Building Care That Fits Its People: Insights from Social Media-Enabled Community-Based Rehabilitation in Thailand
Acarima Nanthanasit, Sophia Ppali, Wan-Jou She, Panote Siriaraya, Thammathip Piumsomboon, Siwaporn Sukittanon, Chee Siang Ang, Alexandra Covaci · 2026 · Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems (CHI '26) · doi:10.1145/3772318.3791965
Summary
This CHI 2026 paper reports on a six-month, multi-platform social media intervention for Community-Based Rehabilitation (CBR) of stroke patients in rural Chiang Mai, Thailand. The authors co-designed content with clinicians, Village Health Volunteers (VHVs), and community members, producing 30 videos and supplementary print materials in four production styles (professional explainer, demonstration with patients, narrative voiceover, and dramatic storytelling). The intervention was deployed across four platforms already embedded in Thai daily life — Facebook, YouTube, TikTok, and LINE — reaching roughly 2,000 villages through 204 sub-districts, and supplemented by hospital outreach at 24 hospitals, TV and radio segments, and local influencer partnerships. The study uses a mixed-methods design: platform analytics (impressions, reach, reactions, shares, click-through) combined with 28 post-deployment semi-structured interviews conducted in northern Thai dialect with one stroke patient, 12 family caregivers, six VHVs, and nine healthcare professionals. Qualitative data was analysed using Reflexive Thematic Analysis and the Critical Incident Technique, presented through six analytic vignettes. Rather than treating CBR as an information-delivery problem, the authors frame digital tools as participants in what they call 'infrastructures of repair' — the locally-assembled socio-technical arrangements (LINE groups, waiting-room conversations, temple visits, VHV rounds) through which care is actually sustained in low-resource settings.
Key findings
Communities appropriated the content in ways the designers never planned. Videos became 'social objects' used to manage family hierarchies: younger caregivers shared clips to introduce advice to elders without violating kreng jai (deferential respect), and VHVs used videos as proxy doctor-authority to earn trust from skeptical patients. Rehabilitation was woven into Buddhist merit-making (tambun): caregivers linked good deeds to patient recovery, and videos were shared as expressions of namjai (generosity) and bunkhun (indebted gratitude). Humour — brief 4-5 second comedic beats in 'Cool VHVs' clips — emerged as a culturally-grounded tool for emotional resilience, whereas prescriptive 'motivation' videos had the lowest engagement (YouTube 4.31%, TikTok 3.2%). Participants preferred private LINE group chats and offline hospital-waiting-room conversations over public comment threads, driven by raksa naa (face-preservation). Platform analytics revealed strong gender asymmetry: women constituted 79.3% of LINE and 74% of Facebook users, mirroring the feminisation of caregiving, though men engaged more with Mental Health Care and Rehabilitation & Exercise content. LINE's push notifications drove a 29.42% click-through rate, vastly outperforming social feed engagement.
Relevance
For accessibility practitioners working on digital health in LMICs — or any context where care is relational, under-resourced, and culturally specific — this paper is a corrective to transmission-model thinking. It argues that HCI should design with waiting rooms, VHV schedules, temple calendars, and household routines as first-order materials, not background context. The 'infrastructures of repair' framing treats appropriation, workarounds, and 'shadow systems' as evidence of design opportunity rather than misuse. Practitioners should note the explicit critique of engagement metrics: views and click-through rates can misclassify a successful intervention as failing if its impact happens offline, in hand-offs between people. Limitations include the single-site rural Chiang Mai setting, five-month deployment window, and lack of assessment of how content handled digital accessibility barriers (connectivity, shared devices, low digital literacy) — gaps the authors acknowledge.
Tags: Community-Based Rehabilitation · digital health · social media · Global South accessibility · stroke recovery · caregiving · LMIC · participatory design · cultural accessibility · Thailand