Iowa Medicaid Enterprise (IME) needed help fighting fraud, waste and abuse in its new managed care program.
IME utilized the flexible analytic capabilities of IBM® DataProbe® software to run specialized algorithms that mined for the highest priority vulnerabilities.
Data base builtwith six years of fee-for-service and managed care data
USD 41.5 millionrecovered fee -for - service over a 2- year period
Provides oversightto Special Investigations Units for each of the MCOs
Business challenge story
Fighting fraud, waste and abuse
In April 2016, IME, part of the Iowa Department of Human Services, launched Iowa Health Link — a managed care program that offers Medicaid members access to physical, behavioral and long-term care benefits from a MCO of their choosing.
The state was still responsible for fighting fraud, waste and abuse in its Medicaid program, but under managed care, the task became more complex. In addition to surveilling traditional provider payments made from the state or MCO directly to providers, managed care added the complexity of a second payment — the capitation payments made to MCOs also had to be reviewed. In addition, under- and over-utilization had to be monitored to ensure that MCOs were not denying or restricting medically necessary treatment.
Strengthening program integrity with robust analytics
IME turned to the program integrity (PI) team at IBM Watson Health™ for help detecting, deterring and recovering Medicaid fraud, waste and abuse.
Using the DataProbe solution, the team built a PI database by integrating and standardizing six years of fee-for-service and managed care data history, including claims (paid and denied), eligibility, provider files and several reference files.
Then, they created a library of custom predictive analytics to identify complex schemes as well as simple billing mistakes. They used both parametric and non-parametric methods for anomaly detection, which included clustering techniques to identify peer groups with normal claims behaviors whose homogeneity has been enhanced not only by specialty designation, but also practice setting, procedure mix fingerprints, geography and other relevant data-driven factors.
The results were presented as straightforward alerts as well as multivariate composite scoring, supporting risk-stratified provider scorecards. This approach provided high-quality leads with fewer false positives than less-structured predictive models.
Getting results — and accolades
The Watson Health PI team helped IME analyze Medicaid funds paid to providers. Several high-priority targets were identified, and a series of desk- and field-based audits and investigations were launched. As a result, over USD 41.5 million was recovered during a 2- year period.
The team also created a comprehensive framework to provide oversight to Special Investigations Units for each of the MCOs operating in Iowa. And, in a recent review of the state’s Medicaid program, the Centers for Medicare & Medicaid Services (CMS) commended IME for its ability to detect and mitigate risks associated with provider enrollment and ongoing verification checks.
Iowa Medicaid Enterprise
IME is the division of the Iowa Department of Human Services (DHS) that administers the Iowa Medicaid program. On April 1, 2016, the DHS transitioned most Iowa Medicaid members to a managed care program called IA Health Link.
About Watson Health
For more than a century, IBM has been creating innovations that matter. And with Watson Health, we are committed to help build smarter health ecosystems. This means working with you to help you achieve simpler processes, better care insights, faster breakthroughs, and improved experiences for people around the world. Our core strengths combined — deep industry expertise in health, data and analytics, actionable insights, and our reputation for trust and security – can enable us to deliver support for our clients. Together with our clients and partners, IBM is putting smart to work. For more information, visit https://www.ibm.com/watson-health
Take the Next Step
To learn more about the IBM solution featured in this story, please contact your IBM representative or IBM Business Partner,