New Data Products: Chronic Condition Warehouse (CCW) & Medicare Current Beneficiary Survey (MCBS) Andy Shatto, Director, Division of Data Development and Services Group Office of Information Products and Data Analytics (Date)
CCW What s New? New Research Enclave Environment * Opening Fall 2013!!! New Interactive Dashboards Beneficiary Summary Dashboard coming soon!!
CCW Data Enclave Benefits ACCESS Increases data access > VPN and Virtual Desktop Utilize personal laptops Controlled virtual access can mean greater flexibility in data policy COST Increases efficiency of data delivery Reduces infrastructure costs for data users SECURITY Increases data security > no shipping of external media Satisfies all privacy and security requirements DATA & ANALYSIS Secure File Transfer System (SFTS) transfers data files efficiently and securely Convenient, quicker, and efficient access to CCW data Perform analyses and data manipulation Technical/Analytic support from CCW staff Statistical data output review
Data Enclave Process Flow DUA Approval by CMS Enclave Security Testing Extract/Transfer of Data Set to Enclave User Analysis Web/VM access User performs analysis Output results Secure File Transfer System (SFTS) File Upload File Download of Approved Output File(s) CCW Output Review Process Statistical Review HIPAA Compliance User Requests Output Review File Transfer Request System (FTRS) Request Data output review
Data Currently Available in the CCW Data Enclave File Name Years Master Beneficiary Summary File 1999-2012 Medicare claims: Inpatient Outpatient Skilled Nursing Facility Home Health Agency Hospice Carrier Durable Medical Equipment 1999-2012 MedPAR 1999-2011 Part D Event Data 2006-2011 Medicare-Medicaid Linked File 2006-2008 Medicaid: 1999-2009 IP, LTC, Other, PS, Rx, Medicaid Summary New Long Term Care Minimum Data Set 1999-2012 Inpatient Rehab Facility 2002-2012 Swing Bed Assessments 2002-2012
Data Enclave User Requirements Standard/Basic DE User Using SAS environment including: Working knowledge/experience with: SAS programming language SAS Enterprise Guide (EG) Review of information on CCW Website www.ccwdata.org/ SAS EG User Guides and CCW 101 Video My CCW > File Transfers page Data Enclave resources CCW workbench Tool
CCW Workbench Sign out of Workbench Create Finder File Upload Finder File Create Extract Access Privacy Policy and Accessibility Statements on ccwdata.org website Access CCW Help Desk contact information Access User Guide
Workbench Query Request Screens File Selection
CCW Statistician Performs Data Output Review To protect the confidentiality of Medicare and Medicaid beneficiaries, CCW performs data output review prior to release from the Data Enclave Avoids disclosure or perceived disclosure of protected data PHI PII Small cell sizes
Beneficiary Summary Dashboard In Development Dashboard will present statistical views of information on the enrollment, utilization and spending for Medicare beneficiaries Report tabs will consist of: Demographics by HRR Demographics by State Cost & Utilization Part D Measures by State Availability: 1999-2011
CCW Data Websites WWW.RESDAC.ORG Research Request Assistance WWW.CCWDATA.ORG CCW Data/Dashboards
Medicare Current Beneficiary Survey (MCBS) The MCBS is a continuous, multipurpose, in-person survey of a representative national sample of the Medicare population. Goals of the MCBS are to: Provide important information on Medicare beneficiaries that is not available in CMS administrative data; Determine expenditures and sources of payment for all services used by Medicare beneficiaries, including copayments, deductibles, and non-covered services; Ascertain all types of health insurance coverage and relate coverage to sources of payment; and Track changes over time, such as changes in health status, spending down to Medicaid eligibility.
Demographics Health status and functioning Cost of care Preventive services Access to, satisfaction with, and usual source of care Health insurance Household or facility characteristics MCBS Contents Survey weights Community/ Institutional timeline Event specific information: Inpatient/Outpatient Provider Drug LTC SNF Dental
MCBS Data Files Access to Care: Point in time Always enrolled No survey-reported cost and utilization included No imputation n =15,500 Cost and Use: Entire calendar year Ever enrolled Include costs and utilization and event-level info. Imputation n = 12,000
New to MCBS & Coming Soon! Ever enrolled Weights for Access to Care 2012 File Will allow researchers to use Access to Care data to produce estimates that not only represent the continuously enrolled, but also those that had limited enrollment during the year (e.g., new enrollees, deceased, etc.). Accountable Care Organizations MCBS will oversample for ACO members in 2013 and hopes to identify ACO members as part of our annual files (starting in Fall 2014 with the release of our Access to Care 2013 file). Data Briefs and MCBS User Training Please visit our website and review data briefs released with each MCBS research file. Coming soon!! Downloadable presentations & training slides on key components of the survey (e.g., MCBS research files, MCBS survey weights, etc.).
New and/or Health Reform-related Content/Questions Care coordination Patient activation Electronic health record use by provider of care Satisfaction, access to care, and usual source of care Knowledge and use of preventive services Mobility and Use of Transportation Services Premiums Out of pocket spending Enhanced race/ethnicity reporting Self-reported health outcomes (e.g., ADLs, health status, etc.)
MCBS Research Files MCBS Data Products Available in.sas or.txt formats Data broken out into segments (or RICs) by content area (e.g. Admin., Health Status, etc.) Files available for the following years: Access to Care (1991-2011); Cost and Use (1992-2009) 2010 Just Released! MCBS Data Tables Includes weighted population statistics on Medicare population and subpopulations using MCBS research file. Useful for program statistics, benchmarking, program monitoring, etc. Questions about and how to obtain MCBS data? ResDAC For general questions and to purchase research files contact CMS Research Data Assistance Center (ResDAC), www.resdac.org. MCBS Website MCBS data tables, codebooks, questionnaires, and other information can be viewed and downloaded at the MCBS website (www.cms.hhs.gov/mcbs).
CMS Interactive Dashboards Office of Information Products and Data Analytics
Chronic Conditions Dashboard Presents 2011 state, Hospital Referral Region, and national comparison data on the prevalence of chronic conditions, as well as Medicare costs and utilization measures for beneficiaries with multiple chronic conditions. Released in December 2012. Available at http://www.ccwdata.org/web/guest/interactive-data/chronicconditions-dashboard Geographic Variation Dashboard Overview Presents information on state-level variation in standardized per-capita costs for the Medicare fee-for-service population. Released in June 2013. Available at http://www.cms.gov/research-statistics-data-and- Systems/Statistics-Trends-and-Reports/Dashboard/CMS-Dashboard- Geographic-Variation-Dashboard.html 19
Chronic Conditions Dashboard - Prevalence Provides information on the prevalence of Medicare beneficiaries, and sub-populations, with 15 specific conditions. This set of conditions corresponds with the conditions suggested by the HHS Strategic Framework on Multiple Chronic Conditions. Allows state-to-state and state-tonational comparisons of condition rates. Displays information on comorbidity for specific conditions. 20
Spending by Chronic Condition Count How does spending compare between Maryland and the nation for females aged 65 74? Beneficiaries with 6 or more conditions in Maryland have per capita standardized In Maryland, costs of 9% $38K, of female compared with beneficiaries $31K nationally. aged 65 74 have 6 or more conditions, but this population accounts for 42% of Medicare spending How do the percent of beneficiaries and percent of spending compare for this group in Maryland? 21
Chronic Conditions Dashboard State and HRR Mapping The mapping tabs allow users to visualize variation in per capita spending, ER visits per 1,000, and 30 day readmission rates across geography and chronic condition count. Bar graphs allow for comparison of key measures across condition counts. 22
Source: Public use files compiled by OIPDA using the Chronic Conditions Warehouse (CCW) Our goal was to create a simplified form of the available data that would allow novice users to compare standardized spending across states and service types. This dashboard was initially released on June 4, 2013. Future iterations may include additional health measures and drilling functionality Demo Medicare Geographic Variation Dashboard 23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
Liberating Claims Data: Medicare Claims Public Use Files (PUFs) & Medicare Data Sharing Program Samuel C. Chris Haffer, Ph.D. Data Development & Services Group Office of Information Products & Data Analytics - CMS Baltimore, Maryland
Context for PUF Project CMS Data are a National Resource Access has been limited Transparency/Open Government Innovation: CER/PCOR Data Entrepreneurs Information Intermediaries
Pilot Project Goal While strictly protecting beneficiary confidentiality. Increase access to CMS claims data through the creation of de-identified data sets While strictly protecting beneficiary confidentiality!!!!
Phase 1: BASIC PUFs Challenge Create a set of BASIC or slim/thin files Comply with the HIPAA safe harbor standard Test the utility of the files for research Outcome 14 Basic PUFs Inpatient, Outpatient, SNF, DME, Hospice, HHA, PDE, Carrier (Reference Years 2008 and 2010)
Phase 2: Profile PUFs Challenge Increase utility of the files for research Comply with the HIPAA safe harbor standard Use More Data Outcome 8 Profile PUFs Chronic Conditions, Prescription Drug, Institutional Provider, Physician/Supplier (Reference Years 2008 and 2010)
Phase 3: Synthetic PUFs Challenge Unmet Needs of Phase 1 and Phase 2 PUFs Comply with the HIPAA safe harbor standard Outcome DE-SynPUF (20 Samples x 8 Files each) Inpatient, Outpatient, Carrier, PDE, Beneficiary Summary (Reference Years 2008, 2009, 2010)
Data Entrepreneurs Synthetic PUF for Medicare Claims Data (DE-SynPUF) GOAL: Increase access to realistic Medicare claims data files in a timely and less expensive manner to spur the innovation necessary to achieve the goals of better care for beneficiaries and improve the health of the population. USES: train researchers on the use and complexity of conducting analyses with CMS claims data prior to initiating the process to obtain access to actual CMS data; support safe data mining innovations that may reveal unanticipated knowledge gains while preserving beneficiary privacy; and, allow data entrepreneurs to develop and create software and applications that may eventually be applied to actual CMS claims data; 44
Data Entrepreneurs Synthetic PUF for Medicare Claims Data (DE-SynPUF) New type of synthetic file useful for data entrepreneurs for software and application development Preserve detailed data structure of key variables at beneficiary and claim levels Data is fully synthetic for disclosure safety Limited analytic utility due to lack of preservation of interdependence between variables Create file that can be certified to be released as a Public Use File 45
DE-SynPUF Contents 5% sample of enrolled Medicare beneficiaries in 2008 3 years of claims (2008, 09, 10) Inpatient Outpatient Carrier PDE Prescription Drugs
DE-SynPUF Description The synthetic beneficiaries have nearly the same number of claims as their seeds, with very similar claim patterns The data structure is very similar to the real CMS files, albeit with a smaller number of variables Programs and procedure designed using the SynPUF should be fully functional when applied to CMS limited data sets
DE-SynPUF Tables Beneficiary Table Rows: 2.4 million Variables: 52 Inpatient Claims Table Rows: 1.8 million Variables: 81 Outpatient Claim Table Rows: 21.4 million Variables: 76 Carrier Claims Table Rows: 120 million Variables: 142 Prescription Drug Event Claims (PDE) Table Rows: 157 million Variables: 8
We want you!! Download and Explore Conduct Exploratory Research Offer Constructive Feedback Encourage Others Use as a Teaching Resource
Medicare Data Sharing Program: What it is. A legal authority, allowing Medicare Parts A and B claims data, and Part D event data to be purchased by organizations certified by HHS as meeting certain requirements for use in producing publicly available provider and supplier report cards using measures of quality, efficiency, effectiveness, and resource use.
Medicare Data Sharing for Performance Measurement NOW FUTURE INSURANCE COMPANIES MEDICARE? INSURANCE COMPANIES MEDICARE? Qualified Entity COMPREHENSIVE CONSISTENT FAIR ACTIONABLE 51
Medicare Data Sharing Program (ACA 10332) Section 10332 of the Patient Protection and Affordable Care Act (ACA) added a new subsection to Section 1874 of the Social Security Act, requiring that the Secretary: 1. Establish a process to allow for the use of standardized extracts of Medicare Parts A, B, and D claims data by Qualified Entities (QEs) 2. who will evaluate and report on the performance of providers of services and suppliers 3. using measures of quality, efficiency, effectiveness, and resource use. 4. Defines QEs as public or private entities that are determined by the Secretary to be qualified to use Medicare claims data to make such evaluations of provider/supplier performance 5. agree to meet specific requirements regarding the transparency of their methods and 6. their use and protection of Medicare data.
Medicare Data Sharing Program (ACA 10332) 7. Requires Medicare claims extracts be combined with other claims data. 8. Specifies the only use of such data and the derived performance information about providers and suppliers be in reports in an aggregate form. 9. Reports must be released and made available to the public, 10. after first making such reports available to any identified provider or supplier and 11. affording an opportunity to appeal and correct errors. 12. Instructs the Secretary to take such actions as she deems necessary to protect the identity of individual beneficiaries, and 13. authorizes her to establish additional requirements that she may specify for QEs to meet, such as ensuring the security of data.
Why is the Medicare Data Sharing Program IMPORTANT? INCREASES ACCESS TO MEDICARE CLAIMS DATA FOR PERFORMANCE MEASUREMENT - a major component in our CMS strategy of improving care for individuals, achieving better health for populations, and lowering costs through improvement. HIGHLIGHTS CMS COMMITMENT TO DATA STEWARDSHIP - protect beneficiary confidentiality - minimize risk of accidental data disclosure
Certified Qualified Entities Name of Lead Entity Region(s) Date of QE Certification Oregon Health Care Quality Corporation Health Improvement Collaborative of Greater Cincinnati Oregon August 31, 2012 Ohio Indiana Kentucky Kansas Missouri August 31, 2012 Kansas City Quality Improvement Consortium September 4, 2012 Maine Health Management Coalition Foundation Maine November 28, 2012 HealthInsight New Mexico January 18, 2013 California Healthcare Performance Information System California February 6, 2013 Pittsburgh Regional Health Initiative Western PA March 27, 2013
FOR MORE INFORMATION: Public Use Files Basic PUFs www.cms.gov/bsapufs SynPUFs - go.cms.gov/synpufs Medicare Data Sharing Program www.cms.gov/qemedicaredata www.qemedicaredata.org
Medicare & Medicaid Research Review The peer reviewed, online research journal published by the Centers for Medicare & Medicaid Services David M. Bott, Ph.D. Editor in Chief Office of Information Products & Data Analytics
Mission Statement Medicare & Medicaid Research Review is a peerreviewed, online journal reporting data and research that informs current and future directions of the Medicare, Medicaid, and Children s Health Insurance programs. The journal seeks to examine and evaluate health care coverage, quality and access to care for beneficiaries, and payment for health services. 58
Why You Should Publish in Highly visible articles on CMS Web site reach key policy makers; 16,000+ MMRR Updates subscribers Accessible research is freely available to anyone with an internet connection No Use Restrictions in the public domain Ease of publication accepts submissions continuously; online submission and peer review 59
Why You Should Read PEER REVIEWED approved by external blinded reviewers; not the company line POLICY RELEVANT actionable results based on testing, not making, assumptions UNIQUE CONTENT willing to dive into details others avoid BUILDING CAPACITY highlighting new data and information products for use in research. 60
Highlights Improved content delivery MMRR Updates: Publication, Research & Data Products Alerts Web Redesign Articles in HTML full text News & Opportunities section Improved navigation Site specific searches Tell us what you would like to see: mmrr info@cms.hhs.gov 61
Data Briefs Purposes Introduce a new/enhanced program data resource Physician and hospital charge data Multiple Chronic Conditions Report observations of interesting phenomena Change in hospital readmission rates Characteristics of beneficiaries without flu vaccines Goal is to inform future research by highlighting resources and raising questions. 62
Next Steps Improved Author/Reviewer Experience Separate submission/review track and example templates for Data Briefs All manuscripts: Speedier acceptance to publication process Special Topics Hospital Acquired Conditions Program Post Acute Care Studies Medicaid Expansion Suggestions/Feedback: email to mmrr info@cms.hhs.gov 63
For more information : http://www.cms.gov/mmrr/ Submit research articles & data briefs to: http://mc.manuscriptcentral.com/mmrr/ Send feedback to: mmrr-info@cms.hhs.gov 64