Data capture in insurance: Thinking beyond the paper metaphor
By Mark Tucker:
Face it--unless you're in the insurance business, you probably don't think much about claims processing. Inextricably linked to some major or minor tragedy, submitting an insurance claim is one of those things we just tolerate, like root canals and annual physicals.
Yet few other business processes, except perhaps income tax filing, affect such a vast number of people in some way. Even if you never get sick, never have a car accident, never have your basement flooded--in other words, never file an insurance claim--the rates you pay for insurance coverage in part depend on how efficiently and effectively your insurer handles data capture for the claims filed by others.
Claims is a data-driven process. The data comes from everywhere, in the form of initial reports, adjuster or examiner reviews, medical provider reports and bills, data from other related claims, even data from the original underwriting of the policy. The nature of the business is that what looks in the end to be a flood of data may actually trickle in over a period of time that varies from case to case.
Claims processing is a classic data collection and distillation operation, primarily focused on the creation, submission, review and archiving of various documents by the claimant, the insurer and the various parties representing the interests of those two parties. Data capture serves two primary purposes in claims processing. At the macro level, captured data represents a record of the events leading to the claim and the explanation of those events--an audit trail if you will. Contained within all those records is event-specific, detailed data that is used to drive transactions, such as the creation of additional correspondence or the payment of the claim itself.
Today's claims processing practices are the legacy of a time when paper was the only medium for business documents, and to a large degree they still reflect the constraints of the paper medium. Each new event in the process is signaled by the submission of a form, letter or some other document, and the requirement to wait for the completion of certain documents by outside parties delays the process repeatedly.
For example, most claims processing operations require the completion of an approved notice-of-claim form, often representing the first delay of the process as the insurer negotiates the delivery of the proper claim form to the customer. Then, depending on the nature of the claim, the insurer and claimant may continue to experience repeated processing delays as the insurer waits for additional documents, such as medical or adjuster reports for workers compensation or property loss claims. What's needed here are methods to speed the information delivery and data capture for those critical reporting steps.
As is often the case whenever we apply technology to long-standing business practices, our earliest attempts at improving the claims processing cycle focused on the desire to make the process run faster--in other words, to speed the flow of paper. The approach we took was to use electronic imaging technology as a repository for the various documents, and workflow as a method to move the documents through the process. Large investments in electronic imaging and workflow as a means of data capture and routing have certainly paid for themselves when applied correctly, but that approach has reached its technological limits.
By now, the performance of those systems has already exceeded the capacity of the people who must work with them, making additional investments in raw performance unwarranted. Instead of fixating on improvements in the paper metaphor for data capture, it's time to rethink the overall process. I am not saying there is no longer any place for imaging in claims processing, just that the whole idea of waiting for documents to initiate and adjudicate the claims process is out of date. The alternative solutions for claims data capture are not waiting on the horizon; they're here today and already being taken advantage of by forward-looking companies.
Initial case intake
Companies have tried to get around the inherent delays in the submission of paper-based forms through the mail, by encouraging the use of telephone intake for collecting information about pending claims. There are many advantages to that approach. First, it shortens the cycle time by eliminating the need to send a claim form to the client, and then wait for the client to fill it out and send it back. It can also improve the accuracy of the submitted information because a trained intake person can immediately ask for clarification of certain statements as well as ensure that all needed information is captured the first time.
One client I've worked with also teaches the telephone intake staff to "listen" for indications that the speaker on the other end of the line may be less than truthful or otherwise not disclosing all of the information about the claim. While they never use such a suspicion as the sole basis for denying a claim, that reflective listening certainly contributes to the company's better-than-average fraud detection record.
Capturing data over the telephone requires both a phone and a place to enter the information, usually some type of online form or data entry screen. Moving from paper to online forms provides opportunities to improve processing, but there are a number of traps to avoid. First, be careful not to automatically assume that the online form should look exactly like its paper cousin. Don't waste a lot of time (and screen real estate) adding borders and shading that enhance the ability of one-time users to read the paper-based form. Second, be aware that the conversational nature of telephone intake does not follow the sequence suggested by the form.
Because we read from left to right, top to bottom, our instinct tells us we should answer the questions on the form in that way, and we may do so if we have the form in front of us. In telephone intake, callers do not have the form to guide them, and so they tend to just tell a story about the claim, in no particular order.
For example, let's say I'm taking workers compensation claims over the telephone, and the first section of our online form includes all of the claimant's identification details--name, address, SSN, etc. What happens if the online form forces me to enter those fields before any others, but the claimant starts the conversation by giving me a long, detailed description of the injury? Can I remember the injury description well enough to key it in after I've taken the claimant back through the questions the system wants answered first? The lesson here is that you have to design the online form in such a way that the person taking the information over the telephone can easily jump around to fill in the information in the order he or she hears it.
Customer completion of forms
When you're thinking about how to implement online forms, remember that data capture does not have to involve a processor in the insurer's office. Think again about the first report of claim forms we've been talking about. We once asked the claimant to fill in the form when it was on paper; why can't we have the claimant fill it in when we have it online? That can easily be initiated over the Internet; a simple link on the company's home page can start the process. Worried about security? OK, be cautious, but don't let it stop you. We're only talking about initial claim forms here, and there are many chances along the way to validate and verify a claimant's identity. Even if you're not ready to Web-enable the form to begin with, you should plan for the day when your customers will demand that capability.
Provider data capture
The capture and exchange of information between insurer and the various providers is an area that has already received a great deal of attention. Much of that results from efforts of the HMOs and other large providers to standardize the reporting procedures for federal government health programs such as Medicare and Medicaid. Initially, those solutions focused on the use of electronic data interchange (EDI) protocols to transmit data between provider and insurer because EDI offered a semi-standard encoding scheme, and it operated in a secure, private network environment.
With the advent of the Internet, and particularly with the impending standards for secure commerce over it, electronic data exchange is coming into the reach of smaller providers and insurers. The attraction to the Internet is partly due to its lower cost. As a "public" network, the point-of-presence costs for getting on the Internet are much lower than are the costs of setting up a private network. The same Web presence will support other aspects of data capture, such as the use of online forms for claims initiation that were mentioned before.
Compound document processing
Documents are not just text, and they are not just paper. Today's business documents are compound electronic documents, virtual objects that consist of text, graphs, photographs, even embedded applications (such as Java applets). Take a look at the Web--most of the sites you find today are perfect examples of a compound electronic document, with scrolling text, active links to other documents (sites), and program applets that do useful things and sometimes not-so-useful things. How about taking advantage of the notion of compound documents to support claims data capture?
Let's consider what happens in the case of a property loss. When writing homeowners insurance, most insurers take one or more photographs of the property as proof of the general condition of the property at the date of coverage. When I built a new house last year, the agent took Polaroid shots of the property during construction and when we finished, and those prints are in a folder in his office (not the insurer's home or regional office).
What does that have to do with claims processing? Potentially a lot. Think of underwriting and claims as a closed system, meaning that the data from one drives the decision-making process in the other. Suppose instead of taking print photos, my agent had used a digital camera. Those digital images could be part of the underwriting record for my property, kept in an electronic file with the rest of my policy application. Then, should I ever file a loss claim on the property, my agent or an adjuster could take another set of digital photographs of the damage and attach those to the electronic claim that goes into the regional examiner's office. Once the damage is repaired, another set of photographs proves that I spent the check on the repairs, and the information loop is complete. My insurer would have a complete record available for analysis, routine auditing of claims adjudication, even for training new claims examiners.
All of the ideas presented here are possible with today's technology. All represent in some way a change in the time-honored process of filling out forms and other paper documents and mailing them to the insurer. At the heart of each is a fundamental shift away from thinking about data capture as a paper-driven process.