Skip to content

About the data.

Healthcards shows verified local public-health statistics for all 153 English upper-tier local authorities. This page states what we cover, where numbers come from, and what we do not claim.

Coverage

  • All 153 English upper-tier local authorities (post-Apr-2023 boundaries): 63 unitary authorities, 36 metropolitan boroughs, 33 London boroughs and 21 counties. Names follow the official statistical areas — counties are labelled “<name> county”, the rest “<name> local authority”. Huddersfield is shown with Kirklees (E08000034) statistics and labelled as such on every page, because Huddersfield is a town, not a statistical area.
  • Boundary versions: statistics use ONS upper-tier area codes; the postcode lookup accepts current equivalents (notably Sheffield postcodes now resolve to E08000039 — same city, newer code) and, in two-tier county areas, maps the postcode’s district to its county.
  • Up to 35 indicator cards across 18 topics per area — including physical activity, long-term conditions, deprivation, winter mortality, sexual health, suicide, happiness, depression and anxiety — plus comparison cards where archived observations exist (5,241 cards across 153 areas).
  • Coverage is England only: Scotland, Wales and Northern Ireland are not included, and postcodes there are reported as unsupported.
  • Postcodes only locate an area; statistics describe the whole local authority, not a street.

Sources & licensing

  • Primary source: OHID Fingertips public-health API (fingertips.phe.org.uk/api), retrieved 8 Oct 2026, with comparison observations rechecked on 10 Oct, the national dataset refreshed on 13 Oct, and supplementary OHID observations retrieved on 14 Oct 2026. All selected OHID observations were reconciled with a fresh official download on 16 Oct 2026.
  • Direct ONS source: local-authority personal well-being estimates, version 4, for happiness and anxiety.
  • Underlying suppliers: ONS (mortality, life expectancy and surveys), Sport England (activity), NCMP/DHSC (children's weight), UKHSA (STI surveillance), Defra/Ricardo (air), and NHS England QOF registers.
  • Fingertips content: © Crown copyright, Open Government Licence. Cite as: Office for Health Improvement and Disparities, Public health profiles, 2026.
  • Postcode lookup: postcodes.io (ONS Postcode Directory). We do not store full postcodes.
  • Artwork: original SVG doodles. Leeds Town Hall photo by Gunnar Larsson; photo and adapted monochrome collage under CC BY-SA 3.0.

How we match & validate

  • England comparisons match indicator ID, sex, age, period, unit and geography. Air-pollution indicators are Not compared by Fingertips — our numerical differences describe modelled concentrations, not statistical significance or overall air cleanliness.
  • Comparison cards use unrounded values. Air difference = 100 × (local PM2.5 / England PM2.5 − 1). Childhood obesity ratio = local Year 6 prevalence / comparator prevalence. Both observations must match indicator, period, sex and age. These are area-level descriptions, not personal risk, causal effects or city-to-city significance tests.
  • Missing or suppressed values are never shown as zero and never coerced; every published card has a real numeric value.
  • IMD uses corrected MHCLG 2025 workbooks and official ranks of average score. Where boundaries match, the prominent rank compares 296 district-level authorities; county councils have only the separate 153-place upper-tier comparison. For upper-tier wording, Ranks 1–70 say “most deprived”; ranks 71–153 use the reversed “least deprived” rank. This is a display convention, not an official deprivation category.
  • All-cause mortality, happiness, anxiety, winter mortality, STI, suicide and recorded-depression ranks are derived from the unrounded published point estimates among upper-tier authorities with usable values. Exact ties share a rank. These are descriptive comparisons, not official publisher league tables; overlapping confidence intervals mean nearby ranks may not be meaningfully different.
  • Diabetes GP-register prevalence includes type 1 and type 2. The separate modelled type-2 estimate has a different year, age band and population; subtracting the two does not estimate undiagnosed cases.
  • Percent vs percentage-point changes are kept distinct; life expectancy vs healthy life expectancy are separate cards.
  • “Better/Worse/Similar” repeats Fingertips' own wording with confidence intervals; colour is decoration, never the message.
  • Source snapshots are retained so each published figure can be traced to its reporting period and population.

Limitations & timing

  • There is no single reporting year: cards span 2021 estimates, 2024, 2024/25, 2025 and multi-year pools. Each card states its own period.
  • 2025 life-expectancy uses provisional deaths/population and will be revised.
  • Heart-attack admissions are pooled to 2020/21 (the latest matched release) and the undiagnosed-hypertension figure is a 2021 modelled estimate — both labelled on their cards.
  • QOF register cards (including hypertension, heart disease, kidney disease, diabetes, asthma, COPD and depression) count diagnosed cases only and are not age-adjusted; a lower register figure can mean under-diagnosis. Patients are assigned by GP practice location, not necessarily their residence.
  • Survey cards (activity, smoking, healthy life, happiness and anxiety) carry confidence intervals — small differences and nearby ranks may be noise.
  • The latest local winter mortality index is 2021/22 and publication has been paused. Happiness and anxiety use the latest ONS local-authority series, 2022/23.
  • The fraction of mortality attributed to air pollution is a modelled estimate, not counted deaths; PM2.5 is population-weighted modelling.
  • Dataset quality-control review: 16 Oct 2026. Failed refreshes never overwrite the last good version (documented in README).
  • Location lookup accepts postcodes, place names and optional browser location; statistics still describe the whole authority.

Questions about methods: see Fingertips guidance on geography and public-health methods, or the per-card source links. These statistics make no causal claims. IMD uses official MHCLG ranks; the seven derived-rank measures clearly label their separately derived point-estimate comparisons and their uncertainty.