Navigating the Healthcare System with AI
A practical workshop for educators to teach migrant learners how to use AI as a communication bridge and information simplifier when navigating local healthcare services.
DigCompEdu Competence Area(s)
Digital resources, Teaching & learning, Empowering learners, Facilitating learners’ digital competence
Time
Preparation time 30 minutes
Implementation time 90 minutes
Educator Progression Level
Explorer
Target groups
Adult migrants A2 – B1 language level Group size: 10-15
Minimum student digital skill level
Basic
Minimum student language level
A2
🎓 Learning Objectives
By the end of this lesson, educators will be equipped to teach learners how to:
- Use AI to translate and simplify complex medical terminology or appointment letters into plain language.
- Role-play describing physical symptoms to a doctor using an AI chatbot to build vocabulary and confidence.
- Identify how to ask AI for local healthcare procedures (e.g. “How do I register with a GP?”) while verifying information with official sources.
Ingredients
| Item / Resource | Quantity | Purpose |
|---|---|---|
| Sample appointment letter (anonymised) | 1 per participant | For simplification practice |
| Symptom description guide | 1 per participant | Basic vocabulary for role-play |
| “Health Navigator” Prompt Book | 1 per participant | Tested prompts for medical context |
| Local health system map | 1 digital copy | Reference for verifying AI information |
| Device with internet access | 1 per participant | To access AI tools |
Utensils
- Smartphone or laptop per participant
- ChatGPT, Claude, or a translation-focused AI like DeepL
- Internet access (required)
- Projector or shared screen for demonstrations
🥣 Preparation – Before Implementation
- Source real materials: Find a standard medical appointment letter or a generic health insurance brochure (fully anonymised).
- Safety briefing: Prepare a clear statement on the limitations of AI – it is NOT a doctor and cannot provide medical diagnosis.
- Localise: Ensure you have the correct local terms for “General Practitioner,” “Emergency Room,” “Specialist Referral,” etc.
- Prepare the “Health Navigator” Prompt Book with tested prompts for medical contexts.
🔥 Implementation Steps
Part 1: The AI Translator (30 minutes)
- (10 min) Present a complex medical letter. Discuss the barrier of “medicalese” and jargon that makes healthcare inaccessible for migrants.
- (20 min) Demo using AI to simplify the letter: “Explain this letter like I am 10 years old.” Show how to extract key information: Date, Time, Location, and Preparation needed.
Part 2: The Symptom Role-Play (40 minutes)
- (20 min) Learners use a prompt to turn the AI into a “Friendly Doctor.” They practise describing a common ailment (e.g. a persistent cough or back pain) using the symptom vocabulary guide.
- (20 min) Partners review the AI’s response. Did the AI understand? Did the learner use the right verbs? Groups share useful phrases with the class.
Part 3: Verification & Next Steps (20 minutes)
- (10 min) Reflection: AI can get answers wrong – why are we only using it for conversational practice and checking letters, and not for diagnosis?
- (10 min) Learners use AI to find local healthcare procedures: “How do I register with a GP in [city]?” Then verify the answer against the local health system map.
🧂 Differentiation & Inclusion
- For low digital skills (CLAVIS 1): Focus on voice-to-text input so learners do not have to type complex medical words.
- For visual learners: Use AI to generate simple diagrams or icons representing symptoms.
- For learners with health anxiety: Keep role-plays light and focused on routine appointments, not emergencies.
- For multilingual groups: Allow learners to use AI in their home language first, then translate.
🍴 Assessment & Feedback
| Criterion | Excellent | Good | Needs Work |
| Letter simplification | Accurately extracts all key information (date, time, location, preparation) from the AI-simplified letter | Extracts most key information with minor gaps | Struggles to identify key information from the simplified letter |
| Symptom role-play | Describes symptoms clearly using appropriate vocabulary; AI understands the description | Describes symptoms with some vocabulary gaps | Struggles to describe symptoms; AI misunderstands the input |
| Critical evaluation of AI | Clearly explains why AI should not be used for medical diagnosis | Shows some understanding of AI limitations | Limited awareness of AI limitations in healthcare contexts |
⚠️ Ethics, Privacy & Safeguarding
- Privacy: Never input real names, ID numbers, or sensitive personal medical history into public AI tools.
- Accuracy: Stress that AI can “hallucinate” medical advice. Always consult a human professional for diagnosis.
- Safeguarding: Be prepared to refer learners to real healthcare services if health concerns arise during the session.
- Data protection: Ensure all sample documents used are fully anonymised before the session.
- Informed use: Teach learners to always verify AI-generated health information against official sources.
🍱 Extensions / Follow‑Ups
- “My Health Dictionary” Project: Learners use AI to create a personalised glossary of medical terms relevant to their health needs.
- Booking Simulation: Use AI chatbots to simulate booking a doctor’s appointment, practising phrases and understanding procedures.
- Community Health Resources: Research local community health initiatives or support groups using AI-powered search tools.
🔗 Cross‑Subject Link
References & Resources
Google Translate: https://translate.google.com
DeepL Translator: https://www.deepl.com
ChatGPT: https://chat.openai.com
Claude: https://claude.ai
NHS (UK): https://www.nhs.uk — for official health information (adapt to local national health service)