The Case for Program Integrity in Medicaid Managed Care

Part Seven: Encounter Data Case Study

In our previous post on encounter data, we offered recommendations for meeting the challenges of problematic encounter data. In this installment, we present a Case Study from one of our own engagements, in which we helped our client transition from PI in the fee-for-service model to the managed care model. We share this real-world example to illustrate those challenges and how they can be remediated with accurate, high-quality data.

Helping Our Client Work with Managed Care Data and Meet its Challenges

The Myers and Stauffer team has performed pharmacy, institutional, and non-institutional fee-for-service (FFS) and managed care claims audits for numerous state clients. In doing so, we have developed test plans, including statistically valid random-sample designs, to create audit approaches focused on the risks of misstatement of Medicaid paid claims. For one client who thought their encounter data was reliable, we identified critical issues that ultimately affected the managed care claim audit process, including test plans, sample design, and extrapolation.

The Problem

At the start of this engagement, the client’s sole focus was auditing FFS claims. However, with the state’s transition to managed care, they needed help adapting their PI approach. In particular, the client, using data analytics, needed to trust the encounter claims data – which was thought to be complete and accurate – to identify risk areas in the Medicaid program, and to perform managed care claim audits. However, while conducting the audits, we discovered the encounter data accuracy was not what was advertised.

Our Proposed Solutions

Upon discovery of inaccurate encounter claims data, we discussed the importance of an increased emphasis on mitigating future audit issues around encounter data by incorporating the following:

Data Validation. Validation of the encounter claims data submitted for audit helps ensure that only final, paid claims are included, thereby excluding prior claim void/adjustment history and duplicate claims. This is crucial for the sample design process because subsequent identification of adjusted and/or duplicated claims in the claims universe can generate additional, unanticipated audit time and costs (sample redesign, etc.). If identified later in the audit process (when sample redesign is not possible), extrapolation of audit results (improper payments) is also less likely, minimizing the effectiveness of the audit.
Coordination with MCOs. Discussions with MCO representatives help ensure correct understandings of encounter data and proper usage for Performing the same analytics on both FFS and encounter data without distinguishing between programs and different coverage rules and practices may lead to identification of false or misleading risk areas and increased audit costs. As one example, if the intention is to identify high-risk durable medical equipment (DME) services for audit, MCOs should confirm the encounter claims properly identify DME claims as DME, as opposed to being waiver services incorrectly billed with a DME procedure code. Ongoing communication with MCOs helps ensure data accuracy and understanding, enabling efficient identification of data elements for effective analytics.

Upon identification of encounter claims data issues that could affect audits activities, our team worked with the client for quick resolution and thorough understanding of the issues to ensure audit success.

Additional Takeaways

By helping clients with managed care audits using encounter claims data, we have learned lessons that ensure audit success, including:

Data Accuracy is Key. Validation procedures are necessary to ensure data excludes voided, adjusted, and/or duplicate claims.
Data Elements Must Be Understood and Accurate. Carefully scrutinizing service-level billing helps confirm the data fields being utilized are correct, thereby ensuring more accurate results. For example, if the objective is to perform a DME claims review, how are DME services identified? Should selection criteria change for FFS versus managed care claims? Should different data fields be used to ensure only the intended services are included? Understanding and using fields correctly means greater accuracy in identifying program risk areas and problematic providers for audit.
Consideration of Data Source. If using encounter claims data for analytics, obtaining claims data directly from the MCO for audit purposes may prove to be more reliable until encounter data in the state repository can be verified as complete and accurate. Although it may take longer to receive paid MCO claims data to begin the audit process, using this data could improve audit accuracy while eliminating additional audit costs that might be incurred if using inaccurate encounter data.

By incorporating these lessons learned, you can better plan for future data analytic processes and perform cost-effective audits.

In our next installment, Managed Care and Program Integrity Oversight of Subcontractors: Meeting the Challenges of Layered Complexity, we explore the relationship between MCOs and their subcontractors and/or delegated vendors – those charged with managing certain contracted dimensions of service delivery – and why they may not be performing PI and what it could mean.

Myers and Stauffer

Purpose driven. Exclusive focus. Government Programs.

Our Benefit/PI program area covers a range of services, disciplines, and areas of focus, including data analytics, root-cause analysis, and encounter data validation. We are here to answer any questions and help with any health and human services needs your agency may be experiencing. Contact a member of our team today.