CMS Encounter Research Identifiable Files – How Are They Different from Fee-For-Service Files?

Encounter Research Identifiable Files: some differences compared to Fee-For-Services

The Centers for Medicare and Medicaid Services (CMS) collects service-level encounter data from Medicare Advantage Organizations (MAOs) for the purpose of recording diagnoses used for risk adjusted payments to MAOs(2). Unlike a fee-for-service (FFS) claim, an encounter record is not a direct record of payment and not used for billing purposes. Thus, Encounter Research Identifiable Files (RIFs) have some unique aspects to consider:

• Chart Review Records Included(3)
  In addition to service records, Encounter files also include chart reviews. MAOs perform retrospective reviews of beneficiaries’ medical records to identify diagnoses that were not originally submitted to MAOs by providers and diagnoses that were submitted by providers in error. MAOs may report diagnoses identified by these reviews to CMS as chart reviews. Their only purpose is to add or delete diagnoses (for more information on chart reviews see the article, “Chart Review Records in Medicare Advantage Encounter Data”).
• Variable Differences
  Some variables available in FFS data are not provided for encounters. The list below contains examples but is not exhaustive. See CCW codebooks for complete lists of variables for each RIF type(4,5).
  • Financial Information: Encounter data files do not include financial variables. MAOs consider it proprietary information.
  • Provider Identifiers: CMS certification number (CCN) is not provided. Fully populated provider identification is provided at the NPI level. An organizational taxonomy field is also included but is not always populated. NPI’s can be linked to the CMS National Plan and Provider Enumeration System (NPPES) file for full taxonomy codes and additional provider information.
  • Provider Specialty Code: Encounter files do not contain a provider specialty code variable.
• Billing Rules
  Billing rules are consistent and publicly available for FFS claims, but MAOs are not obligated to follow CMS billing guidelines nor make theirs publicly available. Using FFS rules to identify specific services in encounter data is a reasonable first approach; but it is prudent to explore your encounter data to ensure that services of interest are thoroughly identified and understood.
• Lag Time
  MAO’s typically have up to thirteen months after the end of the year to submit encounter data to CMS for it to be eligible for risk adjustment calculations(3). Annual Encounter RIFs are released approximately 18 months after the end of the year. For example, preliminary 2023 Encounter RIFs became available in June 2025. This lag is longer compared to annual FFS RIFs, which generally become available after 14 months. Also note that while quarterly FFS RIFs are available 4.5 months after the end of each quarter, quarterly Encounter RIFs are not produced.

Encounter Data Quality

Unlike FFS Medicare claims data sourced directly from Medicare Administrative Contractors (MACs) that adhere to well-defined Medicare data rules and standards, encounter data is based upon data submitted to CMS's Encounter Data System (EDS) by individual MAOs. Although the EDS applies various data checks, encounter data quality remains highly dependent upon the quality of the data submitted by MAOs/plans.

Several reputable publications evaluate encounter data quality. The Medicare Payment Advisory Commission (MedPAC) includes Encounter data quality evaluations within many of its annual reports to Congress(6–9) as do other credible sources(10,11).

Good encounter data research practice includes analyzing research measures by MAO and/or plan to look for outliers. For example, extremely low measures could signal missing records or other issues. If plans with questionable data quality are identified, researchers should consider removing them and/or conducting a sensitivity analysis.

Conclusion

Although Encounter RIFs are quite similar to traditional FFS RIFs, they differ in several ways that may affect analytic plans. Encounter data users are wise to review encounter data documentation thoroughly prior to use. Researchers should also be aware that encounter data quality is not always as reliable as FFS and they may want to evaluate analytic outcomes by MAO or plan to identify quality issues.