Dually enacted stigma among young people and
their caregivers living with HIV:
Challenges and opportunities to reaching
90-90-90 in Zambia
Joseph G. Rosen, Edith S. Namukonda Mwelwa Chibuye,
Lyson Phiri, Bwalya Mushiki, Michael T. Mbizvo, Nkomba Kayeyi
13th INTEREST Conference – Accra, Ghana 14 May 2019
Disclosures
The authors confirm they have no personal or
financial conflicts of interest to disclose.
2
Stigma profoundly disrupts uptake of and
engagement in HIV treatment services
• Threatens progress to
achieving UNAIDS 90-90-90
targets [Levi et al. 2016]
• Young people particularly
vulnerable to stigma in
schools, facilities, and other
community venues [Surkan et
al. 2010, Messer et al. 2010,
Deacon and Stepheny 2007]
3
Need to characterize stigma
within HIV-affected households
to inform more responsive care
and treatment services
Study design
• Analysis of baseline data from 2-year prospective cohort study
▪ ZAMFAM beneficiaries (Central Province)
▪ Non-beneficiary comparison (Eastern Province)
• Assess changes in socioeconomic well-being, health, and HIV service use
4
Measure prevalence and factors
associated with stigma among
YPLHIV and their HIV+ caregivers
Study population from HIV-affected households
5
Caregivers YPLHIV
• Primary guardians of
YPLHIV in household
• Aged 18 or older
• ZAMFAM beneficiaries
(Central Province only)
• Self-reported HIV+
• Aged 5 to 17
• ZAMFAM beneficiaries
(Central Province only)
• Newly diagnosed or linked to ART
(Eastern Province only)
• Know HIV status
(ages 10-17 only)
Sampling strategy (July–October 2017)
• Different sampling approach by province
▪ Central
• 2-stage sampling proportional to YPLHIV population estimate
▪ Eastern
• Convenience sampling using existing registries
• Sampling in districts comparable to Central Province sites by population size and urban/rural locale
• Aimed to recruit comparable numbers of households across provinces
Central
Wards (urban/rural)
were identified
Households sampled within
each ward
Eastern
Registers scanned for
eligible YPLHIV
Households approached for
pre-consent
6
Stigma was operationalized from reports
of any of the following experiences
7
Survey Items [MEASURE Evaluation, 2013] Caregivers
YPLHIV
5-9 10-14 15-17
Treated badly at work or lost work ✓
Treated badly at school or excluded from
school activities✓ ✓ ✓ ✓
Difficulty finding partners ✓ ✓
Family did not provide care when fell ill ✓ ✓ ✓ ✓
Treated badly by health providers ✓ ✓ ✓ ✓
Lost friends ✓ ✓ ✓ ✓
Treated badly by family members or excluded
from family activities✓ ✓ ✓ ✓
Experienced a break-up of a relationship ✓ ✓
Treated like a social outcast by community ✓ ✓ ✓ ✓
Experienced physical violence ✓ ✓ ✓ ✓
Treated badly at church ✓ ✓ ✓ ✓
Categorical outcome variable constructed from
reported experiences with HIV-related stigma
8
DUALLY ENACTED STIGMA
DISCORDANT HOUSEHOLD
STIGMA
NO HOUSEHOLD STIGMA
Both YPLHIV and caregiver
reported at least one
stigma experience
Only one (YPLHIV or
caregiver) reported any
stigma experience
Neither YPLHIV nor
caregiver reported any
stigma experience
Data analysis
● Ordered logistic regression identified correlates of stigma
● Covariates meeting a significance threshold of p<0.1 were introduced into multivariate models• Adjusted for age group, sex, and
province
● Variance Inflation Factor (VIF) scores were calculated to assess covariate saturation
● Data were managed and analyzed in Stata 14.2
9
Recruitment, response, and sample
inclusion rates (%), by province
10
417
264
160
335
264
150
CE
NTR
AL P
RO
VIN
CE
EA
STE
RN
PR
OV
INC
E
79%
57%
63%
61%
Recruitment
Response
Sample
Sample sociodemographics aggregated by
province at baseline, by participant subgroup
11
Caregivers
(n=310)
90% female
Median age: 40 years
Marital status: 56%
married or in union,
26% widowed
5–9 years
(n=143)
53% female
Median age: 7 years
School enrollment*:
77% females,
52% males
60% female
Median age: 13 years
School enrollment:
80% females,
85% males
10–17 years
(n=167)
*Children in Zambia start school at 7 years, so only children 7 and older were included in this
analysis
YPLHIV
One-fourth (28%) of households reported any
stigma experience (7.1% were dually enacted)
12
0 2 4 6 8 10 12 14
Treated badly at church
Experienced physical violence
Treated like outcast by community
Partner dissolved relationship
Treated badly or excluded by family
Lost friends
Treated badly by health providers
Family didn't provide care when ill
Difficulty finding partners
Treated badly or excluded at school
Treated badly at or lost work
Percent (%)
HIV+ caregivers
YPLHIV
N/A
YPLHIV associations with discordant and dually
enacted stigma (N=310)
13
Variables
Categorical stigma (%) Adjusted*
None
(n=205)
Discordant
(n=83)
Dual
(n=22) pOR 95% CI
Continuous age, in years
(mean + SD)10.1 ± 3.4 10.9 ± 3.7 11.0 ± 3.7 1.08 1.01–1.16
Sex
Male
Female
39.0
61.0
50.6
49.4
54.6
45.4
1.00
0.59
Ref.
0.36–0.97
Stopped ART for 30+ cont.
days, past year (n=294)7.3 9.9 27.3 3.89 1.25–12.10
Too sick for daily activities,
past month27.3 46.7 36.4 1.89 1.13–3.15
*Model adjusted for YPLHIV continuous age, sex, province, and other covariates presented in the table
Caregiver associations with discordant and
dually enacted stigma (N=310)
14
*Model adjusted for caregiver age group, sex, province, and other covariates presented in table
Variables
Categorical stigma (%) Adjusted*
None
(n=205)
Discordant
(n=83)
Dual
(n=22) pOR 95% CI
Spent a whole day/evening
without eating at least once
weekly, past month
44.4 54.2 50.0 1.85 1.04–3.31
Female decision-making
authority over any household
financial expense (n=293)
62.9 45.8 68.2 0.73 0.43–1.24
Factors associated with discordant and dually
enacted stigma in sero-concordant households
Variables Caregivers YPLHIV
Older age ✓
Male sex ✓
Non-adherence to ART ✓
Severe illness ✓
Food insecurity ✓
15
✓ statistically significant (p<0.05) correlates of discordant and dually enacted stigma in
multivariate ordered logistic regression analysis
Conclusions and recommendations
• Stigmatizing experiences reported primarily in food-insecure, socially vulnerable households
• Few households reported dually enacted stigma (<10%)
• Stigma experiences may be associated with suboptimal treatment outcomes among YPLHIV
• Novel stigma mitigation approaches addressing households, not just individuals, should be explored
16
Study limitations
• Cross-sectional design
• Dichotomized measures of inherently different stigma experiences
• Small sample size
• Different sampling procedures by province
• Social desirability and response biases
• Self-reported HIV status of caregivers
17
Acknowledgments
• PEPFAR Zambia
• Zambia Ministry of
Health
• Development Aid from
People to People (DAPP)
• Provincial and District
Health Offices (Eastern
Province)
• Study enumerators
18
Project SOAR (Cooperative Agreement AID-OAA-A-14-00060) is made possible
by the generous support of the American people through the United States
President’s Emergency Plan for AIDS Relief (PEFPAR) and United States
Agency for International Development (USAID). The contents of this
presentation are the sole responsibility of Project SOAR and Population
Council and do not necessarily reflect the views of PEPFAR, USAID, or the
United States Government.
Through operations research, Project SOAR will determine how best to
address challenges and gaps that remain in the delivery of HIV and AIDS care
and support, treatment, and prevention services. Project SOAR is producing a
large, multifaceted body of high-quality evidence to guide the planning and
implementation of HIV and AIDS programs and policies. Led by the Population
Council, Project SOAR is implemented in collaboration with Avenir Health,
Elizabeth Glaser Pediatric AIDS Foundation, Johns Hopkins University,
Palladium, and The University of North Carolina.
Thank You
NEWS
RESOURCESRESEARCH
EVENTS
Visit projsoar.org!
19
