This portal needs JavaScript to draw its charts and tables
The evidence itself does not. Everything on these pages is read from a
read-only JSON API on this same server, and every response carries the
source URL, retrieval time and payload hash of the figures in it:
/api/v1/summary — headline counts and their caveats/api/v1/pay — pay evidence/api/v1/council_spend — published council payment lines/api/v1/contracts — contract notices/api/v1/providers — one entry per provider/api/v1/authorities — commissioning authorities/api/v1/geography — evidence coverage per authority/api/v1/boundaries — authority boundary geometry/api/v1/fingertips — treatment demand indicators/api/v1/ndtms — NDTMS estimates with their confidence intervals/api/v1/compare — the existing series for two or more authorities or providers, on shared axes/api/v1/layers — the geography map's overlay layers, each with its own caveats/api/v1/pfd — coroners' Prevention of Future Deaths reports/api/v1/claims — what the campaign says, with the evidence behind each claim/api/v1/freshness — when each source table was last collected/api/v1/export — any of the above as CSV or JSON
/api/v1/export returns the same figures with that provenance
written into the file, because a CSV gets separated from any
accompanying note within a day of leaving the server. An export is
complete for its filters — it is not the window a page was showing.
The API is documented at /api: parameters,
response shapes and example URLs, on a page that needs no JavaScript
either.
Read the caveats before quoting any of it. They lead with
the things that must not be computed from this data — no
claims-per-employee rate, no dividing treatment numbers by workforce
figures, no differencing workforce census years.