"What witchcraft
is this?"
The first published evaluation of ambient voice technology in an NHS ambulance service. Paramedics at LAS saw more patients, in less time, with documentation quality held and cognitive load down.
What the trial measured.
Within-subject pre-post evaluation across two LAS pathways — Hear-and-Treat in the Clinical Hub, and face-to-face care on Double-Crew Ambulances and Solo Response Vehicles.
AVT users scored 98.5% on documentation audit vs 98.2% for non-users. Faster did not mean looser.
NASA Task Load Index median dropped from 5.83 to 4.42 (Z=−3.556, p<.001). Paramedics described the workload reduction in their own words.
Mean combined on-scene plus handover time fell 4.9% (58.4 → 55.6 min) — the figure the operational model is built on.
Solo Response Vehicles, the cohort most starved of admin time, saw the biggest gain: 0.35 → 0.45 patients per hour, median 4 → 5 per shift.
I said to her, just text me when you've used it — and she sent me a GIF which says, "What witchcraft is this?" That was her review.
I genuinely don't think I would be happy losing AVT. I'm still in control as a clinician — it's just helping me with the paperwork side. I'm a lot less stressed.
How the trial was run.
A pragmatic, NHS-led service evaluation. Two pathways, the same Tortus tool, the same clinicians compared against themselves.
- 01Hear-and-Treat (Clinical Hub)April 2025
Two AVT-enabled workstations out of ~50 in the 999/111 Clinical Hub. 11 pilot users. 656 AVT cases analysed against 23,044 non-user comparator cases over the same period.
- 02Face-to-Face (on-scene)Mar–Apr 2025
6-week post-implementation window (5 Mar – 14 Apr 2025) against a 3-month pre-implementation baseline (2 Dec 2024 – 2 Mar 2025). 7 clinicians across Double-Crew Ambulances and Solo Response Vehicles. 344 AVT contacts vs 816 baseline contacts.
- 03MeasuresMixed methods
Assessment duration, patients per hour/shift, on-scene and handover times, documentation-quality audit, NASA Task Load Index, and semi-structured clinician interviews (n=7).
- 04GovernanceRegistered evaluation
Registered as a clinical evaluation study at GOSH DRIVE. Verbal patient and clinician consent. Funded by NHS England (London) Frontline Digitisation and the NIHR GOSH BRC. Tortus AI supplied free of charge.
What the authors flag.
Self-selecting clinician cohort. No full EHR integration at the time of the trial. Only two equipped Clinical Hub workstations. Six-week face-to-face data window — seasonal demand variation may have shaped some of the gains. The preprint is not yet peer-reviewed. Tortus is a collaborator on the work.
Bring the work to your service.
We've published the data. Now let's talk about your hub, your fleet, your shift patterns.