Toolkit: To Help Decrease Improper Payments in Medicare Advantage Through the Identification of High-Risk Diagnosis Codes

What Is the Toolkit?

This toolkit offers Medicare Advantage (MA) organizations information that will enable them to replicate Office of Inspector General's (OIG's) techniques to identify and evaluate high-risk diagnosis codes to ensure proper payments and provide better care for enrollees. This toolkit is meant to be a practical, hands-on device that will help MA organizations improve the accuracy of their submitted diagnoses that are at high risk for being miscoded.

The Centers for Medicare & Medicaid Services (CMS) makes monthly payments to MA organizations for each of their enrollees. These payments are based, in part, on the diagnoses that their enrollees receive from their providers. Using a system of risk adjustment, CMS pays MA organizations more for providing benefits to enrollees with diagnoses associated with more intensive use of health care resources relative to healthier enrollees, who would be expected to require fewer health care resources. Accordingly, CMS relies on MA organizations not only to collect diagnosis codes from their providers and submit the associated codes to CMS, but also to implement an effective compliance program to monitor the accuracy of these diagnosis codes. In this respect, MA organizations generally review medical records in order to detect and correct inaccurate diagnosis codes.

We have identified certain diagnosis codes which, when paired with other circumstances, are at a high risk for being miscoded. This toolkit provides additional information about the circumstances under which these diagnosis codes could be miscoded and the actual programming codes that we used in our audits to identify them.

Why Did OIG Create This Toolkit?

We have performed several audits for which our objectives were to determine whether MA organizations submitted certain diagnosis codes—ones that when coupled with other data indicated that the codes were at high risk for being miscoded to CMS for use in CMS's risk adjustment program in accordance with Federal requirements. Thus far (as of November 2023), we have found that overall, approximately 70 percent of those diagnosis codes were not supported in the associated medical records. Moreover, some diagnosis codes were consistently not supported over 90 percent of the time. Although the audited MA organizations usually disagreed with our reports' recommendations and with various aspects of our audit methodologies, those MA organizations generally did not disagree with our determinations regarding these diagnosis codes, and some expressed a genuine interest in submitting corrections to CMS. Other MA organizations—including ones that we have not audited—have asked us to share with them how we decided which diagnosis codes were at high risk for being miscoded.

The purpose of this toolkit is to provide information as to the steps that we followed as well as the data supporting the process that we followed. Our hope is that the users of this toolkit will, at a minimum, use the information to detect and correct inaccurate diagnosis codes in their own systems. To a greater extent, we hope that MA organizations use this toolkit as a starting point to identify other diagnosis codes that are at high risk for being miscoded and take appropriate measures to prevent, detect, and correct such errors. Thus, our goals are to ensure that CMS makes accurate payments to the MA organizations and that the enrollees receive the correct level of care.

What Does This Toolkit Include?

We have compiled the high-risk groups identified in our audits for incorrect diagnosis codes that were consistently submitted to CMS. We have also included explanations that demonstrate why these specific diagnosis codes, when coupled with other data (including procedure codes and prescription drug events), are at high risk for being miscoded. This toolkit contains the Structured Query Language (SQL)—a programming language used in many software programs-that we used in our audits to query CMS's systems. To provide further information as to what we analyzed, we have included all of the codes used in our programming language. The SQL language and codes included in this toolkit relate to the 2019 payment year.

This toolkit enables an MA organization to adapt our actual programming codes in order to query its internal data systems so that it can more accurately identify the enrollees in the various high-risk groups.

---