A remote SCA study group needs more than a video call and a queue of cases. It needs a clear task, a candidate who has not already seen the answer, an observation method and a debrief that changes the next attempt. Prosomio's combination of solo and colleague-based practice makes that group workflow a more useful subject than another generic AI-patient comparison.
As checked on 26 September 2026, Prosomio advertises solo consultations, remote practice with colleagues, custom scenarios and structured AI feedback. This article is published by iatroX and includes iatroX among the complementary practice options discussed. The session plan below is an original proposal, not a report of a trial demonstrating that either platform improves SCA results.
Solo practice and colleague practice have different jobs
Prosomio's public description, reviewed on 26 September 2026, distinguishes Dialogue Practice and Virtual Dialogue and describes consultation rehearsal with AI patients or colleagues. It also describes domain-based feedback and progress tracking. These are advertised functions, not evidence that every possible role, recording function or group-room arrangement is available in every account. The product page is the source of those descriptions.
Use solo practice when the aim is to repeat a narrow behaviour without requiring everyone to meet. Examples include opening a consultation without prematurely narrowing the agenda, summarising before proposing a plan or explaining uncertainty in plain language.
Use colleague practice when the important information includes another person's experience of the interaction. Did the candidate's response actually address the concern? Was the explanation understandable? Did a supposedly open question invite an answer or subtly close the discussion?
The group can organise these roles through its own agreed arrangements where the platform does not establish the required permissions. Do not assume an integrated observer seat, private case-authoring field or recording library simply because remote practice is advertised.
Agree the observation target before selecting the case
A session labelled "respiratory cases" may produce plenty of clinical discussion but little change in consultation behaviour. A more useful target is observable: identify the patient's main concern before negotiating a plan, or explain a recommendation without implying certainty that the case does not support.
Choose a case that creates a genuine opportunity to demonstrate that behaviour. Do not build an artificial scenario in which the patient repeatedly announces the hidden concern until the candidate repeats the expected phrase. That tests recognition of a script, not exploration.
Keep the clinical task proportionate to the learning objective. A group working on responding to uncertainty does not need an elaborate diagnostic puzzle that consumes the whole debrief. Equally, a communication target should not remove the need to gather the information required for a safe and coherent clinical discussion.
Use fictional material. A recognisable patient story is not necessary for authenticity, and confidential recalled examination content is not an appropriate source for a custom case.
A complete proposed remote session
The following 75-minute plan is a suggested study-group format. Its intervals are chosen for rehearsal and feedback, not presented as the official SCA timetable. The group should separately check the current examination instructions when undertaking a format-faithful mock.
| Time | Activity | Intended output |
|---|---|---|
| First 5 minutes | Confirm roles, technology and the shared learning target | Everyone knows what will be observed |
| Next 15 minutes | Candidate A completes the first consultation | An uninterrupted attempt rather than live coaching |
| Next 10 minutes | Candidate reflection, peer observations and available AI feedback | One agreed behaviour to change |
| Next 15 minutes | Candidate B completes a different case | A fresh attempt without copying the first consultation |
| Next 10 minutes | Structured debrief of the second case | A specific, evidence-linked improvement target |
| Next 10 minutes | Focused replay of difficult moments | An opportunity to try a different response |
| Final 10 minutes | Agree independent practice and the next review | A realistic plan with named follow-up tasks |
Assign a candidate, a patient or case facilitator, and an observer through the arrangements actually supported. With a small group, roles can rotate between cases or meetings. The facilitator should understand the case but should not improvise new clinical facts merely to rescue a stalled consultation.
The observer records examples rather than interrupting. The candidate completes the attempt before receiving advice. A replay is deliberately different: coaching is allowed because the aim has changed from observing performance to practising an alternative.
A custom-case template that protects the candidate's first attempt
Separate the material the candidate sees from the information used to run and assess the case. This is an authoring principle, not a claim that Prosomio has particular private fields or access controls.
| Case component | Information to prepare | Information not to give away |
|---|---|---|
| Candidate brief | Setting, presenting issue and the task | The intended diagnosis or target phrase |
| Patient account | Opening statement, concerns and relevant background | A scripted announcement of every important fact |
| Information revealed when asked | Consistent answers to foreseeable questions | New facts invented to reward a preferred response |
| Clinical boundaries | Facts needed to keep the scenario coherent | Ambiguous details that change the case accidentally |
| Learning objective | One or two observable behaviours | An exhaustive list that turns the case into a checklist recital |
| Debrief reference | Evidence or guidance relevant to disputed clinical points | A claim that the author's preference is official marking guidance |
Prepare the answer to likely questions before the session. When a candidate asks something the case does not specify, the facilitator should acknowledge that limitation and decide whether it is material. Inventing an answer mid-consultation can make the scenario internally inconsistent.
Where platform permissions cannot keep the candidate's view separate, use a separate case brief shared only with the facilitator. Do not let a convenient screen-sharing arrangement expose the entire case to the person being assessed.
An original scenario focused on a patient's concern
Imagine a fictional adult who is reluctant to proceed with an investigation that has been discussed. The learning objective is not to persuade them at all costs. It is to understand the reason for their reluctance and respond to that reason while preserving an accurate explanation of the clinical uncertainty.
The candidate brief states the reason for the consultation and the information already available. The patient brief explains that a previous experience in the family has made the person anxious about what an abnormal result might mean. The concern is not revealed automatically in the opening sentence; it emerges through appropriate exploration.
During the first attempt, the candidate gives a technically clear explanation of the investigation but never asks what worries the patient. The observer records that omission. During the replay, the candidate asks about the concern, summarises it and then checks whether the explanation addresses the issue raised.
This example does not specify which investigation is indicated or what treatment should follow. It demonstrates a consultation behaviour that can be rehearsed across many clinical topics without converting the exercise into a clinical protocol.
Reconcile AI feedback with peer feedback
Start the debrief with the candidate's account. What were they trying to achieve? At what point did the consultation become difficult? This helps separate a knowledge gap from a communication choice or a technical interruption.
Next, the observer provides an example: the question asked, the answer received and what the candidate did with it. Avoid labels such as "poor rapport" unless they are connected to observable behaviour. The patient-role participant can explain how the interaction felt, while recognising that one peer's reaction is not a universal assessment standard.
Then inspect the AI feedback available for the attempt. Ask whether each important judgement can be linked to the interaction. When the AI and the observer disagree, record the disagreement rather than averaging their conclusions into a reassuring score.
| Feedback claim | Evidence to locate | Next action |
|---|---|---|
| A concern was missed | The concern and the candidate's response | Replay the point of misunderstanding |
| An explanation was unclear | The actual wording and the listener's interpretation | Try a shorter explanation and check understanding |
| Important information was omitted | Whether it was relevant and available in the case | Verify the clinical requirement before adding a routine question |
| The plan was not shared | What options and preferences were discussed | Practise negotiating rather than merely announcing |
If a disagreement concerns clinical accuracy or an important safety issue, use an appropriate source or supervisor. A majority vote among peers is not a substitute for resolving the underlying evidence.
Make the next solo session specific
The group should finish with a task small enough to complete. "Improve communication" is not actionable. "In the next unfamiliar case, ask what the patient is most worried about before giving a long explanation, then summarise the answer" is observable.
At the next meeting, assess the behaviour in a different scenario. Repeating the same words in the same case may show rehearsal rather than transfer. Vary the patient's concern or the reason the discussion is difficult while keeping the target skill recognisable.
Per iatroX product information, September 2026, iatroX Simulations combines voice and text practice, pause-and-coach and uninterrupted modes, transcript-linked domain feedback and Tutor-led remediation. Those features can support work between peer sessions; they do not establish that iatroX reproduces Prosomio's colleague-practice arrangements. Check the relevant tracks through iatroX's examination catalogue.
Which workflow should a group choose?
A group whose main difficulty is arranging realistic remote interaction should examine Prosomio's actual colleague-based workflow and test the necessary roles before committing. A learner whose main difficulty is repeating a specific behaviour between meetings may need a useful solo mode rather than more group features.
A group that already practises effectively through another platform or a straightforward peer arrangement does not need to switch merely because a new name appears. iatroX's existing SCA platform articles provide wider comparisons; this article's decision is narrower: whether the tools help the group protect the first attempt, observe what happened and improve the next one.
Frequently asked questions
Does Prosomio support practice with colleagues as well as AI patients?
Its public page checked on 26 September 2026 advertises both solo and remote colleague-based practice. Confirm the actual roles, access permissions and account arrangements required for your group.
Should a study group accept AI feedback when peers disagree?
Neither should automatically overrule the other. Locate the relevant interaction, clarify the criterion and use an appropriate source or supervisor for important unresolved points.
Do custom cases need to reproduce real SCA stations?
No: original cases can target consultation skills without copying confidential examination material. Keep the clinical facts coherent and the candidate brief separate from the case-author's learning objectives.