Department of Health and Human Services

National Committee on Vital and Health Statistics

Working Group on HHS Data Access and Use

September 17, 2013

National Center for Health Statistics, Hyattsville, MD

Meeting Minutes

The Working Group on HHS Data Access and Use was convened on the afternoon of September 17, 2013, at the National Center for Health Statistics in Hyattsville, MD. The meeting was open to the public. Present:

Working Group members

  • Justine M. Carr, M.D., Chair
  • P. Kenyon Crowley, MBA, MS
  • Bruce Cohen, Ph.D.
  • Leslie Pickering Francis, J.D., Ph.D.
  • Mohit Kaushal, MD
  • Joshua Rosenthal, Ph.D.
  • Walter Suarez, MD
  • Leah Vaughan, MD

Absent:

  • Bill Davenhall
  • M. Chris Gibbons, MD
  • Pete Hudson, MD
  • Patrick Remington, MD
  • Kalahn Taylor-Clark, Ph.D.

Lead Staff and Liaisons

  • Marjorie Greenberg, NCHS, Exec. Secretary
  • Jim Scanlon, ASPE
  • Susan Queen, ASPE
  • Lily Bradley, ASPE

Others  

  • Debbie Jackson, NCHS
  • Katherine Jones, NCHS
  • Marietta Squire, NCHS
  • Raj Chanderraj, NCVHS member
  • Alexandra Goss, NCVHS member
  • Larry Green, NCVHS chair
  • Linda Kloss, NCVHS member
  • Vicky Mays, NCVHS member
  • Ob Soonthornsima, NCVHS member
  • Paul Tang, NCVHS member
  • Vicky Boothe, CDC
  • Tammara Jean Paul, NCHS
  • Michele Plorde,  King County WA
  • Susan Kanaan, NCVHS writer

Note:  The transcript of this meeting is linked to the meeting agenda on http://ncvhs.hhs.gov. Use the meeting date to access the agenda.  

EXECUTIVE SUMMARY

In this half-day meeting, Working Group members and other participants focused on conceptualizing and planning a project to explore innovative uses of HHS data and social media data in a community context. The group agreed on these parameters for the project:

  • 90-day timeframe;
  • Small project;
  • Proof of concept rather than an event;
  • A topic on which people use social media;
  • Partnership with a project that is already under way, with the aim of adding to it.

On that basis, they ultimately decided to explore partnering with a project in Louisville, KY (described below) to look into the utility of combining HHS data and social media data to support a community-based project. In addition, Ms. Bradley briefed the group on several HHS innovation activities including an Innovation Map that she is developing.

DETAILED SUMMARY

Project Brainstorming and Planning

After Dr. Carr reviewed the agenda and objectives for the meeting, Dr. Francis reported that a review of social media privacy policies that she is overseeing is under way and is proving to be complicated. She promised a full report in the near future. As a preface to the major agenda item of this meeting, Dr. Carr reminded the group that it decided at its June meeting to identify a community in which HHS data and “new data” such as social media data could be combined in meaningful ways to inform a local project. The activity was provisionally called a “Solvathon.” A letter was sent via email inviting proposals, and one of the respondants, Michele Plorde of King County, WA, will join this meeting by phone to discuss a possible joint project.

Dr. Carr’s query about the goals of the proposed Solvathon initiated a discussion of the purpose, customer(s), design, methodology, and limitations of the exercise that continued for much of the meeting. The points and questions raised and the themes of the discussion included:

  • Who is the major customer HHS, a community/communities, or combinations of these?
  • What mix of people and expertise are needed to design and conduct the project?
  • Do HHS data exist that are granular enough to inform a neighborhood-level project?
  • How would social media data be used? One suggestion: to help define the problem and target population, be part of the intervention itself, and be used in the outcomes assessment.
  • Questions were raised about selection bias and the integrity and validity of social media data. Some members asserted that identifying and understanding such limitations, as well as the potential of social media data, is a reason to do such a project rather than a reason to avoid doing it.
  • What are the tactics? For example, how will implementation be organized, what infrastructure is needed to support the process, and who will parse through the data?
  • What would success look like?
  • How broadly generalizable do the problem and solution have to be to be worth the effort?
  • Are any communities already using social media and/or other non-traditional data along with traditional-source data? If so, are they in the public domain or the private sector?

Dr. Carr observed that it was proving more challenging than anticipated to identify a community and a problem in which such questions can be explored through a targeted project. The group gradually moved away from the idea of a solvathon and toward framing its project as a proof of concept. The concept to be tested, Dr. Carr said, is whether HHS data can be enhanced by other types of data from non-traditional sources. (It was clarified that the question is not whether a specific combination of data can help solve a specific problem.) Dr. Rosenthal added that the concept to be tested includes “how you walk through the stages” and whether any collateral can be produced to help communities solve actionable problems using HHS and non-traditional data. This includes what can be learned regarding gaps, obstacles, and challenges. He noted that sufficient data density is one of the requirements of such a project.

Several different approaches were suggested in the course of the discussion. Dr. Cohen proposed selecting a community and an issue and taking time to identify the relevant actors and data sources and their potential uses and protections.  Dr. Carr proposed bringing a group of people who are already working on this kind of thing together to discuss their learnings and challenges. In this case, Dr. Rosenthal suggested partnering with HDC. Ms. Bradley wondered about the potential for including case studies. Ms. Greenberg and others suggested starting the process with a hearing or virtual hearing, providing a structured setting in which to learn about the experiences of a few communities. Ms. Goss suggested issuing an RFI to gather this information. Dr. Mays suggested picking a topic that tends to be discussed on social media for example, heat strokes, flu, and obesity; Dr. Carr added football head injuries. Dr. Mays noted the need to be creative with respect to the population being reached out to for example, looking at blogs for discussions within the black community. Mr. Soonthornsima pointed out the need to include natural language and not just clinical and technical terminology. Dr. Green suggested looking at more than one site, to generate useful comparisons.

The group explored two possible community projects. They talked by phone with Michele Plorde, Deputy Director for King County, Washington, Emergency Medical Services. Her department is developing strategies to address variations in EMS call volume, especially for several target groups including seniors with limited English. She had responded to the NCVHS email query because of interest in using social media and social media data in its programs. Dr. Carr promised to get back in touch with her after the Working Group decided on next steps.

Dr. Rosenthal presented a project in Louisville, KY, called New Roots, Inc., that is funded by RWJF and led by Ted Smith of Louisville Metro Innovation and Health. It is taking produce into impoverished inner city neighborhoods and collecting data with which to measure outcomes. Dr. Smith is willing to work with NCVHS to explore what data are available to support the intervention, and specifically the utility of social media, social media data, and HHS data.

Ultimately, the group agreed on the following parameters for the project:

  • 90-day timeframe;
  • Small project;
  • Proof of concept rather than an event;
  • A topic on which people use social media;
  • Partnership with a project that is already under way, with the aim of adding to it.

After further discussion and a number of caveats, a decision was made to move ahead with exploring the Louisville project as described above. Dr. Carr noted that the findings (which Ms. Boothe noted would help “elucidate the complexity”) might make an interesting presentation at the 2014 Datapalooza.

Innovation Map and Other Federal Data Activities – Ms. Bradley

Ms. Bradley reported on several HHS data and networking resources that she has been working on, including an Innovation Map of the US. The map includes “doers,” advisors, convenors, developers, tools, and HIEs. The purpose of the map is so everyone can know about everyone else’s activities and skill sets and where they are located, and be able to coordinate. She asked members to provide information to her on others who should be included on it.

She also briefly mentioned several other activities through which HHS is “learning to be innovative.” One is Project Open data, a website built on GitHub, which has tools that NCVHS groups could take advantage of. She described efforts to create a federated model for data clearinghouses and exchanges. A lot of work is being done on metadata resources, including defining schema and standards for the metadata. A project with the group W-3 is starting with a focus on finding civic services. (The National Human Services Interoperability Architecture, run by the Agency for Children and Families, is relevant.) In other news, CDC has just launched data.cdc.gov, and NIH is considering its own data catalog and has launched a data-to-knowledge initiative. Other programs include usability.gov; a second round of the HHS entrepreneurs program; and the project HHS Ignites, which is encouraging innovation within HHS. Some of these programs are designed to facilitate collaboration between HHS and developers.

Dr. Carr observed that the Working Group could identify HHS activities about which it might provide useful advice to the Department. Dr. Mays suggested having regular updates to the Working Group on Departmental activities of this kind, with explanations about what things mean, why they are important, and to whom. As one example of what can be learned, Ms. Bradley cited the fact that the NCVHS website needs to have metadata added to it so that Google and other search engines can find it. In addition, NCHVS could consider posting its reports and other documents to PubMed, which is fully indexed, and/or to the GSA’s FACA page.

Dr. Carr then adjourned the meeting.

I hereby certify that, to the best of my knowledge, the foregoing summary of minutes is accurate and complete.

/s/

Char – Date