Practical AI for CPD
Example prompts for using AI tools in CPD. Copy them, replace the bracketed sections with your own details, and see what you get.
Before you start
Before using any AI tool for CPD work, keep three things in mind.
De-identify first. Never enter patient-identifiable information into an AI tool. Remove names, hospital numbers, dates of birth, and any combination of clinical details that could identify a patient. If the text appeared on a public website, would anyone be identifiable? If the answer is yes, or maybe, generalise further. De-identification happens before the AI sees your text, not during.
Choose your tool thoughtfully. Use a reputable general-purpose AI tool (ChatGPT, Claude, Gemini, or similar). Check your employer's policy on which tools are approved for professional use. Use a dedicated project or workspace for CPD work, and turn off any "use my data for training" setting.
You are the quality-control layer. Think of the AI as a very capable but unreliable junior colleague. It is well-read, writes fluently, works fast, and never complains about restructuring a document for the third time. But it will sometimes make things up, it does not know what it does not know, and it will never push back and say "actually, I think your reasoning is wrong." Every output needs your review, your judgment, and your professional sign-off.
Try these prompts
Copy, paste into your preferred AI tool, replace the bracketed sections with your own details, and see what you get. Each prompt follows the CPR-D method from the guide: Capture your raw input, Process it with AI, Refine the output with your clinical judgment, and Document the result.
1. Reflective writing
You had a significant clinical experience and want to turn it into a structured CPD reflection while the learning is still fresh. Start by talking or typing for two to five minutes about what happened, what surprised you, what you did well or would do differently, and what you want to learn more about. Then paste those rough notes into the prompt below.
Keep my voice and my specific observations. Do not add clinical content, references, or learning points that are not present in or clearly implied by my notes. If something is unclear, flag it as a question for me to address rather than filling in the gap.
My notes:
[paste your captured text here]
Tip: a two-minute voice memo recorded in the car park is more useful than a detailed reflection you planned to write that evening but never did.
2. Conference and workshop notes
You attended a conference, a workshop, a webinar, or listened to a podcast on your commute. You have messy notes you will probably never look at again. AI can organise them into a structured CPD learning summary in about 25 minutes, and you already have the attendance certificate.
Please organise these into a structured CPD learning summary with the following sections:
- Sessions attended (list with speaker and topic where I have noted them)
- Key learnings (grouped by theme, using only what I have written below)
- Surprises or challenges to my current practice
- Planned practice changes
- Follow-up actions
Use only the content I have provided. Do not add clinical detail, evidence, or recommendations beyond what appears in my notes. If a section is thin because my notes are thin, leave it thin.
My notes:
[paste your raw notes here]
Tip: dictate your notes on the drive or train home while the sessions are fresh. Capture and processing are separate steps.
3. Literature digestion
You have journal articles or guideline updates sitting unread in your downloads folder. AI can help you process them, but it is not a replacement for reading the paper. It gives you a structured starting point; you supply the judgment about what matters for your practice.
1. A one-paragraph plain-language summary of what the study investigated and what it found
2. Study design and sample size
3. Key findings (primary outcome and up to three secondary outcomes), with the actual numbers or effect sizes reported in the paper
4. Limitations acknowledged by the authors
5. Any limitations you can identify that the authors did not acknowledge
6. One paragraph on what this might mean for clinical practice in [your setting, e.g., "Australian general practice" or "a regional emergency department"]
Important: Only include information that appears in the article I have provided. Do not add findings, statistics, or references from other sources. If something is unclear or not stated in the paper, say so rather than filling the gap.
Tip: upload the PDF directly if your AI tool supports it. Always check that the key findings and numbers match the actual article.
More use cases in the book
The book covers nine CPD use cases with full workflows, governance notes, and tested prompts for each. These six are covered in detail:
Learning plans
Structure your annual CPD Plan and end-of-year reflection. Every registered practitioner must do it, no patient data is involved, and the output responds to known template questions. The book covers both the Q1 plan and the Q4 reflection as a paired workflow.
Practice improvement activities
Document audits, quality improvement projects, and practice reviews. AI organises messy notes into recognised frameworks (audit cycle, PDSA) while you supply the data and conclusions. The book includes governance guidance for handling aggregate patient data safely.
Case-based learning
Turn M&M meetings, case conferences, and corridor conversations into structured CPD entries. This is the most governance-sensitive use case because the raw material is a real patient encounter. The book covers de-identification in detail.
Peer discussion preparation
Prepare for journal clubs, case presentations, and annual peer review conversations. AI sharpens your existing thinking rather than supplying opinions. The book includes separate prompts for literature-based discussions, case presentations, and structured peer review.
CPD summaries and annual statements
Pull together a year of CPD activities into a coherent narrative for annual reporting. AI tallies hours, maps activities to learning goals, and drafts your end-of-year reflection. The book addresses the key risk: making scattered activities look more deliberate than they were.
Exam preparation and self-assessment
Use AI as an interactive study partner for college exams or knowledge testing. The book covers how to set up structured quiz sessions, verify AI-generated answers against primary sources, and avoid the key pitfall: confidently wrong responses.