RELATIONSHIP DYNAMICS AND PARTNER SUPPORT

ARE ASSOCIATED WITH ART ADHERENCE IN MALAWI

Amy A. Conroy, Allison Ruark, Lynae Darbes, Torsten Neilands, Mallory O. Johnson,

Joep J.G. van Oosterhout, James Mkandawire

Adherence Conference

Miami, June 8-10, 2018

DISCLOSURES

I have no conflicts or disclosures to report.

RELATIONSHIP QUALITY AND HIV CARE ENGAGEMENT

Higher relationship quality is linked to better health.1

Positive relationship dynamics (e.g., unity, satisfaction) are associated with higher uptake of HIV testing and partner disclosure, and lower viral load.2-4.

Negative relationship dynamics (e.g., intimate partner violence) are associated with lower HIV care engagement and ART adherence.5

1. Robles et al., 2014. Psych Bull. 2. Conroy, 2013. J Sex Res. 3. Conroy & Wong, 2015. Soc Sci Med. 4. Johnson et al.,

2012. AIDS Behav. 5. Hatcher et al., 2015. AIDS.

COUPLES-BASED INTERVENTIONS FOR HIV

There is a dearth of research on relationship dynamics and adherence to ART among couples in sub-Saharan Africa

This information is necessary to identify intervention targets for couples living with HIV

In the U.S., couples-based interventions have been effective at reducing risky sex and at improving ART adherence.6-7

6. Burton, Operario, and Darbes, 2010. AIDS Behav. 7. Remien et al., 2005. AIDS.

STUDY AIMS

1. Aim 1 To describe how couple relationships affect

engagement in HIV care and treatment

2. Aim 2 To examine whether relationship dynamics and

partner support are associated with engagement in

care and treatment (i.e. ART adherence)

3. Aim 3 To develop a preliminary intervention for couples

targeting engagement in care and treatment

CONCEPTUAL FRAMEWORKContextual Variables

Structural Social Treatment-related

Relationship Variables

Communication Trust Intimacy

Satisfaction Social support

Violence/conflict Power dynamics

Dyadic capacity to work collaboratively around HIV

Adherence to ART

Individual-level Variables (Partner A)

Knowledge/beliefs

Socio-demographics Alcohol use

Dyadic interaction model

adapted from

Karney et al. 2010

Individual-level Variables (Partner B)

Knowledge/beliefs

Socio-demographics Alcohol use

CONCEPTUAL FRAMEWORK

THE UMODZI M’BANJA PROJECT

“Unity in the Family”

High rates of marriage/cohabitation8

HIV prevalence of 16%9

Test-and-treat rolled out in June of 201610

Zomba

district

8. Malawi Demographic Health Survey, 2016. 9. UNAIDS, 2014. 10. Ministry of Health of Malawi, 2016.

ELIGIBILITY

1. In a non-polygamous union for at least six months

2. Age 18 or older

3. Have at least one partner on ART for two months

(“index patient”)

4. Disclosure of HIV and ART status to primary partner

STUDY PROCEDURES

Two recruitment sites:

An urban clinic at a district hospital

(7000+ patients)

A rural clinic at a private hospital

(4000+ patients)

Index patients were recruited in waiting

rooms when attending appointments

Separate, private interviews (single visit)

MEASURES AND ANALYSIS

Outcome variables:1. Self-reported 30-day adherence (bean method)2. Partner’s estimate11 of patient’s 30-day adherence (bean method)Dichotomized as 90-100% adherence or less

Explanatory variables:1. Relationship dynamics (e.g., intimacy, trust, unity, satisfaction, commitment)2. Social support from partner (general and treatment-specific)Computed couple-level means, continuous variables

Analysis plan: Multivariable logistic regression using generalized estimating equations (GEE), controlling for the clustering of persons within couples

11. Johnson et al., 2011. JAIDS

BACKGROUND CHARACTERISTICS (N=211 COUPLES)

mean

Age 40.5 13.7%

16.6%

2.6%

none

primary

secondary

higher

Education

mean

Relationship

Duration (years)

12.5 33.2%Discordant

Concordant

Couple HIV Statusmean

Time on

ART(years)

4.84

Married

100%

Taking ART

98.9%67.1%

66.8%

N=352

PREVALENCE OF ADHERENCE TO ART

3.9%

90-100% adherence

Less than 90% adherence

4.4%

90-100% adherence

Less than 90% adherence

SELF-REPORTED

ADHERENCEPARTNER-REPORTED

ADHERENCE

95.6% 96.1%

ADJUSTED ODDS RATIOS: RELATIONSHIP DYNAMICS AND

SELF-REPORTED ADHERENCE

Unity Satisfaction Sexual satisfaction

Adjusted For:

gender, age, years of

education, household wealth

index, relationship duration,

couple HIV status, treatment

regimen complexity, and

length of time on ART

2.11

3.88

3.13

1.241.53

1.19

3.62

9.85

8.20

0.00

2.00

4.00

6.00

8.00

10.00

Results not shown:

Intimacy, trust, and

commitment were not

significantly associated

with adherence

ADJUSTED ODDS RATIOS: PARTNER SUPPORT AND

SELF-REPORTED ADHERENCE

General Partner

Support

Adjusted For:

gender, age, years of

education, household

wealth index, relationship

duration, couple HIV

status, treatment regimen

complexity, and length of

time on ARTHIV Treatment-Specific

Partner Support

1.12 1.12

1.02 1.04

1.23 1.21

0.5

0.7

0.9

1.1

1.3

1.5

ADJUSTED ODDS RATIOS: RELATIONSHIP DYNAMICS AND

PARTNER-REPORTED ADHERENCE

4.23

2.663.43

4.19

1.48 1.45 1.16 1.33

12.11

4.86

10.15

13.26

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

Intimacy Unity Satisfaction Sexual satisfaction

Adjusted For:

gender, age, years of

education, household wealth

index, relationship duration,

couple HIV status, and

length of time on ART

Results not shown:

Trust was not significantly

associated with adherence

ADJUSTED ODDS RATIOS: PARTNER SUPPORT AND

PARTNER-REPORTED ADHERENCE

General Partner

Support

Adjusted For:

gender, age, years of

education, household

wealth index, relationship

duration, couple HIV

status, treatment regimen

complexity, and length of

time on ART HIV Treatment-Specific

Partner Support

1.111.17

1.001.04

1.24

1.32

0.50

0.70

0.90

1.10

1.30

1.50

DISCUSSION

Positive relationship dynamics (unity, sexual satisfaction) were

associated with better adherence in Malawian couples

Partner social support was also associated with better

adherence; little difference between general and treatment-

specific support

Findings were consistent for both outcomes of adherence

STRENGTHS & LIMITATIONS

Limitations Cross-sectional data Self-selection bias; potentially higher functioning couples Social desirability bias in self-reported adherence (hair biomarker data

being processed)

Strengths Dyadic data collection and analysis which helps to mitigate bias from a

single partner’s report Adds to the literature on importance of couple relationships and

adherence to ART in sub-Saharan Africa

FUTURE DIRECTIONS

Interventions that incorporate relationship-

building exercises might improve

adherence to ART in couples

Longitudinal research is needed to

examine temporality, change, and

mediating pathways

ACKNOWLEDGEMENTS

Staff and graduate students

THANKS to our research participants!

Nancy Mulauzi (Project Director)

Doreen Kanyika

Chadrick Banda

Martha Tasizizyo

Issac Mponga

Sylvia Jebiwott, MPH

Melissa Saphir, PhD

Funding agencies

National Institute of Mental Health

K01-MH107331