PILOT EVALUATION OF AI-ENABLED AMBIENT SPEECH CAPTURE IN HOME-BASED CARE: REDUCING ADMINISTRATIVE BURDEN AND ENHANCING REAL-WORLD EVIDENCE GENERATION FOR COMPLEX THERAPIES

Author(s)

Katie Duncalf, MSc1, Gary Gallagher, MSc1, Joseph Frost, MSc1, Warren Hart, MSc1, Alison Griffiths, RN1, Michael Applewhaite, .2.
1Sciensus, London, United Kingdom, 2Microsoft, Oxford, United Kingdom.
OBJECTIVES: To evaluate the feasibility of integrating Microsoft Dragon Copilot for ambient speech capture in home-based clinical workflows, focusing on reducing administrative burden for nurses, enhancing patient interactions, and facilitating structured capture of real-world evidence (RWE) from patient conversations in complex conditions.
METHODS: In December 2025, Sciensus launched the 'CareTranscribe' pilot (n=100), integrating Dragon Copilot with its Intouch mobile app and Microsoft Dynamics 365 platform. Clinical nurse specialists in the UK are using the tool during home visits to record consented patient consultations, generating detailed notes and structured summaries while nurses remain focused on the patient. Nurses review and validate AI-generated content as supportive input for medical record completion, retaining full control over documentation. Outcomes assessed include transcription accuracy, nurse usability and workflow integration, time savings, patient comfort and consent, and potential to identify behavioural triggers that might help patients get better medical care and support. The pilot operates under a robust governance framework aligned with NHS information governance, UK/EU legal standards, patient-informed consent, strict anonymisation, and Class I medical device requirements.
RESULTS: Pre-launch qualitative research feedback suggests high nurse acceptance, with reduced documentation time enabling greater focus on attentive, personalised care. The recorded interactions are identifying treatment realities rarely captured in traditional clinical settings or trials. Quantitative evaluation of transcription accuracy, workload reduction, and data quality for RWE is in progress. A full data read out will be available at the congress.
CONCLUSIONS: This AI implementation demonstrates potential to alleviate administrative burdens without disrupting nurse-patient relationships, while preserving high-quality clinical documentation. By capturing contextual insights from home-based care such as adherence barriers, education needs, and practical treatment challenges, it supports enriched RWE generation outside controlled environments. This approach offers scalable benefits for outcomes and evidence measurement in complex therapies, facilitating collaboration with biopharma partners to inform service design and improve patient support.

Conference/Value in Health Info

2026-05, ISPOR 2026, Philadelphia, PA, USA

Value in Health, Volume 29, Issue S6

Code

MSR56

Topic

Methodological & Statistical Research

Topic Subcategory

Artificial Intelligence, Machine Learning, Predictive Analytics

Disease

SDC: Gastrointestinal Disorders, SDC: Musculoskeletal Disorders (Arthritis, Bone Disorders, Osteoporosis, Other Musculoskeletal), SDC: Neurological Disorders, SDC: Oncology, SDC: Rare & Orphan Diseases

Your browser is out-of-date

ISPOR recommends that you update your browser for more security, speed and the best experience on ispor.org. Update my browser now

×