Project

Chicago Healthcare Network

Optimization-Based Decision Support for Equitable Access

In Preparation with Hemanshu Kaul & Kim Erwin

About

This project develops an optimization-based framework for analyzing and improving equitable access to healthcare facilities across Chicago. Using network models and facility location theory, we identify gaps in coverage, model patient travel burdens, and recommend facility placements that improve access for underserved communities.

Interactive Dashboard

The dashboard is designed as a decision-support tool for health sector investors, policymakers, and public officials. It brings together facility coverage, community health indicators, and geographic boundaries to help stakeholders:

  • Identify medically underserved areas and coverage gaps
  • Evaluate candidate sites for new primary care facilities
  • Analyze spatial disparities in health outcomes across neighborhoods
  • Support data-driven resource allocation and investment decisions
Open Dashboard →

Data Sources

Dataset Provider Coverage Used For
Community Health Indicators Chicago Health Atlas 77 community areas 33 health indicator layers
HRSA Health Centers HRSA / data.gov Chicago, IL Federally qualified health centers (170 sites, 43 non-primary-care excluded)
Healthcare Facilities Google Places API Chicago, IL Hospitals, primary care, urgent care (416)
NPPES Providers CMS NPPES Chicago, IL Individual provider counts by specialty per facility (61,460 providers)
Medically Underserved Areas HRSA MUA/P Chicago, IL HRSA-designated underserved areas with IMU scores (35 areas)
Geographic Boundaries City of Chicago / U.S. Census Chicago, IL Community areas, census tracts, health zones

AI Usage Disclosure

This dashboard was developed with assistance from Claude (Anthropic). AI was used for:

  • Deck.gl layer configuration, SvelteKit component structure, and reactive state management to embed the dashboard into this webpage,
  • Querying and interpreting the Chicago Health Atlas REST API to call indicator data,
  • Writing Python scripts for exporting GeoJSON files.

All AI-generated code and content was reviewed and accepted by the author. The underlying data sources (Chicago Health Atlas, HRSA, CMS NPPES, Google Places, U.S. Census Bureau) are independent of AI and are cited with full provenance in DATA_SOURCES.md.

AI was not used for: research design, selection of health indicators, interpretation of health outcomes, or any analytical conclusions drawn from the data.

Let's Talk Data