Journal

The educational opportunity is not simply better explanations. It is maintaining a coherent learner model, running realistic longitudinal cases, and coordinating simulated patient,...

GPT-6 Astra scored 63.4 on length-adjusted HealthBench Professional against 60.5 for GPT-5.6 Sol, with a 1.05 million token context window and genuine agentic capability. What actu...

Geeky Medics holds the largest UK OSCE library. iatroX connects simulation to the question bank and Tutor in one loop across eighteen examinations. Verdicts by learner scenario, an...

There is no single overall winner in AI clinical exam simulation. Evaluated by examination coverage, feedback evidence, full circuits, remediation and price, with honest verdicts b...

The DCE tests synthesising a complex history into a prioritised presentation and defending it under examiner discussion. Simulation supports the thinking and verbal performance; re...

The published protocol for testing which GP appraisal toolkit genuinely takes the least time: a synthetic annual portfolio built in every major platform, measured by clicks, minute...

There is no single FourteenFish alternative, only the right alternative for your specific situation. Eight categories, an honest decision tree, and the critical caveat: several of ...

The published protocol for a blinded, multi-rater benchmark of UK AI virtual-patient platforms: ten test stations spanning risk, safeguarding and difficult communication, fourteen ...

There is no single best AI patient platform for every UK exam. Ranked by examination and workflow rather than one artificial overall winner: the strongest option for the SCA, PLAB ...

PubMed is a literature database. Clinical decision support requires an additional layer that selects, grades, synthesises and contextualises the literature. Five stages, one clinic...

Direct PubMed access improves retrieval and provenance. It does not automatically perform evidence appraisal. What ChatGPT still has to do after finding a paper, and why a citation...

Generative AI is fluent and flexible; traditional diagnostic expert systems like DXplain may still preserve explicit disease-ranking logic better. A 2026 comparison on unpublished ...