The Influence of AI-Generated Health Information on Medical Consultation Decision-Making Among Adults: A Systematic Review
Background: The emergence of large language models (LLMs) such as ChatGPT has transformed how adults access and interpret health information, with approximately 48% of consumers now using generative AI for health-related inquiries. However, the influence of AI-generated health information on medical consultation decision-making remains poorly synthesized. Objective: This systematic review synthesizes evidence on how AI-generated health information influences medical consultation decision-making among adults, examining demographic characteristics of AI users, types of AI tools used, outcomes on consultation behaviour, and moderating factors. Methods: A comprehensive search of PubMed/MEDLINE, Scopus, Web of Science, Cochrane CENTRAL, PsycINFO, IEEE Xplore, and ACM Digital Library was conducted for studies published between 2021 and 2026. Two independent reviewers screened 2,604 records, with 13 studies meeting inclusion criteria. Data extraction and quality assessment were performed using JBI, AXIS, ROBINS-I, and MMAT tools. A narrative synthesis approach was adopted. Results: AI tools significantly influence consultation decisions, with 47.4% of users employing AI to determine consultation necessity, 35.6% requesting referrals, and 31% modifying medications. AI demonstrates potential benefits including improved consultation efficiency (40% reduction in waiting times) and high-quality information provision, yet safety concerns persist with fewer than 40% of AI responses deemed harmless. Patients consistently prefer physician-led care, with AI serving as a complementary tool. Digital health literacy and younger age predict AI engagement, raising equity concerns. Conclusions: AI-generated health information significantly shapes medical consultation behaviour, necessitating risk-stratified deployment strategies, clinician guidance frameworks, and interventions to ensure equitable access and preserve the physician-patient relationship.