“I asked about Ideas, Concerns and Expectations. So why did I still lose marks?”
I hear this frequently from GP registrars preparing for the MRCGP SCA. It feels confusing.
You followed the framework. You asked the questions. You ticked the boxes.
Yet the result came back with poor marks in Relating to Others.
Here is the short answer: most GP registrars elicit ICE but do not integrate it. They complete the first half of the process and miss the second half entirely. And it is the second half that examiners are actually looking for.
After coaching over 357 IMGs and resitters, I have reviewed hundreds of recorded SCA consultations. The ICE problem shows up in almost every one. This article breaks down exactly what is going wrong, why it costs marks across all three SCA domains and what to do differently.
What Is ICE in the MRCGP SCA?
For GP registrars newer to the framework, a quick definition before we go deeper.
ICE stands for:
- Ideas — what the patient believes might be causing their symptoms
- Concerns — what they are worried or frightened about
- Expectations — what they are hoping you will do for them today
ICE was introduced into UK consultation theory by Pendleton and colleagues in the 1980s, who noticed that a small group of GPs consistently achieved better outcomes by focusing on the patient’s experience of illness rather than simply pursuing a diagnosis. That observation became foundational to how RCGP now assesses consultation quality.
In the SCA, ICE matters across all three marking domains:
| SCA Domain | How ICE Affects Your Mark |
|---|---|
| Data Gathering & Diagnosis | ICE reveals the real reason for attendance and narrows your questioning |
| Clinical Management | A plan anchored to the patient’s ICE is patient-centred |
| Relating to Others | Readdressing ICE is the primary evidence of genuine communication |
What ICE is not is a set of three questions you insert into the consultation. That misunderstanding is precisely what costs marks.
The Three ICE Mistakes I See Every Week
Mistake 1: Asking ICE as a Checklist
This is the most common pattern. The trainee reaches a natural pause in their history, often around minute four or five and then shifts gear:
“What do you think might be causing this?”
“And what’s worrying you most?”
“What were you hoping I’d be able to do today?”
Three questions. Brief answers. Consultation continues.
The problem is the rhythm. Those three questions delivered in rapid succession signal to the patient that these are tickbox questions rather than genuine curiosity. The patient responds accordingly: briefly, guarded, without depth.
The examiner hears a doctor moving through a scripted sequence rather than listening to the person in front of them.
In a recent session, I was coaching Taha, a registrar preparing for his June exam. He had developed a sensible workaround: he would write I, C, and E on his whiteboard during the reading time and tick them off as he went. It felt safe.
But in a care home case, the patient’s representative mentioned early on that a close friend of the resident had just been hospitalised after a fall. Taha acknowledged it briefly and moved on to his ICE questions.
He ticked all three boxes. He also missed the most important information in the entire consultation: that a vulnerable man was withdrawing from the world because his closest companion had just disappeared from his daily life. That was the answer to the case. It passed by unremarked because ICE was being collected rather than heard.
What to do instead: Treat ICE as an exploration. Your goal in the first three minutes is to understand three things about this person: what they think is happening, what they fear, and what they want. Let those emerge through active listening and cue-following rather than formulaic questioning. One well-placed open question can surface all three.
Mistake 2: Asking ICE Too Late
The second pattern is timing. Many registrars understand that ICE matters but treat it as something to address once the clinical history feels complete, usually around minute six or seven.
This is too late. And here is why it matters.
Stella came to me having narrowly failed her first SCA attempt. She was capable, warm with patients and genuinely person-centred in her practice. Her recordings showed competent history-taking. But her marks in Data Gathering were inconsistently low.
When I reviewed a syncope case with her, the pattern became clear.
Stella spent the first six minutes taking a thorough biomedical history — symptoms, red flags, family history, medications. At minute seven, she asked what the patient was worried about.
The patient told her she had a three-year-old at home and was terrified about what would happen to her daughter if she had another blackout while they were alone together.
Stella acknowledged it. Then she moved into investigations.
The problem was that this piece of information i.e a mother alone with a young child, frightened of collapsing while caring for her should have shaped the entire consultation from minute two. It told us what mattered most to this patient, which questions to prioritise and how to frame every element of the management plan. It also pointed directly toward the social history that was missing from Stella’s assessment.
ICE elicited at minute seven was too late to do that work. Stella had already spent five minutes on a general history when the patient’s specific story would have guided a more focused and more person-centred history.
What to do instead: Aim to have all three ICE components surfaced by minute three. Often, one well-timed opener“What made you decide to come in today about this?” or “What’s been going through your mind about all of this?” will bring ideas, concerns and expectations into the conversation naturally. Then you build your history around what the patient tells you.
Mistake 3: Eliciting ICE but Not Readdressing It
This is the mistake that most surprises registrars when I point it out. It is also the most expensive one, because it affects marks in all three domains simultaneously.
Maria is a registrar I coached recently. She is warm, clinically strong, and more self-aware than most trainees I work with. When I asked her to reflect on her own weaknesses, she mentioned re-addressing ICE.
We ran a depression case together. A veteran presenting with low mood. Maria took a thoughtful history. She explored the psychosocial context with genuine care. She asked what was worrying him most, and he told her clearly: he was terrified of being re-admitted to a psychiatric unit. He had been there before and the experience had been traumatic.
Maria heard it. She acknowledged it. Then she moved into management.
She explained her clinical thinking. She offered a plan. She addressed his reluctance around medication. The consultation was reasonable.
But when I reviewed it back to her, I asked: “Was the patient’s fear of re-admission addressed?”
She was quiet for a moment.
The management plan had been built around the clinical presentation rather than around the person.
And this is precisely what examiners are trained to spot. When a management plan fails to connect back to the patient’s stated ideas, concerns and expectations, it signals a doctor-centred consultation regardless of how warm the history-taking was.
What to do instead: Before you move into management, run a quick internal check: “What did this patient tell me was worrying them most? Have I built my plan around that? Have I said it explicitly?”
Readdressing ICE in management sounds like this:
“You mentioned earlier that your biggest fear is going back into hospital. What I’m suggesting today is aimed at helping you avoid that.”
Or: “You told me you were worried this might be something serious like your father’s heart condition. I want to come back to that now and be clear about what I’ve found and what it means.”
Those sentences do three things simultaneously. They show the patient they were heard. They show the examiner that the consultation has been genuinely person-centred. And they make the management plan feel personal because it is built around what this specific person told you mattered.
What the RCGP Marking Scheme Is Actually Looking For
Understanding how RCGP examiners assess ICE helps clarify why readdressing matters so much.
The Relating to Others domain assesses whether the doctor communicates effectively, shows genuine empathy and involves the patient as an active decision-maker. Formulaic ICE exploration is explicitly cited as a reason for poor marks in this domain. Examiners are not looking for whether you asked the questions; they are looking for whether the patient’s agenda shaped the consultation.
The Clinical Management domain assesses whether the plan is safe, evidence-based, and tailored to the patient. Generic plans without specificity to the patient’s ICE, context and comorbidities fail this domain even when the clinical content is correct.
The Data Gathering domain is affected by when and how ICE is explored. Asking ICE early and using the patient’s answers to guide targeted questioning is what high scorers do. Asking ICE late, after a generic biomedical template has been completed, produces the kind of unfocused history that misses psychosocial cues and key discriminators.
This is why the ICE loop — elicit, integrate, readdress is the structural mechanism through which you demonstrate competence across all three marking domains in every single case.
The ICE Loop: A Framework for the Full Process
Most training resources describe ICE as a technique for data gathering. What they underemphasise is that ICE is a loop with three distinct phases, each of which requires different skills.
Phase 1: Elicit (Minutes 2–3)
Create the conditions for the patient’s ideas, concerns and expectations to emerge. This means opening with genuine curiosity. Use the patient’s opening statement to guide your first follow-up question.
High-scoring opener: “What’s been going through your mind about all of this?”
Listen actively for cues — a mention of a family member with a similar diagnosis, a hesitation when discussing work, an anxious reference to something they read online. Follow those cues. They are often the route to the concern before you have to ask for it directly.
Phase 2: Integrate (Minutes 3–6)
Use what the patient has told you to guide your history. Their ICE should narrow your questions, not sit as a separate block of data. If a patient tells you they are worried this might be related to their father’s heart attack, you now know which questions matter most. If a patient tells you they are hoping to avoid surgery, you know how to frame every option that follows.
Integration means: your clinical questioning is shaped by what the patient told you.
Phase 3: Readdress (Minutes 6–12)
When you move into explanation and management, return to what the patient shared. Name it explicitly. Build your plan around it. Safety-net in a way that connects to their specific concern.
Readdressing means: the patient hears that what they told you actually changed what you are recommending.
ICE Phrases That Work in the SCA (And the Ones That Don’t)
Phrases to Avoid
These sound mechanical, feel forced, and signal checklist consulting to the examiner:
- “Do you have any ideas about what might be causing this?”
- “Any concerns you’d like to share?”
- “What were your expectations for today?”
- Asking all three in direct sequence within 30 seconds
Phrases That Work
These surface ICE naturally and invite genuine disclosure:
Ideas:
- “What’s been going through your mind about what might be causing this?”
- “Have you had any thoughts about what this could be?”
- “What made you decide to come in today about this particular problem?”
Concerns:
- “What’s worrying you most about all of this?”
- “Is there something specific you were hoping we’d be able to rule out?”
- “You mentioned [X] — tell me more about what’s been going through your mind there.”
Expectations:
- “What were you hoping we might be able to do today?”
- “Is there something specific you were hoping I’d be able to help with?”
- “What would feel like a helpful outcome from today for you?”
For Readdressing in Management:
- “You mentioned earlier that you were worried about [X]. I want to come back to that now — here’s why I’m reassured / here’s how my plan addresses that.”
- “Given what you’ve told me about wanting to avoid [X], the approach I’d suggest is [Y] — and here’s why I think it fits with what you’re looking for.”
- “I’ve taken on board what you said about [concern]. That’s shaped the plan I’m recommending.”
A Common SCA Case Type: How ICE Elevate Your Consultation
To make this concrete, consider one of the most common SCA case types: a patient presenting with a request for something you cannot or should not provide — a specific medication, an inappropriate referral, a test without clinical indication.
Most registrars approach this as a negotiation problem. They hear the request, acknowledge it, and then try to redirect the patient toward a more appropriate management plan.
The problem is that this sequence skips entirely over the ICE that would make the negotiation possible.
Why does this patient want this specific thing? What do they believe it will do for them? What are they frightened will happen if they do not get it? What is driving the request beneath the surface?
Without that information, the “negotiation” is simply two agendas in conflict. The doctor’s clinical judgement against the patient’s stated preference.
When you elicit ICE first — genuinely, the conversation changes. You often discover that the patient’s expectation is not actually the thing they asked for. A patient requesting antibiotics for a viral infection is often asking for reassurance that nothing serious is being missed. A patient requesting a sick note may be asking for permission to take the rest they cannot give themselves. A patient requesting a specific scan may be carrying a fear that was never addressed.
When you understand the concern underneath the expectation, you can address the concern. And when the concern is addressed, the expectation often resolves itself.
This is what high-scoring candidates do. They do not argue with the patient’s agenda. They understand it.
How to Practice ICE for the SCA: A Weekly Routine
The Recording Review Exercise
After each practice consultation, review the recording specifically for ICE. Ask three questions:
- At what minute did I first surface each ICE component?
- How did the patient’s ICE change what I asked in the history?
- Where exactly in management did I return to the patient’s concern?
If you cannot point to the specific moment where you addressed what the patient told you was frightening them, that is your work for the next week.
The Six-Minute ICE Checkpoint
Before you move from data gathering into management, pause internally and confirm: do I know this patient’s idea, concern and expectation? If any of the three is genuinely unclear, take one more minute to surface it before moving on. A management plan built with ICE will be personalised.
The Readdress Drill
Take any management plan you have given in a practice case. Read it back. Then ask: “Could this plan have been given to any patient presenting with these symptoms? Or does it contain something that is specific to this particular person?”
If the answer is the former, the plan needs one addition: a direct, explicit acknowledgement of what the patient told you, and a sentence connecting your recommendation to that acknowledgement.
How did this IMG pass MRCGP SCA?
After a failed SCA attempt with a score of 69, Dr Lawrence’s confidence was crushed. Check out how Dr Lawrence conquered SCA after this major setback.
Frequently Asked Questions
Q: Should I always ask all three ICE questions in every case?
You should always understand the patient’s ideas, concerns and expectations before moving into management. How you surface that information should be responsive to the patient. In some cases a single open question will bring all three elements naturally. In others, you will need to follow specific cues.
Q: What if the patient does not seem to have a strong idea or concern?
If a patient seems to have no strong view, it usually means the question was asked in a way that did not invite disclosure, or was asked too early before trust was established. Try asking later, in context: “Now that we’ve talked through what’s going on — is there anything specific you were worried this might be?” Often, that phrasing after some history-taking will surface something that a direct early question would not.
Q: Does readdressing ICE take up extra time?
No. One well-placed sentence in your management plan “Given what you’ve told me about your concern that this might be related to your father’s condition, I want to reassure you that..” — takes ten seconds. What it does is transform a generic management plan into a person-centred one. The time investment is negligible. The mark difference is significant.
Q: I’m an IMG. Does ICE feel culturally different in UK general practice?
This is one of the most important questions I help IMGs work through. In many healthcare settings outside the UK, the doctor’s authority is clearer and the patient’s role in the consultation is more passive. Patients in the UK are active participants in deciding their own care. Eliciting ICE is the process through which you demonstrate that you understand this contract. It can take time to feel natural. The way to accelerate that process is deliberate practice with targeted feedback.
The Deeper Point About ICE
I want to end with something that goes beyond technique.
Every time I watch a high-scoring consultation I notice the same quality. They are genuinely curious about the person in front of them.
They want to know what this specific patient believes. What they fear. What they hope for. And that curiosity is how they practise in their clinic.
The exam rewards curiosity — the kind that makes patients feel understood and that shapes every clinical decision that follows.
When you approach ICE as a technique, you will always be fighting to remember to ask it, to ask it at the right time, to make it sound natural, to come back to it in management.
When you approach ICE as a compass to orient the consultation. It becomes simply what happens when you listen well.
That is the shift that changes scores.
And it is available to every GP registrar willing to practise deliberately, seek honest feedback, and trust that the doctor they already are in clinic is exactly the doctor the SCA is designed to assess.
Summary: The ICE Loop for the MRCGP SCA
| Phase | When | What It Looks Like | Common Failure |
|---|---|---|---|
| Elicit | Minutes 2–3 | Open questions, cue-following, genuine curiosity | Asked as three rapid sequential questions |
| Integrate | Minutes 3–6 | Clinical questions shaped by ICE | ICE gathered but history continues from template regardless |
| Readdress | Minutes 6–12 | Management explicitly connects to patient’s stated concern | Plan is clinically sound but generic without considering ICE |