Medicare Claims Analysis with HSI
Top Ten Medicare Claims Myths
Myth #1: Medicare claims analysis is a mainframe-only analysis.
HCFA Fact: The entire Medicare claims database for one year is stored on approximately 2,000 data tapes. While the Health Care Administration (HCFA) will never release a 100% sample of their data, they will provide a 5% sample of the claims for under $20,000. The problem is that HCFA WILL ONLY PROVIDE THE DATA ON MAINFRAME TAPE OR CARTRIDGE.
Myth #2: Medicare claims data can only be used by large organizations because of cost.
HSI Fact: We've made the cost Medicare claims analysis at least 10% of what it has been. There are two big costs to using Medicare claims data: loading the data and cleaning/preparing it for analysis. Even if you have access to a mainframe, the task of loading and preparing in excess of 80 tapes per year for a year is time-consuming task. Since HSI staff know the structure of Medicare claims, we can easily generate a core set of analytic files from mainframes tapes using our 3480 and 9-track tape readers. What's more, we have the disk space to assemble larger analytic files because our file server will allow for ten times larger than HCFA's mainframe disk drives!!!! We also save you significant programmer/analyst labor costs by 'cleaning' the data of unwanted data elements for analysis. Finally, the databases we produce from HCFA data are on PC-based media CD-ROM, DVD (coming Spring 1998) and PC-backup tapes so that you can start working on the data right away using a high-end Windows or Macintosh personal computer. How high end? Usually a Pentium 100 works fine, but we prefer using our symmetric multiprocessing Pentium II server that cost us less to build than a used Hyundai.
Myth #3: Medicare claims data cannot be used in a relational database format.
HSI Fact: We've made Medicare claims analysis as convenient as working with any other relational database. Since the DEMPAQ project, Medicare claims data have been organized as relational databases for physician profiling. We are currently working on developing a common file claims file structure that can be applied to any data using Medicare claims as a template. With this technology we a building a episode of care client-server application for use on a stand-alone windows based machine. The goal of technology is to marry the depth of health services research methods with off-the-shelf relational database products to give managers and analysts the tools to generate custom profiles with a few mouse clicks.
Myth #4: Everybody has access to Medicare claims data.
HCFA Fact:ALL Medicare claims data requests must be approved by HCFA. Over the last few years, HCFA has significantly increased the security of its data systems to prevent the release of "individually identifiable information". To acquire Medicare claims data you must request it through the Office of Healthcare Information at HCFA. In your request, you must provide a research protocol defending how your analysis will provide a public health benefit. HSI staff have contacted HCFA personnel regarding this process and help stream-line the application process by acting as the agent for the organization requesting data. After successfully navigating through HCFA's data proposal review process. we would then pre-process the claims data and provide analytic files under the terms outlined in HCFA's data release agreement to be signed by HSI staff and any prospective client.
Myth #5: You can purchase Medicare claims data outside of HCFA.
Painful HSI Fact: Last Fall (1996), we made a formal request to HCFA allow purchase of the Medicare 5% claims file to HSI for re-packing as a commercial HSI product. Our plan was to market an analysis ready version of the 5% sample at roughly half the HCFA purchase price on a multi-volume CD-ROM set designed to be used by SAS and relational database packages. We and another inquiring organization were turned down by HCFA early this year on account of heightened security procedures and HCFA's desire to review every use of line-item Medicare claims data. HSI will continue to engage HCFA in a discussion on the best method to promote ease of use of the data and safeguard the inappropriate release of individually identifiable information.
Myth #6: Organizations that buy Medicare claims must analyze it themselves.
HSI Fact: Other organizations, such as HSI, can be approved by HCFA to act as custodians of the data to create databases and perform analyses by signing the data release agreement between HCFA and the organization that purchased the data from HCFA. All organizations using HCFA data must return the data following the completion of an analysis and NOT use in for any other analysis that was not approved by HCFA (See Myth #9).
Myth #7: Only hospital and physician claims are available.
HSI Fact: From the Medicare National Claims History file, one can purchase:
Myth #8: You can complete Medicare claims analysis through the World Wide Web.
HCFA Fact: To date, HCFA has not established a web based client-server system to allow access to Medicare claims data from the WWW. Based on our experience with the Web and Medicare claims, we believe this technology is possible, provided HCFA either runs or contracts out a subscription service to potential analysts.
Myth #9: You can use purchased Medicare claims for any purpose, once your original purpose is approved.
HCFA Fact: HCFA is especially not fond of organizations that use Medicare claims for purposes other than those approved when the data was provided. HCFA clearly states that any new use of the data must by cleared through another review of research protocol. HSI seeks to operate under the rules laid out by HCFA and would work with an organization to modify their original data approval from HCFA as necessary if the purchase of Medicare claims was facilitated through HSI.
Myth #10: Medicaid data is available in the same format at Medicare claims from HCFA.
HSI Fact:We've worked with both, so we should know. While the data elements are similar, there are significant disparities between Medicaid and Medicare claims because of each state's 'custom' Medicaid claims database. HCFA originally contracted out the task of standardizing several states data into the SMRF (State Medicaid Research Files) format. However, HCFA only has SMRF data available for 1992 and before, with the intervening years between 1997 and 1992 in a non-uniform format. Suffice it to say that HCFA is working or developing a standard database for more recent years that can be applied as easily as the Medicare National Claims File for analysis. The one key advantage of Medicaid claims data over Medicare claims is the inclusion of pharmacy claims. The one clear disadvantage of Medicaid data (other than a lack of uniformity) over Medicare data is that a person may spend only a few months out of year in the Medicaid program and not provide adequate information for research and analysis purposes. Because of the design of the program, Medicare beneficiaries generally provide continuous information unless they just entered or left the program due to age, death or entry into a managed care plan.
sample medicare claims projects
analysis project details
data project details
software project details
welcome about HSI contact us details of services sample projects sales information technology watch reference desk