Key takeaways
- Most surges are predictable: Mondays, the day after a bank holiday, the start of winter and school terms. Plan rotas for them in advance.
- Set trigger levels based on the number of unreviewed requests, with a named action for each level, so nobody has to improvise.
- Reply templates for the commonest outcomes cut handling time without making replies feel automated.
- National online submissions peaked at 9.9 million a month in March and June 2026[1]; local peaks will be sharper.
Why surges catch practices out
Online demand does not arrive evenly. Requests bunch on Monday mornings, after bank holidays, during school holidays when parents have time to submit, and in winter. Nationally, online consultation submissions peaked at 9.9 million a month in March and June 2026 before dipping to 9.5 million in July[1]. Within a single practice the swings between a quiet Thursday and a post-holiday Tuesday are far larger.
A queue of unreviewed requests is a safety risk as well as a workload problem. A worrying symptom in request number 140 is not seen until requests one to 139 have been read. A surge plan exists to keep that queue short and to make sure urgent requests surface first.
Source: British Medical Association, 2026[1]
Show the data
| Item | Value |
|---|---|
| March 2026 | 9.9 million |
| June 2026 | 9.9 million |
| July 2026 | 9.5 million |
Know your own curve
Before changing rotas, pull at least eight weeks of request data by hour and day. Most tools report this. Look for:
- The busiest hour of the week, usually between 8am and 10am on Monday.
- How demand after a bank holiday compares with a normal Monday.
- The share of requests that are admin rather than clinical.
- Average and maximum time from submission to first review on each day.
- The number of requests still unreviewed at noon and at 4pm.
The national picture helps set expectations. In March 2026 practices received 153 online requests per 1,000 patients over the month[2]. For a practice of 10,000 patients that is around 1,500 requests in a month, before counting phone and walk-in contacts. NHS England notes that call volumes are affected by the number of days in the month, public holidays and weekend days[3], and online demand behaves the same way.
Build the triage rota around the curve
| Time | Navigators reviewing online | Clinicians on triage | Notes |
|---|---|---|---|
| 8.00 to 9.00 | 2 | 1 GP, 1 advanced practitioner | Overnight and early requests cleared first |
| 9.00 to 10.30 | 2 | 2 GPs | Peak submissions; urgent requests surfaced |
| 10.30 to 12.30 | 1 | 1 GP | Steady flow; admin queue handled separately |
| 12.30 to 14.00 | 1 | Duty GP | Lunch cover; queue kept below trigger level |
| 14.00 to 16.00 | 1 | 1 GP, 1 pharmacist | Second peak after school run |
| 16.00 to 18.30 | 1 | Duty GP | Final sweep so no request is left overnight |
Short triage blocks of one to two hours, alternating with consultations, are easier to sustain than whole sessions. Put the strongest cover where the data show the queue grows fastest.
Set trigger levels
Trigger levels turn a vague sense of being overwhelmed into a clear decision. Base them on the number of unreviewed requests, or on the age of the oldest unreviewed request, whichever your tool reports more reliably.
Green, normal
Queue below your normal level and oldest request under an hour. Rota as planned.
Amber, rising
Queue above normal or oldest request over two hours. A named second navigator joins online review; the duty GP checks the queue for red flags; admin requests are moved to the admin team's list.
Red, surge
Queue well above normal or oldest request over four hours. A second clinician joins triage for an agreed period; non-urgent clinical admin is paused; the acknowledgement message is updated with a realistic response time.
Black, exceptional
Staff shortage or system failure. The practice manager and duty partner agree actions, which may include asking the primary care network for support and updating the phone message.
Write the triggers on one page and put them where the navigators sit. Name who makes each decision so it happens without a meeting.
Reply templates that save time
Templates speed up the commonest outcomes and keep wording consistent. Keep them short, personal and editable.
- Booked appointment: "Thank you for your request. We have booked you a telephone appointment with Dr Shah on Wednesday between 10am and 12pm. Please keep your phone with you."
- Pharmacy First: "Thank you for your request. A pharmacist can assess and treat this through the NHS Pharmacy First service, without a GP appointment. Here is a list of local pharmacies. If you do not improve or get worse, contact us again."
- Self-referral: "You can refer yourself directly to the local physiotherapy service using this link. If symptoms change, contact us again."
- More information needed: "To help us decide the best next step, please reply with a photo of the rash and tell us how long you have had it."
- Admin completed: "Your fit note has been issued and is available in the NHS App."
- Redirect for urgent symptoms: "Your request describes symptoms that need urgent assessment. Please call the surgery now on 01632 960789. If you feel very unwell, call 999."
Review templates with clinicians every quarter. A template that patients misunderstand creates a second request.
Use the whole team
Surges are easier when work spreads beyond GPs. Route medication queries to the practice pharmacist, simple musculoskeletal problems to first-contact physiotherapists if your network has them, and eligible minor illnesses to community pharmacy. Community pharmacies had delivered 3.05 million Pharmacy First consultations by the end of February 2026[4], so the route is well established. Admin requests should never wait in the clinical queue.
Protect the people doing the review. Workload pressure after the online access changes is widely reported: in a BMA survey, 68% of practices reported more staff stress[5]. Rotate triage duties, keep breaks and debrief after a red day.
Keep phone and online demand in one queue
A surge in online requests often coincides with a busy phone line. If the two are handled in separate lists, staff tend to favour whichever is in front of them, and patients on the other route wait longer. During a surge, that imbalance gets worse.
Have navigators enter phone requests into the same tool, using the same questions, so every request is prioritised by the same rules. Watch both numbers on the same screen: calls waiting and online requests unreviewed. When one rises, move staff between them rather than letting one side fall behind. Patients who cannot go online should never be disadvantaged because the online queue looks more urgent on a screen.
Tell patients what to expect on busy days
Clear communication reduces repeat requests, which are a hidden part of every surge. Patients who hear nothing for hours submit again, or ring to check, and each duplicate adds to the queue.
- Update the acknowledgement message on red days with a realistic response time, and change it back when the queue clears.
- Put a short notice on the website and in the phone message on predictable peak days, such as the first day after a bank holiday.
- Ask patients not to submit the same request twice, and make it easy to add information to an existing request instead.
- Tell patients how they will hear back: text, call or app message, and roughly when.
Track duplicate requests during surges. A high rate is a sign that messaging, not capacity, is the problem.
Plan for the predictable peaks
- Mondays: strongest triage cover of the week, with pre-bookable capacity released for the days that follow.
- After bank holidays: treat the first working day as a double Monday. Christmas 2026 creates a four-day closure for many practices in England and Wales[6], so the first day back needs extra cover.
- Flu season: the adult flu programme in England starts on 1 October, with most eligible groups to be vaccinated by the end of November[7][8]. Questions about vaccination add admin requests; a template and a website page reduce them.
- School term starts and holidays: more requests about children; make sure the paediatric red flag guidance is visible to navigators.
- Supplier or telephony outages: have a paper or offline process ready, and a phone message to explain.
Example: a red day handled to plan
Example, fictional: on the Tuesday after a bank holiday, a practice's queue passed its red trigger by 9.40am. The navigator lead moved a second navigator to online review, the duty GP scanned the queue for red flags and found two requests needing same-day calls, and the pharmacist took medication queries. The acknowledgement message was changed to promise a reply within two working days for routine requests. By 3pm the queue was back to amber and by 6pm every request had been reviewed. At the debrief, the team agreed to add a second clinician to the triage rota for every first day after a bank holiday.
After the surge
Every surge leaves data. Review it within a week while memories are fresh.
- What triggered the surge, and could it have been predicted?
- How long did requests wait at the worst point?
- Were any urgent requests found late?
- Which actions helped most, and which were not used?
- What should change in the rota or triggers for next time?
Record the answers in a short log. Over a year, the log becomes the most useful planning document the practice has.
Frequently asked questions
What should we do when the online queue gets too long?
Use pre-agreed trigger levels: add navigator and clinician review time, move admin requests out of the clinical queue, check the queue for red flags and update the acknowledgement with a realistic response time.
Can we pause the online form during a surge?
Not during core hours, under the current GP contract requirement. Manage the surge through capacity, routing and clear messaging.
How do we spot urgent requests in a long queue?
Use the tool's red flag questions and filters, and have a clinician scan the unreviewed queue at set times on busy days.
Which days are usually busiest for online requests?
Monday mornings and the first working day after a bank holiday, with seasonal rises in winter. Check your own data, because local patterns vary.
References
British Medical Association (2026). Pressures in general practice data analysis (opens in a new tab) ↑
Care Quality Commission (2026). State of Care 2025/26: capacity in primary care (opens in a new tab) ↑
NHS England (2026). Statistical Note: Integrated Urgent Care Aggregate Data Collection (IUC ADC) July 2026 (opens in a new tab) ↑
NHS England (via GOV.UK) (2026). NHS England annual report and accounts 2025 to 2026 (opens in a new tab) ↑
British Medical Association (2026). Pressures in general practice data analysis (opens in a new tab) ↑
GOV.UK (2026). UK bank holidays (opens in a new tab) ↑
GOV.UK (UKHSA, DHSC and NHS England) (2026). National flu immunisation programme 2026 to 2027 letter (opens in a new tab) ↑
GOV.UK (UKHSA, DHSC and NHS England) (2026). National flu immunisation programme 2026 to 2027 letter (opens in a new tab) ↑