HIV Testing, Stigma, Social
Conditions
Keywords
HIV stigma
Brief summary
This project will assess whether a digital crowdsourced intervention can reduce HIV stigma and promote HIV self-testing among adolescents and young adults (AYA). NIH has emphasized the need for research on interventions to reduce HIV-associated stigma and its impact on the prevention and treatment of HIV/AIDS, particularly in low and middle income countries (LMICs). The proposed study has the highest public health significance: it uses a community-based participatory approach to engage local AYA to develop a digital crowdsourced HIV stigma reduction and self-testing intervention to reduce HIV stigma and increase HIV testing. Study Aim 1: To develop a crowdsourced digital HIV stigma reduction and self-testing intervention targeting AYA in Kazakhstan. Using a community-based participatory approach that engages local adolescents and young adults (AYA) and youth organizations, we will launch a national crowdsourcing contest in which AYA will design multimedia content to reduce HIV stigma in order to promote HIV testing among peers. Study Aim 2: To pilot test this crowdsourced HIV stigma reduction and self-testing intervention in a preliminary efficacy trial. We will assess the intervention's feasibility and acceptability and obtain preliminary estimates of its effects on decreasing HIV stigma (primary outcome) and increasing HIV testing (secondary outcome) among AYA in Kazakhstan who received the intervention compared to individuals who did not. Participants (n=168) will be randomized 1:1 to: 1) receive the winning multimedia crowdsourced HIV stigma reduction content and a link for HIV self-testing, or 2) receive standard Kazakhstan Ministry of Health HIV informational materials and a link for HIV self-testing.
Detailed description
This project will assess whether a digital crowdsourced intervention can reduce HIV stigma and promote HIV self-testing among adolescents and young adults (AYA). Globally AYA are at increased risk for HIV acquisition. In Eastern Europe and Central Asia (EECA), new HIV infections among 15-24 year old AYA are expected to increase 28% by 2030. In Kazakhstan, one in four HIV infections occur among AYA. Despite the growing HIV burden among AYA in Kazakhstan, this population has some of the lowest HIV testing rates in the country, largely due to stigma. Yet few efforts in Kazakhstan address HIV stigma and the role it plays as a barrier to HIV testing. Digital technologies and crowdsourcing campaigns (i.e., engaging groups of AYA online to address public health challenges and share solutions) are scalable, cost-effective tools that can increase HIV testing services and reduce HIV stigma in low and middle income countries (LMICs) and other resource- constrained settings. Crowdsourcing may be particularly successful among AYA, given their high levels of social media use and technological literacy. Complementing the crowdsourcing approach, mailing HIV rapid test kits can enable AYA to avoid the stigma associated with attending the AIDS Center and overcome transportation barriers.
Interventions
BEHAVIORALDigital Crowdsourced Intervention to Reduce HIV Stigma among Adolescents and Young Adults
Adolescents and young adults will participate in a crowdsourcing contest to create digital materials to reduce HIV stigma in order to increase HIV testing. Entries will be judged by a community judging panel and expert judging panel. Winning entries will be selected for the digital intervention and presented to participants in the intervention arm.
Sponsors
National Institutes of Health (NIH)
Fogarty International Center of the National Institute of Health
Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD)
Columbia University
Study design
Allocation
RANDOMIZED
Intervention model
SINGLE_GROUP
Primary purpose
DIAGNOSTIC
Masking
SINGLE (Subject)
Intervention model description
Participants will be randomized to receive the digital crowdsourced intervention or standard of care (existing HIV informational materials). We will examine effects of the intervention on reducing HIV stigma and increasing HIV testing.
Eligibility
Sex/Gender
ALL
Age
16 Years to 24 Years
Healthy volunteers
Yes
Inclusion criteria
* Age 16-24 * Report previous sex with another individual * Reside in Kazakhstan
Exclusion criteria
* Not willing to provide consent or not able to understand study procedures
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Mean Score Change in Total HIV Stigma Scale | Baseline and 3 months | The scale/questionnaire is comprised of 17 items on a 5-point Likert scale (score range from 17-85). Higher scores indicate higher levels of stigma. Outcome will be assessed in mean score change from baseline to 3 months follow-up. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Mean Change in Perceived HIV Test Stigma | Baseline and 3 months | This is a subscale of the Total HIV Stigma scale. It contains 2 items, each on a 1-5 point Likert scale, with a range of 5-10. Higher values indicate higher levels of HIV stigma. |
| Mean Change in Perceived HIV Healthcare Stigma Subscale | Baseline and 3 months | This is a subscale of the Total HIV Stigma Scale. It contains two items, each on a 1-5 point Likert scale, with a range of 5-10. Higher values indicate higher levels of HIV healthcare stigma. |
| Mean Change in Fear & Judgement HIV Stigma Sub-scale | Baseline and 3 months | This is a subscale of the Total HIV Stigma Scale. It consists of 5 items, each on a 1-5 point Likert scale, with a range of 5-25. Higher values indicate higher levels of fear & judgement HIV stigma. |
| Mean Change in Perceived Community HIV Stigma Subscale | Baseline and 3 months | This is a subscale of the Total HIV Stigma Scale. It consists of 8 items, each on a 1-5 point Likert-scale, with a range of 6-40. Higher values indicate higher levels of perceived community HIV stigma. |
| Uptake of HIV Self-test | After baseline and by the 3 month follow-up period | The number of participants who ordered an HIV self-test kit after baseline and by the 3-month follow-up. It is a dichotomous outcome (yes/no). |
Countries
Kazakhstan
Participant flow
Recruitment details
Adolescents and young adults were recruited online via websites and social media advertisements and in-person at youth events in Almaty, Kazakhstan from January to August 2023. Those who were interested in study participation completed an online screening survey.
Participants by arm
| Arm | Count |
|---|---|
| Crowdsourced Intervention The digital crowdsourced intervention will be presented to participants in the intervention arm.
Digital Crowdsourced Intervention to Reduce HIV Stigma among Adolescents and Young Adults: Adolescents and young adults will participate in a crowdsourcing contest to create digital materials to reduce HIV stigma in order to increase HIV testing. Entries will be judged by a community and expert judging panel. Winning entries will be selected for the digital intervention and presented to participants in the intervention arm. | 111 |
| Standard of Care Standard HIV informational materials currently used by the Kazakhstan Ministry of Health will be presented to participants in the control arm. | 105 |
| Total | 216 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 |
|---|---|---|---|
| Overall Study | Lost to Follow-up | 13 | 12 |
Baseline characteristics
| Characteristic | Crowdsourced Intervention | Standard of Care | Total |
|---|---|---|---|
| Age, Continuous | 20.1 years STANDARD_DEVIATION 2.4 | 19.4 years STANDARD_DEVIATION 2.4 | 19.7 years STANDARD_DEVIATION 2.4 |
| Race/Ethnicity, Customized Ethnicity Kazakh | 58 Participants | 63 Participants | 121 Participants |
| Race/Ethnicity, Customized Ethnicity Other | 16 Participants | 14 Participants | 30 Participants |
| Race/Ethnicity, Customized Ethnicity Russian | 37 Participants | 28 Participants | 65 Participants |
| Region of Enrollment Kazakhstan | 111 participants | 105 participants | 216 participants |
| Sex: Female, Male Female | 65 Participants | 51 Participants | 116 Participants |
| Sex: Female, Male Male | 46 Participants | 54 Participants | 100 Participants |
| Sexual Orientation Heterosexual | 89 Participants | 77 Participants | 166 Participants |
| Sexual Orientation Sexual Minority | 22 Participants | 28 Participants | 50 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 111 | 0 / 105 |
| other Total, other adverse events | 0 / 111 | 0 / 105 |
| serious Total, serious adverse events | 0 / 111 | 0 / 105 |
Outcome results
Primary
Mean Score Change in Total HIV Stigma Scale
The scale/questionnaire is comprised of 17 items on a 5-point Likert scale (score range from 17-85). Higher scores indicate higher levels of stigma. Outcome will be assessed in mean score change from baseline to 3 months follow-up.
Time frame: Baseline and 3 months
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Crowdsourced Intervention | Mean Score Change in Total HIV Stigma Scale | Baseline | 47.48 units on a scale | Standard Error 0.71 |
| Crowdsourced Intervention | Mean Score Change in Total HIV Stigma Scale | 3 month follow-up | 44.60 units on a scale | Standard Error 0.75 |
| Standard of Care | Mean Score Change in Total HIV Stigma Scale | Baseline | 47.89 units on a scale | Standard Error 0.74 |
| Standard of Care | Mean Score Change in Total HIV Stigma Scale | 3 month follow-up | 47.32 units on a scale | Standard Error 0.79 |
Comparison: Separate multilevel linear mixed models were calculated for the overall stigma score and then for each of the stigma subscale scores. The multilevel linear mixed models incorporated fixed effects for time (categorical), study arm, and their interaction and included intercept as a random effect. Models were also adjusted for pre-intervention stigma levels, whether the participant had submitted content to the study contest, prior HIV testing, age, sex at birth, and sexual orientation.p-value: 0.08395% CI: [-4.9, 0.3]Mixed Models Analysis
Secondary
Mean Change in Fear & Judgement HIV Stigma Sub-scale
This is a subscale of the Total HIV Stigma Scale. It consists of 5 items, each on a 1-5 point Likert scale, with a range of 5-25. Higher values indicate higher levels of fear & judgement HIV stigma.
Time frame: Baseline and 3 months
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Crowdsourced Intervention | Mean Change in Fear & Judgement HIV Stigma Sub-scale | Baseline | 11.12 units on a scale | Standard Error 0.35 |
| Crowdsourced Intervention | Mean Change in Fear & Judgement HIV Stigma Sub-scale | 3 months follow-up | 10.56 units on a scale | Standard Error 0.36 |
| Standard of Care | Mean Change in Fear & Judgement HIV Stigma Sub-scale | Baseline | 11.24 units on a scale | Standard Error 0.36 |
| Standard of Care | Mean Change in Fear & Judgement HIV Stigma Sub-scale | 3 months follow-up | 11.08 units on a scale | Standard Error 0.38 |
Comparison: Separate multilevel linear mixed models were calculated for the overall stigma score and then for each of the stigma subscale scores. The multilevel linear mixed models incorporated random fixed effects for time (categorical), study arm, and their interaction and included intercept as a random effect. Models were also adjusted for pre-intervention stigma levels, whether the participant had submitted content to the study contest, prior HIV testing, age, sex at birth, and sexual orientation.p-value: 0.4895% CI: [-1.49, 0.29]Mixed Models Analysis
Secondary
Mean Change in Perceived Community HIV Stigma Subscale
This is a subscale of the Total HIV Stigma Scale. It consists of 8 items, each on a 1-5 point Likert-scale, with a range of 6-40. Higher values indicate higher levels of perceived community HIV stigma.
Time frame: Baseline and 3 months
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Crowdsourced Intervention | Mean Change in Perceived Community HIV Stigma Subscale | Baseline | 24.65 units on a scale | Standard Error 0.45 |
| Crowdsourced Intervention | Mean Change in Perceived Community HIV Stigma Subscale | 3 month follow-up | 23.28 units on a scale | Standard Error 0.47 |
| Standard of Care | Mean Change in Perceived Community HIV Stigma Subscale | Baseline | 25.05 units on a scale | Standard Error 0.47 |
| Standard of Care | Mean Change in Perceived Community HIV Stigma Subscale | 3 month follow-up | 24.74 units on a scale | Standard Error 0.5 |
Comparison: Separate multilevel linear mixed models were calculated for the overall stigma score and then for each of the stigma subscale scores. The multilevel linear mixed models incorporated random fixed effects for time (categorical), study arm, and their interaction and included intercept as a random effect. Models were also adjusted for pre-intervention stigma levels, whether the participant had submitted content to the study contest, prior HIV testing, age, sex at birth, and sexual orientation.p-value: 0.3795% CI: [-2.6, 0.51]Mixed Models Analysis
Secondary
Mean Change in Perceived HIV Healthcare Stigma Subscale
This is a subscale of the Total HIV Stigma Scale. It contains two items, each on a 1-5 point Likert scale, with a range of 5-10. Higher values indicate higher levels of HIV healthcare stigma.
Time frame: Baseline and 3 months
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Crowdsourced Intervention | Mean Change in Perceived HIV Healthcare Stigma Subscale | Baseline | 5.80 units on a scale | Standard Error 0.14 |
| Crowdsourced Intervention | Mean Change in Perceived HIV Healthcare Stigma Subscale | 3 months follow-up | 5.57 units on a scale | Standard Error 0.15 |
| Standard of Care | Mean Change in Perceived HIV Healthcare Stigma Subscale | Baseline | 5.90 units on a scale | Standard Error 0.15 |
| Standard of Care | Mean Change in Perceived HIV Healthcare Stigma Subscale | 3 months follow-up | 5.89 units on a scale | Standard Error 0.16 |
Comparison: Separate multilevel linear mixed models were calculated for the overall stigma score and then for each of the stigma subscale scores. The multilevel linear mixed models incorporated random fixed effects for time (categorical), study arm, and their interaction and included intercept as a random effect. Models were also adjusted for pre-intervention stigma levels, whether the participant had submitted content to the study contest, prior HIV testing, age, sex at birth, and sexual orientation.p-value: 0.4895% CI: [-0.72, 0.29]Mixed Models Analysis
Secondary
Mean Change in Perceived HIV Test Stigma
This is a subscale of the Total HIV Stigma scale. It contains 2 items, each on a 1-5 point Likert scale, with a range of 5-10. Higher values indicate higher levels of HIV stigma.
Time frame: Baseline and 3 months
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Crowdsourced Intervention | Mean Change in Perceived HIV Test Stigma | Baseline | 5.92 units on a scale | Standard Error 0.13 |
| Crowdsourced Intervention | Mean Change in Perceived HIV Test Stigma | 3 months | 5.19 units on a scale | Standard Error 0.14 |
| Standard of Care | Mean Change in Perceived HIV Test Stigma | Baseline | 5.70 units on a scale | Standard Error 0.14 |
| Standard of Care | Mean Change in Perceived HIV Test Stigma | 3 months | 5.64 units on a scale | Standard Error 0.14 |
Comparison: Separate multilevel linear mixed models were calculated for the overall stigma score and then for each of the stigma subscale scores. The multilevel linear mixed models incorporated random fixed effects for time (categorical), study arm, and their interaction and included intercept as a random effect. Models were also adjusted for pre-intervention stigma levels, whether the participant had submitted content to the study contest, age, sex at birth, and sexual orientation.p-value: 0.03295% CI: [-1.16, -0.18]Mixed Models Analysis
Secondary
Uptake of HIV Self-test
The number of participants who ordered an HIV self-test kit after baseline and by the 3-month follow-up. It is a dichotomous outcome (yes/no).
Time frame: After baseline and by the 3 month follow-up period
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Crowdsourced Intervention | Uptake of HIV Self-test | 30 Participants |
| Standard of Care | Uptake of HIV Self-test | 18 Participants |
Comparison: We used a generalized linear model using a binomial distribution to examine whether the intervention was associated with increased HIV self-testing uptake in the follow-up period. The model was adjusted for history of HIV testing (pre-intervention testing and testing prior to the study), whether the participant had submitted content to the JasSpark contest, and if the participant had a main intimate partner (e.g., girlfriend/boyfriend, spouse).p-value: 0.09995% CI: [0.98, 1.32]Regression, Logistic
Post Hoc
Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale
The Total HIV Stigma scale consists of 17 items, each on a 1-5 point Likert scale, with a range of 17-85. Higher values indicate higher levels of HIV stigma.
Time frame: Baseline and 3 months
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Crowdsourced Intervention | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | 3 months follow-up | 42.39 units on a scale | Standard Error 0.96 |
| Crowdsourced Intervention | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | Baseline | 47.55 units on a scale | Standard Error 0.91 |
| Standard of Care | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | Baseline | 48.24 units on a scale | Standard Error 1.06 |
| Standard of Care | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | 3 months follow-up | 48.17 units on a scale | Standard Error 1.09 |
| Males - Crowdsourced Intervention | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | 3 months follow-up | 47.80 units on a scale | Standard Error 1.16 |
| Males - Crowdsourced Intervention | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | Baseline | 47.44 units on a scale | Standard Error 1.09 |
| Males - Standard of Care | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | 3 months follow-up | 46.41 units on a scale | Standard Error 1.12 |
| Males - Standard of Care | Subgroup Analyses By Sex of Mean Change in Total HIV Stigma Scale | Baseline | 47.63 units on a scale | Standard Error 1.02 |
Comparison: Sex was marginally significant in the initial multilevel mixed models, therefore, we examined potential moderation effects of sex on stigma changes. Fixed effects corresponding to the two-way interactions involving sex, time, and/or arm, as well as the three-way interaction of sex, time, and arm, were added to the models.p-value: 0.01295% CI: [-8.59, -1.58]Mixed Models Analysis
Comparison: Sex was marginally significant in the initial multilevel mixed models, therefore, we examined potential moderation effects of sex on stigma changes. Fixed effects corresponding to the two-way interactions involving sex, time, and/or arm, as well as the three-way interaction of sex, time, and arm, were added to the models.p-value: 0.5695% CI: [-2.26, 5.42]Mixed Models Analysis