Manage vaccinations in schools prototype

Manage vaccinations in schools (Mavis) is a digital tool designed to help school age immunisation service (SAIS) teams manage their vaccination programmes end to end.

This is a prototype of the service, built using the NHS prototype kit.

Requirements

Node.js v22.21

Installation

1. Clone this repository

2. Install the dependencies and create data:
   npm install

3. Start the application:
   npm start

Creating session data

Session data uses pre-compiled JSON files saved to a .data folder.

To regenerate this data run:
npm run create-data

Mostly, the generated data comprise randomised records, with distributions of different statuses governed by weighted probabilities. However, there are also some records generated deterministically to match scenarios we want to use in research.

Adding a scenario

1. Add lib/scenarios/library/<what-it-sets-up>.js:

   export default defineScenario({
     name: 'Flu session at Grange Hill, consent window closing tomorrow', // must be unique
     description: 'Written for researchers and designers, not just devs.',
     run(context, shared) {
       // create sessions / children / replies etc.
       // ...
   
       // return URLs to any useful pages, with meaningful link text
       return {
         startUrl: session.uri, // a deliberate choice of the best demo entry point
         links: [{ label: 'Session', href: session.uri }]
       }
     }
   })

2. Import the scenario from the new file and add it to the scenarios array in lib/scenarios/index.js.

3. Run npm run create-data then check /scenarios to test links, description, etc.

Rules of thumb

• Scenarios are independent; they only use shared, never each other's output. If two scenarios need the same thing (e.g. a school), build it once in shared-resources.js, add it to the SharedResources typedef in types.js, and reserve any IDs in constants.js.
• Create your own children rather than picking from the random pool. Scenarios run after all random generation, so a random child at a given school may already have consent, vaccinations and so on.
• Reuse the existing generators and models (app/generators/*, Model.create) as create-data.js does.
  • Note that generateSession looks schools up in app/datasets/schools.js, so any school used with it must live in that dataset. The scenario schools (see ScenarioSchoolUrns in lib/scenarios/constants.js) do, and create-data.js leaves them out of the random children and sessions it generates.
• Keep capabilities in helpers.js and policy in populate-session.js.
  • lib/scenarios/helpers.js offers scenario-focused building blocks that know nothing about programme statuses, like creating a session, a child with parents or a consent reply, along with options such as isUnder16.
  • lib/scenarios/populate-session.js holds the business logic for a mix of statuses: what backs each status (consent replies, triage decisions, vaccination records) and when to use those options.
  • A scenario file holds only its own settings (e.g. school, date, shares). Dependencies point one way: scenario, then populate-session.js, then helpers.js, then app/.

How your scenario gets incorporated

After creating all of the randomised data, lib/create-data.js calls createScenarioData(), which writes .data/scenarios.json. app/data.js loads that file, and app/models/scenario.js, app/controllers/scenario.js and app/views/scenario/list.njk render the page.