How every MedExamHelper station is checked before you play it

In short: A MedExamHelper station goes live only after our clinician, Dr Manju Raja, has reviewed and endorsed two things: the case itself and its marking checklist. Your encounter is then marked by a council of three examiner judges against that endorsed checklist, and every mark cites your own words.

Key points

Why the checking matters

Practice is only useful if the feedback is right. If a case is unrealistic, you practise the wrong thing. If a checklist is loose, the marks don't tell you what to fix. So the part of MedExamHelper we guard most closely is not the AI patient. It is the case and the checklist behind it.

Two endorsements per station

Each station is reviewed in two parts.

The case. The clinician reads the case content: the patient, their history, what they will and won't say, examination findings and investigation results. The case has to be realistic and clinically sound. This endorsement approves the case version itself.

The checklist. The clinician reviews the marking checklist item by item: the domain each item belongs to, what earns it, and how many points it is worth. Items can be edited, added or removed, and every change is saved as a new verified version.

A station only goes live for students when both the case and the checklist are endorsed.

How your encounter is marked

When you end a station and choose End & grade, a council of three examiner judges reads your transcript against the station's endorsed checklist. You get a total, a score for each marking domain, and a debrief. Every mark cites the line of your encounter it came from. That works both ways: what you said that earned a point, and, for a point you missed, the closest thing you did say. Click a cited line to jump to it in your transcript.

The debrief also tells you the status of its checklist. "Clinician-verified practice feedback" means the checklist is endorsed.

Checking the marking, not just the case

Endorsing a checklist once isn't the end of it. The clinician also looks at graded runs and records, for each checklist item, whether she agrees with the mark the judges gave. That agreement is tracked for each station, so differences between the AI marking and the clinician's judgement are visible and can be fixed, either by tightening a checklist item or by saving a new checklist version.

How the patient is set up

Each AI patient follows the case it was given and the rules of a standardised patient. One of those rules is to never volunteer more than the question asks. The patient answers what you ask, in their own words, and leaves you to find the rest. That is the point of a history station.

What MedExamHelper is, and isn't

MedExamHelper is a practice tool for OSCE-style clinical exams, built for UKMLA-style clinical assessment. It is not a patient, it is not medical advice, and it doesn't replace clinical teaching. It exists so you can run a full station, get specific feedback and try again, whenever you have eight minutes.

Common questions

Who reviews the stations? Every case and checklist is reviewed and endorsed by our clinician, Dr Manju Raja, before the station goes live.

How is my performance marked? A council of three examiner judges marks your transcript against the station's endorsed checklist. Every mark cites the line of your encounter it came from.

Can I see why I lost a mark? Yes. Each missed point links to the moment in your encounter where it was available, and missed points become flashcards so you meet them again.

Is this endorsed by the GMC? No. MedExamHelper is built for UKMLA-style clinical assessment but is not affiliated with, or endorsed by, the GMC or any exam body.

Related reading

Start a free station

← Back to all posts