HIV
Conditions
Brief summary
Engagement in HIV medical care and adherence to HIV medications are both essential in improving health outcomes among people living with HIV (PLH), but PLH living in rural areas-who suffer higher mortality rates than their urban counterparts-can confront multiple barriers to care engagement and adherence, especially as they face the logistical, medical, and social challenges associated with aging. This project will pilot test four intervention components designed to improve care engagement and medication adherence to determine their impact on health outcomes and quality of life among rural, older PLH. The four intervention components, adapted from evidence-based interventions and delivered remotely, are: (1) counselor-facilitated peer social support, (2) HIV stigma reduction, (3) strengths-based case management, and (4) individually-tailored technology use optimization. The investigators hypothesize that components will be acceptable to participants, will be feasible to administer remotely, and will show preliminary impact on (1) the proportion of participants that have viral suppression and (2) health-related quality of life. Results from this study will provide us with tools to improve health outcomes for rural older people living with HIV.
Interventions
BEHAVIORALGroup-Based Social Support
This intervention involves weekly support group calls facilitated by a licensed counselor or therapist for 8 consecutive weeks. The calls will last approximately 90 minutes and will include 5-8 individuals per group. Groups will follow pre-determined topic areas, with participants encouraged to explore their feelings about the difficulties associated with normal aging, being HIV-positive, and living with HIV/AIDS as an older adult. Therapists will facilitate mutual support among group members, encourage greater openness and emotional expressiveness, and help participants to improve their social and family support and enhance their quality of life. This intervention is an adaptation of Telephone Supportive-Expressive Group Therapy (Heckman et al., 2013).
BEHAVIORALHIV Stigma Reduction
This intervention involves weekly support group calls facilitated by a licensed counselor or therapist for 6 consecutive weeks. The calls will last approximately 60-90 minutes and will include 5-8 individuals per group. This intervention is grounded in minority stress theory and will use cognitive-behavioral strategies to help empower participants to cope with stressful and stigmatizing experiences. Intervention components may include minimizing self-stigmatizing attitudes, reducing engulfment, developing a sense of future and hope, and developing and pursuing meaningful life goals. This intervention is an adaptation of Ending Self Stigma (Lucksted et al., 2011).
BEHAVIORALStrengths-Based Case Management (SBCM)
The investigators have adapted an individually-tailored strengths-based case management (SBCM) intervention to help address the multiple structural barriers faced by rural older PLH. The adapted intervention, delivered by trained research staff, will include two 60-minute telephone-based SBCM counseling sessions with shorter follow-up phone calls to check-in on progress and help patients navigate identified barriers. The case manager will provide tailored sessions based on individually-identified needs and proximal life stressors. Capitalizing on participants' personal strengths, case managers will help empower participants to navigate issues related to employment, insurance, mental health, housing, or transportation. This may include assistance understanding, applying for, and accessing benefits or programs.
BEHAVIORALPersonalized Technology Detailing
Participants will be called by a technology-fluent study staff member, who will assess the current state of their technology literacy, access, and use. Detailing will focus on advancing the participant along the technology use cascade (using the internet, possessing a device and service to access internet at home, using the internet to access their electronic health record and pharmacy services, seeking HIV-related information, and finding social support). Each participant will be provided with personalized assistance based on their local and personal circumstances. Because of the individualized advice provided, there will be a range of contacts between 1 and 5, at customized intervals. Detailing protocols include the option of providing the participant with a tablet including cellular service (for 3 months) when in-home internet service is not available or is cost prohibitive.
Sponsors
Medical College of Wisconsin
Study design
Allocation
RANDOMIZED
Intervention model
FACTORIAL
Primary purpose
TREATMENT
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
50 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Age 50 years or greater * Living in a zip code classified as a Small and Isolated Small Rural Town area by Rural-Urban Commuting Area Codes (RUCAs), and/or in a county classified as rural based on RUCAs, and/or in a county with a score of .4 or higher on the index of relative rurality (IRR) * Living in Alabama, Arkansas, Georgia, Kentucky, Mississippi, Missouri, Oklahoma, South Carolina, or Tennessee * Living with HIV * Indicates willingness to participate in support groups * Indicates willingness to self-collect a dried blood spot sample * Has a telephone at home * Able to provide informed consent
Exclusion criteria
* Not meeting eligibility criteria described above
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 3 months following enrollment/baseline survey | Proportion of participants with HIV viral load ≥832 copies/mL, as measured through use of HemaSpot dried blood spot (DBS) testing. At baseline and follow-up, participants were sent dried blood spot (DBS) kits by mail to complete self-collection of blood samples for HIV viral load testing. Following specimen collection, participants placed the filled HemaSpot container into the shipping envelope, which sent the specimen directly to the clinical laboratory for testing. HemaSpot devices with sufficient blood volume were tested using Abbott m2000 RealTime HIV-1 dried blood spot (DBS) quantitative assay, with 832 copies/mL limit of detection. DBS viral load results were classified in three categories: 1. Detected and quantifiable at ≥832 copies/mL; 2. Detected and non-quantifiable, estimated between \<832 and \>300 copies/mL; and 3. Not detected, estimated to be \<300 copies/mL. |
| Health-Related Quality of Life | 3 months following enrollment/baseline survey | Based on full scale scores from the 31-item WHOQOL-HIV BREF (O'Connell & Skevington, 2012), with scores ranging from 0 to 100. Higher scores indicate higher (better) quality of life. Quality of life was assessed with 31 items from the WHOQOL-HIV BREF (O'Connell & Skevington, 2012). Domains assessed include physical health; psychological health; level of independence; social relationships; environmental health; and personal beliefs. Additionally, two individual items focus on overall quality of life and general health. Domain scores were created as described by the WHO, and we created a composite quality of life score by equally weighting the 6 domains, overall quality of life item, and general health item. The composite score was rescaled such that overall scores ranged from 0 to 100, with higher scores indicating better health-related quality of life. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Medication Adherence | 3 months following enrollment/baseline survey | Adherence to HIV antiretroviral medications in the past 30 days was assessed with the 3 items from the Wilson adherence scale (Wilson et al., 2017; αs = .71-.79). This scale was scored in line with Wilson et al., with scores ranging from 0 to 100 and higher scores indicating better adherence. Due to significant skew in this outcome, we created a binary variable indicating perfect adherence to HIV medications. |
| Depressive Symptoms | 3 months following enrollment/baseline survey | Depressive symptoms during the past 2 weeks were assessed with the 9 items from the Patient Health Questionnaire-9 (PHQ-9; Kroenke et al., 2001; αs = .87-.90). Participants indicated how often they had experienced different depressive symptoms (e.g., Feeling down, depressed, or hopeless). Responses ranged from not at all (0) to nearly every day (3). Items were summed to yield scores ranging from 0 to 27, with higher scores indicating more depressive symptoms. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| Social Support + Stigma Reduction + SBCM + Tech Detailing Group-Based Social Support Intervention + Group-Based HIV Stigma Reduction Intervention + Individual Strengths-Based Case Management Intervention + Individual Personalized Technology Detailing Intervention | 4 |
| Social Support + Stigma Reduction + SBCM Group-Based Social Support Intervention + Group-Based HIV Stigma Reduction Intervention + Individual Strengths-Based Case Management Intervention | 4 |
| Social Support + Stigma Reduction + Technology Detailing Group-Based Social Support Intervention + Group-Based HIV Stigma Reduction Intervention + Individual Personalized Technology Detailing Intervention | 3 |
| Social Support + Stigma Reduction Group-Based Social Support Intervention + Group-Based HIV Stigma Reduction Intervention | 3 |
| Social Support + SBCM + Technology Detailing Group-Based Social Support Intervention + Individual Strengths-Based Case Management Intervention + Individual Personalized Technology Detailing Intervention | 4 |
| Social Support + SBCM Group-Based Social Support Intervention + Individual Strengths-Based Case Management Intervention | 4 |
| Social Support + Technology Detailing Group-Based Social Support Intervention + Individual Personalized Technology Detailing Intervention | 4 |
| Social Support Group-Based Social Support Intervention | 4 |
| Stigma Reduction + SBCM + Technology Detailing Group-Based HIV Stigma Reduction Intervention + Individual Strengths-Based Case Management Intervention + Individual Personalized Technology Detailing Intervention | 4 |
| Stigma Reduction + SBCM Group-Based HIV Stigma Reduction Intervention + Individual Strengths-Based Case Management Intervention | 3 |
| Stigma Reduction + Technology Detailing Group-Based HIV Stigma Reduction Intervention + Individual Personalized Technology Detailing Intervention | 4 |
| Stigma Reduction Group-Based HIV Stigma Reduction Intervention | 4 |
| SBCM + Technology Detailing Individual Strengths-Based Case Management Intervention + Individual Personalized Technology Detailing Intervention | 4 |
| SBCM Individual Strengths-Based Case Management Intervention | 4 |
| Technology Detailing Individual Personalized Technology Detailing Intervention | 4 |
| HIV Information Only This arm will not receive any of the 4 intervention components but will receive information on successfully aging with HIV. | 4 |
| Total | 61 |
Baseline characteristics
| Characteristic | Social Support + Stigma Reduction + SBCM + Tech Detailing | Social Support + Stigma Reduction + SBCM | Social Support + Stigma Reduction + Technology Detailing | Social Support + Stigma Reduction | Social Support + SBCM + Technology Detailing | Social Support + SBCM | Social Support + Technology Detailing | Social Support | Stigma Reduction + SBCM + Technology Detailing | Stigma Reduction + SBCM | Stigma Reduction + Technology Detailing | Stigma Reduction | SBCM + Technology Detailing | SBCM | Technology Detailing | HIV Information Only | Total |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age, Continuous | 53 years | 59 years | 55 years | 52 years | 54.5 years | 54 years | 59.5 years | 61 years | 55 years | 62 years | 61 years | 62 years | 61 years | 58.5 years | 54 years | 55 years | 57 years |
| Race/Ethnicity, Customized Race/Ethnicity Black or African American or Caribbean, Non-Hispanic | 2 Participants | 1 Participants | 1 Participants | 0 Participants | 2 Participants | 0 Participants | 0 Participants | 3 Participants | 3 Participants | 1 Participants | 1 Participants | 1 Participants | 0 Participants | 2 Participants | 2 Participants | 0 Participants | 19 Participants |
| Race/Ethnicity, Customized Race/Ethnicity Latino/Hispanic, Any Race | 1 Participants | 0 Participants | 0 Participants | 1 Participants | 0 Participants | 0 Participants | 0 Participants | 0 Participants | 0 Participants | 0 Participants | 0 Participants | 1 Participants | 0 Participants | 0 Participants | 0 Participants | 0 Participants | 3 Participants |
| Race/Ethnicity, Customized Race/Ethnicity White, Non-Hispanic | 1 Participants | 3 Participants | 2 Participants | 2 Participants | 2 Participants | 4 Participants | 4 Participants | 1 Participants | 1 Participants | 2 Participants | 3 Participants | 2 Participants | 4 Participants | 2 Participants | 2 Participants | 4 Participants | 39 Participants |
| Sex: Female, Male Female | 2 Participants | 0 Participants | 1 Participants | 0 Participants | 2 Participants | 1 Participants | 1 Participants | 1 Participants | 1 Participants | 0 Participants | 1 Participants | 1 Participants | 0 Participants | 1 Participants | 2 Participants | 1 Participants | 15 Participants |
| Sex: Female, Male Male | 2 Participants | 4 Participants | 2 Participants | 3 Participants | 2 Participants | 3 Participants | 3 Participants | 3 Participants | 3 Participants | 3 Participants | 3 Participants | 3 Participants | 4 Participants | 3 Participants | 2 Participants | 3 Participants | 46 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk | EG002 affected / at risk | EG003 affected / at risk | EG004 affected / at risk | EG005 affected / at risk | EG006 affected / at risk | EG007 affected / at risk | EG008 affected / at risk | EG009 affected / at risk | EG010 affected / at risk | EG011 affected / at risk | EG012 affected / at risk | EG013 affected / at risk | EG014 affected / at risk | EG015 affected / at risk |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| deaths Total, all-cause mortality | 0 / 4 | 0 / 4 | 0 / 3 | 0 / 3 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 3 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 |
| other Total, other adverse events | 0 / 4 | 0 / 4 | 0 / 3 | 0 / 3 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 3 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 |
| serious Total, serious adverse events | 0 / 4 | 0 / 4 | 0 / 3 | 0 / 3 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 3 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 | 0 / 4 |
Outcome results
Primary
Health-Related Quality of Life
Based on full scale scores from the 31-item WHOQOL-HIV BREF (O'Connell & Skevington, 2012), with scores ranging from 0 to 100. Higher scores indicate higher (better) quality of life. Quality of life was assessed with 31 items from the WHOQOL-HIV BREF (O'Connell & Skevington, 2012). Domains assessed include physical health; psychological health; level of independence; social relationships; environmental health; and personal beliefs. Additionally, two individual items focus on overall quality of life and general health. Domain scores were created as described by the WHO, and we created a composite quality of life score by equally weighting the 6 domains, overall quality of life item, and general health item. The composite score was rescaled such that overall scores ranged from 0 to 100, with higher scores indicating better health-related quality of life.
Time frame: 3 months following enrollment/baseline survey
Population: Analysis included all participants with follow-up survey data (N = 51). Participants were included in outcome analyses in assigned arm regardless of intervention participation/attendance.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Social Support Intervention: Yes | Health-Related Quality of Life | 68.54 score on a scale | Standard Deviation 15.37 |
| Social Support Intervention: No | Health-Related Quality of Life | 64.97 score on a scale | Standard Deviation 15.32 |
| Stigma Reduction Intervention: Yes | Health-Related Quality of Life | 65.83 score on a scale | Standard Deviation 13.96 |
| Stigma Reduction Intervention: No | Health-Related Quality of Life | 67.51 score on a scale | Standard Deviation 16.82 |
| Strengths-Based Case Management Intervention: Yes | Health-Related Quality of Life | 66.50 score on a scale | Standard Deviation 16.75 |
| Strengths-Based Case Management Intervention: No | Health-Related Quality of Life | 66.81 score on a scale | Standard Deviation 13.96 |
| Technology Detailing Intervention: Yes | Health-Related Quality of Life | 68.01 score on a scale | Standard Deviation 15.64 |
| Technology Detailing Intervention: No | Health-Related Quality of Life | 65.00 score on a scale | Standard Deviation 15.04 |
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.36Mixed Models Analysis
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.88Mixed Models Analysis
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.89Mixed Models Analysis
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.73Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.01Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.12Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.08Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.12Mixed Models Analysis
Primary
Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS)
Proportion of participants with HIV viral load ≥832 copies/mL, as measured through use of HemaSpot dried blood spot (DBS) testing. At baseline and follow-up, participants were sent dried blood spot (DBS) kits by mail to complete self-collection of blood samples for HIV viral load testing. Following specimen collection, participants placed the filled HemaSpot container into the shipping envelope, which sent the specimen directly to the clinical laboratory for testing. HemaSpot devices with sufficient blood volume were tested using Abbott m2000 RealTime HIV-1 dried blood spot (DBS) quantitative assay, with 832 copies/mL limit of detection. DBS viral load results were classified in three categories: 1. Detected and quantifiable at ≥832 copies/mL; 2. Detected and non-quantifiable, estimated between \<832 and \>300 copies/mL; and 3. Not detected, estimated to be \<300 copies/mL.
Time frame: 3 months following enrollment/baseline survey
Population: Analysis included all participants with follow-up HemaSpot DBS data (N = 49). Participants were included in outcome analyses in assigned arm regardless of intervention participation/attendance.
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Social Support Intervention: Yes | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 5 Participants |
| Social Support Intervention: No | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 7 Participants |
| Stigma Reduction Intervention: Yes | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 8 Participants |
| Stigma Reduction Intervention: No | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 4 Participants |
| Strengths-Based Case Management Intervention: Yes | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 6 Participants |
| Strengths-Based Case Management Intervention: No | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 6 Participants |
| Technology Detailing Intervention: Yes | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 8 Participants |
| Technology Detailing Intervention: No | Proportion of Participants With HIV Viral Load ≥832 Copies/mL (HemaSpot DBS) | 4 Participants |
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.33Mixed Models Analysis
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.63Mixed Models Analysis
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.78Mixed Models Analysis
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.9Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.38Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.78Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.45Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.91Mixed Models Analysis
Secondary
Depressive Symptoms
Depressive symptoms during the past 2 weeks were assessed with the 9 items from the Patient Health Questionnaire-9 (PHQ-9; Kroenke et al., 2001; αs = .87-.90). Participants indicated how often they had experienced different depressive symptoms (e.g., Feeling down, depressed, or hopeless). Responses ranged from not at all (0) to nearly every day (3). Items were summed to yield scores ranging from 0 to 27, with higher scores indicating more depressive symptoms.
Time frame: 3 months following enrollment/baseline survey
Population: Analysis included all participants with follow-up survey data (N = 51). Participants were included in outcome analyses in assigned arm regardless of intervention participation/attendance.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Social Support Intervention: Yes | Depressive Symptoms | 6.84 score on a scale | Standard Deviation 6.62 |
| Social Support Intervention: No | Depressive Symptoms | 9.33 score on a scale | Standard Deviation 6.74 |
| Stigma Reduction Intervention: Yes | Depressive Symptoms | 8.47 score on a scale | Standard Deviation 6.59 |
| Stigma Reduction Intervention: No | Depressive Symptoms | 7.84 score on a scale | Standard Deviation 7 |
| Strengths-Based Case Management Intervention: Yes | Depressive Symptoms | 8.08 score on a scale | Standard Deviation 7.32 |
| Strengths-Based Case Management Intervention: No | Depressive Symptoms | 8.25 score on a scale | Standard Deviation 6.22 |
| Technology Detailing Intervention: Yes | Depressive Symptoms | 7.50 score on a scale | Standard Deviation 7.43 |
| Technology Detailing Intervention: No | Depressive Symptoms | 8.96 score on a scale | Standard Deviation 5.84 |
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.24Mixed Models Analysis
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.17Mixed Models Analysis
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.66Mixed Models Analysis
Comparison: We fit a linear mixed model using the lmer function in the R package lme4. The model included a random effect for participant and utilized all available data. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.89Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.045Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.07Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.57Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.42Mixed Models Analysis
Secondary
Medication Adherence
Adherence to HIV antiretroviral medications in the past 30 days was assessed with the 3 items from the Wilson adherence scale (Wilson et al., 2017; αs = .71-.79). This scale was scored in line with Wilson et al., with scores ranging from 0 to 100 and higher scores indicating better adherence. Due to significant skew in this outcome, we created a binary variable indicating perfect adherence to HIV medications.
Time frame: 3 months following enrollment/baseline survey
Population: Analysis included all participants with follow-up survey data (N = 51). Participants were included in outcome analyses in assigned arm regardless of intervention participation/attendance.
| Arm | Measure | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|
| Social Support Intervention: Yes | Medication Adherence | 13 Participants |
| Social Support Intervention: No | Medication Adherence | 14 Participants |
| Stigma Reduction Intervention: Yes | Medication Adherence | 13 Participants |
| Stigma Reduction Intervention: No | Medication Adherence | 14 Participants |
| Strengths-Based Case Management Intervention: Yes | Medication Adherence | 14 Participants |
| Strengths-Based Case Management Intervention: No | Medication Adherence | 13 Participants |
| Technology Detailing Intervention: Yes | Medication Adherence | 14 Participants |
| Technology Detailing Intervention: No | Medication Adherence | 13 Participants |
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.049Mixed Models Analysis
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.66Mixed Models Analysis
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.78Mixed Models Analysis
Comparison: We fit a generalized linear mixed model using the glmer function and a logit link in the R package lme4. The model included a random effect for participant. Predictors (fixed effects) included time (baseline vs. follow-up), random assignments, and interactions between assignments and time. We adjusted for mode of survey completion. Significant interactions indicated differential changes in outcomes over time for those randomly assigned vs. not randomly assigned to different interventions.p-value: 0.54Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.09Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.4Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.66Mixed Models Analysis
Comparison: Given that our factorial design resulted in many control participants who were assigned to alternative interventions, for this pilot we also report on changes over time in outcomes for participants randomly assigned to each intervention. These changes tested with mixed models that contained only participants assigned to each intervention. Here, time was the predictor of interest, and we again controlled for mode of survey administration.p-value: 0.87Mixed Models Analysis