After Go-Live
Your surgery is live — here’s how to get the most from the toolkit and keep it running well.
Test Before You Announce
Before telling your full reception team, spend 10–15 minutes testing with one or two people who’ll be using it daily.
Pick 5–10 symptoms your team encounters regularly — for example:
- Sore throat
- UTI in an adult
- Chest pain
- Rash in a child
- Back pain
For each one, check:
- Do the instructions make sense to a receptionist, not just a clinician?
- Do the appointment options match what’s actually available in your appointment book?
- Do highlighted phrases draw attention to the right things?
- Do the high-risk buttons appear for appropriate symptoms?
Fix anything that looks wrong now. It’s much quicker before the whole team is relying on it.
Introducing the Tool to Your Team
Most practices adopt the toolkit fully within 24 hours of going live. A brief 10-minute intro session helps — you don’t need a formal training programme.
What to cover:
- How to find a symptom (search bar and browse)
- What the highlighted text means
- How to use the high-risk buttons
- Who to contact if an instruction doesn’t look right
What not to do: don’t try to walk through every symptom in advance. The tool is designed to be used in the moment — staff learn it fastest by using it on real calls.
Tip: Pin the toolkit URL or add it to your browser bookmarks bar before the intro session so staff can access it immediately.
The Clinical Review Cycle
Clinical review is your ongoing governance loop. It ensures that every instruction your team sees has been checked by a clinician.
When to review:
- After running AI customisation (every AI-edited instruction is set to Pending until approved)
- When a base symptom is updated centrally (you’ll see it appear as Pending)
- When you or a colleague edits an instruction manually
- When a reception team member flags that an instruction doesn’t match your workflow
How it works:
- Open Clinical Review from the admin dashboard
- Work through pending symptoms — check the instruction, check the routing, approve or request changes
- Any instruction you edit resets to Pending — the final version is always what gets approved
There is no shortcut here. The clinical review step exists because these instructions inform real triage decisions. Keep the pending count low — fewer than 10 is a healthy target.
Who should do clinical reviews? Ideally a GP partner or clinical lead. In practice, experienced nurses or ANPs can review instructions within their scope. Whoever does it should understand your local workflows well enough to spot when routing is wrong.
Keeping Instructions Up to Date
Your symptom library needs occasional maintenance — not constant attention, but a light touch when things change.
Review your instructions when:
- Your appointment model changes (new clinicians, new appointment types, retired services)
- A local pathway changes (e.g. a new MSK referral route, a change to your duty doctor model)
- A reception team member tells you an instruction doesn’t match reality
- You notice a symptom is being used differently than the instruction suggests
Re-running AI customisation:
If you’ve made significant changes to your practice profile or appointment model, you can re-run AI customisation at any time. The new run uses your updated profile and overwrites previous drafts — they’ll appear as Pending for clinical review before going live.
You don’t need to re-run customisation for minor wording tweaks — edit instructions directly in the symptom editor instead.
Monitoring Usage
If analytics are enabled for your surgery, the admin dashboard shows:
- Active users — how many team members are using the tool regularly
- Most-viewed symptoms — which conditions your team looks up most often
- Symptom views over time — whether usage is growing or dropping
What to look for:
Low usage after go-live usually means one of three things — staff haven’t been shown the tool, they can’t find it easily, or they don’t trust it yet. A follow-up conversation with your reception lead usually surfaces the real reason quickly.
High usage of a small number of symptoms with many others untouched is normal early on — staff default to what they know. Encourage broader use by highlighting the tool during team meetings when relevant symptoms come up.
AI Suggested Questions
The AI Suggested Questions panel generates triage-style questions to help staff gather consistent information from patients.
These appear alongside symptom instructions and are generated based on the symptom and your practice profile. Examples:
- “When did the symptoms start?”
- “Any red-flag features such as difficulty breathing or chest pain?”
- “Is this for a child or an adult?”
Staff don’t need to ask every question — the panel is a prompt, not a script. It’s particularly useful for less experienced team members who are still building confidence in triage conversations.
Practice Handbook
The Practice Handbook (Admin Toolkit) is a space for surgery-specific guidance that doesn’t fit neatly into the symptom library.
Good uses for the handbook:
- Local contact numbers (out-of-hours, community nursing, mental health crisis line)
- Internal procedures (how to handle a patient who refuses to give symptoms, chaperone policy)
- Frequently asked questions from reception staff
- Induction materials for new team members
The handbook is visible to all logged-in users at your surgery. Keep it current — an outdated handbook is worse than no handbook.
Getting Help
If something isn’t working as expected, or an instruction looks clinically unsafe, contact us directly:
For clinical safety concerns, do not wait — contact us immediately and remove the affected instruction from live use via the clinical review workflow while we investigate.
Last updated: March 2026