Independent information. England-wide.Public data. Clear sources.

REUSABLE EVIDENCE

Understand the evidence behind a care shortlist

Explore reporting providers, inspect source definitions and download selected measures with their dates, populations and reuse information.

Download evidence from a comparison

Use the CSV download on a comparison to retain numerical observations, explicit missing states, source dates, denominators, confidence intervals where published, and source-specific licence information. The file includes the release identifier for citation.

A practice-group median is calculated from available practice values without weighting. It differs from a national patient-weighted estimate. Similar list size or network membership does not adjust for case mix or prove similar needs.

Build a provider shortlist

Public provider API

The website uses the same read-only JSON endpoint documented here. No account or API key is required. The endpoint accepts up to three stable provider IDs from a provider’s comparison selection.

GET /api/providers/?ids=PROVIDER_ID,SECOND_PROVIDER_ID

The response contains a providers array. Each available active record includes identity, canonical category and slug, official contact route, inspection evidence, offers, numerical metrics and source provenance. Unknown or inactive identities are omitted. An empty selection returns an empty array.

GET /api/export/?ids=PROVIDER_ID,SECOND_PROVIDER_ID

The export endpoint returns a UTF-8 CSV of numerical observations. It requires one to three valid IDs and returns HTTP 400 for invalid input, 404 when no active publishable record matches, or 503 if the evidence service is unavailable.

Value states

FieldMeaning
valueA published or explicitly derived number. Null is never converted to zero.
stateAvailable, missing, suppressed, not applicable or stale.
valueLabelPublisher-supported bounded text, where an exact value is unavailable.
period / grain / dimensionsThe time window, reporting entity or population, and source cohort.
denominator / lower / upperSource sample/base and 95% confidence limits where published. Missing fields have no implied value.

Keep requests bounded and cache only as permitted by each source. For national bulk research, use the original publisher datasets linked in the source register. These small-selection endpoints are intended for inspecting care options.

Correcting or verifying a record

Submit the page and evidence through the correction route. A provider claim records a verification request; it does not grant an automatic ability to edit public facts or alter independent comparisons.