← All work
Healthcare Analytics2026 / 15

Hospital Price Variation

A CMS inpatient-pricing analysis that exposes how charge-to-payment gaps vary by service, geography, and hospital-market context.

PythonCMS DataRegressionVisualization
Source & deliverables
2023CMS inpatient data
DRGservice comparison
USstate and regional views
HTMLself-contained report
01 / Context

The question
behind the work.

Hospitals can submit very different charges for comparable inpatient services, and charge figures alone are difficult to interpret without a common payment benchmark and market context.

My role

I prepared the 2023 CMS inpatient data, constructed DRG-level charge and Medicare-payment comparisons, analyzed state and regional variation, modeled relationships, and produced the self-contained report.

02 / Implementation

What I built.

The report uses DRG-level comparisons to make pricing variation traceable, separating a meaningful payment benchmark from a headline charge number.

  • Compared submitted charges and Medicare payments at the DRG and hospital level.
  • Mapped state and regional variation and separated service mix from geographic patterns.
  • Built scatter, gap, and market-context views for large outliers and recurring pricing structures.
03 / In detail

Comparing like with like

I used 2023 CMS inpatient pricing data to focus the analysis on comparable services rather than mixing unrelated procedures. The report centers on a high-volume DRG and uses Medicare payment as a common benchmark for comparing submitted hospital charges across hospitals, states, and regions.

The workflow calculates percentile spread, charge-to-payment ratios, state medians, regional medians, and market context so a reader can separate a general price level from an unusually wide benchmark gap.

What the price spread showed

For the focal service, submitted charges ranged from about $30,936 at the 10th percentile to $133,639 at the 90th percentile. State medians ranged from about $24,611 in Maryland to $148,155 in Nevada.

The strongest payment disconnect appeared in the West, where median submitted charges were about 5.1 times median Medicare payment for the same DRG. Metropolitan hospitals also tended to post materially higher charges than small-town and rural hospitals.

A benchmarking report for the next conversation

I packaged the work as a self-contained HTML report with service-level dispersion, regional comparisons, state views, and charge-versus-payment charts. The sequence starts with the comparable service, then moves into geography and the gap between list prices and the payment benchmark.

The result gives a hospital or market analyst a sharper question to investigate: where does the organization's pricing posture sit against true peers, and what part of the gap comes from market structure rather than the service itself?

04 / Engineering judgment

The decisions
that shaped it.

  1. Used Medicare payment as a consistent benchmark rather than treating submitted charge as realized revenue.
  2. Kept results stratified by service and geography so comparisons do not mix fundamentally different inpatient cases.

Evaluation & results

The report reproduces state, regional, service-level, and hospital-market views from one 2023 CMS source and makes high-gap observations traceable to the underlying comparison level.

05 / Working outputs

See it for yourself.

Select an image to view it at full size.

06 / Artifacts

Take a closer look.

01Interactive HTML report
Next projectCapacity & Mortality