“Typically carriers have left risk management up to the underwriting and loss control executives or have done risk management by committee,” comments Pauli. “But the [CRO] position is going to be surfacing to a great degree over the next couple of years.”

As philosophies and best practices for risk management pervade the enterprise, carriers also are looking at technologies that support various line-of-business processes to see how those systems might augment the risk management process. In their search, claims has been a natural target.

“By its nature, the claims function has a couple of points where it touches ERM,” Light says. One of these points is compliance risk. “Claims is one of the more heavily regulated functions within insurance,” he explains.

The other main point is the risk to corporate image. As a process that involves the policyholder in almost every instance, claims handling has been an area of competitive differentiation for carriers, and poor claims handling the cause of hits on a carrier's reputation, from which recovery is long and difficult.

Content management technologies have proven to be a vital tool carriers use to manage compliance and other enterprise risk. These technologies drive standardization into the documents claims departments generate via templates and make electronic content more easily retrievable for regulatory review than paper-based files.

“The initial goal of [content management] technology may have been focused specifically on something such as going to paperless files, but the end result is you wind up creating a much broader or more detailed capture of data surrounding the claim investigation, which can be used at an enterprise level [for risk management],” says Clint Harris, analyst at Conning Research & Consulting.

Coupled with workflow controls, content management also helps ensure consistent and timely process completion and avoid regulatory criticism. “Through workflow rules, companies can track what processes should have been completed or what forms should have been sent on claims that were open X number of days ago, and they can note that electronically in the record,” Light indicates. “This is in contrast to the poor claims adjuster with 100 claim files saying, 'Now, on which of these do I need to send a letter out today?'”

The Hartford is in the process of deploying a content management system in its claims area, a technology that already exists in other areas of the company. The reason claims was not the initial area targeted for the system was, due to the complexity and volume of claims and the nonstructured nature of claim documents, The Hartford first wanted to develop its content management competencies in data-based transaction processing areas.

Additionally, the company wanted to optimize its claims handling before moving forward. “You don't want to automate bad processes,” explains Karen Chamberlain, The Hartford's CIO for property/casualty claims. “Instead, we looked at our capabilities and defined the workflow we wanted. We then assessed how we needed to get our arms around all the diverse content that flows through the claim process so it gives us visibility, accessibility, and the ability to track, route, and manage that content effectively and consistently. Only then will we digitize content, get it in a central repository, and begin to build workflows around it.”

For The Hartford, there are several risk management goals that will be addressed by the content management initiative. According to Paula Beland, vice president in the property/casualty claims planning and business integration area, the biggest is disaster recovery.

“Vulnerabilities exist across the industry as we are locked into a paper claim file,” she says. “For Katrina-type disasters or fire, there is significant rework to recreate that [claim file] material. Our ability to store and have access to the full digitized claim file–not just the typical data fields from the management system and a few adjuster notes, but the entirety of a file, photos, e-mail, paper mail–will enable something that's significant from a disaster recovery standpoint.”

Another benefit is it will give The Hartford a centralized view of claim content for audits and legal actions. “Our lack of ability to get after all that content certainly creates risk,” Beland points out. “The system will give us visibility into parts of the claim process we just don't have today.”

Equally important, the system will help drive greater standardization into claims. “It's no secret when you have a lot of people executing a process, you're vulnerable to variation and inconsistency. There also are many different types of content–phone calls, mail, e-mail–coming from different sources the workflow system will help manage and control,” says Beland.

Final decisions on software and hardware are pending The Hartford's evaluation of the entire claim process. However, the company expects to be able to leverage existing content management software being used elsewhere in the P&C operation as well as the internal knowledge built from other deployments.

“We've built foundational capabilities at an enterprise level,” Chamberlain says. “The claims IT organization really is acting more as an integrator as opposed to a builder and developer.”

Decision-automation and -support systems, including rules and predictive analytics, are used extensively in fighting fraud (see “Running Up the Flag,” p. 26). Outside fraud, these technologies have been used in the reserve setting and claims adjudication processes, both of which have significant risk management impacts.

“Analytics can put precision into reserving, including identifying claims that might 'blow up' and should be resolved more quickly,” Harris says. “That has significant impact from a risk management perspective.”

There has been a greater uptake of rules- and model-based reserving in the medical arena, including workers' compensation, than in property or liability lines. One insurer that has been using technology in setting reserves is the Ohio Bureau of Workers' Compensation (BWC), which has had MIRA Claims Advisor from Fair Isaac in place for more than a decade.

MIRA software uses statistical modeling based on historical claim data. The system incorporates rules to ensure predicted reserves reflect state laws and the bureau's own reserving guidelines. Calculated reserves are batch uploaded to the bureau's claims administration system.

MIRA's models can produce a reserving adequacy of about 98 percent, Fair Isaac asserts, a figure Elizabeth Bravender, actuarial director for BWC, says has been supported based upon analysis completed at the aggregate level. The bureau plans to reevaluate the efficacy of the platform on a claim-by-claim basis as it considers a migration to version 2.0 of the platform later in 2007.

Like other early adopters, BWC has found ways over time to extend the predictive modeling system to other uses, including those that impact risk management today. “We have grown tremendously as an agency in terms of using MIRA reserves in other processes,” says Bravender.

For instance, the bureau provides its case specialists information they can use to project ultimate claim costs and run “what if” scenarios reflecting changes employers could make to reduce those costs. “Employers don't realize how long workers' compensation claims can last. Being able to show them the predicted duration of a claim is two years unless they do something about it, or a claim is predicted to become a permanent condition, is pretty powerful information,” Bravender reports.

BWC also can use projected reserve data for claim assignment, routing high-cost or high-severity claims to more experienced adjusters. In addition, it has begun using the MIRA system more frequently for claim settlement. “MIRA reserve information is another tool used as a part of the process the settlement teams can check for reasonableness on their formula to support negotiations,” Bravender explains. “We realize there's a long-term financial benefit to settlement of claims. We're working toward cutting down our long-tail liability.”

A health insurer using rules in its claims processing is BUPA Ireland. One of the key enterprise risks the insurer identified within its claim process was breach of contract. If claims aren't paid on time, BUPA Ireland is in violation of service-level agreements with healthcare providers, not to mention the danger of garnering unwanted attention from BUPA corporate.

Speed of payment became a pressing issue as the company experienced rapid growth immediately following its 1997 founding. In 1999, the company implemented HaleyRules software from Haley Systems in its inpatient claims adjudication system, where bills are received directly from providers. In 2003, the insurer incorporated Haley into its outpatient system, as well, where reimbursement requests are received from policyholders.

HaleyRules connects with an Oracle database that maintains data on the types of health products individuals have, providing claims handlers details on what claims are compensable and amounts that should be paid.

“Quality was a major requirement,” says John Foley, project manager at BUPA Ireland. “Prior to the rules, assessors [claims handlers] would have a copy of benefits on the table beside them and would have to go to the specific benefit, read the book, and decide how much to pay. It was a very manual process. To be fair, a lot of them knew the rates, but there was a lot of room for human error.”

The rules-based system has standardized and streamlined the claims-payment process and, according to Foley, given the company a competitive advantage over the majority of BUPA Ireland's competitors that still rely on manual assessment. “It has sped up the process tremendously,” he says. “It's been a major bonus for us.”

Despite the benefit of rules and predictive modeling, application of these technologies to claims outside the medical arena or processes other than fraud detection has been slow. “People are warming up to it, but the applicability to fraud is [more] intuitive,” Pauli says. “I have talked to claims managers who are reluctant to turn reserving over to a predictive model.”

In other words, claims professionals are loath to admit their decision-making expertise can be augmented or automated. However, this likely will change as the impact of technology on both claims and risk management is proven and as decision-support systems pervade the enterprise.

As proof, Pauli points to the progression of these systems in commercial lines, where carriers initially questioned their applicability. “They're bringing predictive models, straight-through processing, and rules engines into those areas, and BOP policies are being [automated]. Commercial lines execs are coming around; now, we just need to convince the claims folks,” she says.

Geographic information systems (GIS) represent another category of technology that has had a place in underwriting and is making its way into claims. In claims, GIS have been deployed most commonly in catastrophe management, something that directly impacts risk management.

“How carriers handle catastrophes has an impact on corporate image,” Light says. “The idea is, who showed up first? The problem of how you mobilize adjusters in a catastrophe has been around forever, but how insurers can respond has been altered through the experience of 2005 [with Hurricane Katrina].”

Faster and more efficient catastrophe response was what Church Mutual targeted when it began using MapInfo's geo-coding software, providing what MapInfo terms “location intelligence” by representing Church Mutual's book of business as a set of geocoded map points. “The initial idea was to identify claims in a spatial manner, which is much easier than using ZIP code,” explains Christopher Graham, CIO at Church Mutual.

Prior to using MapInfo, Church Mutual would have to create custom reports from its claims and policy administration systems to determine the projected impact of a catastrophe. Today, MapInfo presents this information in a visual format while allowing Church Mutual to manipulate data more easily as weather conditions change.

“With geospatial technology, we can not only see the information as it develops and as it changes day to day, but we also can use the spreadsheets that are the basis behind the GIS system to manipulate the data,” says Gene Simon, property claims manager. “It makes it a very clean process that is detached from the normal encumbrances of administration [systems].”

For starters, this helps the insurer predict and plan a response before a catastrophe hits. “If a weather pattern is developing along the coast, and we get projected data [from national meteorological sources] of where it would hit inland, we would be able to go into our system to try to map out our exposure based on where the hurricane was going to hit and determine exposures and staff allocation,” Graham says.

And after disaster strikes, it enables Church Mutual to respond more efficiently by being able to better plan routes of locations to visit. Using geocoding is particularly important post-catastrophe, Graham adds, because “typically when we send people out to churches in those areas, the churches don't exist anymore.”

It also allows the claims department to take a proactive approach to post-catastrophe claims management. For instance, after the hurricanes of both the 2004 and 2005 seasons, Church Mutual used MapInfo to contact churches in affected areas to find out whether damage had occurred rather than waiting for claims to be reported. Responding quickly and proactively has an impact on image and reputation, Simon notes. “Those issues, we've found, are very vital to keeping good customer service and PR.”

Beyond catastrophe management, insurers can use geographic information systems to manage response to any type of claim. For instance, Church Mutual uses the MapInfo system in large single losses where it may have to undergo the added expense of sending specialized claims staff.

“To make [adjusters'] trips worthwhile, we run mapping displays that tell us whether there are other significant losses in the area,” explains Simon. If so, “they can plan a bit longer trip and visit three or four other sites where there are significant but not as extensive losses.”

Church Mutual has bridged external databases, such as from the Property Loss Research Bureau, to the system to identify overlooked areas of concentrated exposure and better manage its overall risk. The insurer also uses the system to assess property concentration for underwriting and reinsurance purchasing and to target more effectively its sales efforts to underserved areas.

Further, GIS can have a role in fraud detection. “If insurers see 50 different claims that all occurred in one location, they can take a look from a fraud standpoint,” Pauli says.

Simon sees a day where GIS are not only modeling and forecasting tools but an integral part of the day-to-day operations of claims. “In the future, getting into your claims and handling [them] may be as easy as going to the map of your territory, finding the area you're going to be in that day, and making decisions about the various claims issues that are going on in that day,” he says.

Both regulatory pressure and developing best practices will drive insurers' need to manage risk better. Claims, like all areas of the company, will be expected more and more to support a company's enterprise risk management efforts.

Part of supplying that support will be deploying new systems and finding innovative ways to leverage existing technology that will provide improved visibility into and control of claim processes. How claims systems fit into a larger enterprise risk management framework is the next technological challenge.

“People understand the need to represent their entire risk portfolio now, but in terms of creating an enterprise solution, carriers are all over the map,” Pauli says.

Additionally, technology is not the most important component of enterprise risk management. Effectively identifying, assessing, and managing risk ultimately relies on a company's ability to create a vigilant, risk-aware attitude among its staff.

“The claims [area] recognizes at a technological level data needs to be in a format that can be housed and used at an enterprise level, but on a working level, claims people still see themselves as claims people, not risk managers,” Pauli says. “Having an enterprise vision [for risk management] takes constant dialogue and is a huge challenge.”