When Medicare Meets Novel Hardware:
CardioNet, iRhythm, and the Recurring Problem of Pricing Proprietary Inputs
Executive Summary:
CardioNet 2010 and iRhythm 2020 show a recurring Medicare pricing problem: breakthrough monitoring technologies often rely on proprietary, vertically integrated components that do not have clear market-clearing prices.
In both cases, CMS hesitated to set national RVU-based payment when the key direct practice-expense input was hard to validate—first a telemetry monitor and later the SD339 Zio patch. The result was delayed national pricing, reliance on (unstable!) contractor rates, and prolonged disputes over invoices and the obscure corners of CMS PE methodology. The broader lesson is that Medicare’s standardized fee schedule fits commoditized inputs but crashes against one-off, service-embedded technologies.
###
CardioNet vintage 2007-2008 remains one of the best examples of how a seemingly booming medtech story can run headlong into the internal logic of Medicare pricing. On the surface, the company looked almost ideal. It had an FDA-cleared product, rising sales, growing physician adoption, Medicare coverage, and eventually dedicated Category I CPT codes for mobile cardiovascular telemetry. Its patient-worn sensor and handheld monitor sent ECG data to a remote monitoring center, and its business ramp looked impressive: revenue rose from $73.0 million in 2007 to $120.5 million in 2008, with MCOT revenue of $100.2 million, or 83.2% of total revenue. Its 2008 IPO raised about $81 million.[1]
That is what made the reimbursement shock so dramatic. For the technical component, CPT 93229, Medicare did not initially set a uniform national Part B rate. Instead, the code was carrier-priced, and Highmark Medicare Services posted a rate of $1,123 in early 2009.[2]
But carrier local pricing may be unstable. Next the floor dropped. Highmark cut the payment to $754, effective September 1, 2009, and CardioNet described the change as roughly a 33% reduction with material adverse effects on the business.[3] Because commercial payors were already looking to Medicare as a benchmark, the Medicare cut threatened to spread beyond Medicare itself.
Bruce Quinn has a recollection that helps explain why the new number became so sticky. Highmark’s local medical director had apparently noted that CMS had already cross-walked the service in the hospital outpatient setting to an ambulatory payment amount of about $753; once that number existed elsewhere inside Medicare, it became easy to defend the local carrier rate as “Medicare-like,” even if it was only a rough analogy (B. Quinn, pers. comm.). That captures something important. These pricing crises do not always begin with an explicit policy manifesto. Sometimes they begin because a contractor spots a nearby number elsewhere in Medicare’s own machinery and decides it is good enough.
CardioNet’s core problem was that national CPT recognition did not automatically produce a favorable national fee schedule price. CMS did not establish a national payment amount for 93229 for 2010, leaving the code on contractor pricing for another year.[4]
When CMS finally addressed the code in the 2011 physician fee schedule process, the real dispute came into view. Commenters wanted CMS to treat a broader centralized monitoring infrastructure as direct practice expense, but CMS refused. CMS said that, for 93229, the only equipment item appropriately treated as a direct PE input was the cardiac telemetry monitoring device worn by the patient; the software and hardware, workstation, webserver, and call-recording system were treated as indirect costs instead.[5]
That leads to the deeper lesson in the CardioNet story. The valuation problem was not just that the service was new. The problem was that its most important physical input was a proprietary device inside a vertically integrated business model. As Bruce Quinn has noted, CMS had reason to be skeptical if the “price” of the telemetry monitor was really an internal company transfer value rather than an open-market price (B. Quinn, pers. comm.). That is where the Medicare PE system becomes uncomfortable. It handles standardized, externally priced inputs reasonably well: an MRI scanner, a microscope, a reagent, a surgical supply with a visible invoice trail. It is much less comfortable when the crucial input is a one-off novel device used inside a service company that largely makes, ships, processes, and interprets the product itself. In that setting, CMS is being asked to trust a price that may look more like an internal accounting construct than a market-clearing price. (In this case, the invoice price exceeded $20,000).
Not surprisingly, the result was that CMS finally set a national price for 93229 in 2011 at only about $739, close to the prevailing contractor price rather than a return to the earlier economics.[3][5] In other words, CardioNet got national pricing, but only after the premium economics had already been broken:
In the final price-setting, CMS granted about $23,000 as the invoice device price, with a 3-year life, and allotted about $1118 per use (nominal dollars before budget neutrality discounts).
Same Problem, Ten Years Later
What makes this story especially interesting is that the same basic problem reappeared more than a decade later with iRhythm Zio. The code was different, the form factor was different, but the institutional problem was strikingly similar. For CPT 93247, the key direct PE driver was not a capital monitor but a disposable supply, SD339, the extended external ECG patch. CMS explained in the 2023 final rule that it had struggled for multiple years to determine an acceptable national price for SD339. At first CMS rejected certain invoices from the clinical-study marketplace because it “typically require[s] an invoice representative of commercial market pricing” to establish a national price for a new supply or equipment item.[8] CMS initially used a proxy supply price, then in the CY 2021 final rule left 93241, 93243, 93245, and 93247 on contractor pricing because of conflicting information and the need for more pricing data.[8]
That did not end the matter. In the CY 2022 cycle, CMS received ten invoices and finalized an updated SD339 price of $200.15, with invoice values ranging from $179.80 to $241.99, but still declined to establish national pricing for the code family for 2022.[8] CMS said it wanted broader feedback before nationalizing payment.[8]
Then, for CY 2023, CMS received additional invoices and settled on $315 as a price for SD339, altough the net payment was around $250 after budget neutrality adjustments.[8]
CMS’s path is revealing because it sounds so much like the old CardioNet problem. Commenters argued that the independent diagnostic testing facility model did not fit comfortably within the traditional physician-office PE methodology, and pointed to costs tied to R&D, AI, and software-as-a-service. CMS’s response was measured but unmistakable: yes, more comprehensive PE data might be useful, but the current methodology is the current methodology, and pricing still turns on the sort of invoice-based evidence CMS can use within that framework.[8] What CardioNet had illustrated in 2010, iRhythm replayed in 2021–2022: when the code depends on a novel, proprietary, vertically integrated input, Medicare hesitates.
Putting the Puzzle Together
That is why the iRhythm episode should be read not as an isolated dispute, but as a mirror of the CardioNet experience. The physical object changed. In CardioNet, the troublesome PE input was the telemetry monitor and its surrounding infrastructure. In iRhythm, it was the disposable patch, SD339. But the logic was the same. CMS was being asked to assign national RVUs to a service whose key input did not look like a standard, commoditized, transparently priced medical article. Instead, it looked like a novel component embedded in an integrated service company. The agency’s reflex in both cases was similar: ask for invoices, reject weak proxies or nonrepresentative evidence, strip out indirect and enterprise-level cost arguments, and delay national pricing until the direct PE input had a price it trusted.
This is more than a coding curiosity. It reflects a recurring collision in Medicare policy between innovative technologies and standardized pricing systems. Medicare’s PE methodology was built to handle routine, decomposable, office-based economics. It works best when the service relies on inputs with relatively visible market prices. It works less elegantly when innovation comes packaged as a vertically integrated service platform with one-off physical components, proprietary software, and internalized logistics. That is why a Medtronic device sold into a distributed market may fit the system more readily than a CardioNet or iRhythm service, where manufacturing, monitoring, processing, and interpretation may all sit under one corporate roof.
Seen that way, CardioNet and iRhythm are not just two remote-cardiology episodes separated by a decade. They are variations on the same Medicare puzzle. Both involved highly innovative monitoring technologies. Both involved direct PE inputs that looked unusually expensive or hard to verify. Both forced CMS to decide whether a novel physical component had a real market price or only an asserted one. And in both cases CMS moved slowly, cautiously, and skeptically. The lesson is not that Medicare opposes innovation. It is that the most innovative products often fit worst into fee schedules designed around standardized, openly priced inputs. CardioNet showed that in 2010. iRhythm showed, all over again, that the lesson still applied in 2021–2022.
References
[1] SEC registration statement for CardioNet, showing 2007–2008 revenue growth and IPO context.
https://www.sec.gov/Archives/edgar/data/1113784/000104746908002538/a2179762zs-1a.htm
[2] CardioNet announcement regarding posted Highmark reimbursement rate for CPT 93229.
https://www.biospace.com/cardionet-inc-announces-that-b-highmark-medicare-services-b-posts-reimbursement-rate-for-cpt-code-93229-mobile-cardiovascular-telemetry-services
[3] CardioNet Form 10-K discussing the reduction of 93229 reimbursement from $1,123.07 to $754 and later national pricing at $739.
https://www.sec.gov/Archives/edgar/data/1113784/000104746912001348/a2207476z10-k.htm
[4] CardioNet update on CMS not establishing a national reimbursement rate for 2010 for mobile cardiovascular telemetry.
https://www.dicardiology.com/content/cardionet-provides-update-mobile-cardiovascular-telemetry-reimbursement
[5] CY 2011 Physician Fee Schedule final rule, discussing CPT 93229 and direct versus indirect practice expense inputs.
https://www.federalregister.gov/documents/2010/11/29/2010-27969/medicare-program-payment-policies-under-the-physician-fee-schedule-and-other-revisions-to-part-b-for
[6] Philadelphia Inquirer article on the market shock around CardioNet reimbursement concerns.
https://www.inquirer.com/philly/blogs/inq-phillydeals/Was_analyst_forced_out_for_being_right_about_CardioNet.html
[7] Philips announcement of the BioTelemetry acquisition, valuing the company at about $2.8 billion.
https://www.philips.com/a-w/about/news/archive/standard/news/press/2020/20201218-philips-to-become-a-global-leader-in-patient-care-management-solutions-for-the-hospital-and-the-home-through-the-acquisition-of-biotelemetry-inc.html
[8] CY 2023 Physician Fee Schedule final rule, discussing SD339 invoice evidence, contractor pricing in 2021–2022, and final national pricing in 2023 for 93241/93243/93245/93247.
https://www.federalregister.gov/documents/2022/11/18/2022-23873/medicare-and-medicaid-programs-cy-2023-payment-policies-under-the-physician-fee-schedule-and-other
