CMS Auditing Series: Identifying High Risk Hospitals

As the Centers for Medicare and Medicaid Services (CMS) continue fraud and abuse enforcement efforts, new technologies are being utilized to identify hospitals at “high risk” for potential fraud and abuse.1 This new methodology may allow CMS to transition from “pay and chase” strategies to early detection of fraudulent claims, before payments are made, improving efficiency for the 4.5 million claims CMS reimburses daily.2 On July 1, 2011, CMS began using predictive modeling software to detect potential Medicare fraud and abuse.3 In part three of the CMS Auditing Series, this article examines the new CMS methodology for prospective identification of high risk hospitals.

Currently, CMS audits utilize “pay and chase” strategies. After a claim is submitted and CMS has provided reimbursement (aka “pay”), providers found guilty of fraud are “chased” down for remunerations.4 This strategy essentially creates a situation where CMS can only fight fraud and abuse after the improper payment.5 Since 2010, CMS has aimed to move away from the inefficiency of “pay and chase” auditing through the fraud and abuse provisions of the Patient Protection and Affordable Care Act.6 The predictive modeling software should allow CMS to identify fraudulent claims before payments are disbursed.7

In order to provide proper support during the implementation and initial usage of predictive modeling technologies, CMS has contracted with Northrup Grumman, a leading global security company, in a $77 million multi-year contract.8 The predictive modeling software CMS uses is similar to software used by credit card companies when detecting and preventing potential fraudulent claims.9  The software sorts data from claims and examines factors such as: “(1) ability of the beneficiary to receive the billed services; (2) geographic distance between health care provider and beneficiary, and (3) perceived likelihood that the services were provided.” The software then generates a risk score for each claim.  Once a risk score reaches a pre-established level, an alert is produced for the federal investigators, who then review the claim prior to sending payment.10 Additionally, CMS plans to establish trends between high and low risk hospitals by examining internal compliance policies and procedures, as well as interviewing and / or surveying hospital leadership and compliance officers, to add contextual information to any possible trends.11

Through the use of real time data and algorithms, predictive modeling software will allow CMS to expand its detection capabilities by identifying potential fraud and abuse.13 At a Philadelphia healthcare fraud summit, Kathleen Sebelius stated, “Medical data [is] in lots of different pots…We wanted to be able to spot the doctor claiming to be in six cities billing for the same procedure on the 16th of June.”12

Although predictive modeling may make fraud and abuse detection more efficient, provider reimbursements may slow high risk fraudulent claims. Under the 2010 Small Business Jobs Act, HHS may waive requirements that claims be paid within 30 days if the claim reaches the alert threshold. Payment may be delayed until CMS can confirm the claim is valid.13 Overall, providers should expect more questions and examination before reimbursement.14

With the implementation of preventative modeling technology, providers may need to take additional precautions to ensure they are not inadvertently committing Medicare fraud.  Providers may choose to reexamine documentation and billing practices to help make sure they can properly respond to any potential CMS requests.15  In the final article of the CMS Auditing Series, HC Topics will examine the possible penalties associated with fraud and abuse claims uncovered by CMS audits.

“Inspector General Plans Stepped-Up Oversight” By Joe Carlson, Modern Healthcare, October 5, 2011, (Accessed 10/17/2011).

“Medicare Turns to Tech to Get Ahead of Fraud” HomeCare, June 20, 2011, reimburse/predictive-modeling-technology-20110620/ (Accessed 10/12/2011); “HHS Employs Predictive Modeling to Fight Medicare Fraud” By Sean Reilly, June 20, 2011, (Accessed 10/12/2011);  “CMS to Use Predictive Modeling Tools to Detect Medicare Fraud” iHealthBeat, June 17, 2011, (Accessed 10/12/2011).

“New Technology to Help Fight Medicare Fraud” Centers for Medicare and Medicaid Services, June 17, 2011, (Accessed 10/12/2011).

“Is There A Statistician In The House?” By Allyson Jones Labban, Smith Moore Leatherwood, Health Care Law Note, July 2011.

“From ‘Pay and Chase’ to ‘Catch and Keep:’ CMS to Introduce Anti-Fraud Predictive Modeling” Bradley Arant Boult Cummings LLP, June 28, 2011, (Accessed 11/17/2011).

“Pay and Chase’ No More: CMS Begins Implementing Health Reform’s Provider Enrollment Provisions” Bass, Berry & Sims, PLC, November 12, 2010, /files/Publication/c47a0998-bf2c-47e8-8f2a09f58a8c99a4/ Presentation/PublicationAttachment /239799d4-fb0f-426e-a2e2-195d7673a0bd/HealthReformImpact-12November2010.pdf (Accessed 11/17/2011).

Bradley, “From ‘Pay and Chase’ to ‘Catch and Keep,’” 2011.

“Medicare Hires Northrup Grumman For Predictive Modeling Anti-Fraud Project” Kaiser Health News, June 20, 2011, medicare-anti-fraud.asp (Accessed 10/12/2011); CMS, “New Technology to Help Fight Medicare Fraud,” 2011.

Kaiser, “Medicare Hires Northrup Grumman For Predictive Modeling Anti-Fraud Project,” 2011; CMS, “New Technology to Help Fight Medicare Fraud,” 2011; “Reducing Improper Payments, Fighting Fraud, and Curbing Waste and Abuse Under the Affordable Care Act,”, October 12, 2011, (Accessed 10/12/2011).

iHealthBeat, “CMS to Use Predictive Modeling Tools to Detect Medicare Fraud” 2011.

“HHS OIG Work Plan: Fiscal Year 2012” Office of the Inspector General, 2011, p. 1-5.

CMS, “New Technology to Help Fight Medicare Fraud,” 2011; Reilly, “HHS Employs Predictive Modeling to Fight Medicare Fraud,” 2011.

HomeCare, “Medicare Turns to Tech to Get Ahead of Fraud,” 2011.

Bradley, “From ‘Pay and Chase’ to ‘Catch and Keep,’” 2011.

Labban, “Is There A Statistician In The House?” 2011.

Healthcare Valuation Banner Advisor's Guide to Healthcare Banner Accountable Care Organizations Banner