A useful CCFP SOO preparation routine gives a study partner and an AI simulation different jobs. Human role-play can reveal whether another person understands and believes your response. Repeatable solo cases let you revisit a specific weakness between observed sessions. Neither should become an exercise in delivering a memorised consultation while ignoring the patient in front of you.
This article is published by iatroX and includes its simulations alongside peer practice. The comparison is about which preparation task each approach can support, not a claim that automated feedback reproduces a CFPC examiner's judgement.
Start with the examination's purpose
The CFPC's preparation material, checked on 3 October 2026, describes SOOs as assessing how candidates understand and manage patients' problems, including their individual experience and a mutually acceptable plan. It provides sample encounters, scripts and marking material. Those are the appropriate starting point for deciding what good practice should look like, rather than an unofficial list of phrases to insert into every conversation. See the CFPC preparation resources.
A candidate can ask about feelings, ideas, function and expectations yet still fail to use the answers. The educational question is what changes after the patient responds. Does the explanation address the actual concern? Does the proposed follow-up fit the circumstances disclosed? Is a second problem integrated rather than treated as an interruption?
Use the official framework to discuss these questions with a partner. Do not convert a practice checklist into a purported official scoring instrument.
Give your study partner a defined role
Before a case, agree whether the session is for learning or uninterrupted rehearsal. In a learning session, the role-player can stop after a missed cue and allow the candidate to try the next exchange differently. During an uninterrupted rehearsal, interruptions belong to the patient role, not to coaching from outside the scenario.
The role-player needs a consistent fictional background, a reason for attending and information that can emerge naturally. They should not reward every broad question by delivering the whole case. Equally, deliberately withholding information after an appropriate enquiry creates an artificial test of guessing the script.
Where a third person is available, ask them to record consequential moments rather than every sentence. For example: the patient disclosed an obstacle; the candidate acknowledged it; the final plan either did or did not change. That sequence is more useful than a global comment that the candidate should be more empathetic.
Work through a disagreement about the plan
Consider a fictional patient, Alex, who agrees that follow-up is important but repeatedly hesitates when an appointment is discussed. The candidate initially explains the clinical rationale more firmly. Further conversation reveals that Alex is caring for a relative and cannot attend at the proposed time.
The practice task is not to invent a convenient clinical alternative. It is to recognise that repeating the same explanation does not address the obstacle, explore what is feasible and negotiate an appropriate next step within the scenario. A partner can describe whether the candidate sounded curious, dismissive or willing to reconsider.
On a second attempt, change the reason for hesitation. Alex might instead misunderstand the purpose of follow-up or believe the problem has already been resolved. The candidate should not automatically reuse the caring-responsibility response. The same outward behaviour can arise from different concerns.
This is an illustrative exercise, not a released CFPC case or an official model answer.
Use solo simulation to repeat the behaviour, not the speech
As described in iatroX's September 2026 launch information, its simulation offering includes a CCFP SOO track, voice and text interaction, a pause-and-coach practice mode and uninterrupted exam mode. Its feedback links observations to the transcript and supports follow-up learning through the Tutor. These are published product features, not evidence that every feedback judgement is correct or that a practice result predicts certification. The simulation launch article describes the wider offering.
A useful solo goal after Alex's case would be to identify and respond to the reason for hesitation in a different encounter. Attempt the conversation before reading the feedback. Then inspect the relevant exchange, including the patient's response, rather than relying only on an overall score or summary.
When feedback appears wrong, identify what the transcript actually shows. A system may overlook a reasonable alternative formulation or give credit for a question whose answer was never used. Bring a consequential disagreement to an experienced human reviewer rather than assuming either the software or your own impression is automatically right.
Compare feedback without averaging it
Suppose a partner says the consultation felt responsive, while automated feedback identifies a missing exploration of expectations. These observations are not necessarily contradictory. The interaction may have felt comfortable while leaving an important uncertainty unexplored.
Ask what each judgement is based on. Locate the moment when the concern could have been clarified, consider whether it was already addressed indirectly, and decide whether a change would improve the consultation. Do not average two scores from different systems into a more authoritative-looking number.
The reverse also matters. A transcript may contain all the expected topics while the partner reports feeling repeatedly interrupted. That is a reason for observed practice, not a reason to dismiss the partner because the checklist looks complete.
Build a week with an observable next step
An example week might start with one observed encounter, continue with short solo rehearsals of the resulting weakness, and finish with an unfamiliar observed case. This is a suggested structure, not a validated training dose or a promise that a particular number of sessions is sufficient.
Record one sentence after each session: what behaviour needs to change and how the next case will test it. Avoid maintaining a growing folder of complete transcripts without a plan to use them. The record should make the next practice session easier to choose.
Keep cases fictional or appropriately de-identified. Neither a study group nor a learning platform is a reason to circulate identifiable patient histories.
Choose the support that solves the remaining problem
For a candidate whose main difficulty is responding naturally to another person, an experienced partner or supervisor may be the most useful next resource. For someone who understands the feedback but needs more opportunities to apply it, solo simulation can fill the gaps between those meetings. For an uncertain clinical decision, return to appropriate Canadian sources rather than expecting role-play to settle the medicine.
As published by iatroX on 3 October 2026, one complete simulation is free. Ongoing access is included with question banks, Socratic Tutor, study planner and CPD tools for £99 paid upfront annually, or £29 monthly. The annual option is equivalent to £8.25 per month billed annually; it is not a monthly payment plan. Choose the commitment around your actual preparation period, not around the number of unrelated examinations in the catalogue.
Frequently asked questions
Can AI simulation replace a CCFP SOO study partner?
It can support repeatable solo rehearsal, but it does not establish how a trained examiner or another person will judge the encounter. Include observed practice when possible.
Should I memorise the official sample SOO scripts?
Use them to understand the task and the basis of feedback, not to learn fixed speeches. Test the same consultation skill in a different fictional case.
What should I do when human and automated feedback disagree?
Inspect the specific exchange and the reasoning behind each observation. Ask an experienced reviewer to resolve important disagreements rather than combining incomparable scores.