AI Patient Support and 6-Month Medication Adherence in a Digital Obesity Program: A Retrospective Analysis

Area of Science:

• Digital Health
• Artificial Intelligence in Healthcare
• Pharmacotherapy Adherence

Background:

• Glucagon-like peptide-1 receptor agonist (GLP-1 RA) therapies have high attrition rates in digital weight loss services (DWLSs).
• Conversational artificial intelligence (AI) is proposed to improve patient support and medication retention at scale.
• Evidence for AI's effectiveness in enhancing GLP-1 RA adherence in DWLSs is limited.

Purpose of the Study:

• To evaluate the impact of an AI patient support agent (Junebot) on 6-month medication adherence in a commercial DWLS.
• To assess the independent relationship between digital engagement and medication retention.
• To identify key predictors of long-term retention in DWLSs.

Main Methods:

• Retrospective analysis of 16,556 adults prescribed semaglutide in an Australian DWLS (May 2024 - Oct 2025).
• Primary endpoint: 6-month medication adherence (≥6 orders fulfilled within 183 days).
• Multivariate logistic regression and propensity score matching (PSM) analyses to evaluate AI integration and engagement effects.

Main Results:

• Six-month adherence was higher post-AI integration (53.2%) versus pre-AI (47.3%).
• However, the AI integration era was independently associated with increased attrition odds (OR: 1.178).
• High initial tracking intensity, program pauses, and higher program costs strongly predicted attrition; moderate digital engagement (50-74.99% of weeks) significantly improved adherence odds.

Conclusions:

• AI integration alone did not independently improve medication adherence in this DWLS.
• Programmatic costs, pauses, and initial tracking anxiety were stronger predictors of retention than AI.
• Prioritizing behavioral risk screening and financial accessibility is crucial for digital health providers, alongside moderate, active digital engagement.