About this tool

What this is. I'm Anne Marie Cunningham, a GP. I built this tool to show how English GP practices compare on access, continuity and capacity, using only data that NHS England and the GP Patient Survey already publish about every practice. It is written for practices working under pressure, to help them see where changes might be worth making. I make and run it in a personal capacity — it is not produced, endorsed or reviewed by NHS England or the NHS, and it is not connected to any organisation I work with.

What this is not. It is not a rating, a ranking or a league table, and nothing here is a judgement of the quality of care at any practice. The comparisons are descriptive associations across England's ~6,000 practices — they say what tends to go together, not what causes what, and not what any one practice should do. It is not medical, legal or business advice.

Parts of this tool use a language model: on the explorer page it turns your question into a database query and writes a plain-English interpretation of the result. Language models make mistakes — including confident-sounding ones. The generated query is always shown so anyone can check exactly what was asked of the data, and every figure can be verified against the published NHS sources named on each card. Please verify before acting on or publishing anything from this tool. If you spot an error, use the feedback boxes or contact me — I would rather know.

The underlying data has known limits, stated on each page: telephony covers only practices whose supplier submits nationally, appointment books do not capture total workload, some online-consultation counts are incomplete, and survey estimates for small practices are noisy.

Everything shown is practice-level information republished from open NHS England, GP Patient Survey and related official publications. There is no information about individual patients anywhere in this tool, and nothing is collected from or about patients. Contains public sector information licensed under the Open Government Licence v3.0; each card names the dataset it draws on.

Practices are organisations, but at the smallest practices published figures can in effect relate to identifiable professionals. This tool adds no evaluative claims about any individual or any named practice beyond what the official statistics state, and presents every figure with its context and caveats. If your practice is shown and you believe a figure is wrong or missing context, contact me and I will review it promptly.

If you quote or republish material from this tool, please keep the caveats attached, attribute the underlying sources, and do not present anything here as a rating or ranking of practices.

The practice page (mypractice.html). Lookups run entirely in your browser: the practice code or name you type is not sent to me or anyone else. If you choose to send feedback, your comment, your rating and the practice code being viewed are sent to my server so I can improve the tool. I delete feedback after 90 days.

The explorer (explore.html), shared demo. Your question is sent to my server (Cloudflare) and on to Anthropic's API to generate the query and — together with the generated SQL and a small sample of the public result rows — the written answer; Anthropic does not use API data to train its models by default. I log questions and answers so I can improve the tool, against a short code derived from your internet address that changes every day; the address itself is never stored. I delete question logs after 30 days and feedback after 90. The demo is rate-limited per visitor, which is why the daily code exists.

The explorer with your own key. Requests go directly from your browser to Anthropic (or to your own local model). Nothing is sent to or logged by me. Your key is stored only in your browser's local storage, on your machine.

No cookies, no analytics. These pages set no cookies and run no analytics or trackers. Your browser's local storage is used only to remember your model settings and key, because the page cannot work without them; clearing your browser data removes it.

Please don't type personal details. Free-text boxes are for comments about the data and the tool. Don't include anything that could identify a living person — yourself, a patient, or a colleague. If something looks like personal details, especially anything about a patient's health, I will delete it from the logs.

Who is responsible and your rights. I am the data controller for the feedback and demo logs (contact me via LinkedIn), and I am registered with the ICO as a data controller [add registration number Z------- when the confirmation arrives]. The lawful basis is my legitimate interest in improving a free tool. Because logs are held against a code that changes daily rather than your name or address, I usually cannot link a log entry to you — but you can contact me about anything you submitted, and you have the right to complain to the Information Commissioner's Office.

I provide the tool free, as is, in the hope it is useful — but without any warranty of accuracy, completeness or availability. Use it with your own judgement; verify figures against the published sources before relying on them. Don't use the tool to present any practice's data as a rating, ranking or endorsement, to attempt to identify individuals, or to overload the shared demo (it is rate-limited, and I may withdraw it at any time). These pages and terms may change; the date above shows the current version. Nothing here affects the terms of the underlying NHS publications, which remain available directly from their publishers.