Wait Time Canada

A clinically defensible Health Systems Observatory for Canadian emergency department wait-time methodology and data quality.

📊 Overview

Wait Time Canada is not a simple wait-time leaderboard. It's a comprehensive health systems observatory that audits and exposes methodological inconsistencies in Canadian emergency department reporting across provinces.

Current Coverage

4 Provinces: Quebec, Ontario, Alberta, British Columbia
380+ Hospitals: Public source monitoring across all regions
Source Refreshes: Scheduled or manually dispatched collection from public provincial sources, with the current cadence posture tracked in the public roadmap
First-in-Canada: ED stretcher occupancy visualization (Quebec)
Ontario Public Resources Module: /resources now serves facilities, OSM-backed AED fallback data, Health Canada recalls, AQHI, analytics-only Ontario EMS system context, Ontario reserve-system long-term drinking water advisories, and an official Ontario naloxone link-out with a first-class source catalog

Why This Matters

Provincial health authorities report ER wait times using fundamentally different methodologies:

Quebec starts the clock at REGISTRATION (administrative check-in)
Ontario starts at TRIAGE (clinical assessment)
Different statistical measures (P90 vs rolling averages)
Different patient populations (all vs mid-acuity)

Direct comparison without methodology awareness is clinically misleading. This observatory makes those differences transparent.

By the Numbers

As reflected in the current runtime and roadmap baseline on 2026-06-13:

4 provinces remain in the active audited source set: Quebec, Ontario, Alberta, and British Columbia
380+ hospitals are represented across the current provincial inventories
4 distinct statistic types are still present across the active provinces: MEAN, ROLLING_AVG, POINT_ESTIMATE, and P90
6 of 6 cross-province source pairs still fail direct comparability on the current ontology tags
1 province currently publishes live stretcher occupancy in the platform: Quebec

These are methodology and operations findings, not performance rankings. They describe what the system is auditing and how the data is structured, not which province or hospital is "better."

Limitations

Scraper freshness reflects scheduled ingestion rather than continuous real-time monitoring.
Provincial methodology labels are inferred from public documentation and can lag unannounced reporting changes by source organizations.
Reported wait times and occupancy values are limited to what provinces publish; the platform cannot surface unreported overcrowding or internal flow constraints.
The equity layer is currently implemented for Ontario only and should not be generalized to other provinces without province-specific source and tract validation.

Public Documentation Boundary

This repository contains public project documentation and reproducible development information. Deployment details, credentials, monitoring configuration, private operational notes, and environment-specific production paths are intentionally excluded from public documentation.

✨ Key Features

🎯 Methodology Transparency

Ontology-Based Architecture: Every measurement tagged with start_event, end_event, statistic_type, patient_scope
Divergence Warnings: Automatic alerts when comparing incompatible metrics
Comparability Matrix: Visual display of cross-province methodology alignment
Deep Linking: Share specific hospital comparisons with methodology context

📈 Data Quality & Freshness

Anomaly Detection: Automated flagging of suspicious measurements
Data Quality Dashboard: Real-time visibility into scraper health, measurement counts, staleness
Freshness Monitoring: Heartbeat checks flag stale or failed source updates
Methodology Change Detection: Tracks when provincial reporting methods change

🚑 Public Health Resources

Facilities & Services: Ontario facility search grounded in MOHSERLO reference-directory data
AED Fallback Map: OpenStreetMap-based AED locations with explicit crowdsourced/incomplete labeling
Safety Alerts: Health Canada recalls and safety alerts with optional Drug Product Database enrichment
Ontario EMS Context: Annual land-ambulance response-time summaries presented as background system context only
Ontario Water Advisories: Official ISC long-term drinking water advisories for Ontario reserve systems with freshness-aware suppression
Official Naloxone Link-Out: Ontario naloxone guidance is linked to the provincial source rather than republished until reuse rights are clearer
AQHI Overlay: Cached GeoMet air-quality snapshots with freshness-aware suppression when upstream data is stale
Source Transparency: Provenance links, last refresh timestamps, and domain-specific caveats on every public-health slice

🏥 Live Occupancy (Quebec)

First-in-Canada: Stretcher occupancy percentage visualization
Color-Coded Indicators:
- 🟢 Green (<90%): Below capacity
- 🟡 Yellow (90-110%): Near capacity
- 🔴 Red (>110%): Overcrowded with pulse animation
Clinical Context: >100% indicates overcrowding and potential extended waits

📊 Analytics & Benchmarking

Peer Benchmarking: Hospital performance vs regional/provincial averages
Temporal Patterns: Hour-of-day, day-of-week, monthly trend analysis
Regional Intelligence: 15 health regions mapped with analytics segmentation
System Trends: Province-wide performance tracking with 90d/6m/1y views

🗺️ Interactive Map

Mapbox Integration: Geographic visualization of all hospitals
Live Data Indicators: Real-time updates highlighted
Distance Calculation: User location-based sorting
Cluster Markers: Efficient rendering of 380+ locations

📤 Data Export

Citation-Ready: CSV/JSON export with methodology metadata
Granularity Control: Raw measurements, hourly, daily, weekly, or monthly aggregates
Research-Grade: Full audit trail with payload hashing and parser versioning

🔍 Access & Equity Insights

Access Burden Estimator: Fuel + parking cost calculations for patient decision-making
Provincial Gas Price Awareness: ON $1.55/L, QC $1.60/L, BC $1.75/L
Distance-Based Analysis: Nearest hospital identification within radius
Equity Layer Scaffold: Foundation for census tract income overlays (future)

🏗️ Technical Architecture

graph TD
    subgraph "Provincial Sources"
        QC[Quebec MSSS<br/>BeautifulSoup]
        ON[Ontario Health<br/>HTTP + HTML tables]
        AB[Alberta AHS<br/>Playwright]
        BC[BC PHSA<br/>JSON/__NEXT_DATA__]
    end

    subgraph "Scheduled Collection"
        CRON[Provincial Scrapers]
        HB[Freshness Checks]
    end

    subgraph "PostgreSQL Database"
        SOURCES[(sources)]
        HOSPITALS[(hospitals)]
        MEASUREMENTS[(measurements)]
        AGGREGATES[(measurement_aggregates)]
        QUALITY[(data_quality_snapshots)]
        STATUS[(scraper_status)]
    end

    subgraph "Next.js 14 Frontend"
        API[API Routes<br/>/api/hospitals<br/>/api/analytics<br/>/api/data-quality]
        PAGES[Pages<br/>Map View<br/>Analytics Dashboard<br/>Methods Page<br/>Resources]
        MAP[Mapbox GL JS<br/>380+ Hospital Markers]
    end

    subgraph "Services"
        DB_SVC[DatabaseService]
        AGG_SVC[AggregationService]
        DQ_SVC[DataQualityService]
        ANOM_SVC[AnomalyDetectionService]
    end

    QC --> CRON
    ON --> CRON
    AB --> CRON
    BC --> CRON

    CRON --> DB_SVC
    DB_SVC --> MEASUREMENTS
    DB_SVC --> STATUS

    HB --> STATUS
    HB -->|State changes| DQ_SVC

    MEASUREMENTS --> AGG_SVC
    AGG_SVC --> AGGREGATES

    MEASUREMENTS --> ANOM_SVC
    ANOM_SVC --> DQ_SVC
    DQ_SVC --> QUALITY

    SOURCES -.-> API
    HOSPITALS -.-> API
    MEASUREMENTS -.-> API
    AGGREGATES -.-> API
    QUALITY -.-> API

    API --> PAGES
    PAGES --> MAP

    style QC fill:#4A90E2
    style ON fill:#4A90E2
    style AB fill:#4A90E2
    style BC fill:#4A90E2
    style CRON fill:#F5A623
    style HB fill:#F5A623
    style MAP fill:#50E3C2

Backend

Language: Python 3.12+
Testing: pytest with broad unit coverage plus database-backed integration tests; some backend integration and smoke checks require DATABASE_URL, and the local pipeline smoke path also expects a frontend server on localhost:3000
Scrapers: 4 provincial scrapers (BeautifulSoup, HTTP/HTML parsing, Playwright, JSON extraction)
Database: PostgreSQL with 15 tables, including the public-health-hub source/resource/alert/system-context lane
Services:
- DatabaseService, AggregationService, DataQualityService
- AnomalyDetectionService, MethodologyChangeDetector
- GeocodingService (Nominatim), HeartbeatService
CLI Tools: Scraper runner, database cleanup, storage stats, seeding, aggregation, region mapping

Frontend

Framework: Next.js 14 App Router + TypeScript
Testing: Vitest + React Testing Library for unit/component coverage, with Playwright browser verification kept CI-first, manual-dispatch, and outside the default push/PR merge gate
Mapping: Mapbox GL JS
Components: 30+ React components with targeted coverage across major user-facing flows
API Routes: Hospitals/comparisons, analytics, data quality, and public resources endpoints
Pages: Home (map + list), /data-quality, /analytics, /methods, /resources, /about

Database Schema (15 Tables)

sources - Provincial data source metadata
hospitals - Facility metadata with verification workflow
measurements - Audit log with ontology tags (payload hashing, not full HTML)
scraper_status - Scraper freshness state
scraper_alert_state - Stateful source-health tracking and recovery suppression
measurement_aggregates - Permanent statistical summaries (hourly/daily/weekly/monthly)
data_quality_snapshots - Daily scraper reliability metrics
methodology_change_events - Detected methodology shifts
regions - Province region metadata for analytics
hospital_regions - Hospital-to-region mappings
public_data_sources - Public-health-hub source catalog and freshness metadata
resource_locations - Normalized facilities and AED locations for /resources
public_health_alerts - Health Canada recall and safety alert records
public_health_system_metrics - Ontario EMS system-context records for analytics-only /resources context cards
public_health_source_alert_state - Stateful public-health ingest status tracking

Automation

Scheduled Workflows: Scrapers, freshness checks, quality snapshots, and bounded retention cleanup
Playwright Browsers: Automated for Alberta and browser-based verification flows
Health Tracking: Source status is reconciled through state-change-aware checks
Cache Guardrails: Shared anonymous read-heavy routes use short cache windows; user-specific and export routes use no-store

Project Impact

Wait Time Canada demonstrates public-interest software development, health-systems thinking, documentation, and responsible handling of health-adjacent information.

The project is designed to:

make provincial wait-time methodology differences visible instead of masking them through unsafe normalization
preserve provenance and source semantics for every measurement
help users understand when direct hospital or province comparisons are invalid
support research, journalism, policy analysis, and public understanding with transparent caveats
provide public-health resource discovery while clearly labeling source freshness, scope, and limitations

This is a non-clinical observatory. It does not provide medical advice, diagnose conditions, recommend care sites, or replace emergency services.

🚀 Quick Start (Local Development)

This section is for reproducible local development. Production deployment details and environment-specific operations notes are intentionally excluded from public documentation.

Prerequisites

Python 3.12+
Node.js 22+
PostgreSQL
Mapbox account (free tier sufficient)

Recommended Database Path: PostgreSQL

Wait Time Canada uses standard PostgreSQL and can be run against any compatible database:

create or provision a PostgreSQL database
copy the connection string
export it into your shell before running backend commands
run the repo migrations and analytics bootstrap commands below

This keeps local setup portable while preserving a reproducible schema and migration workflow.

1. Clone and Setup Environment

git clone https://github.com/yourusername/waittimecanada.git
cd waittimecanada

# Frontend environment
cp frontend/.env.example frontend/.env.local
# Edit frontend/.env.local with NEXT_PUBLIC_MAPBOX_TOKEN

# Backend runtime config
export DATABASE_URL="postgresql://user:pass@host:5432/dbname" # pragma: allowlist secret

backend/.env.local can still be kept as a human convenience template, but the backend runtime expects DATABASE_URL to already be present in the process environment.

2. Backend Setup

cd backend

# Create virtual environment
python3 -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install --upgrade pip
pip install -e '.[dev]'  # Note: use quotes in zsh

# Install Playwright browsers (required for Alberta scraper)
playwright install chromium

# Apply database migrations
python run_migrations.py

# Seed sources and bootstrap analytics
python -m waittime.cli.bootstrap_analytics --days 180

3. Frontend Setup

cd frontend

# Install dependencies
npm install

# Run development server
npm run dev

Open http://localhost:3000 to view the app.

4. Run Scrapers (Optional)

cd backend
source .venv/bin/activate

# Run all scrapers
python -m waittime.cli.scraper --all

# Or run single province
python -m waittime.cli.scraper --source quebec-msss

📚 Common Commands

Backend

source .venv/bin/activate

# Scrapers
python -m waittime.cli.scraper --all              # Run all scrapers
python -m waittime.cli.scraper --source ontario-health  # Single province
python -m waittime.cli.scraper --dry-run --all    # Test without DB writes

# Maintenance
python -m waittime.cli.check_heartbeat            # Verify scraper health
python -m waittime.cli.aggregate --backfill       # Regenerate aggregates
python -m waittime.cli.cleanup --dry-run          # Preview cleanup
python -m waittime.cli.storage_stats              # Inspect measurements table growth

# Testing
pytest tests/unit                                  # Unit tests only
pytest tests/integration                           # Integration tests (requires DATABASE_URL)
pytest tests/ -v --cov=waittime                   # Full suite with coverage; backend smoke tests may skip without DATABASE_URL and a local frontend server

Frontend

cd frontend

# Development
npm run dev          # Start dev server
npm run build        # Production build
npm run start        # Run production build

# Quality
npm run type-check   # TypeScript validation
npm run lint         # ESLint checks
npm run test:unit    # Vitest unit tests

📖 Documentation

Essential Reading

docs/planning/roadmap.md - Active roadmap and milestone status
AGENTS.md - repository agent instructions and project architecture

Deep Dives

docs/adr/ - Architecture Decision Records (25 ADRs)
backend/docs/methodologies/ - Provincial methodology documentation
docs/case-studies/ottawa-gatineau-divergence.md - Cross-border methodology divergence case study
backend/README.md - Backend architecture and testing
frontend/README.md - Frontend architecture and testing

API Reference

docs/API.md - API contracts and examples
docs/reference/data-dictionary.md - Data Dictionary & Schema
docs/architecture/data-flow.md - Data Flow Architecture
Endpoints: /api/hospitals, /api/comparisons, /api/analytics/*, /api/data-quality

🔧 Repository Structure

waittimecanada/
├── backend/
│   ├── src/waittime/
│   │   ├── scrapers/          # Provincial scrapers (QC, ON, AB, BC)
│   │   ├── services/          # Business logic services
│   │   ├── core/              # Models, enums, ontology
│   │   └── cli/               # Command-line tools
│   ├── tests/                 # broad pytest coverage (unit + integration)
│   │                          # plus an opt-in local smoke/E2E path
│   ├── migrations/            # Database schema migrations
│   ├── seed_data/             # Hospital/region seed data
│   └── docs/                  # Backend-specific documentation
├── frontend/
│   ├── app/                   # Next.js 14 App Router
│   │   ├── api/               # API routes
│   │   ├── data-quality/      # Data quality dashboard
│   │   ├── analytics/         # Analytics dashboard
│   │   └── methods/           # Methodology documentation page
│   ├── components/            # React components
│   ├── tests/                 # 419 Vitest unit/component tests
│   └── utils/                 # Utilities (cache, distance, date)
├── docs/
│   ├── planning/              # Roadmap, strategic plans
│   ├── operations/            # Operational guides
│   ├── adr/                   # Architecture decisions
│   ├── methodologies/         # Provincial methodology docs
│   └── architecture/          # System architecture
└── .github/workflows/         # CI/CD automation (16 workflows)

🎯 Operational Workflows

Production Automation

Scraper Runs: Scheduled or manually dispatched collection from public provincial sources, depending on the current quota posture
Freshness Checks: Source health is tracked from scraper status records
Database Cleanup: Manual cleanup keeps a 30-day raw-measurement window; the GitHub Actions maintenance path uses bounded batched deletes and reports storage growth

CI/CD Pipelines

Frontend CI: Type checking, linting, unit tests
Scraper CI: Python tests, coverage reporting
Production Readiness: Pre-deployment validation
Docs CI: Documentation quality checks

Deployment Configuration

Public documentation covers reproducible local development, architecture, and data methodology. Production deployment details, monitoring configuration, and environment-specific paths are kept outside public documentation.

🛡️ Project Guardrails

Clinical Safety

✅ Never provide medical advice or triage recommendations
✅ Always include emergency disclaimer: "Call 911 for emergencies"
✅ Display telehealth routing (811 varies by province)

Data Integrity

✅ Preserve source semantics - never normalize incompatible metrics
✅ Surface divergence warnings when comparisons are invalid
✅ Hash payloads (SHA256) instead of storing full HTML
✅ Tag every measurement with complete ontology metadata

Verification & Quality

✅ Government health authority sources are trusted and auto-approved
✅ Quality enforced via anomaly detection and data quality monitoring
✅ Heartbeat monitoring ensures data freshness
✅ Methodology change detection tracks reporting shifts

Attribution

✅ Link back to official provincial sources
✅ Display data provenance in all visualizations
✅ Citation-ready export formats
✅ Never claim work from automated tools as human authorship

📊 Current Status (as of 2026-06-13)

Milestones Completed

✅ M1-M4: Database foundation, Ontario/Quebec scrapers, methodology warnings, PWA setup
✅ M7-M8: UX polish, SEO, landing page optimization
✅ M9: Public launch foundations (About section, testimonial governance)
✅ M10-M11: Multi-province expansion, Access Burden Estimator
✅ M12: Research infrastructure (citation export and source-health state-change alerts)
✅ M13: Aggregation pipeline (hourly/daily/weekly/monthly)
✅ M14: Data quality & anomaly detection (3 new DB tables)
✅ M15: Analytics & benchmarking (peer comparison, temporal patterns)
✅ M16: Multi-province operationalization (4 provinces, 380+ hospitals, region mapping)
✅ M17: Quebec occupancy implementation (scraper + API)
✅ M18: Occupancy frontend UI (visual indicators on hospital cards)
✅ M28: Ontario real-data equity layer (StatsCan tract integration)
✅ M29: Ontario equity academic rigor hardening (uncertainty + interpretation limits)
✅ M30-M33: Reliability hardening, divergence briefs, deployment readiness, and historical occupancy trends
✅ Operations: Public documentation boundary cleanup and comprehensive verification posture
✅ Operations: CI coverage artifacts retained for both frontend and backend verification

Validation Posture

Backend: broad unit coverage, plus opt-in database-backed integration and local pipeline smoke checks when DATABASE_URL and a local frontend server are available
Frontend: Vitest unit/component coverage is part of the normal workflow, while Playwright browser verification stays CI-first and manual-dispatch to conserve GitHub free-tier minutes
Current source of truth: see docs/planning/roadmap.md for the active validation and deployment posture instead of relying on a fixed test-count snapshot here

Data Freshness

Update Frequency: Scheduled or manually dispatched source refreshes, depending on the current GitHub Actions quota posture
Freshness Thresholds: Source-health checks flag stale or failed updates
Current Status: See the public status and data-quality pages for current source state

💡 Future Roadmap

Planned Enhancements

Evaluate additional provinces only where public source stability supports responsible ingestion
Multi-province equity layer beyond Ontario when province-specific source and tract validation are available
Public research artifacts, including case studies and quantified methodology findings
Privacy-safe usage analytics only after an explicit privacy-policy decision
Smart scheduling (reduce frequency during overnight hours)
Occupancy-based system-pressure indicators

Deferred / Research

Manitoba scraper (data source unclear)
Saskatchewan scraper (no public data available)
Territories expansion (limited data availability)

See docs/planning/roadmap.md for detailed status and next steps.

🤝 Contributing

Contributions are welcome! Please see:

CONTRIBUTING.md - Development workflow and guidelines
CHANGELOG.md - Project history and versioning
CITATION.cff - How to cite this project

Use GitHub issue templates for feature requests and data quality reports.

📄 License

MIT License - See LICENSE for details.

Data Sources:

Quebec: Ministère de la Santé et des Services sociaux (MSSS)
Ontario: Health Ontario
Alberta: Alberta Health Services (AHS)
British Columbia: Provincial Health Services Authority (PHSA)

All data sourced from publicly available provincial health authority websites.

🎯 Mission

Wait Time Canada is a public health data and methods observatory focused on transparent interpretation of Canadian emergency department reporting.

The project exists to improve access to operational health-system information, make methodological differences visible, and support safer interpretation of reported wait-time data by patients, clinicians, researchers, journalists, and policy users.

⚖️ Equity and Access

Wait Time Canada is developed with a health equity, EDIA, and barrier-reduction lens. Public health information is often fragmented, inconsistently defined, difficult to compare, or difficult to find. Those gaps can create barriers for people navigating care, for communities experiencing unequal access, and for decision-makers trying to improve the system responsibly.

The project aims to reduce those barriers by improving transparency, surfacing limitations, documenting methodological differences, and prioritizing public-interest access to operational health-system information.

🛡️ Stewardship

Wait Time Canada prioritizes provenance, comparability, limitations, and responsible reuse of public data. It does not provide medical advice, does not claim to resolve underlying reporting inconsistencies, and does not present invalid cross-system comparisons as if they were directly comparable.

Contact: Use the repository issue tracker or the repository owner's GitHub profile for project inquiries.

📞 Emergency Disclaimer

⚠️ This is a data observatory tool, not a triage service.

For medical emergencies: Call 911 immediately
For health advice: Call your provincial health line (811 in most provinces)
Never delay emergency care based on wait time estimates

Wait time data is for informational purposes only. Clinical decisions should always prioritize patient safety over convenience.

Name		Name	Last commit message	Last commit date
Latest commit History 338 Commits
.github		.github
backend		backend
docs		docs
frontend		frontend
scripts		scripts
.editorconfig		.editorconfig
.gitignore		.gitignore
.nvmrc		.nvmrc
.pre-commit-config.yaml		.pre-commit-config.yaml
.secrets.baseline		.secrets.baseline
.zenodo.json		.zenodo.json
AGENTS.md		AGENTS.md
CHANGELOG.md		CHANGELOG.md
CITATION.cff		CITATION.cff
CLAUDE.md		CLAUDE.md
CONTRIBUTING.md		CONTRIBUTING.md
GEMINI.md		GEMINI.md
LICENSE		LICENSE
Makefile		Makefile
README.md		README.md
SECURITY.md		SECURITY.md
lighthouserc.json		lighthouserc.json
mkdocs.yml		mkdocs.yml

Folders and files

Latest commit

History

Repository files navigation

Wait Time Canada

📊 Overview

Current Coverage

Why This Matters

By the Numbers

Limitations

Public Documentation Boundary

✨ Key Features

🎯 Methodology Transparency

📈 Data Quality & Freshness

🚑 Public Health Resources

🏥 Live Occupancy (Quebec)

📊 Analytics & Benchmarking

🗺️ Interactive Map

📤 Data Export

🔍 Access & Equity Insights

🏗️ Technical Architecture

Backend

Frontend

Database Schema (15 Tables)

Automation

Project Impact

🚀 Quick Start (Local Development)

Prerequisites

Recommended Database Path: PostgreSQL

1. Clone and Setup Environment

2. Backend Setup

3. Frontend Setup

4. Run Scrapers (Optional)

📚 Common Commands

Backend

Frontend

📖 Documentation

Essential Reading

Deep Dives

API Reference

🔧 Repository Structure

🎯 Operational Workflows

Production Automation

CI/CD Pipelines

Deployment Configuration

🛡️ Project Guardrails

Clinical Safety

Data Integrity

Verification & Quality

Attribution

📊 Current Status (as of 2026-06-13)

Milestones Completed

Validation Posture

Data Freshness

💡 Future Roadmap

Planned Enhancements

Deferred / Research

🤝 Contributing

📄 License

🎯 Mission

⚖️ Equity and Access

🛡️ Stewardship

📞 Emergency Disclaimer

About

Topics

Resources

License

Contributing

Security policy

Uh oh!

Stars

Watchers

Forks

Releases 5

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages