# Artificial Intelligence in Outpatient Primary Care: A Scoping Review on Applications, Challenges, and Future Directions

## TL;DR

This paper reviews how artificial intelligence is being used in outpatient primary care, finding that most applications are still in development with limited real-world use.

## Contribution

The study provides a comprehensive scoping review of AI applications in outpatient primary care, highlighting current developmental stages and implementation gaps.

## Key findings

- Most AI studies in primary care focus on model development rather than real-world implementation.
- AI applications are mainly in clinical decision-making and diagnosis, with few addressing workflow automation.
- Only 11 studies reported real-world AI implementations in primary care settings.

## Abstract

Artificial intelligence (AI) has significant potential to impact clinical decision-making and improve patient outcomes in outpatient primary care. However, despite rapid advancements, the extent of AI implementation in outpatient primary care remains unclear. This scoping review explores how AI functions, undergoes trials, or integrates into non-urgent outpatient primary care settings.

This scoping review was conducted in accordance with the Joanna Briggs Institute methodology and reported following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guidelines. We searched MEDLINE, CINAHL, Scopus, and clinicaltrials.gov databases. Eligible studies were peer-reviewed articles published in English between January 2019 and November 22, 2024, examining AI applications in primary care settings with a direct focus on patient care. Studies were excluded if they were not in English, did not address primary care workflows, or if the full text was unavailable. We added clinicaltrials.gov to uncover active protocols that suggested wider potential adoption. We used thematic analysis to synthesize findings related to AI application domains, research stage, and status of implementation.

We screened 3203 manuscripts and found 61 that met the eligibility criteria. Most studies (n = 26; 43%) focused on model development, while eight reported clinical trial results. AI applications included provider support (n = 5; 8%) and radiological disease diagnosis (n = 1; 2%). Most studies examined clinical decision-making, disease diagnosis, and risk prediction, but none addressed provider cognitive support, workflow automation, or risk-adjusted paneling. There were 11 studies of real-world implementations.

Overall, based on this scoping review of peer-reviewed literature, AI in primary care remains in the developmental stage, with minimal real-world use beyond ambient scribing, clinical decision support, and workflow automation.

The online version contains supplementary material available at 10.1007/s11606-025-09938-0.