Fraud

From Clinfowiki
Jump to: navigation, search

Question

Are predictive modeling techniques enough to assist the government’s effort in preventing Medicare fraud?

Introduction

In February 2011, “A trio of doctors altered the diagnoses and medications of thousands of patients to make it look like they qualified for purported group therapy sessions at American Therapeutic's chain of South Florida clinics, costing the taxpayer-funded Medicare program hundreds of millions of dollars.” Two weeks later, Federal authorities announced a bust of unprecedented proportions-- 111 doctors, nurses, and therapists across the nation, arrested for stealing from Medicare. In one case, a Detroit podiatrist billed Medicare about $700,000 for procedures amounting to little more clipping patients toenails.

The estimate of fraud in healthcare spending is estimated to be 3 percent to 5 percent, according to the National Healthcare Anti-Fraud Association.# # The amount of money lost to fraud and abuse for 2006 was estimated to be $48 billion to $80 billion. With annual healthcare spending increasing annually, money lost to fraud is only expected to increase. The Hillestad report which is often cited for support of the advancement of HIT estimated that after a 15-year roll out period, assuming 90% adoption, the EHR could save the health care system $81 billion annually.# $81 billion after 15 years and 90% adoption.# Isn’t it faster to stop the fraud? Maybe faster, but not easier.

Methods

Recently the government, with funds from ARRA, is claiming they have advancements in predicative modeling analytics to help identify potential fraudulent claims before being paid.# Using predictive modeling techniques with large amounts of data shared through interoperable exchange will give RHIO’s and HIE’s the necessary data for advanced analytics. “Interoperability enables the aggregation of rich clinical and financial databases to which advanced analytic techniques are applied to detect patterns of fraud. It is when we have a comprehensive NHIN with interoperable RHIOs or HIEs, that the ongoing evaluation of data and the ability to identify fraud patterns will be achieved. “

Billing systems were created with the intention of honest providers billing for appropriate services with the intent to get paid.# To beat all the industry’s current defenses, all they have to do is bill correctly. If they do that correctly, they are free to lie.#

Limitations

There is no standard predictive model that can be utilized across healthcare settings; primarily because different clinical settings have different predictive modeling techniques.

Conclusion

Predicative modeling techniques while an excellent start at preventing fraud are not standardized enough in the healthcare industry to detect and prevent fraud.# If we are waiting for RHIO’s/HIE’s to supply this rich data it may be too late to prevent the fraud.

Related Articles

Works cited

  1. Feds bust over 100 doctors for Medicare fraud, many from Florida Feb 17, 2011 at 7:09 PM America/New York
  2. Modern Healthcare Supplement, By the Numbers, Crain publication, December 19, 2005, page 8. [1]
  3. Hillestad report, Health Affairs,2005
  4. RHIOs—Build in Healthcare Fraud Management from the Beginning, Susan P Hanson, Bonnie S Cassidy, Journal of Healthcare Information Management — Vol. 20, No. 3
  5. Commentary: A Criminal’s Dream: Automated Payment Systems, Designed for Honest Providers, Are Easy Targets for Fraud by Malcolm K. Sparrow

published in Modern Healthcare, October 6, 1997.

Submitted by Concetta Pryor