United Kingdom version. Set to the United Kingdom: prompts use NHS terms, NEWS2 and trust escalation policy, mmol/L and BNF names, follow the NMC Code on records, and the privacy check masks NHS numbers and postcodes. US version

AI Prompts for Nurses (UK)

Prompt templates for UK nurses: ward handover, NEWS2 escalation calls, care plans and patient leaflets, with NHS number masking before you paste anything.

The interactive tool loads here when JavaScript runs.

About this tool

What is AI Prompts for Nurses?

This version opens with the region set to the United Kingdom, so every prompt carries four UK constraints. The assistant writes British English with the 24-hour clock, keeps glucose in mmol/L, copies medicine names and doses exactly as written (the BNF is where you check them), and talks about the ward, the nurse in charge, the registrar, GP, A&E, NHS 111 and 999 rather than "the provider". Where your observations include a NEWS2 score it uses the trust's escalation thresholds only as you state them, and otherwise leaves a [CHECK LOCAL NEWS2 ESCALATION POLICY] marker.

Two templates change most. The escalation call script is framed for a UK ward: who you bleep, the NEWS2 score if you gave one, and the response time your trust policy sets as a marker for you to fill. The practice-question template, used by many internationally educated nurses preparing for the NMC test of competence, writes single-best-answer questions in UK practice terms and points you to the NMC's own preparation materials rather than claiming to match the real computer-based test.

Before you copy, the privacy check masks what turns up in UK notes: names after "Patient:" or a title, NHS numbers (with or without the label, checked against the modulus 11 check digit), dates of birth, phone numbers and full postcodes. Bed numbers and ward names are not caught, so take those out yourself.

  • Ten nursing templates from SBAR and care plans to incident wording and student practice
  • Teach-back questions built into every patient teaching prompt
  • Identifier scan that masks titled names, MRNs, dates, phones and emails
  • What-not-to-paste list and a nursing review checklist next to each prompt
  • Starred favourites kept in your browser and a Markdown copy of the whole kit
Read the guide: AI for Doctors and Nurses: Safe Prompts for Notes, Letters and Teaching

Updated

How to use AI Prompts for Nurses

  1. Enter your unit or setting, such as medical-surgical ward or community nursing, and pick your region.
  2. Choose a template: SBAR handoff, escalation call script, teach-back teaching plan, discharge education, family update or one of the others.
  3. Type your observations and notes, keeping out names and record numbers, or load the fictional example to see how it reads.
  4. Mask anything the privacy check flags, copy the prompt and paste it into the assistant your employer allows.
  5. Check the draft against your own assessment, the chart and local policy before you say or file anything.

Worked examples

Masking a ward note before handover drafting

Input

Patient: Joan Pryce, NHS number 943 476 5919, bed 12, NEWS2 5 at 14:00, lives at CF14 4XW.

Output

Patient: [NAME-1], NHS number [RECORD-ID-1], bed 12, NEWS2 5 at 14:00, lives at [POSTCODE-1].

Three identifiers are masked: the name, the NHS number and the Cardiff postcode. "Bed 12" stays, because the check does not know ward layouts; delete it by hand. The NEWS2 score and time are clinical facts and are left for the handover.

Escalation call on a surgical ward

Input

Escalation call script (SBAR), region United Kingdom, day 1 after a bowel resection, NEWS2 rising from 3 to 6, heart rate 118, urine 15 ml/hr for two hours

Output

SBAR script with the NEWS2 score stated, a clear ask for a review, the trust response time left as [CHECK LOCAL POLICY] and a read-back line

A score of 5 or 6 usually triggers an urgent review under NEWS2-based policies, but the exact response time and who attends differ between trusts, so the script asks you to fill it from your own policy instead of guessing.

Background and insight

SBAR (Situation, Background, Assessment, Recommendation) originated in military communication and was adapted for healthcare in the early 2000s as a way to make urgent conversations predictable. A model is good at reshuffling rough notes into those four boxes; it is not able to judge whether a patient is deteriorating. The escalation template therefore asks you for the observations and your concern, and it drafts the words you say, with the recommendation in your own terms.

Care plans are where AI drafts need the most care. The care-plan template asks for nursing problems, goals and interventions as a draft for your review and tells the model not to add interventions outside what you list or your unit's protocols. Nursing diagnoses taken from a formal taxonomy should be checked against the version your organisation uses, because models often paraphrase labels.

Patient data rules are the same as for any clinician. In the US a vendor that handles protected health information needs a HIPAA business associate agreement, and in the UK and EU health information is special category personal data under GDPR Article 9. Pseudonymised text is still personal data under GDPR Recital 26, so masking reduces risk but does not make a consumer chatbot an approved place for patient records.

Common use cases

  • Shift handoff

    Turn scribbled notes into a consistent SBAR handoff so nothing important is lost at the 7 o'clock change.

  • Escalating concerns

    Rehearse what to say to a provider or rapid response team when a patient is getting worse, with a clear recommendation.

  • Patient and family teaching

    Write medication sheets, discharge teaching and family updates in plain words with teach-back checks.

  • Nursing education

    Students and educators generate practice questions with rationales and summarise a long policy into a checklist.

Tips and limitations

  • If your trust has approved an AI tool covered by a data protection impact assessment, use that for anything you cannot fully de-identify.
  • Read every observation and score in a draft against the chart or e-obs system; the model only reorganises what you typed.
  • For revalidation reflective accounts, write the reflection yourself and use AI only to tidy structure; the NMC expects it to be your own learning.

Sources and further reading

  1. The Code: Professional standards of practice and behaviour for nurses, midwives and nursing associates, Nursing and Midwifery Council. Section 10 on records and section 5 on confidentiality.
  2. National Early Warning Score (NEWS) 2, Royal College of Physicians
  3. NMC new Test of Competence launched, Nursing and Midwifery Council. The computer-based test and OSCE for nurses trained outside the UK.
  4. Data Protection Act 2018, legislation.gov.uk

Frequently asked questions

Can NHS nurses use ChatGPT for handover notes?

Only within your employer's policy. Patient information is special category data under the UK GDPR and is covered by the common law duty of confidentiality, so a consumer chatbot is not an approved place for it. Use your trust's approved tool or remove every identifier and paste the minimum.

What does the NMC Code say about records written with AI help?

The Code does not mention AI, but section 10 applies: records must be accurate, clear and completed as soon as possible after the event, and you are accountable for what you sign. A draft from an assistant is your note once you enter it.

Will the prompts work out a NEWS2 score for me?

No. The UK rules tell the assistant never to invent a score and to use only the one you give. Calculate NEWS2 from your observations on the chart or e-obs system, then let the model put it into the SBAR wording.

Are the practice questions the same as the NMC CBT?

No. They are written in the CBT's single-best-answer style from your own study notes, and the prompt tells the model not to claim otherwise. Use the NMC's published candidate materials as the reference and check each rationale.

Does masking the NHS number make the text anonymous?

No. It removes one direct identifier. A rare condition, a date and a ward can still point to one person, and pseudonymised data remains personal data under the UK GDPR. Read the masked text before you copy it.

Can I use this for community or district nursing notes?

Yes. Put "district nursing team" or "community nursing" in the setting field. The UK rules still apply, and addresses matter more in community notes, so check that street names are removed as well as postcodes.

Is there an in-depth guide to AI Prompts for Nurses?

Yes. “AI for Doctors and Nurses: Safe Prompts for Notes, Letters and Teaching” covers the background, how it works, common pitfalls and worked examples. Read the full guide.