Analysts using a system leveraging predictive analytics to identify improper Medicare fee-for-service claims need to improve the process for refining and enhancing its models, according to an audit by the Department of Health and Human Services Office of Inspector General.
The Fraud Prevention System (FPS), first implemented by the Centers for Medicare and Medicaid Services in 2011, uses models that predict suspicious behavior with the goal of preventing the payment of fraudulent claims. However, auditors say CMS might not have the capability to trace the savings from administrative actions back to the specific FPS models that generated the savings.
“Without this capability, the department is not able to accurately evaluate an individual FPS model’s performance,” states the OIG report. “Therefore, the department may be limited in how it assesses the effectiveness of its predictive analytics technologies.”
FPS rule-based models use rules to filter fraudulent claims and behaviors, while anomaly detection models leverage thresholds to identify potentially fraudulent behavior. In addition, the system utilizes predictive models based on past known fraud cases as well as social network analysis models that identify links among potentially fraudulent subjects.
However, auditors found that CMS could not track the savings from administrative actions back to the individual FPS model that initiated the investigation because that capability was not built into the system. As a result, they say the agency is unable to determine the amount of savings generated for each model.
“Calculating the savings from the administrative actions that resulted from specific FPS models is an essential step in determining how effectively a FPS model is performing,” adds OIG. “Without that essential information, CMS does not have complete data to make the most effective determinations about which FPS models should be continued, adjusted or retired.”
To address this shortcoming, auditors recommended that CMS make better use of its performance results to refine and enhance the predictive analytics technologies of the FPS models, including: ensuring the redesigned FPS is effective in allowing the agency to track savings from administrative actions back to individual models, and evaluations of model performance consider not only the identified savings but also the adjusted savings.
Identified savings are the actual and projected savings that the FPS identified that might not be prevented or recovered, while adjusted savings are the amounts the system identified of actual and projected savings that reasonably can be expected to be prevented or recovered.
CMS officials were not immediately available for comment. Nonetheless, in its written response to the OIG report, the agency concurred with the recommendations and outlined steps it is taking to implement them.
Specifically, CMS noted that in March 2017 it launched FPS 2.0, which is designed to provide the “capability of tracking an administrative action back to the models that generated the lead and attribute savings accordingly.” In addition, the agency said that FPS 2.0 will also “use adjusted savings to internally evaluate models,” adding that the “combination of identified and actual savings will allow CMS to evaluate how well various administrative actions perform in terms of prevention and recovery.”
Register or login for access to this item and much more
All Health Data Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access