Technology Evolves to Thwart Insurance Fraud

Insurers’ ability to fight insurance fraud has improved dramatically in direct proportion to advancements in technology. Thirty years ago, state-of-the-art data matching meant combing through paper files until you found a match. Since then, systems data analytics have continually evolved, with the most rapid acceleration taking place in the past eight years. The development of capabilities that provide data visualization, claims scoring, and predictive modeling have revolutionized data analysis. Other developments have provided important process and workflow enhancements, such as fast online access to vast amounts of public-records information and automatic compliance reporting. Industry leaders are more determined than ever to tackle insurance fraud by using sophisticated fraud-fighting solutions.

The Single-Database Breakthrough
Claims analysis took a quantum leap forward eight years ago with the development of the first all-claims database. With that development, insurers easily could scan information about claims in all lines of business instead of only one. ISO ClaimSearch® remains the industry’s front line of defense against fraud.

Today the database contains 570 million claims. By reporting claims to the system, adjusters can view matches based on the claimant’s name, address, vehicle identification number, Social Security number, telephone number, and more. System output can reveal numerous fraud indicators: the claimant has an invalid Social Security number, receives correspondence at a post office box, has filed numerous claims with multiple vehicles, or has filed claims under different names. Using fuzzy logic, the system can match John Smith, Smythe, and Smithe, leading to the discovery that one man is attempting to conceal his identity to collect from multiple policies.

In addition, investigators can initiate broad queries of the database based on any data element they designate. This allows considerable flexibility in follow-up claims investigations. The inquiries return instant results online, in real time. Investigators also can tap public-records information through the system’s integrated public-records portal.

Through the relationship between the National Insurance Crime Bureau (NICB) and law enforcement, the system also provides access to millions of records from impound lots, U.S. Customs, vehicle manufacturers, and the FBI’s National Crime Information Center. In addition, access is provided to stolen-vehicle data from the Coordinating Office of Insured Risks and the Mexican Association of Insurance Companies (OCRA-AMIS).

Next Steps: Link Analysis and Data Visualization
With so much information available, the challenge for adjusters and investigators becomes making sense of it all. If a search reveals that a driver has had many accidents, how does an adjuster know if the claimant is part of a fraud ring or simply a bad driver? Further analysis is crucial. The next step in the evolution of fraud-fighting technology was the development of link-analysis and data-visualization tools to assist investigators in evaluating information.

The tools analyze claims by sifting through large quantities of data and uncovering hidden connections within the information. A claimant might be linked to several addresses, telephone numbers, vehicles, and claims. Within seconds, a graphic picture of the relationships appears, pointing toward possible fraudulent activity. Without this software, it could take investigators days of poring over lengthy reports to find suspicious connections. Now they can see at a glance the individuals and service providers that should be investigated further.

While desktop data-visualization systems are highly useful for individual investigators, insurers also can take advantage of more powerful data-visualization systems that can analyze data from multiple sources, including public records, and can investigate both external and internal fraud. To catch dishonest employees profiting from claims fraud, a system can analyze the company’s checks to find suspicious payments, such as checks written to employees’ relatives.

Assigning Numeric Values to Claims
Claims scoring was the next development in the evolution of fraud-fighting capabilities. This technology provides adjusters with a more effective way to distinguish between suspicious and meritorious claims. Claims scoring employs scenario-based models to assign a numerical ranking to a claim based on the likelihood of fraud. Thousands of data points can be identified and captured for claims analysis. Take the example of a policyholder who lowers his deductible and soon after has an auto accident. The close timing of the policy change and the accident would give the claim a higher score. Further investigation of the case could reveal that the accident actually took place before the policy change was requested. Claims scoring systems can compare all of the claim’s attributes against proven fraud indicators, and provide a score based on matching claims information as well as the information on the parties involved. Claims with high scores that meet company-defined thresholds may be referred directly to the Special Investigations Unit (SIU).

While designed as a fraud-fighting tool, claim scoring has the additional benefit of streamlining the claims-handling workflow by assisting in the assignment of claims. Less complicated claims can be assigned to junior adjusters, while claims with more complex attributes can be assigned to senior personnel who have the skills and experience to handle a complicated claim.

Predictive modeling takes advantage of sophisticated statistical models, and uses prior outcomes to predict fraud. Using historical claims information, the system looks for certain sets of variables that previously have been associated with fraud. An example might be claims in which both the Social Security number and the VIN are invalid. Predictive modeling analyzes claims for those prior outcomes and indicates when a claim has a greater likelihood of being fraudulent.

Other Technologies Support Investigations
Once a claim has been referred to the SIU, case management technology can assist investigators in inputting and analyzing information as they develop their cases. The systems also can enhance workflow by facilitating communication between investigators and managers, and streamlining reporting to the NICB and state fraud bureaus.

The Internet is another technology available to investigators that requires no start-up costs and is just a mouse-click away. SIUs can tap into this vast, ever-expanding information resource to uncover information that isn’t available anywhere else.For example, many people post detailed personal information on social networking sites. Investigators easily can find this information to help them develop their cases.

Communicating Fraud to Industry Organizations
The issue of mitigating fraud extends beyond the doors of insurance companies to the National Insurance Crime Bureau (NICB), state insurance fraud bureaus, state fire marshals, state insurance departments, and law enforcement. Those organizations rely on insurers’ information to track fraud on a wider scale, and insurers often are called upon to report claims information to support the organizations in their work. With advancements in online Web-based systems, the pace at which this information moves has accelerated. Today there is an almost instantaneous delivery of claims data from insurance companies to the central repository of the NICB and, where appropriate, to state insurance fraud bureaus.

Detecting Underwriting Fraud
To combat underwriting fraud, it is essential to catch inconsistencies between what the applicant reports and information available from third-party sources. Today technology allows underwriters to quickly access databases and public records online to verify application information, confirm prior coverage, and discover undisclosed drivers.

Industry Efforts Toward Improving Fraud-Fighting Resources
Even the most advanced claims analysis capability is dependent upon the quality of the data being analyzed. In recent years, an industry fraud data working group—composed of senior claims executives and representatives from ISO and the NICB—developed plans to improve industry resources in the fight against fraud by improving the data reporting process. The result of this partnership was the development of fraud data initiatives aimed at enhancing ISO ClaimSearch and improving companies’ claim reporting to make the system’s data more actionable for fraud detection and investigations.

To date, seven major initiatives have been completed. They include adding new fields to the all-claims database for improving fraud predictions, including 15 common fraud indicators across all lines of business in match reports, and encouraging companies to focus on improved reporting by providing benchmarking data against industry wide compliance. To continue the momentum, ISO and SIU directors developed six new initiatives for 2008. Many of the new initiatives seek to make the process of identifying and reporting suspicious claims faster, easier, and more efficient for busy claims handlers.

Today the insurance industry is making significant progress in its efforts to combat insurance fraud. Technology continues to evolve, bringing greater precision to data analysis, identifying claims that have the greater likelihood of fraud, and allowing meritorious claims to be paid quickly. Technology is also helping underwriters confirm information to eliminate application fraud before it occurs. The first industry-wide efforts to strengthen data collection and analysis have been successfully implemented. Those ground-breaking initiatives are being followed by additional goals for 2008. Through coordinated efforts and the strategic use of a wide range of technologies, the industry is making strong headway in thwarting insurance fraud on all fronts.

Vincent Cialdella is an ISO senior vice president.