Transcript Slides - Division of Gender, Sexuality, and Health

*
Toward a Couple-Based Intervention
for Black Men who Have
Sex With Men at Risk for
HIV/STI Transmission
HIV Center Grand Rounds
18 March 2010
Elwin Wu
Columbia University School of Social Work
http://www.socialwork.columbia.edu/sig
Towards Couple-Based Sexual Risk Reduction
for Men who Have Sex With Men (MSM)…
• Building upon the Social Intervention Group’s (SIG’s)
experience in innovation with respect to HIV prevention
and intervention…
– Project Connect
– Project Eban
– Project Connect Two
– Dissemination
of Connect
2
Targeting an At-Risk Population:
“Methamphetamine-Involved, Black MSM couples”
METH-Involved
African
American
MSM
Established
Relationship
3
Connect With Pride:
Objectives
• Primary Aims:
– To adapt an existing relationship-based HIV/STI
preventive intervention (“Connect”) and revise/refine
it for methamphetamine-involved, Black MSM couples.
– To obtain preliminary evidence establishing the potential
promise regarding the efficacy of the revised intervention
• Secondary Aim:
– To enhance the feasibility of future, larger-scale randomized
clinical trials (RCTs) testing the efficacy of the revised
intervention
4
Target Population:
“Methamphetamine-Involved, Black MSM couples”
• Main Inclusion Criteria—Men who:
– Report having a male “main partner” operationalized as:


a male with whom he has had an ongoing sexual relationship over the
prior 6 months, and
a male with whom the participant has an emotional relationship/bond
more than any person;
– Self-identify as African American/Black*;
– Report using methamphetamine at least 1 time in past 60 days*;
– Report at least 1 unprotected act of anal intercourse in the prior
60 days with a male who is not the main partner*; and
– Identify each other as their main partner
*or has male main partner who meets this criterion
5
Formative Work:
Adaptation
• Enrolled 8 couples from the target population who
participated/returned over a series of 6 focus group:
– Provide data on etiological and phenomenological issues
– Feedback and suggestions on revising content/activities
– Informed feasibility and safety/ethical issues
– Recruitment
• Community Input:
– Focus group with service providers
– Program Review Panel
– Community Advisory Board
Qualitative
Data/Findings
Wu et al. (in press). Adaptation of a couple-based HIV intervention for methamphetamine-involved African American men who have sex with men. Open AIDS Journal.
6
Vicious Cycles
Issues affecting
the relationship
Meth use
Methamphetamine
↕
use
HIV risk
Physiological &
psychological sequelae
Arguments
7
METH, HIV, & Black MSM:
Analyzing/Incorporating Experiences and Worldviews
Personal
Interactional
Risk
and
factors
protective
and
unique
structural
factors
to
Broad
cultural
values
and
• Relationship-Oriented Ecological
Perspective:
each
factors
that
impinge
MSM
thatpartners’s
are
upon
partthe
of the
belief
systems
that
shape
targets multiple levels of key
individual,
relational,
immediate
intimate
setting
history
by of
acting
asthe
and ainteract
with
all
and contextual factors that developmental
play
role in
risk
well
relationship
as
external
asanalytical
the
individual
stressors
context
in
orfactors
which
other
levels
behavior|reduction among intimate
partners
posited
sexual activity
buffers
by
on SCT
the likelihood
and
to affect
risk and
risk
of
and
protective
engaging
protective
inbehaviors
risky
behaviors
behavior
occur
Macrosystem
Exosystem
• HIV/STI knowledge
Microsystem use culture among MSM
• Sex and METH/drug
•Couple
Perceived
threat
of HIV/STIs
•Social
condom
•Ontogenic
support
fornegotiation-efficacy
risk
reduction
•Couple
Race/ethnicity:
HIV,race/ethnicity
sexual
identity
• Anti-racist•and
anti-heteronormative
perspective
(as opposed
to
•Peers/community
sexual communication
skillsand
and
sexual identity
•Couple
Condom
use self-efficacy
non-racist •and
non-heteronormative
ideologies)
•Peer
sexual
satisfaction
norms
about
the threat
of HIV
• Race-conscious
(vs.
“raceor
color-blind”)
•
Condom
use
outcome
expectancies
•Peer
Couple
sexual
decision-making
power
norms
about
safer
sex/condom
use paradigm)
• Also •contrast
to
presuming
congruent
ideology
(e.g.,
Afrocentric
• Condom use intentions
Race/ethnicity
the relationship
w/MSM partner
• •Peer
normsuse
about
meth
• Affirmation
acceptance)
ofand
different
sexual identities
•(vs.
METH
outcome
expectancies
• Impact of METH use and sexual risk on relationship 8
Theoretical Framework
• Social Cognitive Theory: identifies/specifies the mediators
or intervention’s targets of change
– Information/Knowledge: accurate appraisal of risks & risk behavior
– Outcome Expectancies: perceived costs/benefits of behaviors of
meth use & sexual risk behavior(s)

Impact of the relationship as well as impact on relationship
– Social and Self-Regulatory Skills: abilities to recognize triggers &
enact risk reduction, including reinforcement

Emphasis on couple-based regulatory skills (e.g., communication, joint
problem-solving/reinforcement)
– Self-Efficacy: belief in the ability to negotiate HIV risk reduction

Emphasis on self-efficacy w/partner as well as of the couple
– Social Support: social influences that increase or decrease risk
behaviors, as well as reciprocal effects
9
Revising the Existing Ix:
The Starting Point
Session
0
•Purpose
SCT Construct
Information
1
•Review of last session
•HIV/STI 101
•Personal vulnerability
•Speaker/listener intro
•Goal-setting
2
•Review of last session
•Myth/facts @ HIV/STIs
•Alternatives to unsafe
sex
•Pros/cons of participating •Protecting ourselves
and protecting our
Outcome
relationship
Self-Efficacy
Social Support
•Commitment contract
…
…
Expectancies
Social &
Self-Regulatory
Skills
...
Revise
•Speaker/listener practice •Speaker/listener review
•Goal-setting
•Condom use skills
•Goal-setting
•Overcoming barriers to
participating
•Taking control of life
•Communicating about
sex
•Couple-based approach
•Speaker/listener intro
•Communicating about
•Speaker/listener practice sex
•Commitment to
relationship
6
•Recap & review
•Social support network
map
•Relapse prevention &
contingency planning
•Relapse prevention &
contingency planning
… •Rewarding behaviors,
ourselves, & relationship
…
•Relapse prevention &
contingency planning
•Social support network
map
… •Relapse prevention &
contingency planning
10
Session
0
SCT Construct
•Purpose
•HIV/STI awareness
Information •METH awareness
1
•Review of last session
•HIV/STI 101
•Personal vulnerability
•Condom use skills
•Goal-setting
2
•Review of last session
•Speaker/listener intro
•Goal-setting
•Pros/cons of participating •Self-care: Protecting
ourselves and protecting
Outcome
our relationship
Self-Efficacy
Social Support
…
…
Expectancies
Social &
Self-Regulatory
Skills
...
6
•Recap & review
•Social support network
map
•Relapse prevention &
contingency planning
•Commitment contract
•Self-care: Protecting
•Communicated messages
•Relapse prevention &
ourselves and protecting about Black MSM
contingency planning
11
our relationship
•Speaker/listener practice … •Rewarding behaviors,
•Goal-setting
•Goal-setting
ourselves, & relationship
•Overcoming barriers to
participating
•Condom use skills
•Couple-based approach
•Self-care: Protecting
•Speaker/listener intro
ourselves and protecting •Speaker/listener practice
our relationship
•Communities & Black
MSM
•Speaker/listener practice
…
•Relapse prevention &
contingency planning
•Social support network
map
… •Relapse prevention &
contingency planning
11
Pilot Test of Connect With Pride
• Design:
– Recruit 30 METH-involved, Black MSM
couples and use a pre-/post-test design
to “test” the intervention
• Actual:
– Recruitment started November 2009
and stopped in April 2010
– Enrolled 34 couples (68 individuals)
– 28 of 34 (82%) couples completed
all 7 sessions of the intervention
– Follow-up assessment completed with
65 participants (96% retention), with at
least 1 partner from 100% of the couples
Recruitment
& Screening
Pre-Test
(“Baseline”)
Intervention
Delivery
2 months
Post-Test
(“Follow-Up”)
12
Pilot Test: Measures
• Primary Outcomes: Sexual risk over prior 2 months
– # of [male] sexual partners
– # of episodes of unprotected AIa with main partnerb
– % of AI episodesa with main partnerb that were condom-protected
a
Combined receptive and insertive anal intercourse (AI)
b At follow-up, if participant had broken up with main partner that was identified at
baseline, used “replacement” main partner
• Secondary Outcomes: Drug use over prior 2 months
– Frequency of METH use
– Frequency of illicit drug use
– # of different types of illicit drugs used
• Additional Measures:
– Sociodemographics (e.g., age, income, education, race/ethnicity)
– HIV status
Feasibility
13
Pilot Test: Analyses
• Hypothesis Testing: Participants will engage is less
sexual risk behavior following receipt of the
couples-based intervention
– Need to consider non-independence of measures


Level 2: Individual(s) who provide repeated measures
Level 3: Couple(s) consisting of pairs of individuals
(reporting on behaviors with each other)
– Generalized
Linear Mixed
Participant
Time 0Models (GLMM)
Time 1


Level 2 and Level 3 nesting treated as random effects
 1 = follow-up)
Time treated as a fixed 
effect (0 = baseline,
X
X

– Inference based upon the coefficient and standard error for
0
1
↕
X0 p-value)X1
the time term (and associated


– Also tested hypothesesXfor
drug
use
(secondary
outcomes)
X1
0
in an analogous manner

14
Pilot Test: Baseline Characteristics
(N = 68 individuals)
Variables
Age (yrs.)
Employed
Has ≥HS Diploma/GED
Income over past 12 mos. (× $1000)
Distribution
<x> = 41.5 (SD = 8.7)
n = 8 (12%)
n = 57 (84%)
<x> = 12.6 (SD = 9.9)
Race
African American/Black
Latino
Mixed
n = 62 (91%)
n = 5 (7%)
n = 1 (2%)
HIV-Positive
n = 64 (94%)
Used METH in past 2 mos.
n = 53 (78%)
15
Pilot Test Results: Sexual Risk Measures
# Sexual Partners
6
5
4
4.5
3
2
1.6
1
0
BL
FU
# Unprotected AI
16
14
12
10
8
6
4
2
0
13.3
2.2
BL
FU
% Protected AI
80%
70%
60%
50%
40%
30%
20%
10%
0%
72%
18%
BL
FU
Notes:
Graphs display means, with associated standard errors indicated by the rising stem
BL = Baseline (N = 68)
FU = Follow-Up (N = 65)
16
Hypothesis Testing: Sexual Risk
# Sexual Partners
# Unprotected AI
% Protected AI
Time
-2.84*** (0.70)
-11.42*** (2.07)
0.54*** (0.06)
Age (yrs.)
-0.08 † (0.05)
-0.34* (0.16)
0.005 (0.004)
Employed
-0.53 (1.19)
-2.42 (3.96)
0.04 (0.09)
Has HS diploma/GED
-1.32 (1.04)
0.78 (3.41)
-0.02 (0.08)
Annual income (× $1000)
0.04 (0.04)
-0.06 (0.13)
-0.005 (0.003)
Is Black/African
American
0.76 (1.67)
5.05 (5.44)
-0.0003 (0.13)
Is HIV-positive
-0.004 (1.64)
5.24 (5.48)
-0.29* (0.13)
1.25 (0. 86)
-3.86 (2.84)
-0.0005 (0.06)
Used METH 2 mos.
prior to enrollment
†p
< .10; *p < .05; **p < .01; ***p < .001
17
Pilot Test Results: Drug Use Measures
Illicit Drug
Use Freq.
METH Use Freq.
3.5
3.0
2.5
2.8
# Types of
Illicit Drugs Used
1.0
3.0
0.8
2.5
0.83
2.0
0.6
2.0
1.5
0.51
0.4
1.0
1.0
0.8
0.5
0.0
BL
FU
1.5
0.2
0.5
0.0
0.0
BL
FU
Frequency Scale
02.3
= Never
1 = Less than once a month
2 = Once a month
3 = 2 or 3 days a month
1.6
4 = Once a week
5 = 2 or 3 days a week
6 = 4 to 6 days a week
7 = Every Day
BL
FU
Notes:
Graphs display means, with associated standard errors indicated by the rising stem
BL = Baseline (N = 68)
FU = Follow-Up (N = 65)
18
Hypothesis Testing: Drug Use
METH Use Freq.
Illicit Drug
Use Freq.
# Types of Illicit
Drugs Used
Time
-1.85*** (0.28)
-0.31** (0.10)
-0.68** (0.26)
Age (yrs.)
-0.04 † (0.02)
-0.01 (0.01)
-0.002 (0.02)
Employed
-0.25 (0.68)
-0.30 † (0.16)
-0.38 (0.45)
Has HS diploma/GED
0.26 (0.57)
-0.03 (0.14)
-0.42 (0.39)
Annual income (× $1000)
0.01 (0.02)
0.002 (0.005)
Is Black/African
American
1.17 (0.86)
0.52* (0.23)
Is HIV-positive
1.40 (0.89)
0.13 (0.22)
†p
-0.01 (0.01)
1.07 † (0.60)
-0.17 (0.60)
< .10; *p < .05; **p < .01; ***p < .001
19
Lessons Learned
• A couple-based approach to behavioral HIV prevention
for methamphetamine-involved, Black MSM couples is
– Promising in reducing sexual risk
– Promising in reducing METH/drug use
– Feasible and safe
• Additional aspects:
– Methamphetamine → Other drugs

Polydrug use
 HIV transmission risk among users of drugs other than METH
– Use of female facilitator(s)
– Importance of affirming the worth, dignity, and rights of those
who are stigmatized, disenfranchised, and/or isolated

Especially for those as the nexus of multiple categories
20
Next Step(s)
• Dissemination of [additional] findings from
developmental and pilot test activities
• Advance the rigor of efficacy testing of the
intervention by addressing key limitations:
– No comparison/control group (↔ random assignment)
– Timecourse/sustainability
– Biological assay for STIs
– Sample size

Statistical power (also for moderators and mediators)
 Generalizability
“Dose” Analyses
21
Thank you
Nabila El-Bassel
L. Don McVinney
Robert Remien
CDC
Leona Hess
Yves-Michel Fontaine
Dale Frett
Jordan White
Community collaborators/colleagues
Study participants
22
*
Target Population:
• New
Male-to-male
MSM
York
are actually
Citysexual
is an
losing
HIV
contact
ground
epicenter
represents
in the the
largest
fight
totransmission
prevent HIVcategory
transmission!
in the US (CDC, 2008)
MSM
– 51% of all HIV/AIDS diagnoses in 2006
– 72% of male HIV/AIDS diagnoses in 2006
– Roughly 5 times that of IDU for men
– 53% of cumulative HIV/AIDS cases to date
– 68% of cumulative male HIV/AIDS cases to date
Back
Figure adapted from El-Sadr etFigure
al. (2010)
AIDS
in Hall
America—Forgotten
but
Notof
Gone.
published
at
www.nejm.org
on 300(5),
February
10,
2010 (10.1056/NEJMp1000069).
Figure
adapted
adapted
from
from
Estimates
et al. (2008)
of NewEstimation
HIV
Infections
HIV
in NEJM,
the
incidence
United
in
States,
the United
CDC
HIV/AIDS
States. JAMA,
Facts,
August
520-529.
2008
24
Need another reason for renewed or new
HIV prevention efforts for MSM?
• Among “best” or “promising”
interventions in 2009 Compendium
of Evidence-Based HIV Prevention
Interventions published by the CDC:
– Heterosexual adults: 32
– High-risk youth: 17
– Drug users: 15
– MSM: 14
Back
25
Target Population:
• African Americans are disproportionately
affected by HIV/AIDS in the U.S.
– Represent ~50% of new infections (CDC, 2008)
– 33% of MSM living with HIV/AIDS
are African American (CDC, 2008)
– 56% of African Americans living with
HIV/AIDS are MSM (CDC, 2008)
MSM
African
American
• In NY (NYCDOHMH)
– 1 in 70 New Yorkers is infected with HIV
– 1 in 40 African Americans.
– 1 in 25 men living in Manhattan.
– 1 in 12 black men age 40-49 years.
– 1 in 10 men who have sex with men.
– 1 in 8 injection drug users.
– 1 in 5 black men age 40-49 in Manhattan
Back
Figure
Table adapted
adapted from
from Hall
Hall et
et al.
al. (2008).
(2008). Estimation
Estimation of
of HIV
HIV incidence
incidencein
in the
the United
United States.
States. JAMA,
JAMA, 300(5),
300(5), 520-529.
520-529.
26
Target Population:
• MSM in more established relationships may
be at elevated risk for HIV transmission
– ~1.5 times as likely to have unprotected
anal sex w/main vs. casual partner (Sanchez et al., 2006)
– ~2.0 times as likely to have unprotected anal sex
with HIV-positive partner if the partner is
a main partner vs. casual partner (Sanchez et al., 2006)
MSM
Established
Relationship
• Majority of HIV transmissions among MSM
are from main sex partners
Back
Table adapted
Figure adapted
from Sanchez
from Sullivan
et al. (2006).
et al. (2009).
Human Estimating
Immunodeficiency
the proportion
Virus (HIV)
of HIVrisk,
transmissions
prevention,from
and testing
main sex
behaviors…Men
partners amongwho
menhave
whosex
have
with
sexmen,
with November
men in five2003-April
US cities. AIDS,
2005. MMWR,
23(9),1153-1162.
55(SS06),1-16.
27
Target Population:
• Methamphetamine use
– Fueling HIV transmission among MSM:




↑ engagement in sexual risk behavior
↑ use by HIV-positive
Impairs immune system (Tallóczy et al., 2008)
↑ viral load (Ellis et al., 2003)
Drug-Involved
MSM
African
American
– Surveys indicate less use among
African American MSM, but…

Estimates of recent meth use among African American MSM
range from 10-16%
 49% of African American MSM who used club drugs
reported meth use (Halkitis et al., 2005)
Back
Table adapted from Halkitis et al. (2005). Explanations for methamphetamine use among gay and bisexual men in New York City. Substance Use & Misuse, 40(9),1331-1345.
28
Pilot Test: Measures cont.
• Feasibility
– Recruitment


Eligibility with respect to population
Eligibility/yield with respect to outreach parameters
(e.g., venue, time, active vs. passive)
– Retention


Intervention attendance
Follow-up/loss rates
– Adverse events


Nature
Frequency
– Participant satisfaction
Back
29
“Meth helps me get my freak on”
-KC [FG5]
• Sexualizing and sexual
stereotypes of Black/African
American men
Macrosystem
Exosystem
Microsystem
Ontogenic
– Dealing with objectification

“He likes me because I have
a big…” -KC [FG3]
– Masculine/“manly” norms vs. being the/a receptive partner

“It’s assumed that you are less of a man when you are the
bottom…yes it’s an issue for black gay men.” – KA [FG4]
• Methamphetamine:
– Psychological disinhibition and cognitive dissociation
– Perceived aphrodisiac effects (intensity, prolongation)
– Physiological effects: Booty bumping, “instant bottoms”
30
“We [Black MSM] are always on
the outside looking in” -SM [FG3]
• Constant vigilance negotiating
and/or projecting identity
Macrosystem
Exosystem
Microsystem
– “Where we used to live,
we definitely couldn’t hold
hands. As soon as we got out
of the subway, that was it.” -AK [FG2]
– “Those homothugs you see in Chelsea, I can’t stand that.”
Ontogenic
– MS [FG1]
• Marginalization/stereotyping
– “What he’s [partner] got is the most important thing: education.” -SM [FG1]
– “Those homothugs aint ever going to amount to nothing” -MS [FG1]
• Methamphetamine:
– Seen as “white” [gay] men’s drug → distancing from “homothugs”
– Higher price of meth (vs. cocaine, marijuana) → SES status
31
“We [Black male couples] are invisible…
it’s as if we don’t exist” -MC [FG5]
• Isolation
– Couples
Macrosystem
Exosystem
Microsystem
“Once we started seeing each
other, that was it…it was just
the two of us.” -KA [FG3]
 “Once we [black gay men] find the
perfect guy, we’re not going to let someone take him away.” -MS [FG4]

Ontogenic
– Community
“We don’t air dirty laundry.” -KA [FG2]
 “What goes on in the house stays in the house.” -SM [FG2]

• Methamphetamine:
– PNP/sex venues as means to connect to/socialize with others
– METH can be “blamed” rather than sex and/or relationship issues
Back
32
Additional Analyses:
Intervention “Dose”
Measure at Follow-Up
Completers
(n=11)
(n=54)
p-value a
# of Partners
<x> = 2.37
SE = 0.61
<x> = 1.41
SE = 0.15
.93
# Unprotected AI
<x> = 5.27
SE = 4.00
<x> = 1.57
SE = .41
.03
% Protected AI
<x> = .31
SE = .16
<x> = .79
SE = .04
.01
METH Use Freq.
<x> = .50
SE = .34
<x> = .89
SE = .27
.11
Illicit Drug Use Freq.
<x> = .62
SE = .19
<x> = .49
SE = .07
.82
<x> = 2.00
SE = 0.54
<x> = 1.57
SE = 0.18
.54
# Types of Illicit Drugs Used
a
Non Completers
p-values from Completed × Time term in GLMM with covariance adjustment as before
Back
33
Additional Analyses:
Intervention “Dose”
Measure at Follow-Up
Completers
(n=11)
(n=54)
p-valuea
# of Partners
<x> = 2.37
SE = 0.61
<x> = 1.41
SE = 0.15
.14
# Unprotected AI
<x> = 5.27
SE = 4.00
<x> = 1.57
SE = .41
.34
% Protected AI
<x> = .31
SE = .16
<x> = .79
SE = .04
.01
METH Use Freq.
<x> = .50
SE = .34
<x> = .89
SE = .27
.69
Illicit Drug Use Freq.
<x> = .62
SE = .19
<x> = .49
SE = .07
.49
<x> = 2.00
SE = 0.54
<x> = 1.57
SE = 0.18
.50
# Types of Illicit Drugs Used
a
Non Completers
p-values from Mann-Whitney U test
Back
34
*
Connect With Pride:
Session 0 - Orientation
• Welcome and introduction to the study
• Getting to know the couple
• Expectations and Ground Rules for respect
and safety
• HIV/STI and METH Awareness
• Closing
36
Connect With Pride:
Session 1 – Risk Reduction as Self-Care
• Welcome & Check-In/Review
• Self Care: Introduction
• Self-Care: Sexual Health & HIV/STIs
– HIV/STI Education
– Risk level identification
– Condom use skills
• Goal Setting
• Closing
37
Connect With Pride:
Session 2 – METH & HIV
• Welcome & Check-In/Review
• METH Knowledge Assessment
• The AURA of METH Use
• METH Harm Reduction
• METH and the Couple
• Goal Setting and Homework
• Closing
38
Connect With Pride:
Session 3 – Couple Communication
• Welcome & Check-In/Review
• Communicated Messages About BGM
• Communication Styles and Techniques
• Strengthening Couple’s Communication Skills
• Goal Setting and Homework
• Closing
39
Connect With Pride:
Session 4 – Couple Sex Negotiation
• Welcome & Check-In/Review
• Relationship Myths
• Unwritten Rules: Power, &
Decision-Making in the Relationship
• Couple’s Sex History and Lives
• Becoming Safer Sex Gourmets
• Goal Setting and Homework
• Closing
40
Connect With Pride:
Session 5 – METH, Drugs, & Alcohol
• Welcome & Check-In/Review
• Perceptions about drug use
• Power and Decision-Making in Relationships:
A Focus on Drug Involvement
• Triggers for Drug Use and Unsafe Sex
• Problem Solving Around Triggers
• Goal Setting and Homework
• Closing
41
Connect With Pride:
Session 6 – Connecting With Pride
• Welcome & Check-In/Review
• Social Support
• Strengthening Positive Support
• Relapse Prevention
• Rewards for Healthy
Behaviors
• Commitment to Future Healthy Goals
• Closing Ceremony
42
*
Pilot Test: Analyses
• Hypothesis Testing: Participants will engage is less
sexual risk behavior following receipt of the
couples-based intervention
– Need to consider non-independence of measures


Level 2: Individual(s) who provide repeated measures
Level 3: Couple(s) consisting of pairs of individuals
(reporting on behaviors with each other)
– Generalized Linear Mixed Models (GLMM)


Level 2 and Level 3 nesting treated as random effects
Time treated as a fixed effect (0 = baseline, 1 = follow-up)
– Inference based upon the coefficient and standard error for
the time term (and associated p-value)
– Also tested hypotheses for drug use (secondary outcomes)
in an analogous manner
Participant
Time0
Time1
44
Pilot Test: Analyses
• Hypothesis Testing: Participants will engage is less
sexual risk behavior following receipt of the
couples-based intervention
– Need to consider non-independence of measures


Level 2: Individual(s) who provide repeated measures
Level 3: Couple(s) consisting of pairs of individuals
(reporting on behaviors with each other)
– Generalized Linear Mixed Models (GLMM)


Level 2 and Level 3 nesting treated as random effects
Time treated as a fixed effect (0 = baseline, 1 = follow-up)
– Inference based upon the coefficient and standard error for
the time term (and associated p-value)
– Also tested hypotheses for drug use (secondary outcomes)
in an analogous manner
Participant
Time 0
Time 1
45
Pilot Test: Baseline Characteristics
(N = 68 individuals)
Age (yrs.)
Employed
Has ≥HS Diploma/GED
Income over past 12 mos. (× $1000)
<x> = 41.5 (SD = 8.7)
n = 8 (12%)
n = 57 (84%)
<x> = 12.6 (SD = 9.9)
Race
African American/Black
Latino
Mixed
n = 62 (91%)
n = 5 (7%)
n = 1 (2%)
HIV-Positive
n = 64 (94%)
Used METH in past 2 mos.
n = 53 (78%)
46
Hypothesis Testing: Sexual Risk
# Sexual Partners
# Unprotected AI
% Protected AI
Time
-2.84*** (0.70)
-11.42*** (2.07)
0.54*** (0.06)
Age (yrs.)
-0.08† (0.05)
-0.34* (0.16)
0.005 (0.004)
Employed
-0.53 (1.19)
-2.42 (3.96)
0.04 (0.09)
Has HS diploma/GED
-1.32 (1.04)
0.78 (3.41)
-0.02 (0.08)
Annual income (× $1000)
0.04 (0.04)
-0.06 (0.13)
-0.005 (0.003)
Is Black/African
American
0.76 (1.67)
5.05 (5.44)
-0.0003 (0.13)
Is HIV-positive
-0.004 (1.64)
5.24 (5.48)
-0.29* (0.13)
1.25 (0. 86)
-3.86 (2.84)
-0.0005 (0.06)
Used METH 2 mos.
prior to enrollment
†p
< .10; *p < .05; **p < .01; ***p < .001
47
Hypothesis Testing: Drug Use
METH Use Freq.
Illicit Drug
Use Freq.
# Types of Illicit
Drugs Used
Time
-1.85*** (0.28)
-0.31** (0.10)
-0.68** (0.26)
Age (yrs.)
-0.04† (0.02)
-0.01 (0.01)
-0.002 (0.02)
Employed
-0.25 (0.68)
-0.30† (0.16)
-0.38 (0.45)
Has HS diploma/GED
0.26 (0.57)
-0.03 (0.14)
-0.42 (0.39)
Annual income (× $1000)
0.01 (0.02)
0.002 (0.005)
Is Black/African
American
1.17 (0.86)
0.52* (0.23)
Is HIV-positive
1.40 (0.89)
0.13 (0.22)
†p
-0.01 (0.01)
1.07† (0.60)
-0.17 (0.60)
< .10; *p < .05; **p < .01; ***p < .001
48
Additional Analyses:
Intervention “Dose”
Measure at Follow-Up
Completers
(n=11)
(n=54)
p-valuea
# of Partners
<x> = 2.37
SE = 0.61
<x> = 1.41
SE = 0.15
.93
# Unprotected AI
<x> = 5.27
SE = 4.00
<x> = 1.57
SE = .41
.03
% Protected AI
<x> = .31
SE = .16
<x> = .79
SE = .04
.01
METH Use Freq.
<x> = .50
SE = .34
<x> = .89
SE = .27
.11
Illicit Drug Use Freq.
<x> = .62
SE = .19
<x> = .49
SE = .07
.82
<x> = 2.00
SE = 0.54
<x> = 1.57
SE = 0.18
.54
# Types of Illicit Drugs Used
a
Non Completers
p-values from Completed × Time term in GLMM with covariance adjustment as before
Back
49
Additional Analyses:
Intervention “Dose”
Measure at Follow-Up
Completers
(n=11)
(n=54)
p-value
# of Partners
<x> = 2.37
SE = 0.61
<x> = 1.41
SE = 0.15
.14
# Unprotected AI
<x> = 5.27
SE = 4.00
<x> = 1.57
SE = .41
.34
% Protected AI
<x> = .31
SE = .16
<x> = .79
SE = .04
.01
METH Use Freq.
<x> = .50
SE = .34
<x> = .89
SE = .27
.69
Illicit Drug Use Freq.
<x> = .62
SE = .19
<x> = .49
SE = .07
.49
<x> = 2.00
SE = 0.54
<x> = 1.57
SE = 0.18
.50
# Types of Illicit Drugs Used
a
Non Completers
p-values from Mann-Whitney U test
Back
50