Back

Clinician Experiences with Ambient AI Scribe Technology in Singapore: A Qualitative Study

Shankar, R.; Goh, A.; Xu, Q.

2026-03-19 health informatics
10.64898/2026.03.17.26348627 medRxiv
Show abstract

BackgroundThe administrative burden of clinical documentation is a recognised contributor to clinician burnout and diminished care quality. Ambient artificial intelligence (AI) scribe technology, which uses large language models to passively record and summarise clinical encounters, has rapidly gained traction internationally. However, no published studies have examined clinician experiences with this technology in the Asia-Pacific region or within Singapores multilingual healthcare system. ObjectiveThis study explored clinician perspectives on ambient AI scribe technology at Alexandra Hospital, Singapore, focusing on perceived benefits, barriers, workflow integration, ethical considerations, and recommendations for sustained implementation. MethodsA qualitative descriptive study was conducted using semi-structured interviews with 28 clinicians across multiple specialties at Alexandra Hospital, National University Health System (NUHS). Participants were purposively sampled for diversity in role, specialty, and usage level. Interviews were analysed using reflexive thematic analysis guided by the RE-AIM/PRISM framework. The COREQ checklist was followed. ResultsFive themes emerged: (1) reclaiming presence in the clinical encounter, (2) navigating accuracy and trust in AI-generated documentation, (3) workflow disruption and adaptation, (4) privacy, consent, and ethical tensions within Singapores regulatory landscape, and (5) envisioning sustainable integration. Clinicians reported improved patient engagement and reduced cognitive burden. Persistent barriers included accuracy concerns, AI hallucinations, limited multilingual functionality, loss of documentation style, and uncertainties around compliance with the Personal Data Protection Act (PDPA). ConclusionsAmbient AI scribe technology holds promise for alleviating documentation burden in Singapores public healthcare system. Realising this potential requires attention to safety validation, multilingual capability, clinician training, and patient-centred consent aligned with local regulatory frameworks.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.