Durable Medical Equipment, Prosthetics, Orthotics, and Supplies Competitive Bidding Program Health Status Monitoring
Summary of Findings thru the Fourth Quarter of 2015
No negative changes in beneficiary health outcomes resulting from the Medicare Durable Medical Equipment, Prosthetics, Orthotics, and Supplies (DMEPOS) Competitive Bidding Program have been observed to date.
The Centers for Medicare & Medicaid Services (CMS) has been actively monitoring the competitive bidding program since it was first implemented in nine competitive bidding areas (CBAs) on January 1, 2011. CMS currently actively monitors Round 2 CBAs and national mail-order program CBAs where competitive bidding was implemented on July 1, 2013, as well as all Round 1 Recompete CBAs where the program was implemented following the end of Round 1 Rebid on December 31, 2013. All Round 1 Recompete and Round 2 CBAs are assigned to one of four DME Medicare Administrative Contractor (MAC) regions, based on their geographic location (Northeast, Midwest, South, and West). This assignment can be found in all workbooks in the “DME Region Map” tab. The national mail-order program CBA includes all parts of the United States, including the 50 states, the District of Columbia, Puerto Rico, the U.S. Virgin Islands, Guam, and American Samoa. CMS monitors three groups of beneficiaries in each of the four DME MAC regions and the national mail-order program CBA.
- “Enrolled Population”—all people in the CBA enrolled in Original Medicare.
- “Utilizers”—Original Medicare beneficiaries in the CBA who have a claim for one of the competitively bid products.
- “Access Groups”— Original Medicare beneficiaries who are likely to use one of the competitively bid products on the basis of related health conditions. In the case of mail- order diabetes supplies, for example, the relevant access group would be composed of beneficiaries with diabetes.
Within these groups, CMS monitors claims rates and a range of health outcomes including deaths, hospitalizations, emergency room visits, physician visits, admissions to skilled nursing facilities, average number of days spent hospitalized in a month, and average number of days in a skilled nursing facility in a month. We also monitor beneficiaries who no longer have claims for a competitively bid item after the program began, beneficiaries who may at some point need the item, and beneficiaries who currently have claims for competitively bid items. The data have not indicated any negative changes in beneficiary health outcomes in any group. Separate workbooks displaying the aggregate level rates for the three groups can be found on the CMS website.
The basic structure of the monitoring efforts considers historical and regional trends in health status. To control for historical trends, each CBA’s historical baseline for each rate is provided, beginning in January 2011. Historical health outcome rates for both Round 2 CBAs and non-CBAs are provided for each of the four DME MAC regions to provide context for the Round 1 Recompete CBA rates.
In general, Round 1 Recompete and Round 2 rates in each DME MAC region track closely to rates in non-CBAs both before and after the implementation of the programs. For mail-order diabetes supplies, we provide national rates, as well as historical rates in Round 1 Recompete and Round 2 regions for each of the four DME MAC regions. To provide context for overall access to diabetes supplies, we similarly display rates for non-mail-order diabetes supplies, although they are currently not a competitively bid product category. Importantly, mortality and morbidity rates commonly display seasonal trends unrelated to the competitive bidding program (e.g. winter months of each year typically have elevated rates of mortality and morbidity). Additionally, rates that appear more variable tend to be based on a smaller number of beneficiaries.
Beginning with the first quarter of 2015, the files have been updated to incorporate the changes to our monitoring approach that commenced at the end of the Round 1 Rebid Program on December 31, 2013, and due to the availability of complete data for the first full year of the Round 1 Recompete program that was implemented January 1, 2014. These changes include:
- -- Monitoring usage and health outcomes related to six additional product categories (Transcutaneous Electrical Nerve Simulator (TENS), Nebulizers, Infusion Pumps, Commode Chairs, Seat Lifts, and Patient Lifts) that were introduced under Round 1 Recompete. We monitor usage and health outcome rates in six corresponding utilizer groups, and seven access groups. For the infusion pump product category, we monitor rates in two separate access groups.
- -- Monitoring both Round 1 Recompete and Round 2 using the list of Healthcare Common Procedure Coding System (HCPCS) codes that are covered by Round 1 Recompete.
- -- Comparing trends between the three groups of beneficiaries (mentioned above) in Round 2 and Round 1 Recompete CBAs, instead of Round 1 Rebid CBAs.
The ICD-10 code set was implemented for claims with dates of service on or after October 1, 2015. As a result, beginning with the fourth quarter 2015 update, access groups include ICD-10 diagnosis codes that are considered highly related to the usage of specific DME products. Due to the lack of available ICD-10 claims data, we redefine each access group by mapping its ICD-9 diagnosis codes to a corresponding set of ICD-10 diagnosis codes. To facilitate this process, we use General Equivalence Mappings (GEM) files to identify all ICD-10 diagnosis codes that link to the ICD-9 diagnosis codes for all access groups. In order to construct the most comprehensive set of ICD-10 codes, we use both the forward and backward mappings included in GEM. Because of the structural differences between the ICD-9 and ICD-10 code sets, many diagnoses have an imperfect match from one code set to another. In particular, there may be major discontinuities in the reporting of patient diseases after the implementation of ICD-10, which could result in substantive changes in the size and composition of the access groups. For future updates, we plan on using ICD-10 claims data to refine our access group definitions to ensure monitoring continuity.
A comprehensive listing of all ICD-9 and ICD-10 codes can be found in the “Downloads” section of the Health Status Monitoring page on the CMS website. For the ICD-9 code set, see workbook: “Access_Group_ICD9_Codebook.xlsx.” For the ICD-10 code set, see workbook: “Access_Group_ICD10_Codebook.xlsx.”