STACKing the odds for adolescent survival: health service factors associated with full retention in care and adherence amongst adolescents living with HIV in South Africa

Journal of the International AIDS Society, Volume 21, No. 9, Year 2018

Interpretation

Introduction: There are two million HIV-positive adolescents in southern Africa, and this group has low retention in care and high mortality. There is almost no evidence to identify which healthcare factors can improve adolescent self-reported retention. This study examines factors associated with retention amongst antiretroviral therapy (ART)-initiated adolescents in South Africa. Methods: We collected clinical records and detailed standardized interviews (n = 1059) with all 10- to 19 year-olds ever initiated on ART in all 53 government clinics of a health subdistrict, and community traced to include lost-to-follow-up (90.1% of eligible adolescents interviewed). Associations between full self-reported retention in care (no past-year missed appointments and 85% past-week adherence) and health service factors were tested simultaneously in sequential multivariate regression and marginal effects modelling, controlling for covariates of age, gender, urban/rural location, formal/informal housing, maternal and paternal orphanhood, vertical/horizontal HIV infection, overall health, length of time on ART and type of healthcare facility. Results: About 56% of adolescents had self-reported retention in care, validated against lower detectable viral load (AOR: 0.63, CI: 0.45 to 0.87, p = 0.005). Independent of covariates, five factors (STACK) were associated with improved retention: clinics Stocked with medication (OR: 3.0, CI: 1.6 to 5.5); staff with Time for adolescents (OR: 2.7, CI: 1.8 to 4.1); adolescents Accompanied to the clinic (OR: 2.3, CI: 1.5 to 3.6); enough Cash to get to clinic safely (OR: 1.4, CI: 1.1 to 1.9); and staff who are Kind (OR: 2.6, CI: 1.8 to 3.6). With none of these factors, 3.3% of adolescents reported retention. With all five factors, 69.5% reported retention. Conclusions: This study identifies key intervention points for adolescent retention in HIV care. A basic package of clinic and community services has the potential to STACK the odds for health and survival for HIV-positive adolescents. In this cross‐sectional study, interviews and clinical records were collected from HIV‐positive adolescents in South Africa. Recruitment took place from March 2014 to September 2015. The study site was a rural, peri‐urban and urban health subdistrict in the Eastern Cape province, an area where the healthcare system experiences high burden, poor infrastructure and human resource challenges 18. All health facilities that provided ART to 5 or more adolescents were included (n = 53, including hospital antenatal, paediatric and ART clinics, community health centres and primary care clinics). In each health facility, all clinical files (paper and computerized) were reviewed to identify all adolescents aged 10 to 19 who had ever initiated ART, irrespective of current or past health service attendance. In order to ensure inclusion of adolescents who were both attending and not attending clinical care (and to avoid selection bias of only including those retained in care), all adolescents identified in these files were traced to 180 communities and interviewed at home. Ethical approval was given by the University of Cape Town (CSSR 2013/4) and Oxford University (SSD/CUREC2/12‐21), as well as the Provincial Departments of Health and Education and participating health facilities. All adolescents and their primary caregivers gave written informed consent for participation, and consent procedures were also read aloud in case of low literacy. No financial incentives were given, but all adolescents received a snack, small gift pack (selected by the project’s Teen Advisory Group and including soap, deodorant and pencils) and a certificate. To prevent these becoming an incentive, adolescents received packs and certificates regardless of whether they consented to participate in the study. In order to prevent stigma or unwanted disclosure, the certificate (and all study materials) did not refer to HIV or AIDS but instead to a study about general health and social needs of adolescents in South Africa. Confidentiality was maintained except in cases of risk of harm: where participants reported abuse, suicidality, rape or severe untreated illness, referrals were made to relevant health or social services (n = 94 referrals in the full sample), and followed up to ensure that services were received. Participants completed tablet‐based questionnaires lasting 60 to 90 minutes, with the support of researchers trained in working with vulnerable adolescents. Questionnaires were designed with adolescents (the study’s Teen Advisory Group) to be engaging and non‐stigmatizing, and were piloted with 25 HIV‐positive adolescents in the Eastern Cape. Measures were translated and back‐translated into the local language (Xhosa) and were completed in the language of participants’ choice. In order to identify potential health service factors that were relevant and modifiable, we collaborated with the South African National Departments of Health, Social Development and Basic Education, the South African National AIDS Council, UNICEF, PEPFAR‐USAID and NGOs including Pediatric Adolescent Treatment for Africa. Full questionnaires are available at http://www.mzantsiwakho.co.za. Full self‐reported retention in care was defined as a combination of attending clinic appointments and adhering to ART, defined as both no missed clinic visits over the past year and 85% adherence over the past week, following WHO recommendations 19. Missed appointments were measured over the past year, and used patient self‐report due to low rates of recording of appointments in patient files, low availability of files to healthcare providers when seeing patients and high rates of adolescent mobility between clinics. ART adherence was measured over the past week in order to minimize recall bias, and to include weekdays and weekend given literature on weekend variation. Adherence items used the standardized Patient Medication Adherence Questionnaire 20, adapted using measures developed in Botswana 4. A validation measure was taken in order to test the reliability of self‐reported retention in care: a detectable viral load was extracted from clinical records and defined as viral load 50 + /ml 21. In total, 11 potential protective health service factors were measured, all using adolescent self‐report. Factors hypothesized to increase access to the clinic were (all reported for the past year): (1) less than one hour travel to the clinic from the adolescent’s home; (2) the clinic is accessible: the adolescent can afford to get to the clinic and feels safe whilst travelling and entering the clinic; (3) the adolescent is accompanied to the clinic (either by someone from home or by clinic support staff); and (4) waiting time at the clinic is less than one hour. Factors hypothesized to improve healthcare experience and ART adherence were: (5) the clinic has a reliable antiretroviral stock (i.e. no stock outs in the past year); (6) the clinic healthcare providers have enough time to talk to adolescents; (7) the adolescent perceives that the clinic healthcare providers are kind to adolescents; (8) the clinic healthcare staff provide adolescents with the information they request; (9) the adolescent feels as though their personal information would be kept confidential; (10) the adolescent attends a regular support group that meets at least monthly and; (11) the adolescent has an identified treatment buddy. All measures were dichotomized. In total, 10 potential covariates were measured and controlled for in all analyses, using adolescent self‐report and clinical records: (1) age (dichotomized as younger adolescents aged 10 to 14 and older adolescents aged 15 to 19); (2) gender; (3) residential location (urban/rural); (4) housing situation (formal/informal) were measured using items based on South Africa’s Census 22; (5) maternal orphanhood and (6) paternal orphanhood were measured using items recommended by UNICEF 23; (7) vertical/horizontal HIV infection was assessed following existing Sub‐Saharan African paediatric HIV cohorts: an age cutoff for initiation (10 years) 24 was validated with a detailed algorithm which evaluated the consistency of the initial designations, with inconsistent designations being recoded when strong evidence was available (i.e. maternal and paternal death); (8) overall adolescent health was self‐reported over the past 6 months using the WHO ICF checklist 25; (9) length of time on ART treatment was measured via self‐report and clinic records (more/less than one year on treatment); (10) type of healthcare facility was recorded by the research team and dichotomized into paediatric care versus adult (primary care, adult or antenatal care). We also measured the level of health facility that is primary (clinics), secondary (community health centres/day hospitals) and tertiary care (hospital). Given that access to tertiary care was highly correlated with paediatric care (0.704, p < 0.001), this study only controlled for access to paediatric care. Analyses were conducted in five stages in SPSS 22 and STATA 14. The first three stages were to check the reliability of the sample and outcome measure, and assess frequencies. The final two stages were to identify factors associated with self‐reported retention in care and potential cumulative associations of combinations of those factors. First, eligible participants included in the study were compared to those excluded (the 9.9% not traceable or refused participation) on the sociodemographic characteristics that were available for both groups (age, gender, urban/rural location) using chi‐square tests. Second, frequency distributions for all outcomes, potential protective provisions and covariates were reported. Third, associations of self‐reported retention in care were tested in multivariate logistic regressions, against a validation measure of detectable viral load, controlling for potential covariates (Table 3). Fourth, we ran sequential logistic regressions following Hosmer and Lemeshow's recommendations 26 to test potential associations between individual clinic‐level protective factors and adolescent retention in care. The first step was a model including all potential covariates and health service factors; the second model retained covariates and health service factors significant at p < 0.1; the third and final model retained covariates and health service factors significant at p < 0.05 (Table 4). In the fifth stage, to test potential cumulative effects of protective health service factors, a marginal effects model was run with all potential combinations of significant protective factors, holding significant sociodemographic and HIV‐related cofactors at mean values (Figure 1). This was plotted with 95% confidence intervals. Percentage probabilities of retention in care (controlling for covariates).

AI Digest

Study Justification:

– There are two million HIV-positive adolescents in southern Africa, and this group has low retention in care and high mortality.
– There is almost no evidence to identify which healthcare factors can improve adolescent self-reported retention.
– This study aims to examine factors associated with retention amongst antiretroviral therapy (ART)-initiated adolescents in South Africa.

Study Highlights:

– The study collected clinical records and conducted detailed standardized interviews with 1059 adolescents aged 10 to 19 who were ever initiated on ART in all 53 government clinics of a health subdistrict in South Africa.
– The study found that 56% of adolescents had self-reported retention in care, validated against lower detectable viral load.
– Five factors (STACK) were associated with improved retention: clinics stocked with medication, staff with time for adolescents, adolescents accompanied to the clinic, enough cash to get to the clinic safely, and staff who are kind.
– With none of these factors, only 3.3% of adolescents reported retention, while with all five factors, 69.5% reported retention.

Recommendations for Lay Reader and Policy Maker:

– The study identifies key intervention points for improving adolescent retention in HIV care.
– A basic package of clinic and community services, including stocked medication, staff availability, accompaniment to the clinic, financial support, and kind staff, can significantly improve retention rates.
– Implementing these factors can greatly increase the odds of health and survival for HIV-positive adolescents.

Key Role Players Needed to Address Recommendations:

– Government health departments
– Healthcare providers
– Community organizations
– NGOs working in HIV care and support

Cost Items to Include in Planning the Recommendations:

– Medication stock management
– Staff training and availability
– Transportation support for adolescents
– Community support programs
– Monitoring and evaluation of intervention implementation

Strength of Evidence

The strength of evidence for this abstract is 8 out of 10.
The evidence in the abstract is strong, as it is based on a cross-sectional study with a large sample size (n = 1059) and includes both clinical records and detailed standardized interviews. The study examines factors associated with retention amongst antiretroviral therapy (ART)-initiated adolescents in South Africa. The associations between full self-reported retention in care and health service factors were tested using multivariate regression and marginal effects modeling. The study identifies five factors associated with improved retention: clinics stocked with medication, staff with time for adolescents, adolescents accompanied to the clinic, enough cash to get to the clinic safely, and kind staff. The study provides actionable steps to improve retention by highlighting the importance of these factors. However, the evidence could be further strengthened by conducting longitudinal studies to establish causality and conducting similar studies in different settings to assess generalizability.

Abstract

In this cross‐sectional study, interviews and clinical records were collected from HIV‐positive adolescents in South Africa. Recruitment took place from March 2014 to September 2015. The study site was a rural, peri‐urban and urban health subdistrict in the Eastern Cape province, an area where the healthcare system experiences high burden, poor infrastructure and human resource challenges 18. All health facilities that provided ART to 5 or more adolescents were included (n = 53, including hospital antenatal, paediatric and ART clinics, community health centres and primary care clinics). In each health facility, all clinical files (paper and computerized) were reviewed to identify all adolescents aged 10 to 19 who had ever initiated ART, irrespective of current or past health service attendance. In order to ensure inclusion of adolescents who were both attending and not attending clinical care (and to avoid selection bias of only including those retained in care), all adolescents identified in these files were traced to 180 communities and interviewed at home. Ethical approval was given by the University of Cape Town (CSSR 2013/4) and Oxford University (SSD/CUREC2/12‐21), as well as the Provincial Departments of Health and Education and participating health facilities. All adolescents and their primary caregivers gave written informed consent for participation, and consent procedures were also read aloud in case of low literacy. No financial incentives were given, but all adolescents received a snack, small gift pack (selected by the project’s Teen Advisory Group and including soap, deodorant and pencils) and a certificate. To prevent these becoming an incentive, adolescents received packs and certificates regardless of whether they consented to participate in the study. In order to prevent stigma or unwanted disclosure, the certificate (and all study materials) did not refer to HIV or AIDS but instead to a study about general health and social needs of adolescents in South Africa. Confidentiality was maintained except in cases of risk of harm: where participants reported abuse, suicidality, rape or severe untreated illness, referrals were made to relevant health or social services (n = 94 referrals in the full sample), and followed up to ensure that services were received. Participants completed tablet‐based questionnaires lasting 60 to 90 minutes, with the support of researchers trained in working with vulnerable adolescents. Questionnaires were designed with adolescents (the study’s Teen Advisory Group) to be engaging and non‐stigmatizing, and were piloted with 25 HIV‐positive adolescents in the Eastern Cape. Measures were translated and back‐translated into the local language (Xhosa) and were completed in the language of participants’ choice. In order to identify potential health service factors that were relevant and modifiable, we collaborated with the South African National Departments of Health, Social Development and Basic Education, the South African National AIDS Council, UNICEF, PEPFAR‐USAID and NGOs including Pediatric Adolescent Treatment for Africa. Full questionnaires are available at http://www.mzantsiwakho.co.za. Full self‐reported retention in care was defined as a combination of attending clinic appointments and adhering to ART, defined as both no missed clinic visits over the past year and 85% adherence over the past week, following WHO recommendations 19. Missed appointments were measured over the past year, and used patient self‐report due to low rates of recording of appointments in patient files, low availability of files to healthcare providers when seeing patients and high rates of adolescent mobility between clinics. ART adherence was measured over the past week in order to minimize recall bias, and to include weekdays and weekend given literature on weekend variation. Adherence items used the standardized Patient Medication Adherence Questionnaire 20, adapted using measures developed in Botswana 4. A validation measure was taken in order to test the reliability of self‐reported retention in care: a detectable viral load was extracted from clinical records and defined as viral load 50 + /ml 21. In total, 11 potential protective health service factors were measured, all using adolescent self‐report. Factors hypothesized to increase access to the clinic were (all reported for the past year): (1) less than one hour travel to the clinic from the adolescent’s home; (2) the clinic is accessible: the adolescent can afford to get to the clinic and feels safe whilst travelling and entering the clinic; (3) the adolescent is accompanied to the clinic (either by someone from home or by clinic support staff); and (4) waiting time at the clinic is less than one hour. Factors hypothesized to improve healthcare experience and ART adherence were: (5) the clinic has a reliable antiretroviral stock (i.e. no stock outs in the past year); (6) the clinic healthcare providers have enough time to talk to adolescents; (7) the adolescent perceives that the clinic healthcare providers are kind to adolescents; (8) the clinic healthcare staff provide adolescents with the information they request; (9) the adolescent feels as though their personal information would be kept confidential; (10) the adolescent attends a regular support group that meets at least monthly and; (11) the adolescent has an identified treatment buddy. All measures were dichotomized. In total, 10 potential covariates were measured and controlled for in all analyses, using adolescent self‐report and clinical records: (1) age (dichotomized as younger adolescents aged 10 to 14 and older adolescents aged 15 to 19); (2) gender; (3) residential location (urban/rural); (4) housing situation (formal/informal) were measured using items based on South Africa’s Census 22; (5) maternal orphanhood and (6) paternal orphanhood were measured using items recommended by UNICEF 23; (7) vertical/horizontal HIV infection was assessed following existing Sub‐Saharan African paediatric HIV cohorts: an age cutoff for initiation (10 years) 24 was validated with a detailed algorithm which evaluated the consistency of the initial designations, with inconsistent designations being recoded when strong evidence was available (i.e. maternal and paternal death); (8) overall adolescent health was self‐reported over the past 6 months using the WHO ICF checklist 25; (9) length of time on ART treatment was measured via self‐report and clinic records (more/less than one year on treatment); (10) type of healthcare facility was recorded by the research team and dichotomized into paediatric care versus adult (primary care, adult or antenatal care). We also measured the level of health facility that is primary (clinics), secondary (community health centres/day hospitals) and tertiary care (hospital). Given that access to tertiary care was highly correlated with paediatric care (0.704, p < 0.001), this study only controlled for access to paediatric care. Analyses were conducted in five stages in SPSS 22 and STATA 14. The first three stages were to check the reliability of the sample and outcome measure, and assess frequencies. The final two stages were to identify factors associated with self‐reported retention in care and potential cumulative associations of combinations of those factors. First, eligible participants included in the study were compared to those excluded (the 9.9% not traceable or refused participation) on the sociodemographic characteristics that were available for both groups (age, gender, urban/rural location) using chi‐square tests. Second, frequency distributions for all outcomes, potential protective provisions and covariates were reported. Third, associations of self‐reported retention in care were tested in multivariate logistic regressions, against a validation measure of detectable viral load, controlling for potential covariates (Table 3). Fourth, we ran sequential logistic regressions following Hosmer and Lemeshow's recommendations 26 to test potential associations between individual clinic‐level protective factors and adolescent retention in care. The first step was a model including all potential covariates and health service factors; the second model retained covariates and health service factors significant at p < 0.1; the third and final model retained covariates and health service factors significant at p < 0.05 (Table 4). In the fifth stage, to test potential cumulative effects of protective health service factors, a marginal effects model was run with all potential combinations of significant protective factors, holding significant sociodemographic and HIV‐related cofactors at mean values (Figure 1). This was plotted with 95% confidence intervals. Percentage probabilities of retention in care (controlling for covariates).

Materials

N/A

Innovations

Based on the information provided, here are some potential innovations that could improve access to maternal health:

1. Stocking clinics with an adequate supply of medication: Ensuring that clinics have a consistent and reliable stock of medication can help improve access to maternal health services for HIV-positive adolescents.

2. Allocating time for adolescents during healthcare visits: Providing dedicated time for adolescents during healthcare visits can help address their specific needs and concerns, improving their overall experience and increasing retention in care.

3. Accompanying adolescents to the clinic: Offering support and accompaniment to adolescents when they visit the clinic can help reduce barriers to accessing maternal health services, such as transportation issues or safety concerns.

4. Ensuring safe transportation options: Providing financial support or transportation services to help adolescents safely travel to the clinic can remove a significant barrier to accessing maternal health services.

5. Promoting kindness and empathy among healthcare providers: Creating a supportive and compassionate environment within healthcare facilities can help adolescents feel more comfortable and engaged in their maternal health care.

These innovations, collectively referred to as the “STACK” factors, were found to be associated with improved retention in care among HIV-positive adolescents in South Africa. Implementing these innovations has the potential to improve access to maternal health services for this population.

AI Innovations Description

The study titled “STACKing the odds for adolescent survival: health service factors associated with full retention in care and adherence amongst adolescents living with HIV in South Africa” explores factors that can improve retention in care among HIV-positive adolescents in South Africa. The study collected clinical records and conducted interviews with adolescents who were initiated on antiretroviral therapy (ART) in 53 government clinics in a health subdistrict. The goal was to identify healthcare factors that can improve adolescent self-reported retention in care.

The study found that five factors, referred to as STACK, were associated with improved retention: clinics stocked with medication, staff with time for adolescents, adolescents accompanied to the clinic, enough cash to get to the clinic safely, and staff who are kind. The study showed that with none of these factors, only 3.3% of adolescents reported retention, while with all five factors, 69.5% reported retention.

The study highlights the importance of these factors in improving access to maternal health for HIV-positive adolescents. By ensuring that clinics are stocked with medication, staff have time for adolescents, adolescents are accompanied to the clinic, there is enough financial support for transportation, and staff are kind, the odds of retention in care and adherence to ART can be significantly improved.

This research provides valuable insights for developing innovative interventions to improve access to maternal health for HIV-positive adolescents. Implementing a basic package of clinic and community services that address these factors can help improve the health and survival outcomes for HIV-positive adolescents.

AI Innovations Methodology

Based on the provided description, here are some potential recommendations to improve access to maternal health:

1. Improve transportation: Ensure that pregnant women have access to safe and reliable transportation to healthcare facilities. This can be achieved by providing transportation vouchers, establishing community transportation services, or partnering with ride-sharing companies.

2. Increase availability of healthcare facilities: Expand the number of healthcare facilities, particularly in rural and underserved areas, to ensure that pregnant women have access to nearby healthcare services. This can involve building new clinics or mobile health units.

3. Enhance healthcare provider training: Provide comprehensive training to healthcare providers on maternal health issues, including prenatal care, childbirth, and postnatal care. This can improve the quality of care provided to pregnant women and increase their trust in the healthcare system.

4. Implement telemedicine services: Utilize technology to provide remote consultations and monitoring for pregnant women, especially in areas with limited access to healthcare facilities. This can include virtual prenatal visits, remote monitoring of vital signs, and teleconsultations with specialists.

5. Strengthen community engagement: Engage local communities in promoting maternal health and encouraging pregnant women to seek care. This can involve community health workers, peer support groups, and community education campaigns.

To simulate the impact of these recommendations on improving access to maternal health, a methodology could include the following steps:

1. Define the target population: Determine the specific population group that will be the focus of the simulation, such as pregnant women in a particular region or demographic.

2. Collect baseline data: Gather relevant data on the current state of access to maternal health services, including factors such as distance to healthcare facilities, transportation availability, healthcare provider capacity, and community engagement.

3. Define indicators: Identify key indicators that will be used to measure the impact of the recommendations, such as the number of pregnant women accessing prenatal care, the rate of maternal mortality, or the percentage of women receiving postnatal care.

4. Develop a simulation model: Create a mathematical or computational model that simulates the impact of the recommendations on the defined indicators. This model should take into account factors such as population size, geographical distribution, and the effectiveness of each recommendation.

5. Input data and run simulations: Input the baseline data into the simulation model and run multiple simulations to assess the potential impact of different combinations of recommendations. This can involve adjusting variables such as the number of healthcare facilities, transportation availability, or the level of community engagement.

6. Analyze results: Analyze the simulation results to determine the potential impact of the recommendations on improving access to maternal health. This can include comparing different scenarios, identifying the most effective recommendations, and assessing the overall impact on the defined indicators.

7. Refine and validate the model: Continuously refine and validate the simulation model based on real-world data and feedback from experts in the field. This will ensure that the model accurately represents the complexities of the healthcare system and provides reliable predictions.

By following this methodology, policymakers and healthcare providers can gain insights into the potential impact of different recommendations on improving access to maternal health and make informed decisions on implementing the most effective strategies.

Author


Dima Health

Statistics:

Citations: 41

Identifiers:

DOI: 10.1002/jia2.25176

Research Areas:

Community Interventions, Health System and Policy, Infectious Diseases, Maternal Access, Maternal and Child Health, Quality of Care, Violence and Injury

Study Countries:

Botswana, Multi-Countries, South Africa

Study Design:

Cohort Study, Cross Sectional Study, randomised-control-trial

Study Approach:

Quantitative