All projects

Evidence synthesisCompleted

Large language models in perioperative care: a scoping review

Mapping 869 studies of language models across the surgical journey

A protocol-driven scoping review mapping 869 studies evaluating how large language models are used in perioperative care, with detailed focus on patient-facing, clinician-facing, medical education, research, and administrative and workflow uses.

Why it matters

Use of large language models in perioperative care is emerging but highly heterogeneous. There is a need for a synthesized appraisal of their use and limitations in this context.

About the project

Perioperative medicine spans the multidisciplinary care of patients from preoperative evaluation through intraoperative management and postoperative recovery. It is an information-intensive, time-critical domain where clinicians must integrate complex patient comorbidities, diagnostic tests, procedural risks, and clinical guidelines to ensure safe outcomes. While large language models offer strong potential to assist patients, clinicians, medical learners, and medical researchers, the rapidly growing literature has remained fragmented, heterogeneous, and unstandardized.

To map this emerging landscape, we conducted a protocol-driven scoping review using Joanna Briggs Institute methodology and PRISMA-ScR reporting guidelines, registered on the Open Science Framework. Comprehensive searches across MEDLINE, Embase, EBM Reviews, and Scopus, from inception through November 14, 2024, with an update through October 1, 2025, identified 16,731 records and ultimately yielded 869 studies.

To enable granular analysis, this parent dataset of 869 records was categorized into focused use-case domains: patient-facing uses, clinician-facing uses, medical education uses, research uses, and administrative and workflow support uses.

Next step

Finalize the living scoping review protocol and disseminate findings to support new and emerging studies evaluating large language models in perioperative care.

One project, informed by a wider field of research.

CAPITAL's research themes connect projects to patient understanding, clinical fit, safety, equity, evidence, and responsible implementation.

Explore the research themes

Judging AI-written clinical text

Open project