Milam, Joel PhD*; Richardson, Jean L DrPH*; McCutchan, Allen MD†; Stoyanoff, Susan MPH*; Weiss, Jony MPH*; Kemper, Carol MD‡; Larsen, Robert A MD§; Hollander, Harry MD∥; Weismuller, Penny DrPH¶; Bolan, Robert MD#
Antiretroviral therapy (ART) is effective in controlling or lowering viral load in many HIV patients,1,2 but a successful response to therapy requires strict adherence to treatment regimens. Prior studies have found mean ART adherence rates of 70% to 80%.2-4 These rates of adherence, however, may be inadequate; studies indicate that taking 95% or more of doses is necessary to ensure a high likelihood of success (eg, achieving a nondetectable viral load).3,5 Because ART involves several medications with potentially unpleasant side effects and complex dosing regimens, adherence is especially challenging.6 Thus, several studies indicate that ART adherence tends to decline over time,7,8 suggesting that patients become fatigued or discouraged because of these difficulties.
Although interventions developed to improve ART adherence have shown promise,9-12 their effects may not be long lasting.13 Interventions are thus needed that not only attempt to increase adherence levels but help patients to maintain high levels during the course of their treatment.14 This may be accomplished with relatively brief (3-5 minutes) physician-directed interventions that are delivered to patients across time (eg, at each clinic visit). The HIV clinic is an ideal setting for a sustainable intervention that can be delivered to a large number of patients by primary care providers. Health care providers in primary care settings have successfully intervened to reduce tobacco,15,16 alcohol,17 and cholesterol18 intake and also to promote exercise,19 a healthy diet,20 and adherence to chemotherapy.21 HIV care providers may be similarly successful in promoting patient adherence to ART, especially when patients are satisfied with their care and trust their provider.22-25
We examined the efficacy of a brief ART adherence intervention for HIV-positive persons that was integrated within routine medical care and given to patients by their providers at each clinic visit. Based on mutual participation models of patient care,26 providers established or reinforced a trusting and committed relationship with their patients by communicating the importance of patients and providers working together as a team to help promote the patient's health. Patients received prevention messages and educational materials (specific to the issue of adherence) in printed form and verbally reinforced by providers. Providers briefly counseled patients on ways to overcome barriers to adherence and, in collaboration with patients, developed tailored pill-taking schedules. The intervention was evaluated in a before and after design using self-reported adherence to ART and medical chart data on plasma HIV RNA (as a biologic marker of virologic control).
The study evaluated an ART adherence intervention and 2 types of safer sex interventions at 6 HIV specialty clinics in California (range: 500-2500 patients per clinic). Two clinics implemented the adherence intervention, and 4 clinics implemented a safer sex intervention. The results from the safer sex protocol have recently been reported,27 and these participants serve as the attention control arm in this evaluation of the adherence intervention arm. All patients attending each clinic received the intervention assigned to their clinic. Assignment of interventions was stratified by the race and/or ethnicity of the clinic's patients (ie, the 3 clinics with the largest population of Hispanic patients were randomly allocated to the intervention arms separately from the remaining 3 clinics).
An assessment cohort was recruited from randomly selected patients at each clinic from 1998 to 1999, and baseline data were collected by means of questionnaire and medical chart review. Questionnaire data were collected by staff unaffiliated with the intervention. After a single baseline assessment, providers and staff were trained to deliver the intervention assigned to their clinic and were blinded to which patients were in the assessment cohort. The intervention was then delivered to patients attending the clinic during a 10- to 11-month period from 1999 to 2000. When the intervention ended, the study team no longer provided supportive materials, including posters, brochures, pill schedules, or booster training. The topics of adherence and safe sex were still discussed during clinic visits, however, because we expected the trained providers to have become habituated to using specific behavioral strategies that were incorporated into their routine medical visits. The cohort received a single follow-up assessment over a period of up to 7 months after the intervention support ended. An incentive payment was provided to participants at the baseline and follow-up interviews. Procedures for the protection of human subjects were approved by the institutional review boards (IRBs) overseeing each clinic and by the IRB at the Centers for Disease Control and Prevention.
Participant Selection Criteria and Recruitment to the Assessment Cohort
Trained interviewers implemented standardized recruitment procedures. At clinics that schedule patients for specific appointment times, study candidates were randomly selected for every 1-hour time block from the daily appointment schedule. At clinics that grouped patients into morning or afternoon sessions, patients were randomly selected from those registered by a specific time. Criteria for inclusion included being aware of one's HIV-positive status for at least 3 months, sexually active during the previous 3 months (mutual masturbation and oral, anal, or vaginal sex), 18 years or older, fluent in English or Spanish, able to provide informed consent, and intending to obtain care at the clinic for the next year. Enrollment continued until approximately 150 cohort patients were recruited at each clinic. Because of their limited numbers, sampling of female patients was given priority until at least 20 women (of the total 150 participants) were enrolled at each site.
A total of 2027 patients were approached to determine eligibility. Nine percent (n = 187) refused to be screened. Of those screened, 562 were ineligible because of no sexual activity in the past 3 months (88.1%), not receiving regular care (ongoing) at the clinic (6.4%), HIV-positive diagnosis less than 3 months ago (6.2%), not speaking English or Spanish (0.7%), and age less than 18 years (0.2%). Of the 1278 who were eligible, 886 (69%) enrolled, with 696 on ART. Those who were eligible but not recruited (n = 392) refused to participate because of lack of time (46.7%), not wanting to be in the study (10.0%), too ill (1.8%), other reasons (2.5%), or no reason given (39.0%). Among patients who were eligible, a significantly (P < 0.05) larger proportion of screened females (vs. males) and Hispanics (vs. other ethnic groups) enrolled. The composition of each clinic sample closely approximated the composition of the clinic population in terms of gender and ethnicity. Interviewers administered the questionnaire in private examination rooms, and none of the medical providers were involved in data collection.
Postintervention Follow-Up Procedures
At follow-up, interviewers attempted to contact all patients who participated in the baseline survey. At baseline, patients indicated whether we could contact them during their clinic appointment and/or by telephone. Most participants agreed to telephone contact, and we attempted to link interview times with their planned clinic appointments. Data collectors were encouraged to perform the interviews before the patient's clinic visit, and only 60 participants (13.7% of the analytic sample, with similar rates for each study arm) were interviewed after a clinic visit. At least 10 attempts were made to contact participants. The status of all participants who were reported as having died by clinic personnel was confirmed by death reports.
To integrate the intervention into the clinic fully, a 4-hour staff training program was delivered at each clinic. The training for the adherence intervention consisted of 6 components: (1) background data and rationale for promoting adherence to ART, (2) behavior change theories and models, (3) communication skill building, (4) how to set up a tailored pill-taking regimen and reinforce medication adherence, (5) role play of adherence counseling (mainly primary care providers, including physicians, physician assistants, and nurse practitioners), and (6) implementing the program in the clinic. Clinic staff also received a “take-home” manual covering these topics. The training was supplemented with on-site booster training 1 month after the intervention was initiated.
The adherence intervention included the following components: (1) brochures that introduced the patient to the partnership concept and messages about ART adherence; (2) posters in the waiting room that conveyed the partnership theme and posters in every examination room that included adherence messages; and (3) communication from the medical provider during the medical examination to establish and solidify the partnership, present adherence messages, and discuss pill scheduling (using the Vertex “Daily Drug Planners”) and adherence goals. At subsequent clinic appointments, patients were given 1-page informational flyers on a monthly basis to support provider messages and cover commonly asked questions about topics, including viral load and CD4 T-cell count, what the HIV medications actually do, tips to help you stick with your medications, how to keep from getting resistant virus, and finding support to help you take your medications. All materials (flyers, brochures, and posters) were presented in English and Spanish and were designed to help providers adhere to the protocol and maintain the integrity of the intervention.
The content of these intervention materials included 3 major components: information (printed and verbal), self-efficacy and skill building, and behavioral cues. The information component included articulation of the patient's regimen (eg, number of pills), potential side effects, and the general importance of ART adherence. The self-efficacy component included problem solving (eg, identifying barriers and ways to overcome them), identifying supportive people who can be encouraging, and efforts to increase patients' confidence that they can adhere to their regimen. The behavioral cue component included tailoring of pill taking and establishing cues for when to take the pills (eg, linking habitual activities with pill taking). This component included creating a detailed daily pill-taking reminder chart for the patient (using the Vertex Daily Drug Planner). This schedule included a personal timeline with stickers picturing common ART medications and stickers showing personal activities (eg, walking a dog, eating breakfast or lunch) that were placed on the schedule together to link pill taking with daily activities. Patients were advised to place this schedule at their residence, where it could be readily seen.
Safer Sex Intervention
The main aspect that systematically differed between adherence and safer sex clinics was the focus and content of the intervention. The procedures and training for implementing the safer sex protocol were similar to those used in the adherence intervention. They were developed by the same research team and used the same types of materials (brochures, posters, and flyers). Patients received a brochure that contained the same patient-provider partnership theme, printed messages on the consequences of unsafe sex and the advantages of safer sex, and tips on how to stay safe. Medical providers communicated the partnership theme, verbally stated safer sex messages to their patients, and discussed safer sex goals. At subsequent clinic appointments, patients received flyers on safer sex issues (the same number of flyers used in the adherence protocol).
Self-Report Measure of Adherence
Self-reported adherence over the previous 7 days was determined for up to 4 ART drugs for each participant. Interviewers gave participants a card with pictures and corresponding names of antiretroviral drugs to aid in recall. Participants answered questions for each antiretroviral medication they were prescribed. They were asked “how many times a day were you told to take this drug,” “how many pills were you told to take each time,” and “how many pills were missed or skipped in the past 7 days.” Seven-day adherence (percentage of pills taken as prescribed) was determined by dividing the total number of pills missed or skipped in the past week by the total number of pills prescribed per week. Because this variable was skewed at baseline and follow-up assessments (eg, 68% of the participants reported 100% adherence at baseline), medication adherence was dichotomized: participants achieving 95% or greater adherence were coded “1,” and those with less than 95% adherence were coded “0.” This cutoff point is consistent with previous research indicating that extremely high levels of adherence are needed for adequate viral suppression.3
At baseline and follow-up, data collectors abstracted the most recent HIV RNA plasma viral load from medical charts. We defined undetectable viral load as 500 copies/mL or less to avoid any discrepancies between the test methods used. Viral load was dichotomized at both time points, with a detectable viral load coded as “1” and an undetectable viral load coded as “0.”
Our analyses examined whether the intervention (1) increased adherence among those who were <95% adherent at baseline and (2) maintained high levels of adherence among those who were 95% or greater adherent at baseline. First, we examined the prevalence of adherence (<95% vs. 95% or greater) at follow-up stratified by adherence (<95% vs. 95% or greater) at baseline using χ2 tests. Next, logistic regression analyses were conducted to assess whether the adherence intervention (relative to the safer sex intervention) affected ART adherence at follow-up after statistically adjusting for preexisting differences in ART adherence at baseline. Participants were used as the unit of analysis for the primary results. Because randomization occurred at the clinic level, we also performed a cluster level analysis to adjust for similarities within each clinic, using the SAS system PROC GENMOD. In this generalized estimating equation (GEE) model, we specified the logit-link function, because the outcome was dichotomous, and an exchangeable correlation matrix indicating that each pair of patients in the same clinic was assumed to be correlated and the correlation remains the same for every pair of patients. Because there were so few clinics in this study (n = 6), we note that this analysis strategy has limited statistical power.28 Viral load was also examined as a dichotomous outcome variable in a similar set of logistic regression and GEE analyses. Before conducting these tests, we examined an array of participant-level variables (gender, sexual orientation, age, income, education, ethnicity, CD4 cell count, viral load, time since testing HIV-positive, adherence to ART, highly active antiretroviral therapy [HAART] or non-HAART regimen, and number of pills in regimen) for inclusion as covariates in the regression models. Variables that showed any indication of difference by intervention arm (using a liberal cutoff point of P ≤ 0.20) were included in the regression equations to adjust for potential confounding effects.
Because the primary outcome was ART adherence, the analytic sample included only participants who were continuously on ART and responded to the baseline and follow-up surveys. Of the 585 participants who responded to the follow-up survey, 438 were consistently on ART (58 were never on ART during the course of the study, whereas 43 stopped ART and 47 started ART during the course of the study). One of the participants always on ART did not have complete adherence data, yielding an analytic sample of 437 participants. Of the 259 patients who were on ART at baseline and were missing from the final analyses, we were able to contact 67 (26%) to determine reasons for not participating. The most common reasons were patient death (n = 16, 24%), wishing not to participate (n = 14, 21%), and moving (n = 12, 18%). Baseline characteristics were compared (χ2 and t tests) between participants in the final analytic sample and those categorized as lost to follow-up (n = 259). These analyses indicated that the analytic sample was similar to the group lost to follow-up with regard to intervention assignment (adherence vs. safer sex), baseline adherence, number of prescribed ART drugs, CD4 cell count, age, gender, ethnicity, income, employment status, and length of time since receiving an HIV-positive diagnosis. Importantly, in the adherence arm and the safer sex arm, baseline adherence (percentage 95% or greater adherent) was similar between the analytic sample and the group lost to follow-up (P > 0.60).
Those lost to follow-up were more likely to have had a detectable viral load at baseline than were those in the analytic sample (P < 0.001), however. Further analysis revealed that this difference between the analytic sample and those lost to follow-up was found only in the safer sex arm (P < 0.001) and not in the adherence arm (P = 0.77). Despite this differential attrition, the prevalence of a detectable viral load at baseline in the analytic sample was highly comparable between the 2 intervention arms (40% vs. 44% for the safe sex and adherence arms, respectively; see Table 1).
Table 1 displays a descriptive summary of the baseline demographic characteristics of the sample (n = 437). Participants were predominantly male and of white or Hispanic ethnicity. Most participants were not employed and had an annual household income in the previous year of less than $15,000 before taxes. The mean age was 39 (SD = 7.9) years, and the mean length of time since testing HIV-positive was 6.4 (SD = 4.2) years. Those in the adherence intervention arm were more likely to be white and employed and to have a higher income compared with those in the safer sex arm. Table 1 also gives summary information on participants' ART regimens. HAART was defined as 2 or more nucleoside and/or nucleotide reverse transcriptase inhibitors with a protease inhibitor or a nonnucleoside reverse transcriptase inhibitor.29 The intervention arms were similar in the percentage on HAART, the absolute number of medications prescribed, and the number of pills prescribed per day.
The extent to which participants received the adherence intervention was assessed in the follow-up questionnaire with 2 questions asked of participants at each of the 6 clinics. After statistically controlling for responses to the same questions at baseline, participants in the adherence (vs. safer sex) clinics were more likely to report that providers had (1) asked them about whether they were taking or having any problems in taking their ART medicines (odds ratio [OR] = 1.62, 95% confidence interval [CI]: 0.97 to 2.71; P = 0.06) and (2) talked about ways to make it easier to take ART medicines (OR = 2.41, 95% CI: 1.47 to 3.96; P < 0.01). These findings indicate that adherence counseling occurred more frequently at the adherence clinics than at the safer sex clinics.
Rates of 95% or greater adherence to ART regimens in the week before the baseline questionnaire were significantly higher in the adherence arm than in the safer sex arm (82.6% vs. 71.5%; χ2 = 6.41, P = 0.01), and this difference became stronger in the week before the follow-up questionnaire (85.9% vs. 69.8%; χ2 = 13.7, P < 0.001). Table 2 provides a more detailed examination of these findings by stratifying the data by baseline level of adherence and intervention arm. Among those who were <95% adherent at baseline, the percentage of persons 95% or greater adherent at follow-up was nonsignificantly higher in the adherence arm compared with the safer sex arm. Among those who were 95% or greater adherent at baseline, however, a significantly greater percentage of participants in the adherence arm stayed highly adherent compared with those in the safer sex arm (91% vs. 75%). Thus, the adherence intervention was more successful at helping patients to maintain high levels of adherence than in increasing levels of adherence.
Table 3 displays the results of logistic regression analyses of 95% or greater adherence at follow-up. The analyses were not done separately for the 2 baseline adherence groups; rather, those groups were included in the model as a covariate. Results indicated that patients in the adherence arm were significantly more likely than those in safer sex arm to be 95% or greater adherent at follow-up after adjusting for differences in baseline adherence (OR = 2.39, 95% CI: 1.40 to 4.09; P = 0.001). This effect was primarily a result of the maintenance of high levels of adherence among participants in the adherence arm, as described previously. We repeated the regression model, this time statistically controlling for potential covariates: income, ethnicity, employment status, AIDS diagnosis, HAART regimen (vs. non-HAART), and number of pills per day (also adjusting for adherence at baseline). The adherence intervention effect remained significant (OR = 2.09, 95% CI: 1.20 to 3.66; P < 0.01). We repeated this analysis using GEE (SAS PROC GENMOD) to account for clustering of patients by clinic, with simultaneous adjustment for all the potential patient covariates. Although these results were consistent with the original logistic regression results, the resulting standard errors were larger, leading to a marginally significant intervention effect (OR = 2.05, 95% CI: 0.92 to 4.56; P = 0.077).
Because there is the possibility that the observed intervention effect is driven by less discussion of adherence among providers in the safer sex clinics, we ran logistic regression models adjusting for the extent to which participants reported whether their providers counseled them on (1) whether they were taking or having any problems taking their ART medicines or (2) ways to make it easier to take ART medicines. The adherence intervention effect remained significant in both of these models (OR = 2.30, 95% CI: 1.34 to 3.95; P < 0.01 and OR = 2.19, 95% CI: 1.27 to 3.78; P < 0.01, respectively).
We further examined whether the “maintenance effect” occurred at each of the 2 clinics that implemented the adherence intervention. These analyses were performed among those who were 95% or greater adherent at baseline. Although the reduced sample size diminished the statistical power of these multivariate tests, each adherence clinic showed a higher prevalence of 95% or greater adherence at follow-up compared with the safer sex clinics pooled. In an uncontrolled analysis, the OR was 5.26 (95% CI: 1.83 to 15.17; P < 0.01) at 1 adherence clinic and 2.26 (95% CI: 0.96 to 5.30; P = 0.06) at the other adherence clinic. Similar findings were obtained after statistically controlling for the full array of covariates (OR = 5.39 [95% CI: 1.54 to 20.50; P < 0.01] and OR = 2.34 [95% CI: 0.76 to 5.08, P = 0.16], respectively).
Finally, we examined viral load (detectable = 1 vs. nondetectable = 0) as a dependent variable representing a biologic marker of adherence. Self-reported 95% or greater adherence at baseline and follow-up was significantly associated with a nondetectable viral load at follow-up (X(1)2 = 6.19, P < 0.05 and X(1)2 = 9.33, P < 0.01, respectively). The percent reduction in detectable viral load was greater in the adherence group (44% to 27% detectable = 39% reduction) compared with the safer sex group (40% to 35% detectable = 13% reduction). Thus, in a logistic regression analysis statistically controlling for detectable viral load at baseline, those in the adherence group were significantly less likely to have a detectable viral load at follow-up than those in the safer sex group (OR = 0.60, 95% CI: 0.37 to 0.98; P = 0.04). When baseline levels of adherence were added to this model, however, the statistical significance of the difference was reduced (OR = 0.63, 95% CI: 0.39 to 1.04; P = 0.07). Further, this finding was nonsignificant when the covariates were added to this model (OR = 0.89, 95% CI: 0.52 to 1.52; P = 0.67) and in a GEE (SAS PROC GENMOD) model adjusting for clustering by clinic and patient level covariates (OR = 0.91, 95% CI: 0.48 to 1.76; P = 0.79). (A similar pattern of results was found when examining a 1-log drop in viral load as the dependent variable.)
Poor adherence to ART regimens can result in failure to suppress viral load and in selection of ART-resistant strains of HIV that can be transmitted to others through sexual activity and needle sharing.30 Because ART reduces the infectivity of patients who are successfully treated,31,32 widespread sustained application of effective counseling could reduce HIV incidence in the United States, where much of the HIV-infected population is under medical care.
This controlled intervention trial at 6 HIV specialty clinics suggests that a brief ART adherence intervention delivered by primary care providers each time they saw patients may help their patients to maintain high levels of ART adherence. The intervention effect was driven largely by the maintenance of high levels of adherence among patients who were 95% or greater adherent at baseline rather than by increases in adherence among those who were initially less than 95% adherent. This finding is not surprising. Preintervention levels of ART adherence in many HIV clinic research populations are relatively high.2-4 This “ceiling effect” makes it difficult for interventions to increase adherence significantly above those levels. Other intervention studies have also shown more success in maintaining high ART adherence than in improving adherence.14 There was some evidence that the adherence intervention maintained the prevalence of nondetectable levels of plasma viral load, but the finding became nonsignificant after statistically controlling for covariates at baseline and clustering by clinic. Because participants with a detectable viral load were more likely to be lost to follow-up in the safer sex group than in the adherence group, this may have inhibited our ability to find a significant intervention effect on viral load. Further, if the intervention had a greater impact on improving adherence, it is likely that we would have seen a stronger effect on viral load. Because the analysis was conducted with patients who were currently on ART (for some, this may have been for years), it is unclear what impact our intervention would have had on patients just beginning treatment. Too few participants initiated ART during the study period to examine the impact of the intervention on this subgroup.
It is possible that the particular type of brief intervention examined here was better suited to help patients maintain high levels of adherence than to increase adherence levels. In other words, our intervention may have functioned better as a tool for providing support and motivation that reinforced positive behavior (ie, high adherence) than as a strategic tool that helped nonadherent patients to change their behavior. If this is true, our results suggest that providers should make a distinction between patients who are highly adherent and those who are not. Although adherent patients may only need ongoing supportive messages, nonadherent patients may need referral for more in-depth counseling or interventions that provide more monitoring, check-ins, and strategies for overcoming barriers to adherence. For example, adherence among HIV-positive drug users has been improved with on-site dispensation of ART doses (directly observed therapy).33 Thus, a 2-pronged approach may be needed: a more intensive client-centered approach to help improve adherence and a brief supportive intervention to help maintain adherence across time.
The study had several limitations. Although there is concern about not using a standard of care control group, our use of an attention control group controlled for the extra attention that is given to the patients as part of the intervention. In other words, if we had used a standard of care control group, any differences between arms could have been a result of mere attention differences. Nevertheless, although we statistically controlled for variables that showed any suggestion of difference between the groups, it is possible that there were other unmeasured differences that may have contributed to these results. Although staff members at the safer sex clinics were told to continue the usual adherence counseling given to their patients, they may have reduced the time they normally spent on adherence counseling to counsel on safe sex. This alternative explanation is less likely, however, because when we included basic measures of adherence counseling in the final logistic regression models, the results remained significant. This suggests that the intervention effect was driven by more than whether or not adherence was discussed with the patient. Nevertheless, a stronger study design for future research is to include an additional standard of care arm to the more intensive adherence intervention arm so as to better control for these possibilities.
These results generalize to relatively healthy, sexually active, mostly men who have sex with men, HIV-positive patients receiving outpatient care. Men who have sex with men have demonstrated a greater likelihood of achieving sustained virologic suppression compared with patients with other risk behaviors. Further, because the intervention consisted of counseling, prevention messages, and pill scheduling as well as provision of educational materials, we cannot determine the contribution of each of these components to the observed effects. Nevertheless, these components complement each another and are best conceptualized as an “intervention package.” Statistically, the study was designed to have sufficient statistical power to detect intervention effects with data pooled across the 2 clinics in the adherence intervention arm. Although the cluster level analysis was only marginally significant, we note that with only 6 clinics participating in the study, statistical power was limited.28 Further, the “maintenance effect” was observed in each of the 2 clinics assigned to the adherence intervention, although it is unclear why the effect was somewhat larger in 1 clinic than in the other.
We assessed adherence with self-reports of pill taking. Self-reports may be inflated relative to other measurement techniques (eg, medication monitoring devices).34,35 In our study, self-report bias might have been stronger in the adherence arm than in the safer sex arm, but we cannot directly test this possibility. We did, however, attempt to minimize its effects by designing the study so that providers would not be involved in collecting questionnaire data, not have access to the self-reports, and not know which patients were in the measurement cohort. We do not believe the accuracy of the adherence measure was severely compromised, because, similar to prior studies,4 we found that self-reported adherence was significantly associated with virologic data (ie, nondetectable viral load).
In summary, high levels of ART adherence among sexually active HIV-positive persons in outpatient care can be maintained with a brief provider-delivered ART adherence intervention offered to patients across time. HIV care providers can be trained to conduct brief counseling sessions with patients, and, further, brief interventions can be readily integrated into HIV clinic settings using visual cues (eg, posters in waiting room and examination rooms), printed information (eg, brochures), and provider communication and behavioral planning (eg, using the Vertex pill schedule) with patients. The intervention was less successful at improving adherence among those who were less than 95% adherent before the intervention. More intensive client-centered approaches may be necessary to improve adherence in these persons. Because many HIV patients fail to adhere to their ART regimens, HIV clinics should strive to implement and sustain adherence counseling programs and other supportive activities. Hopefully, this can improve patients' quality of care and response to treatment, with the added benefit that patients with a low viral burden may be less likely to infect others through risk behavior.
The authors thank Gary Marks for his help with this research effort.
1. Hirsch MS, D'Aquila RT, Kaplan JC. Antiretroviral therapy. In: DeVita VT, Hellman S, Rosenberg SA, eds. AIDS: Biology, Diagnosis, Treatment and Prevention. 4th ed. Philadelphia, PA: Lippincott-Raven; 1997.
2. Press N, Tyndall MW, Wood E, et al. Virologic and immunologic response, clinical progression, and highly active antiretroviral therapy adherence. J Acquir Immune Defic Syndr. 2002;31(Suppl 3):S112-S117.
3. Paterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection [comment]. Ann Intern Med. 2000;133:21-30 (erratum in Ann Intern Med 2002;136:253).
4. Haubrich RH, Little SJ, Currier JS, et al. The value of patient-reported adherence to antiretroviral therapy in predicting virologic and immunologic response. California Collaborative Treatment Group. AIDS. 1999;13:1099-1107.
5. Low-Beer S, Yip B, O'Shaughnessy MV, et al. Adherence to triple therapy and viral load response. J Acquir Immune Defic Syndr. 2000;23:360-361.
6. Kleeberger CA, Phair JP, Strathdee SA, et al. Determinants of heterogeneous adherence to HIV-antiretroviral therapies in the Multicenter AIDS Cohort Study. J Acquir Immune Defic Syndr. 2001;26:82-92.
7. Stone VE, Clarke J, Lovell J, et al. HIV/AIDS patients' perspectives on adhering to regimens containing protease inhibitors. J Gen Intern Med. 1998;13:586-593.
8. Liu H, Golin CE, Miller LG, et al. A comparison study of multiple measures of adherence to HIV protease inhibitors [comment]. Ann Intern Med. 2001;134:968-977 (erratum in Ann Intern Med. 2002;136:175.)
9. Simoni JM, Frick PA, Pantalone DW, et al. Antiretroviral adherence interventions: a review of current literature and ongoing studies. Top HIV Med. 2003;11:185-198.
10. McPherson-Baker S, Malow RM, Penedo F, et al. Enhancing adherence to combination antiretroviral therapy in non-adherent HIV-positive men. AIDS Care. 2000;12:399-404.
11. Tuldra A, Wu AW. Interventions to improve adherence to antiretroviral therapy. J Acquir Immune Defic Syndr. 2002;31(Suppl 3):S154-S157.
12. Goujard C, Bernard N, Sohier N, et al. Impact of a patient education program on adherence to HIV medication. J Acquir Immune Defic Syndr. 2003;34:191-194.
13. Fogarty L, Roter D, Larson S, et al. Patient adherence to HIV medication regimens: a review of published and abstract reports [comment]. Patient Educ Couns. 2002;46:93-108.
14. Tuldra A, Fumaz CR, Ferrer MJ, et al. Prospective randomized two-arm controlled study to determine the efficacy of a specific intervention to improve long-term adherence to highly active antiretroviral therapy. J Acquir Immune Defic Syndr. 2000;25:221-228.
15. Hollis JF, Lichtenstein E, Bogt TM, et al. Nurse-assisted counseling for smokers in primary care. Ann Intern Med. 1993;118:521-525.
16. Kottke TE, Battista RN, DeFriese GH, et al. Attributes of successful smoking cessation interventions in medical practice: a meta-analysis of 39 controlled trials. JAMA. 1988;259:2882-2889.
17. Ockene J, Adams A, Hurley T, et al. Brief physician- and nurse practitioner-delivered counseling for high-risk drinkers. Arch Intern Med. 1999;159:2198-2205.
18. Keyersling T, Ammerman A, Davis C, et al. A randomized controlled trial of a physician-directed treatment program for low-income patients with high blood cholesterol: the Southeast Cholesterol Project. Arch Fam Med. 1997;6:135-145.
19. Calfas K, Sallis J, Oldenburg B, et al. Mediators of change in physical activity following an intervention in primary care: PACE. Prev Med. 1997;26:297-304.
20. Okene J, Hebert J, Ockene J, et al. Effect of training and structured office practice on physician-delivered nutrition counseling: the Worchester-area trial for counseling in hyperlipidemia. Am J Prev Med. 1996;12:252-258.
21. Richardson JL, Shelton DR, Krailo M, et al. The effect of compliance with treatment on survival among patients with hematologic malignancies. J Clin Oncol. 1990;8:356-364.
22. Ickovics JR, Meisler AW. Adherence in AIDS clinical trials: a framework for clinical research and clinical care. J Clin Epidemiol. 1997;50:385-391.
23. Stall R, Hoff C, Coates TJ, et al. Decisions to get HIV tested and to accept antiretroviral therapies among gay/bisexual men: implications for secondary prevention efforts. J Acquir Immune Defic Syndr. 1996;11:151-160.
24. Murphy DA, Roberts KJ, Martin DJ, et al. Barriers to antiretroviral adherence among HIV-infected adults. AIDS Patient Care STDS. 2000;14:47-58.
25. Roberts KJ. Barriers to and facilitators of HIV-positive patients' adherence to antiretroviral treatment regimens. AIDS Patient Care STDS. 2000;14:155-168.
26. Szasz T, Hollender M. A contribution to the philosophy of medicine: the basic models of the doctor-patient relationship. Arch Intern Med. 1956;97:585-592.
27. Richardson J, Milam J, McCutchan A, et al. Effect of brief safer-sex counseling by medical providers to HIV-1 seropositive patients: a multi-clinic assessment. AIDS. 2004;18:1179-1186.
28. Klar N, Donner A. Current and future challenges in the design and analysis of cluster randomized trials. Stat Med. 2001;20:3729-3740.
29. Prevention CDC. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. MMMWR Morb Mortal Wkly Rep. 1998;47:1-82.
30. Wainberg MA, Friedland G. Public health implications of antiretroviral therapy and HIV drug resistance. JAMA. 1998;279:1977-1983.
31. Quinn TC, Wawer MJ, Sewankambo D. Viral load and heterosexual transmission of human immunodeficiency virus type 1: Rakai Project Study Group. N Engl J Med. 2000;342:921-929.
32. Vernazza PL, Troiani L, Flepp MJ. Potent antiretroviral treatment of HIV-infection results in suppression of the seminal shedding of HIV. AIDS. 2000;14:117-121.
33. Sorensen JL, Mascovich A, Wall TL, et al. Medication adherence strategies for drug abusers with HIV/AIDS. AIDS Care. 1998;10:297-312.
34. Bartlett JA. Addressing the challenges of adherence. J Acquir Immune Defic Syndr. 2002;29(Suppl 1):S2-S10.
35. Wagner GJ, Rabkin JG. Measuring medication adherence: are missed doses reported more accurately then perfect adherence? AIDS Care. 2000;12:405-408.
© 2005 Lippincott Williams & Wilkins, Inc.