RDC Heflin Data. Data available from the MU Research Data Center

Transcription

1 RDC Heflin Data Data available from the MU Research Data Center A Branch of the Kansas City RDC and a unit of the Federal Statistical Research Data Centers (FSRDC) MU Population, Education and Health Center (PEHC) Colleen Heflin, Co-Director(presenter) Peter Mueser, Co-Director October 30, 2015

2 Presentation Outline Background: PEHC and the MU RDC Overview of data available through RDC 3 other presentations this fall Focus on health data Review of particular datasets and variables available Brief review of procedures for obtain access to MU RDC 2 PEHC

3 Background January 2013: Population, Education and Health (PEHC) Seminar Series starts Summer 2013: Submission of NSF proposal to establish Federal Statistical Research Data Center (FSRDC) in Kansas City with branch site at MU Facilitates research access to Census confidential data MU is in consortium with KC Fed, Kauffman Foundation, KU and UMKC 3 PEHC

4 Background June 2013: Census Bureau Site Visit Over 40 MU researchers meet with Census Bureau staff August 2014: NSF Accepts Proposal MU commits $1 million over 5 years to support MU RDC RDC in KC and branch at MU to be in place by November PEHC

5 5 PEHC

6 What is a Census RDC? A secure environment where researchers can undertake analysis of non-public data from federal agencies MU will be one of 24 RDC sites nationwide by the end of Use of RDC allows for research that would be difficult otherwise

7 PEHC Support of RDC Doctoral Fellowships Grant Writer to help obtain outside funding Small Grants Program Support work preparing proposal for RDC data access and analysis Most grants $10,000-$15,000 Deadline Nov. 15, 2015 For more information: 7 PEHC

8 National Vital Statistics System Data received from 57 vital registration jurisdictions (50 states, DC, NYC, 5 territories) Births and information on the pregnancy and delivery Deaths and information on cause of death Produces state and local level data

9 National Health Interview Survey In person, in home survey collects data on about 100,000 persons in 40,000 responding households per year representative of the non-institutionalized civilian U.S. population Data on: Health status, health conditions, functioning and disability Insurance coverage Access to and use of health services Immunization Health-related behaviors and risk factors Special topics

10 National Health and Nutrition Examination Survey Standardized physical examinations, laboratory tests, personal interviews of around 5,000 persons representative of the non-institutionalized civilian U.S. population Data on: Disease/condition prevalence Nutrition Body measures Growth and development Health-related behaviors Risk factors

11 National Survey of Family Growth In-person interview with men and women years of age, representative of the non-institutionalized civilian U.S. population Data on: Reproductive health Fertility/infertility Contraception Pregnancy Sexual activity Family formation Marriage, divorce, cohabitation

12 State and Local Area Integrated Telephone Survey National Survey of Children s Health (2003, 2007, ) National Survey of Children with Special Health Care Needs (2001, , ) National Survey of Adoptive Parents (2007) National Survey of Adoptive Parents of Children with Special Health Care Needs (2008) Survey of Pathways to Diagnosis and Services (2011) National Survey of Children in Nonparental Care (2013)

13 National Immunization Survey Telephone interview of households and mail survey of doctors and other vaccination providers to monitor immunization coverage for children (NIS: months) and adolescents (NIS-TEEN: years). Data on: Types of immunizations Dates of administration Facility characteristics of vaccination providers

14 National Health Care Surveys Ambulatory and hospital care surveys National Ambulatory Medical Care Survey (NAMCS) Physician offices Community Health Centers National Hospital Ambulatory Medical Care Survey (NHAMCS) Outpatient Clinics Emergency rooms National Hospital Care Survey (NHCS) Inpatient care -- hospitals

15 National Health Care Surveys Long-term care surveys National Survey of Residential Care Facilities (NSRCF) National Home and Hospice Care Survey (NHHCS) National Nursing Home Survey (NNHS) National Study of Long-Term Care Providers (NSLTCP) Bi-annual data collection starting in 2012 Administrative data: home health agencies, nursing homes, hospices Questionnaire data: adult day services centers, residential care communities

16 NCHS Data Linkage Linked data enable broader analyses of factors that influence health and health outcomes Surveys are linked with administrative data such as The National Death Index Claims data from the Centers from Medicare & Medicaid Services Social Security Benefit History data Linked data are accessed through Public-use files NCHS Research Data Center (for restricted use files)

17 Public Types of Data and Data Access Data files: collapsed categories, top and bottom coded Reports Restricted: Direct Identifiers Name, SSN, address Used to create the Linkage Products and removed Never available Restricted: Indirect Identifiers Items that when combined with other information (from NCHS or other sources) could lead to identification Available through the RDC

18 Examples of Restricted Variables Geography, exact dates, more detailed versions of variables (race/ethnicity, industry/occupation) National Health Interview Survey: Preliminary Quarterly Files National Health and Nutrition Examination Survey: Sensitive information for the youth population (e.g., conduct disorder, drug use, sexual behavior) National Survey of Family Growth: Contextual data, Interviewer data State and Local Area Integrated Telephone Survey: Links between the surveys National Health Care Surveys Provider/facility characteristics, contextual information

19 NCHS Linked Data Linked to Mortality NHIS NHANES I Epidemiologic Followup Study (NHEFS) NHANES II, III, SOA, LSOA II NNHS NHHCS 2007 Linked Social Security Benefit History NHIS NHEFS NHANES III, NNHS 1985 LSOA II NNHS Linked to Medicare NHIS NHEFS NHANES III LSOA II Linked to Medicare NHIS NHANES NNHS 2004 Linked to Medicaid NHIS NHEFS NHANES III, NNHS LSOA

20 Steps to gain access Step 1: Develop your proposal. NCHS is happy to help. Step 2: your proposal to Peter Meyer, NCHS RDC Director at who assigns it to an NCHS RDC Analyst. Step 3: The NCHS RDC Analyst reviews the proposal to ensure completeness and clarity. The analyst may contact you with questions and comments.

21 Steps to gain access Step 4: The NCHS RDC Analyst sends the proposal to the Review Committee (4-6 individuals). The analyst may contact you with questions and comments. Step 5: The NCHS RDC Analyst will contact you and the University of Missouri Census Administrator when the proposal has been approved and will send you an invoice to complete the data management ($750 a day) for your project.

22 Steps to gain access Step 6: The NCHS RDC Analyst gathers the restricted variables while you: For NCHS: create the public use and any non-nchs datasets, complete and return the confidentiality training/paperwork, send payment For Census: apply for Special Sworn Status and complete Account and Data Request Form (ADRF). Step 7: RDC administrator creates and transfers the dataset to the Census RDC

23 What kind of information does RDC want to see in a proposal? A. Abstract B. Research Question C. Background D. Public Health Benefit E. Data Requirements F. Methodology G. Output H. Data Dictionary I. References J. Other Authors K. Resumes Keep these sections short Make the questions specific methods, output, and data requirements should align with these questions Disclosure risk Technical feasibility Public health benefit

24 Are there any good examples? See RDC Developing the Proposal webpage: 1) Proposal Format (PDF and Word formats) A lot of information included to help researchers 2) Sample Proposal NHANES 3) Creating the Data Dictionary Instructions and examples

25 Any mistakes to avoid? Make sure the proposal is complete before you submit it to the RDC Do not request too many or too few variables than you need for your project Really helpful to start working with the data and create public dataset before submitting the proposal

26 Once a proposal is received, when can you start doing analyses? NCHS tries to have a decision on your proposal in 6-8 weeks SSS can take 3 months after the proposal has been approved Also, depends on you: Responding to our questions/comments Creating the public use datasets Sending payment for data management fees Completing NCHS and Census paperwork

27 Who do I contact to schedule time at the RDC? Please contact our Census RDC Administrator to schedule time in the lab. Please note that you are only authorized to use the lab when a Census RDC Administrator or Designee is present.

28 Is there anything else I need to do? Follow the NCHS RDC Publishing Guidelines: 1. Adhere to any additional requirements specified in the approval 2. Do not reveal any information that could identify an individual or establishment 3. Do not reveal any information that could be used to identify geographic areas where respondents live or were sampled unless you specifically have permission to make estimates for those areas 4. Do not reveal any information about specific dates 5. Do not reveal any information from external sources of data that have been merged to NCHS data based on temporal or geographic components that could facilitate the identification of areas, dates, or individuals 6. Ask the RDC staff for help any time you have questions

29 RDC Proposal Process

30 Overview 1. Researchers interested in using restricted-use Census datasets will need to go through a proposal process and obtain Special Sworn Status (SSS) 2. Jake Cronin is the RDC Administrator that works full-time on campus to help researchers through the proposal process.

31 Special Sworn Status After the proposal has been approved by Census Bureau, researchers will need to obtain Special Sworn Status (SSS) in order to physically access the RDC. SSS entails a background check and an oath to project the confidentiality of the data forever Requires completing paperwork, getting fingerprints, online trainings, and identification verifications.

32 Working in an RDC Only researchers with approved projects and Special Sworn Status are allowed in the RDC. Researchers will work on thin client computers that tunnel into secure Census servers. No data is stored on local machines All analysis must be done in the RDC. Output cannot be removed from the RDC prior to a formal disclosure avoidance review

33 Disclosure Avoidance Review There is a formal process to remove aggregated output out of the RDC. That process typically takes about three weeks. This is the only approved process for moving results from inside the confidential RDC to the outside public domain. The general idea is to ensure that no identifiable information on an establishment, firm, person, or household is disclosed. In the proposal phase, it is important to think ahead about the output of the research and if it poses any disclosure risk

34 More Information For information on restricted-use data available through the RDC: MU RDC Administrator: (when RDC opens)

35 Further Information Population, Education, and Health Center Colleen Heflin, co-director, Peter Mueser, co-director, Resources Center for Economic Studies (CES) AHRQ jsp NCHS