Secondary Logo

Journal Logo

Clinical Science

Antidepressant Treatment Improves Adherence to Antiretroviral Therapy Among Depressed HIV-Infected Patients

Yun, Lourdes W. H MD, MSPH*†; Maravi, Moises BS, MSc*; Kobayashi, Joyce S MD*‡; Barton, Phoebe L PhD; Davidson, Arthur J MD, MSPH*†§

Author Information
JAIDS Journal of Acquired Immune Deficiency Syndromes: April 1, 2005 - Volume 38 - Issue 4 - p 432-438
doi: 10.1097/01.qai.0000147524.19122.fd
  • Free


Medication adherence is key to obtaining optimal benefit from any effective drug regimen. Adherence, defined as the extent to which a person's behavior coincides with medical advice,1 is a multifactorial process involving the individual patient, the treatment regimen characteristics, and the quality of the patient-provider interaction.2-4 Nonadherence is a significant problem; estimates of therapeutic regimen adherence range between 18% and 80%,5-7 with half of chronically ill patients exhibiting some degree of nonadherence.8-10

Aggressive treatment with combination antiretroviral therapy (ART) successfully suppresses HIV viral loads, preventing the onset of full-blown AIDS.11 Near-perfect adherence is required, however, because suboptimal drug levels have been associated with decreased viral suppression and development of antiretroviral resistance.12 Resistance prevention is important to avoid cross-resistance to other ART agents within each of the 3 major drug classes now available.13 Therefore, nonadherence to ART not only decreases current treatment effectiveness but may lead to permanent treatment ineffectiveness because of the development of drug-resistant mutations.

For most diseases, maintaining an adherence level greater than 80% of prescribed doses is associated with positive therapeutic outcomes14; however, this level of ART adherence may be insufficient for viral suppression. Although the adherence threshold necessary to ensure viral suppression has not been established, persons with high adherence rates (ie, >95%) were significantly more likely to have undetectable viral loads.12-15 Such rates are higher than those generally observed in clinical settings16,18,19

AIDS was once considered a rapidly fatal disease. Effective ART, improved HIV clinical care, and prophylaxis for preventable AIDS-associated opportunistic conditions have increased survival time, redefining AIDS as a chronic and manageable disease. Improved survival rates, in turn, imply a longer period of vulnerability for the development of psychiatric comorbidity. Psychiatric disorders have been associated with many chronic and/or serious medical illnesses, with an estimated lifetime prevalence of 42%.20

Among HIV-infected patients, the estimated lifetime prevalence of at least 1 psychiatric disorder is 38% to 75% (compared with 33% for the general population), with higher rates among HIV-infected homosexual men (range: 80%-88%).21,22 Lifetime depression prevalence rates among HIV-infected patients are estimated to be between 22% and 45% (compared with 15% for the general population),23-26 with current clinical depression observed in 4% to 20%.24,27-35 Although depression may occur at any HIV disease stage, incidence increases with disease progression and is correlated with development of AIDS.33

This association between psychiatric disease and HIV infection is complex, with interactions that may result in suboptimal outcomes for either. For example, patients with psychiatric disorders often exhibit behaviors such as self-destructiveness, impulsiveness, and/or substance abuse that may increase the risk of acquiring HIV infection, poor self-care, and subsequent transmission to others.36 Psychiatric disorders, particularly depression, have been associated with medication nonadherence among patients with HIV infection and other medical conditions.12,37-41 Despite the high prevalence of depressive disorders among HIV-infected patients, psychiatric treatment and antidepressant medication are not universally used because of patient, provider, and/or system barriers to care. Primary care providers may underdiagnose psychiatric disorders or misclassify a number of depressive symptoms that overlap with those of HIV infection (ie, poor appetite, weight loss, loss of energy, insomnia), resulting in delayed treatment.33,42 Once diagnosed, depression is effectively treated with antidepressant and short-term interpersonal therapy; among those treated, 85% respond to medication and 50% exhibit complete recovery.33

For depressed HIV-infected patients, effective treatment of underlying depression may correlate with improved ART adherence. Understanding the contribution of depression and its subsequent treatment on ART adherence might direct clinicians toward earlier identification and more aggressive treatment among this population. Although prior research has demonstrated that depression may be effectively treated among HIV-infected patients,33 no prior study has evaluated the effect of antidepressant medication therapy in improving ART adherence. This study examines the impact of antidepressant treatment (ADT) on ART adherence and assesses the temporal relation of ADT treatment on ART adherence in a population of depressed HIV-infected patients.


Study Population and Study Design

This retrospective cohort study is based on Denver Public Health Department surveillance data derived from the Centers for Disease Control and Prevention (CDC)-funded multicenter Adult and Adolescent Spectrum of HIV Disease (ASD)43 and Supplement to HIV/AIDS Surveillance (SHAS)44 projects (Colorado Multiple Institutional Review Board Protocol 98-838 and 98-273, respectively). All HIV-infected patients older than 12 years of age who received clinical services from Denver Health (DH) were enrolled in the study.

DH is an integrated health care system that provides care to 120,000 primarily indigent and under- or uninsured citizens of Denver County (20% of the population) and consists of a municipal safety net public hospital and a network of 11 community health centers throughout the county. Comprehensive HIV clinical services are provided at 3 early intervention clinics and an infectious diseases/AIDS clinic funded through Ryan-White Title I, II, and III funds.

Data Collection

Using standard forms, ASD data from January 1997 through December 2001 were collected by trained abstractors who retrospectively reviewed electronic images of medical charts for the 12 months before enrollment and at 6-month intervals through December 31, 2001, or until patients were lost to follow-up or death. These standard forms were developed initially in 1990 and revised periodically as part of the CDC-funded multisite study. All abstractors were trained and followed a standard CDC-defined protocol for chart abstraction. New abstractors had complete review (dual abstraction) of all charts by an experienced abstractor for quality and accuracy for a minimum of 3 months. Thereafter, randomly sampled (5%) charts were reabstracted for quality review and consistency. Only after any new abstractor was properly trained and there was sufficient concordance was dual abstraction discontinued. Data collected included basic demographics; mode of HIV exposure; occurrence of AIDS-defining conditions; prescribed use of ART medications; and other medical conditions, including depression and alcohol and substance use (intravenous and nonintravenous drugs). Since 1997, all DH outpatient encounter forms and hospital charts have been electronically scanned within 24 hours for review by authorized individuals over a wide area network. Access is nearly instantaneous, and virtually 100% of all records are retrievable.

Medical and psychiatric visit-specific utilization data for the same period were retrieved from administrative information systems, including date of visit, International Classification of Diseases (ICD)-9 diagnosis code(s), service area, and health care provider. A measure of annualized medical (primary and urgent care) and psychiatric (with identified psychiatric providers) visits was calculated based on the number of visits by a patient divided by the duration of observation from the first visit to the last visit in the system (minimum of 6 months of observation). Psychiatric visit rates were stratified based on a minimal expected number of visits for patients in psychiatric care (ie, 2 visits per year of observation). Agency-wide outpatient electronic pharmacy data for the same period were retrieved, including type and dose of medication, date a drug was dispensed, number of pills dispensed, and expected number of days for which that prescription was supplied. Because the dedicated HIV pharmacy provides virtually all AIDS Drug Assistance Program medications at a significantly discounted rate, nearly all prescriptions were filled within this system.

Additional socioeconomic variables (ie, income, education, employment status, living status [patient living with family members, partner, or friends]) used in the analyses were collected from a single-time SHAS questionnaire-based interview conducted with these same patients. Information on selected variables from this questionnaire was available for approximately half of the enrolled patients.

Adherence Measure

For this study, we use the term adherence to indicate the availability of medication retrieved from a DH pharmacy. Pharmacy records were used to calculate ART and ADT adherence rates. Most pharmacy-based drug adherence studies calculate rates for 1 or more simultaneously prescribed drugs. ART regimens usually involve frequent changes (eg, drugs may be added or subtracted over time); thus, calculated ART adherence rates were modified from methods described in the literature for other medications.14,45 ART adherence was defined as the sum of dispensed days for each ART drug divided by the sum of the observation periods (from first to last refill plus days dispensed) for each ART drug. A similar process was used to calculate ADT adherence. ART and ADT adherence rates were computed for patients with at least 2 refills for the same medication during the 6-month minimal medical care observation period. If all medications were picked up during the observation period, the adherence rate would be 100%. Patients were considered adherent if the rate was equal to or greater than the median adherence value for ADT drugs and if adherence was equal to or greater than 95% for ART drugs.

Diagnosis of Depression

Depression was considered present if it had been identified in any of the 3 potential data sources: (1) administrative data (diagnosis established by a psychiatric provider meeting formal Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) criteria46 and ICD-9 coded for Major Depression and Bipolar Disorders [ie, 296.20-296.36 and 296.50-296.66]), (2) chart review (primary care provider assessment or diagnosis in medical charts) in which patients may have met formal Major Depressive Disorder criteria but never followed through with psychiatric referral, or (3) electronic pharmacy records (prescription of ADT). Patients who had received antidepressant medications for other medical conditions (eg, amitriptyline for peripheral neuropathy, trazodone for sleep) were identified on the basis of the low-dose range routinely used for those conditions and were considered not to have received ADT.

Statistical Analysis

SAS (Statistical Analysis Software, Cary, NC), version 8.1, was used for all analyses. Because all patients did not complete the SHAS interview, denominator data are included to indicate the portion of the sample included in the interview analyses. Bivariate analysis assessed the independent effect of age, race, gender, education, income, employment, living or not living with someone, alcohol and drug use, number of primary care provider visits, number of psychiatric provider visits, and ADT adherence or ART adherence. Logistic regression analyses assessed confounding (P < 0.25) and effect-modifying factors (P < 0.05) while constructing a final model that controlled for variables and interactions that might affect the association between ADT adherence and ART adherence. Independent variables and interactions included in the final model were those of significance in the bivariate analysis and shown to alter the final model significantly, respectively.

To assess the temporal effect of ADT prescription and ADT adherence on ART adherence levels, paired t tests were performed to compare the difference in ART adherence for 6 months before versus 6 months after ADT initiation among ADT-adherent and ADT-nonadherent patients. We also compared the temporal association of ART adherence among those who were depressed but never received any ADT (from pharmacy records); for those individuals, the ART adherence levels were compared for 6-month intervals preceding and following the date of first depression diagnosis to the mean time to initiation of ADT among those who started on ADT.


Between January 1, 1997 and December 31, 2001, 1713 HIV-positive patients with a minimum of 6 months of care within DH were seen for 5432 person-years of follow-up. The median number of medical visits, which includes primary care and psychiatric visits, was 12 per year. An ART prescription (with at least 2 refills) was dispensed to 818 (48%) patients (Fig. 1). Among those receiving ART, the mean adherence rate was 72%, with 210 (26%) achieving optimal ART adherence >95%. Of the total group, 981 (57%) had a diagnosis of depression. An ADT prescription (with at least 2 refills) was dispensed to 450 (46%) of these patients (mean adherence rate was 79% and median adherence rate was 87%).

Flow diagram for antiretroviral therapy (ART) adherence analyses among HIV-infected patients, Denver Health, 1997 through 2001. ADT indicates antidepressant treatment. Calculated ART adherence compared depressed versus nondepressed patients. Among depressed patients, comparison was made between (1) receipt versus no receipt of ADT, (2) ADT adherence versus ADT nonadherence, (3) ADT adherence versus untreated and ADT nonadherence, and (4) before and after ADT initiation stratified by ADT adherence status.

Demographic and clinical characteristics of HIV-infected patients by depression diagnosis are shown in Table 1. Compared with nondepressed persons, depressed patients tended to be older, to be white non-Hispanic, to have more than 12 medical and at least 2 psychiatric visits per year, to be drug and/or alcohol users, to be unemployed, to have a lower annual income, to have received ART, and to be less adherent to ART. Factors associated with ART adherence among depressed patients are described in Table 2. Compared with adherent patients, ART-nonadherent patients had more psychiatric visits per year and were more likely to be alcohol users, to be unemployed, and to have a lower annual income.

Characteristics of HIV-Infected Patients by Diagnosis of Depression, DH (1997-2001)
Demographic and Clinical Characteristics for Depressed, HIV-Infected Patients by ART Adherence Status, DH, (1997-2001)

To assess the effect of ADT prescription and adherence on ART adherence, we compared the proportion of depressed HIV-infected patients who were adherent to ART (ie, ≥95%) among those who were and were not prescribed ADT and, further, by ADT adherence status (greater or less than the mean ADT adherence value) among those receiving any ADT (Table 3). The proportion of depressed patients adherent to ART was significantly lower among those not receiving ADT compared with those who received ADT (65% vs. 35%, respectively; P = 0.01). In addition, when ART adherence was stratified by ADT adherence status, the proportion of depressed patients adherent to ART was significantly higher among ADT-adherent patients compared with ADT-nonadherent patients (69% vs. 31%, respectively; P = 0.001).

ART Adherence by ADT Receipt Adherence for Depressed HIV-Infected Patients, DH (1997-2001)

To understand the potential temporal effect of ADT receipt and adherence on ART adherence, we compared ART adherence rates for a 6-month period before and after ADT initiation (Table 4). Comparing the 6-months before and after ADT initiation, ART adherence for all 3 groups (ADT-adherent, ADT-nonadherent, and no ADT received), significantly increased (ADT-adherent and ADT-nonadherent, P < 0.0001; no ADT received, P < 0.0003). When the pre-ADT versus post-ADT initiation mean differences in ART adherence were compared between the 3 groups (ADT-adherent, ADT-nonadherent, and not on ADT), statistically significant (P < 0.005) differences were observed in mean change in ART adherence among those not receiving ADT (0.10) compared with ADT-adherent patients (0.42) and ADT-nonadherent patients (0.37); no differences were observed between the ADT-adherent and ADT-nonadherent groups.

ART Adherence Stratified by Pre- and Post-ADT Time Frame for ADT-Adherent and ADT-Nonadherent Patients for Depressed HIV-Infected Patients, DH (1997-2001)

Table 5 shows multivariate analysis results for the effect of ADT adherence and other selected demographic and socioeconomic variables on ART adherence. After controlling for other variables, ART nonadherence was significantly more likely to be found in patients nonadherent to ADT (P = 0.0019) and in alcohol users (P = 0.01).

Adjusted Odds Ratios and 95% Confidence Intervals for Nonadherence to ART by Selected Demographic and Clinical Characteristics for HIV-Infected patients, DH (1997-2001)


A high prevalence rate of depression was observed among HIV-infected individuals, for whom ART adherence improved after prescription of ADT and reached even higher levels among individuals who were adherent to ADT. Although depression was significantly more common in white and slightly older individuals, this analysis focused on adherence rather than on the determinants of depression prevalence. Neither race nor age was associated with adherence in the multivariate analysis. Our analysis supports the importance of routine assessment of depression, prompt initiation of ADT when indicated, and monitoring of ADT adherence status. Those individuals adherent to ADT were also more adherent to ART. Although suggestive of a positive effect of ADT toward improved ART adherence, this might reflect an individual's intrinsic adherence behavior to any regimen.

For depressed HIV-infected individuals, a temporal association with higher ART adherence in the 6-month period after ADT initiation (compared with a similar period preceding ADT initiation) was observed for ADT-adherent and ADT-nonadherent patients. The temporal trend for ART adherence among depressed patients who did not receive ADT, using a pseudoinitiation date, showed improvement over time. Although the mean ART adherence improved over time for all 3 groups (ADT-adherent, ADT-nonadherent, and no receipt of ADT), comparison of mean differences (pre-ADT vs. post-ADT initiation) in ART adherence demonstrated a lower mean ART adherence increase among those who never received ADT compared with those who had. Other studies suggest that adherence rates decrease with longer follow-up periods (eg, 1-2 years).47,48 This study was limited to a shorter before and after comparison, given the relative limited time frame before ADT initiation in some HIV-infected patients.

A major strength of this study was the ability to combine a variety of data (ie, chart abstraction, diagnostic coding, utilization [visit and pharmacy]) from multiple electronic information systems. The pharmacy system permitted calculation of ART and ADT adherence measures and assessed the temporal relation of ART adherence based on initiation of and adherence to ADT. Adherence measures for ART are complex, given the variability of regimens and the frequency of changes over time. The calculated observation period for ART (denominator) summed all periods (from the first to last dispensing date) for each drug (ie, nucleoside analogue, protease inhibitor, nonnucleoside reverse transcriptase inhibitor) and thus accounted for the fact that not all antiretroviral drugs were prescribed simultaneously.

Prior studies suggest that those with high ART adherence rates (ie, ≥95%) are more likely to show undetectable viral loads.12,15 Among the 818 patients who received any ART in this study, only 26% reached this high level of adherence. We used this optimal ART adherence cutoff for the analysis, although a lower mean ART adherence rate was observed, consistent with other published observations from clinical settings (62%-80%).16-18 Although bivariate analysis suggested decreased ART adherence with concomitant depression, this variable was excluded from multivariate analysis as a result of limited dispersion, because the analysis sample was restricted to depressed patients. Previous reports have mixed results showing no association between depression and drug adherence,49 with others studies suggesting such an association.12,39,40 Methodologic differences may be important, because depression was measured by researchers rather than established clinically by health care providers and there was no way to measure the impact of ADT among untreated subjects.40 This study attempted to identify previously diagnosed depression through medical review charts, ICD-9-coded administrative data, or receipt of ADT dispensed by the pharmacy. Among depressed patients on ART, 72% received ADT, potentially attenuating depression-associated ART nonadherence. After controlling for other factors, ADT-adherent patients demonstrated improved ART adherence.

Although our data suggest that ADT adherence is associated with ART adherence, there is no measured evidence that depression improved with ADT adherence. Information used in the analysis, derived from various data sources, lacked any clinical outcome assessment as a result of a specific ADT therapy. This is clearly a limitation of real-world clinical studies in which specific diagnostic or research tools for depression are not routinely implemented. Alternative hypotheses may explain the improved ART adherence over time, such as increased interaction over time with providers (medical and psychiatric) who routinely encourage patients to adhere to all regimens. Others50 have suggested improved mental health status among HIV-infected patients receiving highly active antiretroviral therapy (HAART) through the promise of extended survival and better quality of life, because HAART therapy decreases opportunistic infections and other HIV symptoms.

More than a quarter of the depressed patients did not receive any ADT. Some of these individuals may have been prescribed medications but never filled those prescriptions at the pharmacy. Others identified as depressed by their primary care providers may not have met full Major Depressive Disorder criteria requiring ADT. Some “depressed” patients with bipolar depression, although included in the analysis, would have routinely received only mood-stabilizing medication.

Assessing medication adherence is complex and carries its own limitations. Medication adherence calculated from pharmacy records may not completely reflect patient adherence, because the quantity of medication dispensed is only a surrogate for actual drug use. Prior studies found significant associations between pharmacy refills and other adherence measures as well as measures of drug presence (serum drug levels) or physiologic drug effects,45,51,52 however, suggesting that prescription refill is a good proxy indicator of drug adherence. Because patients may obtain drug refills before depleting their supply, adherence rates are best determined across several refills, at least over periods longer than 60 days,14 consistent with the calculated time frames in this study.

The study suggests that ART adherence among HIV-infected depressed patients may be increased by (1) enhanced identification of depression, (2) ADT initiation, and (3) improved ADT adherence. At the provider level, failure to diagnose depressed patients is a hindrance to adequate treatment. Diagnosis of depression may be difficult among HIV-infected patients, suggesting the need for health care provider education and more active screening for depression among all HIV-infected patients, particularly among ART-nonadherent individuals. HIV primary care providers need to be aware of the range of effective ADT available as well as the importance of prompt initiation of ADT. Mental health evaluation should be an integral health care component of all HIV-infected patients receiving medical care. Improved depression diagnosis and management could be accomplished by implementation of educational programs directed to primary care providers; such training should be part of every clinical setting that provides medical care to HIV-infected patients.

Patient level interventions include culturally and educationally53 appropriate materials and programs with a focus on depression and its treatment. Measures tailored to different disease and treatment stages include increased frequency of health care provider contact and support during acute treatment, regular monitoring during ongoing treatment, and establishment of a long-term relationship with those patients who have a history suggesting vulnerability to relapse.54 These efforts to improve ADT adherence should enhance HIV management, including improved ART adherence.

At a systems level, improved patient adherence depends on adequate feedback mechanisms. Clinicians need rapid near-real-time methods of assessing patient-specific adherence rates for ART and ADT to identify those patients with whom further intervention is necessary. These methods of provider feedback have previously been shown to be effective across a spectrum of clinical activities.55-57 Reporting features of electronic pharmacy data systems useful for tracking dispensed medications should be enhanced so that providers and patients receive real-time information to assist with treatment adherence efforts.

This retrospective study demonstrated a potential benefit of ADT on ART adherence among depressed HIV-infected patients. To corroborate these important findings, prospective studies are needed to assess the accuracy of diagnosis of depression, the correlation between adherence to ADT and clinical improvement of depression, and the effect of the improved clinical depression on ART adherence.


The authors acknowledge John Frank for data collection assistance, Paula Breese for analytic support, and Paul Anderson for provision of data.


1. Haynes R. Compliance in Health Care. Baltimore: Johns Hopkins University Press; 1979.
2. Sacket D. The magnitude of compliance and noncompliance. In: Sacket DL, Haynes RB, eds. Compliance with Therapeutic Regimens. Baltimore: Johns Hopkins University Press; 1976:11-27.
3. Ickovics J, Meisler A. Adherence in AIDS clinical trials: a framework for clinical research and clinical care. J Clin Epidemiol. 1997;50:385-391.
4. Williams A, Friedland G. Adherence, compliance and HAART. AIDS Care. 1997;9:51-58.
5. Besch C. Compliance in clinical trials. AIDS. 1995;9:1-10.
6. Eraker S, Kirscht J, Becker M. Understanding and improving patient compliance. Ann Intern Med. 1984;100:258-268.
7. Stewart R, Cluff L. A review of medication errors and compliance in ambulant patients. Clin Pharmacol Ther. 1972;13:463-468.
8. Sacket D. A critical review of the determinants of patient adherence. In: Sacket DL, Haynes RB, eds. Compliance with Therapeutic Regimens. Baltimore: Johns Hopkins University Press; 1976:26-39.
9. Conway S, Pond M, Hamnett T, et al. Compliance with treatment in adult patients with cystic fibrosis. Thorax. 1996;51:29-33.
10. Epstein L, Cluss P. Behavioral medicine perspective on adherence to long-term medical regimens. J Consult Clin Psychol. 1982;50:950-971.
11. Macilwain C. Better adherence vital in AIDS therapies. Nature. 1997;390:326.
12. Paterson D, 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.
13. Hecht F, Chesney M. Adherence to HIV therapy. In: Sande M, Volberding P, eds. The Medical Management of AIDS. Philadelphia: WB Saunders; 1999:117-120.
14. Christensen D, Williams B, Goldberg HI, et al. Assessing compliance to anti-hypertensive medications using computer-based pharmacy records. Med Care. 1997;35:1164-1170.
15. Shelton M, Esch L, Hewitt R, et al. The impact of patient-reported adherence with antiretroviral therapy on virologic response. Paper presented at the 38th Meeting of the Interscience Annual Conference: September 24-27, 1998; San Diego, CA.
16. Bangsberg D, Hecht F, Charlebois E, et al. Adherence to protease inhibitor, HIV-1 viral load, and development of drug resistance in an indigent population. AIDS. 2000;14:357-366.
17. Bozek P, Purdue B, Forest-Smith M. Maintaining protease inhibitor therapy in a clinic setting. Paper presented at the 5th Conference in Retroviruses and Opportunistic Infections: February 1-5, 1998; Chicago, IL.
18. Hecht F, Colfax G, Swanson M, et al. Adherence and effectiveness of protease inhibitors in clinical practice. Paper presented at the 5th Conference in Retroviruses and Opportunistic Infections: February 1-5, 1998; Chicago, IL.
19. Gordillo V, Amo J, Soriano V, et al. Sociodemographic and psychological variables influencing adherence to antiretroviral therapy. AIDS. 1999;13:1763-1769.
20. Wells KB, Golding JM, Burnam MA. Psychiatric disorder in a sample of the general population with and without chronic medical conditions. Am J Psychiatry. 1988;145:976-981.
21. Chuang H. Psychiatric morbidity in patients with HIV infection. Can J Psychiatry. 1992;37:109-115.
22. Lyketsos C, Fishman M, Hutton H, et al. The effectiveness of psychiatric treatment for HIV-infected patients. Psychosomatics. 1997;38:423-432.
23. McDaniel J. An assessment of rates of psychiatric morbidity and functioning in HIV disease. Gen Hosp Psychiatry. 1995;17:346-352.
24. Brown G. Prevalence of psychiatric disorders in early stages of HIV infection. Psychosom Med. 1992;54:558-601.
25. Morris R, Schaerf F, Brandt J, et al. AIDS and multiple sclerosis: neural and mental features. Acta Psychiatr Scand. 1992;85:331-336.
26. Brown G, Rundell J. A prospective study of psychiatric aspects of early HIV disease in women. Gen Hosp Psychiatry. 1993;15:139-147.
27. Gala C, Pergami A, Catalan J. The psychosocial impact of HIV infection in gay men, drug users, and heterosexuals: a controlled investigation. Br J Psychiatry. 1993;163:651-659.
28. Maj M, Jansen R, Starace F. WHO neuropsychiatric AIDS study, cross-sectional phase I. Study design and psychiatric findings. Arch Gen Psychiatry. 1994;51:39-49.
29. Hinkin C, van Gorp W, Satz P, et al. Depressed mood and its relationship to neuropsychological rest performance in HIV-1 seropositive individuals. J Clin Exp Neuropsychol. 1992;14:289-297.
30. Atkinson JH, Jr, Grant I, Kennedy CJ, et al. Prevalence of psychiatric disorders among men infected with human immunodeficiency virus. A controlled study. Arch Gen Psychiatry. 1988;45:859-864.
31. Perkins D, Stern R, Golden RN, et al. Mood disorders in HIV infection: prevalence and risk factors in a nonepicenter of the AIDS epidemic. Am J Psychiatry. 1994;151:233-236.
32. Williams J, Rabkin J, Remien R. Multi-disciplinary baseline assessment of homosexual men with and without human immunodeficiency virus infection II: standardized assessment of current and lifetime psychopathology. Arch Gen Psychiatry. 1991;48:124-130.
33. Treisman G, Lyketsos CG, Fishman M, et al. Psychiatric care for patients with HIV infection. The varying perspectives. Psychosomatics. 1993;34:432-439.
34. Rosenberg P, Bornstein R. Psychopathology in human immunodeficiency virus infection. Compr Psychiatry. 1993;34:150-158.
35. Kelly B. Psychiatric disorder in HIV infection. Aust NZ J Psychiatry. 1998;32:441-453.
36. King V. Influence of psychiatric comorbidity on HIV risk behaviors: changes during drug abuse treatment. J Addict Dis. 2000;19:65-83.
37. Blackwell B. From adherence to alliance-a quarter century of research. Neth J Med. 1996;48:140-149.
38. Singh N, Squier C, Sivek C, et al. Determinants of compliance with antiretroviral therapy in patients with human immunodeficiency virus: prospective assessment with implications for enhancing compliance. AIDS Care. 1996;8:261-269.
39. Catz S, Kelly J, Bogart L, et al. Patterns, correlates, and barriers to medication adherence among persons prescribed new treatments for HIV disease. Health Psychol. 2000;19:124-133.
40. DiMatteo M, Lepper H, Croghan T. Depression is a risk factor for noncompliance with medical treatment: meta-analysis of the effects of anxiety and depression on patient adherence. Arch Intern Med. 2000;160:2101-2107.
41. Tucker JS, Burnam MA, Sherbourne CD, et al. Substance use and mental health correlates of nonadherence to antiretroviral medications in a sample of patients with human immunodeficiency virus infection. Am J Med. 2003;114:573-580.
42. Michaels R, Marzuk P. Progress in psychiatry: part II. N Engl J Med. 1993;329:628-638.
43. Farizo K, Buehler JW, Chamberland ME, et al. Spectrum of disease in persons with human immunodeficiency virus infection in the United States. JAMA. 1992;267:1798-1805.
44. Buehler JW, Diaz T, Hersh BS, et al. The supplement to HIV-AIDS surveillance project: an approach for monitoring HIV risk behaviors. Public Health Rep. 1996;111(Suppl 1):134-137.
45. Steiner J, Prochazka A. The assessment of refill compliance using pharmacy records: methods, validity, and applications. J Clin Epidemiol. 1997;50:105-116.
46. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington, DC: American Psychiatric Association; 1994.
47. Enlund H. Measuring patient compliance in antihypertensive therapy-some methodological methods. J Clin Hosp Pharm. 1982;7:43-51.
48. Ostrop N, Gill M. Antiretroviral medication adherence and persistence with respect to adherence tool usage. AIDS Patient Care STDS. 2000;14:351-358.
49. Conn V, Taylor S, Miller R. Cognitive impairment and medication adherence. J Gerontol Nurs. 1994;20:41-47.
50. Chan KS, Orlando M, Joyce C, et al. Combination antiretroviral therapy and improvements in mental health: results from a nationally representative sample of persons undergoing care for HIV in the United States. J Acquir Immune Defic Syndr. 2003;33:104-111.
51. Inui TS, Carter WB, Pecoraro RE, et al. Variations in patient compliance with common long-term drugs. Med Care. 1980;18:10.
52. Choo P, Rand C, Inui T, et al. Validation of patient reports, automated pharmacy records, and pill counts with electronic monitoring of adherence to anti-hypertensive therapy. Med Care. 1999;37:846-857.
53. Kalichman S, Ramachandran B, Catz S. Adherence to combination antiretroviral therapies in HIV patients of low health literacy. J Gen Intern Med. 1999;14:267-273.
54. Frank E. Enhancing patient outcomes: treatment adherence. J Clin Psychiatry. 1997;58:11-14.
55. McDonald C, Hui S, Smith D, et al. Reminders to physicians from an introspective computer medical record. Ann Intern Med. 1984;100:130-138.
56. Tierney W, Miller M, McDonald C. The effect on test ordering of informing physicians of the charges for outpatient diagnostic tests. N Engl J Med. 1990;322:1499-1504.
57. LeBaron C, Chaney M, Baughman A, et al. Impact of measurement and feedback on vaccination coverage in public clinics, 1988-1994. JAMA. 1997;277:631-635.

adherence; compliance; depression; antiretroviral therapy; antiretroviral adherence; antidepressant therapy adherence; HIV infection

© 2005 Lippincott Williams & Wilkins, Inc.