An Intervention to Reduce HIV-Related Stigma Among African American Communities in South Carolina

1
An Intervention to Reduce HIV-Related Stigma
Among African American Communities in South
Carolina
Presentation at the 2005 National HIV Prevention
Conference

• John B. Pryor, Ph.D.
• Department of Psychology
• Illinois State University

Bambi Gaddist, DrPH South Carolina HIV/AIDS
Council
Letitia Johnson-Arnold, MSPH South Carolina
HIV/AIDS Council
6/13/2005, 330-500 PM Session
NumberM3-C17-05, LocationEmbassy-Hong
Kong Funded by the Academy for Educational
Development
2
Outline of Todays Talk

• Background - HIV in South Carolina
• South Carolina HIV/AIDS Council
• General Survey 2003-2004
• Some comparisons of the General Survey to a
  national probability sample
• A conceptual model of stigma
• An Intervention Evaluation
• No evidence for an immediate impact of the
  intervention
• Implications of intervention survey for factors
  affecting stigma avoidance
• Some connections between stigma and prevention
  behavior

3
Background

• Approximately 201 of every 100,000 adults and
  adolescents in South Carolina are living with HIV
  and another 185 per 100,000 are living with AIDS
  (CDC, 2003).
• The number of AIDS cases among African Americans
  in South Carolina is higher than that among any
  other ethnic/racial group.
• The prevalence among Blacks is ten times that of
  Whites in South Carolina (CDC, 2004).

4
South Carolina HIV/AIDS Council

• Three-prong initiative to reduce HIV-related
  stigma among African American communities
  throughout South Carolina
• Organizing educational town hall meetings in
  rural communities
• Developing and staging an educational theatre
  production concerning HIV/AIDS stigma and
  discrimination
• A statewide media campaign

5
General Survey 2003-2004

• Survey instrument was based upon national
  telephone surveys conducted by Herek and his
  colleagues in 1991, 1997, 1999 (Herek,
  Capitanio, Widaman, 2002)
• SCHAC General Survey sample was recruited from
  attendees at town hall meetings and theatrical
  performances organized by SCHAC
• Herek 1999 national probability sample 669
  adults
• SCHAC 2003-2004 sample 403 African American
  adults in South Carolina

6
more avoidance
SCHAC survey items not in Herek survey
Bars in pink represent 95 confidence intervals
for national probability survey
7
Bars in pink represent 95 confidence intervals
for national probability survey
8
more misconceptions
Bars in pink represent 95 confidence intervals
for national probability survey
9
Bars in pink represent 95 confidence intervals
for national probability survey
10
Bars in pink represent 95 confidence intervals
for national probability survey
11
Summary

• Compared to the national probability sample, the
  SCHAC sample of African American adults in SC
• Displayed more avoidant intentions in 2 out of 3
  measures
• Were less likely to blame PLWHA
• Displayed more misconceptions about transmission
  in 3 out of 5 measures
• Indicated much less comfort with PLWHA in only 1
  out of 4 measures
• Indicated somewhat less negative emotions in 2
  out of 3 measures

12
A conceptual model of HIV-related stigma
13
Conceptual Model of Psychological Reactions to
Stigma
Cognitive Responses
Approach/ Avoidance Behaviors
Emotional Reactions
14
Components of Stigma in General Survey
Cognitive
Affective
Behavioral
Blame
Avoidance of PLWHA
Comfort- PLWHA
Belief in Casual Contact
Negative Emotions- PLWHA
15
Hierarchical Multiple Regression
Predicting Avoidance Intentions
STUDY 1
Step 1
Step 2
Blame
Comfort- PLWHA
ß.26
Avoidance of PLWHA
ß-.43
ß.29
Belief in Casual Contact
Negative Emotions- PLWHA
ß.34
R2.25, F(2,367)59.65, plt.01
R2.49, Fchange(2,365)85.46, plt.01
plt.05
16
Hierarchical Multiple Regression
Predicting Avoidance Intentions
STUDY 1
Step 1
Step 2
Blame
Comfort- PLWHA
ß-.48
Avoidance of PLWHA
ß.04
ß.11
Belief in Casual Contact
Negative Emotions- PLWHA
ß.30
R2.48, F(2,367)166.76, plt.01
R2.49, Fchange(2,365)3.51, plt.01
plt.05
17
Summary

• Avoidance intentions are related to both
  cognitive and affective components of stigma
• Affective components seem to account for more
  unique variance in avoidance intentions

18
An intervention evaluation
19
Intervention

• Play about HIV infection in African American
  families
• Themes
• Forgiveness
• Family support
• Tolerance
• It could happen to you

20
Design of Intervention Evaluation
Survey
Play
Discussion
Recruitment From the African American Community
Delay
Play
Discussion
Survey
21
Constructs Measured in the Intervention Survey

• Empathy for persons living with HIV/AIDS (PLWHA)
  (?.82)
• Negative emotions for PLWHA (?.77)
• Blame (?.50)
• Beliefs in transmission via causal contact
  (?.82)
• Avoidance tendencies regarding PLWHA (?.82)
• Attitudes about being tested (?.89)
• Motivation to control prejudice regarding PLWHA
  (?.79)
• Superstitious contagion beliefs (??)
• Attitudes toward MSM and WSW (?.87, ?.93)
• Personal contact with MSM and WSW (??)
• Positive religious beliefs regarding PLWHA
  (?.67)
• Support for coercive social policies (?.80)

22
Summary of the Analyses from the Intervention
Survey

• There were no statistically significant
  differences between the two intervention
  conditions across any of the constructs
• Participants in intervention study were from
  Richland County (pop334,609)
• 3/4 of General Survey sample were from counties
  with less than 100,000 population
• Comparisons between measures common across the
  Intervention Survey and the General Survey
  revealed that participants in the intervention
  blamed PLWHAs less than people from counties less
  than 100,000 they also indicated weaker
  avoidance intentions than people from counties
  less than 100,000
• Ironically, participants in the intervention
  displayed certain negative emotions more than all
  others from the General Survey

23
Examining the HIV-related stigmaconceptual
model
24
Components of Stigma in Intervention Survey
Cognitive
Affective
Behavioral
Blame
Avoidance of PLWHA
Empathy- PLWHA
Belief in Casual Contact
Negative Emotions- PLWHA
25
Hierarchical Multiple Regression
Predicting Avoidance Intentions
STUDY 2
Step 1
Step 2
Blame
Empathy- PLWHA
ß.22
Avoidance of PLWHA
ß-.23
ß.33
Belief in Casual Contact
Negative Emotions- PLWHA
ß.07
R2.06, F(2,72)2.16, P.12
R2.19, Fchange(2,70)5.85, plt.01
plt.05 plt.07
26
Hierarchical Multiple Regression
Predicting Avoidance Intentions
STUDY 2
Step 1
Step 2
Blame
Empathy- PLWHA
ß-.24
Avoidance of PLWHA
ß.05
ß-.01
Belief in Casual Contact
Negative Emotions- PLWHA
ß.35
R2.19, F(2,72)8.41, plt.01
R2.19, Fchange(2,70)lt1, NS
plt.05
27
Summary

• Affective components were more strongly related
  to variance in avoidance intentions than
  cognitive components of stigma

28
Why should prevention researchers be concerned
with HIV-related stigma?
29
Hierarchical Multiple Regression
Predicting Attitudes toward HIV Testing
STUDY 2
Step 1
Step 2
Blame
Empathy- PLWHA
ß-.30
Attitude Toward HIV Testing
ß.38
ß.01
Belief in Casual Contact
Negative Emotions- PLWHA
ß.11
R2.09, F(2,77)3.34 Plt.05
R2.22, Fchange(2,75)6.21, plt.01
plt.05
30
Hierarchical Multiple Regression
Predicting Attitudes toward HIV Testing
STUDY 2
Step 1
Step 2
Blame
Empathy- PLWHA
ß.43
Attitude Toward HIV Testing
ß.-.19
ß.06
Belief in Casual Contact
Negative Emotions- PLWHA
ß-.04
R2.19, F(2,77)8.88, plt.01
R2.22, Fchange(2,75)1.47, NS
plt.05
31
Summary

• Empathy for PLWHAs was a relatively strong
  predictor of attitudes toward testing
• People who felt empathy/compassion for PLWHAs
  held more positive attitudes toward being tested
  themselves or encouraging their family friends
  to be tested

32
Future Directions

• Our analyses of survey data suggest that
  anti-stigma interventions might focus upon
  encouraging empathy for PLWHAs
• Interventions that encourage empathy/compassion
  for PLWHAs may also have an impact upon attitudes
  toward HIV testing
• A focus upon empathy/compassion may be well
  received in faith communities

33
Special thanks to the following people who helped
in data analyses for this project

• Jamie Hughes
• Leah Pryor

34
Correlation Matrix from Intervention Study
Correlations in yellow are statistically
significant, p lt .05