Key takeaways
- Combine reviews for patients with several conditions into one prepared visit, with bloods and observations done first.
- Run a fixed invite sequence: bloods link, review link, reminder, second invitation and a call for higher-risk patients.
- QOF achievement is measured on 31 March, so spread recalls across the year rather than leaving a winter backlog.
- Match each part of the review to the right role: HCA for tests, nurse or pharmacist for most reviews, GP only where needed.
- Record non-response properly after the invitations, and keep trying differently for patients who never respond.
Why recall design matters more each year
The number of patients needing regular reviews keeps growing. The population aged 65 and over rose in every UK nation in the year to mid-2025, including by 2.0% in England[1], and the UK now has 285 people of pensionable age for every 1,000 people of working age[2]. More older patients means more people with two, three or four long-term conditions, each with its own review.
If each condition gets its own recall, the same patient may be invited five times a year, have bloods taken twice and see three different clinicians who each ask the same questions. That wastes appointments and the patient's patience. Good recall design treats the patient, not the condition, as the unit of booking.
The QOF calendar shapes the work
For practices in England, the Quality and Outcomes Framework sets much of the review timetable. In 2026/27 the clinical domain is worth 437 points and the public health domain 145[3]. Achievement is measured on the last day of the financial year, 31 March, for patients registered on that date, and where an indicator refers to activity 'in the preceding 12 months', the activity must fall in the 12 months ending on that 31 March[3].
Source: North and South Essex Local Medical Committees, 2026[3]
Show the data
| Item | Value |
|---|---|
| Clinical domain | 437 pts |
| Public health domain | 145 pts |
That deadline creates a familiar trap: reviews drift through the summer and autumn, then pile into January to March just as winter pressure peaks. Spreading recalls evenly across the year, with a monthly target, protects both the clinics and the patients. Our guide to birthday-month recall compares ways of spreading the load.
Combine reviews into one prepared visit
| Bundle | Typical conditions | Before the visit | Who and how long |
|---|---|---|---|
| Cardiometabolic | Diabetes, hypertension, CKD, CHD, raised cholesterol | Bloods and urine, blood pressure, weight | Practice nurse, 30 minutes |
| Respiratory | Asthma, COPD | Questionnaire sent by text, inhaler list | Nurse or respiratory-trained pharmacist, 20 to 30 minutes |
| Mental health | Serious mental illness | Physical health checks booked first | Nurse plus GP or mental health practitioner, 30 minutes |
| Frailty and medicines | Multiple conditions, 10 or more medicines | Medication list check, bloods | Clinical pharmacist, 30 minutes, GP if needed |
| Single condition | One condition only | Tests as required | Most appropriate clinician, 15 to 20 minutes |
Build each bundle as a slot type with its own length, so invite links offer the right appointment automatically. A patient with diabetes and hypertension should receive one invitation for one 30-minute cardiometabolic review, not two invitations for two 15-minute slots.
An invite sequence patients act on
Week 0: tests first
Send a self-book link for bloods and observations with the HCA, two to three weeks before the review window.
Week 1: review link
Once bloods are booked, send the review link for a slot after the results will be back.
Week 2: reminder
Remind anyone who has not booked either appointment, with a phone option.
Week 4: second invitation
Send a second invitation by a different route, such as the NHS App if the first was SMS, or a letter for patients without a mobile.
Week 6: phone call
Reception or the recall coordinator calls higher-risk patients who still have not responded, such as those with poor diabetes control.
Record the outcome
If invitations have gone unanswered, record it in the way the QOF business rules require; patient non-response to invitations is one of the recognised personalised care adjustment reasons[3].
Check the current QOF guidance and business rules for the exact recording requirements before relying on non-response coding. Treat the adjustment as a record of what happened, not as a substitute for trying again.
Set up the searches once, then reuse them
Most of the effort in a good recall system is front-loaded. Once the searches, slot types and templates are built, each month's recall becomes a routine task rather than a project. Spend a few sessions getting these right with your clinical system lead.
- One search per review bundle that finds patients due in the coming month, excluding those already booked, recently reviewed or with a recorded reason not to invite.
- A flag for patients with conditions in more than one bundle, so they get one combined invitation, routed to the longest bundle.
- Slot types for each bundle and for pre-review tests, with the right length and clinician, bookable by invite link.
- Message templates for each step of the sequence, stored in the messaging tool with clear names.
- A monthly report showing invited, booked, completed and not responded, by bundle.
Name one person as recall coordinator, with protected time each week. Recall slips fastest when it belongs to everyone and therefore to no one.
Messages that get bookings
- Tests first: 'Hawthorn Medical Centre: it is time for your yearly health check. First, please book a 10-minute blood test and blood pressure check here: [link].'
- Review: 'Hawthorn Medical Centre: thank you for booking your blood test. Now choose a time for your 30-minute review with the practice nurse, after your results are back: [link].'
- Second invitation: 'Hawthorn Medical Centre: we have not been able to arrange your yearly review. It helps us keep your treatment safe. Please book here or call us: [link].'
- Keep condition names out of texts; 'yearly health check' or 'annual review' is enough.
Patients are increasingly comfortable acting on digital messages: 30.8% last contacted their practice online in 2026[4]. But keep a letter and phone route for patients who do not use text or apps.
Use the whole team
Most review work does not need a GP. In general practice, 44.4% of appointments are now delivered by GPs and the rest by other practice staff[5]. Design the review so that each part goes to the right role: HCAs for bloods, blood pressure, weight and foot checks; practice nurses for most diabetes, respiratory and cardiovascular reviews; clinical pharmacists for medication reviews and polypharmacy; and GPs for complex decisions and patients whose care plans need changing.
Ring-fence review sessions in the template, especially from October to March, so they are not swallowed by same-day demand during winter. A protected weekly session for each review bundle is easier to defend than ad hoc slots.
Write a one-page protocol for each bundle that says which tests are needed, who does each part and when a patient must be passed to a GP. The protocol lets a newly recruited nurse or pharmacist run the review safely from their first week, and it gives reception confidence about which slot to book.
Patients who never respond
- Look for a pattern: age, language, housing, work hours. Non-responders are often the patients who need the review most.
- Offer an evening or Saturday review through the PCN, or a combined visit when they next attend for anything else.
- Flag the record so that any clinician who sees them opportunistically can complete parts of the review.
- For patients with a learning disability or serious mental illness, ask the key worker or carer to help arrange the visit.
- Ask once, kindly, whether there is a reason; some patients are frightened of results or have had a bad experience.
Example: a month of recalls
Example (fictional): a 14,000-patient practice targets about 180 combined reviews a month. In April it sends bloods links to 220 patients due in May. Within two weeks, 150 have booked bloods and 120 have booked their review. Reminders and second invitations bring the totals to 175 and 160. The recall coordinator calls 30 higher-risk non-responders and books 14. By the end of May, 174 reviews are done, the nurse sessions run at 92% utilisation and no patient has had to attend twice for separate condition reviews.
What to measure
- Reviews completed per month against the monthly target.
- Share of patients with several conditions seen in a single combined review.
- Response rate at each step of the invite sequence.
- Share of reviews where bloods were back before the appointment.
- DNAs for review appointments, by bundle.
- Patients with no review for more than 15 months.
To fill review gaps when patients cancel, use a short-notice cancellation list. For the year-end push, see QOF year end, and for dental recall intervals, see applying NICE guidance on dental recall.
Frequently asked questions
How do GP practices invite patients for annual reviews?
Usually by text or NHS App message with a self-book link, often with a blood test first and the review afterwards, followed by reminders, a second invitation and sometimes a phone call.
Can several long-term condition reviews be done in one appointment?
Yes. Many practices combine reviews for related conditions, such as diabetes, hypertension and kidney disease, into one longer appointment with tests done beforehand.
What happens if a patient ignores review invitations?
The practice should try different routes and times. If invitations go unanswered, the practice records this as QOF rules allow, but should keep trying for higher-risk patients.
References
Office for National Statistics (2025). Population estimates for the UK, England, Wales, Scotland and Northern Ireland: mid-2025 (opens in a new tab) ↑
Office for National Statistics (2025). Population estimates for the UK, England, Wales, Scotland and Northern Ireland: mid-2025 (opens in a new tab) ↑
North and South Essex Local Medical Committees (2026). Quality and Outcomes Framework (QOF) guidance for 2026/27: LMC summary (opens in a new tab) ↑
Ipsos (GP Patient Survey provider for NHS England) (2026). 2026 GP Patient Survey results released (opens in a new tab) ↑
British Medical Association (2026). Pressures in general practice data analysis (opens in a new tab) ↑