Introduction to CMS Data 02/28/2013 Presented by Erin Mann, ResDAC - PowerPoint PPT Presentation

Introduction to CMS Data 02/28/2013 Presented by Erin Mann, ResDAC Technical Advisor

About ResDAC  Centers for Medicare and Medicaid (CMS) contractor  Offer free assistance to researchers interested in using Medicare and Medicaid data for research  Staffed by a team of public health and health services research specialists  Provide a range of services related to CMS data ˗ Assistance Desk ˗ Workshops and Outreach 2

CMS Data – Where to Start?  Nearly 100 data files that can be requested or downloaded  Many other resources on the CMS website including summary statistics, reports, and payment system methodology files  Often difficult for both new and seasoned CMS data users to know where to start  ResDAC is here re to help. p. 3

Overview of CMS Data Files  Goals for today ˗ Provide an overview of how data files are classified ˗ Requirements for release ˗ Examples of data files ˗ Information about accessing data files  More detailed information about data file contents will be presented at later webinars 4

CMS Data File Privacy Levels  All CMS data files are organized into three privacy levels. ˗ Non-identifiable or Public Use (PUF) ˗ Limited Data Sets (LDS) ˗ Research Identifiable (RIF) 5

Non-Identifiable Data What are non-identifiable files?  Also known as Public Use Files (PUFs)  Files have been edited and stripped of all information that could be used to identify beneficiaries  The files contain aggregate level data on Medicare beneficiary or provider utilization  Some files contain no utilization information, but include characteristics about providers 6

Non-Identifiable Data What are examples of non-identifiable data?  Provider of Services File ˗ Information about institutional providers ˗ Name, facility characteristics  Part B Summary Files ˗ Organized by procedure code ˗ Total procedures billed, payment amounts  Medicare Cost Reports ˗ Annual reports submitted to CMS ˗ Utilization statistics, costs, charges, financial information 7

Non-Identifiable Data How can I access non-identifiable data?  Access will depend on the file, but data can either be downloaded for free from the CMS website OR purchased and the data will be sent on hard media  Non-identifiable data can also be shared  No data use agreement is required 8

Limited Data Sets (LDS) What are LDS files?  LDS data contain beneficiary level data  Files include protected health information, but selected variables have been removed, encrypted, or ranged ˗ No zip code, only state or county ˗ No exact DOB, either age year of 5-year range ˗ No true physician identifiers; institutional facilities CAN be identified  Only available as standard extracts ˗ 5% random sample or 100% files (depends on file) ˗ Not possible to customize based on geography, diagnosis/procedures codes, or researcher-submitted identifiers (Medicare numbers, SSNs for known cohort) 9

Limited Data Sets (LDS) What are examples of LDS files?  Standard Analytic Files (SAFs) ˗ Fee-for-service claims data ˗ Files organized by utilization type  Denominator File ˗ Demographic and enrollment information ˗ Often used with the SAFs  MedPAR ˗ Separate files for hospital, SNF, and LTCH ˗ Cannot be linked to the SAFs  Medicare Current Beneficiary Survey (MCBS) 10

Limited Data Sets (LDS) How can I access LDS files?  Requestors must submit an application to establish a Data Use Agreement (DUA) ˗ New Use – purchase the data new ˗ Reuse – request to use data already at your organization for a new purpose  Applications involve 4 documents  Requests typically take up to 8 weeks to process  Wide range of data uses 11

Research Identifiable Files (RIF) What are RIFs?  RIF data contain beneficiary level protected health information ˗ Most RIF data files do not contain direct identifiers, but do include information such as zip code, date of birth, and exact dates of service  Data is at the beneficiary level  Standard extracts are available (5% random sample), but custom cohorts (diabetics in Minnesota) are also possible 12

Research Identifiable Files (RIF) What are examples of RIFs?  Medicare Data ˗ Fee-for-service claims data ˗ MedPAR ˗ Part D event data ˗ Assessment data ˗ Master Beneficiary Summary File  Medicaid Data ˗ MAX Files 13

Research Identifiable Files (RIF) How can I access RIF data?  RIF data are only released for research studies  Requestors must submit an application to establish a Data Use Agreement ˗ New Use – purchase the data new ˗ Reuse – request to use data already at your organization for a new study  Packets typically involve at least 8 documents and documents are reviewed by ResDAC prior to submission to CMS  RIF requests are reviewed by the CMS Privacy Board 14

Summary of Privacy Level Differences RIF LDS Non-identifiable Beneficiary Yes Yes No level data? Data files Yes No No customizable to specific cohort? Linkable to Yes No No non-CMS data at the beneficiary level? Require Privacy Yes No No Board Review? Require a DUA Yes Yes No 15

New File Series: DE-SynPUFs  Data Entrepreneurs' Synthetic Public Use Files  Series of linkable files that were released in February  File is classified as a PUF since the data is synthetic, but created from LDS files  Purposes of the DE-SynPUF: Allow data entrepreneurs to develop and create software ˗ and applications Train researchers on data complexities prior to accessing ˗ actual CMS data ˗ Support data mining while preserving beneficiary privacy 16

Resources for Help  ResDAC Assistance Desk ˗ Email: resdac@umn.edu ˗ Phone: 1-888-973-7322 ˗ Web: http://www.resdac.org  ResDAC Website ˗ Search for files by Privacy Level ˗ Access request materials for RIF requests ˗ Browse ResDAC KnowledgeBase 17

Resources for Help  CMS Website - Good resource for Non-identifiable and LDS files - CMS.gov  Research, Statistics, Data and Systems  Files for Order - ResDAC website is more appropriate for RIF data  CMS Data Navigator ˗ http://dnav.cms.gov ˗ Menu-driven search tool for locating CMS data and information resources 18

Upcoming Webinars  03/19 – Non-Identifiable Data  04/04 – Cost Reports  04/09 - DE-SynPUFs  04/10 – Limited Data Sets  04/25 – Research Identifiable Data  05/02 – Utilization Data 19