Department of Health and Human Services
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
February 25-26, 2009
Hubert H. Humphrey Building
The National Committee on Vital and Health Statistics was convened on February 25-26, 2009, at the HHS Building in Washington, D.C. The meeting was open to the public. Present:
- Harry Reynolds, Chair
- Jeffrey Blair, M.B.A.
- Justine M.Carr, M.D.
- Leslie Pickering Francis, J.D., Ph.D.
- Larry A. Green, M.D.
- Mark Hornbrook, Ph.D.
- Garland Land, M.P.H.
- Carol J. McCall, F.S.A., M.A.A.A.
- Blackford Middleton, M.D.
- Sallie Milam, J.D.
- J. Marc Overhage, M.D., Ph.D.
- William J. Scanlon, Ph.D.
- Donald M. Steinwachs, Ph.D.
- Walter Suarez, M.D.
- Paul Tang, M.D.
- Judith Warren, Ph.D., R.N.
- John P. Houston, J.D.
Staff and liaisons
- Marjorie Greenberg, NCHS/CDC, Executive Secretary
- James Scanlon, ASPE, Executive Staff Director
- Jorge Ferrer, VA liaison
- J. Michael Fitzmaurice, AHRQ liaison
- Debbie Jackson, NCHS
- Katherine Jones, NCHS
- Michael DeCarlo, BlueCross BlueShield Assn.
- Allison Viola, AHIMA
- Henry Heffernan, NIH
- Frank Kyle, Amer. Dental Assn.
- Mairi Breen Rothman, Amer. College of Nurse-Midwives
- Catherine Schenck-Yglesias, Johns Hopkins Univ.
- Dan Rode, AHIMA
- Melvyn Greberman, Public Health Resources, LLC
- Shobha Srinivasan, NCI
- Carol Bickford, Amer. Nurses Assn.
- Melak Tadesse, Catholic University
- Maria Friedman, SureScripts-RxHub
- Ken Whittemore, SureScripts-RxHub
- April Falconi, Academy Health
Note: The transcript of this meeting and speakers’ slides (when used) are posted on the NCVHS Web site, http://ncvhs.roseliassociates.com. Use the meeting date to locate them.
The Committee unanimously passed a motion to approve sending a letter to the Secretary, after further wordsmithing, that encourages adequate funding and resources for the nation’s health statistics enterprise to ensure the nation’s capacity to measure and realize the return on the new investments in health IT, comparative effectiveness, and prevention and wellness.
Update on the Department, the Transition—Jim Scanlon, ASPE
Mr. Scanlon briefed the Committee on the transition, the current budget and plans for the next one, and (in some detail) relevant aspects of The American Recovery and Reinvestment Act, which provides a large infusion of funds to HHS and its agencies and activities. The discussion period with NCVHS members focused on next steps.
CMS Update—Karen Trudel
Ms. Trudel reported on the Centers’ work on e-prescribing and PHR pilots in SC and AZ (both of which she described as exciting). She said she would provide a fact sheet on the PHR project to the Committee. She also reported that CMS published the final rules on ICD-10 code sets and new 5010 transactions in January. Committee members commented on ways to frame the transition to ICD-10 as a strategic advantage, the possible use of incentives, and the role of SNOMED CT in the transition to ICD-10-CM and ICD-10-PCS. Regarding the Recovery Act, Ms. Trudel said the adoption incentives raise a number of “interesting issues” CMS will have to tackle, such as how to define “meaningful use.” The regulation agenda is being redefined by the new administration.
National EHealth Collaborative Update (NeHC) —Laura Miller, Interim Executive Director
Ms. Miller reported on the early days of NeHC. She noted the public/private nature of its mission to help accelerate progress toward the effective use of health IT. It is bringing together health care stakeholders in a membership organization to identify needs, focus expertise and resources on prioritizing and addressing the needs, and ensure extensible solutions that can be integrated. She described the national interoperability prioritization process, whereby stakeholders submit value cases, prioritize their needs, and submit them for national action. Dr. Tang, who chairs its Value Case Committee, explained that the value case is one of the most important propositions of the new organization. NeHC is also working on the NHIN governance framework. Mr. Blair serves on a broadly representative workgroup on that subject, convened jointly with ONC. Ms. Miller said it is essential to have a governance framework that can support both the emergent and mature NHIN.
ONC Update—Rob Kolodner, MD
Dr. Kolodner briefed the Committee on the Health Information Technology (“HITECH”) Act and the status of standards acceptance. The HITECH Act has five sections, the first of which establishes the Office of the National Coordinator and establishes the HIT policy and standards committees (FACAs) and a process for adopting standards. Other sections have to do with NIST, infrastructure, privacy, and incentives for Medicare and Medicaid providers (which start in 2011). All of this has to happen quickly, because the initial set of standards is to be adopted by the end of 2009. The work on standards to date lays the groundwork for the next stage. Even these large amounts of money are not enough to cover what is needed, and the Department must figure out how to structure use of the funds to get “ten dollars of impact for every dollar we put out.” He also noted the importance of creating an appropriate and constructive standards process that “builds toward a more common structure.”
Nationwide Privacy and Security Update—Jodi Daniel, ONC
Ms. Daniel reviewed the privacy and security pieces of the new law. The most significant change in HIPAA rules concerns how they apply, especially with respect to business associates of covered entities. The Department will be figuring out how to incorporate the new changes into existing regulatory structures. The statute explicitly states that health information exchange organizations, RHIOs, e-prescribing gateways and certain specified PHR vendors are business associates under the current HIPAA regime, and thus HIPAA rules on business associates apply to them. In addition, it has provisions about PHR breach notification. Ms. Daniel also described a Privacy and Security Framework that HHS put out in December 2008 as high-level principles for people establishing health information exchanges. It is designed to “overlay” the patchwork of Federal and state laws. ONC and OCR also have developed tools to clarify the relationship between the privacy rules and health IT and information exchange.
De-Identification and Uses of Anonymized Data under HIPAA —Deven McGraw, Center for Democracy and Technology (CDT)
To address questions about de-identification, CDT convened a workshop on September 26, 2008. There were expert panels on current de-identification practices, common uses of de-identified data, and policy implications. Dr. Carr was on the second panel, and former NCVHS member Mark Rothstein was on the third. CDT is releasing a paper based on the workshop, and Ms. McGraw reviewed some of the key points. A core point is that the current HIPAA standard for de-identification has limited utility for many data uses. One option in this area (recommended by Mr. Rothstein) is a requirement to use “the least identifiable data possible.” The goal of ensuring a low risk of re-identification is becoming more difficult, partly because of the increased availability of data. Ms. McGraw said CDT would like to continue to work with NCVHS on this issue, as part of its contribution to the regulatory and guidance process. Dr. Carr noted that the complementarity between CDT and NCVHS in this area of common concern is “a great way of working together.”
Data Stewardship Primer—Mr. Reynolds and Ms. Kanaan
Mr. Reynolds and Ms. Kanaan discussed plans for a primer on health data stewardship, to be produced by the Committee and written by Ms. Kanaan. The idea is to help people in the field understand the key concept and principles of data stewardship and know how to find the essential resources on the subject. Members praised the idea of a primer as a potentially useful NCVHS contribution, and offered suggestions on content and dissemination.
Committee Discussion: Strategic Planning
NCVHS members and staff discussed how the Committee can make the greatest contribution in the current environment. In view of the rapid pace of developments, they considered what should be done in the short term, and how the customary long and broad NCVHS view can be most helpful. Clarifying the respective roles of the advisory committees and identifying where NCVHS offers to take the lead were identified as priorities, along with seizing the timing to recommend funding for essential surveys and data systems.
Implementing Standards: A View from the UK and IHTSDO —Prof. Martin Severs
Prof. Severs is Chairman of the National Health Service (NHS) Information Standards Board for Health and Social Care (ISB HaSC), and the UK representative to the Management Board of the International Health Terminology Standards Development Organization (IHTSDO) as well as its Chairman. He gave perspectives from both the UK body and the IHTSDO, and offered personal conclusions. The UK’s Information Standards Board focuses on implementability, interoperability, and safety, and it manages the tension between major stakeholders. He stressed that “the value of the standard lies only in effective implementation.” He then gave a brief overview of the IHTSDO, a Danish Association that owns the intellectual property in SNOMED-CT. He described its governance structures and mechanisms, stressing the importance of governance, and outlined its major accomplishments during its two years of existence. Prof. Severs’ personal conclusions include the point that “fundamental standards” (one of three types he outlined) should be international and internationally controlled; as few as possible; and the owners should have a set of attributes that he outlined. He stressed the importance of sharing risks, costs, and benefits and the need to think through information strategies and policy for clinical practice, and he noted the need for “an end-to-end national information description.”
Health Statistics for the 21st Century—Dan Friedman, PhD
Dr. Friedman was a co-author of the 2002 NCVHS report, Shaping a Health Statistics Vision for the 21st Century, developed collaboratively by NCVHS, NCHS, and the HHS Data Council. He and co-author Gibson Parrish, MD, have been asked to take a new look at the report, assess progress on the recommendations, and focus on updating and revising the vision in light of national and international developments since 2002. He reviewed the timetable for the project, which plans a draft working paper in September, and asked members to rank the original 36 recommendations, as informal input for the Subcommittee on Population Health.
NCHS/BSC Update on Vital Statistics—Dr. Scanlon, Mr. Land
The NCHS Board of Scientific Counselors recently discussed the use of EHRs in the vital statistics system. Mr. Land said NAPHSIS staff have been analyzing how to import medical record data into electronic birth and death systems. The area contains both potential and many issues, some of which he described. Ms. Greenberg agreed with the issues he raised, and said she believes EHRs can be useful in informing birth and death records without replacing them. The important thing is to design EHR systems to meet these information needs, and to design electronic birth and death records to use the information.
Emerging and Innovative Sources of Health Data
1. Matthew Holt, Health 2.0
Mr. Holt owns and operates The Health Care Blog and is the co-founder of Health 2.0. He talked about Web 2.0, Health 2.0, and the early uses of online activity for data collection and research. Web 2.0, which began to emerge around 2005, is characterized by social networks such as Wikipedia or Facebook and tool-sharing. To explain Health 2.0, he used two lists of constituent parts, noting that his view of the phenomenon is evolving because it is a moving target. His latest version states that patients are increasingly guiding their own care—the locus of control of health care is shifting. The components are:
- Personalized search that looks into the long tail, but cares about the user experience;
- Communities that capture the accumulated knowledge of patients and caregivers and clinicians, and explain it to the world;
- Intelligent tools for content delivery and transactions; and
- Better integration of data with content.
He offered several examples of online communities that are using their assets to compile data on “what works and what doesn’t,” and he commented on the importance of integrating such activities with traditional health care. His presentation stimulated a wide-ranging discussion with NCVHS members. Asked about the key ways to shape the NHIN to capitalize on these developments, he stressed the importance of integrating information from all possible sources, including patient communities, and of using these sources to capture a broad range of data about people’s lives and experiences.
2. John Halamka, MD, CareGroup and Harvard Medical School
Dr. Halamka listed 14 non-traditional and emerging data sources and noted that many are more clinically oriented, and can better assist quality analysis, than traditional sources. He gave several examples of innovative uses of data and novel data sources and exchanges, many of them in his region. In summary, he said a number of interesting non-traditional approaches to gathering data and measuring quality are happening in his region. He predicted that the PHR, especially the aggregation of data by patients and sharing that data, and the use of device data will be very important to emerging sources of quality measures.
Letter by the Subcommittee on Population Health
(See ACTIONS, above.)
UPDATE ON THE DEPARTMENT, THE TRANSITION—Jim Scanlon, ASPE
This is the first NCVHS full Committee meeting since President Obama was sworn in. While HHS awaits a Secretary, some political staff and leaders are already in place at HHS. All HHS agencies and offices are headed by senior career deputy officials with a lot of experience. This year’s budget process is more complicated than usual. Congress is acting on an Omnibus Budget Bill for the rest of FY2009, which is “reasonable for HHS.” For FY2010, the President has given an overview of the planned initiatives and announced that health reform will be a priority. That budget probably will be submitted in April.
The American Recovery and Reinvestment Act (“the Recovery Act”) provided a large infusion of funds to HHS and other agencies. Most of it is one-time spending, intended to “do good” and also “be transformative” over a short time period. There is tremendous pressure to get the money out to the states. HHS received about $137 b. over two years, and the Act provides $20 b. for health IT, of which $2 b. goes to ONC and about 17.6 b. comes downstream through Medicare and Medicaid to incent EHR adoption, beginning in 2011. There is a lot of work to be done before that, in a short period of time. Mr. Scanlon gave more of the details of the Act, which codifies and establishes ONC in the statute. It also sets out a process to promote, adopt and approve standards, building on the existing foundation. It creates a health IT policy committee and a standards committee, both FACAs; and it funds regional education on health IT use. $300 m. is earmarked for regional or sub-national efforts, and there are new privacy and security provisions. On population health, the Act creates a prevention and wellness fund in HHS for $1 b.; the Department has not yet determined how much of the money will go to NCHS. The Act also funds comparative effectiveness research at $1.1 b., and $10 b. goes to NIH.
In addition to several detail questions about the Recovery Act, members raised questions about the anticipated roles and duties of the several advisory committees. Mr. Scanlon recommended waiting for further developments and clarifications. Mr. Reynolds said he had talked with Dr. Kolodner, and this question would be discussed later in the meeting. Dr. Green called attention to the idea of regional “extension offices,” which could help overcome barriers to health IT adoption. Mr. Scanlon said this would be funded out of the $2 b. allocated to ONC. Ms. McCall reiterated his advice about doing lots of listening, adding that NCVHS should focus on where it can add the most value. She asked about funding for the health statistics enterprise for the 21st century, noting that progress depends on data as much as on technology. Mr. Scanlon agreed, adding that while health statistics is not mentioned explicitly, it is implied in several areas. The Committee will have a role in interpreting and implementing the legislation. Mr. Reynolds noted the relevance of the high-level discussions at the previous day’s Standards Subcommittee hearing, to which several thought leaders contributed. (See the hearing summary on the NCVHS Website.)
Returning to his presentation, Mr. Scanlon noted that the Subcommittee on Population Health was convening a panel on February 27 to look at the capacity for modeling for health policy data. The Data Council is working on the same subject, and looks forward to the Subcommittee’s recommendations.
CMS UPDATE—Karen Trudel
Ms. Trudel began with a briefing on e-prescribing. CMS is working on a pilot (conducted by Rand) to test two standards, RxNorm terminology and the structured and codified sig. CMS has determined that more work is needed on electronic prior authorization. It is working with the DEA, ONC and others on e-prescribing for controlled substances.
She described the Center’s work on personal health records (PHRs) as very exciting. The South Carolina project, MyPHRSC, is finishing its first year of operation with 4,000 registered users (of 100,000 beneficiaries). In January it added a partnership with TRICARE and rolled out a new PHR pilot, PHRChoice, in Arizona and Utah. Unlike the SC project, this one allows people to choose among four pre-qualified vendors—NoMoreClipboard, PassportMD, Healthtrio, and GoogleHealth. There are 700 registrants so far. ASPE is helping with the evaluations. In view of NCVHS interest, Ms. Trudel offered to provide a fact sheet on the project to the Committee.
On HIPAA, CMS published the final rules on ICD-10 code sets and new 5010 transactions in January. It benefited from both extensive analysis in industry comments and from NCVHS recommendations. The final rule significantly changes the timetables for the 5010, to January 2012, and ICD-10 code sets, to October 2013, and it is starting to see a coalescence around those dates. She acknowledged that this is in ways an unfunded mandate. Ms. Greenberg added that as organizations are developing EHRs, it is helpful to know that the country is moving to ICD-10 code sets. The funding for EHRs could facilitate mappings with SNOMED CT and related vocabulary activities. The goal is to make the transition to ICD-11, anticipated for around 2020, as seamless as possible.
Committee members commented on ways to frame the transition to ICD-10 code sets as a strategic advantage, the possible use of incentives, and the role of SNOMED CT in the transition.
Regarding the Recovery Act, Ms Trudel said the adoption incentives raised a number of “interesting issues” CMS would have to tackle, such as how to define “meaningful use.” Moreover, she does not consider the incentives “as a one-time shot”; rather, it is clear that Congress “intended us to raise the bar over time in terms of meaningful use.” She agreed with members’ comments about potential synergies over time between health IT adoption and the transition to ICD-10 code sets. Dr. Warren commented on the good feedback loop in the e-prescribing pilot and the importance of continuing to have such feedback loops in the future. She also noted Dr. Detmer’s emphasis in the Standards hearing on communications standards. Asked about other proposed rules, Ms. Trudel said the regulation agenda is being redefined by the new administration.
Regarding the transition to ICD-10 code sets, Dr. Scanlon asserted that the revenues in the health sector have been so large that “it’s not unreasonable to say, ‘do more.'” Nor is it unreasonable to say that people cannot participate unless they do certain things. He cautioned against timidity in “what we’re asking for.”
NATIONAL EHEALTH COLLABORATIVE UPDATE (NeHC) —Laura Miller, Interim Executive Director
Ms. Miller said NeHC is ready to help with the new mandates from the President. She stressed the public/private nature of its mission to help accelerate progress toward the effective use of health IT. It is bringing together health care stakeholders to identify needs, focus expertise and resources on prioritizing and addressing the needs, and ensure extensible solutions that can be integrated. NeHC has been a membership organization for just a few weeks, and there is already evidence of interest. It has committees on policy, mission development, and value case—the latter chaired by Dr. Tang.
Ms. Miller described the national interoperability prioritization process, whereby stakeholders submit value cases, prioritize their needs, and submit them for national action. Value cases can go beyond standards harmonization to other factors that can also be barriers to implementation, including model processes, best practices, and frameworks. The purposes of prioritizing the value cases are to articulate how the initiatives increase the quality and/or reduce the cost of care, to generate meaningful value propositions, to address the business case, and to drive health IT adoption. She described the process involved and showed a template for value case proposals.
Dr. Tang explained that the value case is one of the most important propositions of the new organization. He stressed that the process is stakeholder-initiated and driven by clinical and business needs, based on the interoperability problems encountered in the field. The expectation is that it will not be primarily about standards. The idea is to identify common problems and “solve them as a nation” through breakthrough strategies developed by groups of stakeholders. As an example, he cited quality measures derived from EHRs.
Continuing her presentation, Ms. Miller said NeHC is also dealing with the NHIN governance framework. Mr. Blair serves on a broadly representative workgroup on that subject, convened jointly with ONC. She said it is essential to have a governance framework that can support both the emergent and mature NHIN. The core principle is that governance should support the needs of participants and be collaborative. The governance objectives include maximizing the likelihood of preserving privacy and security, fostering effective use of health IT, enhancing trust, and creating a sustainable business model. A set of principles have been identified to guide decisions about functions and structures. They include distributed and representative governance, transparency, responsiveness, and accountability. Seven key functions for the governance framework to accomplish have been identified, starting with developing strategic directions. These functions will be fleshed out in the governance structure.
The next steps for NeHC include getting feedback from the federal Interim Interagency Coordinator Group, briefing key stakeholder groups, and releasing a draft governance framework for public feedback. After this and review by ONC, the framework will be finalized.
In conclusion, Ms. Miller, noting that “we’re at the tipping point,” said we need wise investment and accountability to transform health care in the U.S. NeHC expects to issue a call for value cases by the end of February.
ONC UPDATE—Rob Kolodner, MD
Dr. Kolodner briefed the Committee on the Health Information Technology (“HITECH”) Act—especially how ONC plans to approach it—and the status of standards acceptance. The HITECH Act has five sections, the first of which establishes the Office of the National Coordinator and establishes the HIT policy and standards committees (FACAs) and a process for adopting standards. Other sections have to do with NIST, infrastructure, privacy, and incentives for Medicare and Medicaid providers (which start in 2011). ONC must set up two FACAs. Congress was explicit about who nominates the members of the policy one; the Department has discretion on the standards committee. There are also statements about the role of NeHC. All of this has to happen quickly, because the initial set of standards are to be adopted by the end of 2009, at least through interim final rulemaking. Dr. Kolodner reviewed the status of standards that have been through the AHIC and acceptance process to date, some of which will provide the basis for moving forward.
Funding is in six grant areas, including one for NIST to spark innovation related to health information exchange. ONC will work closely with NIST on it. There is immediate funding to strengthen the health IT infrastructure, using existing agencies’ grants offices as vehicles. ONC will also have a small grants office for accountability and reporting. The bill talks about HIT architecture training and dissemination of information on best practices to integrate IT. One part of the grant is for setting up regional “extension centers,” with a parent center for support, training, and technical assistance. Their number and structure are yet to be determined. There are state grants for health information exchange; ONC intends to look at many models regarding how to use those funds. There is funding for strategic planning by nonfederal participants, consistent with the Federal HIT strategic plan. Another grant type is for revolving loan programs for states and Indian tribes, plus education and demonstration programs. Another is for investment in medical informatics education for health and/or IT professionals.
As a general point, Dr. Kolodner said that even these large amounts of money are not enough to cover what is needed, and the Department must figure out how to structure use of the funds to get “ten dollars of impact for every dollar we put out.” He also stressed the importance of creating an appropriate and constructive standards process that “builds toward a more common structure.”
NATIONWIDE PRIVACY AND SECURITY UPDATE—Jodi Daniel, ONC
Ms. Daniel reviewed the privacy and security pieces of the statute. Most of the provisions modify existing HIPAA rules, but in some areas Congress took a different approach or added new things (notably, breach notification). The most significant change in the HIPAA rules concerns how they apply, especially with respect to business associates of covered entities. The Department will be figuring out how to incorporate the new changes into existing regulatory structures. The statute explicitly states that health information exchange organizations, RHIOs, e-prescribing gateways and certain specified PHR vendors are business associates under the current HIPAA regime, and thus HIPAA rules on business associates apply to them. The new law also has provisions about PHR breach notification. The Department will work with the Federal Trade Commission on these issues.
The privacy section also has enhanced enforcement provisions for HIPAA, related to both the approach to enforcement and the penalty structure. There is a lot of work to be done to make enforcement practices consistent across the board. There is also a requirement that the Department do education on the privacy of health information. Finally, there is guidance on defining unsecured protected health information for purposes of the breach notification provisions, and on the most appropriate technical safeguards.
The Department put out a “Privacy and Security Framework” in December 2008 to provide a set of high-level principles for people establishing health information exchanges. The Framework is designed to “overlay” the patchwork of Federal and state laws. The developers struggled to get the right balance between appropriate access to information and protection of information. They looked at a wide range of existing principles and fair information practices, including some international principles.
The Framework’s preamble covers the purpose and background, ONC’s methodology in developing the principles, and summarizes the principles. Part 2 of the document presents and discusses the principles themselves: individual access; correction; openness and transparency; individual choice; collection, use, and disclosure limitation; data quality and integrity; safeguards; and accountability. Ms. Daniels briefly discussed each of them, noting that individual choice and the limitations on collection, use and disclosure go hand in hand. She explained that the Framework is offered as guidance, not a mandate, in the hope that it will be useful and “used universally” in moving toward a “trust framework.”
ONC also developed tools, working with OCR, to clarify areas of confusion about the relationship between the privacy rules and health IT and information exchange. They include a security guide to help small health care providers adopting new technologies and a draft PHR model privacy notice modeled on the nutrition facts labels on food products.
In the discussion period for the two ONC briefings, members had questions and comments on clarifying “business associate” and other aspects of the new privacy provisions. Ms. Daniel and Dr. Francis agreed to talk later about guidance for the Subcommittee on Privacy and Security as it plans future activities. Dr. Green wondered about how HIPAA can be fixed to “align with what we’re about to accelerate”; Ms. Daniel observed that this is a question for Congress.
Dr. Kolodner acknowledged that ONC knows that the respective roles of NCVHS and the new FACAs have to be worked out and clearly communicated. Given the National Committee’s statutory authority in the areas of HIPAA and e-prescribing and their bearing on the effective use criteria under development, there are questions to be worked out about the relationship between the new law and the movement of claims data and other things that are part of HIPAA.
Dr. Tang expressed appreciation for the HHS Privacy Framework and suggested that the NCVHS data stewardship report and its letters to the Secretary on privacy and security, as well as new contributions by NCVHS, may be useful as this area is fleshed out. Echoing Dr. Kolodner, Ms. Daniel affirmed that ONC expects these to be collaborative efforts. Ms. McCall pointed out that there will be “new players” or “entities” in the health ecosystem, and what is important is to anticipate and embrace that fact. She recommended use of the concept of a data steward.
Mr. Reynolds told Ms. Miller that NCVHS would watch for ways to cooperate with and support the work of the NeHC, for maximum effectiveness of both. He noted that Dr. Tang is well positioned to be “the translator” between the two bodies.
DE-IDENTIFICATION AND USES OF ANONYMIZED DATA UNDER HIPAA —Deven McGraw, Center for Democracy and Technology (CDT)
Ms. McGraw said CDT believes health IT has tremendous potential to improve health care quality, reduce costs, and save lives. Consumers have serious privacy concerns, and the organization is trying to develop and promote workable privacy and security policy solutions for electronic personal health information that protect privacy while enabling information use and exchange. To this end, it typically explores “some of the more challenging issues”; for example, it put out a paper on the role of consent. Ms. McGraw recited a list of questions about de-identification that CDT wanted to address. For that purpose, it convened a workshop on September 26, 2008. There were expert panels on current de-identification practices, common uses of de-identified data, and policy implications. (Dr. Carr was on the second panel, and former NCVHS member Mark Rothstein was on the third.)
CDT is releasing a paper based on the workshop. Ms. McGraw reviewed some of the key points. First, the HIPAA de-identification standard needs to be reviewed on an ongoing basis and updated to reflect ongoing changes in usage and needs. The general sense at the workshop was that at least some entities that hold data need an easy-to-apply standard to assure that the data are truly de-identified. Publicly beneficial uses of de-identified data should be permitted. However, the full universe of uses is unknown. We need better methods for holding data use “at the ends” accountable. There is no public accountability if the data are re-identified and used for purposes other than those originally intended. The current HIPAA standard for de-identification has limited utility for many data uses; and this is true as well for “limited data sets.”
There are a few options for data anonymization in addition to those in HIPAA, which do not serve many purposes. De-identified data are often not useful for research, public health, and quality purposes; and limited data sets are still rigid and not useful for many important purposes. Another option (recommended by Mr. Rothstein) is a requirement to use “the least identifiable data possible.” The goal of ensuring a low risk of re-identification is becoming more difficult, partly because of the increased availability of data. One suggestion at the workshop was to require more transparency to the public on how the data are used, and perhaps more tracking and monitoring, as well as data stewardship entities. There needs to be an ongoing review of the safe harbor, and more rigor in the statistical method. There was general agreement at the workshop that more options are needed for data anonymization. Ms. McGraw noted that the 2009 Recovery Act provides further opportunities to probe this issue, including direction to the new advisory committees to consider NCVHS recommendations. She said CDT would like to continue to work with NCVHS on this issue, as part of its contribution to the regulatory and guidance process.
Dr. Carr commented that this was a “terrific conference” that covered a great deal of ground in a short period. She noted that this complementarity between CDT and NCVHS is “a great way of working together” on a matter of common concern.
Dr. Suarez reported that HITSP has produced constructs establishing harmonized standards for anonymizing and pseudo-anonymizing data, and they have been published and recognized. Dr. Frances noted that different uses of data will probably require different levels of anonymization or identifiability. Dr. Steinwachs mentioned data centers as one mechanism, and Ms. McCall imagined that this could evolve into a type of service entity. Mr. Reynolds noted what a fast-moving and evolving set of concepts and issues this is, presenting challenges for educating the industry, the public, and other stakeholders. Dr. Green noted the varied levels of risk to the person attached to the data, and levels of severity of consequences for them. He cautioned against a “one-size-fits-all” approach and advocated some form of stratification based on levels of risk of harm to the individual. He and Ms. McGraw discussed notions of risk and harm, and she and Ms. Bernstein referenced the work of Cynthia Dwork regarding risk. Ms. McCall noted the importance of defining “what we mean by privacy,” since the term is used to mean many different things.
DATA STEWARDSHIP PRIMER—Mr. Reynolds and Ms. Kanaan
Mr. Reynolds and Ms. Kanaan discussed a draft outline for a primer on health data stewardship, to be produced by the Committee and written by Ms. Kanaan. Observing that NCVHS is known for making difficult subjects clear, Mr. Reynolds explained that the idea behind this project is to help people in the field understand the key concept and principles of data stewardship and know how to find the essential resources on the subject by NCVHS and others. He added that it is not an easy concept to understand. The intended audience encompasses the general health care industry, from small doctors’ offices to large institutions—and particularly “naïve readers”—to “help an industry learn.” He emphasized that this is intended as a resource, not to convey NCVHS recommendations or positions.
Ms. Kanaan described the current plans for the primer and invited suggestions from Committee members. The comments and suggestions were as follows:
- Start with what data stewardship is and why it is important, including the idea of coexisting tensions.
- Note the sense of urgency with the acceleration of HIT.
- Include a call to action for the reader.
- Cover the harm that can come to people from use of the wrong data or sending data to the wrong place; i.e., data stewardship is not just about privacy but also about the correct use of data and the idea of data integrity.
- Address how we assure that a data holder is being a good steward—i.e., something on governance (if any of the source documents cover this).
- Include privacy frameworks from around the world. (see ISTPA report).
- The principles should “appeal to a reasonable man.”
- Include an example/scenario to aid understanding, as IOM reports do.
- Be explicit about security, and that part of stewardship is ensuring secure procedures and policies.
- Include the concept that successful communication is part of stewardship, involving respect for people at both ends of the communication line.
- Keep it simple.
Several members praised the idea of a primer as a potentially useful NCVHS contribution and possible “market niche” for its products. Dr. Warren noted that students are a potential audience for the primer. Members discussed the possibility of creating other simplified vehicles on data stewardship, such as fact sheets and even “refrigerator magnets” for computers (“Do unto the data of others as you would have done unto yours,” per Dr. Tang), or a special Website, to help get the word out. A suggestion to publish the primer in a journal was deferred until it is completed.
COMMITTEE DISCUSSION OF STRATEGIC PLANNING
Mr. Reynolds invited members to discuss “how we play and what we play” with respect to the developments heard earlier in the day.
Dr. Carr recalled Mr. Scanlon’s advice a year ago not to get “locked into agendas until we saw what was unfolding”; now the time of “unfolding” has arrived. The NCVHS themes that have reemerged include HIPAA, stewardship, and PHRs.
Dr. Green commented on the great value and richness of the previous day’s Standards Subcommittee hearing. Noting the high degree of overlap among the subcommittees’ domains these days and the significance and complexity of the subject matter, he wondered about the Committee’s ability to subdivide into small groups and “do our work” and then come back and “make sense of it” in short periods of time. He proposed, instead, functioning “much of the time as a Committee of the whole in order to bring all these perspectives into play.” Dr. Warren agreed, adding that the Standards Subcommittee would share the hearing summary with all NCVHS members so they can educate themselves on the content. Mr. Reynolds urged members to follow Dr. Green’s example and attend the meetings of other subcommittees.
Dr. Suarez pointed to the opportunity to “make bold statements” about the issues at hand, noting the time challenges given the timeframes mandated in the Recovery Act. He urged finding a way to make recommendations in the next six to eight weeks. Mr. Blair said there are plans to share the preliminary findings with ONC after each of the four Standards hearing days. Responding to Dr. Suarez, Dr. Warren reminded the group of the Committee’s “ability for the long vision,” which should not be lost out of concern for “what’s happening in the next 90 days.”
Ms. Greenberg noted the abundant evidence that there is a role for NCVHS in the current environment. One potential contribution, for example, is to broaden the prevailing discussion from health care to health and population health. She agreed with Dr. Suarez that the Committee should move rapidly to be relevant in areas where “the train is moving,” such as by clarifying NCVHS roles vis a vis the new advisory groups and suggesting where NCVHS is willing to take the lead.
Mr. Land expressed concern about seizing any available opportunities in the stimulus funding to save national surveys, and he suggested writing a letter to that effect. Mr. Reynolds welcomed this suggestion and said he’d like to see a list of areas where NCVHS could “make bold statements” like this one, based on prior work. He pointed out that ONC leaders were present for the entire Standards hearing, which probably had an impact on them. Mr. Blair added that ONC and NCVHS have a trusting working relationship. Finally, Mr. Reynolds urged members and subcommittees to “put their passions into action” and decide how to use the current opportunities.
The Committee then recessed into breakout sessions, to reconvene the following day.
IMPLEMENTING STANDARDS: A VIEW FROM THE UK AND IHTSDO —Prof. Martin Severs
Prof. Severs, among other roles, is Chairman of the National Health Service (NHS) Information Standards Board for Health and Social Care (ISB HaSC), and the UK representative to the Management Board of the International Health Terminology Standards Development Organization (IHTSDO) as well as its Chairman. In his remarks to the Committee, after setting the context, he gave the perspectives from the ISB HaSC and the IHTSDO, and then offered some personal conclusions.
The UK thinks in terms of three levels of standards: the framework, the high-level overarching structure from which standards at other levels can be derived and developed; the fundamental standard, which supports operational standards and will have many instantiations; and operational standards, which are detailed and precisely defined for use within specific areas of the NHS. They believe operational standards should have clearly defined and approved requirements and be associated with organizational, technical, and clinical and user implementation guidance. There also have to be conformance criteria to establish that they are being used properly.
The ISB HaSC is a “quasi-independent” advisory body that has evolved since it was set up in 2000. (Prof. Severs stressed that approval systems should evolve with time and learning.) It is the only information standards approval body for the NHS and social care in England, and approves standards based on appraisal that is based on published criteria. The evidence demanded is from implementation evaluation—i.e., does it work? The appraisals are in six domains, to represent different perspectives. The Information Standards Board focuses on implementability, interoperability, and safety, and it manages the tension between major stakeholders. Prof. Severs stressed that “the value of the standard lies only in effective implementation.” He noted that this emphasis on implementability on the part of a “standards approver” differs from that of a standards developer. The Information Standards Board’s workload of standards it has assessed has increased over the years; he showed a chart illustrating that “demand is going up.” He showed two examples of UK operational standards.
Prof. Severs then gave a brief overview of the IHTSDO, which is a Danish Association. He noted the progress in the governance area in recent years, as called for in a December 2002 NCVHS report analyzing various terminologies. The Association owns the intellectual property in SNOMED CT. Individual members (countries, including the U.S.) govern the Association through the Articles of Association. Members do not have a property right in SNOMED CT that they could use to disadvantage others. Most ongoing decision-making is done by the Management Board, which is accountable to the General Assembly, guided by Vision and Values that are part of the explicit governance of the Association. Members pay fees based on a wealth-based World Bank metric. Within a member country, anyone can use SNOMED CT.
In its two years, IHTSDO has maintained the timely release of SNOMED CT twice a year and a smooth transition from old to new owners; set up the IHTSDO as a Danish Association; designed, acquired and will soon implement a replacement tooling infrastructure for design and building a SNOMED CT; and focused the support service effort on improving the quality of SNOMED CT. Prof. Severs stressed that translation and cultural differences are important to address in an international standard. IHTSDO demonstrates a primary citizen focus and intends to continue innovation around that focus. It is attracting new members, and demonstrates a commitment to harmonization and financial balance. Finally, it is demonstrating openness, transparency and fairness. He said the UK is starting to roll out SNOMED CT in primary care systems.
As his personal conclusions, Prof. Sever began by asking what processes are in place to ensure that a country has confidence in its own data; and also whether variation in data reflect differences in practice, information standards, or both—and “how would you know?” On fundamental standards, he said they should be international and internationally controlled; as few as possible; and the owners should have certain attributes (his slides spell out 17 of them, including a strong governance process and an explicit funding stream). He stressed the importance of sharing risks (which should decrease over time), costs, and benefits. He said a “big learning point” in the UK is that “professional practice information management is a bit of a dark art.” He stressed the need to think through information strategies and policy for clinical practice, and noted the need for “an end-to-end national information description.” In conclusion, he reiterated the importance of implementation in its technical, organizational, and user dimensions, and the importance of governance.
Prof. Severs’ presentation stimulated many questions and rich discussion. For example, Dr. Tang pointed out the importance of clinical governance and clinical participation, and he noted the challenge of getting busy clinical users to participate in standards development. Prof. Severs noted that “the science of clinical practice has outstripped the science of recordkeeping.” The challenge, he said, is not with individuals but at the national level, to get “formal professional authority.” Dr. Warren stressed the importance of IHTSDO structures and processes in alleviating people’s initial fears. This in turn, she said, led to a far better product. Asked for his recommendations for the U.S., Prof. Severs suggested a listing of the fundamental standards, managing interrelationships, and emphasizing implementation, to “help people develop things that can be used in real life.” Dr. Middleton commented on the challenges of implementing SNOMED in a clinical information system infrastructure, such as those related to knowledge representation and management. Dr. Severs reiterated his commitment to facilitating implementation, as well as the need for clinical governance to resolve many questions.
HEALTH STATISTICS FOR THE 21ST CENTURY—Dan Friedman, PhD
Dr. Friedman, a former NCVHS member, was a co-author of the 2002 NCVHS report, Shaping a Health Statistics Vision for the 21st Century, which was developed collaboratively by NCVHS, NCHS, and the HHS Data Council at the request of Dr. Sondik. The three-year process involved consultations at ten professional meetings, a workshop, a series of papers, and four regional hearings. The final report included NCVHS recommendations.
Having found a dearth of definitions, the authors began the report by defining health statistics as “numerical data that characterize the health of a population and the influences and actions that affect the health of a population.” They also developed a concept and defined the health statistics enterprise, with three key functions and attributes. The report also includes two conceptual frameworks and models that formed the backbone of the report and the recommendations—one on the influences on the populations health, and the other on the health statistics cycle. The conceptual framework on population health influences focuses on the context of communities, and helps maintain a focus on the data needed for the health statistics enterprise. Dr. Friedman called attention to the fact that the framework on the health statistics cycle includes, among other stages, a focus on evaluating the extent to which user needs are met. It has at its center an “integrating hub.”
The consultative process led to the conclusion that in the US, both health statistics and the health statistics cycle have “a great deal of unrealized potential,” partly due to too narrow a definition that focuses on individual rather than community health and on disease rather than functional status and well-being as well as disease, among other reasons. The result is abundant data but too little information. Also, enterprise-wide coordination and integration efforts are ineffective due to insufficient connections among components.
The Committee and its partners took time to develop ten guiding principles. On that basis, it developed 36 recommendations through an iterative, consultative process with participants.
Dr. Friedman and co-author Gibson Parrish, MD, have been asked to reconsider the health statistics report, assess progress on the recommendations, and focus on updating and revising the vision in light of national and international developments since 2002. The first phase is a literature view, followed by expert interviews and a review of the findings with NCVHS members and staff. Dr. Friedman reviewed the timetable for the project, which includes a draft working paper in September. To that end, he supplied members with a rating sheet and asked them (individually) to rate the 36 recommendations in terms of their continued importance and whatever other criteria they deem relevant. The members then engaged in this informal prioritization exercise, which Dr. Steinwachs said would provide input to the Subcommittee on Population Health.
NCHS/BSC UPDATE ON VITAL STATISTICS—Dr. Scanlon, Mr. Land
The BSC recently discussed the use of EHRs in the vital statistics system. Mr. Land said NAPHSIS staff have been developing a functional analysis of how to import medical record data into electronic birth and death systems. The area contains both potential and many issues, some of which he described. It seems unlikely that all information can be transported from an EHR into a birth certificate. One issue is how to edit the data, a function that is very important because the birth certificate is both a legal and a statistical document. On the death side, the medical record typically does not include the underlying cause of death. Several state experiments along these lines have not been successful. However, there are efforts to set up HL7 standards for the transmission of medical record content to birth and death certificates. He believes this process should go forward and be evaluated. He recommended caution, and added that regulations to ensure that states conform to certain standards would help.
Ms. Greenberg expressed agreement with the issues raised by Mr. Land. Her view is that EHRs can be useful in informing birth and death records, short of replacing them. The important thing, she said, is to design EHR systems to meet these information needs, and to design electronic birth and death records to use the information.
Dr. Sondik commented on several things that must be dealt with and invested in, including electronic vital statistics, “to build a system that really is 21st century.” NCHS had not had the resources to do that, but there is potential to use Stimulus funds for this.
Dr. Green observed that the nation’s vital statistics system is in jeopardy, and that the Stimulus Bill provides an opportunity to resurrect this issue, since there can be no accountability without vital statistics. Thus, this is a timely political issue. He urged that the Committee take immediate action to address this set of conditions.
Dr. Scanlon said the Subcommittee on Population Health plans to bring a draft letter addressing these issues before the full Committee before the end of this meeting. It goes beyond the vital statistics system, which in itself is essential but not sufficient. The general point is that the investment in HIT systems should go hand in hand with an investment in health statistics and vital statistics.
EMERGING AND INNOVATIVE SOURCES OF HEALTH DATA
1. Matthew Holt, Health 2.0
Mr. Holt owns and operates The Health Care Blog and is the co-founder of Health 2.0. (This summary will be more meaningful in conjunction with his 32 slides, which are posted on the NCVHS Website.) He has been a health services researcher and also worked at the Institute for the Future, and has studied the impact of the Internet on health care. He holds a conference twice a year for Health 2.0. He said he would talk about Web 2.0, Health 2.0, and the “very early-stage issues of using what’s going on online … as a venue for data collection and research.” He stressed that at this point, much of this is conjecture.
He observed that the worlds of health care IT and Web 2.0 are surprisingly unaware of each other. Web 2.0, which began to emerge around 2005, is characterized by social networks such as Wikipedia or Facebook and tool-sharing. These changes have happened very rapidly, with enormous uptake. He contrasted Web 2.0 with the WorldWideWeb (born 1994-95), which was characterized by publishing, searching and reading, facilitated by content management and regulated by Webmasters. Web 2.0 provides a new way for people to “put their own stuff up,” with new rules of guidance and new terminologies. Tim O’Reilly identifies seven “core competencies” in a book on Web 2.0. Of the seven, Mr. Holt stressed “control of unique, hard to recreate data sources that get richer as more people use them” and “leveraging the long tail [i.e., rare incidents] through customer self-service” as especially relevant to health data collection and research.
To explain Health 2.0, he used two lists of constituent parts, noting that his view of the phenomenon is evolving because it is a moving target. His latest version states that patients are increasingly guiding their own care—the locus of control of health care is shifting. The components in that version are:
- Personalized search that looks into the long tail, but cares about the user experience;
- Communities that capture the accumulated knowledge of patients and caregivers and clinicians, and explain it to the world;
- Intelligent tools for content delivery and transactions; and
- Better integration of data with content.
He noted that presenting data in an integrated and useful way to all users is particularly important in the context of his subject for NCVHS, along with knowledge capture by communities. Simplified, the components are search, tools, and social networks; and the intersection among the three is “where the really interesting stuff is starting to happen.” He added that health care has not done a good job of putting together content and transactions, and these tools are intended to bridge that gap. He then showed screenshots of several examples of Health 2.0 components—personalized search, intelligent communities (e.g., ACOR and patientslikeme), and intelligent presentation of content. (See details in transcript and slides.) He predicted that the coming years will see “more and more people throwing computing power at content and information.”
The Edelman Health Engagement Barometer reports that 22 percent of U.S. adults are “Health Info-entials” who both look for health information online and disseminate it to others. Such consumers, Mr. Holt asserted, “in some ways are running ahead of their providers.” The issue for online communities is to use technologies to “get the really important stuff to bubble to the top.” The most useful information is personalized and supports decision-making and transactions. He noted that some health care provider groups are starting to use such tools to support and engage their patients; and new EMR companies are looking into how to use the tools.
He then showed a schematic of a continuum of Health 2.0, moving from left to right: user-generated health care > users connect to providers > partnerships to reform delivery > data drives discovery. We are still on the left side of the continuum, he said, with little impact on the last two “boxes”; and it will probably be necessary to change “the way we pay for things” to generate benefits and impact in using data to change discovery and therapy. He described the activities of people using online communities such as PatientsLikeMe and CureTogether for research to determine “what works and what doesn’t.” A broad challenge now is to integrate these phenomena “back with what’s happening in the health care facilities where the care is taking place.”
Finally, he offered suggestions for how to spend Stimulus money. He encouraged work on partnerships with Health 2.0 networks; funding for key online patient communities; attention to integration to connect the big databases with disparate innovative data sources; and reaching out to patients everywhere they are, to get information and reward out to them and to enforce “reporting back.”
Mr. Holt’s presentation stimulated a long and lively discussion with NCVHS members. They had comments and questions on online health decision support for consumers; the importance of the interpersonal relationship in health care; privacy issues and emerging Health 2.0 models of protection; theory of evidence issues about the data from social networks; the need to understand and prepare for these trends; the use of Health 2.0 for health management as well as disease management; the need for a valid critique of these developments and the possible role of the Committee in that regard; and key ways to shape the NHIN so it can capitalize on these developments. To the last question, Mr. Holt stressed the importance of integrating information from all possible sources, including patient communities, and of using these sources to capture a broad range of data about people’s lives and experiences.
2. John Halamka, MD, CareGroup and Harvard Medical School
Before talking about non-traditional data sources, Dr. Halamka defined “traditional data” as mostly administrative claims data, aggregated from hospital-based, payer-based, physician organizations, and health data consortia sources. He listed 14 non-traditional and emerging data sources (see slides), noting that many are more clinically oriented and can assist quality analysis. For example, providers are creating business intelligence “datamarts” and repositories, and moving from a transactional system into a more longitudinal relational system that can be queried. He noted that these work especially well for large care groups. For solo practices, the Massey Health Collaborative has created a quality datamart for 600 practices, and the hope is that many will contribute to a quality data warehouse. Dr. Halamka described a pilot envisioned using HITSP standards, the Massey warehouse and HITEP measures to demonstrate how a region can do quality measurements from EHRs and hospital information systems.
As another example, he noted that payers are realizing that administrative claims data are not adequate; they want lab data and many others, and every payer wants something different. His organization is working on a way to submit a uniform set of lab data to payers through RHIO-based transactions from providers to payers. Another novel data source is longitudinal medication lists created by RxHub and Surescripts, which are being used for various analytic purposes. He gave several more examples of innovative uses of data and novel data sources and exchanges.
Noting the issues with the accuracy of patient-entered data, Dr. Halamka said an emerging use of PHRs is to enable patients to aggregate their own data from multiple data sources, with privacy flags to mask sensitive data. This enables access to more comprehensive data on patients. Another Google tool is Google Trends, which can be used to see every Web search going on (e.g., fever), which could serve as an early indicator. His region uses Aegis to gather data from all local emergency departments to look at influenza outbreaks. And Catalyst, a clinical translational science awards site, enables data mining activities.
In summary, he said a number of interesting non-traditional approaches to gathering data and measuring quality are happening in his region. He predicted that the PHR, especially the aggregation of data by patients and sharing that data, and the use of device data will be very important to emerging sources of quality measures.
NCVHS members had questions about privacy issues and solutions related to these new trends; PHR adoption and usage patterns for CareGroup patients; and semantic interoperability and natural language processing.
LETTER BY THE SUBCOMMITTEE ON POPULATION HEALTH
Dr. Hornbrook presented and read aloud a draft letter to the Secretary that encourages adequate funding and resources for the nation’s health statistics enterprise to ensure the capacity to measure and realize the return on the new investments in health IT, comparative effectiveness, and prevention and wellness initiatives. Committee members offered editorial suggestions. The Committee unanimously passed a motion to approve sending the letter, after further wordsmithing by the Subcommittee.
Mr. Reynolds then adjourned the meeting.
I hereby certify that, to the best of my knowledge, the foregoing summary of minutes is accurate and complete.
/s/ June 10, 2009