One opportunity to apply the platform model of innovation is in the use of data. From education to transportation, virtually every state agency is looking to data to improve performance. Combating Medicaid fraud is particularly appealing given the abundance of data and enormous potential savings. States manage a lot of funds — New York’s Medicaid program, for example, pays out $55 billion each year to more than 100,000 providers, and these funds are a prime target for fraud. Nationally an estimated $19 billion of Medicaid funds are spent each year on improper payments. To address this problem, almost every state attorney general’s office has set up a Medicaid Fraud Control Unit (MFCU) with the federal government providing 75 percent of the funding and the rest coming from matching state funds.
While MFCUs are useful — in 2013, MFCUs recovered $2.5 billion from their investigations — more can be done, especially with many states expanding their Medicaid programs under Obamacare. Until recently, the overwhelming volume of transactions has made it difficult for auditors to find violations. However, last year the U.S. Department of Health and Human Services ruled that states could use federal funds for data mining programs to fight fraud. This change means more states can develop and deploy data analytics programs to identify fraudulent activities, prioritize investigations and uncover complex criminal activity. Some states have begun to prove the value of these investments. The TexasHealth and Human Services Commission, for example, spent about $20 million to build a fraud analytics program that has already more than paid for itself. States like Massachusetts and Illinois have also made notable progress in using analytics to reduce fraud.
By sharing their data, states can encourage competition between vendors focused on delivering the best analytical capabilities. Some states may want to explore alternative contract arrangements where analytics companies receive a percentage of the improper payments recovered or Netflix Prize-style competitions where states offer rewards for significant improvements in the accuracy of their algorithms. The goal of these efforts is to make data the locus of innovation.
States will need to make some changes to maximize their savings from better analytics use. For instance, most states use a “pay-and-chase” system, which results in wasted effort recovering money that’s already gone out the door. Real-time analytics can help identify suspicious payments at the outset, but states must update their payment policies before this can be implemented. States should also consider data sharing agreements with private insurers and other states to improve the quality of their analytics programs. More comprehensive data may help investigators detect new schemes to defraud Medicaid programs and ensure that bad actors identified by one state don’t move on to another.
Importantly, state Medicaid programs should recognize that the goal isn’t just to implement a one-off analytics program for fraud, but to develop a multipurpose analytics platform that can be used to generate predictive intelligence, cut costs and streamline operations. Just as Apple built the iPhone to be more than a communications tool, so too should states design their analytics platforms to be reused for other purposes.
While the early payoff may come from reducing fraud, long-term gains may come from using data to streamline the health-care system and improve health-care outcomes. By creating a platform for analytics, states can make progress on the immediate goal of reducing fraud, and they can also use it to unlock a new wave of data-driven innovation.
