RECONSTRUCTING ADAPTIVE GLP-1 RECEPTOR AGONIST PATIENT JOURNEYS FROM ONLINE COMMUNITY NARRATIVES
Author(s)
Omar Dabbous, MD, MPH1, Noreen Sullivan, BSc2, Howard Hussein Dabbous, MD3.
1Clearwater Beach, FL, USA, 2MHP Inc, Clearwater FL, MA, USA, 3Dartmouth College, Hanover, NH, USA.
1Clearwater Beach, FL, USA, 2MHP Inc, Clearwater FL, MA, USA, 3Dartmouth College, Hanover, NH, USA.
OBJECTIVES: To characterize longitudinal GLP-1 receptor agonist treatment pathways, patient-centered outcomes, adherence, persistence, continuation, discontinuation, switching, re-initiation, and motivation to continue using sequential discussions from online patient community forums.
METHODS: A retrospective qualitative longitudinal study analyzed publicly available online community forum discussions from 2021 to 2026. De-identified patient identifiers linked posts and comments from the same individual chronologically, and timestamps were used to reconstruct treatment timelines. Inclusion required multiple patient-authored updates on distinct dates with sufficient temporal evidence to establish a journey. Large language model-assisted classification supported extraction of journey stages, patient-reported outcomes, barriers, facilitators, and pathway signals; researcher-reviewed coding determined final classifications. Two qualitative researchers independently reviewed outputs, resolving disagreements by consensus. Causal inference was not attempted.
RESULTS: Among 17,547 GLP-1-relevant patient-authored comments, 1,069 patients met longitudinal inclusion criteria. Median follow-up was 5.1 months. Treatment switching was observed in 372 journeys (34.8%; median follow-up, 6.2 months), access or reimbursement-related interruption in 244 (22.8%; 7.0 months), successful discontinuation in 241 (22.5%; 5.8 months), successful continuation in 228 (21.3%; 6.0 months), adverse-event discontinuation in 228 (21.3%; 5.6 months), and re-initiation after interruption in 133 (12.4%; 5.9 months). Journeys rarely followed a simple start, respond, and stay-on-treatment sequence. Instead, patients commonly started treatment, adjusted dose, experienced benefit or barriers, and then switched, interrupted, restarted, or continued. Motivation to continue was reinforced by perceived response, reduced food-related cognitive burden, improved mobility, daily functioning, emotional well-being, and fear of regain.
CONCLUSIONS: Community-derived longitudinal narratives identified distinct GLP-1 treatment journey patterns, including continuation, switching, interruption, discontinuation, and re-initiation, showing that persistence was adaptive rather than linear. These journey distinctions provide patient-centered context often missing from structured real-world data and may inform adherence research, journey mapping, patient support design, digital health methods, and hypothesis generation around treatment continuation, interruption, switching, and re-initiation.
METHODS: A retrospective qualitative longitudinal study analyzed publicly available online community forum discussions from 2021 to 2026. De-identified patient identifiers linked posts and comments from the same individual chronologically, and timestamps were used to reconstruct treatment timelines. Inclusion required multiple patient-authored updates on distinct dates with sufficient temporal evidence to establish a journey. Large language model-assisted classification supported extraction of journey stages, patient-reported outcomes, barriers, facilitators, and pathway signals; researcher-reviewed coding determined final classifications. Two qualitative researchers independently reviewed outputs, resolving disagreements by consensus. Causal inference was not attempted.
RESULTS: Among 17,547 GLP-1-relevant patient-authored comments, 1,069 patients met longitudinal inclusion criteria. Median follow-up was 5.1 months. Treatment switching was observed in 372 journeys (34.8%; median follow-up, 6.2 months), access or reimbursement-related interruption in 244 (22.8%; 7.0 months), successful discontinuation in 241 (22.5%; 5.8 months), successful continuation in 228 (21.3%; 6.0 months), adverse-event discontinuation in 228 (21.3%; 5.6 months), and re-initiation after interruption in 133 (12.4%; 5.9 months). Journeys rarely followed a simple start, respond, and stay-on-treatment sequence. Instead, patients commonly started treatment, adjusted dose, experienced benefit or barriers, and then switched, interrupted, restarted, or continued. Motivation to continue was reinforced by perceived response, reduced food-related cognitive burden, improved mobility, daily functioning, emotional well-being, and fear of regain.
CONCLUSIONS: Community-derived longitudinal narratives identified distinct GLP-1 treatment journey patterns, including continuation, switching, interruption, discontinuation, and re-initiation, showing that persistence was adaptive rather than linear. These journey distinctions provide patient-centered context often missing from structured real-world data and may inform adherence research, journey mapping, patient support design, digital health methods, and hypothesis generation around treatment continuation, interruption, switching, and re-initiation.
Conference/Value in Health Info
2026-11, ISPOR Europe 2026, Vienna, Austria
Value in Health, Volume 29, Issue 12S
Code
PCR60
Topic
Patient-Centered Research, Real World Data & Information Systems
Topic Subcategory
Patient Behavior and Incentives
Disease
Diabetes/Endocrine/Metabolic Disorders (including obesity)