Objective: To determine if HIV/AIDS-specific quality of life (QOL) predicts adherence to antiretroviral therapy (ART).
Methods: HIV-infected outpatients on efavirenz plus 2 or 3 nucleoside analogue reverse transcriptase inhibitors and with HIV viral loads <75 copies/mL were followed until the censoring event of detectable viremia or 1 year of follow-up. QOL was assessed at baseline with the HIV/AIDS-Targeted Quality of Life instrument (HAT-QoL), as were depression symptoms, stress levels, social support, and substance use. Follow-up high (≥95%) versus low (<95%) adherence was measured for 90 days before the censoring event.
Results: Fifty-six (48%) of 116 recruited participants had low adherence. Baseline financial worries (from the HAT-QoL) were greater in those with low versus high adherence (P = 0.02). Those with low versus high adherence also were more likely to use alcohol (P = 0.01) and other drugs (P = 0.02) currently at baseline. Regression analysis led to a model that included only current alcohol use (odds ratio [OR] = 2.65, 95% confidence interval [CI]: 1.20 to 5.87)) and financial worries (OR = 1.16, 95% CI: 1.03 to 1.310, for each 10-unit rise).
Conclusions: Baseline financial worries were associated with antiretroviral adherence later in time. Questions about paying bills and financial ability to care for oneself may be clinically useful in identifying patients who will have suboptimal adherence.
From the *Center for Health Equity Research and Promotion, Philadelphia Veterans Affairs Medical Center, Philadelphia, PA; †Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, Philadelphia, PA; ‡Division of General Internal Medicine, Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA; and the §Division of Infectious Diseases, Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA.
Received for publication April 14, 2007; accepted August 1, 2007.
Research support provided by National Institutes of Health through the University of Pennsylvania Center for AIDS Research Clinical Core (P30-AI45008) and an Agency for Healthcare Research and Quality (AHRQ) Centers for Education and Research on Therapeutics cooperative agreement (HS10399). Research activities also were supported, as was R. Gross, by a National Institute of Mental Health Career Development Award (K08MH01584) and by a contract with Bristol-Myers Squibb and a Young Investigator award from GlaxoSmithKline. W. C. Holmes was supported by a Research Career Development award from the Health Services Research and Development Service of the Department of Veterans Affairs (RCD 03-029).
The sponsors of this study had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.
Correspondence to: William C. Holmes, MD, MSCE, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, 733 Blockley Hall, 423 Guardian Drive, Philadelphia, PA 19104-6021 (e-mail: email@example.com).
Antiretroviral therapy (ART) adherence is a major determinant in the development of HIV resistance and virologic failure, AIDS-defining events, and death.1-5 Identifying predictors of ART adherence has been a research priority, and numerous studies have assessed a myriad of candidate variables. Few of these have considered quality of life (QOL). Most of those that have were cross sectional in design, and thus unable to identify predictors of future adherence.6-11 These studies have suggested, however, that a number of QOL dimensions-most all from disease-specific instruments-possibly may predict adherence. These include the Cherishing the Environment dimension from the Living with HIV instrument;6 the Current Health Perceptions dimension from the Medical Outcomes Study instrument;9 and the Financial Status, Cognitive Functioning, and Medical Care10 dimensions from the Multidimensional Quality of Life-HIV instrument, respectively.
Only 2 longitudinal studies have assessed the relation between QOL measured at a single point in time and ART adherence subsequent to that time.12,13 Neither revealed an association. Both of these studies used generic QOL instruments, however, and the lack of an identified association between QOL and adherence over time, although it could indicate that QOL does not predict future adherence, may have been attributable to not having used a QOL measure more relevant to HIV/AIDS and/or attributable to other features common to generic instruments (when used in HIV/AIDS populations) such as floor and ceiling effects.14
We pursued these uncertainties surrounding whether QOL predicts future ART adherence by using a disease-specific measure-the HIV/AIDS-Targeted Quality of Life instrument (HAT-QoL)-to assess the association between baseline QOL and subsequent adherence. The HAT-QoL was developed using grounded theory and including content provided solely by people living with HIV/AIDS.14-16
Participants were enrolled in a prospective cohort study of HIV-infected individuals on ART, the primary goal of which was to determine the relation between adherence and virologic breakthrough.17 Individuals ≥18 years old with viral loads <75 copies/mL on efavirenz-based regimens containing 2 or 3 nucleoside or nucleotide reverse transcriptase inhibitors but no protease inhibitors or other nonnucleoside reverse transcriptase inhibitors were eligible. Participants had to be on their regimen for at least 3 months.
Between January 1999 and May 2004, those meeting study criteria were offered participation at 4 HIV clinics affiliated with the University of Pennsylvania and at 1 HIV clinic each at Hahnemann University Hospital and Temple University Hospital. Participants also had to be living outside of a setting that provided medications (eg, nursing home) but could have a partner who assisted them in taking medications. Participants were followed up to a virologic load of ≥1000 copies/mL or 1 year, whichever came first. All study procedures were approved by the University of Pennsylvania and Philadelphia Veterans Affairs Medical Center Committees on Studies Involving Human Beings; all participants provided written informed consent.
Enrollment Visit Procedures
Routine information about sociodemographics, HIV/AIDS history, and QOL were obtained verbally at the enrollment visit by the study coordinator, who was not a member of the clinical care team. The HAT-QoL captured information on overall function (using 6 items), sexual function (2 items), health worries (4 items), medication worries (5 items), disclosure worries (5 items), financial worries (3 items), HIV mastery (2 items), life satisfaction (4 items), and provider trust (3 items) for the previous 4 weeks.14-16 The same 5-point Likert-style response option (“all of the time,” “a lot of the time,” “some of the time,” “a little of the time,” or “none of the time”) is used for all items. Each dimension score was transformed to a scale from 0 to 100, with higher scores indicating better dimension-specific QOL.
Other variables possibly related to QOL, including depressive symptoms, stress, and social support, were assessed. Depressive symptoms were assessed using the 20-item Center for Epidemiologic Studies-Depression scale (CES-D), which uses a 4-point Likert-style response option (“never/rarely,” “sometimes,” “often,” or “mostly or always”).18 CES-D scores can range from 0 to 60, with a higher score indicating more depressive symptoms and/or more frequent depressive symptoms. Stress was assessed using the Perceived Stress Scale (PSS), which measures the degree to which situations in one's life are appraised as stressful and used a 5-point Likert-style response option (“never,” “almost never,” “sometimes,” “fairly often,” or “very often”).19 PSS scores can range from 10 to 50, with a higher score indicating higher stress. Social support was assessed using the 6-item version of the Interpersonal Support Evaluation List (ISEL), which taps the tangible, belonging, self-esteem, and appraisal dimensions of support and uses a 4-point Likert-style response option (“definitely false,” “probably false,” “probably true,” or “definitely true”).20 ISEL scores can range from 6 to 36, with a higher score indicating greater perceived support.
Follow-Up Visit Procedures
Participants were provided with a microelectronic processor and bottle (Medication Event Monitoring System [MEMS]; Aardex, Zug, Switzerland) at their baseline visit. At that time, participants were instructed to keep their efavirenz in the monitored bottle exclusively and only to open the bottle to take doses. They also were instructed to transfer new supplies of efavirenz, when a refill was obtained, into the monitored bottle at the routine time they were scheduled to take efavirenz. If they accidentally opened the bottle at other times (eg, to check to see how many pills were left), participants were instructed to record the date and time. If participants were instructed by a provider not to take their efavirenz (eg, surgery), they were instructed to record these days.
Participants were seen monthly; at each visit, blood was drawn, MEMS data downloaded, and medication diaries (eg, erroneous openings) obtained. Plasma viral loads were assayed every 3 months initially using the Quantiplex HIV-1 RNA 3.0 assay (Chiron Diagnostics, Emeryville, CA) and then, after November 2002, the Versant HIV-1 RNA 3.0 bDNA assay (Bayer Diagnostics, Berkeley, CA). Adherence was calculated for the 90-day period before censor date and expressed as percent of prescribed doses taken. Adherence was dichotomized as “high” and “low” using 95% as a cutoff.2
Comparisons of adherence subgroups were performed using 2-tailed t tests for nonskewed continuous variables, Mann-Whitney U tests for skewed continuous variables, and χ2 methods for categoric variables (or the Fisher exact test when the expected frequency for a subgroup was <5). All variables associated with adherence (using a P level of ≤0.10) were entered into backward stepwise logistic regression to identify a parsimonious model of predictors for adherence. Results were verified by assessing whether the same variables entered into forward stepwise regression yielded the same solution. SPSS 12.0 for Windows (SPSS, Chicago, IL) was used to manage and analyze data.
Of 116 participants recruited into the study, most were middle-aged men and African American (Table 1). Fewer than half were educated beyond high school, and fewer than one third were employed; most had an income below the federal poverty line. Nearly half of the participants in this cohort were men who have sex with men (MSM), and nearly one quarter had a history of injection drug use (IDU). The median length of known HIV infection was 5 years, and most had an AIDS diagnosis. Most were on their first antiretroviral (ARV) regimen.
Most HAT-QoL dimension scores indicated QOL to be high at baseline (Table 2), with two thirds of dimension scores being >75 (out of 100). Nearly all dimension scores were consistently higher than those reported for HAT-QoL developmental studies (completed at a time just after the beginning of the highly active antiretroviral therapy [HAART] era).14-16
Other baseline QOL-related findings indicated that all mean CES-D, PSS, and ISEL scores were at the lower end of the range of possible scores (see Table 2). For CES-D and PSS assessments, this was a favorable finding, indicating that depression symptoms were less than the threshold of 16 (the cutoff found to be correlated with a diagnosis of depression) and that stress was similarly low. Conversely, social support was, on average, low for the sample.
Only half (52%) of the cohort were fully adherent with their ART regimen. Adherence subgroups did not differ on the basis of any of the baseline sociodemographic and HIV/AIDS-related characteristics (see Table 1). More high- than low-adherence participants had reported no current alcohol (53% vs. 30%; P = 0.01) or other drug use (87% vs. 69%; P = 0.02) at baseline, however.
QOL and Adherence
Only the baseline assessments of the HAT-QoL health worries, medication worries, and financial worries dimensions were associated with adherence over time (and only assessment of financial worries was significant at P ≤ 0.05; see Table 2). Low-adherence participants were more worried about their health, medication, and finances than were high-adherence participants.
Baseline CES-D and PSS assessments also were associated with adherence over time, although both only met the less stringent P value criterion of ≤0.10. Depression symptoms and stress levels were greater in those with low versus high adherence.
When all variables associated with adherence in bivariable analyses were entered into backward stepwise logistic regression, only alcohol use in the past year and financial worries were associated with adherence (Table 3). The odds of future high adherence were found to increase 2.65 times for those who had had no alcohol in the past year. The odds of future high adherence also were found to increase 1.15 times for every 10-fold increase in the financial worries score (which translates into a decrease in worries related to finances). These 2 variables explained 14% of the variability in the adherence variable. Forward stepwise regression yielded the same solution.
Although this is the third longitudinal study to explore whether QOL measured at a single point in time predicts ART adherence subsequent to that time, it is the first to do so using HIV/AIDS-specific QOL assessment.12,13 Similar to the other 2 studies, most QOL dimensions did not predict adherence to ART over time. Unlike prior studies, the HAT-QoL financial worries dimension predicted adherence over time, and was 1 of only 2 variables to remain in a parsimonious model predicting adherence. The amount of variability in adherence behavior explained by the model was high considering the number of other variables known and/or believed to be associated with adherence. Alcohol use was the other variable in this model, which, given that alcohol use has been one of the most consistently identified predictors of ART adherence,21-28 imparts a degree of external validity to the model.
Prior studies have looked directly at the relation between income and adherence.2,12,13,29,30 Of these, only the longitudinal study by Kleeberger et al13 found income to be associated with (<100%) adherence in a multivariable analysis. These investigators stressed that their finding did not identify difficulties of individuals who had low incomes but rather those of individuals from the general population who had middle incomes.13 As such, they may have been identifying individuals who experience a mismatch between what their income provides them and what their expenditures require, particularly when the costs for medications are added to them. These individuals may be those who are not below the poverty line but those who could manage to afford their lives if it were not for out-of-the-ordinary costs of living with HIV/AIDS. We may have found the association we report because we captured just this construct regarding mismatch, which goes beyond what a quantitative assessment of monthly or yearly income can impart.
Parsing the content of items that make up the financial worries dimension of the HAT-QoL affirms this possible interpretation. The 3 items assess the frequency of recent (past 4 weeks) worries about “having to live on a fixed income,” about “how to pay my bills,” and that “money has been too tight for me to care for myself the way I think I should.” The latter 2 items clearly highlight the implications of income/expenditure mismatches. Even the income question suggests this as well, asking about a “fixed” income, which implicitly may be read (in the context of the other 2 items) as the likelihood that one is not going to be able to meet future increases in expenditures. The third item, in particular, suggests how these financial worries may be most likely to affect adherence-income/expenditure mismatches may lead one to limit self-care and/or receipt of care in an attempt to minimize expenditures. Doing so may lead to self-rationing of medications and medical care as was suggested when people living with HIV/AIDS who had competing subsistence needs were found to have poorer access to medical care.31
Four limitations warrant specific mention. First, we were unable to collect data on refusal rates, because providers did not refer potential participants to the study if patients were not interested. The demographic characteristics of the sample were representative of the urban Philadelphia HIV-seropositive population at the time of the study, however, providing some reassurance that the sample is representative of the population of inference.32 Second, most participants were on their first round of ART, and all were willing to participate in a research study of adherence; thus, generalizability of results may be limited to this subpopulation. Third, we measured adherence using the MEMS, and it is possible that some participants may not have used the MEMS for the entire period, leading to misclassification of our adherence variable and likely biasing all findings toward the null (thus, our findings may underestimate the actual relations between variables and adherence). Fourth, subgroup numbers were small, which may have limited the statistical power of analyses to identify independent effects of some factors previously identified as predictors of adherence (eg, drug use, social support, depression).9,21,23,25,33-38 Also, some factors previously identified as predictors of adherence were not assessed; thus, they were not included in analyses (eg, self-efficacy, coping strategies, hopefulness, posttraumatic stress disorder).34,39-42
Nonetheless, this study of whether baseline HIV/AIDS-specific QOL was associated with subsequent ART adherence indicated that financial worries were a strong and independent predictor. This suggests that asking patients about how frequently they experience worries about financial matters (eg, fixed income, paying bills, caring for themselves as they think they should), at initiation and periodically while on ART, might provide useful information about their likelihood for high-level adherence in the future. We do not suggest obtaining this information in an effort to identify individuals for whom clinicians should curtail prescribing ART. Rather, if confirmed, these findings should form the basis of a screening tool for identifying higher risk individuals. Further, financial worries should be addressed in behavioral interventions designed to improve adherence.
1. Gross R, Bilker WB, Friedman HM, et al. Effect of adherence to newly initiated antiretroviral therapy on plasma viral load. AIDS
2. Paterson DL, Swindells S, Mohr J, et al. Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med
3. Bangsberg DR, Hecht FM, Charlebois ED, et al. Adherence to protease inhibitors, HIV-1 viral load, and development of drug resistance in an indigent population. AIDS
4. Bangsberg DR, Perry S, Charlebois ED, et al. Non-adherence to highly active antiretroviral therapy predicts progression to AIDS. AIDS
5. Wood E, Hogg RS, Yip B, et al. 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
6. Holzemer WL, Corless IB, Nokes KM, et al. Predictors of self-reported adherence in persons living with HIV disease. AIDS Patient Care STDS
7. Murri R, Ammassari A, Gallicano K, et al. Patient-reported nonadherence to HAART is related to protease inhibitor levels. J Acquir Immune Defic Syndr
8. Gifford AL, Bormann JE, Shively MJ, et al. Predictors of self-reported adherence and plasma HIV concentrations in patients on multidrug antiretroviral regimens. J Acquir Immune Defic Syndr
9. Wilson TE, Barron Y, Cohen M, et al. Adherence to antiretroviral therapy and its association with sexual behavior in a national sample of women with human immunodeficiency virus. Clin Infect Dis
10. Carballo E, Cadarso-Suarez C, Carrera I, et al. Assessing relationships between health-related quality of life and adherence to antiretroviral therapy. Qual Life Res
11. Ganz PA, Schag CAC, Kahn B, et al. Describing the health-related quality of life impact of HIV infection: findings from a study using the HIV Overview of Problems-Evaluation System (HOPES). Qual Life Res
12. Singh N, Berman SM, Swindells S, et al. Adherence of human immunodeficiency virus-infected patients to antiretroviral therapy. Clin Infect Dis
13. 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
14. Holmes WC, Shea JA. Two approaches to measuring quality of life in the HIV/AIDS population: HAT-QoL and MOS-HIV. Qual Life Res
15. Holmes WC, Shea JA. Performance of a new, HIV/AIDS-targeted quality of life (HAT-QoL) instrument in asymptomatic seropositive individuals. Qual Life Res
16. Holmes WC, Shea JA. A new HIV/AIDS-targeted quality of life (HAT-QoL) instrument: development, reliability, and validity. Med Care
17. Gross R, Bilker WB, Friedman HM, et al. High level, but not low level HIV breakthrough on efavirenz-based regimens associated with poor adherence. Presented at: 22nd International Conference on Pharmacoepidemology and Therapeutic Risk Management, International Society for Pharmacoepidemiology; 2006; Lisbon.
18. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas
19. Cohen S, Kamarck T, Mermelstein R. A global measure of perceived stress. J Health Soc Behav
20. Cohen S, Mermelstein R, Kamarck T, et al. Measuring the functional components of social support. In: Sarason I, Sarason B, eds. Social Support: Theory, Research, and Applications
. Dodrecht, The Netherlands: Martinus Nijhoff Publishers; 1987:73-94.
21. Holstad MK, Pace JC, De AK, et al. Factors associated with adherence to antiretroviral therapy. J Assoc Nurses AIDS Care
22. Peretti-Watel P, Spire B, Lert F, et al. Drug use patterns and adherence to treatment among HIV-positive patients: evidence from a large sample of French outpatients (ANRS-EN12-VESPA 2003). Drug Alcohol Depend
. 2006;82(Suppl 1):S71-S79.
23. Murphy DA, Marelich WD, Hoffman D, et al. Predictors of antiretroviral adherence. AIDS Care
24. Heckman BD, Catz SL, Heckman TG, et al. Adherence to antiretroviral therapy in rural persons living with HIV disease in the United States. AIDS Care
25. Mellins CA, Kang E, Leu C, et al. Longitudinal study of mental health and psychosocial predictors of medical treatment adherence in mothers living with HIV disease. AIDS Patient Care STDS
26. Howard AA, Arnsten JH, Lo Y, et al. A prospective study of adherence and viral load in a large multi-center cohort of HIV-infected women. AIDS
27. Golin CE, Liu H, Hays RD, et al. A prospective study of predictors of adherence to combination antiretroviral medication. J Gen Intern Med
28. Mathews WC, Mar-Tang M, Ballard C, et al. Prevalence, predictors, and outcomes of early adherence after starting or changing antiretroviral therapy. AIDS Patient Care STDS
29. Eldred LJ, Wu AW, Chaisson RE, et al. Adherence to antiretroviral and Pneumocystis
prophylaxis in HIV disease. J Acquir Immune Defic Syndr
30. Gao X, Nau DP, Rosenbluth SA, et al. The relationship of disease severity, health beliefs and medication adherence among HIV patients. AIDS Care
31. Cunningham WE, Andersen RM, Katz MH, et al. The impact of competing subsistence needs and barriers on access to medical care for persons with human immunodeficiency virus receiving care in the United States. Med Care
32. City of Philadelphia Department of Public Health. Philadelphia AIDS Statistical Update
. Philadelphia, PA: AIDS Activities Coordinating Office; December 2005.
33. Williams PL, Storm D, Montepiedra G, et al. Predictors of adherence to antiretroviral medications in children and adolescents with HIV infection. Pediatrics
34. Weaver KE, Llabre MM, Duran RE, et al. A stress and coping model of medication adherence and viral load in HIV-positive men and women on highly active antiretroviral therapy (HAART). Health Psychol
35. Kleeberger CA, Buechner J, Palella F, et al. Changes in adherence to highly active antiretroviral therapy medications in the Multicenter AIDS Cohort Study. AIDS
36. Hinkin CH, Hardy DJ, Mason KI, et al. Medication adherence in HIV-infected adults: effect of patient age, cognitive status, and substance abuse. AIDS
. 2004;18(Suppl 1):S19-S25.
37. Ingersoll K. The impact of psychiatric symptoms, drug use, and medication regimen on non-adherence to HIV treatment. AIDS Care
38. Gordillo V, del Amo J, Soriano V, et al. Sociodemographic and psychological variables influencing adherence to antiretroviral therapy. AIDS
39. Nilsson-Schonnesson L, Diamond PM, Ross MW, et al. Baseline predictors of three types of antiretroviral therapy (ART) adherence: a 2-year follow-up. AIDS Care
40. Godin G, Cote J, Naccache H, et al. Prediction of adherence to antiretroviral therapy: a one-year longitudinal study. AIDS Care
41. Wilson KJ, Doxanakis A, Fairley CK. Predictors for non-adherence to antiretroviral therapy. Sex Health
42. van Servellen G, Chang B, Garcia L, et al. Individual and system level factors associated with treatment nonadherence in human immunodeficiency virus-infected men and women. AIDS Patient Care STDS