Skip to main content

GeraClinic / Health Access Index / Methodology

Gera Health Access Index — Methodology

The full, reproducible recipe. Every number on the index pages traces back to a real NHS England or ONS dataset — the novelty is the join and the composite, never an invented value.

1. The five inputs

The Gera Health Access Index (HHAI) is a composite. It does not introduce any new measurement; it joins five datasets that are already published and verified, each licensed under the Open Government Licence v3.0:

  • A&E (4-hour) — NHS England A&E monthly statistics (March 2026). Input: Type-1 4-hour performance %. Higher = better.
  • GP access — NHS England Appointments in General Practice (April 2026). Input: same-day % + face-to-face % − DNA %, from the real regional totals. Higher = better.
  • Ambulance (Cat 1) — NHS England Ambulance Quality Indicators / AmbSYS (May 2026). Input: Category-1 mean response time in seconds, incident-weighted from trust to region (target 7m 00s). Faster = better, so the value is inverted before normalising.
  • Elective (RTT 18-week) — NHS England consultant-led RTT waiting times (April 2026). Input: % of incomplete pathways within 18 weeks. Higher = better.
  • Life expectancy — ONS life expectancy at birth (2017-19). Input: combined (male+female)/2 life expectancy. Higher = better.

2. The area-code join (the crux)

The five datasets are published at four different geographies — NHS acute trusts, Integrated Care Boards, ambulance trusts, and ONS local authorities. The only geography with complete coverage across all five is the NHS England region. We join on the region slug, which is byte-identical across the A&E, GP, RTT and ambulance datasets.

  • Ambulance: per-trust AmbSYS rows are rolled up to their NHS region by an incident-weighted mean (Σ c1Mean×c1Incidents ÷ Σ c1Incidents).
  • Life expectancy: the nine ONS statistical regions are mapped to the seven NHS regions (East Midlands + West Midlands → Midlands; North East + Yorkshire and The Humber → North East and Yorkshire); where two ONS regions form one NHS region their combined-sex life-expectancy values are averaged.

A region is included in the index only when all 5 inputs resolve — no missing input is silently treated as zero. 7 of 7 NHS England regions have all 5 inputs. 7 regions are published.

3. Normalisation and weights

Each input is min-max normalised to a 0–100 sub-score across the seven regions, oriented so higher always means better access:

sub_score = 100 * (value - min) / (max - min)
ambulance uses (-seconds) so faster = higher

HHAI = 25% * A&E
     + 25% * GP
     + 20% * Ambulance
     + 15% * Elective(RTT)
     + 15% * LifeExpectancy

The weights are a documented Gera editorial choice that prioritises the access points patients reach most often (A&E and GP), then urgent response (ambulance), then planned care (elective) and population outcome (life expectancy). They sum to 100%. On every index page the interactive checker lets you re-weight the five components yourself and see a personalised score — the index is transparent, not a black box. Lower access produces a lower score.

4. Reproducibility

The five-step method below is published as machine-readable HowTo structured data on this page. Anyone with the same five public datasets can reproduce the index exactly.

  1. Collect five real open datasets. Take five Open Government Licence v3.0 datasets already published by NHS England and the ONS: A&E 4-hour performance, GP appointment access, ambulance Category-1 response, RTT 18-week elective waits, and life expectancy at birth.
  2. Join on the NHS England region. These datasets are published at different geographies (NHS trusts, ICBs, ambulance trusts, ONS local authorities). Join them on the single common key with full coverage — the seven NHS England regions. Ambulance trusts are incident-weighted up to region; ONS regions are mapped to the NHS seven-region structure. Include a region only when all five inputs resolve.
  3. Normalise each input 0–100. Min-max normalise each input across the seven regions, oriented so higher always means better access (ambulance response is inverted so faster scores higher).
  4. Apply the documented weights. Combine the five 0–100 sub-scores with fixed weights — A&E 25%, GP 25%, ambulance 20%, elective 15%, life expectancy 15% — to produce the Gera Health Access Index (0–100, higher = better access).
  5. Re-compute on each release. Re-run the join and the index whenever a source dataset publishes a new period (the NHS components are monthly), and re-date the pages.

5. Limitations

  • The index is at NHS region level — it describes whole-system access in a region, not any single hospital, GP practice or individual visit.
  • The five source releases cover different reference periods (NHS components are recent monthly data; ONS life expectancy is the latest published period, 2017-19). The freshness label on each page states the periods.
  • Min-max normalisation is relative to the seven regions in the current release, so a sub-score reflects rank within England, not an absolute target.
  • Component figures are area averages and are not predictions for any individual patient.

Sources

← Back to the Gera Health Access Index

Informational/educational only — not a substitute for professional medical advice; a clinician interprets results.