About the GP Patient Survey

The GP Patient Survey is an independent survey run by Ipsos for NHS England. Every year it asks a random sample of patients at every GP practice in England about their experience. All the figures on this site come from its 2026 results.

Who is asked

The survey goes out every year in January. For each practice, Ipsos draws a random sample of patients from the NHS’s national Personal Demographics Service (PDS). Anyone aged 16 or over who has been registered with the practice for at least 6 months can be picked. You cannot volunteer: to keep the survey valid, Ipsos only invites people selected at random.

For 2026, 2,968,415 questionnaires were sent and 654,714 were completed between 1 January and 13 April 2026, a response rate of 22.1%. People could answer on paper or online, including online in British Sign Language and 14 other languages, and also by phone, in large print or in Braille. The results were published on 9 July 2026.

What it asks

The questions cover how easy it is to contact the practice, the care patients got from healthcare professionals, their experience of NHS services when the practice is closed, and their general health. There are also questions on NHS dental and pharmacy services. Most questions are about the last 12 months.

The questionnaire was first developed with the universities of Exeter and Cambridge and is reviewed every year; changes are tested with patients. It changed a lot in 2024, which started a new series, so most results can only be compared from 2024 onwards.

The survey measures patients’ experience. It does not measure clinical quality, such as how well a condition is treated.

Weighting

Some groups of patients answer more often than others. If a practice has many young patients but few of them reply, the replies would mostly reflect older patients’ views. Ipsos corrects for this by weighting: answers from groups that replied less than their share of the practice count for more, and answers from groups that replied more count for less. The weighting takes account of age, gender, the type of neighbourhood and the region, along with features of the local area such as deprivation.

All published results are weighted. Most outputs also give the number of people who actually answered each question (the unweighted base). That is the number of answers we show for each question.

Why every result has a margin of error

The survey hears from some of a practice’s patients, not all of them. So a practice’s figure could be a bit different if other patients had answered. Ipsos publishes a 95% confidence interval for each result: the range the true figure is estimated to lie in, meaning the result you would get if every patient had answered. Fewer answers make the range wider.

When the intervals of two practices overlap, Ipsos says the difference is not statistically significant and may be down to chance. When they do not overlap, the difference is significant. Results based on fewer than 10 people are not published at all.

Rankings

The survey’s FAQ says it is not designed to rank GP practices, and that Ipsos and NHS England do not advise using it for league tables or ranking lists. It gives three reasons: there is no single measure that ranks practices well, every result carries uncertainty, and practices serve different populations with different resources.

This site has no list of the lowest results. Every figure is shown with its interval and its number of answers, and a label is given only when the interval supports it. We do list the best-rated practices: in each area, and for England the 50 with the highest score among those clearly above the England average. Practices with the same score share a place, and each place has its interval and number of answers beside it. Two places are often within the margin of error of each other, and the lists say so. How the rating works sets out the rules.

The question we use

The survey part of our Practice score is one question: “Overall, how would you describe your experience of your GP practice?” (question 32): the share of patients who describe their practice as good, the measure NHS England uses. It counts for 70% of the score; the other 30% is the share of appointments within two weeks of booking, from NHS appointment data. It is the figure Ipsos reports first for each practice, and NHS England has used it as its indicator of patient experience of GP services in the NHS Outcomes Framework. We take it exactly as Ipsos publishes it, with its 95% interval, and rate a practice only when at least 30 people answered it.

Using that one question means we do not decide which parts of a visit matter most. Every practice page also shows twelve more questions, from getting through on the phone to seeing the same GP, each with its own interval and number of answers. They are not part of the rating.

If you are sent the survey

Taking part is voluntary. The invitation usually comes by post, and reminders can come by text, email, WhatsApp or the NHS App. The survey never asks for your name, address or bank details, and no one will come to your home or phone you to ask you to take part. Ipsos has a free helpline on 0800 819 9135 (8am to 9pm Monday to Friday, 10am to 5pm on Saturdays, as the survey’s FAQ gives them). More is on the survey’s own website.