NCVHS Subcommittee on Population Health
“Meeting Information Needs for Health and Health Care”
February 9, 2010 Teleconference
Summary and Notes
The NCVHS Subcommittee on Population Health held this teleconference in lieu of a planned meeting that was canceled due to heavy snow. The objective of the meeting was to continue the work on an NCVHS vision for meeting information needs to improve health and health care.
- Harry Reynolds, NCVHS Chair
- William J. Scanlon, Ph.D., Subcommittee Co-Chair
- Donald M. Steinwachs, Ph.D., Subcommittee Co-Chair
- Justine M.Carr, M.D.
- Leslie Pickering Francis, J.D., Ph.D.
- Larry A. Green, M.D.
- Anthony Rodgers, M.S.P.H. (also a presenter)
- Walter Suarez, M.D. (also a presenter)
- Judith Warren, Ph.D., R.N.
Staff and Liaisons
- Marjorie Greenberg, NCHS, Executive Secretary
- Debbie Jackson, NCHS
- Virginia Cain, NCHS BSC Executive Secretary
- Jennifer Madans, NCHS
- Nancy Breen, Ph.D., NCI (also a presenter)
- Carrie Klabunde
- Mike Repp
- Doug Boenning
- Susan Kanaan
- Dr. Jane Sisk, NCHS Division of Health Care Statistics
Agenda, Purpose and Goals
As noted, the purpose of this meeting was to continue the work recently conducted by Drs. Dan Friedman and Gib Parrish to revisit the 2002 NCVHS vision for health statistics in the 21st century. The current project focuses on finding ways to 1) improve information on health and health care, 2) take advantage of new health information technologies, and 3) influence current policy initiatives so the entire U.S. population can benefit. The focal timeframe is the next five to ten years, during which the Committee believes real progress can be made. NCVHS will present a report on these topics at the Committee’s June, 2010, 60th Anniversary Symposium and celebration. The report is intended to provide a high-level framework for approaching these topics, in order to help guide national information policy for improving population health and health care and coordinate the work of NCVHS and its subcommittees over the next few years.
The meeting featured presentations on population health, health care provider surveys, and quality of care reporting, along with many opportunities for discussion. (Speakers’ slides are available online; raw notes on the proceedings, in lieu of a transcript, are available on request.)
Health Measurement and Contributing Factors – Mr. Rodgers & Dr. Breen [slides]
The presenters began by outlining the objectives to be served by a fully realized health information infrastructure improving population health, identifying and combating disparities, establishing and maintaining quality standards, providing evidence-based services at fair prices, and ensuring coverage for the U.S. population. They reviewed the ways population health is currently measured (primarily through national surveys) and the sources of complementary information on the social determinants of health (SDOH). Much information on SDOH is already collected, but not linked to health or health care information. The presenters stressed the importance of local data and the need to augment and calibrate the data using reliable national and state survey data. They presented a vision of a local, state and national health survey system, grounded in a state-level health information exchange platform architecture that combines multiple data sources diagnostic, prescription history, administrative/survey data, EHRs, and clinical lab data.
Mr. Rodgers noted that data flow enabled by health information technology (IT) presents new opportunities to collect, exchange, manage and analyze data and use the information to improve the health of sub-populations and the population as a whole. However, privacy issues have to be resolved to make this possible. He showed and described an enterprise-level data repository and decision support infrastructure. In the presenters’ vision, an advanced medical home plays a pivotal role as the vanguard of population and community health assessment, evolved from current versions of the medical home to include links with community resources, translational research, and other activities. Data can be rolled up from the locality to state and national levels, with calibration by surveys. Mr. Rodgers stressed that data from local sources must generate useful information back to those sources, for use in clinical and population health interventions.
The speakers concluded that health information exchange (HIE) and survey data need to be integrated into the same infrastructure; data are needed at local, state and national levels; and clinical care needs to be understood in the context of population health and the SDOH. A critical next step in improving population health, they said, is standards and guidelines to ensure that survey data represent the U.S. population and that this information is integrated with information from health information exchanges and on all the determinants of health.
In the discussion period, the group discussed the gap between current health care delivery and the ideal depicted by the presenters. Mr. Rodgers said that technology will help create the advanced medical home. (Some participants suggested alternate terms either “health care home” or the standard “usual source of care.”) Various ways of storing and accessing data, including a distributed model, were discussed. Dr. Scanlon noted that a critical issue is the rules for accessing and protecting data. The Committee’s work on health data stewardship was referenced.
A key point is that right now, an infrastructure is being formulated with no specification on how population health management will use the infrastructure. Mr. Reynolds stressed that current timing is critical, because vendors are building solutions and need to be given specifications that enable progress for population health. Dr. Frances noted the challenges related to the different regulatory regimes that apply to different arenas of data use.
The group discussed several models of participatory, community-based research (e.g., CTSA and NIH’s From Cells to Society) and the potential represented by such activities. There was interest in learning more about these models.
Dr. Suarez and others reiterated the importance of this meeting in relation to current national health policy initiatives. Of particular interest is the opportunity to contribute to the evolution of the Meaningful Use concept, to ensure that data captured by EHRs are useful for population health and that developers have the necessary specifications to make that happen.
National Health Care Surveys – Dr. Sisk [slides]
The National Center for Health Statistics collects, analyzes and disseminates data on the use, access, quality, and cost of health care and the organizations and professionals who deliver it. Dr. Sisk described the several surveys on ambulatory and hospital care and on long-term care. (See slides for details.) Data from these surveys have been used to assess quality, including safety and disparities, conduct comparative effectiveness research, track the diffusion of technology and the effects of policy changes, and study the epidemiology of specific conditions, among other things.
Current data collection methods include in-person collection from the medical record, abstractions from medical records, CAPI and CATI methodologies (for long-term care surveys), paper forms, and mailed surveys. As examples of how data are linked, Dr. Sisk said that nursing home data have been linked to Medicare data for 2004, and in 2007 the same will happen for home and hospice survey data. Hospital discharge survey will be linked to the national death index; and information in the ambulatory files are linked to geographic characteristics such as the poverty level in a given zip code.
Dr. Sisk discussed the data on the diffusion of EHRs among physician offices, noting that less than half had some kind of electronic system in early 2009 and far fewer had “the features that matter most,” such as decision support, order entry, and lab results. She added that health IT would enable NCHS to release data sooner and may lead to more accurate and complete data. A major challenge is that EHRs are not pervasive enough to support nationally representative surveys, and it may be a long time before that happens; so NCHS will continue to rely on mixed modes of data collection and mixed sources. NCHS hopes to hire a consultant to support electronic reporting, starting with receiving encounter and abstract data. They are monitoring the development of health information exchanges (HIEs), which would enable them to get information on many providers from HIEs instead of going to individual providers, and would generate more population data.
In the discussion period, participants discussed the fact that while it will probably take a long time for electronic sources to develop for these purposes, they are clearly inevitable. The question is how to handle the transition period. Dr. Carr made a “plea” for the continued use of hybrid data, including administrative data, because most providers’ EHRs are likely to be incomplete for a long time to come. Dr. Green noted the potential synergies between this project and the new investments in human resources and regional extension centers.
Mr. Reynolds praised Dr. Sisk’s slide giving examples of provider-based data, and he urged that NCVHS offer similarly pragmatic examples to help industry understand what kind of information is needed to illuminate the population health dimension. The group affirmed its interest in helping to shape the next stages of work on Meaningful Use, and to influence the relationship between providers and vendors to increase the population focus and identify what is needed for national surveillance.
Quality of Care Reporting – Dr. Suarez
Dr. Suarez reported that the Population Health Subcommittee’s ad hoc workgroup on information for quality looked at how quality information is obtained today and how health IT will affect how it is obtained and exchanged in the future. It looked at national surveys, NQF activities, other federal initiatives, and what the private sector is doing. All in all, there is a convergence of measurement initiatives into the concrete metrics being incorporated into expectations for Meaningful Use that organizations must meet to qualify for incentives. This represents more concrete ability to measure quality in real time. However, there are still too many measures some 147 in all.
The workgroup identified issues in several areas. First, there is no framework for federal health care quality of the kind that exists for the federal health architecture and NHIN privacy and security. Second, drilling down to more granular information and looking at population segments will be important in enabling improvements in population health. The group also noted issues related to privacy and security considerations, especially related to granularity and segmentation; EHR capabilities; the alignment of incentives; and the consumer dimension of quality reporting; and it stressed the importance of thinking about the goal of quality measurement, which is to improve the health of the public.
In the discussion, Dr. Steinwachs added the question of how we construct longitudinal and cross-population views when most of today’s quality information is at a point in time. Dr. Carr noted the recent NCVHS letter calling for a coordinated national quality measurement strategy. She added that this meeting is enabling her to see that “This can work at a population level; we have the building blocks.”
Dr. Green cited Maintenance of Certification as another lever moving clinicians toward measuring and reporting quality. He stressed the importance of getting granular building blocks as soon as possible.
Participants mentioned some of the things we need to know, including whether those who can get better are getting better, who is and is not receiving services, how the system is doing in preventing conditions, and people’s functional status.
Review and Discussion of Next Steps
Dr. Scanlon asked the participants what stood out for them in the meeting. These are some of the responses:
- The medical home as a fundamental building block for population health is a powerful and unifying approach.
- We need to articulate the ideal and plan for the gaps. What happens to the portion of the population without access to a medical home? A vision linking the medical home and population health would require universal coverage.
- We need to keep privacy in the loop as we consider different data sources.
- Make sure the standards being adopted by HHS are promulgated throughout the system. There can be a big gap between clinical care and population health in this regard. If we focus on phases 2 and 3 of Meaningful Use, we could look at what standards exist and what ones are needed to support this vision.
- The integration for the future is about stitching together multiple information sources to provide a single rich information base. Surveys will be needed to fill some gaps. The vision can help show how the sources complement each other and fit together.
- The presentation by Dr. Breen and Mr. Rodgers shows the possibilities and the vision; many participants felt it provided a useful and reliable framework. The question is how to get from today’s fragmented system to the one it envisions, and how NCVHS can help move things in that direction.
- More work is probably needed on defining what a medical home consists of.
- The Committee needs to avoid overextending itself. Is getting into the structure of the delivery system and workforce issues beyond its scope?
- The group affirmed that the scope of the new NCVHS project is health and health care, not just the latter.
- The tools discussed today are “additive, not substitutive”; i.e., they can be added to traditional population health data tools.
As future actions, Dr. Breen offered to revise and distribute the slides on health measurement. Ms. Greenberg noted that the other Subcommittee Co-Chairs need to be brought up to speed on what happened in this meeting.
In summary, Dr. Scanlon observed that the day’s discussions had enriched the preliminary outline for the report. He challenged the group to figure out how to use the meeting content, or part of it, for the June deliverable and symposium. Dr. Steinwachs added that the vision that is evolving and becoming more real is that the three traditional data sources (surveys, administrative data, and EHRs) can be used in combination to achieve information of unprecedented richness at local, state, and national levels. Dr. Scanlon noted that the adoption of EHRs is being accelerated, and “it’s our job to make sure they’re useful.”