An OSCE circuit is built from five station types. Observed stations put an examiner in the room with the candidate and a patient. Unobserved stations set a written task with no examiner. Technology-enhanced stations use a manikin or a part-task trainer. Linked stations run two consecutive tasks on one case. Rest stations hold a slot with no task. Each type assesses and asks of the circuit something different.
The AMEE guide on OSCE organization lists the first four as the station types a writer chooses from. Rest stations are a circuit device rather than an assessment, but they occupy slots, and they have rules, so they belong in the same table. Whole-exam formats such as the OSPE or the sequential OSCE are covered on the variants page.
Why the type is chosen before the station is written
The type decides what the station can measure. A history cannot be assessed without an examiner watching it. A prescription cannot be assessed by watching; it has to be read. A procedure on a manikin measures technique, not the patient. Choosing the type is therefore a blueprint decision: the cell specifies which competency is sampled, and the type specifies how.
The type also determines the station's cost. An observed station requires an examiner, usually a simulated patient, and a room for 8 to 10 minutes per candidate. An unobserved station needs a desk and a marker afterward. The budget page prices the difference. A circuit of twelve observed stations and a circuit of eight observed plus four unobserved are different exams with different bills.
Which type assesses what?
Type | What it assesses | Examiner and patient | Circuit constraints |
|---|---|---|---|
Observed | History, examination, communication, explanation, procedures on a person | One examiner in the room; a simulated or real patient | The standard slot; the examiner scores live |
Unobserved | Data interpretation, prescribing, written management plans, results letters | No examiner; marked afterward from the candidate's written work | Quiet room; collection and secure storage of scripts; marking time after the day |
Technology-enhanced | Procedural technique on part-task trainers, resuscitation on manikins, examination of signs a person cannot present | An examiner; a technician for reset; no patient, or a patient beside the trainer | Reset time between candidates; spares for every device; a rehearsed fallback if the device fails |
Linked | A longer task in two parts: take the history, then present it; examine, then interpret the results | Usually two examiners, one per part; the patient in the first part only | The pair runs in order for every candidate; no one starts the circuit on the second part or ends on the first |
Rest | Nothing; the slot is a break | None | Never first or last; interspersed among the live stations; excluded from any global rating |
When should a station be observed?
Whenever the competency is a behavior. History taking, examination technique, explaining a diagnosis, breaking bad news, obtaining consent: all of these exist only while they are happening. The examiner is the measuring instrument, and the marking scheme is the examiner's guide to what to look for.
The observed station is also the most expensive and the most exposed to examiner variance. That is why the people section exists. The type is right for the competency; the cost is examiner training.
When is an unobserved station the better choice?
When the competency leaves a written product. Interpreting an ECG, writing a prescription, drafting a discharge letter, choosing the next investigation from a set of results. The candidate completes the task alone under timed conditions, and a marker later scores the product using a key.
Two cautions. First, an unobserved station is a short written exam inside the OSCE. The blueprint must say so, or the exam claims to sample a performance it only sampled on paper. Second, the marking still needs a scheme and a marker briefing. "Marked afterward" is not "marked loosely".
What does a technology-enhanced station add?
A manikin or a part-task trainer allows the exam to assess what a simulated patient cannot present: cardiac arrest, a cannula, a suture, a catheter, or an abnormal heart sound. The trainer standardizes the sign for every candidate. The examiner observes the technique.
The costs are reset after failure. Every trainer has to be returned to the same state between candidates, which is a technician's job and a slot in the timing. Every device can fail, so the equipment page calls for spares. The station also needs a written rule for what happens to the candidate in the room when the device stops.
How does a linked station work in the circuit?
A linked station consists of two consecutive stations in one case. The first is usually observed: take a history, examine a patient. The second uses the first: present the case, interpret the investigations, write the plan. The variants page explains what the link buys.
The circuit rule is simple and easy to break. Every candidate must meet the two parts in order. So no candidate can start the circuit on the second part, and no candidate can finish on the first. With one linked pair in a twelve-slot circuit, two of the twelve starting positions are unavailable, and the rotation has to be built around that. A linked pair also concentrates risk. A candidate who loses time in the first part has nothing to present in the second. The incident log has to record how that was handled.
Where do rest stations go?
A rest station is a slot with no task. It gives a circuit with more candidates than live stations a position for everyone. It also gives candidates a break in a long circuit. It collects no score and is excluded from any global rating.
The AMEE guide states that rest stations are interspersed among the live stations and are never first or last. A rest station at the start is a candidate waiting; a rest station at the end is a candidate leaving early and talking. The delivery section covers rest stations as part of circuit timing and the quarantine of cohorts.
Limits
The types are not exclusive. An observed station can include a trainer for one step. A linked pair can have an unobserved second part. The table describes the pure forms so that the blueprint can name what each station is. When a station is a hybrid, the blueprint and the marking scheme should specify which competency each part assesses, so the coverage claim remains honest.
Frequently asked questions
What are the types of OSCE stations?
Observed stations with an examiner and a patient; unobserved stations with a written task and no examiner; technology-enhanced stations using manikins or part-task trainers; and linked stations that run two consecutive tasks on one case. Rest stations are the fifth kind of slot, with no task and no score.
What is an unobserved OSCE station?
A station where the candidate completes a written task alone, such as interpreting results or writing a prescription, and a marker scores the product afterward with a key. It samples a written product rather than a performance, and the blueprint should reflect that.
Can a rest station be the first station in an OSCE circuit?
It should not be. The AMEE organization guide places rest stations among the live stations and never first or last. A rest station at the start is a candidate waiting to begin; one at the end is a candidate who finishes early and can talk to others. Place rest stations midway along the circuit.
How do linked stations affect the rotation?
Every candidate must complete the two parts in order, so no candidate can start on the second part or end on the first. Those starting positions are removed from the rotation, and the timetable is built around the pair. A problem in the first part also spoils the second for that candidate, which the incident procedure has to cover.
Does a manikin station need a simulated patient?
Not for the technique. A part-task trainer standardizes the sign or the procedure, and the examiner scores the technique. Some stations put a simulated patient beside the trainer so that the candidate also explains and consents, in which case the marking scheme scores both parts separately.