Transcript Open slides - CTN Dissemination Library

Relationship Power &
Sexual Risk in a Sample of
Drug-Involved Women
Aimee Campbell, MSSW
Columbia University School of Social Work
Social Intervention Group
Symposium Presentation
American Psychological Association
Annual Convention
August 16, 2008 | Boston, MA
Co-Authors
 Susan
 Shari
Tross, PhD
(Principal Investigator)
L. Dworkin, PhD
 Jennifer
Manuel, PhD
 Martina
Pavlicova, PhD
 Mei-Chen
 Edward
Hu, PhD
V. Nunes, MD
Background: Power & HIV Risk
Heterosexual transmission most prevalent
route of HIV infection for women
 Less condom use in steady relationships –
interferes with intimacy, trust, commitment
 Women traditionally at social, economic and
emotional disadvantage to male partners
 Drug use compounds HIV risk
 Limited research on how power operates
within heterosexual relationships of druginvolved women

Study Design & Methods
Secondary Analysis using Baseline Data
from NIDA Clinical Trials Network Safer
Sex Skills Building Protocol (0019)
 Sub-sample of women with main male
sexual partner
 Outcome is unprotected vaginal or anal
sex with main male partner


Data collected via audio computer-assisted
interview, Sexual Experiences & Risk Behavior
Assessment Schedule (Meyer-Bahlburg et al., 1991)
Sexual Relationship Power Scale
(SRPS)

Pulerwitz, Gortmaker, & DeJong, 2000

Derived from Theory of Gender & Power
(Connell, 1987) and Social Exchange Theory
(Emerson, 1976)

Validated with a primarily Latina community
sample

Power as a multi-dimensional construct

Relationship Control

Decision Making Dominance
SRPS Items

Relationship Control (15 items)





Consequences of asking partner to use condoms
Partner has more say
Feeling trapped or stuck
Partner gets more from relationship
Decision Making Dominance (8 items)

Who has more say about friends, whether to
have sex, what you do together, types of sexual
acts
Hypothesis

Increased relationship power would be
associated with an increase in safer sexual
practices

Operationalized as a decrease in unprotected
sexual occasions
Data Analysis

Mixed Effects Modeling (MEM) to test the
relationship between the SRPS subscales and
unprotected sexual occasions




MEM was considered optimal for analyzing effects, while
estimating random effect due to site and accommodating
for missing data
Poisson link function used in the generalized linear
model (GLM)
Partner abuse, Living arrangement, substance use,
SRPS subscales were treated as fixed effects
Age, race/ethnicity, and education included as
covariates
Sample Characteristics (N=396)
Age (yrs)
M or %
SD
38.6
8.6
Race/Ethnicity
White
African American
Hispanic/Latina
Other
56.8
24.0
9.6
9.6
< High School
High School
> High School
28.5
37.9
33.6
Education
Living w/ Main Partner
45.5
Sample Characteristics (N=396)
M or %
Current Partner Abuse
SD
29.0
Alcohol Use (# days/30)
2.0
5.1
Drug Use (max days/30)
4.9
8.8
Unprotected Sexual
Occasions (# times/90d)
22.3
32.1
SRPS – Total Score
2.8
0.6
SRPS – Relationship Control
3.0
0.6
SRPS – Decision Making
Dominance Subscale
2.5
0.7
GLM Model of Power on
Unprotected Sexual Occasions
AOR
95% CI
Relationship Control
1.24*
1.18-1.29
Decision Making
0.76*
0.73-0.80
Living Together
1.18*
1.13-1.24
Current Partner Abuse
0.85*
0.80-0.89
Alcohol Use (30 days)
1.02*
1.02-1.03
Drug Use (30 days)
1.01*
1.01-1.01
* p<0.01
Conclusion

Mixed support for the hypothesis that higher
relationship power would be associated with
reduced sexual risk behavior in drug involved
women

Decision Making Dominance purported to assess
‘power over,’ may be a better mechanism through
which women could reduce risky sexual behavior

Examining different types of relationship power
can be useful in better understanding the
association between power and sexual risk
behaviors in a sample of women in drug treatment
Limitations

Did not control for intention to use
condoms (desire to use)

Other variables may also help to explain
sexual risk behavior (e.g., partner
communication, perceived partner fidelity)

The SRPS does not include items
addressing economic or employment
variables
Future Research

Examine the association between
relationship power and sexual risk stratified
by primary substance or program type

Further refine interpersonal power
measurement tools based on contextual
and cultural factors

Further examine women’s goals around
condom use with main male partners
Thank You

We would like to acknowledge the work of
staff at the 12 treatment programs and
the participants involved in the study

Contact Information:

Aimee Campbell

[email protected]