Transcript Open slide - CTN Dissemination Library

Behavioral Intention and Partner
Type on Condom Use Among
Men in Drug Treatment
Yong S. Song, PhD, University of California, San Francisco
Donald Calsyn, PhD, University of Washington
Suzanne Doyle, PhD, University of Washington
Rhodri Dierst-Davies, MPH, Friends Research Institute
TeChieh Chen, BA, University of Washington
Background

Many substance users enrolled in treatment
continue to engage in high-risk sexual
behaviors

HIV risk reduction interventions delivered in
drug treatment programs are effective in
reducing HIV risk behaviors among drug
users
Background

Positive attitudes and expectancies about
condom use are associated with more
consistent condom use

Perceptions of HIV and STI risk are
associated with actual condom use
Purpose

Secondary outcome analysis from NIDA
CTN 018, “Safer Sex Skills for Men”

Explored the effects of the following on
Condom Use:


Behavioral Intentions to Use Condoms
Sexual Partner Type
Method
Participants: Men enrolled in a NIDA Clinical Trials Network
(CTN18) gender specific HIV-prevention intervention
• 7 MMT & 7 outpatient psychosocial programs
• Interventions: Randomly assigned to either:
• HIV-Ed: Single-session standard HIV education
intervention
• REMAS: Five-session gender-specific HIV prevention
intervention
• Measurement:
• ACASI Assessments at baseline, 3- & 6-mos f/u
Sample Demographics:
Randomized (N=590)
Age
Education
Income (Mo.)
Marital
Never
Married
Div/Wid/Sep
HIV-ED
m
(sd)
39.4
(10.4)
12.4
(2.0)
511.4 (1241.3)
n
137
57
105
(%)
(45.8)
(19.1)
(35.1)
REMAS
m
(sd)
38.7
(10.6)
12.0
(1.7)
530.8 (1114.3)
n
142
55
94
(%)
(48.8)
(18.9)
(32.3)
Sample Demographics:
Randomized
Ethnicity
Black
White
Hispanic
Native Am.
Asian
Other
HIV-ED
n (%)
91 (30.4)
173 (57.9)
29 ( 9.7)
2 ( .7)
3 (1.0)
1 ( .3)
REMAS
n (%)
74 (25.4)
169 (58.1)
37 (12.7)
6 ( 2.1)
2 ( .7)
3 ( 1.0)
Predicting Condom Use at Most Recent
Sexual Event from Intention to Use Condoms
with Regular Partners
INTENTION
RATING
RECENT
CONDOM USE
N
PHI
Z
OR
(95% CI)
Baseline
RP
3-mos
RP
377
0.30781**
15.01**
3.69
(1.91, 7.14)
Baseline
RP
6-mos
RP
226
0.24868**
11.73**
3.12
(1.62, 5.99)
3-mos
RP
6-mos
RP
231
0.36677**
27.15**
6.40
(3.18, 12.86)
RP=Regular partner
**p<.001
Predicting Condom Use at Most Recent Sexual
Event from Intention to Use Condoms with
Casual Partners
INCREASE USE
RATING
RECENT
CONDOM USE
N
PHI
Z
OR
(95% CI)
Baseline
CP
3-mos
CP
76
-0.00869
0.01
0.96
(0.37, 2.47)
Baseline
CP
6-mos
CP
75
0.18310
2.48
2.19
(0.83, 5.56)
3-mos
CP
6-mos
CP
59
0.30653*
5.32*
3.68
(1.22, 11.15)
CP = Casual Partner
* p<.05
Data Analytic Plan: Condom use
by Regular vs. Casual Partner
Primary Outcome (dependent) Variable: Condom Use with Regular or
Casual partner (Binary variable)
Approach: Repeated Measures Logistic Regression Model
Independent variables used in models:
Intervention Condition (HIV-Ed or REMAS)
Time (Baseline, 3-mos, 6-mos)
Treatment x Time
Time-varying covariate used in the model:
Frequency of vaginal and anal intercourse occasions with either
regular or casual partners at baseline, 3-mos, and 6-mos follow-up
points
Condom Use with Regular Partner
(Female & Male) N=413
Model Results of Repeated Measures Logistic Regression
Variable
Parameter Standard z-value
p
Estimate
Error
Intercept
-1.451
0.169
-8.55 <.0001
Covariate
0.007
0.002
3.42 0.0006
Treatment
-0.029
0.225
-0.13 0.8949
Time
0.117
0.099
1.17 0.2413
Treatment X Time 0.179
0.146
1.22 0.2212
Condom Use with Regular Partner
(Male & Female) N=413
Treatment x Time z= 1.22, p=0.22
Observed Proportions
35
30
25
20
HIV-Ed
REMAS
15
10
5
0
Baseline
3-Mos F/U
6-Mos F/U
Condom Use with Casual Partner
(Female & Male) N=261
Model Results of Repeated Measures Logistic Regression
Variable
Parameter Standard z-value
p
Estimate
Error
Intercept
-0.532
0.198
-2.74 0.0062
Covariate
0.012
0.003
3.65 0.0003
Treatment
0.112
0.252
0.44 0.6575
Time
1.842
0.503
3.66 0.0002
Time2
-0.932
0.253
-3.68 0.0002
Treatment X Time -1.773
0.694
-2.56 0.0106
Treatment x Time2 1.003
0.353
2.84 0.0045
Condom Use with Casual Partner
(Female & Male) N=261
Treatment x Time z=-2.56, p=0.01
Treatment x Time2 z= 2.84, p=0.004
Observed Proportions
70
60
50
40
HIV-Ed
REMAS
30
20
10
0
Baseline
3-Mos F/U
6-Mos F/U
Conclusions

Behavioral intentions to use condoms at
baseline were better predictors of condom use
with regular partners than with casual sexual
partners.

However, behavioral intentions to use condom
ratings at the follow up data points were
stronger predictors of condom use than
baseline ratings for both causal and regular
sexual partners.
Conclusions

Neither intervention appeared to affect actual
condom use with regular partners.

For casual partners, the single-session HIV-Ed
intervention appeared to have a time limited
effect on actual condom use.

For the REMAS treatment condition there was
in incremental increase in condom use with
Casual Partners at both 3- and 6-months.
Limitations

Reliability of self-report outcome data

No Post-intervention data point for
behavioral intentions to use condoms

Other factors may account for the difference
in condom use
Acknowledgements
 NIDA U10DA015815 (James Sorensen, PI)
 NIDA U10DA13714 (Dennis Donovan, PI)
We thank the CTN018 lead node team, and the 14
sites participating in this clinical trial
RRTC Teams
Site PIs
Site Coordinators
Research Assistants
Therapist & Therapist Supervisors