Correction in progress, 2026-09-10. The headline survives a re-derivation and one figure does not. This article was produced by a pipeline that reduced each provider to a single rate using a median which, on an even number of rates, returns the midpoint of the two middle values — a number that need not appear anywhere in the payer's file. We have re-derived this article's figures using observed rates only.
What survives, and it is the claim the article is about. The 4.23x constant between the top and bottom standard imaging schedules re-derives exactly on observed rates — 4.2335 to 4.2341 across the nine codes, a spread of 0.0006, on a different vintage and under a different tier rule. The schedule rates themselves ($81.74, $346.05 and the rest) are shared to the cent by thousands of providers each, so they are observed values by construction and are unaffected. Schedule assignment, not line-by-line negotiation, is unchanged.
What is wrong, and it is overstated. The sentence that "the typical provider lands modestly above Medicare — a median of roughly 1.1x to 1.4x across these codes" was computed with the defective estimator. Re-derived on observed rates, the per-code medians run 0.96x to 1.26x of Medicare, not 1.1x to 1.4x. The direction is down on every code (by 0.1% to 10.5%), and on the lumbar-spine X-ray the typical provider sits below Medicare, not above it — so "modestly above Medicare" is not true across the board.
A third item, separate from the estimator and not fixed by re-running anything. The article names screening mammography as the one code that breaks the 4.23x pattern (at 3.90x) and explains it as the payer pulling its top schedule down on a service federal law requires plans to cover at no cost. That exception is an artifact of our own construction, not a property of the schedule. It appears only because we defined "top" and "bottom" as the highest and lowest of the five most common rates. Under a rule not conditioned on the top five — highest and lowest rate tier holding at least 500 providers — all ten codes obey a single constant (7.094x), mammography included. The paragraph explaining the exception was explaining something that is not there. Tracked as DED-1008, and corrected in the body on 2026-09-17.
Body corrected 2026-09-17. The body no longer carries the two figures above: the Medicare median now reads 0.96× to 1.26×, and the invented mechanism for mammography's apparent exception has been removed and replaced with its actual cause. The original figures are named in this notice on purpose — a correction that deletes what it corrects is not checkable. Nothing else in the article moved.
Pull one insurer's published imaging rates for a single metro and look at what it pays for the most ordinary scan on the list — a head CT without contrast, CPT 70450. In BlueCross BlueShield of Tennessee's in-network home-network file for the Nashville area, the two most common standard rates for that scan are $81.74 and $346.05. Same code, same payer, same city, same component of the service. One is about 20% below Medicare. The other is more than three times Medicare. And each of those two prices is carried, to the cent, by thousands of different providers — about 2,700 at the low rate and about 3,000 at the high one.
That is not the footprint of negotiation. Thousands of unrelated providers do not independently negotiate their way to the identical rate down to the penny. It is the footprint of a fee schedule — a rate table the payer assigns providers onto. We took an earlier look at a single one of these shared schedules in this file. The fuller picture is more useful, and more unforgiving: there isn't one standard schedule. There are several, stacked at wildly different levels, and which one you land on decides almost everything about what you're paid.
Five standard rates, one identical scan
For each of ten common imaging codes, the five most frequent rates are each shared by roughly 1,800 to 3,600 providers. Here is the lowest and highest of those five standard rates for each code, graded against the 2026 Medicare fee schedule:
| Procedure | Bottom schedule | Top schedule | Top ÷ bottom | Bottom ×Medicare | Top ×Medicare |
|---|---|---|---|---|---|
| CT head, w/o | $81.74 | $346.05 | 4.23× | 0.77× | 3.25× |
| MRI brain, w/ & w/o | $254.25 | $1,076.45 | 4.23× | 0.80× | 3.40× |
| CT thorax, w/ | $129.50 | $548.27 | 4.23× | 0.78× | 3.29× |
| X-ray, lumbar spine | $36.01 | $152.47 | 4.23× | 0.67× | 2.85× |
| MRI lumbar spine, w/o | $151.89 | $643.07 | 4.23× | 0.79× | 3.35× |
| MRI joint, lower extremity | $159.80 | $676.55 | 4.23× | 0.78× | 3.31× |
| CT abdomen + pelvis, w/ | $237.04 | $1,003.59 | 4.23× | 0.79× | 3.34× |
| Ultrasound, abdomen | $87.37 | $369.91 | 4.23× | 0.76× | 3.24× |
| Ultrasound, retroperitoneal | $78.63 | $332.91 | 4.23× | 0.74× | 3.15× |
| Screening mammography | $93.94 | $366.00 | 3.90× | 0.74× | 2.90× |
Read the middle column first, because it is the whole story. On nine of the ten codes the top standard schedule is 4.23 times the bottom one — not "about 4×," but 4.234 on every one of them (the individual ratios run only from 4.2335 to 4.2341) across MRI, CT, ultrasound, and plain film alike. A ratio that stable across procedures that have nothing physically in common is not a coincidence and it is not ten separate negotiations landing in the same place. It is one base rate table multiplied by a single number. The payer built a base imaging schedule, and the "top" schedule is that base scaled up by 4.23 — the same 4.23 whether you're pricing a $36 X-ray or a $254 brain MRI.
Which means the real variable isn't any single procedure's rate. It is the level of the schedule you land on — and because the schedules are proportional, the distance from the lowest common schedule to the highest is the same 4.23× on every code, from the cheapest X-ray to the priciest MRI.
Mammography looks like the one code that breaks the pattern — 3.90× instead of 4.23× in the table above. It does not break it, and the exception is ours rather than the payer's. "Top" and "bottom" in that column are the highest and lowest of the five most common rates on each code, and that cut is arbitrary: it is a rank, not a schedule. Under a rule that is not conditioned on the top five — the highest and lowest rate tier holding at least 500 providers — all ten codes obey a single constant of 7.094× (individual ratios 7.0940 to 7.0948), mammography included. That second constant is not manufactured by the 500-provider threshold either: the number of qualifying tiers differs by code (18, 19, 20 and 22) and the ratio holds regardless.
The two numbers are not in competition, and each one belongs to its own rule: 4.23× is the span of the five most common rates; 7.094× is the span of every tier holding at least 500 providers. The first is the one this article is built on, and it is the narrower, more conservative claim.
We originally explained mammography's 3.90× as the payer pulling its top schedule down on the one service federal law requires plans to cover at no cost to the patient. That was a mechanism invented to account for an artifact of our own boundary, and we have removed it. The finding is stronger without it: across the whole span of this schedule family, every one of the ten codes is an exact scalar multiple of the others, with no exception to explain.
Four schedules at or below Medicare, one at triple
Now read the two right-hand columns. The bottom standard schedule pays 0.67× to 0.80× of Medicare — below the government rate on every single code, and only two-thirds of it on X-ray. The three middle schedules (not shown) sit just above that, still landing between roughly two-thirds of Medicare and Medicare itself on every code. Then the top schedule jumps to 2.85× to 3.40× Medicare.
These five rates are the most common, but they are not the whole file: together they cover only about a third of the providers listed on each code. The rest sit on rates in between — and re-derived on observed rates, the typical provider's median runs 0.96× to 1.26× of Medicare across these codes. That is not "modestly above Medicare" across the board: on the lumbar-spine X-ray the median provider sits below the government rate. So this isn't a claim that every provider sits at one of two extremes. It is that the payer's dominant, penny-identical schedules span a 4.23× range on the identical scan, and where a provider falls in that range is set by which schedule it's placed on — not by how hard it negotiated line by line. The "average commercial imaging rate" that anchors most benchmarking blends all of it into one number and hides the schedule underneath.
The multiple against Medicare is not fixed either
Here is the trap for anyone who models commercial imaging as a flat markup over Medicare. The top schedule is a constant 4.23× the bottom schedule — but it is a moving multiple of Medicare: 2.85× on the lumbar X-ray, 3.40× on the brain MRI, 2.90× on the mammogram. The payer's own internal relativities — how it prices an MRI relative to an X-ray — are not the same as Medicare's. So the same top schedule looks like a 2.85× markup on one line and a 3.40× markup on the next.
Both of these are true at the same time, and they are the two facts a finance team has to hold together:
- Between the payer's own schedules, the ratio is fixed. The distance from the bottom schedule to the top is the same 4.23× on every code.
- Against Medicare, the ratio is not fixed. It swings by nearly a full turn depending only on which scan you're pricing.
A rule of thumb like "commercial pays about 3× Medicare for imaging" will overstate what you should be paid on X-ray and mammography and understate it on the high-dollar MRIs — while completely hiding that the payer's most common low schedules pay below Medicare and its top schedule pays roughly triple it.
How to read your own rate
For a provider org or the consultant grading its contracts, the practical lesson is short and a little uncomfortable:
- Your imaging rate is an assignment, not an achievement. When thousands of providers share your exact rate to the penny, you were placed on a schedule, not rewarded for a negotiation. The first question isn't "is my MRI rate good?" It's "which of this payer's schedules am I on, and what does the next one up pay?"
- The prize is a multiplier, not a line item. Because the schedules are proportional, the gap between the bottom schedule and the top is the same multiple on every code — 4.23× in this file. That fixed step, not any single rate, is the size of the conversation worth having.
- Never benchmark against one anchor. "We're at 1.2× Medicare" and "we're at 0.3× the top schedule" can be the same rate. The Medicare comparison tells you how high the market sits off the floor; the schedule comparison tells you where you rank inside a table the payer built. You need both to know which conversation to have — and you need it code by code, because the flat multiple is a fiction on both axes.
None of this required inside information. It is one payer's public transparency file, read against the public Medicare fee schedule, for ten common imaging codes in one city. The prices were always there. The only work was refusing to average them into a single number that hides the schedule underneath.
Methodology: rates are the negotiated global (undivided professional + technical) amounts for the ten CPT codes shown, drawn from BlueCross BlueShield of Tennessee's in-network home-network machine-readable file for the Nashville area (series 890, vintage 2026-07-27), compared component-for-component (global against global). "Standard schedule" rates are the most frequent negotiated rates on each code — each shared, to the cent, by roughly 1,800–3,600 distinct provider NPIs after expanding the file's provider-reference groups; "bottom" and "top" are the lowest and highest of the five most common. Those five most-common rates together cover roughly a third of the providers listed on each code; the remainder sit on intermediate rates, and the top-to-bottom ratio is a property of the schedules themselves, not of how many providers occupy them. "×Medicare" is the ratio to the 2026 national Medicare Physician Fee Schedule global rate for the code (non-facility total RVU × the 2026 conversion factor of $33.4009, released December 2025). No provider or customer names are used. Every figure is reproducible from the source file and the published fee schedule.