Journal

TORTUS and Tandem both claim a first Class IIa certification, and both are right. They were certified under different regimes: TORTUS under the UK's UKCA regime, Tandem under EU MD...

On 24 June 2026, TORTUS became the first ambient voice technology certified UKCA Class IIa, assessed by Scarlet (UK Approved Body 8536). What the jump from self-declared Class I to...

The real signal in Midjourney Medical isn't the scanner — it's a consumer AI-native company entering regulated healthcare hardware. Why that shift matters, where it's risky, and wh...

A wave of free, ad-free, citation-backed clinical answer tools has arrived — Heidi Evidence, OpenEvidence in the US, scribes bundling answers into the consultation. The AI 'answer ...

Clinical AI tools increasingly carry a regulatory class — Class I, Class IIa, sometimes higher — and the number is widely misread as a ranking of quality or safety. It isn't. Class...

Tandem and Tortus are both deployed across the NHS at scale through major distribution partners, and both are climbing the medical-device regulatory ladder. They differ on regulato...

Both Tandem and Heidi started as AI scribes and have moved into clinical decision support. Their CDS products take different routes on regulation, sourcing and — critically for UK ...

Clinical reference is not new. What AI changes is the interface: instead of searching manually, a clinician can ask a question and review a source-linked answer. The intended use s...

Clinical AI is not one category. Accurx, Heidi, Tortus, Tandem, OpenEvidence and iatroX occupy different layers of the workflow — communication, documentation, coding, decision sup...

As scribes add clinical evidence and decision support, buyers need to separate documentation quality, recommendation quality, regulatory scope and real workflow value.

What Tandem's reported Class IIa certifications mean after the UK CE-recognition decision, and why market access, local assurance and clinical performance remain separate.

PSA failure is fixed by practising safe prescribing workflows under time pressure, not by reading more pharmacology.