Interstate Variation and Progress Toward Balance in Use of and Expenditure for Long-Term Services and Supports in 2009. d. Data Anomalies and Exclusions


As with most administrative files of similar size and scope, MAX data contain a variety of data anomalies. A list of data anomalies associated with MAX LTSS measures used in this analysis for 2009 is provided in Appendix C.

The analyses presented in this report capitalize on the strengths of MAX while taking into account the aforementioned limitations of the data. Our analyses represent Medicaid enrollees from 38 states that we believe have reliable data.21 We excluded states with extensive missing data or data that vary significantly from summary measures reported for other data sources. In our previous study, we used MAX 2006 data and excluded 11 states from the analysis due to data quality concerns. Two of the states excluded in 2006 (New Hampshire and Texas) could be included in the analysis of MAX 2009 data, because data quality concerns from 2006 had been resolved. We excluded 13 states from all analyses in this study (Arizona, Hawaii, Maine, Massachusetts, Michigan, Minnesota, Montana, New Mexico, Oregon, Pennsylvania, Rhode Island, Tennessee, and Wisconsin). We list data anomalies for these and all other states in Appendix C. We based exclusion decisions on comparisons with statistics prepared by Eiken et al. (2011) that reflect CMS Form 64 data, comparisons with waiver statistics reported by Kaiser (2012), and knowledge about the structure of state Medicaid programs in terms of their institutional and community LTSS provisions, comparison of community LTC expenditures in MAX and in, comparison of Section 1915(c) waiver enrollees in MAX with Form 372 data, and information in the MAX anomaly tables about major data quality limitations in relevant data fields (such as missing claims data or missing waiver enrollment). In addition, we excluded the District of Columbia, Vermont, and Washington from our analyses of compared system performance for people with physical disabilities and those with ID/DD. In these states, we could not adequately differentiate people with ID/DD from those with physical disabilities.

We note that not all excluded states are known to have problematic LTSS data in MAX. We excluded states from the analysis when statistics obtained using MAX varied substantially from other published reports and the accuracy of MAX data could not be confirmed. MAX data for such states indeed may be accurate but are not included in our results.

21 In this report, the use of the word "states" encompasses the 50 states and the District of Columbia.

View full report


"ProgBal.pdf" (pdf, 1.6Mb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®