Help

BI’s Article search uses Boolean search capabilities. If you are not familiar with these principles, here are some quick tips.

To search specifically for more than one word, put the search term in quotation marks. For example, “workers compensation”. This will limit your search to that combination of words.

To search for a combination of terms, use quotations and the & symbol. For example, “hurricane” & “loss”.

Login Register Subscribe

Technological advances have made it easier to detect insurance fraud

Reprints
Technological advances have made it easier to detect insurance fraud

Technology used by insurers and service providers to collect claims information has become more advanced in pinpointing data that shows potential for insurance fraud, sources say.

But while the information can make it easier for claims adjusters to spot and intervene in problematic cases, it also can be difficult for claims handlers to sort through various warnings raised by increasingly sophisticated software, said Paul Braun, managing director of casualty claims for Aon Global Risk Consulting in Los Angeles.

“Technology is good if it's used properly, but it also can overwhelm people,” Mr. Braun said.

Workers compensation claims, in particular, provide a wealth of data about claimants that can be analyzed based on their potential for insurance fraud, said Claude Yoder, New York-based director of global analytics for Marsh Inc. That information includes “structured” data, such as a claimant's age and occupation, and “unstructured” data, which includes notes about a claimant from claims examiners or physicians.

“The claim is typically the most robust source of information you have,” Mr. Yoder said.

By comparing a claimant's information with historical claims that have turned out to be fraudulent, companies can find claims that have a “statistically significant” chance of being illegitimate, Mr. Yoder said. And he says insurers and service providers are developing technology that is increasingly accurate at identifying such cases.

%%BREAK%%

Aon's Mr. Braun said such claims data is being electronically filed by insurers and claims handlers into an index operated by Jersey City, N.J.-based Insurance Services Office Inc. While companies previously submitted paperwork for the database, Mr. Braun said there has been a push among insurers and service providers in recent years for claims data to be filed electronically.

It has become easier for insurers and adjusters to be notified of potential fraud as the database has become increasingly digitized, Mr. Braun said. For instance, adjusters can receive almost instant notification if a person has filed more than one claim for a single accident.

“Technology made (the index) a more effective product,” Mr. Braun said.

The development of better fraud detection technology has been helpful particularly in dealing with organized crime groups that specialize in commercial insurance fraud, said Chris Story, Austin, Texas-based claims services manager with Computer Sciences Corp.

“As far as your major crime rings ... that do fraudulent activity for a living, they do know which companies use technology in some ways and which ones do not, and they're going to put their efforts toward the ones that do not,” said Mr. Story, whose company developed a Fraud Analytics Suite used by commercial and personal insurance companies.

Despite advances in claims fraud identification, experts say there are disadvantages to processing and reviewing claims that have a wealth of information.

While predictive models are beneficial for guiding the data collection process, such systems still need experienced claims handlers to sort through the information and decide how best to intervene, said Richard Pankhurst, an Austin, Texas-based director in the insurance advisory practice with PricewaterhouseCoopers L.L.P.

%%BREAK%%

“Predictive analytics are only as good as the data that's fed into them,” Mr. Pankhurst said.

Claims software should guide claims adjusters to ask targeted questions that provide specific information for fraud identification, Mr. Pankhurst said.

Mr. Braun said models that flag potential fraud run the risk of inundating claims adjusters with warnings that eventually could end up being ignored.

“If you see the same thing 15 times and then you find out it's sort of a false alarm, you tend to not respond to it,” Mr. Braun said.

Still, Marsh's Mr. Yoder believes newer technology is helpful in allowing adjusters and examiners to concentrate their efforts on a select number of troublesome claims.

“The focus that comes through that predictive analytics exercise ... helps create a more efficient way to use limited resources,” Mr. Yoder said.

Read Next