A clinical case review passes your Educational Supervisor's Report when it shows your reasoning, not when it reads like polished prose. Supervisors and ARCP panels are looking for evidence that you thought about the case, learned something, and can link it to the RCGP capabilities, which is a different thing from a well-written paragraph. Get the structure right, keep the reflection genuinely yours, and use AI only to support the mechanical parts. Here is how.
Key takeaways
- The ESR rewards demonstrated reasoning and learning, not fluent writing.
- Use a consistent structure: brief description, reflection, capability linking, and a learning need.
- Keep the brief description short; the reflection is where the marks are.
- Anonymise rigorously, since identifiable patient data is an information governance breach.
- Use AI to structure and to check management, but write the reflection yourself.
What the ESR is actually looking for
Your Educational Supervisor is trying to see how you think, not how you write. A clinical case review is evidence that you encountered a real case, reflected on it honestly, identified what you did not know, and linked the learning to the capabilities the RCGP expects you to demonstrate. A beautifully written entry with no genuine reflection is weaker evidence than a plainly written one that shows real thought. So write for insight, not polish, and make your reasoning visible on the page.
The structure that works
Use the same structure every time, because consistency speeds up writing and ensures nothing is missing: a brief description, then reflection, then capability linking, then a Doctor's Educational Need. Think of it as a bottom-heavy pyramid, with the least space on the description and the most on the reflection and learning. When you open a new clinical case review, keep the description to three to five sentences of context, then spend your effort on what you thought, what you would do differently, which capabilities the case demonstrates, and what you will learn next. That shape is what supervisors are trained to expect.
Keep the brief description short
Most trainees write too much here, and it is the most common structural mistake. The brief description exists only to give enough context for the reflection to make sense, so three to five sentences is plenty. It should not be a full case presentation with every investigation and value. A crisp description that sets up the reflection reads better and leaves room for the part that actually earns marks. Remember too that a clinical case review is for a case you personally saw and managed; if you only read about or observed it, use a different entry type such as CPD or supporting documentation.
Anonymise rigorously
This is the medico-legal non-negotiable. Strip anything that could identify the patient: name, date of birth, precise dates, unusual occupations, rare co-occurring conditions, and specific locations. Refer to the patient by age and relevant clinical detail only. The FourteenFish portfolio has an AI feature that flags potentially identifiable information before your supervisor sees it, which is a useful safety net, but the responsibility is yours. An entry that leaks identifiable data is an information governance breach in your own portfolio, and it is entirely avoidable.
Link a case to a capability convincingly
Capability linking is where many entries are thin. Do not just tag a capability; justify it with the specific thing in the case that demonstrates it. For a worked example, take a consultation with an anxious parent about a child's symptoms. To link it to communication and consultation skills, write that you elicited the parent's specific fear, checked their understanding, and reached a shared plan they were comfortable with. To link it to managing medical complexity or clinical management, write how you weighed the risk, decided on investigation or reassurance, and safety-netted. One case can evidence more than one capability if you justify each with a concrete detail rather than a label.
Use AI well, not lazily
AI can help, but the failure mode is outsourcing the reflection. The safe pattern is three steps: first, get the correct, guideline-grounded management from a knowledge tool so the clinical substance is right; second, write the reflection yourself, in your own words, because that is the part panels can tell you did or did not do; third, let a portfolio tool help structure and tidy the entry. If you invert this and let a general chatbot write the reflection, you produce a generic, polished entry that supervisors increasingly recognise, and you learn nothing. Ask iatroX gives you free, NICE and CKS-grounded answers to confirm the management in your case before you write it up. Try iatroX through its free questions, and see our comparison of portfolio AI tools.
Frequently asked questions
What makes a clinical case review pass the ESR? Demonstrated reasoning and genuine reflection, linked to specific RCGP capabilities, with an identified learning need. Fluent writing without real thought is weaker evidence than plain writing that shows how you reasoned.
How long should the brief description be? Three to five sentences. It only needs to give enough context for the reflection to make sense. Most trainees write too much here and too little in the reflection, which is where the marks are.
How do I anonymise a case review? Remove names, dates of birth, precise dates, unusual occupations, rare condition combinations, and locations. Use age and relevant clinical detail only. Identifiable data in your portfolio is an information governance breach.
Can I use AI to write my case review? Use it to check management and to structure the entry, but write the reflection yourself. Outsourced reflection is generic, detectable, and defeats the purpose of the exercise.
How do I link a case to a capability? Justify the capability with a specific detail from the case, not just a label. Describe the exact thing you did that demonstrates it, and one case can evidence several capabilities if each is justified concretely.