NHS Clinicians Reclaim Hours Each Week with AI Voice Documentation

TL;DR:

• Microsoft’s Dragon Copilot is recording NHS consultations and auto-generating clinical notes, with one specialist reporting two hours saved per week.
• The AI scribe is now listed on the NHS England Ambient Voice Technology Registry, clearing the path for wider adoption across trusts.
• Clinicians at Guy’s and St Thomas’ and Imperial College Healthcare report less stress and more meaningful patient interactions — though governance frameworks remain essential.

Clinicians at two major London NHS trusts are using Microsoft’s Dragon Copilot to transcribe patient consultations in real time and draft clinical letters, freeing them from hours of post-clinic typing.

Dr Rachel Hilton, a consultant nephrologist at Guy’s and St Thomas’ NHS Foundation Trust, estimates the tool saves her around two hours of admin each week. Working part-time at 20 hours a week, that represents a 10% time saving she can redirect toward patient care.

“Clinics are way, way less psychologically hard work for me,” Hilton said. “The contact I’m having with the patients is way more meaningful.”

How It Works

Dragon Copilot runs via a phone app synced to the clinician’s desktop. It records consultations, generates structured transcripts organised by clinical topic, and drafts GP referral letters in an appropriate tone — all for the doctor to review and sign off.

At Imperial College Healthcare NHS Trust, consultant gynaecologist Deirdre Lyons noted an unexpected benefit: because the AI only captures spoken words, clinicians must verbalise observations that might otherwise go unrecorded. A patient’s nod becomes a spoken confirmation — encouraging more precise documentation.

Governance First

Both trusts stress that the technology supplements rather than replaces clinical judgement. Imperial established an AI Steering Group to set governance boundaries before deployment, and Microsoft’s Dr Simon Wallace confirmed Dragon Copilot is fully compliant with NHS Ambient Voice Technology standards.

The tool joins a growing ecosystem of AI documentation assistants in UK healthcare. NHS England’s AVT Registry, launched to standardise adoption, now serves as a quality gate that trusts increasingly rely on when evaluating new tools — a pattern that mirrors the broader push toward structured AI governance across public services.

Looking Forward

Wallace sees potential well beyond hospital consultants, arguing that nurse practitioners, speech therapists, physiotherapists, and other allied health professionals could all benefit. If adoption scales as Microsoft intends, the aggregate admin time returned to frontline care across the NHS could be substantial — though each trust will need to fine-tune language models for its own clinical specialities.