Skip to main content

Build with AI

The Nurse Citizen Developer

Illustration of diverse NHS nurses collaborating around a whiteboard and table to prototype clinical workflows
Build with AICreator → Builder → ArchitectClinical Safety & Governance First
The Shift

Nurses don’t just use technology; they shape and build it — providing clinical logic while AI assists with code.

Safe Prototyping Scope

Start with educational quizzes, staff study aids, and non-patient workflow prototypes using synthetic data.

Hard Safety Boundary

Never use a self-built calculator or app for live patient care without CSO (DCB0160) and MHRA SaMD approval.

Vision

"The future nurse won't just use technology; they will build it."

For decades, if a nurse had an idea for a digital tool (e.g., a better shift planner, a revision simulator, a patient education guide), they had to hire a developer or wait for IT.

Generative AI has changed this forever.

Tools like Claude Artifacts, v0.dev, and ChatGPT Canvas act as a "Technical Co-founder" for every nurse. You provide the clinical logic, and the AI handles the code.

The 3 Tiers of Nurse Builders​

We can think of this new capability in three tiers:

TierRoleToolingExample Project
1. The CreatorGenerating ContentCopilot / ChatGPT / Claude"Create a patient scenario quiz."
2. The BuilderBuilding Interactive Educational PrototypesClaude Artifacts / Canvas"Build an educational simulation prototype for practising fluid balance maths."
3. The ArchitectDesigning Logical SystemsAI Agents / Cursor"Design a logic tree for an educational sepsis escalation trainer."

Getting Started: Your First Educational Prototype​

You don't need to learn Python or HTML first. You start by learning how to describe the exact clinical logic and safety boundaries.

Citizen Developer Exercise · Educational Prototype OnlyClaude Artifacts / ChatGPT Canvas

Build a Paediatric Fluid Calculation Study Trainer

Write a simple React educational web app for nursing students to practise checking daily maintenance fluid calculations using the Holliday-Segar formula.

Banner required at top of UI:
"EDUCATIONAL SIMULATION PROTOTYPE ONLY — NEVER FOR CLINICAL USE WITH REAL PATIENTS."

Inputs:
- Fictional child weight (kg)

Logic to display step-by-step so students can verify their manual maths:
- 100 ml/kg/day for the first 10 kg
- 50 ml/kg/day for the next 10 kg (11–20 kg)
- 20 ml/kg/day for each kg above 20 kg

Output:
- Show the manual step-by-step working first, then the 24-hour total and hourly rate.
- Include a reminder to check NICE NG29 (IV fluid therapy in children and young people in hospital).

Mandatory Clinical Safety & Governance Check

  • Verify the maths manually: Test edge cases (e.g. 8 kg, 15 kg, 25 kg) against manual calculation and NICE NG29.
  • Why this cannot be used on a ward: Any software that calculates doses or clinical scores for real patients is regulated as Software as a Medical Device (MHRA SaMD) and requires formal DCB0129/DCB0160 clinical safety engineering.

What You Can Build Safely vs What Needs Full Regulation​

Green · Safe to Build & Share

Educational & Non-Data Tools

  • Interactive OSCE revision flashcards and glossary explorers.
  • Fictional case-study generators and team workshop templates.
Amber · Needs Local IG & Manager Sign-Off

Internal Operational Templates

  • SharePoint/Copilot agents that search approved, public-facing trust policies.
  • Non-patient shift handover or audit checklists built in your M365 tenant.
Red · Stop (Regulated Medical Device / IG Breach)

Live Clinical Calculators or Patient-Data Apps

  • Using a self-built dose, fluid, or NEWS2 calculator on real patients without MHRA SaMD & CSO sign-off.
  • Connecting any personal app or database to real patient or staff records.

Why This Matters for Education​

Teaching "Citizen Development" to nursing students is about empowerment:

  1. Problem Solving: It teaches them that digital problems are solvable, not fixed.
  2. Logic & Safety: To build an educational tool, you must understand the logic perfectly. If you can't explain the formula and its clinical exceptions, you can't verify the tool.
  3. Innovation: It unlocks the creativity of the frontline workforce.

The "Safety First" Rule​

Just like we double-check a drug calculation, we must double-check our tools.

Clinical Safety & Automation Bias

Never use a self-built tool for actual patient care without:

  1. Validation: Testing it against known manual calculations.
  2. Governance: Getting approval from your clinical safety officer (CSO).
  3. Disclaimer: Clearly marking it as an "Educational/Prototype Tool Only".

Beware the Polished Output: According to the Anthropic AI Fluency Index (2026), when AI generates functional, beautiful outputs (like web apps or code artifacts), users become less evaluative. They stop checking facts and questioning the model's reasoning.

If a tool looks highly professional, your brain will try to convince you the clinical logic inside it must be correct. It might not be. Do not let a beautiful UI turn off your clinical critical thinking.

The Future: AI Agents​

We are moving towards a world of AI Agents (such as coding and research agents). These are not just chatbots; they are systems that can:

  • Research complex guidelines.
  • Test their own code.
  • Deploy secure applications with human oversight.

For the Nurse Citizen Developer, this means the technical barrier to prototyping drops dramatically — making clinical judgement, ethics, and safety governance more important than ever.