AI Corner
Chat GPT 5.4
(Essay is generated by AI and discusses hypotheticals) (Prompt at bottom)
(See Sidebar, Fraud in Genomics Industry)
The Centers for Medicare & Medicaid Services (CMS) has long struggled with large-scale fraud schemes involving suppliers and laboratories billing for implausible volumes of services, particularly in areas such as durable medical equipment and genetic testing. One well-known example in the laboratory sector involved certain high-complexity sequencing codes—such as CPT 81408 for full gene sequencing—being billed in medically implausible quantities by questionable laboratories, particularly concentrated in states like Texas and Florida. Medicare spending on these codes briefly reached hundreds of millions of dollars per year before collapsing once medical necessity controls and enforcement actions were implemented. Against this backdrop, CMS has launched a broad anti-fraud initiative known as CRUSH (Comprehensive Regulations to Uncover Suspicious Healthcare), intended to shift the program away from the traditional “pay-and-chase” model toward proactive fraud prevention using analytics, screening, and real-time enforcement tools. The initiative explicitly seeks stakeholder input on how CMS can strengthen its program integrity toolkit across Medicare, Medicaid, and related programs.
The motivation for this initiative is understandable. CMS reports that it has already suspended billions of dollars in suspected fraudulent payments using advanced analytics and has revoked the billing privileges of thousands of providers engaged in improper activity. At the same time, the agency has identified laboratory testing—particularly molecular and genetic tests—as an area of growing financial exposure. Although genetic tests represent only a small share of laboratory test volume, they account for a large proportion of Medicare laboratory spending. As a result, molecular diagnostics has become a focal point for program integrity discussions.
However, the tools CMS uses to combat fraud typically operate system-wide rather than targeting only the bad actors, which creates a significant risk of unintended consequences for legitimate laboratories. Large and well-established, well-run precision medicine companies could at least hypothetically experience what might be called “bystander injury” if aggressive anti-fraud tools are deployed broadly. Several mechanisms within the CRUSH initiative illustrate how this could occur.
One major risk is the expansion of pre-payment review and algorithmic claims screening. CMS is attempting to move away from retrospective audits toward real-time fraud detection based on data analytics and predictive modeling. In practice, such systems often identify statistical anomalies rather than fraud itself. Molecular diagnostics laboratories are particularly vulnerable to this approach because their tests are expensive and specialized. Rapid adoption of a new technology, regional concentration of specific cancers, or rapid growth of a successful test could all trigger automated fraud alerts. If these triggers lead to pre-payment claim holds or intensive review, a financial impact on laboratories occur.
A second area of concern involves medical record documentation requirements during audits. Program integrity enforcement frequently relies on post-payment review of medical records and the authority to suspend payments when credible allegations of fraud exist. Yet molecular diagnostic laboratories are typically reference laboratories that do not maintain complete clinical records for patients. They receive specimens from oncologists or hospitals and often possess only partial documentation—such as a pathology report or requisition form. If CMS or its contractors require comprehensive oncology records, imaging studies, treatment plans, and physician notes during audits, laboratories may struggle to obtain these materials from ordering providers. The result could be audit failures even when the underlying tests were medically appropriate. This mismatch between audit expectations and laboratory workflows is already a well-recognized friction point in some coverage programs, including MolDx.
Another potential unintended consequence involves the use of artificial intelligence and machine learning tools to detect fraud patterns. CMS is actively exploring AI systems capable of reviewing large volumes of claims data and assisting in medical record coding analysis. While such technologies promise efficiency, they also carry well-known limitations. Machine learning models typically flag unusual statistical patterns, but unusual patterns do not necessarily indicate fraudulent activity. Laboratories operating at the technological frontier of medicine—introducing novel genomic assays or rapidly expanding clinical adoption—may naturally produce unusual utilization patterns. If automated systems generate fraud alerts based on these patterns, legitimate laboratories could be subjected to cascades of audits, denials, or payment delays.
A related risk is the potential expansion of prior authorization requirements for high-cost laboratory tests. Although CRUSH does not explicitly mandate prior authorization, anti-fraud initiatives frequently evolve in that direction as policymakers search for ways to control spending on expensive services. If Medicare were to require prior authorization for molecular tests above a certain price threshold—such as $1,000 or $2,000—the operational burden on laboratories and oncology clinics could increase dramatically. Laboratories would need to build large administrative teams to manage authorization requests, physicians would face additional workflow hurdles, and patient care could be delayed while approvals are obtained. For tests used in time-sensitive clinical decision making, such as therapy selection or recurrence monitoring, such delays could have real clinical consequences.
Another area where overreach could occur involves provider risk classification and enrollment controls. CMS is exploring whether certain provider types should be designated as higher-risk categories for program integrity purposes. Given the high cost and rapid growth of genomic testing, molecular laboratories could easily fall into such a classification. If that occurs, laboratories could face stricter enrollment requirements, enhanced screening procedures, fingerprinting and background checks for additional personnel, higher surety bond requirements, and more frequent revalidation of Medicare enrollment. While these measures may be justified for certain high-risk supplier categories, they could impose significant compliance burdens on legitimate diagnostics companies.
Payment suspension authority also poses a potential hazard for legitimate laboratories. CMS already has the authority to suspend Medicare payments when there is a credible allegation of fraud or reliable indication of overpayment. In some past cases across the healthcare sector, this authority has been applied broadly during investigations. If statistical anomalies or algorithmic signals trigger such suspensions, laboratories could experience abrupt interruptions in cash flow while disputes are resolved. For companies performing high-volume oncology testing, where Medicare revenue can reach tens or hundreds of millions of dollars annually, even temporary payment suspensions could have serious financial implications.
Finally, the CRUSH initiative raises questions about ownership transparency and identity verification requirements for Medicare-enrolled entities. CMS has expressed concern that some fraud schemes involve complex ownership structures or foreign actors who operate outside the reach of U.S. law enforcement. The agency is therefore considering enhanced identity verification and even citizenship or residency requirements for owners of Medicare providers. While aimed at fraudulent shell companies, such rules could also affect legitimate diagnostics firms that rely on global venture capital or international investment. Many biotechnology and diagnostics companies operate with multinational investor bases, and additional ownership restrictions could complicate financing structures.
In summary, the CRUSH initiative addresses a genuine problem. Fraud schemes involving phantom laboratories, telemarketing genetic testing scams, and abusive billing practices have imposed significant costs on the Medicare program. However, the enforcement mechanisms used to combat fraud—data analytics, pre-payment review, enrollment screening, documentation audits, and ownership restrictions—often operate broadly across entire sectors rather than targeting only the most egregious offenders. For legitimate molecular diagnostics laboratories, the resulting “bystander effects” could include increased audit risk, expanded documentation requirements, algorithmic payment holds, administrative burdens from prior authorization, and stricter enrollment controls. Policymakers therefore face a delicate challenge: designing fraud-prevention systems that successfully eliminate clearly fraudulent actors without inadvertently creating substantial compliance friction for the very laboratories responsible for advancing precision medicine.
SIDEBAR - Lab Genetics Fraud ( source here)
Fraudulent uncontrolled payments circa a billion dollars for 81408. This chart shows 81408; the same labs billed much additional money under adjacent codes 81407, 81406. The outpouring of bogus funds continued til 2022; I wrote about it from fall 2020 onward. Click to enlarge.
One of the biggest Medicare codes in 2024 was "Other pathogen 87798," at $450M, an explosion of +200M in 24 months.
##
Prompt
CMS has bad problems with fraud - fraudulent DME, or fraudulent labs in Texas and Florida billing huge amounts of dollars on medically unbelievable codes like 81408, long gene full sequencing, which reached $300M per year in CMS and then plummeted to zero with medical necessity controls. CMS is rolling on a big program CRUSH as in PDFs. This SHOULD be designed to attack only obviously fraudulent nonsensicla labs in TX and FL. but what are the risks of problems and overreach affecting good bona fide labs like say natera, guardant, caris, etc. One risk is random audits which require lots of layers of medical records the labs don't have. Another risk is punshing prior auth with AI on all lab tests over $1000. These are nightmare scenarios for legit labs. plesae discuss risks for legit labs (unintended consequences or bystander injuries so to speak)
https://chatgpt.com/c/69b06aec-4f70-83e8-a952-0c5ef76d919e
