Ambient AI documentation has so far lived in consulting rooms and outpatient clinics. On 8 July 2026 it moved into the ambulance: Ortivus announced that its UK subsidiary has partnered with TORTUS AI to integrate ambient documentation directly into MobiMed, the pre-hospital electronic patient record platform used by thousands of paramedics. This matters beyond one deal. Ambulance documentation is a genuinely harder problem than clinic documentation, the announcement cites NHS evaluation figures worth reading carefully, and embedding a scribe inside the record that already flows from scene to hospital raises the most interesting question in this category: what happens when the documentation chain becomes continuous. Here is what was announced, what the evidence shows and does not, and what to watch.
In brief: Ortivus UK and TORTUS AI announced a strategic partnership on 8 July 2026, integrating TORTUS's ambient voice technology into the MobiMed pre-hospital ePR, with an initial UK introduction already live and international rollout anticipated. The announcement cites figures from the Great Ormond Street and NHS England Phase 4 evaluation of ambient voice technology, covering more than 16,000 patient encounters across nine NHS sites: 15 per cent more patient assessments per hour, a 35 per cent reduction in clinicians feeling overwhelmed by note-taking, and a 4.8 per cent reduction in combined on-scene and handover time in fast response settings. These are implementation metrics from an NHS evaluation, not randomised outcome evidence, and the pre-hospital environment will test them further.
Key takeaways
- TORTUS's ambient documentation is being embedded into MobiMed, Ortivus's pre-hospital ePR platform.
- MobiMed is used by over 12,000 paramedics in more than 2,700 emergency vehicles, including major UK deployments.
- The cited figures come from the GOSH and NHS England Phase 4 ambient voice technology evaluation.
- The deal follows TORTUS becoming the first AVT to achieve UKCA Class IIa certification, weeks earlier.
- Documentation flowing structured from scene into the ePR points towards a continuous scene-to-hospital chain.
What was announced
The substance is an embedding deal, not a pilot press release. Ortivus AB's UK subsidiary, Ortivus UK Ltd, has entered a strategic partnership with TORTUS AI to integrate ambient AI documentation into the MobiMed platform: TORTUS listens, transcribes and drafts the clinical documentation during the patient interaction, and the output transfers automatically into the MobiMed ePR in both structured and unstructured form, where it can be used downstream in the care pathway. The solution has already had its initial introduction in the UK, and Ortivus signals a broader international rollout. Context makes the partner choice logical: Ortivus is a Swedish pre-hospital specialist founded in 1985, with a UK subsidiary since 1998, and MobiMed is a modular platform used by over 12,000 paramedics in more than 2,700 emergency vehicles, spanning real-time monitoring and an electronic patient record. Its UK footprint is substantial, including Europe's largest ambulance digitalisation project, a managed-service ePR deployment across 1,200 UK vehicles, and selection by the Northern Ireland Ambulance Service. The platform itself is a regulated medical device, classified Class IIb under the EU MDR and registered with the MHRA.
The timing is the story: certification, then distribution
Two weeks before this announcement, TORTUS became the first ambient voice technology to achieve UKCA Class IIa certification, assessed by an independent UK approved body, a step we track in our AI scribe regulatory status tracker. The Ortivus release leans on exactly that: the partnership is framed as offering trusts a certified medical integration embedded directly into the ePR. The sequence is worth noticing because it is the emerging playbook in this category. Independent certification converts a scribe from a productivity app into something a regulated platform can safely embed, and embedding into a platform converts certification into distribution, in this case across an installed base of thousands of vehicles. For TORTUS, a UK company that built its profile on NHS trials and governance, this is the certify-then-distribute move executed quickly, and it extends the company from clinics and wards into an entirely new setting.
Why ambulance documentation is a harder problem
It is worth being clear-eyed about why this setting is difficult, because it is where the claim will be tested. Ambulance documentation differs from clinic documentation in almost every hard dimension. The acoustic environment is hostile: road noise, sirens, radios, and conversations happening over the patient. There are multiple speakers, crew, patient, family, bystanders, and other responders, whose words must be attributed correctly. Care and documentation happen simultaneously and under interruption, rather than in a tidy consultation arc. Connectivity is intermittent on the move. Observations change rapidly and need structured, time-stamped capture, not just narrative prose, which is precisely why pre-hospital records are built around structured fields. And the encounter ends in a handover, a moment where information loss is a recognised patient-safety problem. A scribe that performs well in a quiet consulting room has not yet proven anything about the back of a moving ambulance, which is why an integration designed for this environment, inside the pre-hospital record, is a more serious proposition than pointing a generic scribe app at the problem.
The figures, read properly
The announcement's numbers deserve both weight and care, and the first thing to establish is where they come from. They are cited to the Great Ormond Street Hospital and NHS England Phase 4 evaluation of ambient voice technology, covering more than 16,000 patient encounters across nine NHS sites, which makes them NHS evaluation findings rather than loose vendor marketing, a meaningful distinction. The headline results: a 15 per cent increase in patient assessments per hour, a 35 per cent reduction in clinicians feeling overwhelmed by the burden of note-taking, and, in fast response settings, a 4.8 per cent reduction in combined on-scene and handover time, framed as enabling an additional patient per shift. Now the care. These are implementation metrics, throughput, workload perception and time, not clinical outcomes, and they come from service evaluation rather than randomised comparison, so they show what happened where the technology was deployed, not what it causes everywhere. Perceived workload is self-reported. And the nine NHS sites of the evaluation are not identical to routine ambulance operations at scale, so the pre-hospital rollout is the test of transfer, not its confirmation. The honest reading is that this is genuinely encouraging NHS evidence for ambient documentation reducing burden, now being extended into a harder environment where it must prove itself again. The same discipline applies here as to any performance claim, and we set out the full checklist in how to read medical AI evaluation claims.
The continuous documentation chain
The most interesting part of this deal is structural. MobiMed is not just an onboard record: it already shares live data with the receiving hospital, including a clinical workstation view where hospital staff can see incoming patients' information and estimated arrival. Put ambient documentation inside that pipe, transferring structured and unstructured output into the ePR, and you get the outline of a continuous documentation chain: the record begins at the scene, travels with the patient, and arrives before them, informing the handover rather than being reconstructed after it. The opportunities are real, better handover quality, less repetition of the story the patient has already told twice, and a cleaner audit trail of what was observed and done when. The risks deserve equal billing. Documentation errors made at the scene propagate downstream with the authority of the record, so verification, and the realistic capacity of a crew to review AI-drafted notes mid-emergency, becomes a safety question. And consent and transparency, which NHS ambient-scribe guidance handles through explicit patient consent in clinics, need rethinking for emergency settings where the patient may be unable to consent to anything. Clinical safety cases and local governance will have to answer those questions before the chain is trusted end to end, and how they are answered will shape the whole category.
What it signals for the market
Strategically, this is a data point for a thesis we have argued before: the moat in clinical documentation is shifting from ambient capture, which is commoditising, to workflow continuity and governance, which are not. TORTUS's move fits that pattern exactly, pairing first-mover certification with embedding into a regulated platform's installed base, and taking the category into a setting competitors have not reached. For how the major documentation players' strategies differ, see four different bets on clinician workflow in Europe. The near-term things to watch are which UK ambulance services deploy it, what the pre-hospital-specific evaluation shows, and how consent and review are operationalised in emergency care.
Where iatroX fits
This partnership is about documentation, and documentation is only one layer of pre-hospital clinical work. The other layer is reasoning: the assessment, the differential, the decision to convey or refer, and the knowledge underneath it, and that is the layer iatroX serves. Ask iatroX answers clinical questions grounded in NICE, CKS, SIGN and the SmPC with the source attached, as a UKCA-marked, MHRA-registered clinical tool, and for pre-hospital clinicians building that underlying capability, iatroX's adaptive question bank includes preparation for pre-hospital qualifications such as the Diploma in Immediate Medical Care, alongside the wider exam catalogue. A scribe writes down what happened; the reasoning layer helps you decide what should happen, and the two are complementary, not competing. Try it with free sample questions at iatroX.
Frequently asked questions
What did Ortivus and TORTUS announce? A strategic partnership, announced 8 July 2026, integrating TORTUS's ambient AI documentation into the MobiMed pre-hospital electronic patient record. TORTUS listens, transcribes and drafts documentation during the encounter, and the output transfers into MobiMed ePR in structured and unstructured form.
What is MobiMed? Ortivus's modular pre-hospital platform, combining patient monitoring and an electronic patient record, used by over 12,000 paramedics in more than 2,700 emergency vehicles, with major UK deployments including a 1,200-vehicle managed service and the Northern Ireland Ambulance Service. It is a Class IIb medical device under the EU MDR.
Where do the performance figures come from? From the Great Ormond Street and NHS England Phase 4 evaluation of ambient voice technology, across more than 16,000 patient encounters at nine NHS sites: 15 per cent more assessments per hour, 35 per cent less note-taking overwhelm, and a 4.8 per cent on-scene and handover time reduction in fast response settings.
Do those figures prove it works in ambulances? Not yet. They are implementation metrics from an NHS service evaluation, not randomised outcome evidence, and the ambulance environment, with its noise, multiple speakers, interruptions and connectivity gaps, is harder than the settings where ambient scribes first proved themselves. The pre-hospital rollout is the real test.
Why does TORTUS's Class IIa certification matter here? Because it is what makes the embedding possible. TORTUS became the first ambient voice technology with UKCA Class IIa certification weeks before the deal, and Ortivus frames the partnership as offering trusts a certified integration inside a regulated platform. Certification converts a scribe into something a Class IIb platform can safely embed.