Talking about claims and using Decision Management to improve claims I recently caught up with Donald Light of Celent. Donald recently wrote a new report for Celent – Claims Systems Vendors, North American P/C Insurance 2011. This is the most recent update of the Celent study that evaluates vendors in the claims processing space. The report ranks core systems (in this case claims systems for P&C insurance) on 4 different dimensions and does so every 2 years.
Obviously these core claims processing and adjudication systems have various levels of workflow and rules built in (as well as standard features like support for reserving, claims adjuster desktops). According to Donald there is a definite trend for these core systems to now come with some kind of rules-based capability. Some of them are reasonably robust and sophisticated (though not on the level of a dedicated BRMS), others much less so. Most of this rules capability is focused on claims validation and routing but there is a growing interest by insurers on automating certain decisions and workflows, but not the entire end-to-end process. True straight through processing (STP) or auto-adjudication is focused currently on simpler claims (like windshields in auto for instance) and slowly moving up the complexity curve.
In my view, for now it seems that organizations wanting to auto-adjudicate a lot of claims will need to be thinking in terms of adding their own decisioning capabilities to these core systems – combining business rules with analytics (to detect fraud and abuse) to get high rates of auto adjudication without creating the opportunity for fraud.
Interestingly, and somewhat different from other industries, Donald said that support for process management or workflow is not at the same level yet. Mostly these packages are apparently focused on screen flow and managing dynamic forms and data capture rather than on true process management.
Finally, in terms of analytics, all the packages have reporting and it is increasingly common to have BI-like functionality such as dashboards, drill down, graphical displays and configuration of client-specific metrics and reports. That said, there is not much around data mining/predictive analytics yet and this remains something you will have to add on.
So if you like the sound of claims decision management, it looks to me like you are probably going to be buying and implementing a decision engine or business rules management system and building/embedding your own fraud models. For some suggestions on how to go about this, check out our Decision Management Solutions white paper on Next Generation Claims Systems over at decisionmanagementsolutions.com/resources.
