An International Publisher for Academic and Scientific Journals
Author Login
SAS Journal of Medicine | Volume-11 | Issue-10
Artificial Intelligence Scribes in Primary Care: A Narrative Review of Promise, Pitfalls, and Practicality
Imran Khan, Abid Sabir
Published: Oct. 24, 2025 |
324
174
Pages: 1021-1024
Downloads
Abstract
Introduction & Background: Within the primary care setting, one of the major issues has been how a physician can meet the clinical demands of a consult and also document effectively. The documentation has multiple purposes, including an accurate record of the consultation, and can be used in multiple future contexts, e.g. audits, medico-legal cases. The emergence of artificial intelligence (AI) tools has meant that ambient scribes are now able to synthesise a record of the consultation whilst the physician focuses on the clinical aspects. This new role of software means time saved, and also has benefits for the doctor and patient satisfaction. Objectives: This literature review will look at current evidence available on AI scribes used in primary care. In particular, focus on benefits, challenges, ethics, legal issues and future direction. Methods: The review was conducted on the following databases - PubMed, Scopus and Google Scholar the aim is to look at AI scribe technologies, pilot studies and implementation reports within the primary care setting. Results: The use of AI scribes has shown signs that documentation time can be reduced, allowing for the physician to focus on their interaction with the patient. This can mean a more comprehensive consult where the patient feels more engaged. In addition, from a physician’s perspective, the relative burden of having to ensure a comprehensive record of the consult, in a short space of time, can help reduce anxiety and burnout. There are, however, significant challenges to seamless adoption. In particular, accuracy, integration within existing Electronic Patient Record (EPR) systems, data privacy, legal issues and costs. The patient consents to having their medical information recorded within an accepted EPR; however, an AI scribe is likely not covered by this consent process as it sits outside of the accepted information technology (IT) infrastructure. The challenges going forward are how scribes can be accepted by th


