Adolescent clubs and self-efficacy linked to better HIV outcomes

Adolescents living with HIV (ALHIV) face psychosocial challenges that could affect HIV treatment outcomes. Peer support networks and aspects of well-being, including self-efficacy, self-esteem, and social capital, could ameliorate these challenges. This retrospective analysis describes participation in existing facility-based adolescent clubs and the associations between club attendance, adolescent well-being, and HIV treatment outcomes. Data were collected through interviews with a sub-sample of adolescents aged 10-19 years and medical record abstraction of all adolescents attending HIV services at seven clinics in Tanzania. Independent variables included adolescent club attendance, self-efficacy, self-esteem, symptoms of depression/anxiety, social capital, and other health utilization or HIV experience characteristics. Study outcomes included visit adherence, viral suppression (<1000 cp/ml), and retention. Of 645 adolescents, 75% attended clubs at least once, with a median of eight club sessions attended over a two-year period. Mental distress, or symptoms of depression and anxiety, was prevalent, with 67% of the adolescents scoring above a recognized cut-off of >5. Adolescents who attended 10 or more clubs, compared to those not attending any clubs over a two-year period, were at an almost three-fold increased odds of having good visit adherence (odds ratio [OR] 2.72, 95% confidence interval [CI]: 1.25, 5.94). Club attendance was also strongly associated with retention in the following year, with adolescents attending some clubs (<10) having three-times the odds of being retained (OR 3.01; 95%CI: 1.86, 4.87) and adolescents attending more frequently (10+) having over seven-times the odds (OR 7.29; 95%CI: 4.34, 12.22). Among the sub-sampled adolescents who were interviewed, being in the top self-efficacy tertile was positively associated with viral suppression (OR 3.04, 95%CI: 1.08, 8.60) and retention (OR 4.44, 95%CI: 1.19, 17.40). Attending the HIV clinic with a guardian/treatment supporter (OR 3.29, 95%CI: 1.17, 9.22) was also associated with viral suppression, and social capital was associated with club attendance (B1 3.24, 95%CI 0.64, 5.85). This study points to the need for comprehensive psychosocial support interventions for ALHIV. Self-efficacy, social capital, and treatment support are important characteristics that facilitate better health outcomes among adolescents. However, many ALHIV either never attended or did not regularly attend clubs, mitigating the real-world impact of such interventions. Further research is needed to identify barriers to club attendance and interventions that promote adolescent well-being, resilience, and guardian/treatment supporter engagement. Data were collected through interviews with a sub-sample of adolescents aged 10-19 years and medical record abstraction of all adolescents attending HIV services at seven clinics in Tanzania. Data were analyzed using STATA 16.1. # Adolescent clubs and self-efficacy linked to better HIV outcomes Dataset DOI: [10.5061/dryad.cnp5hqc5n](10.5061/dryad.cnp5hqc5n) ## Description of the data and file structure We conducted a retrospective analysis of adolescents living with HIV attending seven health facilities with well-established adolescent clubs as part of their HIV clinical services. Data sources included (1) abstracted data on HIV/ART clinical visits from 2015 to 2019, (2) abstracted data on club attendance records from 2016 to 2018, and (3) data collected from structured interviews with a subset of attending adolescents in 2018. Participant and siteid IDs are unlinkable to any personal identifiers. Neither ID alone nor in combination with other variables in the dataset can be used to identify any individual or geographic location within the study country, Tanzania. Two age variables, while treated as continuous in the analyses, were binned to declassify them as "indirect identifiers." The remaining indirect identifiers are HIV status, years on ART, and gender. The data dictionary describes where there are missing values, and which variables were obtained from interviews versus medical record abstraction. The related manuscript describes in more detail how variables were defined. ## Code/software Stata (version 16.1, [stata.com](stata.com)) was used for analysis. Data are deposited here in csv files. ## Access information NA ## Human subjects data Among those recruited for an interview, caregivers of minors and adolescents who were either emancipated or age 18 years or older provided written informed consent. Adolescent minors (age 10–17) provided verbal assent. For adolescents whose medical record information was included in the analysis but who were not recruited for interview, a waiver of informed consent was granted under a separate protocol approved by the National Research Ethics Committee of the National Institute for Medical Research in Tanzania and the US-based IRB, Advarra. All data were fully anonymized using participant and site ID numbers unlinked to any personal identifiers.