GP Institute of Australia · AGPT selection: the Casper SJT · Lesson 12 of 16
Clinical reasoning, uncertainty and shared decisions: judgement when you can’t be sure
Recognise the thinking traps that lead to snap judgements, handle uncertainty honestly and safely, weigh evidence against what matters to people, and make decisions with them rather than for them.
The RACGP national entry assessment is Acuity Insights’ Casper test: 11 scenarios with two open-ended questions each, four answered on video (one minute per question) and seven typed (3.5 minutes for both questions). It assesses nine competencies, not clinical knowledge, so this lesson teaches judgement and communication rather than clinical facts.1
Where are you starting from?
Rate yourself honestly. There are no marks here; each “Not yet” or “Somewhat” points you to the part of this lesson that closes that gap.
I can name common thinking traps such as anchoring and confirmation bias.
I can explain uncertainty to someone honestly without alarming them.
I know what safety-netting is and why it matters when I’m not sure.
I can explain evidence-based practice as more than ‘what the research says’.
I can run a shared decision-making conversation.
“Not yet” answers: . Your choices are not saved.
Learning objectives
- Recognise fast and slow thinking and the common cognitive biases.
- Avoid premature judgements in scenarios by gathering information first.
- Manage and communicate uncertainty with honesty, safety-netting and follow-up.
- Combine evidence, experience and a person’s values in decisions.
- Use shared decision-making and know when to ask for help.
How this shows up in Casper
- Scenario types: a decision made on incomplete information; a person relying on online claims; disagreement about a test or treatment; a group decision under uncertainty.
- Typical question stems: “How would you decide…?”, “How would you communicate your decision?”, “How do you decide what information to trust?”
- Common trap: assuming facts the scenario doesn’t give you, or sounding certain when the honest answer is ‘it depends on…’.
| ★★★★ | Must master: shapes almost every answer on this theme |
| ★★★★ | High yield: frequent scenario types and common traps |
| ★★★★ | Know the principles: background that sharpens your judgement |
| ★★★★ | Background: context |
Ratings are GP Institute’s editorial judgement, not an official RACGP or Acuity Insights weighting. Any competency can appear in any scenario.
1. Why judgement under uncertainty matters ★★★★
General practice is full of uncertainty: early, undifferentiated symptoms, incomplete histories, tests that are neither clearly normal nor abnormal. Casper is uncertain by design: scenarios leave out key facts so that raters can see whether you assume, or whether you ask. The thinking habits that make a good GP (gathering information, holding several possibilities, being honest about what you don’t know, involving the person in decisions) are exactly the habits Casper rewards. This lesson turns the source material on diagnostic reasoning into those transferable habits; Casper does not test clinical knowledge.1
| Habit | In general practice | In a Casper answer |
|---|---|---|
| Gather before deciding | History, examination, collateral information | “I’d first find out…” |
| Hold several possibilities | A differential diagnosis | “It could be that…, or that…” |
| Name what would change your mind | Red flags and review criteria | “If I learned…, I would…” |
| Be honest about uncertainty | Explain what is known and not known | Avoid false certainty |
| Decide with the person | Shared decision-making | Involve those affected in the plan |
2. How we think, and how thinking goes wrong ★★★★
Psychologists describe two broad modes of thinking: fast, intuitive pattern recognition, and slower, deliberate analysis. Experienced doctors rely heavily on fast thinking, which is efficient and usually right, but it is also where cognitive biases live. Croskerry and others have described how errors in diagnosis often arise from predictable biases rather than lack of knowledge.4 The same biases affect how we judge people in everyday situations, which is why Casper scenarios are written to tempt them.
| Bias | What happens | Casper example | Antidote |
|---|---|---|---|
| Anchoring | Sticking with the first impression | Deciding a colleague is lazy from one comment | Ask what else could explain it |
| Confirmation bias | Seeking evidence that fits your view | Only hearing one side of a conflict | Hear every party before judging |
| Availability | Overweighting vivid or recent cases | Assuming the worst because of a story you heard | Ask how common it really is |
| Premature closure | Stopping thinking once an answer appears | Choosing an action in the first line | Gather information first |
| Attribution error | Blaming character rather than circumstances | “She’s careless” rather than “she’s overloaded” | Look for situational causes |
| Framing | Being swayed by how information is presented | Accepting a senior’s account as the whole story | Reframe from another perspective |
“Clearly he was drunk” or “she obviously didn’t care” adds facts the scenario never gave you. Raters treat this as poor judgement. Use conditional language: “If he had been drinking, I would…; if not, I’d want to understand…”
3. Managing uncertainty ★★★★
Uncertainty cannot be eliminated, only managed. Good GPs use time as a diagnostic tool, explain what is and isn’t known, agree what would prompt review, and make sure follow-up actually happens. In Casper, the equivalent is to make a sensible decision with the information available, say what you would do if new information emerged, and build in a check.
| Poorly | Well |
|---|---|
| “It’s nothing, don’t worry.” | “It’s most likely X, but I can’t be completely sure yet, so here’s what we’ll watch for.” |
| “I don’t know.” (and nothing more) | “I don’t know yet; here’s how we’ll find out.” |
| Ordering every test to remove all doubt | Explaining the trade-offs of testing and deciding together |
| Long lists of every possibility | The likely explanation, the important one not to miss, and the plan |
One “if/then” sentence shows a rater that you have thought beyond the obvious: “If she tells me she’s been unwell, I’d focus on supporting her; if it’s a pattern, I’d involve our supervisor.” It is the single most efficient way to show judgement under uncertainty.
4. Evidence-based practice ★★★★
Evidence-based practice is often misunderstood as doing whatever the research says. Sackett and colleagues defined it as integrating the best available research evidence with clinical expertise and the patient’s values and circumstances.5 All three matter. Evidence without values becomes paternalism; values without evidence can lead to harm.
Swipe to see the whole figure →
| Question | Why it matters |
|---|---|
| Who is behind it, and do they gain from it? | Commercial or ideological interests |
| What is the evidence: a study, many studies, or a story? | Anecdotes can’t show cause and effect |
| Is it consistent with reputable bodies? | Guidelines summarise the weight of evidence |
| Is it current? | Advice changes as evidence accumulates |
| Does it fit this person? | Evidence from one group may not apply to another |
5. Shared decision-making ★★★★
The Australian Commission on Safety and Quality in Health Care describes shared decision-making as the discussion before consent in which patient and clinician talk through the options and decide together on the approach that best matches the patient’s preferences and values; its Ask Share Know resources support it.3 Shared decision-making matters most when there is more than one reasonable option and the right choice depends on what the person values.
| Question | What a good answer covers |
|---|---|
| What are my options? | Every reasonable option, including waiting |
| What are the possible benefits and harms of those options? | Both, honestly, in plain words |
| How likely are the benefits and harms to happen to me? | Numbers where possible, tailored to the person |
In scenarios where someone wants something you wouldn’t recommend, a strong answer lays out options and trade-offs and invites their view, rather than either agreeing or refusing outright.
Think of a time you had to make a decision without all the information you wanted. How did you decide, how did you communicate it, and what did you do when more information arrived?
6. Humility: knowing when to ask ★★★★
Part of good judgement is knowing the limits of your own. Registrars who ask for help early are safer than those who never ask. In Casper, a short line such as “I’d check this with my supervisor” or “I’d want advice before deciding” shows self-awareness, as long as it comes alongside your own reasoning rather than instead of it.
Recall a time you changed your mind after asking someone else for their view. What did they see that you didn’t?
7. Worked scenarios
Try each scenario under the real time limit before you open the debrief. Your first instinct on the page is what you are training.
CommunicationProblem-solvingEthics
You are a GP registrar. Mr Peter Collins, 55, has come in asking for a blood test to screen for prostate cancer. He has no symptoms. His wife, who has come with him, has been urging him to have it after a friend’s diagnosis. Your supervisor looks in between patients.
Peter Collins (patient): My wife’s been on at me since Dave got diagnosed. Just do the test so she’ll stop worrying.
Dr Anita Rao (your supervisor): Just tick the box and order it. It’s quicker than a long conversation, and people expect it.
Helen Collins (Mr Collins’s wife, who has come in with him): Please just do it. I couldn’t bear it if we missed something, like with Dave. Is there any reason he shouldn’t?
- How would you respond to Mr Collins?
- How would you respond to Dr Rao’s suggestion?
Open the debrief
What is really going on
A screening decision with genuine trade-offs, where the right choice depends on the person’s values. The request is driven partly by his wife’s worry and a friend’s story (availability). The supervisor suggests skipping the conversation for efficiency. His wife has invited a discussion by asking whether there’s any reason not to.
Perspectives to acknowledge
- Mr Collins: wants to reassure his wife; the decision is his.
- His wife: frightened by a friend’s illness.
- Dr Rao: time pressure; probably well-meaning.
- You: responsible for an informed decision.
A reasonable approach
- Acknowledge his wife’s worry and the friend’s diagnosis.
- Explain, in plain language, that the test has possible benefits and possible harms, and that it’s a personal decision.
- Ask what matters most to him; offer a decision aid or written information they can read together; acknowledge Helen’s fear while keeping the decision his.
- Respect his decision either way and record the discussion.
- With Dr Rao, acknowledge the time pressure and suggest a brief shared decision, or booking a follow-up if time is short.
“He’s asked for it and Dr Rao agrees, so I’d order the test; it’s his choice and it’s only a blood test.”
Treats a request as an informed decision; skips the trade-offs the patient asked about.
“I’d say it’s understandable after Dave’s diagnosis, and that it’s a good question. I’d explain the test has some possible benefits and some possible downsides, so it’s worth deciding what suits him, and offer him some reading or a decision aid to share with his wife. I’d respect whatever he chose. I’d tell Dr Rao I’ll keep it brief, or book him back if needed.”
Answers the question honestly, centres his values, involves his wife appropriately, and handles the supervisor respectfully.
These are illustrative patterns, not scripts to memorise: many different responses can score well.
Ticked: of 5
Problem-solvingCollaborationCommunication
Outside work, you lead a volunteer bushwalking group of eight people on a day hike in a national park.
At the halfway point, at noon, the sky to the west is darkening. The morning forecast gave a 40% chance of afternoon storms. Continuing on the planned loop takes about three hours over an exposed ridge; turning back takes about two and a half hours through forest. Three experienced members want to push on and say storms usually miss this area. One member, Grace, is slower than expected and quietly tells you her knee is sore. Mobile reception is patchy. Another member says “you’re the leader, just decide”.
- How would you decide whether to continue or turn back?
- How would you communicate your decision to the group?
Open the debrief
What is really going on
A decision under uncertainty with safety at stake: incomplete weather information, an exposed route, a member with a sore knee, and social pressure from confident members (anchoring on ‘storms usually miss’). As leader you are responsible for the group’s safety.
Perspectives to acknowledge
- Grace: sore knee, reluctant to slow the group.
- Experienced members: confident, disappointed to turn back.
- The group: wants clear leadership.
- You: responsible for everyone’s safety.
A reasonable approach
- Gather quick information: check the weather if reception allows; check with Grace privately about her knee.
- Weigh options: exposed ridge with possible storm and a sore knee against a longer but sheltered return.
- Choose the safer option given uncertainty (turning back through the forest) and say why.
- Communicate clearly and respectfully: acknowledge disappointment and the experienced members’ views; don’t single out Grace.
- Plan pace, breaks and check-ins; report the decision to the club afterwards.
“The experienced members know the area, so I’d go with the majority and continue, and ask Grace to try to keep up.”
Lets confidence and majority override safety; ignores the injured member.
“I’d quickly check the weather if I could and ask Grace quietly about her knee. With storms possible, an exposed ridge and a sore knee, I’d turn back through the forest. I’d tell the group I know it’s disappointing and that the experienced members might be right, but I’m not willing to bet on the ridge in a storm. I wouldn’t single Grace out, and we’d set an easy pace.”
Gathers information, weighs uncertainty, prioritises safety, and communicates with respect.
These are illustrative patterns, not scripts to memorise: many different responses can score well.
Ticked: of 5
CommunicationEmpathyEthics
Your close friend Lina, 32, has been taking prescribed medication for anxiety for a year and has been doing well.
Over dinner, Lina tells you she stopped her medication two weeks ago after reading online about a herbal supplement that “cures anxiety naturally”. A wellness influencer she follows says prescription medicines are “poison”. She has been sleeping badly this week but says that is “just detox”. She mentions she hasn’t told her GP because she “felt judged last time”. She asks what you think, adding that she knows you’re a doctor but she doesn’t want a lecture. You have known her since school and she has always trusted you.
- How would you respond to Lina?
- How do you decide what health information to trust?
Open the debrief
What is really going on
A friend has made a health decision based on unreliable information, and may be experiencing effects of stopping medication. She has asked for your view but not a lecture. You are her friend, not her doctor, so your role is to be honest and supportive and to encourage her to see her own GP.
Perspectives to acknowledge
- Lina: wants control over her health; wary of being judged.
- You: friend with medical knowledge, but not her treating doctor.
A reasonable approach
- Thank her for telling you and acknowledge her wish to feel better naturally.
- Ask what she has read and how she has been feeling, without judgement.
- Share honestly and briefly that stopping some medicines suddenly can cause problems, and that the poor sleep could be related.
- Encourage her to talk to her own GP soon about her plans; offer to go with her.
- Respect her choice while staying available; check in again.
“I’d tell her the influencer is talking rubbish and she needs to restart her medication straight away.”
A lecture she asked not to receive; acts as her doctor; may push her away.
“I’d thank her for telling me and say I get wanting something natural. I’d ask how she’s been since stopping, then say honestly that stopping some medicines suddenly can cause problems like poor sleep, and that it’s worth talking to her GP soon. I’d offer to go with her. For trust, I look at who’s behind the information, what the evidence is, and whether reputable bodies agree.”
Respectful, honest, keeps the friendship, points to the right care, and gives clear criteria for trust.
These are illustrative patterns, not scripts to memorise: many different responses can score well.
Ticked: of 5
Tick each item you could do in a real Casper sitting without looking back at this lesson.
- Not yet? Re-read section 2 →
- Not yet? Re-read section 2 →
- Not yet? Re-read section 3 →
- Not yet? Re-read section 3 →
- Not yet? Re-read section 4 →
- Not yet? Re-read section 5 →
- Not yet? Re-read section 6 →
Ready: of 7. Your ticks are not saved.
8. Key takeaways
- Casper leaves out facts on purpose: ask, don’t assume.
- Fast thinking is efficient but carries biases: anchoring, confirmation, availability, premature closure, attribution, framing.
- Use conditional language rather than inventing facts.
- Manage uncertainty: acknowledge, rule out danger, plan, safety-net, follow up.
- An ‘if/then’ sentence is the most efficient way to show judgement.
- Evidence-based practice combines research evidence, clinical expertise and the person’s values.
- Judge information by its source, evidence, consistency, currency and fit.
- Shared decision-making lays out options, benefits and harms, and decides with the person.
- Offer options, not orders, when someone wants something you wouldn’t recommend.
- Asking for help is good judgement when it accompanies your own reasoning.
Practice plan
| When | Task |
|---|---|
| Today | Complete the three worked scenarios under time |
| Day 3 | Add an ‘if/then’ sentence to five old practice answers |
| Day 7 | Four typed and two video scenarios with missing information |
| Day 14 | Add a ‘decision without all the information’ entry to your experience bank |
| Week before your test | Mixed timed set; underline every assumption you made and remove it |
Your next step
Choose the line that matches your self-ratings and scenario self-checks in this lesson.
References
- Royal Australian College of General Practitioners. National entry assessment. https://www.racgp.org.au/a-career-in-general-practice/become-a-gp/national-entry-assessment (accessed 17 Nov 2026).
- Medical Board of Australia. Good medical practice: a code of conduct for doctors in Australia (effective 1 October 2020). https://www.medicalboard.gov.au/Codes-Guidelines-Policies/Code-of-conduct (accessed 17 Nov 2026).
- Australian Commission on Safety and Quality in Health Care. Shared decision-making. https://www.safetyandquality.gov.au/clinical-topics/shared-decision-making (accessed 17 Nov 2026).
- Croskerry P. The importance of cognitive errors in diagnosis and strategies to minimize them. Acad Med 2003;78(8):775–80.
- Sackett DL, Rosenberg WM, Gray JA, Haynes RB, Richardson WS. Evidence based medicine: what it is and what it isn’t. BMJ 1996;312(7023):71–2.