Introduction
AI scribes—sometimes called ambient scribes—are digital tools that assist healthcare professionals in documenting clinical work. Using speech-to-text and large language models, they can capture conversations during patient consultations and automatically generate clinical notes and other records.
In the recent year, the adoption of AI scribes has accelerated, first among general practitioners and now also in hospitals through a recent large-scale investment in products from four different companies.
Why this matters
AI scribes promise to improve efficiency and the quality of clinical documentation. Early experiences suggest they may reduce administrative workload and improve accuracy, but we still know very little about their actual impact on everyday clinical practice. This includes how clinical notes function as tools for communication, collaboration, and sensemaking among health professionals.
At the same time, the rapid introduction of AI scribes challenges existing models for testing, procurement, implementation, and scaling of new technologies in healthcare. These processes, which are usually slow and cautious, may need to be rethought to keep up with the pace of innovation.
Possible research questions
A master’s project could, for example, investigate:
- How AI scribes influence clinical work, collaboration, and communication among healthcare staff.
- How existing models for acquiring and introducing new digital tools in healthcare are affected by the fast uptake of AI scribes—and how these models could be adapted.
Methods
We suggest a qualitative approach, e.g. interviews and/or observations with end-users and other stakeholders, but the choice of methods is open to discussion.
Who is this project for?
The project is particularly well-suited for students in the Master’s Program in Digitalization of the Healthcare Sector (DIGHEL). It is relevant both for students with a background in healthcare and those with an informatics or information systems perspective. The project can accommodate up to 4 students, ideally working in groups.