@Article{info:doi/10.2196/66564,
author="Tarfa, Adati
and Pecanac, Kristen
and Shiyanbola, Olayinka
and Liebert, Cameron
and Dietz, Sarah
and Miller, Rebecca
and Westergaard, Ryan P",
title="Support Community Formation on a Mobile App for People Living With HIV and Substance Use Disorder: A Computer-Mediated Discourse Analysis",
journal="JMIR Form Res",
year="2026",
month="Jan",
day="15",
volume="10",
pages="e66564",
keywords="HIV; mobile app; social support; computer-mediated; discourse analysis; virtual community; people living with HIV; substance use disorders; SUD; online support; online communities; mHealth; mobile health; HIV care; opioid use disorder; qualitative analysis; OUD",
abstract="Background: People living with HIV and substance use disorders (SUDs) have complex health care needs requiring adaptive and effective support systems. While mobile health apps can foster virtual communities grounded in shared lived experiences, little is known about the dynamics within these digital spaces. Objective: We examined the formation of a virtual community on the Addiction Comprehensive Health Enhancement Support System (A-CHESS; The Center for Health Enhancement Systems Studies, University of Wisconsin-Madison College of Engineering) message board, a mobile app designed to support HIV care engagement among individuals living with HIV and SUD. Methods: We conducted a computer-mediated discourse analysis of A-CHESS message board posts to examine communication patterns, interaction structures, and engagement dynamics. Quantitative comparisons were used to assess differences between posters and nonposters using t tests and chi-square tests. We then applied qualitative coding to categorize messages by type, speaker, and function to understand how staff and participants coconstructed a supportive virtual environment. Results: Among 208 participants, 87 (42{\%}) posted at least once on the A-CHESS message board, contributing 1834 messages between April 2019 and May 2021. Posters and nonposters did not differ significantly in age (t206=--0.64; P=.52), gender ($\chi${\texttwosuperior}1=0.14; P=.71), or race ($\chi${\texttwosuperior}1=0.52; P=.47). We identified 3 message types: premeditated, adlib, and participant-driven. Staff initially led with premeditated messages (eg, recovery stories, HIV risk information, and ``Thought of the Day'' inspiration), which participants often interpreted and adapted to their own SUD recovery. Over time, staff incorporated adlib messaging styles using personalized narratives and polls to sustain engagement. Participants then developed their own posts using similar formats, incorporating Alcoholics Anonymous literature, sharing legal and personal challenges, and suggesting new app features (eg, medication check-ins to support adherence). Conclusions: A-CHESS staff adapted communication styles to increase engagement, while participants appropriated the app's message board to reflect personal goals and lived experiences. Mobile health interventions may benefit from design elements that support participant-led discourse and customization, fostering ownership, support, and relevance within virtual care communities. Trial Registration: ClinicalTrials.gov NCT03873129; https://clinicaltrials.gov/study/NCT03873129 ",
issn="2561-326X",
doi="10.2196/66564",
url="https://formative.jmir.org/2026/1/e66564",
url="https://doi.org/10.2196/66564"
}