Key takeaways
- Size triage from counted requests across every channel, not from last year's rota or appointment book.
- Online request volumes have climbed sharply since core-hours access became a contract requirement, so older baselines understate demand.
- Plan clinician contacts against a safe daily cap and agree the overflow routes before the session starts.
- Ring-fence time for continuity so triage does not quietly absorb every routine slot.
Why triage capacity needs its own plan
In a total triage or modern access model, triage is no longer a side task for whoever is free. It is the front of the clinical day, and every other session depends on it finishing on time. Since 1 October 2025, practices in England have had to keep online consultation tools open throughout core hours, 8am to 6.30pm, for routine and admin requests[1]. That change moved a large share of demand into a queue that never closes during the working day.
The volumes are not small. The Care Quality Commission reports that practices received 9,617,628 online requests in March 2026, 69% more than in August 2025 before the requirement began[2]. That works out at 153 online requests per 1,000 patients in a single month[2]. A triage rota built on figures from two winters ago will run late, and the cost shows up as triagers finishing after 7pm, routine allocations slipping into the next week and patients ringing to chase.
Source: Care Quality Commission, 2026[2]
Show the data
| Item | Value |
|---|---|
| Aug 2025 | 5,690,082 |
| Mar 2026 | 9,617,628 |
The fix is not simply more triage sessions. It is a plan that links counted demand, realistic triage minutes, a safe contact cap and named overflow routes, then gets reviewed every month.
Count demand across every channel first
Run a two-week count before you change anything. Pick a fortnight without a bank holiday, then repeat the count around one later, because NHS England notes that call volumes are affected by the number of days in a month, public holidays and weekend days[3]. The same pattern shows up on practice lines and online forms, especially on the first working day after a closure.
Count requests, not appointments. Appointments only show the demand you managed to meet. A request is any contact that needs a clinical or admin decision, whatever door it came through.
- Online consultation and admin forms, including those started in the NHS App.
- Phone calls that needed clinical review rather than a simple answer.
- Walk-in requests taken at the front desk.
- NHS 111 direct bookings and community pharmacy referrals back to the practice.
- Internal clinical tasks that compete for triage time, such as abnormal results and discharge letters that need a decision.
For each request, record the time it arrived, the channel, whether it was clinical or admin, how long the triage decision took and where it ended up: self-care, pharmacy, another role in the team, a same-day clinician contact or a routine booking. Most online consultation systems export this; the phone and walk-in figures usually need a simple tally sheet for two weeks.
Turn requests into clinician hours
Forecast requests by weekday
Average your count for each weekday and keep Monday separate. Use the measured Monday figure, not a rule of thumb, and add a post-bank-holiday forecast from your second count.
Time the triage decision
Time a sample of fifty triage decisions across two clinicians. Split simple requests, such as a medication query, from complex ones, such as a new symptom in a frail patient. Use the average for each type, not the fastest triager's pace.
Apply your outcome mix
From the count, work out what share of requests needed a same-day clinician contact, what share went to other roles and what share became routine bookings. This is the number that sizes the urgent sessions behind triage.
Convert contacts into sessions
Divide same-day clinician contacts by a safe number of contacts per clinician per day. The result is the minimum number of clinicians you need on the urgent side, before leave and training.
Add a buffer you can defend
Hold back capacity for home visits, safeguarding and the patient who needs forty minutes. Write the size of the buffer into the rota so it is not spent by 10am.
| Line in the plan | Figure | How it was worked out |
|---|---|---|
| Online requests per Tuesday | 90 | Two-week count, average of two Tuesdays |
| Phone and walk-in clinical requests | 55 | Tally sheets at reception |
| Total requests to triage | 145 | Sum of the two lines above |
| Triage time | About 7 hours | 145 requests at an average of 3 minutes |
| Needing a same-day clinician contact | 55 contacts | Measured outcome mix applied to 145 |
| Urgent-side clinicians needed | 3 | 55 contacts against a cap of 25 each, rounded up |
| Routed to pharmacy or other roles | 40 | Measured outcome mix |
In this Example, triage alone needs one clinician for most of the day. If that clinician is also expected to see the same-day patients, the plan has already failed before the phones open.
Set a safe cap and plan the overflow
The BMA and the European Union of General Practitioners recommend not more than 25 contacts per day per GP for safe care[4]. The BMA guidance suggests an action plan to reach safe numbers, using care coordinators and suitably trained reception staff to direct patients to other services under agreed protocols, and notes that NHS 111 can book up to one same-day slot per 3,000 patients for practice triage[4].
A cap only works if everyone knows what happens when it is reached. Agree the overflow order in a clinical meeting, write it down and put it on the triage screen.
- Community pharmacy for conditions covered by Pharmacy First and minor illness.
- Other roles in the practice or PCN, such as a clinical pharmacist, first contact physiotherapist or paramedic.
- The PCN same-day hub, if your network runs one.
- A planned clinician call the next working day for requests the triager judges safe to wait.
- For the practice's closed hours, clear advice on NHS 111 and 999.
Set one fixed checkpoint, for example 11am, when the triage lead compares requests so far with the forecast. If demand is running above plan, the overflow routes open then, not at 4pm when the urgent list is already full.
Protect continuity inside the triage model
Triage that sends every request to the next free clinician is efficient for the queue but costly for patients who need to see the same person. Allocation decisions also shape who patients meet: 44.4% of standard appointments in August 2026 were delivered by a GP, with the rest delivered by other practice staff[1].
Build continuity into the triage rules rather than leaving it to individual triagers. Flag patients who benefit most, such as those with several long-term conditions, frailty, complex mental health needs or care at the end of life. For flagged patients, the default is the named clinician's next available slot unless the triager judges the request urgent.
Example rule: if the patient's usual GP is working within three working days and the triager judges it safe to wait, offer that GP first and record the reason if you do not.
Then protect those slots. Each clinician keeps a small number of routine slots a week that triage cannot fill with new patients. Without that ring-fence, continuity disappears in the first busy fortnight.
Review the plan every month
NHS England's access plan expects patients to know on the day how their request will be handled, while their preference for a call, visit or online message is respected[5]. A monthly review checks that the capacity plan is delivering that promise, not just filling the rota.
- Requests per thousand patients per week, by channel.
- Share of requests given an outcome on the day they arrived.
- Time the last triage decision was made each day.
- Same-day contacts per clinician against the agreed cap.
- How often each overflow route was used, and whether it was suitable.
- Patients who contacted the practice again within 48 hours about the same problem.
If the last triage decision regularly lands after 4pm, the triage session is undersized or the triager is being pulled into other work. If re-contacts rise, look at the outcome messages patients receive, not only at capacity.
Common planning mistakes
- Sizing triage from the appointment book, which hides unmet demand.
- Leaving phone and walk-in requests out of the count because they are harder to measure.
- Asking one clinician to triage and run the urgent list with no cap on either.
- Planning a normal Monday and treating the day after a bank holiday as a surprise.
- Counting only GP capacity and forgetting the other roles that can take a share of requests.
- Changing the model and the rota at the same time, so nobody can tell which change helped.
Frequently asked questions
How many clinicians should triage each day?
Work it out from your own count: requests per day multiplied by average triage minutes gives triage hours. Most practices find triage is at least one clinician's main task on busy weekdays, separate from the urgent list.
Should the triager also see the same-day patients?
Only in small practices with low request numbers, and then with a cap on both tasks. In most practices combining the two roles is why triage overruns.
Does a contact cap mean turning patients away?
No. A cap means the extra demand goes to routes agreed in advance, such as pharmacy, other roles, a PCN hub or a planned next-day call, rather than being squeezed into an unsafe list.
How often should we recount demand?
Do a full two-week count once a year and around winter, and check the weekly request figures every month. Recount after any change to your online form or phone menu.
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) ↑
British Medical Association (2026). Safe working in general practice in England: daily working contacts (opens in a new tab) ↑
NHS England and DHSC (2023). Delivery plan for recovering access to primary care (opens in a new tab) ↑