Global Trends of Artificial Intelligence Applications in Family Medicine: A Comprehensive Scopus-Based Bibliometric Analysis (2016–2026)

Aim: To map recent artificial intelligence research in family medicine and primary care against an earlier 2015–2024 bibliometric baseline. Methods: Scopus was searched on 9 July 2026 for 2016–2026 publications. One researcher screened 3452 articles and reviews; 132 topical and 13 language exclusions left 3307 records. Manual title, DOI, and Scopus-identifier checks identified no duplicates. VOSviewer 1.6.21 and Biblioshiny (Bibliometrix 5.4.1; R 4.6.1) supported descriptive analysis and mapping. Results: Output increased from 531 in 2024 to 792 in 2025 (49.2%). The 2025 and partial 2026 component contained 1407 records (42.5%); 615 were dated 2026. The USA led publication and raw citation rankings (711 publications; 18,671 citations), while China preceded the United Kingdom in publication count. The Journal of Medical Internet Research (78 publications) and Scientific Reports (69) led the venues; seven top ten venues were shared with the earlier study. Language model terminology coexisted with later-positioned explainability, risk stratification, and diagnostic accuracy terms. Conclusions: This updated map quantifies post-2024 activity and identifies continuity and corpus-dependent differences in geography, dissemination, and evaluation-related vocabulary. Single-database and English language coverage, single-researcher screening, unadjusted citations, and incomplete 2026 coverage limit inference. The findings describe research activity, not clinical adoption, safety, effectiveness, or causality.