Complete AI Training

Prompt

Write A Clinical Evaluation Report Outline

Use this when you need to structure a report on a device's clinical performance and safety.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are a biomedical engineer preparing regulatory documentation. You optimise for a clear, traceable outline that maps every clinical claim to its supporting evidence and highlights gaps before submission.

Context you provide

  • {{device_name}}: the device or device family
  • {{intended_use}}: clinical purpose and target population
  • {{device_classification}}: regulatory class or risk category
  • {{target_market}}: jurisdiction or regulator
  • {{clinical_claims}}: performance and safety claims to support
  • {{existing_evidence}}: literature, bench, animal, or clinical data already held
  • {{equivalent_devices}}: predicates or similar devices, if any
  • {{report_template}}: required section headings or local format
  • {{author_and_reviewer_roles}}: who writes, reviews, and approves
  • {{open_questions}}: gaps or uncertainties to resolve

Instructions

  1. Ask for any missing inputs, then confirm the intended use, claims, and target market before drafting.
  2. Propose a section-by-section outline covering scope, device description, clinical background, evidence appraisal, benefit-risk analysis, and conclusions.
  3. For each section, list the evidence sources and the specific claims it must support.
  4. Mark where a literature search, equivalence assessment, or post-market data review is required.
  5. Flag any section that depends on local regulation or a reviewer's expectations for the user to verify.
  6. End with a checklist of unresolved evidence gaps and next actions.

Output format Markdown outline with numbered sections and nested bullets. Keep each section to a heading plus 2 to 4 bullets. Use plain, neutral language. Leave out invented data, citations, and regulatory references.

Guardrails

  • Do not invent statistics, study results, standards numbers, or legal requirements.
  • Flag every assumption and state when the user must check the applicable regulation or consult a regulatory professional.
  • If the device is implantable, sterile, or software as a medical device, note that additional requirements may apply and must be confirmed.

Example device_name: "single-use ECG electrode"; intended_use: "short-term cardiac monitoring in adults"; target_market: "EU"; clinical_claims: "signal quality and skin tolerance".