Artificial intelligence and machine learning for organizational and system-level functions in primary health care: A...

Primary health care (PHC) is the cornerstone of health systems worldwide, yet faces growing pressures from aging populations, workforce shortages, and constrained resources. While artificial intelligence (AI) and machine learning (ML) are increasingly used for clinical tasks such as diagnostics and decision support, their potential to address organizational and system-level processes remains underexplored. This scoping review maps AI and ML applications targeting non-clinical, organizational, and system-level functions in PHC, examining techniques employed, application purposes, data sources, and implementation maturity, and identifying evidence gaps to inform future research.

This scoping review will be conducted using the Arksey and O'Malley framework and reported according to PRISMA-ScR. We will search Ovid MEDLINE, EBSCO CINAHL, Ovid Embase, Cochrane Library (Wiley), Ovid PsycINFO, and IEEE Xplore, as well as grey literature sources, from January 2010 to December 2025, with a pre-submission update. Two independent reviewers will screen titles, abstracts, and full texts. Eligible studies include empirical research describing, evaluating, implementing, or developing AI and ML applications at the meso-level (organizational, e.g., care scheduling, population stratification) and macro-level (system-wide, e.g., workforce planning, funding allocation) of PHC. Studies on micro-level clinical applications, non-empirical research, and secondary literature will be excluded. Applications will be classified using a two-layer taxonomy (AI technique family and core algorithm) and mapped to PHC building blocks adapted from WHO operational frameworks. Synthesis will examine technique-building block intersections, application purposes, data sources, implementation maturity, and evidence gaps. The protocol is registered in Open Science Framework (osf.io/wzj5x).

This review will provide a timely synthesis of AI and ML applications at organizational and system levels of PHC, identifying which building blocks have been addressed, which remain underexplored, and where the field stands in terms of implementation readiness. The findings will inform a targeted research agenda and provide actionable evidence for decision-makers seeking to leverage AI and ML for PHC strengthening.