Highly active antiretroviral therapy (HAART) use has significantly remitted or slowed HIV disease progression to AIDS and death by dramatically containing virus replication1,2 and has been demonstrated to be very cost-effective by lowering hospitalization days and returning patients to workforce.3,4 The clinical success of HAART has turned HIV/AIDS in developed countries into, in most cases, a manageable chronic infectious disease requiring long-term therapy. However, such success was not achieved without a price in terms of quality of life (QOL) because of complex regimens and adverse effects associated with antiretroviral drugs.5 Thus, with the availability of more efficient and less toxic regimens, maximizing quality-adjusted survival becomes a high priority of effective management of HIV disease in the HAART era.6 To reach this goal, major modifiable predictors of lower QOL need to be first identified.
In describing how different factors affect quality of life, different conceptual models have been proposed.7-9 For example, Wilson et al suggested a model to specify relationships among 5 levels of health outcomes: biological and physiological factors, symptoms, functioning, general health perceptions, and overall quality of life, with other variables such as personal and environmental characteristics interacting with them.9 Although these conceptual models do not detail every causal relationship, they do provide a basis for establishing a predictive model to assess the relative importance of factors associated with QOL.
The objective of this study was to identify the predictors of lower QOL and to help assess important and alterable factors for possible interventions to maximize patients' QOL among HAART-using study participants in the Multicenter AIDS Cohort Study (MACS), a well-characterized cohort with standardized follow-up.
The MACS is a prospective study of the natural and treated history of HIV disease among 6973 homosexual and bisexual men in the United States. A detailed description of MACS has been published elsewhere,2,10 and only methods relevant to the current study are presented. Briefly, participants at the 4 sites-Baltimore, Chicago, Los Angeles, and Pittsburgh-were followed every 6 months with detailed questionnaire-based interviews, physical examinations, and medical history review. Blood samples were collected for concomitant laboratory measurements and storage in both local and national repositories. Enzyme-linked immunoabsorbent assays with confirmatory Western blot tests were used to determine HIV-1 seropositivity. The institutional review board at each site approved the MACS study protocol and informed consent was obtained from each participant. For this study, we restricted our analyses to a nested cohort of 636 participants who enrolled before year 2001 (5622 in total), initiated HAART during the follow-up (744 in total), and had at least one QOL measurement before HAART initiation and at least 2 consecutive QOL measurements after HAART initiation.
Quality of Life
Health-related quality of life measurement was incorporated into the MACS protocol starting with visit 21 (October 1994) using the standard SF-36 Form (SF-36). SF-36 was selected because it is one of the most widely used generic QOL instruments and the QOL findings from this cohort can be compared with its established benchmarks from the normal populations and to results from other disease populations. This form contains 36 items and addresses 8 QOL subscales: general health perceptions, physical functioning, role limitations by physical problems, pain, social functioning, energy/fatigue, emotional well-being, and role limitation by emotional problems. The subscale scores were transformed linearly to a possible range of 0 to 100 according to a standardized algorithm,11 with higher values indicating better functioning and well-being. In addition, physical and mental health summary scores were constructed to represent these 8 subscales according to published scoring recommendations.12 For easier result interpretation and simplified statistical analysis, the 2 QOL summary scores after HAART initiation were the outcomes in this analysis.
According to the conceptual models, prior studies, and available data, individual characteristics (sociodemographics and individual risk behaviors), social support, and clinically related indicators (biological markers, HIV-related medication use, and clinical outcome variables) were assessed as possible predictors of lower QOL in this analysis.
Level of education completed at the time of study entry was recoded as no college, some college, and college and above. Self-reported race/ethnicity was categorized as white/non-Hispanic, black/non-Hispanic, Hispanic, or other. The annual gross income in the past year ascertained at each visit was dichotomized as less than $20,000 or not, as this cutoff statistically differentiated the QOL scores in a previous study from MACS.13 Employment status at each visit was defined as currently employed or not.
Individual Risk Behaviors
The number of male sexual partners in the past 6 months was categorized as none, no more than 5, and 6 and above. Smoking status at each visit was coded as never smoked, past smoker, and current smoker. The average drinking pattern in the past 6 months was recoded as none, light (<3 drinks per week), moderate (3-13 drinks per week) and heavy (≥14 drinks per week). In addition, the self-reported recreational drug use at each visit was dichotomized as use or nonuse of any specific drug. HAART adherence was defined as 100% adherence to current medications over the past 4 days.14 HAART interruption or drug holiday was defined as missing at least 2 consecutive days of all antiretroviral drugs in the past 6 months, and HAART discontinuation was termed as not using HAART at current visit.15
Number of persons around to talk to from the self-administered questionnaire was selected as a representative indicator of social support, and it was categorized into 5 subgroups: none, 1 only, 2 to 3, 4 to 5, 6 and more.
Plasma HIV RNA levels were determined using a polymerase chain reaction assay with a lower limit of detection of 50 copies/mL (Roche Diagnostics, Nutley, NJ). The HIV RNA level at each visit was dichotomized as detectable or not. CD4+ lymphocyte count was measured using the standardized flow cytometry performed at each visit.16,17
Information about the use of antiretroviral drugs and medications for prophylaxis and treatment of opportunistic infection (OI) were obtained by face-to-face interviews at each visit. HAART was defined following the 2002 Department of Health and Human Sciences/Kaiser Panel guidelines.18 The peak incidence of HAART initiation in the cohort occurred in 1997.19
Clinical Outcome Indicators
Number of HIV-related symptoms, outpatient visits, hospitalizations and non-AIDS comorbidities, history of AIDS diagnosis, and depression were used as indicators for disease severity. At each visit, participants were asked if they had certain symptoms since their last visit. The questions for symptoms were updated periodically, and only symptoms with sufficient data were counted for this analysis, including persistent fatigue, skin rash, diarrhea, persistent fever, headache, night sweats, enlarged gland, unusual bruise, and unintentional weight loss of more than 10 pounds. AIDS was defined according to a modified version of the 1993 Centers for Disease Control and Prevention criteria by excluding any AIDS individuals who met only the immunological definition for AIDS (ie, CD4+ lymphocyte count less than 200/mm3).20 Non-AIDS defining comorbidities were also reported in MACS and their number was counted at each visit. The Center for Epidemiological Studies Depression Scale21 was used to assess depressive symptoms at each visit and a cutoff value of 16 was used to define depression.21 At each visit, participants were also asked how frequent they use health-care services since their last visit.
To control for the heterogeneity of starting QOL, baseline QOL was determined using data from the 2 visits immediately before HAART initiation. If 2 QOL measurements were available, their average was used. If only 1 QOL measurement existed, the available one was applied. To avoid temporal ambiguity in the predictive models, factors at or since the (i−1)th visit (Vi−1) were assessed as predictors for QOL at the ith visit (Vi). Because relatively few participants died during the follow-up and informative dropout due to death was minor, we used a standard traditional model rather than a complicated joint modeling technique. Time since HAART initiation was not included as a predictor in this analysis because it mainly reflects aging effect after controlling other time-dependent variables. The associations of sociodemographics, individual risk behaviors, social support, biological markers, HIV-related medication use, and clinical outcome indicators with QOL were assessed first univariately using a random-effects linear model to control for within-individual correlation. To further evaluate the independent associations of the variables in predicting physical health summary score (PHS) or mental health summary score (MHS) after controlling for confounding effects, the predictors significantly (P < 0.05) related to either PHS or MHS in the univariate analyses, together with the baseline PHS or MHS before HAART initiation, were included in the corresponding random-effects multivariate models. Only predictors significantly associated with study outcomes were maintained in the final parsimonious multivariate models. As the predictors were selected from the conceptual models and prior studies, no correction for multiple comparisons was made. All analyses were carried out using SAS version 9.0 (SAS Institute, Cary, NC).
The 636 MACS HAART-using participants contributed 5285 person visits after HAART initiation through visit 40 (April 2004). The median (interquartile range: 25%, 75%) age of the subjects at the first visit of HAART use was 43.3 (39.0, 48.2) years. Among the participants, 84% were white, non-Hispanic and about 88% had at least some college education at study entry. Before HAART initiation, 133 of the study subjects were diagnosed with AIDS, and the unadjusted means of PHS and MHS were 48.8 and 47.7, respectively. The median (interquartile range) time from HAART initiation to last follow-up visit was 5 (4.5-7.0) years. During the follow-up, 66 participants were diagnosed with AIDS and 70 died (42 of them AIDS-related). The PHS showed a significant decrease trend over time [beta = −0.12 per year; 95% confidence interval (CI): −0.20, −0.04; P = 0.006], whereas MHS did not (beta = −0.09 per year; 95% CI: −0.20, 0.02; P = 0.122).
The significant univariate associations of the individual characteristics (sociodemographics and individual risk behaviors), social support, and clinical-related indicators (biological markers, HIV-related medication use, and clinical outcomes variables) with either PHS or MHS are shown in Tables 1 and 2, respectively. Unemployment, lower income, smoking, having no male sexual partners, not drinking alcohol; missing at least 2 consecutive days of antiretroviral drugs (or drug holidays), having more symptoms, outpatient visits, and non-AIDS comorbidities, and depression in the past 6 months were associated with lower PHS and MHS. No overall differences in QOL were observed for different education levels, race/ethnicity groups, and lower adherence to and discontinuation of HAART regimens at the prior visit.
After controlling for possible confounders, the independent associations of the significant predictors with QOL are presented in Table 3. Not surprisingly, baseline QOL was significantly related to subsequent QOL. Among the sociodemographic indicators, older age (P < 0.001), lower income (P < 0.0001), and unemployment (P < 0.0001) were negatively related to PHS. Men reporting less male sexual partners or having no drinks had lower PHS. Indicators of worse health status, such as lower CD4+ cell counts, taking OI medications (P < 0.05), and increased number of HIV-related symptoms (P < 0.0001), outpatient visits (P < 0.0001), hospitalizations (P < 0.0001), or non-AIDS comorbidities (P < 0.0001) were significantly associated with lower PHS. However, history of an AIDS diagnosis by itself was no longer significantly related to PHS after controlling for variables such as HIV symptoms and current use of OI medications. The significant decrease trend of PHS over time disappeared after controlling for many time-dependent variables.
For mental health QOL, after controlling for the values before HAART initiation, sociodemographics were not independently associated with MHS. However, those reporting using recreational drugs (P < 0.05) or interrupting their antiretrovirals in the past 6 months (P < 0.05) had lower mental health QOL, whereas participants with more social support as represented by the number of people around to talk to had better MHS. Among the clinically related indicators, taking amprenavir (P < 0.001), more outpatient visits (P < 0.0001), and having depression symptoms (P < 0.0001) were significantly associated with decreased MHS.
The effects of HAART in containing HIV and reducing HIV-related morbidity and mortality have been demonstrated in many studies.1,2 Thus, in addition to maintaining HIV RNA at undetectable level and preventing patients from further disease progression, maximizing quality of life has been high on the agenda of effectively managing HIV-infected individuals in the HAART era. To meet this challenge, we need to first determine the predictors for lower QOL to help identify targets for possible interventions. In our study, we collected data on individual characteristics (sociodemographics and individual risk behaviors), social support indicator, and clinically related variables (biological markers, HIV-related medication use, and clinical outcome indicators) to identify major predictors for lower QOL, a combination of functional status, general health perceptions, and overall QOL as described in the Wilson conceptual model.4
Among the sociodemographic variables, older age was significantly associated with worse physical functioning, most likely due to declining physical and physiological status of the human body over time.22 Consistent with other studies,23,24 we found that men currently employed or having higher incomes had significantly better PHS. Although higher socioeconomic status implies more dispensable resources or an indirect indicator of better functioning status, it was also legitimate to hypothesize that HIV-infected individuals might benefit from working itself.25 None of the sociodemographic characteristics in our study had an independent impact on MHS.
Of the individual risk behaviors, light-to-moderate drinking and having more male sexual partners were related to better PHS, whereas recreational drug use was significantly associated with worse MHS. As the ability to engage in riskier behaviors itself is an indicator of better functional status and these behaviors are highly correlated over time,26 we are cautious in making any causal inference even though we avoided temporal ambiguity. Although HAART has demonstrated efficacy in decreasing HIV disease progression, its effectiveness largely depends on individual level of adherence to drug regimens. Previous studies14,15 have shown that participants in the MACS reported a very high level of adherence to HAART. This lack of heterogeneity might possibly explain why we did not observe significant associations of lower HAART adherence level or HAART discontinuation with lower QOL. In addition, the significant association between drug interruption and lower mental QOL might be due to the fact that they share the same causal factors, such as depression or adverse effects of medication.15,27
In our study, some environmental factors, such as social support, were also very important for patients' mental health QOL. The number of persons around to talk to showed a clear dose-response relationship with MHS, which implies that more social support might be very helpful in improving patients' mental QOL if it is demonstrated as a modifiable protective factor.28,29 However, considering that some HIV-infected individuals tend to isolate themselves from any support group, proactive outreach of community social support programs to these individuals will be critical to improve patients' mental QOL.
Clinical indicators, including symptoms, health-care use, and OI medication use, are very closely linked to QOL because disease status directly affects patients' general functioning and well-being. In addition, some QOL measurements also contain clinical information as an integral component, such as the pain subscale in the SF-36 form. Similar to previous studies30,31 and congruent with clinical experience, we found that having more HIV-related symptoms, non-AIDS comorbidities, and health-care use were significantly associated with decreased PHS. Although history of an AIDS diagnosis did not independently predict lower PHS in our analysis, it is not surprising because (1) other clinical outcome variables such as clinical symptoms through which AIDS mediates its effect on QOL were adjusted at the same time and (2) the effect of AIDS on QOL was further diluted after controlling for baseline PHS, as about two thirds of AIDS cases were diagnosed before HAART initiation. Among the markers of HIV disease stage, only CD4+ cell counts remained independently associated with PHS, which is consistent with findings by others that CD4+ cell count is more prognostic for HIV disease progression than HIV RNA level while using HAART.32,33
Consistent with previous studies,22,34 the effects of clinical outcomes on MHS were not as significant as they were on PHS in our study. Except for the number of outpatient visits, no other clinical indicators were independently associated with lower MHS. One possible explanation for this might be that the number of outpatient visits is not only a surrogate for disease severity, but also an indicator reflecting individual health-care-seeking preferences. As expected, depressive symptoms were associated with dramatically lowered MHS. In addition, taking amprenavir, which represents salvage treatment after drug resistance,35 was also significantly related to lower MHS.
Our analysis showed different predictor profiles for PHS and MHS. However, to maximize quality of life, further efforts should be made. First, we should identify causal risk factors for QOL as targets for future interventions from among these identified predictors. For instance, the significant association between having more male sexual partners and higher PHS was likely due to the fact that the participants with better functional status had more sexual partners, rather than that having more sexual partners enhanced PHS. Although confirmation of risk factors for QOL might be very complicated, it is still feasible to discern them from prior studies, clinical experience, and pilot intervention tests. Second, we should evaluate whether these risk factors are modifiable. In contrast to demographics, individual risk behaviors such as recreational drug use and HAART use behaviors are well modifiable, and intensive education targeting high-risk populations will likely improve their QOL. In addition, the feasibility or cost-effectiveness of proposed intervention program should also be assessed. Third, the importance of the QOL predictors should not be judged solely based on statistical associations. For example, studies have found that abacavir use can cause severe hypersensitivity symptoms and affect the functional status of susceptible individuals.36,37 Although abacavir was not found to be a statistically significant indicator after controlling for factors such as clinical symptoms, its clinical importance should not be dismissed. Closely monitoring for possible adverse effects of HAART regimens remains a priority for clinicians to avoid decreasing patients' QOL. Last, but not least, comprehensive intervention strategies for improving QOL should reflect combined efforts from clinicians, patients, and the community. With more and more effective HAART regimens available, clinicians now have more options to optimize their patients' QOL through active treatment of HIV-related diseases and non-AIDS comorbidities and through appropriate management of clinical symptoms. Patients can enhance their own QOL through modifying their risk behaviors with help from clinicians and the community. In addition, the community can improve the patients' QOL through active social support.
In our analysis, we used data from a large longitudinal cohort study to identify important predictors for lower QOL in the HAART era after carefully avoiding possible temporal ambiguity. Although we tried to be inclusive in studying multiple dimensions of QOL, some important unmeasured variables, such as spiritual factors and coping styles, were not assessed, although their effects might be reflected through some factors in our study such as social support. In addition, because the participants in our study do not fully represent the larger population of individuals infected with HIV in terms of sex, race/ethnicity, cause of infection, and age, more research among different populations will be needed to further explore predictors for lower QOL. In summary, many variables can serve as predictors for QOL. Maximizing QOL requires targeted and comprehensive interventions for modifiable risk factors, with combined efforts from clinicians, HIV-infected individuals, and the community.
The Multicenter AIDS Cohort Study (MACS) includes the following: Baltimore-The Johns Hopkins University Bloomberg School of Public Health: Joseph B. Margolick (Principal Investigator), Haroutune Armenian, Barbara Crain, Adrian Dobs, Homayoon Farzadegan, Joel Gallant, John Hylton, Lisette Johnson, Shenghan Lai, Justin McArthur, Ned Sacktor, Ola Selnes, James Shepard, Chloe Thio. Chicago-Howard Brown Health Center, Feinberg School of Medicine, Northwestern University, and Cook County Bureau of Health Services: John P. Phair (Principal Investigator), Joan S. Chmiel (Co-Principal Investigator), Sheila Badri, Bruce Cohen, Craig Conover, Maurice O'Gorman, David Ostrow, Frank Palella, Daina Variakojis, Steven M. Wolinsky. Los Angeles-University of California, UCLA Schools of Public Health and Medicine: Roger Detels (Principal Investigator), Barbara R. Visscher (Co-Principal Investigator), Aaron Aronow, Robert Bolan, Elizabeth Breen, Anthony Butch, Thomas Coates, Rita Effros, John Fahey, Beth Jamieson, Otoniel Martínez-Maza, Eric N. Miller, John Oishi, Paul Satz, Gaetano Vaccaro, Harry Vinters, Dorothy Wiley, Mallory Witt, Otto Yang, Stephen Young, Zuo Feng Zhang. Pittsburgh-University of Pittsburgh, Graduate School of Public Health: Charles R. Rinaldo (Principal Investigator), Lawrence Kingsley (Co-Principal Investigator), James T. Becker, Robert L. Cook, Robert W. Evans, John Mellors, Sharon Riddler, Anthony Silvestre. Data Coordinating Center-The Johns Hopkins University Bloomberg School of Public Health: Lisa P. Jacobson (Principal Investigator), Alvaro Muñoz (Co-Principal Investigator), Haitao Chu, Stephen R. Cole, Christopher Cox, Stephen J. Gange, Janet Schollenberger, Eric C. Seaberg, Sol Su. NIH-National Institute of Allergy and Infectious Diseases: Robin E. Huebner; National Cancer Institute: Geraldina Dominguez; National Heart, Lung and Blood Institute: Cheryl McDonald. Web site located at http://www.statepi.jhsph.edu/macs/macs.html.
1. Hammer SM, Squires KE, Hughes MD, et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 cell counts of 200 per cubic millimeter or less. AIDS Clinical Trials Group 320 Study Team. N Engl J Med
2. Detels R, Munoz A, McFarlane G, et al. Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration. Multicenter AIDS Cohort Study Investigators. JAMA
3. McGinn F, Gahagan J, Gibson E. Back to work: vocational issues and strategies for Canadians living with HIV/AIDS. Work
4. Mouton Y, Alfandari S, Valette M, et al. Impact of protease inhibitors on AIDS-defining events and hospitalizations in 10 French AIDS reference centres. Federation National des Centres de Lutte contre le SIDA. AIDS
5. Max B, Sherer R. Management of the adverse effects of antiretroviral therapy and medication adherence. Clin Infect Dis
. 2000;30(suppl 2):S96-S116.
6. Wu AW. Quality of life assessment comes of age in the era of highly active antiretroviral therapy. AIDS
7. Eller LS. Quality of life in persons living with HIV. Clin Nurs Res
8. Johnson RJ, Wolinsky FD. The structure of health status among older adults: disease, disability, functional limitation, and perceived health. J Health Soc Behav
9. Wilson IB, Cleary PD. Linking clinical variables with health-related quality of life. A conceptual model of patient outcomes. JAMA
10. Kaslow RA, Ostrow DG, Detels R, et al. The Multicenter AIDS Cohort Study: rationale, organization, and selected characteristics of the participants. Am J Epidemiol
11. Ware JE, Snow KK, Kosinski M, et al. SF-36 Health Survey: Manual and interpretation guide
. Boston: The Health Institute, 1993.
12. Ware JE, Kosinski M. Interpreting SF-36 summary health measures: a response. Qual Life Res
13. Liu C, Ostrow D, Hu Z et al. Impacts of HIV infection and HAART use on quality of life. Qual Life Res
. In Press.
14. 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
15. Li X, Margolick JB, Conover CS, et al. Interruption and discontinuation of highly active antiretroviral therapy in the multicenter AIDS cohort study. J Acquir Immune Defic Syndr
16. Giorgi JV, Cheng HL, Margolick JB, et al. Quality control in the flow cytometric measurement of T-lymphocyte subsets: the multicenter AIDS cohort study experience. The Multicenter AIDS Cohort Study Group. Clin Immunol Immunopathol
17. Schenker EL, Hultin LE, Bauer KD, et al. Evaluation of a dual-color flow cytometry immunophenotyping panel in a multicenter quality assurance program. Cytometry
18. DHHS/Henry J. Kaiser Family Foundation Panel on Clinical Practices for the Treatment of HIV infection. Guidelines for the use of antiretroviral agents in HIV-infected adults and adolescents. April 2005 revision. Available at: http://aidsinfo.nih.gov
19. Jacobson LP, Li R, Phair J, et al. Evaluation of the effectiveness of highly active antiretroviral therapy in persons with human immunodeficiency virus using biomarker-based equivalence of disease progression. Am J Epidemiol
20. Center for Disease Control. Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep
21. Radloff LS. The CES-D scale: a self-report depression scale for research in the general population. Appl Psychol Meas
. 1977;Vol 1:385-401.
22. Hays RD, Cunningham WE, Sherbourne CD, et al. Health-related quality of life in patients with human immunodeficiency virus infection in the United States: results from the HIV Cost and Services Utilization Study. Am J Med
23. Bing EG, Hays RD, Jacobson LP, et al. Health-related quality of life among people with HIV disease: results from the Multicenter AIDS Cohort Study. Qual Life Res
24. Blalock AC, McDaniel JS, Farber EW. Effect of employment on quality of life and psychological functioning in patients with HIV/AIDS. Psychosomatics
25. Worthington C, Krentz HB. Socio-economic factors and health-related quality of life in adults living with HIV. Int J STD AIDS
26. Gange SJ, Jacobson LP, Munoz A. Parametric models to assess tracking of categorical high-risk sexual behaviour for HIV infection. Stat Med
27. Krentz HB, Gill MJ. The impact on health-related quality of life of treatment interruptions in HIV-1-infected patients. AIDS
28. Bastardo YM, Kimberlin CL. Relationship between quality of life, social support and disease-related factors in HIV-infected persons in Venezuela. AIDS Care
29. Clingerman E. Physical activity, social support, and health-related quality of life among persons with HIV disease. J Community Health Nurs
30. Murri R, Fantoni M, Del BC, et al. Determinants of health-related quality of life in HIV-infected patients. AIDS Care
31. Wilson IB, Cleary PD. Clinical predictors of functioning in persons with acquired immunodeficiency syndrome. Med Care
32. Anastos K, Barron Y, Cohen MH, et al. The prognostic importance of changes in CD4+
cell count and HIV-1 RNA level in women after initiating highly active antiretroviral therapy. Ann Intern Med
33. Tarwater PM, Gallant JE, Mellors JW, et al. Prognostic value of plasma HIV RNA among highly active antiretroviral therapy users. AIDS
34. Burgess A, Dayer M, Catalan J, et al. The reliability and validity of two HIV-specific health-related quality-of-life measures: a preliminary analysis. AIDS
35. Quercia R, Garnier E, Ferre V, et al. Salvage therapy with ritonavir-boosted amprenavir/fosamprenavir: virological and immunological response in two years follow-up. HIV Clin Trials
36. Hetherington S, McGuirk S, Powell G, et al. Hypersensitivity reactions during therapy with the nucleoside reverse transcriptase inhibitor abacavir. Clin Ther
37. Kessler HA, Johnson J, Follansbee S, et al. Abacavir expanded access program for adult patients infected with human immunodeficiency virus type 1. Clin Infect Dis
Keywords:© 2006 Lippincott Williams & Wilkins, Inc.
quality of life; predictor; HAART