Beating Workers’ Compensation Fraud With Technology NCSI 2009 Annual Meeting May 18, 2009 1

Beating Workers’ Compensation FraudWith Technology

NCSI 2009 Annual MeetingMay 18, 2009

1

Agenda

• Workers Comp fraud overview– Fraud Stats– Types of Fraud– Fraud in the News

• Current anti-fraud efforts– Fraud Awareness– Fraud Enforcement

• Tools for fighting WC fraud– WC Fraud Indicators – Red Flags– Industry-wide Databases – Public Records– Data Analytics– Scoring/Predictive Analytics– Premium Audit Model– New tools for fighting WC fraud

• Legislative Update - Medicare Secondary Payer Reporting

2

Workers Compensation Fraud

• Workers Comp fraud costs $6 billion per year. Coalition Against Insurance Fraud

• One in three adults in U.S. condone exaggeration of claims.

Insurance Research Council

• Studies show that 10% of P&C claims and 36% of BI claims involve fraud or inflation of otherwise legitimate claim.

USAA Magazine

3

Workers Compensation Fraud Types

•Employee/claimant fraud

•Provider fraud

•Employer (premium) fraud

4

Workers Compensation Fraud Types

•Claimant fraud– False or exaggerated injury claims

– Claims for injuries not received on the job

– Collecting benefits while working other jobs

– Reduction in workforce results in increased workers comp claims

5

Workers Compensation Fraud

6

Workers Compensation Fraud Types

•Provider fraud– Exaggerating treatments for minor injuries

– inflating and

– billing for treatments not provided

7

8

Workers Compensation Fraud Types

•Employer premium fraud– Under-reporting payroll amounts

– Misrepresenting job classifications

– Misrepresenting employees as independent contractors

9

10

11

12

Industry Anti-Fraud Efforts

• Fraud awareness

• Investigation/enforcement

13

Fraud Awareness

• Industry Awareness– Fraud training for adjusters/underwriters

– Support for industry-sponsored organizations

• Public Awareness– Insurers/state funds

– Coalition Against Insurance Fraud

– NICB

– State Fraud Bureaus

14

Fraud Enforcement

Example: State Fraud Bureau Activity

California District Attorneys’ WC Fraud Program

Fiscal Year 2006-2007 Summary• 549 arrests

• Prosecuted 1,115 cases with

1,224 suspects

• 499 convictions

• Restitution of $24,953,650

ordered; $8,639,562 collected

• Total chargeable fraud was

$260,292,381

15

New Tools for Detecting Fraud: New Tools for Detecting Fraud: Data and AnalyticsData and Analytics

16

Fraud Indicators“Red Flags”

Industry-wide Databases

Case Management

Public Records

Tools for CombatingWC Fraud

Data Analysis&

Visualization

Scoring/Predictive Analytics

17

WC Fraud Indicators – “Red Flags”

18

WC Fraud Indicators – “Red Flags”

Claimant FraudClaimant was a seasonal worker at the time of the injury

Injury occurred shortly after hire

Notice of Injury occurred after employee was terminated

Claimant immediately secured attorney representation

Delay in reporting injury to employer

No witnesses to injury

Claimant has visited multiple medical providers in connection to the injury

19

WC Fraud Indicators – “Red Flags”

Provider FraudTreatment regimen is inconsistent with injury severity

“Cookie cutter” treatments and billing records

High incidence of drug prescriptions

Claimant immediately secured attorney representation

Delay in reporting injury to employer

No witnesses to injury

Claimant has visited multiple medical providers in connection to the injury

20

WC Fraud Indicators – “Red Flags”

Employer (Premium) FraudInability to verify tax/unemployment reports

Insured refuses or delays access to records for audit

Claimants not reported on entity’s unemployment returns

• Multiple related businesses operating from same address

Insured selects a lowest-rated classification for exposure (e.g., oil or gas lease work vs. oil or gas well drilling)

Certificates of Ins. issued without corresponding payroll or subcontractor expense

High experience modifications with low premium exposure

Excessive use of “independent contractor” classification when experience rating

Equipment and vehicles not consistent with job classifications21

Industry-Wide Databases

Fraudsters are often repeat offenders!!

• Looking at activity across both insurers and lines of business can add perspective about claimants, providers and employers

– Claims activity by employees– Billing activity by medical providers– Policy activity by employers

22

What is ISO ClaimSearch?Background

• ISO ClaimSearch – the first and only all-claims

database for the property and casualty industry

– helps improve the claims handling process

– provides state-of-the-art resources used to fight fraud

23

ISO ClaimSearchDatabase Content

Casualty> Workers Compensation> Automobile Liability> Medical Payments> Personal Injury Protection> Auto Medical Payments> Homeowner’s Liability> General Liability> Disability> Personal Injury> Employment Practices> D&O / E&O> Fidelity and Surety

Property> Homeowners> Farm Owners> Fire> Allied Lines> Commercial> Ocean Marine> Inland Marine> Burglary and Theft> Credit > Livestock

Auto> Theft> Theft Conversions> Shipping & Assembly> Salvage> Impound> Rental Vehicles> Export> Vehicle Claim System> Int’l Salvage &Thefts

Volume: Over 602 Million Claims*

*11% annual growth rate24

Casualty

Property Auto

Self-Insureds 650 8 15

TPAs, IAs, MGAs 480 25 24

Insurers* 1,760 1,450 1,246

State Funds 24 - -

Accident/Disability 28 - 1

Types of Companies

•The insurance companies represented in this slide are responsible for approximately 94% of the annual DWP in the United States

ISO ClaimSearchMembership

25

• Claims Inquiry†

– Single Party and Two Party Searches

– OFAC LookUp– IQ Download (New Data Initiative)– AMA Physician Search*

– License Plate Reader*† Core Service for Insurance Companies. Optional Service to all others*Restricted to NICB members

ISO ClaimSearchCore Services

26

New Format for Universal Format Members– Industry-standard red flags

Date of hire/termination/lay-off to date of loss (Workers’ Comp.) Day of loss – Monday/Friday; Day after holiday (Workers’ Comp.)

ISO ClaimSearchOutput – Match Reports

27

Public Records

External data can help complete the picture!

• Public records– Individual information– Business information

• Criminal and Civil records

• Professional licenses

• Vehicle records– Registration information– Motor Vehicle Reports (MVRs)

28

Data Analysis and Visualization

Technology can help make sense of large data sets!

•Improvements in data storage capabilities

•Better off-the-shelf and custom software tools

•Data visualization software

29

Data Analysis and Visualization Tools

A picture is worth a thousand words!

30

Claim Scoring / Predictive Analytics

Claim

31

• Automation of “red flag” rules

• Scoring of individual claims with industry data

• Advanced analytic methods to identify fraud patterns

– Regression analysis– Social network analysis– Text mining

• Many WC applications:– Claimant fraud– Medical provider fraud– Employer fraud (including premium audit)

Scoring/Predictive Analytics

32

Advantages of Predictive Modeling over Rules-based Systems

Predictive Modeling can …

• More efficiently examine more possible predictors

• Take into account interactions between predictors

• Give different predictors different relative importance

• Efficiently examine and use all the historical data available

• Result in superior predictions!

33

Premium Audit Model Development

• By combining historical audit results with additional data and advanced predictive modeling techniques, an accurate prediction of net AP/RP can be developed for each WC account.

HistoricalAuditResults

ExternalDataResources

ExpertInsuranceRiskModelers

Advanced Analytics

Accurate Prediction of Audit Results

34

Premium Audit Model Development

• Example inputs to the model

– Comparison of class codes and business SIC

– Comparison of payroll size and length of time as a business entity

– Comparison of payroll to sales, SIC, and geography

– History of large APs for a particular producer

– Consistency of accident descriptions and class codes

– Hospital beds per capita

35

Implementation

• The Premium Audit model can be used to optimize three operational areas

1) Decide which accounts to audit based on expected additional premium generated (where allowed by state rules)

2) For those accounts that are audited, determine the most efficient allocation of mail, telephone, and physical audits

3) Optimize the order of audits so that the largest premiums due are collected first

36

Summary

• Workers compensation fraud is a continuing problem for companies and society

• New tools are available to help combat WC fraud of all types

• Take action now to stop WC Fraud in your organzation

37

Legislative Update

• Medicare Secondary Payer Reporting

38

Medicare Secondary Payer – Section 111Requirements

• The Medicare Secondary Payer legislation, section 111, requires insurers and self insurers (Responsible Reporting Entities) to report all claims involving Medicare- eligible claimants to the Center for Medicare and Medicaid Services (CMS).

• Lines of business include Workers Comp, Liability and No-Fault claims, considered “Non-Group Health Plan” (NGHP).

• Quarterly reporting involves all Medicare-eligible claimants– Recurring payments (WC and no-fault): report at first

payment or acceptance of coverage and at end of “ongoing payment responsibility” (ORM)

– Single payment liability claims: report only at settlement, judgment or award by Total Payment Obligation to the Claimant (TPOC) date

39

Medicare Secondary Payer – Section 111Registration Overview

• What is a Responsible Reporting Entity (RRE)?A responsible reporting entity is that company that assumes the risk of paying

the insured or claimant medical benefits or compensation for an injury for which the company or their insured is legally obliged to respond

• What companies are RREs?– Insurers– Self insurers (assume risk for line of business or state)– Companies with self insured retentions under which the

company pays the insured or claimant for loss.

40

Medicare Secondary Payer – Section 111ISO Estimated Deliverables

• Testing on new CMS fields with ISO May, 2009

• Assist companies with CMS testing, including Acknowledgments & Rejections

and Query file July, 2009

• Companies start production reporting

and querying to CMS October, 2009

• Final production deadline for all January 1, 2010

companies

41

Questions?

Thank you for your time!

John Swedo

Vice President Claims – AISG Group of ISO

(201) 469-3100

jswedo@iso.com

www.iso.com

42