RUG Refinement

The aim of the "RUG Refinement" is to improve the ability of the existing RUG-III classification system to explain non-therapy ancillary (NTA) costs. The Urban Institute re-examined prior efforts by Abt Associates (2000) and Dr. Brant Fries (2003) using data from our 2001 analysis files and replicated and validated the research. Under this approach, the existing RUG-III classification system continues to be used while adding categories that account for beneficiaries who receive both extensive medical services and rehabilitation, i.e. the medically complex. See download section below for the final report by the Urban Institute, Options for Improving Medicare Payment for SNFs (March 2007), which discusses the refinement analyses. The download section below also contains tables showing the updated RUG-53 unadjusted case-mix indices and the updated 2001 distribution of days for the RUG-44 and RUG-53 models.

Final Report

This is the final report from the Urban Institute (March 2007) that describes data sources and methodological approaches used in the research of the RUG-III refinements. The report provides background information on the RUG-III classification system and the areas for potential refinement of the system, discusses the data and methods used in the analysis, discusses various potential approaches to classifying and reimbursing non-therapy ancillary costs, and provides results and implications for research and policy. Areas for research that could possibly improve the quality of data used to determine the SNF PPS payment include: MDS variables on service use; claims information on ancillary service use, diagnoses, and dates of service; and hospital discharge data.  In addition, the report discusses some potential approaches for developing a high-cost outlier payment policy for the SNF PPS.

Data

• Updated RUG-53 Unadjusted Case Mix Indices--The updated nursing and therapy case mix index chart shows the distributional impact of adding 9 new groups to the RUG-III hierarchy. The addition of the new groups requires a recalibration of the case mix indices for all the RUG-III groups. This was accomplished by running the original staff time measurement data utilized in determining the case mix indices for the 44-group model through the RUG-58 grouper (see below). The unadjusted case mix weights and column descriptors are in the download section below.
• Updated 2001 distribution of days for the RUG-44 and RUG-53 models. This distribution was used to calculate the nursing case mix indices. In calculating the distribution, 861 unclassifiable records (BC1) were deleted from the file.
• RUG-44/RUG-53 Distribution Chart, based on CY 2006 data, is available for download from the Downloads section below.

Resources

• Report to Congress on Patient Classification under Medicare's Prospective Payment System for Skilled Nursing Facilities (2006)
• Dr. Brant Fries Presentation (2003)--These are slides from the May 2003 Presentation by Dr. Fries. The "Testing Revisions of the RUG-III System for Non-Therapy Ancillary Cost" Presentation was done during the May 2003 Urban Institute Technical Advisory Panel.
• Abt Associates Report (2000)--The goal of this study was to review the RUG-III classification system with particular emphasis on the care needs of medically complex Medicare beneficiaries and the variation in non-therapy ancillary services within RUG-III categories.
• FY 2001 SNF PPS Proposed Rule--April 1, 2000: Vol. 65, No. 69
  This rule proposed to refine the RUG-III classification system based on research by Abt Associates. This also includes a Technical Appendix which provides the supporting analyses.
• RUG-58 Grouper--This Statistical Analysis Software (SAS) program is used to generate the 58 groups that provide the base structure for the new RUG-53 model. This includes the original 44 RUG-III groups plus 14 additional groups including MDS assessments that would qualify for one of the original 14 rehabilitation groups.