Secondary Logo

Journal Logo

Feature

Antiretroviral Therapy Adherence

Testing a Social Context Model Among Black Men Who Use Illicit Drugs

Phillips, Craig J., PhD, LLM, RN, ARNP, PMHCNS-BC, ACRN1

Author Information
Journal of the Association of Nurses in AIDS Care: March-April 2011 - Volume 22 - Issue 2 - p 100-127
  • Free

Abstract

“Step into my parlor said the spider to the fly.”

—Mary Howitt (1829)

In many parts of the world, persons living with HIV (PLWH) who receive treatment and “optimal care” are living longer and healthier lives (Joint United Nations Programme on HIV/AIDS [UNAIDS], 2008a). This is true for most individuals living in developed countries and is beginning to be true for some who are living in developing countries as well (UNAIDS, 2008a). However, the ability to access the resources and services required to achieve longer life is not enjoyed by all PLWH. Achieving a longer life for all individuals is a challenge for developed as well as for developing countries. For many vulnerable PLWH, social contexts are constructed, whereby they are marginalized and shut out from the opportunities that are enjoyed by other members of a civil society (Burris et al., 2004; Franks, Muennig, Lubetkin, & Jia, 2006; Halkitis, Kutnick, & Slater, 2005; Halkitis, Parsons, Wolitski, & Remien, 2003; Lane et al., 2004; Szreter & Woolcock, 2004).

Social context (ecosocial) challenges have resulted in a call for a human rights-based approach to manage HIV disease (UNAIDS, 2008b). Such approaches taken for the management of HIV disease will require intersectoral cooperation and collaboration between affected individuals, their communities, and policy-makers at local, national, and international levels. Antiretroviral therapy (ART) adherence is an essential aspect of HIV disease management that highlights the relevance of a human rights-based approach to managing HIV. The ease with which viral resistance develops, even with brief or intermittent nonadherence, and yields adverse treatment outcomes (Chesney, 2006; Chesney, Farmer, Leandre, Malow, & Starace, 2003), has necessitated the implementation of effective clinical intervention at the individual level and sound policy at the population level (Halkitis et al., 2005). As compared with ART, treatment options for other chronic disease states are more accommodating with respect to variability in adherence patterns (Chesney et al., 2003). Devising strategies to study the complex interactions across ecosocial levels (individual, interpersonal, and social context) of influence may offer insights into ways to enhance existing strategies or develop new ones for facilitating individual ART adherence and to optimize the beneficial effect of these drugs in the individual and the population (Lima et al., 2008; Montaner et al., 2006).

The purpose of this study was to develop a theoretical model to understand the effects of ecosocial context factors (individual, interpersonal, and social capital) that influence ART adherence among HIV-infected Black men who use illicit drugs. To develop the theoretical model for this study, a review of previously published data and a social epidemiologic risk assessment of the social environment within which these men reside were conducted. The influence of social context is geographically widespread with local and global implications; however, this study focused only on a specific vulnerable population of illicit drug users residing in the southeastern United States.

The terminology used to describe ethnic and/or racial groups that are discussed in this article reflects those that have been used in the scientific literature. The usage of the term “Black” is based on a broad category that includes individuals from Africa, people residing in the Caribbean basin, and others who self-identify as Black. The term “African American” is used as a more specific term to identify individuals who were born in the United States or its territories, or those who were born abroad to parents who are citizens of the United States and self-identify as African American. The idea behind using these terms is to clarify that different experiences might occur among some of the members of these groups in the United States which might not reflect the overall experiences felt by all the members of these groups.

Ecosocial Theory and the Social Epidemiology of HIV

Ecosocial theory and social epidemiology have provided the theoretical underpinnings and methodology for understanding the complex social phenomenon of ART adherence among Black men who use illicit drugs. Ecosocial theory stems from the contemporary epidemiology canon that multiple factors contribute toward the spread of disease within a population (Krieger, 1994). However, ecosocial theory views the causative effects as occurring across social levels simultaneously and with observable effects at multiple levels. This perspective is in contrast to other epidemiologic approaches that view cause and effect relationships on the basis of proximity of the agent to the individual, with factors termed as proximal (downstream) and distal (upstream; Krieger, 2008). A proximal–distal explanatory approach may contribute toward perceptions that individuals are unable or unwilling to engage in health promoting behaviors (biomedical individualism) and also that social context has a lesser overall effect as compared with individual behavior and accountability (decontextualized “lifestyle” analyses). A proximal–distal approach is insufficient in the case of health because it does not consider the fact that individual behavior, social context, and social relations, which influence health and disease, occur simultaneously (Krieger, 2008). The causative agent of HIV yields illness at the individual level. At the population level, HIV has amplified effects within specific populations (e.g., illicit drug users), such as those affected by poverty, food insecurity, inadequate health care access, and those engaged in high-risk behaviors that contribute toward the spread of the disease. HIV disease and ART adherence among illicit drug-users provide a means by which ecosocial theory and the methods of social epidemiology are empirically tested.

Social Epidemiologic Framework for ART Adherence

The theoretical framework proposed by Poundstone, Strathdee, and Celentano (2004) provided the analytical structure for our study. We needed to adapt to their theoretical model to study the ecosocial levels that influence ART adherence among HIV-infected Black men who use illicit drugs (Figure 1). The theoretical model used in our study differed from that proposed by Poundstone et al. (2004) in two fundamental ways. First, their model explored ART adherence as being critical for the success of “treatment as prevention,” an essential element of secondary HIV prevention efforts among populations at risk for HIV (Lima et al., 2008; Montaner et al., 2006). Second, it excluded “war and militarization” because this construct, as defined in the original framework, is not a typical issue in the United States.

Figure 1. Theoretical model for ART adherence.
Figure 1. Theoretical model for ART adherence.:
Note.ART = antiretroviral treatment; Freq. = frequency; Soc = social.

Empirical Evidence for the Theoretical Model: Literature Review

Previously published data available on adherence are extensive; however, this review regarding research on adherence was limited to only those studies that included Black and African American men who reside in the United States and to those that provided theoretical evidence to support our theoretical model. The 16 studies on the basis of which the theoretical model was created are summarized in Table 1. In this study, the review of the published data on ART adherence included articles published between 1996 (the year during which combination ART became the standard of care for HIV treatment) and 2009 obtained from databases, such as Computerized Index of Nursing and Allied Health Literature, MedLine, PsycINFO, Health and Psychosocial Instruments, and Sociological Abstracts. Single-match or combination keywords used in the search included the following: adherence, adherence intervention, African American, AIDS, Black, culture, drug use/abuse, homelessness, housing status, HIV, medication adherence, psychological distress, psychological state, social capital, social context, and substance use/abuse. Electronic versions of tables of contents for HIV and social sciences journals were scanned for relevant articles (e.g., Journal of Acquired Immune Deficiency Syndrome, Journal of the Association of Nurses in AIDS Care, AIDS and Behavior, International Journal of STD and AIDS, Social Science and Medicine, and others). Further sources of literature were obtained by reviewing the reference sections of articles which were relevant to the understanding of ART adherence and the factors that influence adherence. The 16 articles used for the review in this study were included because they met the following inclusion criteria: (a) they included discussion on factors that influenced adherence; (b) identified adherence, nonadherence, treatment failure, or treatment discontinuation among members of the study sample; (c) reported Black or African American race and/or ethnicity and gender differences among members of the study sample; and (d) included only studies located in the United States. Studies that reported ART adherence interventions were excluded because the focus of this study was to identify factors that influence adherence among HIV-infected Black men who use illicit drugs.

Table 1
Table 1:
Factors Influencing ART Adherence Among HIV-infected Black Men Who Use Illicit Drugs in the United States
Table 1
Table 1:
Continued
Table 1
Table 1:
(Continued)
Table 1
Table 1:
Continued
Table 1
Table 1:
(Continued)
Table 1
Table 1:
Continued
Table 1
Table 1:
(Continued)

Adherence Research With Black Men

The unique ART adherence challenges and the ease with which viral resistance develops underline the need for effective ART adherence interventions. Factors identified during either adherence studies or other behavioral change studies that have a profound influence on adherence include psychosocial factors and other patient-related factors. Bodenlos et al. (2007), in a study related to appointment attendance, identified certain psychosocial factors that influenced adherence outcomes and which were also evident in ART adherence studies, such as stressful life events, depression, anxiety, hopelessness (Gonzalez et al., 2004; Moss et al., 2004; Sledjeski et al., 2005; Ware et al., 2005); low adherence self-efficacy (Johnson et al., 2006a); low familial1 or social support (Feaster et al., 2010; Szapocznik et al., 2004; Wagner et al., 2002); inadequate patient knowledge regarding HIV disease and its treatment (Bogart et al., 2001; Bogart et al., 2010; Johnson et al., 2006a); and use of recreational substances or substance abuse (Altice et al., 2001; Johnson et al., 2006; Moss et al., 2004; Ware et al., 2005). These factors have been studied primarily at an individual level or a social network (interpersonal and familial) level, with research largely ignoring the social and structural context within which adherence occurs. The result of this oversight was a lack of understanding of the full range and influence of social and structural context factors on adherence.

Research spanning multiple populations has identified several factors that contribute toward an individual's experience with ART and the ability to consistently adhere to a prescribed regimen (Chesney et al., 2000; Oggins, 2003; Ware et al., 2005). Identifying relationships between psychological distress in relation to ART adherence and social and structural context may help in informing clinicians, researchers, and policy makers about the role played by the environment in the psychological state of an individual and, in turn, the influence on other health behaviors, including ART adherence. Individuals who use illicit drugs exhibit high rates of ART discontinuation and this particular area of concern needs to be researched further (Moss et al., 2004). African American ethnicity was reported to be associated with lower average rates of adherence, a critical factor when targeting ART adherence interventions for specific patient populations (Hartzell, Spooner, Howard, Wegner, & Wortmann, 2007; Moss et al., 2004). Experiencing side effects, with the associated influence on activities of daily living and quality of life, may lead to treatment fatigue and ART nonadherence (Johnson et al., 2006; Ware et al., 2005). Methodological challenges in previous studies included the lack of research specific to Black men. Studies conducted in the United States typically include heterogeneous samples (Gonzalez et al., 2004; Hartzell et al., 2007; Johnson et al., 2006; Moss et al., 2004; Ware et al., 2005). This limits our understanding of the full range and effects of factors influencing ART adherence unique to Black men or the development of interventions perceived as relevant by them.

Interpersonal and Social Factors Influencing ART Adherence

Ecosocial factors influencing ART adherence among HIV-infected Black men who use illicit drugs include factors that are proximal and distal to the individual's sphere of control. These factors include human relational concepts found in the availability and accessibility of social supports including partners and significant others, health care and social services providers, and the ability of individuals to use social capital (Altice et al., 2001; Bogart et al., 2001, 2010; Hartzell et al., 2007; Wagner et al., 2002). These relational concepts are complex social constructions that include historical, cultural, and economic underpinnings that influence an individual's ability to access and gain optimal benefit from each social level factor influencing ART adherence (Bogart et al., 2001, 2010; Hartzell et al., 2007; Wagner et al., 2002). Additionally, housing and food security as human development factors and determinants of health represent complex challenges for ART adherence because they influence both internal and external environments that contribute toward an individual's ability to achieve optimal health and well-being (Halkitis et al., 2003; Hartzell et al., 2007; Moss et al., 2004; Yun et al., 2005). Studying social factors that influence ART adherence behaviors among HIV-infected Black men who use illicit drugs is an essential part in the attainment of health care as a human right for all.

Methods and Theoretical Model Testing

On the basis of research completed to date, a theoretical model was specified to explore the ecosocial factors that influence the ART adherence of HIV-infected Black men who use illicit drugs. This model was used to conduct the research reported in this article.

Theoretical Framework

The independent variables in the theoretical model represented factors present at each ecosocial level, that is, individual factors (psychological state, ART experience, and housing status), interpersonal factors (partner status [having a partner vs. no partner], patient–provider relationship), and social context factors (social capital). Psychological state included studying the individual's state of mind, psychological distress, and frequency of illicit drug use. ART experience included studying the individual's stage of ART (newly initiated, reinitiated, or consistently on ART) and tolerability of ART. Social capital consisted of three domains: (a) groups and networks, represented the social support available to the person; (b) trust and solidarity, which characterized the nature of the relationships between the individual and his neighbors, service providers, and strangers as well as perceived changes in these relationships over time; and (c) social cohesion and inclusion, identified community-level differences, their management, and who was excluded from public services (Grootaert, Narayan, Jones, & Woolcock, 2004). Therefore, the model (Figure 1) represented hypothesized effects of each factor on the ART adherence of an HIV-infected Black man who uses illicit drugs. The hypothesized directions have been indicated by arrows for each factor in the model to provide clarity of understanding of the relationships between factors and they are consistent with the belief that individuals are personally responsible for their actions and have ultimate responsibility in choosing whether or not to adhere to ART. Factors such as the individual's psychological state, ART experience, housing status, partner status, patient–provider relationship, and social capital were expected to have an effect on ART adherence among HIV-infected Black men who use illicit drugs. The direction of effect (positive or negative) and theoretically-derived rationale for each hypothesized relationship are described below.

Psychological state.

Psychological state was hypothesized to have direct and indirect effects on ART adherence among HIV-infected Black men who use illicit drugs. The direct effect was expected to be the result of either the presence of acute or chronic levels of psychological distress or the presence of psychopathology such as depression (Sledjeski et al., 2005). Previous research has shown the presence of a positive association between the presence of a positive state of mind and ART adherence (Gonzalez et al., 2004). Illicit drug use was postulated to affect ART adherence directly and indirectly, with higher frequency of current drug use expected to have a greater adverse effect on ART adherence (Moss et al., 2004; Johnson et al., 2006; Ware et al., 2005). The psychological state of an individual was expected to indirectly affect ART adherence through its influence on interpersonal and social context factors.

ART experience.

ART experience (stage of ART and tolerability of ART) was hypothesized to have direct and indirect effects on ART adherence among HIV-infected Black men who use illicit drugs. Direct effects on ART adherence were expected to result from the stage of ART and tolerability of the ART regimen (Johnson et al., 2006; Oggins, 2003; Ware et al., 2005). ART experience was posited to indirectly influence ART adherence through its effect on the patient–provider relationship (Bogart et al., 2001; Johnson et al., 2006). Housing status was hypothesized to directly influence ART adherence because stable housing offers the individual a safe and secure environment within which he/she can reside and store ART agents (Moss et al., 2004). Partner status was expected to have direct effects on ART adherence, social capital, and housing status (Wagner et al., 2002). Whether an individual had a partner or not was also expected to indirectly influence ART adherence through its effects on social capital and housing status.

Patient–provider relationship.

The patient–provider relationship was hypothesized to directly influence ART adherence by facilitating procurement of ART prescriptions and other treatment-related services. The patient–provider relationship was hypothesized to have indirect effects on the relationship between psychological state and ART adherence. It was also hypothesized to indirectly influence housing status by facilitating access to additional health care and social services, including housing. Moreover, it was hypothesized to have indirect effects on the relationship between ART experience and ART adherence by providing clinical expertise, technical assistance, and knowledge related to ART adherence and expectations after consuming antiretroviral agents (Altice et al., 2001; Bogart et al., 2001, 2010; Dawson-Rose et al., 2005; Johnson et al., 2006).

Social context.

Social context (i.e., social capital) was postulated to directly influence ART adherence. Social capital was expected to provide the individual with community-level support and access to programs that would facilitate ART adherence (Altice et al., 2001; Bogart et al., 2010). It was also hypothesized to have indirect (mediator) effects on the relationships between psychological state, partner status, and ART adherence (Stirratt et al., 2006; Wagner et al., 2002).

Study Design and Sampling Procedures

Using a cross-sectional study design, a sample including 160 HIV-infected Black men who used illicit drugs were recruited from physician offices, health care and social services agencies, drug treatment programs, and homeless shelters. Inclusion criteria included men who: (a) self-identified as being Black, African American, Caribbean Black, or African Black; (b) were living with HIV; (c) had been prescribed ART; (d) used illicit drugs (current drug use did not disqualify participation); and (e) were of at least 18 years of age. The exclusion criterion was not being able to understand written or spoken English. The sample included men who had been prescribed ART; however, not all participants in this study followed the prescribed ART regimen. This point has been addressed in more depth in the Results section.

The study was approved by the institutional review board at Florida International University before the study participants were recruited. Flyers were posted at physician offices, health care and social services agencies, drug treatment programs, and homeless shelters. Participants were also referred by case managers and other key stakeholders in the community and through community outreach. Chain referral sampling (Penrod, Preston, Cain, & Starks, 2003), also known as snowball sampling, was also used. Eligible participants were asked to sign the informed consent document and complete the study questionnaire, either at recruitment or immediately after recruitment, at a convenient time and place. Approximately an hour was taken to complete the questionnaires, which were in most cases self-administered unless the participant requested the presence of an interviewer during the initial screening process. Participants received $10 as an incentive when they stated that they had completed the study questionnaire.

Sample Size and Power

Adequacy of the sample size for the planned analyses was assessed using PASS: Power Analysis and Sample Size software 2005 (Hintze, 2004). The model was tested by using path analysis, an analytical technique that uses multiple regression models to assess presumed causal relationships between variables simultaneously. Testing of the model was carried out with the subsample of men (n = 105) who reported being on ART at the time of completion of the survey. Additional statistical analyses could be performed for the sample of 160 Black men but were not reported in this study. The initial power analysis determined that 105 cases could produce 97.5% power to detect an adjusted R2 of .20, with control variables accounting for R2 of .15 and the addition of 12 independent variables accounting for a .05 increase in R2, using an F-test with a significance level (α) of .05 (Cohen, 1988; Hintze, 2004).

Instrumentation

Instruments used in the study are described in Table 2. Reliability estimates for each instrument have been previously reported. In general, reliability estimates for the instruments used were acceptable for the sample and consistent with the results reported in other samples. Exceptions included the Social Capital subscale, groups and networks that had a poor reliability estimate (α = .44). Investigator-developed instruments derived primarily from questionnaires related to demographics are also summarized in Table 2.

Table 2
Table 2:
Model of ART Adherence—Instruments and Reliability Estimates
Table 2
Table 2:
Continued
Table 2
Table 2:
(Continued)

Statistical Analysis

Data were analyzed with SPSS 15 for descriptive data and MPLUS version 6 (Muthén & Muthén, 2010) for path analysis with α = .05. Path analysis was used to estimate the model parameters using a mean and variance-adjusted weighted least squares estimation method suitable for categorical data. Parameter estimates were obtained with standard bootstrapping with 1,000 bootstrap draws. Bias-corrected bootstrap confidence intervals for the model parameter estimates were obtained. Measures to assess model constructs were derived from demographic questions or through the usage of instruments previously developed by other authors; these are summarized in Table 2. Dummy variables were created to represent ART adherence, partner status, housing status, and tolerability of ART because each of these variables was grossly skewed. Partner status was dichotomized. The housing status variable coding scheme (homeless, marginally housed, stably housed) was dichotomized to unstably housed (homeless or marginally housed) versus stably housed. The two indicators for ART experience were also dichotomized (stage of ART—“ART naïve” and “ART experienced” and tolerability of ART—“tolerate ART” and “not tolerate ART”) for analysis.

Patient–provider relationship was analyzed using the Patient Practitioner Orientation Scale (PPOS) and Trust in Physician Scale (TiPS) scores separately and also as a composite score. Social capital subscale scores for the Social Capital Integrated Questionnaire (SC-IQ) domains of groups and networks, trust and solidarity, and social cohesion and inclusion were analyzed separately. The groups and networks domain had a low reliability estimate (Table 2) and was not used for analysis. The trust and solidarity domain resulted in a two-factor solution, that is, trust of health care and social services providers and trust of government institutions. As a result, two trust subscales representing these constructs were used for analysis. The internal consistency of scales and subscales for each instrument was assessed with Cronbach's coefficient alpha and have been reported for the subsample who are on ART (n = 105) and also the full sample (N = 160; Table 2). Theoretical model testing used path analysis to test the hypothesized causal relationships. This process used two phases. In the first phase, all variables in the proposed theoretical model were tested as follows: psychological state (state of mind, psychological state, and current illicit drug use), ART experience (ART experience and tolerability of ART), and housing status were regressed on patient–provider relationship (TiPS, PPOS); psychological state and partner status were regressed on social context (trust of health care and social services providers, trust of governmental institutions, and social cohesion); and finally all the independent variables were regressed on ART adherence. The second phase of the path analysis yielded a trimmed model where variables were removed one at a time from the initial model with those variables having the highest p-values being removed first. The cut-off point for removal included those variables from the initial model that had a p value of >.3. This resulted in partner status and trust of government institutions being trimmed from the final path model because they had the least predictive effect on the outcome—ART adherence.

Results

Sample Demographic Characteristics

The Black men in this sample ranged in ages from 24 to 63 years (44.6 ± 8.6 years), were fairly well educated but unemployed (53%), and with low personal and household income (see Table 3). Most of them did not consider themselves to be in a partnered relationship (n = 142, 89%). Those who reported having a partner were in their current relationship for an average length of <2 years (1.5 ± 4.5 years). Average duration of current relationship for those currently on ART (1.7 ± 5.17 years) was approximately twice that of those who were not on ART (.8 ± 2.41 years). The men reported engaging in high-risk sexual and drug use behaviors. Half of them reported being involved in sex for money (n = 80, 50%) and the other half reported being involved in sex for drugs and/or alcohol (n = 68, 43%). The majority of the men believed that they had acquired HIV through sexual activities, primarily with a woman who was infected with HIV (n = 103, 64%). Approximately one fourth (n = 43, 27%) of the men in this study reported acquiring HIV through a nonsexual mode of transmission (e.g., shared injection equipment, blood transfusion and/or medical procedure, other). Sharing injection equipment with an HIV-infected individual was perceived to be the most likely nonsexual mode of HIV transmission (n = 26, 16%). Men on ART were significantly older as compared with those who were not on ART (t [148] = 2.95, p = .004, r = .24). Although educational qualifications were similar for men on ART and those not on ART, a significant number of men on ART were unemployed and disabled as compared with men who were not on ART (χ2 (3, n = 153) = 9.65, p = .02).

Table 3
Table 3:
Demographic and Background Characteristics of the Samplea

Reasons provided for not being on ART included medication side effects (n = 11), health care provider stating that ART is not required (n = 7), unable to purchase ART from the pharmacy (n = 3), and choosing to not take ART even with the knowledge of the health care provider (n = 3). One man reported that he chose to not take ART and his health care provider was unaware of his decision. The remaining men reported “other” as being the reason for not taking ART, but did not specify the exact reason.

Path Analysis

Model testing analysis was limited to participants who reported that they were currently on ART (n = 105). Multiple regression was used to determine the ability of study variables in the theoretical model to predict self-reported ART adherence. This entailed testing the following research question: Do social context (i.e., social capital) and characteristics of the patient–provider relationship mediate the effects of psychological state, partner status, housing status, and ART experience on self-reported ART adherence among HIV-infected Black men who use illicit drugs? Two hypotheses were generated to determine the relationships between the study variables.

Hypothesis 1.

Among HIV-infected Black men who use illicit drugs, characteristics of the patient–provider relationship will mediate the relationship between adherence and (a) psychological state, (b) housing status, and (c) ART experience. Multiple regression analysis was used to test the hypothesized relationship between the outcome variable (patient–provider relationship) and each of the independent variables—psychological state (psychological distress, positive state of mind, and frequency of illicit drug use), ART experience (stage of ART and tolerability of ART), and housing status. A patient–provider relationship composite score was computed using the total scores from the PPOS and the TiPS. Tolerability of ART was the only independent variable significantly related to the proposed mediating variable patient–provider relationship (Table 4).

Table 4
Table 4:
Multiple Regression Summary for Variables Predicting ART Adherence (n= 105)

Hypothesis 2.

Among HIV-infected Black men who use illicit drugs, social context will mediate the relationship between adherence and (a) psychological state and (b) having a partner versus not having a partner. Partner status yielded the least predictive effect in the initial model tested and was the first predictor that was removed and not included in the final path model. The three social capital domains (groups and networks, trust and solidarity, social cohesion and inclusion) were not significantly related to adherence. The regression model testing the effects of the social cohesion and inclusion domain was predicted by current illicit drug use (β = .40, [−.13, .80], p = .07); however, this relationship did not achieve significance.

Analysis.

The most parsimonious path analysis used the following regression models: (a) patient–provider relationship composite score was regressed on psychological state (positive state of mind and current illicit drug use) and ART experience (tolerability of ART), (b) social cohesion was regressed on psychological state (positive state of mind and current illicit drug use), (c) trust of health care and social services providers was regressed on psychological state (current illicit drug use), and (d) ART adherence was regressed on psychological state (positive state of mind, psychological distress, and current illicit drug use), housing status, ART experience, tolerability of ART, social cohesion, trust of health care and social services providers, and social cohesion (Table 4). The path diagram (Figure 2) represents the observed relationships derived from the study's theoretical model and demonstrates the direction of each relationship. The independent variables representing individual level variables in the final model (i.e., positive states of mind, psychological distress, frequency of illicit drug use, housing status, ART naïve vs. experienced, and tolerability of ART) were entered in block 1, and patient–provider relationship, trust of health care and social services providers, and social cohesion were entered in block 2 of the regression. This approach yielded a model in which housing status significantly contributed toward ART adherence (χ2 = 10.5, df = 15, p = .79; Table 4).

Figure 2. Path diagram for ART adherence.
Figure 2. Path diagram for ART adherence.:
Note.PSMS = positive states of mind scale; CDrU = current illicit drug use; K10 = psychological distress; Home = unstable housing; ARTexp = ART experienced; tART = tolerability of ART; SocCoh = social cohesion; Trust HCP = trust of health care and social services providers; PPRCmp = patient–provider relationship composite; Adhere = mean number of ART doses missed in the last 4 days. ∗p< .10, ∗∗p≤ .01.

Direct, indirect, and total effects were calculated. Direct effects were the β (path) coefficients and the indirect effects were the sum of the products of the indirect paths. Total effects were the sum of the direct and indirect effects (Asher, 1983; see Table 5). Comparing the direct, indirect, and total effects of the independent variables in the model provided insights into which variables influenced ART adherence. The independent variable with the greatest influence on adherence was housing status (β = −5.12, [−6.70, −4.32], p < .01). Tolerability of ART was observed to have a greater indirect effect on ART adherence in comparison with a direct effect (β = −.24, [−43, −.08], p = .01). Positive state of mind (β = −.46, [−1.27, .05], p = .06) and current drug use (β = .40, [−.13, .80], p = .07) were also observed to have a greater indirect effect on ART adherence than a direct effect, although neither effect achieved significance.

Table 5
Table 5:
Summary of Path Coefficient Effects for Variables Predicting ART Adherence (n= 105)

Discussion

The initial impetus for this study was a particular research nurse who observed that research participants reported that they were unable to bring back study pill bottles for pill counts because the police had confiscated them. This was disconcerting to the research nurse when the first participant reported it, but was initially dismissed as a one-time occurrence that probably would not occur in the future. However, over the next few months, several other research participants at various times reported that the police had confiscated their pill bottles. After further exploration, it was discovered that the rationale for the police confiscating medications was that they could not differentiate between illegal drugs, legal drugs being in the possession of individuals for illegal use, or drugs being in the possession of an individual for legal use. These field research observations highlight only one of the many ecosocial environmental complexities in maintaining optimal health and ART adherence faced by the HIV-infected Black men in this study who used illicit drugs.

Additional ecosocial challenges may include the challenge of navigating a patchwork of health care and social service agencies to obtain the necessary services for these men. Drug treatment facility policies that limit admission to individuals who are not actively using substances at the time of initial intake and long waiting periods for drug treatment beds create additional barriers against access to required services. Because Black men and other men of color disproportionately receive public assistance, perhaps “…there exists a disparity and inequality of medical service provision in our society” (Halkitis et al., 2005, p. 554). Previously, physician-prescribing practices may have been influenced by stereotyping of minority groups and illicit drug users rather than through assessment of a patient's ability to adhere and it might have contributed to stigmatization and discrimination among members of these groups (Bogart et al., 2001). Buchanan (2000) argued that some members in the society believed that such treatment was acceptable as it was retribution for the individual failings of those needing to use the services of such health care facilities. He described the response of vulnerable individuals as indicative of the presence of a perception of nihilism among them, contributing to a state of despair and disregard about the consequences of their actions. These nihilistic beliefs most likely contribute toward crime, violence, and illicit drug use, and could also influence health-related behaviors including ART adherence. More recent research has explored the importance of patient–provider engagement and the use of comprehensive interventions to facilitate HIV treatment among injection drug users (Knowlton et al., 2010). Findings from such research studies provide further evidence of the importance of taking an ecosocial approach that includes intersectoral and multilevel strategies to understand the complex environments of vulnerable groups such as HIV-infected Black men who use illicit drugs. In cases when the goal is a human rights approach to HIV that includes the fundamental human right of health for all, we would then require intersectoral cooperation and sound policy to guide the progressive realization of this right.

This study explored the methodological challenges of viewing ART adherence from an ecosocial perspective using a social epidemiologic approach to gain simultaneous understanding of the factors that influenced ART adherence among HIV-infected Black men who use illicit drugs. This study enhanced our understanding of some but not all hypothesized relationships in the proposed model. Nurses, as members of a practice discipline and recognized as trusted allies with patients, are uniquely situated within the health care delivery system to collaborate with patients to enhance ART adherence. This requires for them to collaborate with patients in a manner that facilitates adherence and may extend beyond traditional understanding of factors that influence a patient's adherence to a prescribed therapeutic regimen such as ART. Nurses have been involved in many ART adherence intervention approaches and have helped develop many strategies that can be used to facilitate adherence by helping patients to change behaviors through motivational interviewing (Dilorio et al., 2003, 2009) and maximize abilities of nurses to identify patient challenges and engage with patients to tailor optimal ART adherence interventions (Bakken et al., 2005; Holzemer, Henry, Portillo, & Miramontes, 2000). These strategies are important and have established foundations for enhancing ART adherence among many vulnerable populations. However, when the complexity of the lives of vulnerable populations are added into the ART adherence equation, these efforts have focused on the individual patient's behavior and how the nurse or other health care provider can influence the individual's ability to adhere to an ART regimen, with lesser regard for the social context within which the patient resides. In fact, these strategies might instill a false sense of hope for patients facing structural challenges beyond the direct control of the individual or the health care provider and may reinforce perceptions that the patient is incapable of ART adherence and destined to fail.

Acceptance of environments within which the patient is expected to adhere to a prescribed medication regimen, regardless of whether it is a realistic possibility in the ecosocial environment where he resides, is problematic. This approach contributes toward health care provider perceptions of patients from vulnerable groups as having innate character flaws in cases where they do not adhere to the ART regimen, even when adherence is hindered by ecosocial factors beyond the patient's direct sphere of control. Nurses must question whether focusing efforts on individual adaptation to the broader society is a sufficient approach toward managing a complex disease such as HIV or whether they must advocate changing the status quo of the vulnerable populations. A human rights approach to HIV requires for the States (nations) to create systems for the progressive realization of basic human rights for all those individuals living within their geographic boundaries. This understanding of human rights has been criticized as being insufficient to avert the most egregious human rights violations; however, a human rights approach provides a realistic initial approach with international consensus and legal norms that constitute the foundation on which nursing practice is built.

ART adherence is a complex challenge in the best of circumstances. For individuals belonging to vulnerable groups in society, these challenges may be magnified by the complexities inherent in the ecosocial environment within which they reside. This study tested a causal model of factors influencing ART adherence among HIV-infected Black men who use illicit drugs. Path analysis using multiple regression techniques was used to test and derive evidence for the proposed theoretical model. It was proposed that factors from each of the three levels (individual, interpersonal, and social context) would be interrelated and would contribute toward the ability of the Black man to adhere to ART and that there would be direct relationships between adherence and each of the following conceptual variables: (a) psychological state (positive states of mind, psychological distress, and frequency of illicit drug use), (b) partner status, (c) housing status (analyzed as marginally or stably housed), (d) ART experience (stage of ART [naivety] and tolerability of ART), (e) patient–provider relationship (patient–provider orientation scale and trust in physician's scale), and (f) social capital (groups and networks, trust and solidarity, and social cohesion and inclusion). The model also proposed indirect influences on adherence through the patient–provider relationship and social capital.

Individual Ecosocial Level

Psychological state.

None of the psychological state factors influencing ART adherence had a significant direct effect. Men who had a negative state of mind and who used illicit drugs more frequently were less adherent to their ART regimen; however, the effect was indirect. These results are consistent with those of other researchers who observed lower adherence among individuals with altered psychological states (Gonzalez et al., 2004; Halkitis et al., 2003, 2005; Hartzell et al., 2007; Sledjeski et al., 2005; Tucker et al., 2004) and those who use illicit drugs (Halkitis et al., 2003, 2005; Tucker et al., 2004). Patient–provider relationship and social capital were hypothesized to mitigate state of mind, psychological distress, and frequency of illicit drug use effects on ART adherence among the Black men studied. These hypothesized relationships were observed in this study and the men who had a positive state of mind reported a more effective patient–provider relationship. This finding stands to reason that men with a positive state of mind were potentially more engaged in their health care needs and therefore should be able to collaborate more effectively with their health care and social services providers. Those men who reported current illicit drug use also reported more social cohesion and inclusion. This finding may indicate that the sample studied was part of a cohesive group, which may lend itself to ART adherence interventions that require group cohesion.

ART experience.

Stage of ART, whether a man was newly initiated on ART, reinitiated on ART, or had been on it continuously, had no direct or indirect effects on ART adherence among the men in this study. Men who tolerated ART better perceived that they had an effective patient–provider relationship. This may be the result of the men being able to express their concerns in an open and honest manner with their care providers. This finding was consistent with previously published data that highlighted the importance of an effective patient–provider relationship (Dawson-Rose et al., 2005; Johnson et al., 2006a). It stands to reason that tolerability of ART would be enhanced when there was a trusted and competent care provider available to address the concerns and clinical challenges that arose during the management of a long-term chronic illness such as HIV disease.

Housing status.

A direct causal effect was observed among these men indicating the importance of stable housing for improved ART adherence. An indirect effect of the patient–provider relationship between housing and ART adherence was not observed and may indicate that further interventions to address the critical housing needs of HIV-infected Black men who use illicit drugs are warranted (Kidder et al., 2007).

Individual/Interpersonal Ecosocial Level

Partner status.

Partner status did not contribute toward the initial model tested in a meaningful way and was therefore removed from the final model. The importance of having a supportive partner to manage HIV disease and ART adherence has been previously documented and the need to remove this factor from the model tested was counterintuitive in light of the work by other researchers (Wagner et al., 2002). This may have been the result of the small number of men who reported being in a partnered relationship (n = 7) or may have been the result of short relationship durations reported by them. The average relationship duration reported for the men on ART was <2 years. This may not have been a sufficient amount of time for the men to develop a trusting relationship with a partner. Factors related to family dynamics (family hassles) may have also influenced ART adherence among these men (Feaster et al., 2010; Szapocznik et al., 2004). Additionally, there may have been ecosocial structural factors that contributed toward the men not being able to rely on their partners for support. For example, Lane et al. (2004) characterized the challenges faced by the Black community with regard to HIV prevention efforts. These included increased incarceration rates and decreased pools of “marriageable” men within the Black community. Further research is required to determine whether these factors influence ART adherence and the effects that they have on intimate relationship dynamics within the context of HIV and ART adherence. Finally, intimate partner violence could contribute toward the challenges of managing complex therapeutic regimens such as ART. Regardless of the cause for this counterintuitive observation, nurses can refer Black men to couples counseling when they observe problematic relationships.

Interpersonal Ecosocial Level

Patient–provider relationship.

The effect of the patient–provider relationship was not statistically significant, and the indirect effects were minimal. Perhaps this was because of the recruitment sites for the men who may have developed strategies to access health care and social services in their communities. As hypothesized, tolerability of ART was significantly related to the patient–provider relationship. This finding was consistent with that of previous research that highlighted the importance of an effective patient–provider relationship for long-term ART adherence (Bogart et al., 2001; Oggins, 2003; Ware et al., 2005). Black men who reported higher levels of tolerability of ART may have been more satisfied with their health care provider because they were experiencing fewer adverse effects attributed to ART. Additionally, the man's experience with tolerability of ART may have included a series of negotiations that may have resulted in the perception of a more collaborative and positive patient–provider relationship.

Social Context Ecosocial Level

Social context.

The social context variables did not have a direct influence on ART adherence in this sample. However, as previously mentioned, there was a relationship observed between current drug use and social cohesion.

Study Limitations

Study limitations include the use of a convenience sample of HIV-infected Black men who use illicit drugs recruited from a homeless shelter and drug treatment facilities where the principal investigator was known as an integral part of the community because of long-standing employment as a research nurse in the community. Although this allowed for recruitment of the study sample from a hard-to-reach population, it may have resulted in social desirability bias because participants may have been led to believe that their responses were the desired responses. Instruments with evidence of reliability and validity in other samples were used to counterbalance this limitation. Including the subsample of men currently taking ART may have contributed toward the study being underpowered for the analyses conducted; however it was consistent with the research question. Considering the men as adherent to the directives of their primary health care provider would have increased the sample size but would have addressed a different research question.

As is the case for most tests of behavioral theories, this study was limited to testing directed relationships that were hypothesized to be nonzero in the population; therefore, only the consistency of the estimated direction of the effects was tested. Other competing theories are plausible and have not been ruled out by this approach (Cohen, Cohen, West, & Aiken, 2003). Accordingly, this may not be the correct model, especially because there may have been spurious relationships resulting from sources not included in the model. In addition to specification error, identification error may be of concern, that is, the magnitude of some of the reported parameters may have been wrongly estimated. This often results from sampling error or failure to meet some statistical assumptions. In this particular study, the use of ordinal regression is a strength in that the distributions of the variables were appropriately handled. However, the size of the sample may have resulted in significant sampling error and concern that the reported coefficients were not good estimates of the population values. The standard errors for the direct effects were of a reasonable size, as evidenced by the relatively narrow confidence intervals, which provided some confidence that these estimates are relatively precise (Table 5). However, several of the indirect effects have very wide confidence intervals and many that were not found to be significant should be further studied with larger samples. Additional social-context-level limitations are discussed further.

Social Context and the Conundrum of Perceived Social Capital

This study's lack of significant results related to the influence of the social context variable (social capital) may be explained by the ability of the Black men studied to muster the social capital required to navigate the maze of health care and social services to meet their basic needs. The use of chain-referral (snowball) sampling, chosen for its efficacy in accessing hard-to-reach populations, may have resulted in oversampling from a homogeneous sample, thereby reducing sample variability. This may account for the lack of findings related to social context in this study because there may not have been enough variability in the sample to detect differences. The instrument used to measure social capital in this study was developed by the World Bank and was chosen because it had been used in multiple international contexts (Mitchell & Bossert, 2007; Szreter & Woolcock, 2004). The guidelines for analyzing the findings of the SC-IQ included a strong warning that it required complex methods of analysis (Grootaert et al., 2004). The lack of significant findings from the social capital instrument used for our study may be partially explained by research design. Several items on the SC-IQ did not lend themselves well to cross-sectional, observational studies. For example, items on the groups and networks subscale asked participants to quantify the number of friends and groups with which they were associated. Reliability assessment for items such as this would best be evaluated using test-retest reliability estimates that require a longitudinal study design. Additionally, the groups and networks subscale items may not be closely enough related conceptually to make a cogent subscale. This conceptual discord among the items of the subscale may have accounted for the low reliability estimate (α = .44) observed with this sample. During analysis of the data reported here, confirmatory factor analysis was used with the trust and solidarity domain, as specified by the developers of the SC-IQ. This analysis suggested that a two-factor solution would yield the best results, which suggests that the domain is actually measuring two separate concepts. Therefore, the approach used was to separate the domain into two trust subscales (trust of health care and social services providers and trust of government institutions). Perhaps the use of a social capital instrument requiring less complex methods of analysis would yield different findings in similar samples. For example, an instrument that uses fewer different types of item response formats might have been better. The SC-IQ uses dichotomous items, continuous items, and Likert-type items all on the same subscale, which increases the complexity of analysis and interpretation of results obtained from the instrument. It would also be important to compare different social capital instruments on the basis of their psychometric properties to determine whether they measured similar constructs and which items and conceptual domains of each instrument most closely corresponded with each other.

Structural factors in the U.S. health care system and societal limitations imposed on many HIV-infected Black men who use illicit drugs complicated the prospects of adherence to complex medication regimens such as ART (Halkitis et al., 2005; Krawczyk, Funkhouser, Kilby, & Vermund, 2006; Lane et al., 2004; Szreter & Woolcock, 2004). These structural factors may have limited the likelihood of these men claiming human rights that were available to them under international, national, and local legal instruments. There is a need for continued dialogue between HIV-infected Black men who use illicit drugs, community leaders, and policy stakeholders so as to redress the structural challenges that influence health and ART adherence for these men.

Implications for Research, Practice, and Policy

Echoing the call from UNAIDS (2008b) for a human rights-based approach to HIV management, ART adherence among HIV-infected Black men who use illicit drugs provides an example of both successes and challenges for the realization of health as a human right within the United States (Carmalt & Zaidi, 2004). From a human rights perspective, the problem is perpetuated by actions or inactions in relation to achieving health as a human right for vulnerable populations. The evidence presented in this study provides a model for assessing broader contextual elements that contribute toward an individual's ability to perform a seemingly simple behavior such as following the directives of a health care provider and taking ART as prescribed. Ecosocial theory and social epidemiologic approaches can guide understanding of the complex factors that influence ART adherence. This understanding can facilitate active collaboration between researchers, health care providers, and HIV-infected Black men who use illicit drugs to improve social contexts and facilitate attainment of health as a human right for these men.

Clinical Considerations

Nurses working with HIV-infected Black men who use illicit drugs should conduct the following:

  • Assess social factors influencing behavior, including psychological state, and intervene when they observe psychological distress or the presence of substance abuse.
  • Assess housing status, income, employment, and psychological state (e.g., psychological distress or psychopathology) status to determine appropriate strategies for facilitating ART adherence among these men.
  • Facilitate an effective patient–provider relationship to mitigate tolerability issues inherent with some ART.

Disclosure

The author reports no real or perceived vested interests that relate to this article (including relationships with pharmaceutical companies, biomedical device manufacturers, grantors, or other entities whose products or services are related to topics covered in this manuscript) that could be construed as a conflict of interest.

Acknowledgments

This study was funded in part by a 2007 American Nurses Foundation Grant and the author was a 2007 Gloria Smith, RN Scholar, and American Nurses Foundation Grant recipient. The author would like to thank the men who participated in this study. The author would also like to thank his dissertation chair Dr. JoAnne Youngblut and committee members, Dr. Dorothy Brooten, Dr. Sande Gracia Jones, and Dr. Guillermo Prado.

References

Altice, F. L., Mostashari, F., & Freidland, G. H. (2001). Trust and the acceptance of and adherence to antiretroviral therapy. Journal of Acquired Immune Deficiency Syndrome, 28, 47-58.
Asher, H. B. (1983). Causal modeling (2nd ed.). Beverly Hills, CA: Sage.
Bakken, S., Holzemer, W. L., Portillo, C. J., Grimes, R., Welch, J., & Wantland, D. (2005). Utility of a standardized nursing terminology to evaluate dosage and tailoring of an HIV/AIDS adherence intervention. Journal of Nursing Scholarship, 37, 251-257. doi:10.1111/j.1547-5069.2005.00043.x
Bodenlos, J. S., Grothe, K. B., Whitehead, D., Konkle- Parker, D. J., Jones, G. N., & Brantley, P. J. (2007). Attitudes toward health care providers and appointment attendance in HIV/AIDS patients. Journal of the Association of Nurses in AIDS Care, 18, 65-73. doi: 10.1016/j.jana.2007.03.002
Bogart, L. M., Catz, S. L., Kelly, J. A., & Benotsch, E. G. (2001). Factors influencing physicians{L-End} ' judgments of adherence and treatment decisions for patients with HIV disease. Medical Decision Making, 21, 28-36. doi: 10.1177/0272989x0 102100104
Bogart, L. M., Wagner, G., Galvan, F. H., & Banks, D. (2010). Conspiracy beliefs about HIV are related to antiretroviral treatment nonadherence among African American men with HIV. Journal of Acquired Immune Deficiency Syndrome, 53, 648-655. doi: 10.1097/QAI.0b013e3181c57dbc
Buchanan, D. R. (2000). An ethic for health promotion: Rethinking the sources of human well-being. New York, NY: Oxford University Press.
Burris, S., Blankenship, K. M., Donoghoe, M., Sherman, S., Vernick, J. S., Case, P., … Koester, S. (2004). Addressing the “risk environment” for injection drug users: The mysterious case of the missing cop. Milbank Quarterly, 82, 125- 156. doi: 10.1111/j.0887-378X.2004.00304.x
Carmalt, J., & Zaidi, S. (2004). The right to health in the United States of America: What does it mean? Madrid, Spain: Center for Economic and Social Rights. Retrieved from http://www.cesr.org/downloads/Right%20to%20Health%20in%20USA%202004.pdf
Chesney, M. A. (2006). The elusive gold standard: Future perspectives for HIV adherence assessment and intervention. Journal of Acquired Immune Deficiency Syndrome, 43(Suppl. 1), 149-155.
Chesney, M. A., Farmer, P., Leandre, F., Malow, R., & Starace, F. (2003). Human immunodeficiency virus and acquired immune deficiency syndrome. In: World Health Organization, Adherence to long term therapies: Evidence for action. Geneva, Switzerland: World Health Organization. Retrieved from http://whqlibdoc.who.int/publications/2003/9241545992.pdf
Chesney, M. A., Ickovics, J. R., Chambers, D. B., Gifford, A. L., Neidig, J., Zwickl, B., & Wu, A. W. (2000). Self-reported adherence to antiretroviral medications among participants in HIV clinical trials: The AACTG adherence instruments. AIDS Care, 12, 255-266. doi:10.1080/09540120050042891
Cohen, J. (1988). Statistical power analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Erlbaum.
Dawson-Rose, C., Shade, S. B., Lum, P. J., Knight, K. R., Parsons, J. T., & Purcell, D. W. (2005). Health care experiences of HIV positive injection drug users. The Journal of Multicultural Nursing and Health, 11, 23-30.
Dilorio, C., McCarty, F., DePadilla, L., Resnicow, K., Holstad, M. M., Yeager, K., … Lundberg, B. (2009). Adherence to antiretroviral medication regimens: A test of a psychosocial model. AIDS and Behavior, 13, 10-22. doi:10.1007/s10461-007-9318-4
Dilorio, C., Resnicow, K., McDonnell, M., Soet, J., McCarty, F., & Yeager, K. (2003). Using motivational interviewing to promote adherence to antiretroviral medications: A pilot study. Journal of the Association of Nurses in AIDS Care, 14, 52-62. doi:10.1177/1055329002250996
Feaster, D. J., Brincks, A. M., Mitrani, V. B., Prado, G., Schwartz, S. J., & Szapocznik, J. (2010). The efficacy of structural ecosystems therapy for HIV medication adherence with African American women. Journal of Family Psychology, 24, 51-59. doi:10.1037/a0017954
Franks, P., Meunnig, P., Lubetkin, E., & Jia, H. (2006). The burden of disease associated with being African-American in the United States and the contribution of socio-economic status. Social Science and Medicine, 62, 2469-2478. doi: 10.1016/j.socscimed.2005.10.035
Gonzalez, J. S., Penedo, F. J., Antoni, M. H., Duran, R. E., Fernandez, M. I., McPherson-Baker, S., & Schneiderman, N. (2004). Social support, positive states of mind, and HIV treatment adherence in men and women living with HIV/ AIDS. Health Psychology, 23, 413-418.
Grootaert, C., Narayan, D., Jones, V.N., & Woolcock, M. (2004). Measuring social capital: An integrated questionnaire. World Bank Working Paper No. 18, Washington, DC: The World Bank.
Halkitis, P. N., Kutnick, A. H., & Slater, S. (2005). The social realities of adherence to protease inhibitor regimens: Substance use, health care and psychological states. Journal of Health Psychology, 10, 545-558. doi:10.1177/1359 105305053422
Halkitis, P. N., Parsons, J. T., Wolitski, R. J., & Remien, R. H. (2003). Characteristics of HIV antiretroviral treatments, access and adherence in an ethnically diverse sample of men who have sex with men. AIDS Care, 15, 89-102. doi:10.1080/095401221000039798
Hartzell, J. D., Spooner, K., Howard, R., Wegner, S., & Wortmann, G. (2007). Race and mental health diagnosis are risk factors for highly active antiretroviral therapy failure in a military cohort despite equal access to care. Journal of Acquired Immune Deficiency Syndrome, 44, 411-416.
Hintze, J. (2004). NCSS and PASS. Kaysville, UT: Number Cruncher Statistical Systems.
HIV InSite. (2006). Dealing with drug side effects. Retrieved from http://hivinsite.ucsf.edu/insite?page=pb-treat-04-00
    Holzemer, W. L., Henry, S. B., Portillo, C. J., & Miramontes, H. (2000). The client adherence profiling-intervention tailoring (CAP-IT) intervention for enhancing adherence to HIV/ AIDS medications: A pilot study. Journal of the Association of Nurses in AIDS Care, 11, 36-44. doi:10.1016/S1055-3290 (06)60420-2
    Horowitz, M., Alder, N., & Kegeles, S. (1988). A scale for measuring the occurrence of positive states of mind: A preliminary report. Psychosomatic Medicine, 50, 477-483.
    Howitt, M. (1829). The Spider and the Fly (Poem). Retrieved from http://www.love-poems.me.uk/howitt_the_spider_and_the_fly_funny.htm
    Johnson, M. O., Chesney, M. A., Goldstein, R. B., Remien, R. H., Catz, S., Gore-Felton, C., … Morin, S. F. (2006). Positive provider interactions, adherence selfefficacy, and adherence to antiretroviral medications among HIV-infected adults: A mediation model. AIDS Patient Care and STDs, 20, 258-268. doi:10.1089/apc.2006.20.258
    Johnson, M. O., Gamarel, K. E., & Dawson-Rose, C. (2006). Changing HIV treatment expectancies: A pilot study. AIDS Care, 18, 550-553. doi:10.1080/09540120500241439
    Joint United Nations Programme on HIV/AIDS. (2008a). AIDS outlook: World AIDS Day 2008. Geneva, Switzerland: UNAIDS. Retrieved from http://data.unaids.org/pub/Report/2008/20081128_aids_outlook09_en.pdf
    Joint United Nations Programme on HIV/AIDS. (2008b). Report on the global HIV/AIDS epidemic 2008. Geneva, Switzerland: UNAIDS. Retrieved from http://data.unaids.org/pub/GlobalReport/2008/JC1510_2008GlobalReport_en.zip
    Kessler, R. C., Andrews, G., Colpe, L. J., Hiripi, E., Mroczek, D. K., Normand, S. L. T., … Zaslavsky, A. M. (2003). Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychological Medicine, 32, 959-976. doi:10.1017/ S0033291702006074
      Kidder, D. P., Wolitski, R. J., Royal, S., Aidala, A., Courtney- Quirk, C., Holtgrave, D. R. …: Housing and Health Study Team. (2007). Access to housing as a structural intervention for homeless and unstably housed people living with HIV: Rationale, methods, and implementation of the housing and health study. AIDS and Behavior, 11(Suppl 2), 149-161. doi:10.1007/s10461-007-9249-0
      Knowlton, A. R., Arnsten, J. H., Eldred, L. J., Wilkinson, J. D., Shade, S. B., Bohnert, A. S.,… Purcell, D. W. (2010). Antiretroviral use among active injection-drug users: The role of patientprovider engagement and structural factors. AIDS Patient Care and STDs, 24, 421-428. doi:10.1089/apc.2009.0240
      Krawczyk, C. S., Funkhouser, E., Kilby, J. M., & Vermund, S. H. (2006). Delayed access to HIV diagnosis and care: Special concerns for the Southern United States. AIDS Care, 18(Suppl. 1), 35-44. doi:10.1080/09540120600839280
      Krieger, N. (1994). Epidemiology and the web of causation: Has anyone seen the spider? Social Science & Medicine, 39, 887- 903. doi:10.1016/0277-9536(94)90202-X
      Krieger, N. (2008). Proximal, distal, and the politics of causation: What{L-End} 's level got to do with it? American Journal of Public Health, 98, 221-230. doi:10.2105/AJPH.2007.111278
      Krupat, E., Putnam, S. M., & Yeager, C. (1996). The fit between doctors and patient: Can it be measured? Journal of General Internal Medicine, 11(Suppl. 1), 134.
      Lane, S. D., Rubinstein, R. A., Keefe, R. H., Webster, N., Cibula, D. A., Rosenthal, A., & Dowdell, J. (2004). Structural violence and racial disparity in HIV transmission. Journal of Health Care for the Poor and Underserved, 15, 319-335.
      Lima, V. D., Johnston, K., Hogg, R. S., Levy, A. R., Harrigan, P. R., Anema, A., & Montaner, J. S. G. (2008). Expanded access to highly active antiretroviral therapy: A potentially powerful strategy to curb the growth of the HIV epidemic. Journal of Infectious Diseases, 198, 59-67. doi:10.1086/588673
      Mitchell, A. D., & Bossert, T. J. (2007). Measuring dimensions of social capital: Evidence from surveys in poor communities in Nicaragua. Social Science and Medicine, 64, 50-63. doi:10.1016/j.socscimed.2006.08.021
      Montaner, J. S. G., Hogg, R., Wood, E., Kerr, T., Tyndall, M., & Levy, A. R. (2006). The case for expanding access to highly active antiretroviral therapy to curb the growth of the HIV epidemic. Lancet, 368, 531-536.
      Moss, A., R., Hahn, J.A., Perry, S., Charlebois, E.D., Guzman, D., Clark, R., & Bangsberg, D.R. (2004). Adherence to highly active antiretroviral therapy in the homeless population in San Francisco: A prospective study. Clinical Infectious Diseases, 39, 1190-1198. doi:10.1086/424008
      Muthén, B., & Muthén, L. (2010). Mplus (Version 6). Los Angeles, CA: Statmodel.
      Oggins, J. (2003). Notions of HIV and medication among multiethnic people living with HIV. Health and Social Work, 28, 53-62.
      Penrod, J., Preston, D. B., Cain, R. E., & Starks, M. T. (2003). A discussion of chain referral as a method of sampling hard-to-reach populations. Journal of Transcultural Nursing, 14, 100-107.
      Poundstone, K. E., Strathdee, S. A., & Celentano, D. D. (2004). The social epidemiology of human immunodeficiency virus/ acquired immunodeficiency syndrome. Epidemiologic Reviews, 26, 22-35.
      Sledjeski, E. M., Delahanty, D. L., & Bogart, L. M. (2005). Incidence and impact of posttraumatic stress disorder and comorbid depression on adherence to HAART and CD4+ counts in people living with HIV. AIDS Patient Care and STDs, 19, 728-736. doi:10.1089/apc.2005.19.72
      Stirratt, M. J., Remien, R. H., Smith, A., Copeland, O. Q., Dolezal, C., & Kreiger, D.: the SMART Couples Study Team. (2006). The role of HIV serostatus disclosure in antiretroviral medication adherence. AIDS and Behavior, 10, 483-493. doi:10.1007/s10461-006-9106-6
      Szapocznik, J., Feaster, D. J., Mitrani, V. B., Prado, G., Smith, L., Robinson-Batista, C., … Robbins, M. S. (2004). Structural ecosystems therapy for HIV-seropositive African American women: Effects on psychological distress, family hassles, and family support. Journal of Consulting and Clinical Psychology, 72, 288-303.
      Szreter, S., & Woolcock, M. (2004). Health by association? Social capital, social theory, and the political economy of public health. International Journal of Epidemiology, 33, 650-667. doi:10.1093/ije/dyh013
      Thom, D., Ribisl, K. M., Stewart, A. L., & Luke, D. A. (1999). Further validation and reliability of the Trust in Physician Scale. Medical Care, 37, 510-517.
      Tucker, J. S., Orlando, M., Burnam, A., Sherbourne, C. D., Kung, F., & Gifford, A. L. (2004). Psychosocial mediators of antiretroviral nonadherence in HIV-positive adults with substance use and mental health problems. Health Psychology, 23, 363-370.
      Wagner, G. J., Remien, R. H., Carballo-Diéguez, A., & Dolezal, C. (2002). Correlates of adherence to combination antiretroviral therapy among members of HIV-positive mixed status couples. AIDS Care, 14, 105-109. doi:10.1080/095 40120220097973
      Ware, N. C., Wyatt, M. A., & Tugenberg, T. (2005). Adherence, stereotyping and unequal HIV treatment for active users of illegal drugs. Social Science and Medicine, 61, 565-576. doi:10.1016/j.socscimed.2004.12.015
      Yun, L. W., Maravi, M., Kobayashi, J. S., Barton, P. L., & Davidson, A. J. (2005). Antidepressant treatment improves adherence to antiretroviral therapy among depressed HIV-infected patients. Journal of Acquired Immune Deficiency Syndrome, 38, 432-438.

      1 The work of (Feaster et al., 2010; Szapocznik et al., 2004) extends the use of Structural Ecosystems Therapy (a type of family therapy) as an ART adherence intervention that was effective for improving ART adherence. Although these results are promising for African American women, their efficacy has yet to be reported with Black or African American men.
      Cited Here

      Keywords:

      causal modeling; ecosocial theory; HIV; path analysis; social epidemiology

      © 2011Elsevier, Inc.