You have been typing "Act as a patient presenting with chest pain to a GP" into ChatGPT and practising consultations. It works — sort of. But the gap between ChatGPT and purpose-built SCA tools is significant.
What ChatGPT Can Do
Generate case scenarios on any clinical topic. Role-play a patient with basic symptom responses. Provide general feedback on consultation structure. Help brainstorm management plans and differentials.
What ChatGPT Cannot Do
Mark against SCA criteria — it does not know the three-domain marking framework. Provide GP-trainer-level feedback — it cannot assess whether your data gathering was appropriate or your safety-netting was specific. Simulate realistic emotional responses — the patient does not push back, get angry, or express genuine concern. Track performance over time — each session is isolated.
The IG Risk
Creating SCA scenarios from real cases risks entering patient-identifiable information into a US-hosted consumer tool. Even "anonymised" scenarios based on real patients carry re-identification risk. Purpose-built tools generate cases without needing your clinical input.
Better Alternatives
MedTutor AI — voice-based, SCA-marked, GP-trainer reviewed. ~£10/3 scenarios. SCA Prep — AI patients with RCGP-criteria marking. From £14.95/month. SCA Revision — AI patients alongside 350+ written cases. From £11.99/month. Clinitalk — feedback on your real consultations. From £5/month.
The cost difference between ChatGPT (free) and purpose-built tools (£10-15/month) is minimal relative to the SCA exam fee (£1,207).
When ChatGPT Is Acceptable
Generating case ideas for peer practice (not from real patients). Brainstorming management plans post-practice. Exploring clinical topics after a simulation session. These are supplementary uses — not primary SCA preparation.
Where iatroX Fits
For clinical knowledge exploration, Ask iatroX provides guideline-grounded, citation-first answers — more reliable than ChatGPT for UK clinical content, and MHRA-registered for clinical use.