Pediatric Asthma Desert Mapping

ISEF Category: Biomedical and Health Sciences

Subcategory: Other  ·  Difficulty: Intermediate  ·  Setup: Home Setup  ·  Time: 1 to 2 Months

The Hook

A child’s asthma flare-up can start long before the ambulance ride. Dirty air, long drives to care, and uneven clinic access can stack up in the same neighborhood. You can map those gaps with public data and look for places where risk hits hardest.

What Is It?

Geospatial analysis means putting data on a map and comparing layers. Think of it like stacking clear sheets on a table. One sheet shows pediatric asthma ER visits, one shows air pollution, and one shows clinic access. When the hot spots line up, you start to see where kids may face the biggest risk.

An "asthma desert" is not a formal medical label. For this project, you can define it as a place with high asthma burden, higher pollution, and weak access to nearby care. That lets you test a real pattern instead of just describing one map. You are asking whether geography itself helps explain who ends up in the ER.

Why This Is a Good Topic

This topic works well because public data already exist, and the question is easy to test with maps and statistics. You can connect a real health problem to a real-world fix, like mobile clinic placement or targeted outreach. You also learn skills that matter in research, like cleaning data, joining datasets by location, and checking whether a pattern holds up after you control for population size and demographics.

Research Questions

• How does pediatric asthma ER visit rate change with local air pollution levels?
• What is the effect of clinic distance on pediatric asthma ER visit rates after accounting for pollution?
• Does combining AirNow pollution data with Census access data improve prediction of asthma ER hotspots?
• To what extent do high-risk tracts cluster near both low clinic access and high pollution?
• Which geographic unit, ZIP code or census tract, separates asthma desert areas most clearly?
• How does seasonal air quality change the overlap between pollution spikes and pediatric asthma ER visits?

Basic Materials

• Laptop with internet access.
• Spreadsheet software such as Google Sheets or Excel.
• QGIS, a free mapping program.
• Python with pandas, geopandas, matplotlib, and statsmodels, or R with sf and ggplot2.
• CDC or state pediatric asthma ER visit data.
• EPA AirNow air quality data.
• US Census Bureau tract or ZIP code boundary files and American Community Survey variables.

Advanced Materials

• ArcGIS Pro with spatial analyst tools.
• Python with geopandas, pysal, and scikit-learn, or R with sf, tmap, and mgcv.
• Geocoded clinic and hospital location data.
• High-resolution EPA CMAQ or satellite-based PM2.5 surfaces.
• Travel-time network data for drive-time access modeling.
• Bootstrap or permutation-testing workflow for spatial validation.

Software & Tools

• QGIS: Maps census tracts, clinic sites, pollution layers, and ER visit hotspots.
• Python: Cleans public datasets, joins locations, and runs regression models.
• R: Fits spatial statistics and checks whether cluster patterns stay strong after controls.
• Google Colab: Lets you run Python notebooks without installing anything on your computer.
• GeoDa: Tests spatial autocorrelation and helps you spot clusters, outliers, and hot spots.

Experiment Steps

1. Define the map unit you will study, such as census tracts, ZIP codes, or counties.
2. Choose one outcome, one pollution measure, and one access measure before you build the map.
3. Build a clean table that joins ER visits, AirNow data, and Census variables by location.
4. Set a baseline model so you can compare pollution-only, access-only, and combined predictions.
5. Plan a map that flags asthma desert hot spots and checks whether they cluster in space.
6. Decide how you will test your model on held-out places so the pattern is not just fitting noise.

Common Pitfalls

• Mixing county-level ER data with tract-level Census layers, which blurs the geography and weakens the map.
• Using raw ER counts instead of rates, which makes large-population areas look falsely high risk.
• Comparing pollution data from different averaging windows, which turns a real exposure signal into noise.
• Ignoring clinics just outside the study boundary, which makes edge neighborhoods look farther from care than they are.
• Calling every high-pollution area an asthma desert, which skips the access side of the problem.

What Makes This Competitive

A stronger version goes past a heat map. You compare pollution-only, access-only, and combined models, then test which one really predicts pediatric ER burden. The best projects also check spatial autocorrelation, so clustered hotspots do not fool the statistics. That turns a simple map into a real risk model.

Project Variations

• Swap ER visits for hospitalization rates to see whether the same hotspots hold.
• Use travel time by road instead of straight-line distance to measure clinic access.
• Compare one city across different seasons to see whether asthma desert patterns stay stable.

Learn More

• CDC National Environmental Public Health Tracking Network: Search for asthma, air quality, and neighborhood health indicators by place.
• EPA AirNow: Find daily air quality data and learn how AQI categories are defined.
• US Census Bureau American Community Survey: Pull neighborhood-level demographics, poverty, insurance, and housing variables.
• NIH PubMed: Search review articles on pediatric asthma, air pollution, and health access.
• MIT OpenCourseWare: Search for free lectures on GIS, statistics, and data analysis if you want extra background.