AI Pilot Program Proposed to Detect Medicare Fraud

Artificial Intelligence (AI) is a current hot topic of conversation, no matter what industry you’re in, and durable medical equipment (DME) and diagnostic laboratories are not excluded. In January, a bill was introduced to the U.S. Senate that would assess the technology’s effectiveness in detecting Medicare fraud related to clinical diagnostic laboratory tests and DME. The bill, which has now been assigned to the Committee on Finance, aims to develop and test a pilot program using “a predictive risk-scoring algorithm” to detect and decrease Medicare fraud.¹

As proposed, this bill would amend Title XI of the Social Security Act to allow for the development of the pilot program, which would use the algorithm to detect suspicious activity in claims submitted on behalf of Medicare enrollees for DME and clinical diagnostic laboratory tests. The program would evaluate how effective such an algorithm is in overseeing relevant transactions. Another key element of the proposal is the inclusion of a quick alert system that would ask Medicare patients to confirm or deny the suspicious activity before Medicare reimbursement is released.

Highlights of the Medicare Transaction Fraud Prevention Act

The proposed bill’s text ends with a few important definitions lab leaders should take note of:¹

“Applicable item or service” refers explicitly to clinical diagnostic laboratory tests and an item of DME, as defined in Section 1861(n) of the Social Security Act.

“Relevant transaction” is a claim for payment for said testing and/or equipment.

The pilot program would only apply to Medicare beneficiaries who opt in to receive electronic Medicare Summary Notices and involve transactions for diagnostic lab testing and/or DME. Participants may opt out at any time.

The program will consider the following factors when determining which relevant transactions are deemed suspicious:

• • A lack of prior relationship between the beneficiary and the provider of services.

• • Billing patterns that depart from an acceptable standard.

• • Electronic funds transfer (EFT) changes.

• Changes in ownership to a provider’s practice and/or a supplier’s company.

This is not to say that the above factors are disallowed or that they will immediately trigger suspension of remuneration for services provided; rather, these are the risk factors that the AI program will utilize to trigger a wider investigation of the claims filed for an applicable item or service. The bill’s text states that there will be a human review process of all suspended accounts and transactions.

In addition, the bill proposes that industry representatives will work with the U.S. Department of Health and Human Services (HHS) secretary on the development and implementation of the pilot program, though it does not expressly define what qualifications such representatives will have.

How the pilot program will work

The requirements of the pilot program are as follows:¹

1. A predictive risk-scoring algorithm will be adopted.

2. Prior to implementation of the adopted algorithm:

• • the algorithm will be sufficiently tested, evaluated, and reviewed according to the Blueprint for an AI Bill of Rights issued by the White House Office of Science and Technology Policy in October 2022,²

• • a method for notifying applicable beneficiaries and providers of services and suppliers who may be impacted will be developed, and

• methods for communicating with the Office of the Inspector General of the HHS will be established.

3. For those transactions identified as risky by the algorithm, the HHS secretary will:

(i) determine if the reimbursement should be suspended based on the response received in (ii) below

(ii) offer the beneficiary the opportunity to respond via email or phone call to correct a falsely elevated risk score by accurately updating underlying data, and confirm the transaction:

(iii) if the transaction is suspended:

(a) trigger an automatic alert to the beneficiary,

(b) require that all relevant transactions be sent electronically to beneficiaries and in two-week intervals for three months after the first alert is sent, and

(c) include instructions on how beneficiaries can report suspected fraud to the appropriate law enforcement agency

4. The HHS secretary will assume authority to determine when it is appropriate to terminate and/or issue a new Medicare card to a beneficiary

Potential issues with the bill and its chances of passing into law

As the bill was only first introduced in January, it is hard to determine its likelihood of passing. Senator Mike Braun (R-IN) is the bill’s sponsor, joined by only one co-sponsor: Senator Bill Cassidy (R-LA), who is on the Committee on Finance that the bill has been assigned to. The committee will research, discuss internally, and make changes to the bill before voting on whether or not to send it back to the Senate floor for a formal vote or let it die.

Senator Braun posted a news story on the bill to his website on February 1, 2024 noting that similar relevant technology has been deployed successfully in monitoring credit card fraud and has the potential to save hundreds of millions of dollars in fraudulent reimbursements. Based on information the story cites from the HHS Centers for Medicare & Medicaid Services (CMS) in 2019, Medicare scrutinizes fewer claims than commercial insurers do and processes 99.7 percent of Medicare fee-for-service claims with payment being issued within 17 days without any medical review.³

One big question that the Finance Committee will need to address is the fact that the bill does not authorize funding for the pilot project, and there has been no estimate of its costs.

Another public criticism of the proposal has been that it should not fall on the shoulders of Medicare enrollees to verify the risky transactions identified by the algorithm and that CMS should root out fraud itself. However, Taghi M. Khoshgoftaar, PhD, a researcher who studies AI use to detect Medicare fraud at Florida Atlantic University and who was cited in the article, seemed to support the bill and this approach of involving Medicare enrollees in the protection of their own benefits.³

What might this mean for laboratories and suppliers of medical equipment?

While the bill is in its early stages and may be significantly changed before potentially becoming law, providers and suppliers will want to know how this may impact them and their operations should it eventually pass. One of the best ways to ensure your Medicare claims are correct and thus help avoid suspension or denial of reimbursement is to conduct regular billing audits of your services (see “Can Conducting a Billing Audit Increase My Lab’s Revenue Stream?” in the February 2024 issue of Lab Industry Advisor).

Additionally, since the algorithm will base its risk assessment on certain factors detailed above, maintaining proper records of provider or supplier relationships with beneficiaries, filing claims under acceptable standards, and not making unnecessary changes to your EFT account or the ownership of your provider practice or supplier company can help keep your organization from being flagged as risky.

References:

1. 1. https://www.congress.gov/bill/118th-congress/senate-bill/3630/text

1. 1. https://www.whitehouse.gov/ostp/ai-bill-of-rights/

1. 1. https://www.braun.senate.gov/news/in-the-news/federal-bill-seeks-ai-tools-to-stop-medicare-fraud/