Controlling Claims Leakage through Technology

Definitions

As with any topic involving insurance and technology, proper definitions are in order, especially if an acronym is lurking nearby. Claims Leakage is simply defined as lost dollars through claims management inefficiencies that ultimately result from failures in existing processes (manual and automated). Or, as one unnamed claims executive said "the difference between what you did spend and what you should have spent on a claim."

Those differences can be explained by any or all of the following.

Process Issues

Inefficient claim processing
Improper/errant payments (which can result in fines and penalties in a workers compensation issue)

Human Issues

Poor decision making (missing the opportunity of settling a bad claim early in its life cycle)
Poor customer service (unreturned phone calls, longer claim processing time)
Fraud

Identification of CL can be accomplished through an audit of closed claim files. Analyzing the results of the closed claim files (i.e., how they were settled or closed) by comparing them against a number of leakage factors to determine the accuracy of the closure will show how bad the leakage is. For example, how consistent were the settlement decisions made when compared across the claim organization? Were proper reserving and settlement guides used? Was investigation sufficient to reduce the likelihood of fraud? Were subrogation/third-party recovery attempts made consistently and effectively across this sample of claim files?

Once this review has been conducted, and an idea of the extent of the leakage has been determined, what is the next step? Of course, hiring good claims people utilizing efficient processes is a logical and time-tested solution for the problem of claim leakage. But this article's focus is on the tools that can be brought to bear.

Technology Solutions

Following is a discussion of the best technology solutions that can be leveraged.

State of the Art Risk/Claims Administration System

From the perspective of the claims professional, a well-designed claims system is essential in both preventing and discovering CL. Simple human error can be prevented by the claim system's self-auditing features. For example, it would be impossible to process a payment in excess of reserves left available due to the built-in edit of most updated systems.

Unfortunately, many TPAs and insurers still have older, legacy-based systems with little interaction with other systems (i.e., nurse case management, medical management, underwriting, loss analysis/safety software, etc.). That makes it more difficult to spot trends and prevent foolish errors.

A more advanced claim system, however, will provide

a more flexible and extensible claim data model;
simplified claim function through consolidation of claims data, integrated systems, and reporting platform;
improved data-based decision monitoring;
improved monitoring of claims professionals (adjusters, examiners, payment processors); and
improved self-auditing function to prevent regulatory expenses (fines).

This is both good for the claims professional and the ultimate end user: the client. Some clients who are large enough to self-administer their claims may take advantage of some of the most advanced claims administration systems, such as Aon eSolutions' IVOS, Marsh STARS' Enterprise or Professional Edition, and CSC's RiskMaster. Each of these systems, as well as a host of others, provides these single, interfacing claims system solutions.

Reporting Analysis

Another feature of these newer claim systems is a reporting capability to analyze the vast amounts of claims, policy, and ancillary data within the database. It is the reporting tool that can be used to develop metrics for claim personnel performance as well as the analysis of claims processing. Many insurers and larger TPAs have developed client oriented claims systems that serve as the window of the claims program to the client. Some of the more sophisticated programs are Travelers' eCarma system, Chartis's IntelliRisk, and Sedgwick CMS's ViaOne.

These automated tools will equip the risk manager with the ability to analyze the claims data in such a fashion to identify potential and actual CL. For example, the following template reports are generally available through these systems and can identify some problem areas:

Average settlement cost. Looking at this trend over time can give evidence on claim values. If it is trending upward, more investigation is necessary to determine if there is CL involved.
Total caseload/adjuster. By itself, it is not an indicator of CL. But if the caseloads start increasing, one can ask whether all of the investigation and analysis is being done by an overworked claim examiner.
Number of reserve changes per claim. This is another indicator of potential claim leakage. Many reserve changes in a relatively short time frame may indicate stair-step reserving which is generally an inefficient method. It can mean potential CL.
Average time of first contact of claimant. Commonly known as a lag time indicator, this statistic is a leading indicator of CL. The longer the average first time of contact, the greater the chance of the claim increasing in value over what it should have been.
Claim turnover ratio. This is simply the number of new claims in during a month divided by the number of claims closed (either by closure, settlement, or denial). If the numerator greatly exceeds the denominator, or if the trend of the ratio continues to rise over time, it is a sign that there is a workload problem. That usually translates to CL.
Subrogation recovery trend. This statistic simply shows the success rate of recoveries against outgoing claim payments. A shrinking trend of recovery can definitely indicate CL.

These are generally available reports available through most claim information systems. However, there is an even better and surer method of determining the origins and extent of CL: predictive analytics.

Predictive Analytics Reporting Capability

Predictive analytics (PA) or modeling is the discipline that analyzes current and historical facts to make predictions about future events. It is a much deeper dive than traditional reports. And PA involves more than just the claims data as it also seeks information from as many divergent sources affecting claims as possible.

As to claims leakage, PA will focus on areas such as fraud detection: either missed through inadequate investigation or identification of actual fraud (i.e., revealing a coterie of law firms, chiropractors, autobody shops, and professional claimants working together).

Many insurers and system vendors are providing this capability to claim professionals and risk management professionals. Some vendors, such as Marsh STARS and Aon eSolutions, are using Predictive Analytics techniques to identify CL. Deloitte has developed a practice utilizing PA and modeling using not just claims data but other ancillary data to identify problem areas for CL.

Conclusion

Controlling claim leakage is an important method in reducing a company's ultimate cost of risk. And for the claim organization, it is a way to differentiate itself from other providers as a superior organization. Proper usage of claim systems, basic reporting analysis, and advanced tools such as predictive analytics can be valuable allies in reducing this cost.

¹ The Albert Risk Management Consultants claims management team (Stuart Cowan, Lisa Hartman, William Quinn, Jr., and David A. Tweedy) contributes articles on claims topics.