Virtual scribing and AI-assisted clinical documentation: effects on clinician time, burnout, note quality and patient interaction — a scoping review protocol (version 1.0)
Protocol (version 1.0, 14 November 2026) for a scoping review mapping what is known about human virtual scribing, AI-assisted/ambient scribing, and hybrid scribing models — their effects on documentation time, clinician burnout, note quality, and patient interaction — with specific attention to remote/offshore scribing and low- and middle-income settings. Review question (PCC): What is known about the effects of human virtual scribing, AI-assisted/ambient scribing, and hybrid scribing models on documentation time, clinician burnout, note quality, and patient interaction — and what evidence exists from remote/offshore and low- and middle-income settings? Methods: JBI scoping-review methodology (Peters et al. 2020) reported with PRISMA-ScR (Tricco et al. 2018). No meta-analysis, no GRADE, no formal critical appraisal (by design). Databases: PubMed/MEDLINE, Cochrane Library (CENTRAL), OpenAlex, Google Scholar (supplementary). No Embase/Scopus/Web of Science access (stated limitation). Single human reviewer (principal investigator) with calibration set and 10% re-check of exclusions. English only. Full eligibility criteria, search strings, screening, charting and synthesis plans are in the attached protocol file; the final executed search strings are in the attached search-strings file. Timing and honesty note: this deposit is a retrospective public record, not a prospective registration. The protocol was written before searching began (v1.0, 14 November 2026). The formal searches ran on 14 November 2026 before any registration was filed, under the principal investigator's accelerated instruction (Amendment A1); screening (5,260 records), charting (121 included studies) and the manuscript draft were completed on 14 November 2026 before this deposit. PROSPERO does not accept scoping reviews. The protocol was first prepared for OSF Registries; OSF sign-in proved unusable and the registration was moved to Zenodo. Amendments recorded in the protocol file: A1 (searches before registration), A2 (human in-person scribes included as a labelled comparator group), A3 (simulated-encounter evaluations included when they report a clinician-rated note-quality or safety outcome). Funding: none. Conflicts of interest: none declared. The principal investigator works daily as a virtual healthcare assistant in US haematology-oncology; declared as context, not a financial conflict.
Authors
- Haseeb Tariq
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-14
- DOI
- https://doi.org/10.5281/zenodo.24223775
- Primary Topic
- Electronic Health Records Systems
- Type
- preprint