Fraud and the abuse of healthcare services in the U.S. cost an estimated $125-175 billion annually. This represents the second largest component of the $600-850 billion surplus in healthcare spending. Healthcare organizations and government agencies must leverage big-data collections of patient records and financial billing to identify and eliminate system abuses.

Examples of fraud and abuse in healthcare range from:

• Intentional misrepresentation of services that result in higher payments
• Billing of unperformed services
• The deliberate delivery of unnecessary and inappropriate services for the express purpose of receiving the payment

Most of the above-mentioned abusive practices not only add excess cost to the healthcare system, but also can potentially expose patients to health risk.

Fighting Abuse At All Levels

Due to the large dollar amounts and the number of companies and people involved in the healthcare system there is a huge potential for abusive behaviors at all levels. Patients sometimes knowingly participate in fraud by billing for services never provided. For example, drug addicted patients have a history of misusing Medicaid by enrolling for services in multiple states. Payers can easily misrepresent the cost of care by insurers to group plan sponsors. Providers have been known to receive kick-backs for referrals for unnecessary services.

[Big data and public health, part 2: Reducing unwarranted services.]

To combat fraud, the Attorney General and the Secretary of the Department of Health and Human Services (HHS) have joined forces. As a result of the national Health Care Fraud and Abuse Control Program (HCFAC), the Department of Justice (DOJ) and HHS recovered $4.1 billion in funds from healthcare fraud cases in fiscal year (FY) 2011 with $2.5 billion going directly to Centers for Medicare & Medicaid Services (CMS).

The HHS and the DOJ HCFAC Program Annual Report for FY 2011 provides specific details on detected fraud items. The perpetrators of these abuses include pharmaceutical and device manufacturers, medical equipment suppliers, hospitals, clinics, pharmacies, physicians, nursing homes, home health organizations, and transportation companies. In other words, perpetrators represent a cross-section of the entire healthcare industry.

The successful identification of these fraud and abuse cases is only the beginning of systematically identifying and reducing healthcare fraud and abuse of which big-data tools will play a key role.

Big data, predictive analysis and early detection

To date, only 3 to 5 percent of fraud is actually detected—and usually late in the payment cycle. Additionally, only a fraction of money that could have been used to provide care is recovered. Some cases are the result of billing and coding errors, which with prevention practices in place, could have been entirely avoided.

In a similar manner to how credit card fraud is identified, our healthcare system needs to constantly and automatically seek out suspicious activities at all levels. Pattern tracking, anomaly detection, and link analysis techniques need to be perfected and should include real-time and near real-time tracking applications. Reviewing and correlating records months after the fraud has been committed is the equivalent to closing the barn door after the horses have already left.

Model simulation with scenario analysis can be used to predict fraudulent behavior. These models can quantify the impact of fraud and abuse on the system based on different policy models and help support policy change decisions.

Big-data tools can be used to review large healthcare claims and billing information to target the following:

1. Assess payment risk associated with each provider
2. Over-utilization of services in very short-time windows
3. Patients simultaneously enrolled in multiple states
4. Geographic dispersion of patients and providers
5. Patients traveling large distances for controlled substances
6. Likelihood of certain types of billing errors
7. Billing for “unlikely” services
8. Pre-established code pair violation
9. Up-coding claims to bill at higher rates

How agencies can prevent fraud

Examining healthcare fraud with the application of link analysis, including looking at the social networking patterns of known bad actors, is a promising new approach. Similar to how email spammers almost never reply to comments, a fraudster’s social behavior pattern also differs from legitimate users.

The Center for Medicare and Medicaid Services (CMS) has put together a centralized approach to using predictive analytics. As part of their Fraud Prevention System and Medicare’s anti-fraud program, CMS is building a repository of algorithms to target specific claim and provider types. The goal of the program is to keep individuals and companies that intend to defraud out of the system. It also equips CMS with the tools to recognize fraudulent claims and eliminate payment errors.

[Part 1: How to harness Big Data for improving public health.]

The Fraud Prevention System was launched on June 30, 2011. Medicare claims run through a predictive modeling tool similar to that used by credit card companies. Near real-time data is examined, providers are given a risk score, and if it’s high enough, are subject to payment delays and a follow-up visit from CMS.

CMS should be applauded for their use of big data predictive analytics to identify fraud and abuse of healthcare services. Now, this effort needs to be expanded both within CMS and to other agencies across the government who face significant fraud and abuse problems.

In the next article, I will address how big data can be used to reduce administrative inefficiencies in healthcare systems.

Roger Foster is a Senior Director at DRC’s High Performance Technologies Group and advisory board member of the Technology Management program at George Mason University. He has over 20 years of leadership experience in strategy, technology management and operations support for government agencies and commercial businesses. He has worked big data problems for scientific computing in fields ranging from large astrophysical data sets to health information technology. He has a master’s degree in Management of Technology from the Massachusetts Institute of Technology and a doctorate in Astronomy from the University of California, Berkeley. He can be reached at rfoster@drc.com, and followed on Twitter at @foster_roger.