Abstract (75 words). Medicare physician payment looks straightforward: work RVUs, malpractice RVUs, and practice expense RVUs. But practice expense contains a quiet and consequential twist. CMS begins with visible direct inputs—clinical labor, supplies, and equipment—and then adds indirect overhead, yet the final PE RVU often ends up materially lower than those listed costs suggest. The reason is a set of internal scaling and budget-neutrality adjustments. This essay explains that hidden shrinkage and why it affects nearly every CPT code.
The Hidden Shrinkage Inside Medicare Practice Expense RVUs
Everyone in Medicare policy eventually learns the basic catechism of the physician fee schedule. A payment rate is built from work RVUs, practice expense RVUs, and malpractice RVUs. Work is intuitive enough: it reflects physician time, effort, skill, and stress. Malpractice is the smallest and most straightforward component. Practice expense, by contrast, looks simple at first and then becomes strangely elusive. It appears to rest on three visible building blocks—clinical labor, supplies, and equipment—and yet, when you trace many codes from those listed inputs to the final practice expense RVU, the answer often comes out materially lower than a naïve dollar-for-dollar calculation would suggest. That is not a mistake. It is a feature of the system. (1)(2)
CMS and MedPAC both describe the fee schedule as a relative value system, not a literal cost reimbursement system. MedPAC summarizes the three RVU components and notes that practice expense reflects the costs of maintaining a practice, including supplies, equipment, and nonphysician staff. CMS’s practice expense methodology report is even more explicit: direct PE for each procedure is built from clinical staff, equipment, and supplies, and those direct inputs are then run through a multi-step allocation method. (1)(2)
That is where the mystery begins. Suppose a service has accepted direct inputs of $20 clinical labor, $20 supplies, and $20 equipment, for a visible total of $60. If one assumes, just for convenience, a conversion factor of $30, the simple expectation would be 2.0 RVUs. But that is not how Medicare practice expense works. CMS does not simply total the direct inputs and divide by the conversion factor. Instead, it first creates a direct PE pool for the entire fee schedule, then rescales the direct inputs across all services so that the aggregate direct practice expense fits that pool. In CMS’s own step-by-step description, Step 4 is to calculate a “direct PE scaling adjustment” and apply it to each service’s direct costs. (2)
That scaling step is easy to miss, but it is fundamental. CMS says the direct inputs are scaled so that aggregate direct PE RVUs equal the direct PE RVU pool, while preserving the relative relationship between services. In other words, Medicare is not promising that a service with $60 of listed direct inputs will emerge with PE RVUs equivalent to $60. It is promising only that the service will keep its relative position within a nationally budgeted practice expense system. (2)
Then comes indirect practice expense, which makes the story even less transparent. CMS allocates indirect PE using specialty survey data and service-level allocators based on direct PE RVUs, clinical labor PE RVUs, and work RVUs. For most services, the indirect allocator formula includes both direct PE and work RVUs. For some technical or low-work services, clinical labor can substitute for work when it is greater. CMS also applies an indirect PE adjustment so that aggregate indirect allocations do not exceed the available indirect PE pool. So the visible trio of labor, supplies, and equipment is only the beginning. The code is also being marked up for overhead through a separate indirect-expense methodology. (2)
And yet—this is the fascinating part—even after that indirect-expense upweighting, the final PE RVU can still be lower than the visible direct-input total would suggest. CMS’s final step is bluntly labeled: “Add the direct PE RVUs… to the indirect PE RVUs… and apply the final PE budget neutrality (BN) adjustment.” CMS says this final PE budget-neutrality adjustment is calculated by comparing the sum of direct and indirect PE RVUs to proposed aggregate work RVUs scaled by the current PE-to-work ratio. That final step is the quiet shrinkage mechanism. It is not usually described in plain English as a “deflator,” but that is how it behaves. (2)
So return to the toy example. A code has $60 in direct practice expense inputs. One might expect that to become 2.0 PE RVUs at a $30 conversion factor. In reality, Medicare first rescales the direct inputs, then allocates indirect overhead, then applies a final PE budget-neutrality adjustment. The result might be something like 1.5 PE RVUs, equivalent to $45, even though that $45 already includes some share of indirect expense. That is the paradox that makes the policy so hard to see: the code is first marked up for overhead and then, at the system level, compressed back down. (2)
This happens because practice expense is governed by pools, not by isolated code-level arithmetic. CMS calculates aggregate direct and indirect PE pools and makes the code-level values fit those pools. The agency is preserving internal relativity across the fee schedule, not reimbursing each service’s nominal listed inputs dollar for dollar. Put differently, PE RVUs are relative values constrained by a national budget framework, not a pass-through invoice for labor, tubing, gauze, and amortized equipment. (2)
One reason this topic is so confusing is that there are actually two different kinds of budget neutrality in the physician fee schedule. MedPAC explains that when work RVUs change enough to trigger the statutory threshold, CMS generally makes those changes budget neutral through the conversion factor. But changes in practice expense RVUs do not generally change the conversion factor; instead, they are made budget neutral by adjusting PE RVUs up or down for other services. That is the distinction many people miss. The global conversion-factor controversy and the internal PE compression are related ideas, but they are not the same mechanism. (3)
CMS’s own public documentation can actually deepen the confusion. On the fee-schedule documentation page, CMS still shows a visible “Budget Neutrality Adjustor” for 2007 and 2008, but lists it as N/A for 2009–2025 in the payment formula display. Read casually, that looks as if budget neutrality disappeared. It did not. What changed is that, under MIPPA, CMS stopped applying that older separate work-RVU budget-neutrality adjustor in the same way and instead applied it through the conversion factor. CMS said so explicitly in 2008: MIPPA required that a prior separate budget-neutrality adjustment to work RVUs be applied instead to the conversion factor. That historical change helps explain why policy people can search the CMS files and come away thinking the adjustor vanished, when in fact PE still has its own internal scaling and budget-neutrality machinery. (4)
There is another wrinkle. The PE methodology still depends heavily on old empirical infrastructure. In the 2026 proposed physician fee schedule, CMS said that PE methodology continues to rely primarily on the AMA’s 2008 Physician Practice Information survey, and that updated survey efforts submitted in 2025 were not being adopted for 2026 because of concerns about sample size, response rates, representativeness, and data quality. That means PE values are not only scaled and redistributed; they are also built on an aging empirical foundation. This adds to the sense, familiar to many policy wonks, that practice expense is partly cost-based and partly a highly managed artifact of pool mechanics. (5)
So what should one call the hidden shrinkage? CMS does not give it one tidy popular label. Instead, the effect is spread across the phrases direct PE scaling adjustment, indirect PE adjustment, and final PE budget neutrality adjustment. But in plain English, the phenomenon is simple enough: the listed practice expense inputs for a code are not the final payment value of the technical component. They are raw ingredients that are scaled, allocated, and then normalized to fit the overall PE structure of the physician fee schedule. That is why a service with visible practice expense inputs of $60 can end up with PE RVUs worth only $45. And that is why this seemingly nerdy topic is actually rather interesting: it reveals that Medicare’s technical component is not a direct cost-accounting exercise at all, but a national relativity system hiding behind the reassuring vocabulary of minutes, supplies, and equipment. (2)
Endnotes
(1) MedPAC, Physician and Other Health Professional Services Payment System 2025.
https://www.medpac.gov/wp-content/uploads/2024/10/MedPAC_Payment_Basics_25_Physician_FINAL_SEC.pdf
(2) CMS / RAND, Practice Expense Methodology and Data Collection Research and Analysis: Interim Phase II Report.
https://www.cms.gov/files/document/cy-2021-pfs-practice-expense-methodology-and-data-collection-research-and-analysis-report.pdf
(3) MedPAC, Report to the Congress: Medicare and the Health Care Delivery System, June 2025, Chapter 1.
https://www.medpac.gov/wp-content/uploads/2025/06/Jun25_Ch1_MedPAC_Report_To_Congress_SEC.pdf
(4) CMS, Physician Fee Schedule Search: Documentation and Files.
https://www.cms.gov/medicare/physician-fee-schedule/search/documentation
(5) CMS, Calendar Year 2026 Medicare Physician Fee Schedule Proposed Rule (CMS-1832-P) Fact Sheet.
https://www.cms.gov/newsroom/fact-sheets/calendar-year-cy-2026-medicare-physician-fee-schedule-pfs-proposed-rule-cms-1832-p
