Relationships Between Demographic, Clinical, and Health Care Provider Social Support Factors and Internalized Stigma in People Living With HIV

Feature

Relationships Between Demographic, Clinical, and Health Care Provider Social Support Factors and Internalized Stigma in People Living With HIV Nara Jang, PhD, MPH, RN Suzanne Bakken, PhD, RN, FAAN, FACMI* Internalized HIV stigma (IHS) threatens people living with HIV (PLWH) and the public. Our purpose was to identify relationships between PLWH perceptions of IHS and demographic, clinical, overall health, health-related quality of life, and perceived health care provider (HCP) social support. Using survey data from PLWH (n 5 292) in an urban HIV clinic, we first examined the reliability and validity of an existing IHS measure. Exploratory factor analysis revealed that the IHS score was composed of three factors: (a) Perception of Negative Societal Beliefs, (b) Fear of Disclosure, and (c) Perception of Negative HCP Beliefs, which were used as dependent variables in the multivariate analysis. In multivariate regression models, gender, education, quality of life, Getting Needed Care, and Trust in HCP were significantly related to at least one IHS factor. Our findings advance the measurement of IHS and provide a foundation for intervention development to reduce IHS in care settings. (Journal of the Association of Nurses in AIDS Care, -, 1-11) Copyright Ó 2016 Published by Elsevier Inc. on behalf of Association of Nurses in AIDS Care Key words: health care provider communication, health care provider trust, health-related quality of life, internalized HIV stigma, social support, quality care

Goffman (1963) initially defined the concept of stigma as a negative attribute given to an individual

by so-called ‘‘normals,’’ and further categorized it as stigma of character traits (e.g., weak will, unnatural passions), physical stigma (e.g., physical deformities of the body), and stigma of group identity (e.g., race, nation, religion). In contrast, other researchers characterized stigma according to the perspective of the individual experiencing it, including (a) enacted stigma, or the experience of unfair treatments and/or prejudice from others (Scambler & Hopkins, 1986); (b) anticipated stigma, or the fear of unfair treatments or prejudice from others in the future (Markowitz, 1998); and (c) internalized stigma, or internalization or acceptance of negative perceptions from a society (Link, 1987). HIV stigma emerged from initial disease transmission in the homosexual community and a dominant social antigay sentiment. HIV-related stigma has been epidemic over the past 3 decades despite stigma reduction programs through education, health care provider (HCP) training, increased exposure of the general public to persons living with HIV (PLWH) through mass media, and antidiscrimination legislation (Pulerwitz, Michaelis, Weiss, Brown, & Mahendra, 2010). HIV stigma is a barrier to Nara Jang, PhD, MPH, RN, is an Adjunct Instructor, Ewha Womans University, Seoul, Korea. Suzanne Bakken, PhD, RN, FAAN, FACMI, is a Professor, Columbia University School of Nursing and Department of Biomedical Informatics, New York, New York, USA. (*Correspondence to: [email protected]).

JOURNAL OF THE ASSOCIATION OF NURSES IN AIDS CARE, Vol. -, No. -, -/- 2016, 1-11 http://dx.doi.org/10.1016/j.jana.2016.08.009 Copyright Ó 2016 Published by Elsevier Inc. on behalf of Association of Nurses in AIDS Care

2 JANAC Vol. -, No. -, -/- 2016

accessing HIV treatment (Kinsler, Wong, Sayles, Davis, & Cunningham, 2007) and is associated with poor adherence to antiretroviral therapy (ART; Rao et al., 2012). Consequently, stigmatized PLWH have lower overall health status and poorer quality of life (QOL; Slater et al., 2015; Vyavaharkar, Moneyham, Murdaugh, & Tavakoli, 2012). Internalized HIV stigma (IHS), a multidimensional construct that captures stigma related to treatment and other aspects of HIV, is a critical issue for PLWH and the general public for several reasons (Sayles et al., 2008). In terms of PLWH, higher levels of IHS are associated with decreased likelihood to access appropriate medical care (Sayles, Wong, Kinsler, Martins, & Cunningham, 2009), suboptimal adherence to antiretroviral therapy (ART; Waite, Paasche-Orlow, Rintamaki, Davis, & Wolf, 2008), poor QOL (Slater et al., 2015; Vyavaharkar et al., 2012), and lower levels of physical and mental health, including depression (Wolitski, Pals, Kidder, Courtenay-Quirk, & Holtgrave, 2009). The last is important because depression has been associated with nonadherence to ART (Gonzalez, Batchelder, Psaros, & Safren, 2011), compromised immunity (Miller, 2010), and increased risk of HIV-related mortality (Leserman, 2008). Moreover, IHS has the potential to threaten the public health due to a decreased tendency for PLWH with high levels of IHS to disclose their seropositive status. For example, Overstreet, Earnshaw, Kalichman, and Quinn (2013) reported that Black men with HIV infection in the United States who also have sex with men and higher levels of IHS were less likely to have disclosed their HIV status to their last sexual partners or to family members. Consequently, Fear of Disclosure may translate into a public health threat, resulting in an increased incidence of HIV infection. Social support is one approach to address IHS. Research outside of the HIV domain has indicated that social support (e.g., emotional, informational, instrumental) has buffering effects (Cassel, 1976; House, 1987), which suggests that the negative effects of IHS on health may have lessened or nullified IHS through critical sources of social support for PLWH, such as friends, family, and HCP. HCP social support may be experienced as

effective communication (informational support), getting needed care (instrumental support), and trust (emotional support). For example, HCP communication skills positively influence PLWH health outcomes and levels of IHS (Kaai et al., 2007). One mechanism of this positive relationship between communication skills and health outcomes is through improving adherence to ART (Sayles et al., 2009). In a Chinese study, high levels of trust in HCP led to improved health status and better resistance to the deleterious effects of IHS (Wang & Wu, 2007). This finding suggests that trust in HCP (i.e., emotional support) may play a key role in buffering the harmful effects of IHS on health. The primary purpose of our study was to identify relationships between PLWH perceptions of IHS and demographic characteristics, clinical factors, overall health, health-related QOL, and perceived HCP social support. As a preliminary step to examining these relationships, we conducted an exploratory factor analysis (EFA) of an existing IHS instrument (Sayles et al., 2008) to examine validity and reliability in the study sample. Thus, the two research questions for our study were: (a) What is the construct validity and internal consistency reliability of the IHS? and (b) What are the relationships between PLWH perceptions of IHS and demographic characteristics, clinical factors, overall health, healthrelated QOL, and perceived HCP social support (HCP Communication Skills [informational support], Getting Needed Care [instrumental support], and Trust in HCP [emotional support])?

Methods Setting and Sample The study setting was New York-Presbyterian System SelectHealth, an HIV Special Needs Plan. The convenience sample was composed of SelectHealth members who completed baseline surveys as part of the My Health Profile project (Gordon et al., 2012). The My Health Profile project focused on the development of a Web-based system that allowed PLWH, case managers, and clinicians to view data related to continuity of care.

Jang, Bakken / Internalized HIV Stigma

Procedures The study was approved by the Columbia University Medical Center Institutional Review Board. After SelectHealth members were made aware of the My Health Profile project through a letter from SelectHealth leadership, bilingual staff recruited study participants through telephone and face-toface contact. At 6-month intervals, cohorts of approximately 100 SelectHealth members completed Audio Computer Assisted Self-Interview (ACASI) surveys in English or Spanish that included questions related to demographics, medical history, medical care, continuity of care, other services, satisfaction with care, quality of life, trust, stigma, willingness to share personal health information, and computer literacy. To support the collection of reliable data, bilingual research personnel were available to assist with the ACASI process. Consistent with community norms for research participation, PLWH were compensated with a $20 gift card for their time to complete the survey and a round-trip subway card for their participation. Data for our study were from the first time point contact for each individual. Measures Short form of the internalized HIV stigma. To evaluate levels of participant IHS, we employed the short form of the IHS instrument adapted from the measurement developed by Sayles and colleagues (2008). The instrument included 14 items with a 5-point Likert response scale ranging from 1 (none of the time) to 5 (all of the time). Previous uses of the IHS reported two factor scores, HIV Blame and HIV Disclosure. For our study, a score was generated for each IHS component based on the results of the EFA in our study sample. Demographic characteristics. Demographic variables included age, gender, ethnicity, race, and education. For the gender variable, male-to-female transgender was classified as female, and female-tomale transgender was classified as male to reflect gender identity. Race was collapsed into three groups for the analysis: White, Black or African American, and Other (Asian, Native Hawaiian/Pacific Islander, American Indian or Alaska Native, or Others). The

3

education variable was also categorized into three levels: less than high school graduate, high school graduate or General Education Development completion, and associate’s degree or higher. Clinical factors. AIDS diagnosis and CD41 T cell count were included with the latter, categorized as: , 200 cells/mm3, $200 to ,500 cells/mm3, and $500 cells/mm3. Overall health. Participants rated their overall health, answering: ‘‘In general, would you say your health is .’’ with response options from 1 (excellent) to 5 (poor). Health-related QOL. To assess health-related QOL, we employed the short form of the Health Survey (SF-12v2) composed of 12 items, each rated on a 5-level Likert scale using a 4-week recall period. The eight sub-scales (physical functioning, role limitation [physical perspective], pain, general health, vitality, role limitation [emotional perspective], social functioning, and mental health) were reported as two components, a physical component summary (PCS) and a mental component summary (MCS). We computed t-scores per the SF-12v2 scoring guide; a score of 50 equaled the approximate mean of the U.S. general population, and 10 points indicated one standard deviation (Quality Metric Inc., 2005). HCP social support: HCP who Communicate Well. HCP informational support was measured with the HCP who Communicate Well instrument. This instrument was used in the Consumer Assessment of Health Plans Study surveys (Hargraves, Hays, & Cleary, 2003) and consisted of four items (listening carefully, explaining clearly, respecting patient stories, and spending enough time with patients) rated on a 4-point Likert scale ranging from 1 (never) to 4 (always). Responses to the four items were summed for a total score, ranging from 4 to 16. Higher scores represented better communication skills (i.e., higher informational support). The internal consistency reliability of the instrument in our study population was 0.91. HCP social support: Getting Needed Care. The Getting Needed Care instrument was used to evaluate

4 JANAC Vol. -, No. -, -/- 2016

the perceptions of receiving needed care (i.e., HCP instrumental support; Hargraves et al., 2003). Each item in the four-item instrument was rated on a 3-point Likert response scale ranging from 1 (a big problem) to 3 (not a problem). The responses were summed for a total score ranging from 4 to 12. Higher scores represented better outcomes. The internal consistency reliability in our study population was 0.77.

QOL [PCS and MCS]); and HCP social support (HCP Who Communicate Well [informational support], Getting Needed Care [instrumental support], and Trust in HCP [emotional support]). All statistical analyses were performed with SAS, University Edition (SAS Institute Inc., Cary, NC).

Results HCP social support: Trust in HCP. The four-item Trust in HCP scale was used as a measure of HCP emotional support. Two items came from the HIV Cost and Services Utilization Study at the University of California, Los Angeles, and the other two items came from Hall and colleagues (2002). All items had a 5-point Likert response scale with responses summed for a total score that ranged from 4 to 20. Higher scores represented lower trust. The internal consistency reliability in our study population was 0.88. Data Analysis Exploratory factor analysis of IHS. To address the first research question, we conducted an EFA using principal component analysis and Varimax rotation to examine the construct validity of the IHS from the perspective of its factor structure. Items with an Eigenvalue greater than 1 were retained in the analysis. To evaluate internal consistency reliability, Cronbach’s alpha was computed for each factor. Mean factor scores were used in correlational and multivariate regression analyses. Descriptive, correlational, and multivariate regression analyses. Participant characteristics were summarized with descriptive statistics. A correlational analysis was used to identify significant bivariate associations between the IHS factor scores and other study variables. To address the second research question, three multivariate regressions were performed to identify significant correlates of IHS factor scores that served as dependent variables in the analyses. Correlates (i.e., predictor variables) included demographic characteristics (age, gender, ethnicity, race, education, and individual income); clinical factors (AIDS diagnosis and CD41 T cell count); other health-related variables (overall health,

Profiles of Participants The mean age of the participants was 46.7 years, ranging from 20 to 65 years (Table 1). The majority of the respondents was male (65.1%), non-Hispanic (75.3%), African American (54.5%), and had a high school diploma or less education (75.7%). Most (82.2%) had less than a $10,000 income in the previous 12 months. Almost 18% of participants reported a last CD41 T cell count of less than 200 cells/mm3, and 46.9% of participants had been diagnosed with AIDS. The mean health status was 7.58 (range 5 1 to 10). Both PCS and MCS QOL scores were approximately one standard deviation below the U.S. general population. In terms of HCP social support variables, participants reported high mean scores on HCP Communication (informational support) and Getting Needed Care (instrumental support), and low mean scores (i.e., high emotional support) on Trust in HCP. Exploratory Factor Analysis With Principal Component Analysis The initial EFA of the IHS instrument generated four factors, but there was a lack of theoretical consistency between the items that loaded on each factor, and one item, ‘‘My family is comfortable talking about my HIV,’’ did not load on any factor. Consequently, the EFA was repeated without the item and the remaining 13 items loaded on three factors, which we named Perception of Negative Societal Beliefs, Fear of Disclosure, and Perception of Negative HCP Beliefs based on the items in the factors. These three factors accounted for 60.8% of the variance in the IHS score (Table 2). The number of items, mean factor scores, the percent of variance explained, and Cronbach’s alphas (internal consistency

Jang, Bakken / Internalized HIV Stigma Table 1.

5

Participant Characteristics (n 5 292) N (%) or M ± SD (Range)

Characteristics Age, y Gender Male or transgender male Female or transgender female Ethnicity Non-Hispanic Hispanic Race White Black or African American Other (Asian, Native Hawaiian/Pacific Islander, American Indian or Alaska Native, or Others) Education None or less than high school High school diploma or GED, AA or Associate’s Degree, Junior or 2-year College BA, BS, Bachelor’s, 4-year College Degree, or Graduate or Professional Degree Individual Income , $10,000 $ $10,000 Most recent CD41 T cell count ,200 cells/mm3 $200 cells/mm3 to , 500 cells/mm3 $500 cells/mm3 AIDS diagnosis No Yes Overall health Health-related QOL PCS MCS HCP who Communicate Well (HCP informational support) Getting Needed Care (HCP instrumental support) Trust in HCP (HCP emotional support)

46.7 6 9.1 (20-65) 190 (65.1) 102 (34.9) 219 (75.3) 72 (24.7) 36 (12.3) 159 (54.5) 97 (33.2) 106 (36.3) 115 (39.4) 71 (24.3) 240 (82.2) 52 (17.8) 49 (17.6) 120 (43.2) 109 (39.2) 121 (53.1) 107 (46.9) 7.58 6 1.94 (3.00-10.00) 38.08 6 9.72 (22.04-56.86) 41.10 6 7.60 (10.43-56.77) 14.14 6 2.74 (4.00-16.00) 10.73 6 1.87 (4.00-12.00) 6.13 6 3.03 (4.00-20.00)

Note. N 5 number of participants; SD 5 standard deviation; Gender: A transgender male was included in the male category and a transgender female was included in the female category; GED 5 general education development; AA 5 Associate of Arts; BA 5 Bachelor of Arts; BS 5 Bachelor of Science; Individual Income for a year; Overall health, score range from 0 to 10; QOL 5 quality of life; PCS 5 physical component summary; MCS 5 mental component summary; T scores (mean 5 50, standard deviation 5 6 10 [i.e., 61]); HCP 5 health care provider; HCP who Communicate Well, possible scores range from 4 to 16, higher is better; Getting Needed Care, possible scores range from 3 to 12, higher is better; Trust in Health Care Provider, possible scores range from 4 to 20, lower is better.

reliabilities) for each factor were: Perception of Negative Societal Beliefs (7 items, M 5 2.50, SD 5 0.88, 25.67%, and .83), Fear of Disclosure (4 items, M 5 2.77, SD 5 1.18, 22.24%, and .81), and Perception of Negative HCP Beliefs (2 items, M 5 1.91, SD 5 0.98, 12.92%, and .77). Bivariate Correlation Table 3 reports correlations between the three dependent variables (IHS scores), and the continuous

and the dichotomous independent variables. There were no significant correlations between demographic characteristics and IHS scores. In terms of clinical variables and overall health, the Perception of Negative Societal Beliefs score was negatively correlated with AIDS diagnosis (r 5 0.14, p 5 .034) and positively associated with overall health (r 5 0.20, p 5 .001). The significant correlations between health-related QOL measures (PCS and MCS) and IHS scores varied. Perception of Negative Societal Beliefs and Fear of Disclosure

6 JANAC Vol. -, No. -, -/- 2016 Table 2.

Factor Loadings for Exploratory Factor Analysis of the Short Form of Internalized HIV Stigma Scale (n 5 288)

N Factor 1. Perception of Negative Societal Beliefs (a 5 .83) Stigma 2 People blame me for having HIV Stigma 4 People think you can’t be a good parent if you have HIV Stigma 5 People treat me as less than human now that I have HIV Stigma 8 People I am close to are afraid they will catch HIV from me Stigma 9 I feel like I am an outsider because I have HIV Stigma 3 Medical providers assume people with HIV sleep around Stigma 1 Society looks down on people who have HIV Factor 2. Fear of Disclosure (a 5 .81) Stigma 10 I feel ashamed to tell other people that I have HIV Stigma 7 I am concerned that, if I am sick, people I know will find out that I have HIV Stigma 6 I am concerned that, if I go to an AIDS organization, someone I know might see me Stigma 12 It is important for a person to keep HIV a secret from co-workers Factor 3. Perception of Negative HCP Beliefs (a 5 .77) Stigma 13 Nurses and doctors treat people who have HIV as if they are contagious Stigma 14 Nurses and doctors dislike caring for patients with HIV Total variance explained

M (SD)

1

Factor 2

3

2.50 (0.88)

Variance Explained 25.67%

288

2.51 (1.39)

.738

288

2.51 (1.32)

.711

288

2.16 (1.25)

.695

288

1.83 (1.25)

.641

288

2.31 (1.36)

.633

288

2.54 (1.26)

.594

288

3.60 (0.94) 2.77 (1.18)

.451 22.24%

288

2.77 (1.52)

.800

288

2.49 (1.51)

.791

288

2.40 (1.46)

.723

288

3.43 (1.40) 1.91 (0.98)

.686 12.92%

288

2.03 (1.16)

.875

288

1.80 (1.03)

.839 60.74%

Note. HCP 5 health care provider; a 5 Cronbach’s a. Response to each item is on a 5-point categorical response scale: 1 5 none of the time to 5 5 all of the time; Stigma 11, My family is comfortable talking about my HIV, did not load on any three factor, so it was removed.

scores were negatively correlated with MCS, but not with PCS. In contrast, Perception of Negative HCP Beliefs was negatively correlated with PCS (r 5 20.162, p 5 .01). There were significant correlations between all three IHS scores and the three HCP social support variables (HCP who Communicate Well [informational support], Getting Needed Care [instrumental support], and Trust in HCP [emotional support]). Correlations were negative for the first two (i.e., more support and lower IHS), but positive for Trust in HCP because more trust was reflected by a lower score.

Multivariate Regression With IHS Factor Scores from EFA as Dependent Variables In the multivariate regressions, the percentages of explained variance for IHS scores were Perception of Negative Societal Beliefs (32%), Fear of Disclosure (14%), and Perception of Negative HCP Beliefs (22%; Table 4). Five variables were significantly related to at least one IHS factor score: gender, education level, MCS, Getting Needed Care (instrumental support), and Trust in HCP (emotional support). Higher education levels, higher MCS score,

Jang, Bakken / Internalized HIV Stigma Table 3.

7

Bivariate Correlations Between Dependent Variables (IHS Scores) and Independent Variables Perception of Negative Societal Beliefs

Age Gender Ethnicity Race Education CD41 T cell count AIDS diagnosis Overall health Health-related QOL PCS MCS HCP who Communicate Well Getting needed care Trust in HCP

Fear of Disclosure

Perception of Negative HCP Beliefs

2.10 2.03 .07 .06 2.05 2.09 .14* 2.20**

2.10 .03 .07 .13 2.11 2.02 .06 2.09

.08 .11 2.06 .07 2.03 2.06 .05 2.11

2.05 2.33** 2.26** 2.35** .25**

2.04 2.26** 2.12* 2.14* .12*

2.16** -.10 2.31** 2.33** .33**

Note. *p , .05; **p , .001; Gender: a transgender male was included in the male category and a transgender female was included in the female category; Ethnicity: non-Hispanic and Hispanic; Race: White or Caucasian, Black or African American, Asian, Native Hawaiian/ Pacific Islander, American Indian or Alaska Native, or Others; Education: None or less than high school, High school diploma or GED 5 General Education Development; AA 5 Associate of Arts or associate degree, junior, 2-year college; BA 5 Bachelor of Arts; BS 5 Bachelor of Science, bachelor’s, 4-year college degree, or graduate or professional degree; Most recent CD41 T cell count (cells/mm3): .200, #200 to .500, #500; receipt of AIDS diagnosis in the past; Perception of Negative Societal Beliefs, higher scores represent more negative perceptions; Fear of Disclosure, higher scores represent greater Fear of Disclosure; Perception of Negative Health Care Provider Beliefs, higher scores represent more negative perceptions; QOL 5 quality of life; PCS 5 physical component summary, higher scores represent better physical health-related quality of life; MCS 5 mental component summary, higher scores represent better mental health-related quality of life; health care providers (HCP) who Communicate Well, higher scores represent better communication skills (HCP informational support); Getting Needed Care, higher scores represent higher levels of perceptions that needed care was received (HCP instrumental support); Trust in HCP, higher scores represent lower levels of trust (HCP emotional support); N varies due to pair-wise deletion.

and higher Getting Needed Care scores were associated with lower scores on Perception of Negative Societal Beliefs. Only higher education levels and higher MCS scores were significant correlates of lower scores on Fear of Disclosure. Two variables, male gender and Trust in HCP (emotional support), were significant correlates of Perception of Negative HCP beliefs.

Discussion In regard to the first research question, the EFA results provided strong evidence for internal consistency reliability and construct validity of the three IHS factors. Moreover, with a possible range of mean scores of 1-5, the mean scores of all three factors below 3 (some of the time) suggested that the PLWH in our study did not report high levels of

IHS as measured by Perception of Negative Societal Beliefs, Fear of Disclosure, and Perception of Negative HCP Beliefs. The development and use of the three IHS factor scores, as compared to the original two scores (HIV Blame and HIV Disclosure), provided the foundation for distinguishing aspects of IHS, including a component related to HCP, and their relationships with other variables. The second research question focused on the correlates of IHS. Although the sample of PLWH was diverse in gender, race, ethnicity, education, and, to a lesser extent, income, only gender and education had a significant relationship with at least one component of IHS in the multivariate analysis. Although the ratio of males to females was different, with our study having more males, our study finding related to gender was consistent with the findings of Sorsdahl, Mall, Stein, and Joska (2011) in a South African study, where males were less likely than females to

8 JANAC Vol. -, No. -, -/- 2016 Table 4.

Summary of Multivariate Regression Analysis (n 5 211) Perception of Negative Societal Beliefs B SE t

Age Gender Female Male Ethnicity Non-Hispanic Hispanic Race White Black Others Education , HS 5 HS $ AA CD41 T cell count , 200 $ 200-, 500 $ 500 AIDS diagnosis No Yes Overall health HRQOL PCS MCS HCW CARE TRUST R2

Fear of Disclosure B SE

t

Perception of Negative HCP Beliefs B SE t

0.0006

0.0006

0.10

20.0086

0.0092

20.95

0.0132

0.0074

1.78

Reference 0.1092

0.1122

0.97

Reference 0.0143

0.1696

0.08

Reference 20.2760

0.1371

22.01*

Reference 0.0626

0.1518

0.41

Reference 20.1649

0.2294

20.72

Reference 20.0407

0.1856

20.22

Reference 0.1237 0.0592

0.1625 0.1664

0.76 0.36

Reference 0.1966 0.2969

0.2457 0.2516

0.80 1.18

Reference 0.3387 0.2003

0.1988 0.2035

1.70 0.98

Reference 20.3854 20.3059

0.1230 0.1383

23.13** 22.21*

Reference 20.5850 20.5000

0.186 0.2091

23.15** 22.39*

Reference 20.1948 20.0883

0.1504 0.1691

21.30 20.52

Reference 20.0624 0.0019

0.1441 0.1552

20.43 0.01

Reference 20.0637 0.1942

0.2178 0.2345

20.29 0.83

Reference 0.088 20.118

0.1762 0.1897

0.50 20.62

Reference 0.2008 20.0312

0.1107 0.0336

1.81 20.93

Reference 0.1774 20.0115

0.1674 0.0507

1.06 20.23

Reference 0.1723 0.0161

0.1354 0.0410

1.27 0.39

0.0080 0.0062 0.0301 0.0396 0.0239

20.39 23.24** 0.03 24.93** 20.35

0.0121 0.0093 0.0456 0.0598 0.0361

20.04 22.04* 20.34 21.72 20.06

0.0097 0.0076 0.0368 0.0484 0.0292

20.04 0.31 20.69 21.91 2.79**

20.0031 0.0200 0.0009 20.1950 20.0084 0.32

20.0005 20.0191 20.0156 20.1027 20.0020 0.14

20.0004 0.0024 20.0254 20.0922 0.0816 0.22

Note. *p , .05; **p , .01; SE 5 standard error; Male 5 male and transgender male; Female 5 female and transgender female; Gender: 0 5 female, 1 5 male; Race: Others 5 Asian or Pacific Islander, American Indian or Alaska Native, or others; Education: High school (HS) 5 HS or GED (general education development); $ AA 5 Associate of Arts or associate degree, junior, or 2-year college; Bachelor of Arts, Bachelor of Science, bachelor’s degree, 4-year bachelor’s, college degree, graduate or professional degree; Perception of Negative Societal Beliefs, higher scores represent more negative perceptions; Fear of Disclosure, higher scores represent higher fear levels; Perception of negative HCP (health care providers) Beliefs, higher scores represent more negative perceptions; HRQOL 5 health-related quality of life; PCS 5 physical component summary; MCS 5 mental component summary; HCW 5 health care providers (HCP) who Communicate Well, higher scores represent better communication skills (HCP informational support); CARE 5 Getting Needed Care, higher scores represent getting better needed care (HCP instrumental support; Trust in HCP, higher scores represent lower levels of Trust in HCP (HCP emotional support).

perceive IHS. However, in our study, male gender significantly influenced only one of the three IHS factor scores, Perception of Negative HCP Beliefs. On the other hand, our findings related to level of education contrasted with the Sorsdahl and colleagues (2011) finding that more educated participants were more likely to report higher levels of IHS than less

educated participants. We found that higher education level was significantly associated with lower perceived IHS as measured by Perception of Negative Societal Beliefs and Fear of Disclosure. The difference between the findings was likely related to overall differences in education level, with 83% of the South African study participants having less than a

Jang, Bakken / Internalized HIV Stigma

high school education as compared to about 36% in our New York City study. While AIDS diagnosis and health status were significant correlates of Perception of Negative Societal Beliefs in the bivariate analysis, there were no significant clinical or health status correlates of the three IHS scores in the multivariate analysis. This was also true for the PCS component of QOL, which was only significantly related to an IHS score (i.e., Perception of Negative HCP Beliefs) in the bivariate, but not multivariate analysis. The relationship was in the expected direction: those with higher PCS QOL scores reported lower Perception of Negative HCP Beliefs. Slater and colleagues (2015) reported a similar finding using a different QOL instrument, the HIV/AIDS Targeted Quality of Life instrument, which lacked an explicit mental health component, in a sample of 60 homosexual men ages 50 to 65 years living with HIV. However, our findings related to MCS QOL were significant in the multivariate analysis. PLWH with better MCS QOL reported lower Perception of Negative Societal Beliefs and Fear of Disclosure. This was consistent with Sayles and colleagues (2009), who demonstrated a significant association between total IHS score and MCS among 202 PLWH in Los Angeles County. However, in their study, IHS was not categorized into three factors as in our study. The distinction between the IHS variables and their relationships to QOL is a novel addition to the literature. In terms of social support variables, the high mean scores on HCP Communication and Getting Needed Care and low mean scores on Trust in HCP indicated that PLWH perceived high levels of HCP social support in our study. All HCP social support variables had significant relationships in the bivariate analysis. However, in the multivariate regressions, only two HCP social support variables (Getting Needed Care [instrumental support] and Trust in HCP [emotional support]) had significant relationships (i.e., higher support associated with lower IHS) with two IHS measures, Perception of Negative Societal Beliefs and Perception of Negative HCP Beliefs, respectively. The lack of a relationship between any of three IHS factor scores and informational support in the multivariate analyses was in contrast to Slater and

9

colleagues (2015), who found that receiving informational support decreased IHS. One possible explanation for this finding was that the levels of informational support were quite high in our study. Another would be that informational support was measured differently in the two studies, with our study focusing on HCP-provided informational support. The significant findings of our study must be considered in terms of its limitations. We used a convenience sampling method to recruit PLWH during the course of treatment in one HIV Special Needs Plan in New York City. Consequently, findings may not be generalizable to other samples and settings. With respect to threats to internal validity, we employed self-reported questions, and thus the potential for social desirability bias existed. However, the ACASI method of survey administration has been designed to minimize this bias.

Conclusions The existing body of literature has suggested that IHS threatens quality of care, PLWH QOL, and public health. Although levels of IHS were relatively low in our sample of PLWH from an HIV Special Needs Plan, we made novel contributions in two areas. First, the development of three IHS factor scores supports distinct examination of IHS components and may also contribute to the development and evaluation of targeted interventions for reducing IHS. Second, the conceptualization of social support from the perspective of relationship with the HCP may be particularly useful for interventions aimed at reducing IHS, particularly related to Perception of Negative HCP Beliefs in the care setting.

Disclosures The authors report no real or perceived vested interests related to this article, including grantors, or other entities whose services are related to topics covered in this manuscript, that could be construed as a conflict of interest.

10 JANAC Vol. -, No. -, -/- 2016

Key Considerations  Internalized HIV stigma (IHS) is multifaceted and includes Perception of Negative Societal Beliefs, Fear of Disclosure, and Perception of Negative Health Care Provider (HCP) Beliefs.  Interventions should consider the different aspects of IHS.  Instrumental support (Getting Needed Care) and emotional support (Trust in HCP) were associated with two aspects of IHS, Perception of Negative Societal Beliefs and Perception of Negative HCP Beliefs, respectively.  Care settings and HCP should consider implementation of interventions that increase levels of instrumental and emotional support.

Acknowledgments Data collection was supported by the Health Services and Resources Administration (New YorkPresbyterian Hospital/Select Health CCR Demonstration Project, H97HA08483, Peter Gordon, Principal Investigator). Analysis and manuscript preparation were supported by the Center for Evidence-Based Practice in the Underserved (P30NR010677, Suzanne Bakken, Principal Investigator).

References Cassel, J. (1976). The contribution of the social environment to host resistance: The Fourth Wade Hampton Frost Lecture. American Journal of Epidemiology, 104(2), 107-123. Goffman, E. (1963). Stigma: Notes on the management of spoiled identity. Englewood Cliffs, NJ: Prentice-Hall. Gonzalez, J. S., Batchelder, A. W., Psaros, C., & Safren, S. A. (2011). Depression and HIV/AIDS treatment nonadherence: A review and meta-analysis. Journal of Acquired Immune Deficiency Syndromes, 58(2), 181-187. http://dx.doi.org/10. 1097/QAI.0b013e31822d490a Gordon, P., Camhi, E., Hesse, R., Odlum, M., Schnall, R., Rodriguez, M., . Bakken, S. (2012). Processes and outcomes of developing a continuity of care document for use as a personal health record by people living with HIV/AIDS in

New York City. International Journal of Medical Informatics, 81(10), e63-e73. Hall, M. A., Zheng, B., Dugan, E., Camacho, F., Kidd, K. E., Mishra, A., & Balkrishnan, R. (2002). Measuring patients’ trust in their primary care providers. Medical Care Research and Review, 59(3), 293-318. Hargraves, J. L., Hays, R. D., & Cleary, P. D. (2003). Psychometric properties of the Consumer Assessment of Health Plans Study (CAHPS) 2.0 adult core survey. Health Services Research, 38(6 Pt 1), 1509-1527. http://dx.doi.org/10.1111/j. 1475-6773.2003.00190.x House, J. S. (1987). Social support and social structure. Sociological Forum, 2(1), 135-146. http://dx.doi.org/10.1007/ BF01107897 Kaai, S., Sarna, A., Luchters, S., Geibel, S., Munyao, P., Mandaliya, K., . Rutenberg, N. (2007). Changes in stigma among a cohort of people on antiretroviral therapy: Findings from Mombasa, Kenya. Horizons Research Summary. Nairobi: Population Council. Kinsler, J. J., Wong, M. D., Sayles, J. N., Davis, C., & Cunningham, W. E. (2007). The effect of perceived stigma from a health care provider on access to care among a lowincome HIV-positive population. AIDS Patient Care and STDS, 21(8), 584-592. http://dx.doi.org/10.1089/apc.2006. 0202 Leserman, J. (2008). Role of depression, stress, and trauma in HIV disease progression. Psychosomatic Medicine, 70(5), 539-545. http://dx.doi.org/10.1097/PSY.0b013e3181777a5f Link, B. G. (1987). Understanding labeling effects in the area of mental disorders: An assessment of the effects of expectations of rejection. American Sociological Review, 52(1), 96-112. http://dx.doi.org/10.2307/2095395 Markowitz, F. E. (1998). The effects of stigma on the psychological well-being and life satisfaction of persons with mental illness. Journal of Health and Social Behavior, 39(4), 335347. http://dx.doi.org/10.2307/2676342 Miller, A. H. (2010). Depression and immunity: A role for T cells? Brain, Behavior, and Immunity, 24(1), 1-8. http://dx. doi.org/10.1016/j.bbi.2009.09.009 Overstreet, N. M., Earnshaw, V. A., Kalichman, S. C., & Quinn, D. M. (2013). Internalized stigma and HIV status disclosure among HIV-positive Black men who have sex with men. AIDS Care, 25(4), 466-471. http://dx.doi.org/10. 1080/09540121.2012.720362 Pulerwitz, J., Michaelis, A., Weiss, E., Brown, L., & Mahendra, V. (2010). Reducing HIV-related stigma: Lessons learned from Horizons research and programs. Public Health Reports, 125(2), 272. Quality Metric Inc. (2005). How to score version 2 of the SF-12 health survey (with a supplement documenting version 1). Boston, MA: Health Assessment Lab. Rao, D., Feldman, B., Fredericksen, R., Crane, P., Simoni, J., Kitahata, M., & Crane, H. (2012). A structural equation model of HIV-related stigma, depressive symptoms, and medication adherence. AIDS and Behavior, 16(3), 711-716. http://dx.doi.org/10.1007/s10461-011-9915-0

Jang, Bakken / Internalized HIV Stigma Sayles, J. N., Hays, R. D., Sarkisian, C. A., Mahajan, A. P., Spritzer, K. L., & Cunningham, W. E. (2008). Development and psychometric assessment of a multidimensional measure of internalized HIV stigma in a sample of HIV-positive adults. AIDS and Behavior, 12(5), 748-758. http://dx.doi. org/10.1007/s10461-008-9375-3 Sayles, J. N., Wong, M. D., Kinsler, J. J., Martins, D., & Cunningham, W. E. (2009). The association of stigma with self-reported access to medical care and antiretroviral therapy adherence in persons living with HIV/AIDS. Journal of General Internal Medicine, 24(10), 1101-1108. http://dx.doi.org/ 10.1007/s11606-009-1068-8 Scambler, G., & Hopkins, A. (1986). Being epileptic: Coming to terms with stigma. Sociology of Health and Illness, 8(1), 26-43. http://dx.doi.org/10.1111/1467-9566.ep11346455 Slater, L. Z., Moneyham, L., Vance, D. E., Raper, J. L., Mugavero, M. J., & Childs, G. (2015). The multiple stigma experience and quality of life in older gay men with HIV. Journal of the Association of Nurses in AIDS Care, 26(1), 24-35. http://dx.doi.org/10.1016/j.jana.2014.06.007 Sorsdahl, K. R., Mall, S., Stein, D. J., & Joska, J. A. (2011). The prevalence and predictors of stigma amongst people living with HIV/AIDS in the Western Province. AIDS Care, 23(6), 680-685. Vyavaharkar, M., Moneyham, L., Murdaugh, C., & Tavakoli, A. (2012). Factors associated with quality of life among rural women with HIV disease. AIDS and Behavior, 16(2), 295303. http://dx.doi.org/10.1007/s10461-011-9917-y Waite, K. R., Paasche-Orlow, M., Rintamaki, L. S., Davis, T. C., & Wolf, M. S. (2008). Literacy, social stigma, and HIV medication adherence. Journal of General Internal Medicine, 23(9), 1367-1372. http://dx.doi.org/10.1007/s11606-008-0662-5 Wang, X., & Wu, Z. (2007). Factors associated with adherence to antiretroviral therapy among HIV/AIDS patients in rural China. AIDS, 21(Suppl. 8), S149-S155. http://dx.doi.org/10. 1097/01.aids.0000304711.87164.99 Wolitski, R. J., Pals, S. L., Kidder, D. P., Courtenay-Quirk, C., & Holtgrave, D. (2009). The effects of HIV stigma on health, disclosure of HIV status, and risk behavior of homeless and unstably housed persons living with HIV. AIDS and Behavior, 13(6), 1222-1232. http://dx.doi.org/10.1007/ s10461-008-9455-4

11