Identifying Fraud: Trends, Predictors and Exposures

Summary by: Joseph C. Monahan

What does the General Manager of a Major League Baseball team have in common with the Manager of an insurance carrier’s Special Investigations Unit or SIU?  If you guessed that both often employ data analytics, you are correct.  In their panel discussion entitled Identifying Fraud:  Trends, Predictors and Exposures, moderated by John Finnegan of Chadbourne & Parke, LLP, CNA’s Bo Barber, The Hartford’s James Hopkins, and Kevin Miller from Travelers explained the utility of data analytics in an insurance fraud investigation and provided other insights into the current state of SIU investigations. 

The panel described the evolution of SIU from the days when such units were typically composed of either claims personnel, or former members of law enforcement, to their current state, where they often include experts in medical billing codes and other specialists.  The panel explained that medical bills are a particular focus given that such expenses account for such a large percentage of claims (e.g., as much as 70% in the worker’s compensation context).  The panel also emphasized that the analysis is only as good as the data with which the investigator is working, and that the adage “garbage in, garbage out” holds true.  Likewise, the analytical results should be viewed as theory, with the investigative team using good investigation skills to determine if that theory is supported by the facts they uncover.

The fraudsters are typically not the insureds, but rather providers and vendors.  In many cases, they are sophisticated operators using software to help them determine what bills were previously rejected and how to avoid detection in submitting the next bill, leading to a cat and mouse game with the SIU investigator. 

While the sharing of information between carriers regarding different fraudulent billing  practices in the market place can allow for more proactive identification of fraud, the panel indicated that antitrust concerns cause carriers not to share an excessive amount of data with each other, particularly regarding a particular provider, unless they are in one of thirty states  allowing carriers to share information where fraud is suspected.  Carriers can also share  information through the National Insurance Crime Bureau, which grants antitrust immunity to reporting insurers. 

While the typical goal of  the SIU is loss mitigation, where there is enough evidence to support prosecution, investigators will sometimes refer a matter to the authorities for prosecution.  Even where criminal charges are not pursued, SIU may involve medical boards, chiropractic boards, and other professional organizations to pursue available sanctions against a provider as appropriate. 

The panel noted that computer programs like Access and Excel are very powerful tools for the well-trained investigator, such that there is little need to resort to expensive outside vendors to aid in the investigation.  The panel identified some red flags that could trigger an investigation in the medical context, including a change in the severity of billing or in the frequency of billing over time.  However, no change in a billing pattern over a period of time could signal fraudulent “template billing”, where bills are generated on a regular schedule regardless of the actual treatment.   

The panel estimated that between 10% and 17% of every dollar spent on insurance claims is attributable to fraud, and opined that in the run-off context, that percentage is likely higher,  partly because it can be more difficult to prove fraud on older claims.