Keep one row per nurse with the register number, expiry or revalidation date, last check date, who checked and the evidence. A person does every register lookup. AI helps build the sheet, flag dates falling due and draft reminders, but never replaces the check.
This is about the ongoing job for employers: knowing, on any given day, that every nurse working for you is currently registered, and having proof. It is not about a one-off lookup when someone applies. A tracker turns a nervous annual scramble into a ten-minute weekly routine.
What the tracker is for, and what it is not
A small care provider might employ a handful of registered nurses directly, use bank or agency nurses, and rely on a clinical lead to supervise. Whoever employs or places a nurse is responsible for making sure they are entitled to practise. If a registration lapses and nobody notices, the consequences can range from an awkward conversation to a serious safeguarding and insurance problem.
The tracker has three jobs. It holds the facts: who is registered, to what, until when. It prompts action: who needs rechecking and when. And it proves the checks happened: a dated record an inspector, commissioner or insurer can see. It is not a replacement for the regulator's register. The register is the source; your tracker is a record of what you saw there.
The lookups themselves are done by a person. For a nurse registered with the Nursing and Midwifery Council (NMC), that means searching the NMC's online register, which is free; searching by the nurse's PIN (their unique registration code) gives the most accurate result, and the NMC says a nurse should tell you their PIN when asked. For nurses licensed by a state board, it means using that board's own lookup or a verification service it participates in, such as Nursys QuickConfirm, which is free and returns licence and discipline status from participating boards (registered nurse and LPN/VN licences only). Systems differ, and rules change, so write down the exact method each time, and confirm it on the regulator's own current guidance rather than from memory or from an AI answer.

The Complete Excel Course with AI Integration
From formulas and pivot tables to Power Query, DAX, VBA and AI with Copilot, ChatGPT and Claude — hands-on from the first lesson.
By Sagnik Bhattacharya and Virginia Thorn, Primetime Emmy Award winner
Buy nowStep 1: design the columns
Build the tracker as a shared spreadsheet or a simple database in a tool you already use. Restrict who can edit it, because it holds personal data. The columns that earn their place:
| Column | What goes in it | Why |
|---|---|---|
| Staff ID | An internal ID, not the full name in alert emails | Keeps alerts light on personal data |
| Name and role | As on the register | Matches your lookup to the person |
| Regulator or board | The register you check | Different nurses may sit on different registers |
| Registration number | As shown on the register | Lets anyone repeat the check |
| Registration type or scope | For example nurse, midwife, or a specific category | Duties must match what they are registered for |
| Status at last check | Active, conditions, restriction, lapsed | Any entry other than active triggers escalation |
| Expiry or revalidation date | As shown on the register (the NMC's results include an expiry date) | The date alerts are built from |
| Last checked on / by | Date and initials | The audit trail |
| Evidence | Link to a saved screenshot or PDF of the lookup result | Proof for inspection |
| Next check due | Formula from the dates above | Drives the weekly review |
| Employment type | Employed, bank, agency | Agency nurses often need their agency's confirmation as well |
| Notes | Conditions, conversations, escalations | Context for the clinical lead |
Two columns deserve extra thought. "Status at last check" must be a fixed list (a dropdown), not free text, so a formula can detect anything unusual. And "Evidence" should link to a stored file in a folder with restricted access, named consistently, such as staffID_registerName_2026-11-04.pdf.
Build the dropdown from the words your register actually uses, so the person doing the lookup never has to translate. The NMC's glossary (checked November 2026) shows how many states there are beyond "registered": not currently practising (the nurse has told the NMC they have stopped but stays on the register until the current period expires), conditions of practice order, interim conditions, undertakings, caution order, warnings, suspended by a fitness to practise panel, interim suspension order, sanction pending, removed and voluntarily removed. Some of these still allow practice. A nurse with a caution order, for example, can practise without restriction, but the clinical lead should still know about it. So record the exact register term in the Notes column, and map it to the dropdown: "Active" only for a plain registered entry with nothing recorded; "Conditions" or "Restricted" for anything recorded against a nurse who may still practise, which goes to the clinical lead; "Suspended", "Lapsed" or "Not found" for anyone who may not practise, which stops duties that need registration. Because the formatting rule in the next section turns every non-"Active" row red, nothing recorded on the register can sit unnoticed.
Building the sheet for a ten-nurse home nursing provider
Say a small home nursing and care provider has six directly employed nurses, three bank nurses and one agency nurse, all with registrations that renew at different times. The clinical lead currently keeps dates in her head and a diary.
She lists the ten people and their details on a sheet, pastes the headings from the table above, and asks an AI tool to build the formulas. She gives it the headings only, with no names or numbers:
I have a spreadsheet with these columns: A Staff ID, B Name, C Register,
D Registration number, E Scope, F Status at last check, G Expiry date,
H Last checked on, I Checked by, J Evidence link, K Next check due,
L Employment type.
Write formulas for:
1. Column K: next check due = the earlier of (a) 90 days after H and
(b) 14 days before G. If G is blank, use 90 days after H.
2. A helper column M "Days until due" = K minus today.
3. Conditional formatting rules: red if M is below 0, amber if 0 to 14,
green if more than 14. Also red if F is anything other than "Active".
4. A data-validation list for F: Active, Conditions, Restricted,
Lapsed, Suspended, Not found.
Give the formulas for Google Sheets and for Excel.
The AI returns formulas along these lines (illustrative): =IF(G2="",H2+90,MIN(H2+90,G2-14)) for K, =K2-TODAY() for M, and a set of conditional-formatting rules. She pastes them in, tests with fake dates, and finds a bug: when H is blank, K shows a date near 1900. She asks the AI to wrap the formula in IF(H2="","Check now",...), which fixes it. Total build time: about 45 minutes, with the data entry taking the longest.
The test matters. Before relying on any AI-written formula, type in invented rows that should come out red, amber and green and see whether they do. Then delete the test rows.
For Excel specifics, the ChatGPT Excel guide shows how to describe a formula problem so the answer is usable.
Step 2: set the check rhythm in writing
Decide your rules and put them in one paragraph of policy, so a new manager follows the same rhythm. A reasonable starting point (adapt to your regulator and insurer):
- Check registration before a nurse starts, with a saved result.
- Recheck every 90 days for bank and agency nurses, and at least before each renewal date for employed nurses.
- Check again on the day after any renewal date.
- Check straight away if you hear of a concern, complaint or incident that relates to practice.
- For agency nurses, obtain the agency's written confirmation of their checks and do your own register lookup too.
These are example intervals, not requirements. Your regulator, insurer or commissioner may set a minimum, and you should confirm theirs. What matters is that your policy says what you do, and that you do it.
Step 3: the weekly ten-minute routine
Pick a day and a named person. Each week:
- Open the tracker and sort by "Days until due".
- For every row red or amber, do the lookup on the official register. Do not take the nurse's word, and do not use a search engine result or an AI answer.
- Save a screenshot or PDF of the result to the evidence folder, using the naming pattern.
- Update status, expiry date, last checked and checked by.
- For anything other than "Active", follow the escalation steps below the same day.
With ten nurses, most weeks the list is one or two names. In the weeks around a cluster of renewals it may be five. That is why the amber window matters: it spreads the work.
Step 4: alerts that people actually read
A tracker nobody looks at is a spreadsheet. Add prompts that come to people rather than waiting for them to open a file. Options, from simplest to most automated:
Let the register send the alert
Before building your own reminders, check what the registers already offer employers. For state-board licences, Nursys e-Notify is free for institutions: you enrol your nurse list and it emails updates on licence expirations and status changes from participating boards. Board participation differs between Nursys services, so check the participating-boards list for each board your nurses hold licences with. For NMC registrants, the NMC's Employer Confirmations service lets an employer with a login (you request one from the NMC) check up to 10 PINs at once, which turns the weekly lookup for a small team into one search instead of ten.
Neither replaces your tracker. An alert tells you something changed; the tracker records that a named person looked, what they saw and what they did. When an alert arrives, treat it as a red row: do the lookup, save the evidence and update the sheet the same day.
A calendar entry per renewal
For each nurse, create a calendar event 30 days before expiry with the staff ID in the title and the tracker link in the notes. It is manual, but it works and needs no tools.
A weekly digest email
Use your spreadsheet tool's built-in notification rules, or a simple automation, to email the clinical lead and registered manager every Monday with the rows that are red or amber. Keep the email to staff IDs and dates, not full names or registration details, so a misdirected message reveals little.
AI-drafted reminders to the nurse
A nurse is responsible for renewing their own registration, and a polite nudge helps. Ask the AI to draft the message once; then send it from your own system. Example (illustrative):
Hello [first name], our records show your registration is due for renewal on [date]. Please send us confirmation when it is renewed, or let us know if anything is delayed so we can plan your shifts. Thank you, [your name]
Check that the draft does not say or imply anything about the nurse's competence or standing, and keep it neutral. Send reminders at 60 and 30 days, and a firmer, still courteous, one at 14 days that explains what happens to shifts if registration cannot be confirmed.
Step 5: what to do when something is wrong
Write the escalation steps before you need them. A short version (illustrative; get it reviewed by your clinical lead and, where appropriate, your adviser):
- Status shows lapsed, suspended, restricted, conditions, or "not found": tell the registered manager and clinical lead straight away.
- Stop the nurse doing any duty that requires current registration until the status is cleared. If conditions apply, follow them exactly.
- Speak to the nurse. They may have renewed and the register may not have updated, or there may be a genuine problem. Ask for written confirmation and look again.
- Record what was found, when, by whom, what was decided and when it was resolved.
- Consider whether any care delivered during a lapse needs review, and ask your adviser or regulator what to report, if anything.
Here is a realistic case. A weekly check shows a bank nurse's status as "not found". The clinical lead rings the nurse, who says she renewed last week and sends her confirmation email. A second search a day later shows an active entry. The sheet records both lookups with dates, and a note: "Register updated one day after renewal; shifts paused for 24 hours". Nothing dramatic, but the trail shows exactly what was done. Without it, the same incident would be a gap that nobody could explain.
A filled-in extract, so you can see the shape
Here are three rows from an invented tracker, with dates chosen to show each colour. Staff IDs are placeholders, and the last-checked dates assume today is 4 November 2026.
| ID | Type | Status | Expiry | Last checked | Next due | Colour |
|---|---|---|---|---|---|---|
| N-02 | Employed | Active | 30 Nov 2026 | 10 Sep 2026 | 16 Nov 2026 | Amber, 12 days |
| N-04 | Bank | Active | 14 Mar 2027 | 1 Aug 2026 | 30 Oct 2026 | Red, overdue 5 days |
| N-09 | Agency | Active | 20 Jun 2027 | 28 Oct 2026 | 26 Jan 2027 | Green |
N-04 is the instructive row. The nurse's registration is fine until March, but the 90-day recheck rule has run out, so the row is red. Without a recheck rule independent of the expiry date, this nurse would have looked safe until March while nobody had looked at the register since August. A lapse, suspension or new condition can appear between renewal dates, which is exactly what the recheck interval is for.
Three small providers, three set-ups
A dental practice with one visiting nurse
A dental practice that uses a nurse for sedation support a few days a month needs only a few rows. The sheet can live beside the other staff compliance records, with a calendar reminder for each date. The weekly routine collapses to a monthly one, but the evidence file matters just as much.
A physiotherapy clinic that hires nurses for health checks
A clinic running occasional health screening days brings in nurses from a bank. The tracker needs the employment-type column and a pre-shift check box: "Register checked today: yes or no". A shift is not confirmed until the box is ticked, and the person who ticks it adds initials.
A recruitment agency placing nurses with care homes
An agency that supplies nurses holds many more rows and has clients who ask for proof. It needs a separate view per client, a rule for sending evidence on request, and likely a dedicated compliance system once the numbers grow past what a shared sheet handles comfortably. The warning signs: duplicate rows, edits that overwrite each other, and a weekly review that now takes more than an hour.
A common mistake: the expiry date that nobody confirmed
A provider copies an expiry date from the nurse's own certificate, not from the register, and never checks it again. The certificate is out of date because the nurse has since renewed to a different date, and the tracker shows red for a nurse who is perfectly fine. The reverse is worse: a certificate shows a future date, but the register shows a restriction. The fix is a rule that the register is the only source for status and dates, and that a nurse's paperwork is a prompt to look, not proof.
Formulas, reminders and summaries, but never the register check
Three jobs suit AI well: building and debugging the spreadsheet formulas, drafting neutral reminder and escalation messages, and summarising the tracker into a short report for the manager (staff IDs only). Here is a quick example of the third. You paste a version of the sheet with only IDs, status and dates, and ask:
Summarise this tracker for a registered manager in 120 words.
List: how many rows are red, amber, green; the staff IDs that are
red; any row where status is not "Active"; any row with a blank
"Last checked on" date. Do not infer anything about individuals.
The output (illustrative) might say: "10 nurses. 1 red (N-04, check overdue by 6 days), 2 amber (N-02, N-07, due within 14 days), 7 green. No non-active statuses. No blank last-checked dates." What to fix: the sheet says N-04 is 5 days overdue, not 6. The model assumed a different "today", which is a common slip because chat tools do not see your spreadsheet's clock. Put today's date in the prompt, and trust the sheet's own "Days until due" column over any summary. It is still a handy readout, and the manager still opens the sheet to act.
Three jobs do not suit AI. It cannot reliably tell you whether a named person is currently registered. It may state outdated rules about renewal or revalidation, and it may mix up regulators. And it should not hold the master record: personal data in a general chat tool is exposure you do not need. The habits that keep staff details out of free tools are covered in how to stop staff pasting personal data into free AI tools.
Data protection for the tracker itself
The tracker holds personal data about staff, so treat it that way. Limit access to the registered manager, clinical lead and whoever does the checks. Keep it in your business storage with sign-in protection, not in someone's personal drive. Share only what recipients need: alert emails with IDs and dates, not registration numbers. Decide how long you keep records of past nurses, following your retention policy, and record the reason. If you use a spreadsheet's built-in AI features, check what the vendor does with the content on your plan before turning them on.
Agency and bank nurses: the extra check
If a nurse arrives through an agency, the agency is normally responsible for its own checks, but you remain responsible for the care in your setting. Add two habits. First, ask the agency for written confirmation of the nurse's registration, training and any restrictions before the first shift. Second, run your own register lookup, record it and file it, even if the agency's paperwork looks complete. A shift filled in a hurry is exactly when a shortcut is tempting, so make the lookup part of the shift confirmation step, not an optional extra.
The same expiry-and-recheck logic works for other renewals a small provider tracks. If you also manage payer or insurer credentialing, tracking credentialing renewals with AI applies it to applications and re-credentialing dates.
A quarterly spot-check of three random nurse rows
Once a quarter, run a short test. Choose three rows at random. For each, repeat the register lookup, compare it with the sheet and check the evidence file opens and matches. Then check that the alerts fired correctly for the last three renewals, and that nobody was working past an expiry date. Record the results and any fixes.
Realistic findings from a first test (illustrative): one evidence link points to a deleted file; one expiry date was copied from an old certificate; the weekly email goes to a manager who has left. None is serious, and all are the kind of thing that quietly breaks a manual process. Fixing them keeps the tracker something you can show an inspector with confidence.
It is useful to include the tracker in your regular audit round; a monthly care home audit checklist shows how to log items like this and action them. And the onboarding side, where the first check happens, is in the caregiver onboarding checklist. If you want help setting up the whole routine, my AI implementation consultation covers exactly this kind of job.
Questions about tracking nurse registration
Can AI check the register for me?
It should not be your source of truth. A person should look up each nurse on the official register or board lookup and record the result with the date. AI can help build the tracker, draft reminders and summarise the sheet, but it can be out of date or wrong about an individual.
How often should I recheck a nurse?
At hiring, at each renewal date and on a regular schedule you set in policy, such as every three months for agency or bank staff. Also check whenever you hear of a concern. Your regulator or insurer may state a minimum, so confirm it.
Is a spreadsheet enough?
For a team of up to a few dozen nurses, a well-built shared sheet with clear owners and an audit trail is enough, especially alongside a free register alert service where your boards take part. Move to dedicated compliance software if you struggle with volume, multiple sites or several registers at once.
What if a nurse's registration has lapsed?
Follow your written escalation steps: remove them from clinical duties that need registration, tell the manager and record what happened. Do not rely on a verbal assurance. Ask for evidence of the renewed registration and recheck the register before they return to those duties.
Further reads
- AI Tools for Nurses Who Run a Team: Rotas, Handover and Audits — Other jobs AI can take off a nurse manager's desk.
- Home Care Agency Policies and Procedures: A Review-First AI Method — Write the escalation steps into a policy.
- Home Care Agency Employee Handbook: Draft With AI, Then Review — Make staff duties on registration clear in writing.
- AI for Care Homes: A Registered Manager's 90-Day Implementation Plan — See where compliance tracking sits in a wider AI plan.
- AI in Healthcare Staffing: What a Small Agency Can Automate — Agency-side view of nurse compliance paperwork.
- Caregiver Retention: Plan Check-Ins and Exit Interviews With AI — Keep the nurses you have just checked.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: NMC pages Search the register, How to search the register (results fields and status glossary), Employer confirmations and Employers' responsibilities; the Nursys home page on QuickConfirm and e-Notify. All checked November 2026.