CMS 102: Conducting Research with Medicaid Claims Data

Presented at the University of Minnesota (Mayo Building) - Minneapolis, MN

The purpose of ResDAC is to assist CMS in increasing the number of new researchers skilled in accessing and using CMS data for studies. The overall goal is to improve the Medicare and Medicaid programs and add value to current CMS activities. ResDAC faculty and research staff have extensive knowledge of many research applications using CMS data, as well as data limitations and assumptions. They also are knowledgeable about the content of CMS data files, CMS Privacy Act requirements, CMS data release policies and CMS's data release procedures.

CMS has revised the format of the Medicaid data files that they create and release for researchers. ResDAC's CMS102 workshop provides information on the development, structure and application of the Medicaid Analytic Extract (MAX) files. The new CMS MAX files replace the SMRF (State Medicaid Research Files) for data from 1999 forward. If you are working with older CMS data in the SMRF format, this workshop will still be pertinent. Please note: CMS no longer releases SMRF files; only MAX data from 1999 forward is available.

Some new features of the MAX files, compared to previous SMRF data include:

Data collection mandatory for all states, available files include the District of Columbia
Link to Medicare enrollment database / Dual eligibility status
Eligibility codes improved
Increased numbers of diagnosis and procedure codes
Improved coding for services, including Emergency Department visits
Encounter-level data for HMO-enrollees is available for some states

Medicaid is an increasingly important source of coverage for underserved populations. Medicaid covers 36% of all births, 50% of all HIV patients, and 68% of all nursing home patients in America. Medicaid claims data allow health care utilization to be tracked for an individual including, for example, doctor's office visits, prescription claims, emergency room use, hospital stays and long-term care facilities stays.

In addition to information on the MAX files, workshop sessions will also focus on tools and resources that support research using Medicaid administrative data. Two such tools are risk-adjustment models. One of these models is the Chronic Illness and Disability Payment System (CDPS), a diagnostic-based classification system developed to assess risk and adjust capitated payments to health plans enrolling Medicaid beneficiaries. The second is Medicaid Rx, a risk assessment and adjustment model that uses prescription claims data.

Conducted in a computer-based classroom, the workshop will feature hands-on computer work, data anlaysis exercises, short lectures and demonstrations, and interactive learning throughout. Students will have their own computer station and will be working with specially-designed Medicaid data modules developed from real Medicaid data.

Researchers working primarily with non-MAX data files such as MSIS or state-created files should contact ResDAC to discuss their educational needs prior to registration. Familiarity with SAS software will be helpful, but is not required for the course.

Please note: The CMS MAX files are currently only available as Research Identifiable Files. CMS policy states: "CMS cannot support research that will lead to the creation of a product or tool that the researcher intends to market." The criteria used to evaluate the release of CMS Research Identifiable Data can be found at www.cms.hhs.gov/PrivProtectedData/02_Criteria.asp.

Workshop Design and Goals:

This workshop will familiarize attendees with the use of Medicaid administrative data for research. The course is designed for researchers, analysts, and practitioners from any health-related field who are interested in research questions that can be answered using Medicaid claims data. While there are no formal prerequisites for the workshop, attendees will be expected to be comfortable working in a Microsoft Windows environment, and to have some knowledge of basic statistics or analytic methods.

Educational Objectives:

At the end of the workshop participants will:

Understand the general design of the Medicaid program and the implications for research on this population.
Know how the data migrates from states to the final CMS Medicaid Analytic Extract (MAX) files.
Understand the structure of the files that compose MAX data and have a general understanding of the variables available in each file type.
Be aware of some of the data limitations.
Know where to locate resources and tools to support work with Medicaid data.
Understand how to use enrollment and claims data to define a study population.
Be able to generate basic descriptive demographic and utilization information on a Medicaid population including eligibility counts, physician visits, inpatient admissions, and lengths of stay.
Understand how to use the CDPS and MRx classification systems.
Be familiar how Medicaid data in other forms (MSIS, MMIS, state proprietary files) may differ from the CMS MAX data. Given these differences, understand the ramifications on potential research questions and which tools and resources are applicable.
Know the process for requesting MAX data from CMS.

Workshop faculty:

Todd Gilmer, Ph.D.
Associate Professor
Department of Family and Preventive Medicine
University of California, San Diego

Gerrie Barosso, R.D., M.P.H. M.S.
Technical Advisor
Research Data Assistance Center
University of Minnesota

This workshop is supported by CMS contract #500-01-0043; Spike Duzor, Project Officer

For questions about our workshops, please contact beaud005@umn.edu

Last Modified August 26, 2008