Healthcare fraud very seldom happens in a vacuum. Therefore, data matching — the ability to identify, match, and merge records that correspond to the same entities — is essential to healthcare data analysis.
The Centers for Medicare & Medicaid Services issues a unique identification number called a national provider identifier (NPI) to each healthcare provider. Today, many healthcare providers practice through an incorporated entity such as a professional corporation (PC), professional association (PA), limited liability corporation (LLC), or other structure. As a consequence, even though the NPI number was envisioned as a unique identifier, many providers have at least two NPI numbers — one corporate and one personal. In Medicare claims data, these are most commonly seen as the billing NPI and the rendering NPI. Additionally, providers enrolled in both Medicare and Medicaid may have different NPI numbers and other identifying information for each program.
When an investigation is opened, it is important to determine the universe of potentially involved providers and what NPI numbers and health care programs may be involved. Data matching allows us to find all identifiers associated with a fraudulent provider in both Medicare and Medicaid and identify potentially related providers.
The National Plan and Provider Enumeration System (NPPES) database can be a useful tool to identify additional NPI numbers associated with a provider of interest and to identify potentially connected providers. Using the NPPES database allows a provider’s name to be compared against corporate names, which is very effective in finding corporations named after the provider. Any candidates returned by these matches should, of course, be verified by address and other means.
When searching for providers who are members of the same facility, matching based on the fax number can be very effective. Matching on fax number identifies providers in the same office on a functional basis. Providers may have individual phones in their offices, and they may carry individual cell phones, but it is less likely that they would have individual fax machines.
Matching providers based on organization name or address is typically more tedious, but can yield more results. There are many ways to write or abbreviate an address. There are also many abbreviations of the words “corporation”, “company”, “incorporated”, etc. which complicate the matching of organization names. These variations make finding exact matches difficult, but SAS and other data matching tools include features that allow users to compare multiple text values and score their similarity. This score can be used to determine the likelihood that two addresses or organization names with some variations are actually a match.
IntegrityM can provide the support needed to enhance your Medicare and Medicaid fraud data analysis. We have developed an effective tool to scan the NPPES data set using an NPI number as input for potentially matching or related providers using name, address, and telephone or fax numbers. We have also developed data matching algorithms for effectively matching addresses and organization names despite variations. Using these tools, we frequently identify multiple NPI numbers associated with a person, organization, or practice, achieving a better picture of the provider of interest, which may span multiple programs.
For additional information on how we can enhance your data matching strategies, call us at 703-683-9600 or contact us online.