A system that analyzes hospital pricing data to help business owners
make better health insurance decisions for their employees
Small business owners told me they lacked the information to confidently choose employee health insurance.
Comparing insurers’ medical prices requires time and resources these owners don’t have.
The price of buying coverage
Different insurers negotiate different prices for the same care
Through deductibles and coinsurance, depending on their plan
The project
A system that turns millions of fragmented healthcare data points into clear comparisons of what insurers pay for the same outpatient hospital care. Enabling business owners to assess potential healthcare costs for employees.
My role
I defined the question, chose the data sources, set the rules for cleaning the data, and decided how to compare insurers fairly. Anthropic tools handled the coding and data processing.
I started with Alabama, analyzing prices across nearly all of the state’s major hospitals.
pays 30% less than the average insurer
pays 15% less than the average insurer
pays 3% more than the average insurer
pays 7% more than the average insurer
pays 19% more than the average insurer| Insurer | vs. average insurer |
|---|
Minus means lower prices. Plus means higher prices.
Select an insurer to compare prices across hospitalsSelect an insurer to see the analysis
In Alabama, Blue Cross’s consistently lower outpatient hospital prices make it a strong option across the state.
Viva Health’s particularly low prices at UAB may make it attractive for businesses whose employees live near Birmingham and are likely to receive care there.
As the Alabama results illustrate, two plans can cover the same hospitals yet expose employees to very different costs for the same care.
Hospitals publish the prices they negotiate with insurers. But comparing those prices means analyzing an enormous volume of data.
Source: Trilliant Health MRF Data Directoryoria-data.trillianthealth.com
Preparing the data for accurate comparisons across insurers required substantial cleaning and standardization.
For every procedure at every hospital, I compared what insurers pay for the same care and measured the differences.
Blue Cross bundled services at five hospitals, making prices hard to compare. I uncovered the matching service codes in a New York State document, adding 25% of Alabama’s hospital market.
Source: Trilliant Health MRF Data Directoryoria-data.trillianthealth.com
Source: New York State EAPG Crosswalkemedny.org/Crosswalk/
Common procedures were weighted more heavily to reflect how often they occur.
With limited public data on procedure frequency, I used Minnesota’s all-payer claims database as a proxy.
Source: Minnesota All-Payer Claims Databasehealth.state.mn.us/data/apcd/publicusefiles
Larger hospitals handle more procedures, so they were given more weight using patient revenue data from federal reports.
Source: Medicare Hospital Provider Cost Reportdata.cms.gov/provider-compliance/cost-reports/hospital-provider-cost-report
Leveraging Anthropic tools, I identified and evaluated public healthcare data sources, analyzed hospital prices, shaped the narrative, and designed this website.
This project demonstrates how I use AI to take a complex business question from initial research through to a finished product.
Explore my AI portfolio