Advances in HIV care and the widespread availability of antiretroviral medications have resulted in dramatic reductions in HIV-associated morbidity and mortality. Not all groups have benefited equally from these advances, however. Multiple studies have identified lower likelihood of receipt of antiretroviral therapy (ART) among women and members of racial/ethnic minority groups in the United States.1-3 Independent of race and gender disparities, comorbid mental health and substance use disorders also seem to reduce the likelihood of a patient receiving ART,3-6 perhaps because of providers' or patients' decisions to defer ART until perceived barriers to medication adherence have been addressed.7,8
Disparities have also been identified in HIV-related clinical outcomes, particularly related to psychosocial factors. Mental illness and substance (including alcohol) use disorders affect most HIV-positive patients9 and predict more rapid discontinuation of ART,10 delayed virologic response and more rapid virologic failure in those initiating ART,11-13 and increased disease progression and mortality.14,15 An emerging body of literature also documents the high prevalence of psychosocial trauma, such as childhood sexual or physical abuse in HIV-positive patients,16,17 and associations of trauma history with reduced ART adherence and worse clinical outcomes.18,19 Meanwhile, greater psychologic resources, such as positive affect and expectancy about health outcomes, may protect against disease progression and mortality.20
In addition to the growing attention to psychosocial mediators of HIV outcomes, considerable interest has emerged in better understanding the HIV epidemic in the southeastern US.21,22 For some years, the Southeast has been the US region with the fastest-growing HIV epidemic: in 6 contiguous southeastern states whose epidemics share similar demographics and transmission characteristics (North Carolina, South Carolina, Georgia, Alabama, Mississippi, and Louisiana), 23 reported new cases of AIDS increased 36% from 2000 to 2005 compared with a 2% increase in the rest of the South and a 6% decrease in other regions of the United States.24,25 The HIV epidemic in the Southeast is characterized by relatively high incidence in persons exposed by means of heterosexual sexual intercourse and in women, minorities, the poor, and those living in rural areas. Relatively little research on patterns of ART use and effectiveness has been conducted in the US Southeast. Of 26 studies of ART utilization recently reviewed by Palacio and colleagues,1 no studies were located exclusively in this region and only 1 multisite study, dating from 1991, included multiple recruitment sites in the Southeast.
In this article, we present results from the Coping with HIV/AIDS in the Southeast (CHASE) study, a longitudinal study following 611 HIV-positive individuals engaged in clinical care from 5 southeastern states for over 3 years. The CHASE study primarily sought to examine the influence of sociodemographic and psychosocial characteristics on processes of care and clinical outcomes. Our goal with this analysis was to examine sociodemographic and psychosocial predictors of ART utilization, discontinuation of ART, and virologic failure.
Sample and Procedures
The CHASE study recruited HIV-infected patients who received care at 1 of 8 infectious disease clinics in 5 southern states (Alabama, Georgia, Louisiana, North Carolina, and South Carolina) and has been described in detail previously.18 Briefly, the study's primary goal was to examine the sociodemographic characteristics, coping strategies, life histories, and care utilization patterns of persons receiving HIV-related care outside the 3 major cities (Atlanta, Charlotte, and New Orleans) in the Southeast. Consecutive sampling of patients was used to minimize recruitment/volunteer bias and to yield a study cohort reflective of patients engaged in clinical care in this region. From December 2001 through April 2002, 611 patients enrolled in the CHASE study, representing 79% of patients approached. Sociodemographic and psychosocial information was obtained through structured interviews by trained and field-certified interviewers using validated instruments at enrollment and after 9, 18, and 27 months. Interviews lasted approximately 2 hours and were conducted in a private location of the participant's choosing, given the sensitive nature of the material. Trained health care providers recorded clinical information from patients' medical records on standardized abstraction forms for the 527 CHASE participants (82%) who provided consent for chart abstraction. Medical record data covered the entire follow-up period. This study was approved by Institutional Review Boards at Duke University and all recruitment sites.
We considered 3 outcome measures: (1) of those eligible for ART at enrollment, the probability of being on ART; (2) of those on ART at enrollment, the time from enrollment to the first interruption of ART; and (3) of those on ART and with an HIV RNA plasma viral load (VL) <400 copies/mL at enrollment, the time from enrollment to virologic failure (first VL ≥400 copies/mL). For comparability with prior studies3,26 and in accordance with clinical recommendations in effect at the time of recruitment,27 we defined as ART eligible all patients who (1) were on ART at enrollment or (2) had a CD4 count <350 cells/mm3 or VL >55,000 copies/mL within the 12 months before enrollment. Complete ART histories and laboratory values were collected by means of medical record abstraction. ART histories were reviewed for consistency by a board-certified infectious diseases physician, and queries were resolved through consultation with the original chart.
Sociodemographic variables obtained from the baseline interview included self-reported age, gender, race/ethnicity, and health insurance status.
Clinical characteristics, including CD4 cell count and VL at enrollment, whether the current ART regimen was the participant's first, and length of time on the current regimen, were determined from the medical record. Antiretroviral medication adherence was measured by patient self-report using a modified version of the Adult AIDS Clinical Trials Group Adherence Instrument.28 For the purposes of this analysis, patients were considered nonadherent if they reported missing any doses of their HIV medications during the previous 7 days.28 This definition has been previously demonstrated to correlate with virologic response in this sample.18
Psychosocial measures have been described in detail previously.18,29 In brief, the psychosocial trauma measure, adapted from prior research,30,31 assigns 1 point for each of 15 types of possible lifetime traumatic experiences (eg, sexual abuse, physical trauma, childhood physical and emotional neglect, non-HIV-related life-threatening illnesses or injuries, murder of a close family member, death of a child, and death of a spouse/partner). The measure could range from 0 to 15. This specification of the number of types of traumatic events experienced has been used widely and predicts multiple negative health outcomes.18,29,32 Depressive symptoms were measured with the depression subscale of the Brief Symptoms Inventory (BSI); this subscale has adequate internal consistency (Cronbach α = 0.85) and test-retest reliability (Cronbach α = 0.84).33 Subscale scores were converted to gender-specific T scores ranging from 0 to 100, with a change of 10 units corresponding to a shift of 1 standard deviation in the normative population.33 Recent severe stressful life events during the 9 months preceding the interview were measured using a modified version of the Life Events Survey (LES),34 modified to include only those events that were considered moderately to severely stressful based on our previous studies with interviewer-based objectively rated stresses.35 For the current analysis, we focused only on severe stresses in the past 9 months, including divorce or separation from spouse/partner, death or serious illness of immediate family member, major financial problems (eg, foreclosure, not enough money for basic necessities), physical or sexual assault, or more than a week in prison. The number of recent severe stressful events could range from 0 to 8. Recent alcohol and drug use was measured with the Addiction Severity Index (ASI); standard scoring procedures were followed to generate alcohol and drug composite scores ranging from 0 to 100.36 The reliability and validity of the ASI are well established.36,37 Social support was assessed with the Medical Outcomes Study Social Support Survey,38 and the Pearlin Self-Efficacy Scale was used to measure self-efficacy.39 Both measures have been shown to have high internal and construct validity.
We modeled the probability of being on ART among ART-eligible participants using bivariable contingency tables and multiple logistic regression. We modeled the time from baseline to (1) first ART discontinuation and (2) virologic failure using survival analysis techniques, including Kaplan-Meier functions, log-rank tests, and Cox proportional hazard models. In modeling time to ART discontinuation, participants who did not discontinue were censored at the end of follow-up. In modeling time to virologic failure, participants who maintained virologic suppression were censored at the end of follow-up, ART discontinuation, or at the last VL measure before a gap of 12 months or more in VL measures. For each multivariable model, we included sociodemographic (age, gender, race, and health insurance status), clinical (time on current regimen and whether the current regimen was the participant's first), and psychosocial (depression, alcohol and drug use, lifetime trauma, recent stress, social support, and self efficacy) covariates deemed of a priori interest. We observed moderate correlation between some pairs of psychosocial variables (ρ ranged in absolute value from 0.01 to 0.35; the most highly correlated pairs were social support and depressive symptoms [ρ = −0.35], alcohol and drug scores [ρ = 0.31], and trauma and stress [ρ = 0.260]; all other ρ correlations were <0.25). We therefore compared the full model with a set of separate models, each containing 1 psychosocial variable and all sociodemographic and clinical covariates. Finally, in considering ART discontinuation and virologic failure, we compared models with and without controlling for adherence to assess its role in mediating observed relations.
We included site fixed effects in all models to account for variation in prescribing patterns and covariates across sites. For all models, we confirmed that key distributional assumptions were met. In particular, for the proportional hazard models, we verified the assumption of proportionality in the log hazard by examining plots of Schoenfeld residuals for each covariate in the final model and evaluated interaction terms with the natural logarithm of analysis time for covariates that seemd to violate the proportional hazard assumption.40
In sensitivity analyses, we assessed whether results changed substantively when modeling probability of being on HAART (instead of being on ART) and time to HAART discontinuation (instead of ART discontinuation). Here, HAART was defined as a combination of at least 2 nucleoside reverse transcriptase inhibitors in addition to at least 1 of the following: protease inhibitor, nonnucleoside reverse transcriptase inhibitor, or abacavir, consistent with clinical recommendations effective at the time.27 We also considered an alternate definition of virologic failure that excluded “blips”: a single VL measure ≥400 copies/mL but <1000 copies/mL immediately followed by a VL <400 copies/mL.
Of 611 CHASE participants, 84 did not provide consent for medical record abstraction, 46 were of uncertain ART status at study enrollment, and 7 had no CD4 cell count or VL data available for the 12 months before enrollment (Fig. 1). Additionally, 39 individuals were not on ART and were not medically eligible for ART at enrollment. Analysis 1 thus includes 435 individuals on ART or considered ART eligible at enrollment, including 385 participants on ART and 50 participants not on ART but ART eligible. Most participants on ART were on HAART (n = 338 [88%]). Analysis 2 models time from enrollment to ART discontinuation among the 385 participants on ART at enrollment. Analysis 3 models time from enrollment to virologic failure among the 194 participants on ART and with a VL <400 copies/mL at enrollment.
Participants were primarily 30 to 50 years old (mean ± standard deviation: 40.4 ± 8.6) (Table 1). Thirty percent of the participants were female. Sixty-one percent were African American, with the remainder primarily white. Fifty-four percent were on public health insurance, and 21% were uninsured. The sample was fairly treatment experienced: of those on ART at enrollment, only 39% were on their first ART regimen and 47% had been on their current regimen for a year or more. Lifetime traumatic experiences (mean ± standard deviation: 3.1 ± 2.4) and recent severe stressful events (mean ± standard deviation: 1.4 ± 1.2) were common; although comparable statistics for the general population are not available, we note that the prevalence of childhood sexual abuse in this sample (35% of women and 29% of men) was high relative to general population estimates (20% of women and 5% of men).41 Those included in the analysis were comparable to those not included on all measured demographic and psychosocial characteristics (P > 0.05 for all comparisons; data not shown).
Eighty-one percent of all participants (385 of 474) and 89% of medically eligible participants (385 of 435) were receiving ART at enrollment. Compared with all those eligible for but not receiving ART, those receiving ART at enrollment were on average older and were more likely to have private health insurance, and they had higher CD4 cell counts, less stress, less alcohol use, and greater perceived self-efficacy (Table 2). Of all those on ART at enrollment, those with suppressed VL were more likely to be male and white and to have private health insurance, and they had higher CD4 counts, fewer depressive symptoms, and less stress on average than those with unsuppressed VL.
Probability of Being on ART
Of the 435 participants on or eligible for ART, probability of being on ART was associated in multivariable analyses with greater age (adjusted odds ratio [aOR] = 1.92 per 10 years, 95% confidence interval [CI]: 1.23 to 3.1), fewer recent stressful life events (aOR = 0.68 per event, 95% CI: 0.51 to 0.92), less alcohol use (aOR = 0.64 per 10 points on 100-point scale, 95% CI: 0.46 to 0.90), and greater perceived self-efficacy (aOR = 2.82, 95% CI: 1.41 to 5.62) (Table 3).
Of all participants on ART at baseline, those of minority race/ethnicity discontinued ART more rapidly than white non-Hispanics (adjusted hazard ratio [aHR] = 2.44, 95% CI: 1.33 to 4.49; Fig. 2A). Trends were observed for faster ART discontinuation in women compared with men (aHR = 1.49, 95% CI: 0.91 to 2.45; P = 0.11) and for slower ART discontinuation in those with private health insurance compared with those with public or no insurance (aHR = 0.58, 95% CI: 0.31 to 1.09; P = 0.09). No clinical or psychosocial characteristics were associated with ART discontinuation. In a separate model additionally controlling for adherence, no point estimate changed appreciably but better adherence was associated with a slower rate of ART discontinuation (aHR = 0.58, 95% CI: 0.34 to 0.99).
Of all participants on ART and with suppressed VL at baseline, those of minority race/ethnicity demonstrated virologic failure more rapidly than white non-Hispanics (aHR = 2.01, 95% CI: 1.09 to 3.71; see Fig. 2B). No clinical or psychosocial characteristics were associated with time to virologic failure. In a separate model additionally controlling for adherence, no point estimate changed appreciably and adherence was not associated with virologic failure.
When estimating a set of separate models each containing 1 psychosocial variable in addition to all sociodemographic and clinical variables to address potential collinearity, no estimate changed appreciably. Consistent with the primary analysis, lifetime trauma, depression, drug use, and social support did not predict any outcome whereas recent stress, alcohol use, and self-efficacy predicted only the probability of being on ART. When attention was restricted to participants receiving or eligible for HAART rather than ART, the observed associations with stressful life and events, alcohol consumption were no longer statistically significant (P = 0.10 and P = 0.12, respectively) whereas self-efficacy remained positively associated with HAART receipt; those with private insurance were more likely to be on HAART than those without (aOR = 2.54, 95% CI: 1.10 to 5.87; data not shown). When considering time to discontinuation of HAART rather than ART, the trend toward a disadvantage of women relative to men became statistically significant (aHR 1.83, 1.14 to 2.95, P = 0.01) while the differences by race became a trend (aHR = 0.61, 95% CI: 0.36 to 1.09; P = 0.07). When modeling virologic failure but permitting “blips,” the advantage of whites relative to nonwhites became a trend (aHR = 0.54, 95% CI: 0.27 to 1.09; P = 0.09).
The present study contributes to the existing literature on ART utilization and effectiveness in several ways. First, we examined 3 important treatment-related outcomes in a single cohort (ART use, ART duration, and virologic failure). Further, we evaluated a range of key sociodemographic and psychosocial predictors of these treatment outcomes. In addition, we report data from a cohort representing an understudied and epidemiologically important US population, those living with HIV in the rural and semiurban Southeast.
Overall, a high proportion (89%) of patients medically eligible for ART were receiving ART. We identified important psychosocial predictors of receiving ART among those who were eligible. Those with less alcohol use, greater perceived self-efficacy, and lives marked by fewer severe stresses were more likely to be on ART if eligible. Interestingly, we did not observe statistically significant differences in ART receipt according to current depressive symptoms, current drug use, lifetime trauma exposure, or strength of social support. We do note that all point estimates for these variables were in the expected direction, although the modest sample size resulted in relatively wide confidence intervals. Previous research has generally identified a disadvantage in receipt of ART for individuals with concurrent mental health and substance use disorders,3-6 although comparison with other studies is difficult because of differing time frames, source populations, and measures.
Contrary to expectations, no psychosocial factors were associated with discontinuing ART among those on ART or with virologic failure among those on ART with suppressed VL. This finding may indicate that patients and providers are making effective decisions about who initiates ART. Additionally, those on ART are likely to be more heavily engaged in clinical care than those not on ART; therefore, patients with depression or substance abuse may be more likely to be receiving appropriate treatment (although this hypothesis could not be tested in the current study). We note that psychosocial characteristics were measured not at the time of regimen initiation but at the time of CHASE study enrollment, and the cohort may therefore reflect a “survivor's bias” (those remaining on ART at the time of enrollment are not representative of all those who started ART at comparable times).
Several prior studies have reported increased discontinuation and virologic failure among participants with concurrent psychiatric comorbidity including alcohol use, although conflicting reports exist.10-13 Of particular note, a report from a large HIV clinic offering on-site psychiatric services found that patients with a documented and treated psychiatric disorder were more likely to receive ART and to continue it for at least 6 months.42 This finding may reflect the fact that the “treated psychiatric disorder” group was likely more heavily engaged in clinical care, and thus predisposed to succeed on ART.
We considered multiple psychosocial measures in this analysis, some of which were moderately correlated. Results were consistent whether psychosocial dimensions were considered jointly or separately, suggesting that the inclusion of multiple measures did not lead to unstable results as a result of collinearity or to distal effects being masked by the inclusion of more proximal mediators.
A lower proportion of women and racial/ethnic minorities were on ART relative to men and white non-Hispanics in this sample, although the differences were small and not statistically significant (86% of women vs. 90% of men; 87% of minorities vs. 91% of whites). Most published studies have reported racial and gender disparities in ART receipt, although the magnitude of disparities in recent studies has not been large.1 Gender and racial disparities in ART receipt were reported from a multisite cohort recruited in 2001 (comparable in time to the CHASE study enrollment period); the modest absolute differences in utilization (80.9% of women vs. 84.9% of men; 82.2% of African Americans vs. 83.7% of whites) were statistically significant in this large sample (n > 10,000).3 Two of the 10 sites in this study were large metropolitan areas in the South.
In the present study, racial and ethnic minorities did demonstrate faster rates of ART discontinuation if on ART at enrollment and faster rates of virologic failure if virologically suppressed at enrollment. Some but not all previous studies have identified disadvantages of minority populations in maintenance of HAART and virologic suppression.42,43 A report from AIDS Clinical Trials Group (ACTG) 5095, a randomized clinical trial, indicated that African-American non-Hispanics initiating efavirenz-based regimens were significantly more likely to experience virologic failure and had faster rates of discontinuation of their initial regimen than white non-Hispanics, differences that did not seem to be attributable to differential toxicity or viral mutations.43
Taken together, the findings reported here of small differences by race in ART utilization but important disadvantages for minorities in ART maintenance and virologic failure may reflect barriers to accessing a continuous supply of antiretrovirals in the South. The challenge of sustaining effective HIV treatment for this population is suggested by the fact that only 50% (194 of 385) of those patients on ART had a VL <400 copies/mL at enrollment. The high proportion of participants on the expensive antiretrovirals is perhaps surprising, given that only one quarter of participants had private health insurance and one quarter had no insurance coverage at all at the time of enrollment, and likely reflects the effective safety net role played by payers such as Medicaid, state-run AIDS drug assistance programs (ADAPs), clinical trials, and charity care programs run by pharmaceutic companies and hospitals. Patients who rely on such sources for their medication may be more likely to face gaps in coverage, however, with resulting inability to obtain antiretrovirals. For example, many patients on Medicaid periodically lose coverage (as frequently as every 6 months) and enter what is known as the “spend-down” period, during which they must incur a certain amount of out-of-pocket medical expenses before becoming eligible for Medicaid again. These spend-down policies lead to a predictable pattern of high medical expenses in the first and last months of Medicaid coverage and low expenses (ie, gaps in care) during spend-down months.44 Thirty-one states have Medicaid spend-down provisions, including 4 of 5 CHASE study states. Lack of private health insurance disproportionately affected racial/ethnic minorities in this sample: 83% of minorities lacked private health insurance at enrollment compared with 60% of whites. Although we controlled for insurance type in this analysis, we only had a measure of insurance status at the time of enrollment, which may not have adequately captured the longitudinal dynamics of insurance coverage over the follow-up period.
Some caution should be used in interpreting the results of this study. First, the analysis of probability of being on ART is cross-sectional and should not be interpreted etiologically. Second, the statistical significance of results showed some sensitivity to certain analytic choices, for example, when modeling HAART receipt rather than ART receipt, although point estimates generally remained consistent in direction. We decided it was appropriate to model ART receipt, given most patients receiving ART were receiving HAART and the remainder may have been receiving optimized therapy, given their particular clinical scenario, because a substantial fraction of CHASE study participants were heavily ART experienced at enrollment.
It should also be noted that the analyses of time to treatment discontinuation and virologic failure used as their starting point the date of CHASE study enrollment, which occurred at varying points in patients' HIV treatment history. Because the CHASE sampling frame was defined at a given point in calendar time rather than at ART initiation for each patient, time-to-event analyses that used the (historical) date of ART initiation as their starting point would face a left-censoring challenge. To address the use of date of enrollment as our starting point, we considered measures of prior treatment experience and treatment duration at enrollment; these variables did not predict outcomes and did not affect coefficients of other variables in the models.
Although the availability and use of antiretroviral medications have expanded greatly in the past decade, important groups of patients continue to face disadvantages in their use. HIV clinical care systems should continue to strive to address the psychosocial needs of their patients to ensure their ability to remain engaged in clinical care and fully benefit from ART. Continued attention should focus on the dynamics determining differential use and effectiveness of ART in minority populations, particularly sustaining continuous access to expensive medications for those with unstable insurance coverage.
The authors thank the CHASE study patients for their participation. The authors have no relevant conflicts of interest to report.
All authors had full access to all study data. B. W. Pence takes full responsibility for the integrity of the data and the accuracy of all data analyses. B. W. Pence, J. Ostermann, V. Kumar, K. Whetten, N. Thielman, and M. J. Mugavero were responsible for study concept and design. K. Whetten and N. Thielman were responsible for acquisition of data. B. W. Pence, J. Ostermann, V. Kumar, K. Whetten, N. Thielman, and M. J. Mugavero were responsible for analysis and interpretation of data. B. W. Pence was responsible for drafting of the manuscript. J. Ostermann, V. Kumar, K. Whetten, N. Thielman, and M. J. Mugavero were responsible for critical revision of the manuscript for important intellectual content. B. W. Pence was responsible for statistical analysis. K. Whetten obtained funding.
1. Palacio H, Kahn JG, Richards TA, et al. Effect of race and/or ethnicity in use of antiretrovirals and prophylaxis for opportunistic infection: a review of the literature. Public Health Rep
. 2002;117:233-251 [discussion 231-232].
2. Shapiro MF, Morton SC, McCaffrey DF, et al. Variations in the care of HIV-infected adults in the United States: results from the HIV Cost and Services Utilization Study. JAMA
3. Gebo KA, Fleishman JA, Conviser R, et al. Racial and gender disparities in receipt of highly active antiretroviral therapy persist in a multistate sample of HIV patients in 2001. J Acquir Immune Defic Syndr
4. Mocroft A, Madge S, Johnson AM, et al. A comparison of exposure groups in the EuroSIDA study: starting highly active antiretroviral therapy (HAART), response to HAART, and survival. J Acquir Immune Defic Syndr
5. Fairfield KM, Libman H, Davis RB, et al. Delays in protease inhibitor use in clinical practice. J Gen Intern Med
6. Turner BJ, Fleishman JA, Wenger N, et al. Effects of drug abuse and mental disorders on use and type of antiretroviral therapy in HIV-infected persons. J Gen Intern Med
7. Arnsten JH, Demas PA, Grant RW, et al. Impact of active drug use on antiretroviral therapy adherence and viral suppression in HIV-infected drug users. J Gen Intern Med
8. Sternhell PS, Corr MJ. Psychiatric morbidity and adherence to antiretroviral medication in patients with HIV/AIDS. Aust N Z J Psychiatry
9. Bing EG, Burnam MA, Longshore D, et al. Psychiatric disorders and drug use among human immunodeficiency virus-infected adults in the United States. Arch Gen Psychiatry
10. 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
11. Parienti JJ, Massari V, Descamps D, et al. Predictors of virologic failure and resistance in HIV-infected patients treated with nevirapine- or efavirenz-based antiretroviral therapy. Clin Infect Dis
12. Pence BW, Miller WC, Gaynes BN, et al. Psychiatric illness and virologic response in patients initiating highly active antiretroviral therapy. J Acquir Immune Defic Syndr
13. Miguez MJ, Shor-Posner G, Morales G, et al. HIV treatment in drug abusers: impact of alcohol use. Addict Biol
14. Golub ET, Astemborski JA, Hoover DR, et al. Psychological distress and progression to AIDS in a cohort of injection drug users. J Acquir Immune Defic Syndr
15. Ickovics JR, Hamburger ME, Vlahov D, et al. Mortality, CD4 cell count decline, and depressive symptoms among HIV-seropositive women: longitudinal analysis from the HIV Epidemiology Research Study. JAMA
16. Cohen M, Deamant C, Barkan S, et al. Domestic violence and childhood sexual abuse in HIV-infected women and women at risk for HIV. Am J Public Health
17. Whetten K, Leserman J, Lowe K, et al. Prevalence of childhood sexual abuse and physical trauma in an HIV-positive sample from the deep south. Am J Public Health
18. Mugavero M, Ostermann J, Whetten K, et al. Barriers to antiretroviral adherence: the importance of depression, abuse, and other traumatic events. AIDS Patient Care STDS
19. Mugavero MJ, Pence BW, Whetten K, et al. Predictors of AIDS-related morbidity and mortality in a southern US cohort. AIDS Patient Care STDS
20. Ickovics JR, Milan S, Boland R, et al. Psychological resources protect health: 5-year survival and immune function among HIV-infected women from four US cities. AIDS
21. Aral SO, O'Leary A, Baker C. Sexually transmitted infections and HIV in the southern United States: an overview. Sex Transm Dis
22. Whetten K, Reif S. Overview: HIV/AIDS in the deep south region of the United States. AIDS Care
. 2006;18(Suppl 1):S1-S5.
23. Reif S, Geonnotti KL, Whetten K. HIV Infection and AIDS in the Deep South. Am J Public Health
24. Centers for Disease Control and Prevention (CDC). HIV/AIDS Surveillance Report
, vol. 13 (2001). Atlanta, GA: CDC; 2002.
25. Centers for Disease Control and Prevention (CDC). HIV/AIDS Surveillance Report
, vol. 17 (2005). Atlanta, GA: CDC; 2006.
26. Cohen MH, Cook JA, Grey D, et al. Medically eligible women who do not use HAART: the importance of abuse, drug use, and race. Am J Public Health
27. Panel on Clinical Practices for Treatment of HIV Infection. Guidelines for the Use of Antiretroviral Agents in HIV-Infected Adults and Adolescents, August 13, 2001
. Department of Health and Human Services (DHHS); Washington, DC. 2001.
28. Chesney MA, Ickovics JR, Chambers DB, et al. Self-reported adherence to antiretroviral medications among participants in HIV clinical trials: the AACTG adherence instruments. Patient Care Committee and Adherence Working Group of the Outcomes Committee of the Adult AIDS Clinical Trials Group (AACTG). AIDS Care
29. Leserman J, Whetten K, Lowe K, et al. How trauma, recent stressful events, and PTSD affect functional health status and health utilization in HIV-infected patients in the south. Psychosom Med
30. Bernstein D, Fink L. Childhood Trauma Questionnaire Manual
. San Antonio, TX: The Psychological Corporation; 1998.
31. Kilpatrick D, Resnick HS. A description of the posttraumatic stress disorder field trial. In: Davidson J, Foa, EB, eds. Posttraumatic Stress Disorder: DSM-IV and Beyond
. Washington, DC: American Psychiatric Press; 1993.
32. Felitti VJ, Anda RF, Nordenberg D, et al. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. Am J Prev Med
33. Derogatis LR, Melisaratos N. The Brief Symptom Inventory: an introductory report. Psychol Med
34. Sarason IG, Johnson JH, Siegel JM. Assessing the impact of life changes: development of the Life Experiences Survey. J Consult Clin Psychol
35. Leserman J, Petitto JM, Gu H, et al. Progression to AIDS, a clinical AIDS condition and mortality: psychosocial and physiological predictors. Psychol Med
36. McLellan AT, Kushner H, Metzger D, et al. The fifth edition of the Addiction Severity Index. J Subst Abuse Treat
37. Leonhard C, Mulvey K, Gastfriend DR, et al. The Addiction Severity Index: a field study of internal consistency and validity. J Subst Abuse Treat
38. Sherbourne CD, Stewart AL. The MOS social support survey. Soc Sci Med
39. Pearlin LI, Lieberman MA, Menaghan EG, et al. The stress process. J Health Soc Behav
40. Carpenter CC, Cooper DA, Fischl MA, et al. Antiretroviral therapy in adults: updated recommendations of the International AIDS Society-USA panel. JAMA
41. Finkelhor D. Current information on the scope and nature of child sexual abuse. Future Child
42. Himelhoch S, Moore RD, Treisman G, et al. Does the presence of a current psychiatric disorder in AIDS patients affect the initiation of antiretroviral treatment and duration of therapy? J Acquir Immune Defic Syndr
43. Gulick RM, Ribaudo HJ, Shikuma CM, et al. Three- vs four-drug antiretroviral regimens for the initial treatment of HIV-1 infection: a randomized controlled trial. JAMA
44. Whetten K, Zhu CW. Do United States-based Medicaid spend-down programmes make public sense for persons with HIV/AIDS? AIDS Care