CMS Data · Scans by Region

Why do expensive scans vary so much from place to place?

Two patients with the same complaint can get very different scans depending on where they live. There's a public record of how often each doctor orders MRI, CT, and PET scans, tagged with where it happened. The raw record isn't built to show regional patterns. Added up by area, it maps exactly where these scans run hot, where they run cold, and where the differences look like overuse.

MRI, CT & PET scansBy regionSigns of overuseAnswer in seconds
Oshri Cohen, CMS healthcare data made useful
Oshri CohenDigital products delivered
The problem

The scans are counted, the map isn't.

The public record counts how many MRI, CT, and PET scans each doctor delivered and notes where the work happened. Everything needed to study regional patterns is there, but the record is organized doctor by doctor, not place by place, so the regional signal stays scattered until you add it up by area. That's the work almost nobody does at scale.

Roll the scans up by area and the differences become visible. You can compare scan rates across states and metro areas, keep MRI separate from CT and PET, weigh it against how many people live there so you're measuring intensity rather than headcount, and flag the areas whose rates sit far above the norm, the classic sign of overuse rather than sicker patients.

What it answers

Questions you can finally ask.

Each is a question you simply ask and get an answer to, not a three-week analysis project.

Which regions scan the most?

Add up MRI, CT, and PET scans by area to rank states and metros on how heavily they use these expensive scans.

How does MRI use differ from CT and PET?

Break the differences out by scan type, since the regional pattern for MRI can look nothing like the one for CT or PET.

Where is scanning high per person?

Weigh against how many people live there to measure intensity rather than raw counts, isolating true overuse signals.

Which regions look like outliers?

Flag areas whose scan rates sit far above the national norm, the sign of differing habits rather than sicker patients.

Is high scanning from many doctors or a few?

See whether a region's high rate comes from broad habits or a handful of doctors who scan heavily.

Where is scanning unusually low?

Surface regions with low rates, the flip side, which can signal gaps in access rather than restraint.

What goes into it

What the answer pulls together.

How often scans were done

The count of how often each doctor delivered MRI, CT, and PET scans, the raw record of activity.

Just the expensive scans

The specific MRI, CT, and PET scans pulled out from everything else, so the answer covers advanced scans, not all imaging.

Where each scan happened

The location attached to each scan that lets the activity be added up by state, metro, and region.

Scanning that varies fivefold across regions with similar patients isn't a medical difference, it's a difference in habit, and the map shows you exactly where to look.

Oshri Cohen · On CMS data
Common questions

What people ask about this.

Does high scanning really mean overuse?

Not by itself, which is why the answer weighs against how many people live there and compares regions rather than declaring a verdict. Some of the difference reflects genuinely sicker populations or referral habits. The value is in surfacing the outliers, regions far above the norm for similar patients, as candidates for a closer look, not as proof on their own.

Does this count the scan, the reading, or both?

It reflects the scanning activity that's billed, and it stays clear about whether a given answer covers the full service or just one part. That matters because the same scan can be billed in pieces, and you want to compare like with like.

How current is the answer?

It stays current on its own. When new information is published, it's already in there, so you're asking against today's picture, not a year-old extract.

Want this answered
for your region?

Whether you're a payer, a health system, or a researcher, I can get you the exact scan-variation answer you care about.