PACES is the examination in this whole category where the gap between what a voice-based AI patient can do and what the assessment actually requires is widest, and the honest thesis this article works from is precise rather than dismissive: Simsbuddy, and any comparable conversational platform, may genuinely help with PACES communication, consultation structure and clinical-judgement rehearsal, while no voice-only patient can reproduce a full PACES examination, because a meaningful share of what PACES assesses is not conversational at all.
Mapping the platform against every official PACES skill
The official examination requires candidates to perform physical examinations and identify real signs across respiratory, cardiovascular, neurological and abdominal encounters, alongside clinical communication, differential diagnosis, clinical judgement, managing patient concerns and maintaining patient welfare, seven distinct skill domains worth checking individually rather than assuming a platform's overall PACES category covers all of them equally. Physical examination: the domain requiring hands-on technique, positioning and sequence, structurally outside any conversational system's reach. Identifying physical signs: recognising an actual murmur, an actual abdominal mass, an actual neurological deficit, which requires the sign to exist and be perceptible, not merely described in text. Clinical communication: territory a well-built conversational platform can genuinely rehearse, explaining a diagnosis, addressing a worried patient, structuring an explanation under time pressure. Differential diagnosis: reasoning that can be practised conversationally once findings are supplied, though the findings themselves arrive by description rather than discovery. Clinical judgement: decision-making that a well-designed scenario can test through what a candidate does with information, genuinely simulable territory. Managing patient concerns and maintaining patient welfare: interpersonal and professionalism domains squarely within conversational simulation's actual strength.
The questions worth asking of any PACES-labelled product
Can the learner inspect or hear genuine signs, or are physical findings simply supplied in text, "auscultation reveals a pansystolic murmur", a description rather than a discovery, removing the actual skill of finding the sign in the first place. Does the platform assess examination sequence, or only a verbal description of what the candidate says they would do, a materially different and easier task than performing the sequence correctly under observation. Can it recognise an unsafe or technically incorrect examination approach, something a candidate describes doing that a real examiner would flag as dangerous or improper technique, a detection capability conversational text analysis is not well positioned to provide. Does it practise integrated consultations at the stations combining history, examination and management discussion, and does it cover the more purely communication-focused encounters separately, since PACES structures these differently across its stations and a platform's coverage should map to that structure explicitly rather than treat every station as equivalent. And is feedback explicitly limited to the domains the platform can actually observe, an honesty test worth applying directly: a platform that gives confident feedback on examination technique it never actually assessed is offering false reassurance dressed as analysis.
What conversational simulation can genuinely rehearse
Consent, positioning instructions given verbally, the stated examination sequence, verbalisation of findings as a candidate would narrate them to an examiner, explanation to the patient, interpretation of supplied findings, and differential diagnosis built from those findings, a genuinely useful set of skills that overlap substantially with what integrated consultation stations partly assess, worth practising deliberately even knowing the physical component is absent from the rehearsal.
What cannot currently be reproduced
Palpation, percussion and auscultation themselves, the tactile and acoustic skills no text or voice interface can supply. Clinical inspection in three dimensions, the visual assessment of a real patient that no described scenario replicates. Recognition of subtle physical signs, a skill built through repeated exposure to real and simulated physical findings, not through reading a description that states the finding exists. And patient comfort and physical examination technique, the practical, hands-on competence that only bedside practice, skills-lab work and supervised real encounters develop.
The verdict
Use Simsbuddy, Geeky Medics or any comparable platform for communication, explanation and integrated verbal reasoning, genuinely useful rehearsal for a meaningful share of what PACES assesses. Use bedside teaching, real patients and human examiner feedback for physical signs and examination technique, the domains no current conversational platform reaches, a division of labour this cluster's broader analysis of physical-examination-station AI applies consistently across every UK clinical examination containing a hands-on component, not to PACES alone.
Frequently asked questions
Should a PACES candidate use AI patient practice at all?
Yes, specifically for the communication and integrated-reasoning stations and the verbal components of examination stations, while treating it as one component of a preparation plan that must also include supervised physical-examination practice.
How much of PACES can genuinely be prepared through conversational AI alone?
A meaningful minority rather than a majority: the communication-focused stations and the reasoning components of integrated stations are well served, while the physical-examination-heavy stations require the hands-on practice no current platform substitutes for.
Does a platform's own marketing usually make this distinction clearly?
Not consistently, which is precisely why this article recommends checking any PACES-labelled product against the seven skill domains individually rather than trusting a general "PACES preparation" label to mean full-examination coverage.