. Enhancing antiretroviral adherence: a review of published reports of RCTs and ongoing NIH-funded research. In: Trafton J, Gordon W, editors. Best Practices in the Behavioral Management of Chronic Disease, Volume II: Other Medical Disorders. Los Altos, CA: Institute for Brain Potential; 2005. pp. 70–95.
15. Fisher JD, Fisher WA. Changing AIDS risk behavior. Psychol Bull 1992; 111:455–474.
16. Fisher JD, Fisher WA, Misovich SJ, Kimble DL, Malloy TE. Changing AIDS risk behavior: effects of an intervention emphasizing AIDS risk reduction information, motivation, and behavioral skills in a college student population. Health Psychol 1996; 15:114–123.
17. Haubrich R, Best B, Witt MGM, Kemper C, Larsen R, Diamond C, et al
. Determinants of the need for therapeutic drug monitoring (TDM): rates and predictors from CCTG 578
. XII Conference on Retroviruses and Opportunistic Infections
, Boston, MA, February 2005 [abstract 640].
18. Brick JM, Kalton G. Handling missing data in survey research. Stat Meth Med Res 1996; 5:215–238.
19. Remien RH, Stirratt MJ, Dolezal C, Lobozzo JS, Wagner GJ, Carballo-Dieguez A, et al
. Couple-focused support to improve HIV medication adherence: a randomized controlled trial. AIDS 2005; 19:807–814.
20. Rigsby MO, Rosen MI, Beauvais JE, Cramer JA, Rainey PM, O'Malley SS, et al
. Cue-dose training with monetary reinforcement: pilot study of an antiretroviral adherence intervention. J Gen Intern Med 2000; 15:841–847.
21. Knobel H, Carmona A, Lopez JL, Gimeno JL, Saballs P, Gonzalez A, et al
. Adherence to very active antiretroviral treatment: Impact of individualized assessment. Enferm Infect Microbiol Clin 1999; 17:78–81.
22. Goujard C, Bernard N, Sohier N, Peyramond D, Lancon F, Chwalow J, et al
. Impact of a patient education program on adherence to HIV medication: a randomized clinical trial. J Acquir Immune Defic Syndr 2003; 34:191–194.
23. Rawlings MK, Thompson MA, Farthing CF, Brown LS, Racine J, Scott RC, et al
. Impact of an educational program on efficacy and adherence with a twice-daily lamivudine/zidovudine/abacavir regimen in underrepresented HIV-infected patients. J Acquir Immune Defic Syndr 2003; 34:174–183.
24. Mannheimer S, Friedland G, Matts J, Child C, Chesney M. The consistency of adherence to antiretroviral therapy predicts biologic outcomes for human immunodeficiency virus-infected persons in clinical trials. Clin Infect Dis 2002; 34:1115–1121.
25. Howard AA, Arnsten JH, Lo Y, Vlahov D, Rich JD, Schuman P, et al
. A prospective study of adherence and viral load in a large multi-center cohort of HIV-infected women. AIDS 2002; 16:2175–2182.
26. Wagner GJ. Placebo practice trials: The best predictor of adherence readiness for HAART among drug users? HIV Clin Trials 2003; 4:269–281.
27. Wagner GJ, Iguchi M, Schneider S, Scott J, Anderson D. Placebo practice trials: a tool to assess and improve adherence readiness. HIV Clin Trials 2002; 3:475–481.
28. Liu H, Miller LG, Hays RD, Wu T, Golin CE, Wenger NS, et al
. Repeated measures longitudinal analyses of HIV virologic response as a function of percent adherence, dose timing, genotypic sensitivity, and other factors. J Acquir Immune Defic Syndr 2006; 41:315–322.
29. Bangsberg DR, Moss AR, Deeks SG. Paradoxes of adherence and drug resistance to HIV antiretroviral therapy. J Antimicrob Chemother 2004; 53:696–699.
30. Wagner GJ, Kanouse DE. Assessing usual care in clinical trials of adherence interventions for HAART. J Acquir Immune Defic Syndr 2003; 33:276–277.
Keywords:© 2006 Lippincott Williams & Wilkins, Inc.
adherence; antiretroviral; HIV; intervention; cognitive-behavioral