Immunity Map

Immunity Map is an MMR vaccination coverage tracker for England. It groups UKHSA COVER records by the postcode district of each GP practice and presents descriptive coverage bands in a black/red/cream civic dashboard style.

Repository note: the GitHub repository remains HerdWatch, but the public-facing product name is now Immunity Map.

Current status

Immunity Map has been rebuilt as a clean editable React + TypeScript + Vite project and now uses rebuilt real data files rather than the earlier placeholder trend/scaffold state.

The Explorer is a real MapLibre geographic view. It plots every represented practice-postcode district at a lightweight ONS-derived reference centroid, draws a generalised official England outline, distinguishes small samples, and keeps the complete filterable table as an accessible fallback.

Current generated metadata reports:

• Example data: false
• Source files: data/raw/areas.csv and data/raw/trends.csv
• Postcode-district area records: 1,857
• Duplicate postcode districts: 0
• Like-for-like quarterly trend points: 2
• Well below target areas (below 90%): 856
• Below target areas (90% to below 95%): 594
• Areas meeting the target (95% or higher): 407

The public dashboard is an explanatory public-health data interface, not medical advice or an official NHS/UKHSA service. Practice postcode districts must not be interpreted as the home location of registered children.

Important data note

The editable source project now includes normalised generated data under both src/data/generated/ and public/data/.

The raw official source downloads themselves are not committed because the original source/reference files can be large and are reproducible through the data pipeline. The committed normalised CSV/JSON outputs are what the React app builds from.

Before presenting the project as fully production-grade public-health infrastructure, keep the source methodology visible and periodically refresh the data from official NHS/UKHSA sources.

Programmatic local SEO

The production build now creates static route entrypoints for every generated postcode-district page under:

/town/{postcode-district}/

Each generated town page gets route-specific search metadata:

• unique <title>;
• unique meta description;
• index, follow robots tag;
• canonical URL;
• Open Graph and Twitter metadata;
• JSON-LD WebPage structured data;
• a noscript local coverage summary using the same generated area data.

The build also generates dist/sitemap.xml from the current route list and generated area data. The sitemap includes the core public routes and every /town/{postcode-district}/ page.

public/robots.txt points crawlers to the generated sitemap.

A final build-time branding pass applies the public product name Immunity Map across generated HTML/JS/CSS/XML/TXT output so older standalone static pages do not leak the previous public brand.

Tech stack

• Vite
• React
• TypeScript
• Netlify static deployment
• Node data-build scripts
• Node static SEO route generation
• Node sitemap generation
• Node public branding pass
• Python COVER conversion scripts
• MapLibre GL JS geographic explorer

Local development

npm install
npm run dev

Production build

npm run build
npm run preview

The production build runs the data pipeline, builds the Vite app, creates static route entrypoints, creates the sitemap, injects the AdSense meta tag, then applies the public branding pass.

Netlify settings are included in netlify.toml:

[build]
  command = "npm run build"
  publish = "dist"

Data pipeline

The app reads generated JSON files from:

src/data/generated/areas.json
src/data/generated/trends.json
public/data/areas.json
public/data/trends.json
public/data/metadata.json
public/data/map-centroids.json
public/data/england-outline.geojson

These files are built from CSV inputs under:

data/raw/areas.csv
data/raw/trends.csv

Run the standard data pipeline with:

npm run data:build

The script will:

• parse the raw CSV files;
• calculate risk status from coverage;
• validate duplicate postcode districts;
• validate coverage ranges;
• reject records where vaccinated children exceed eligible children;
• write generated JSON for the React app;
• write public JSON for static/route use;
• write metadata to public/data/metadata.json.

Expected area CSV columns:

postcode_district,region,practice_count,total_eligible,total_vaccinated,coverage

Expected trend CSV columns:

year,england_mmr1,england_mmr2,target

If coverage is blank in the area CSV, it will be calculated from total_vaccinated / total_eligible * 100.

Geographic map data

The deployable geography files are derived from the official ONS Postcode Directory (May 2026) and ONS Countries (December 2025) ultra-generalised boundary. Each map point is the arithmetic mean of live England postcode-unit centroids in an outward postcode district represented by the COVER data.

The point is a district reference location only. It is not a patient location, exact GP-practice coordinate, postcode boundary or resident-population estimate.

The 235 MB ONSPD source archive is deliberately not committed. After downloading it from the ONS Open Geography Portal, refresh and validate geography with:

npm run map:geography -- --onspd-zip /path/to/ONSPD_MAY_2026.zip
npm run map:validate

Every production build runs map:validate and will fail if current coverage districts and map coordinates fall out of sync.

Real COVER source workflow

The repo includes scripts for refreshing from official COVER-style source files:

npm run data:download
npm run data:cover:areas
npm run data:cover:trends
npm run data:cover:all

data/raw/source/, data/raw/ref/ and processed reports are ignored because they are reproducible/downloaded build inputs.

Project structure

src/
  App.tsx
  main.tsx
  styles.css
  types.ts
  data/
    areas.ts
    trends.ts
    generated/
      areas.json
      trends.json
data/
  raw/
    areas.csv
    trends.csv
    areas.example.csv
    trends.example.csv
  processed/
    .gitkeep
scripts/
  apply-public-branding.mjs
  build-data.mjs
  build-cover-areas.py
  build-cover-trends.py
  build-map-geography.py
  validate-map-data.mjs
  create-sitemap.mjs
  create-static-route-entrypoints.mjs
  download-cover-sources.mjs
  inject-adsense-meta.mjs
public/
  data/
    areas.json
    trends.json
    metadata.json
    map-centroids.json
    england-outline.geojson
  _redirects
  _headers
  404.html
  editorial.css
  favicon.svg
  og-immunity-map.png
  robots.txt
index.html
netlify.toml
vite.config.ts
tsconfig.json

Potential next upgrades

1. Add local authority/ICB summary pages for search and public usefulness.
2. Add a /rankings/ page for lowest coverage, biggest estimated unvaccinated counts and areas closest to the 95% target.
3. Add screenshots to this README.