CMS announced on June 17 that it will deploy predictive modeling technology to pre-screen claims. The goal is to identify fraudulent claims before they are paid.

Medicare has typically paid whatever claims a participating provider submits, and then sought to recover erroneous claims through retrospective review. This “pay and chase” approach is good for providers and keeps the cost of participating in Medicare at reasonable levels. It can be bad for taxpayers though because fraudulent claims will be paid just as quickly as valid claims.

Predictive modeling technology is widely used in private industry. It is used by credit card companies to detect fraud, insurers to evaluate risk, and financial analysts to evaluate business scenarios, among other things.

CMS has contracted with Northrop Grumman to develop the technology:

Northrop Grumman will deploy algorithms and an analytical process that looks at CMS claims – by beneficiary, provider, service origin or other patterns — to identify potential problems and assign an “alert” and assign “risk scores” for those claims. These problem alerts will be further reviewed to allow CMS to both prioritize claims for additional review and assess the need for investigative or other enforcement actions.

Northrup Gruman has a number of software development contracts with CMS, including the contract to develop and maintain the National Level Repository for use in assessing whether a provider has met the meaningful use standard.

If the new technology works, it could be good for providers and taxpayers by detecting truly fraudulent claims and eliminating criminal abusers. If it does not work and yields too many false positives, providers could become bogged down in justifying valid claims that should be paid without any question.