Skip to main content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.


The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Minutes of the January 6-7, 1999 NCVHS Work Group on National Health Information Infrastructure Meeting

Public Health Service


Work Group on National Health Information Infrastructure

January 6-7, 1999
Washington, D.C.

- Minutes -

The Work Group on National Health Information Infrastructure of the National Committee on Vital and Health Statistics was convened on January 6-7, 1999, at the Hubert H. Humphrey Building, Washington, D.C. The meeting was open to the public. Present:

Work Group members

  • Don B. Detmer, M.D., Chair
  • Jeffrey S. Blair, M.B.A.
  • Daniel Friedman, Ph.D.
  • Richard K. Harding, M.D.


  • Clement Joseph McDonald, M.D.

Other NCVHS members:

  • Robert Gellman, J.D.
  • Barbara Starfield, M.D.

Staff and liaisons

  • Mary Jo Deering, Ph.D., OPHS
  • Sandy Haydock, HCFA
  • Hetty Khan, NCHS
  • Steve Steindel, Ph.D., CDC
  • Marjorie Greenberg, Executive Secretary, NCHS
  • William Braithwaite, ASPE
  • Dave Riley (representing Col. Ray, DOD)
  • Erin O'Connell, ASMB
  • Maureen Williams, ASMB
  • Jackie Adler, NCHS

  1. The Work Group approved a revised charge and affirmed that the word "National" should be retained in its name.
  2. Staff was asked to look at the information plans and models of Australia, the U.K., and Canada, et al., in relation to the Committee's vision document.
  3. Staff was asked to organize presentations on those three national systems at the February full Committee meeting. Dr. Detmer was asked to offer his synthesis at that meeting, as well.
  4. Members approved production of a paper covering the international models and where they are pointing (based on or incorporating the staff prepared document outlined above)
  • major drivers in the U.S.
  • recommendations for the private and public sectors

5. Dave Riley promised the following to the Work Group:

  • the latest version of the DII master plan
  • a description of a model of collaboration he is developing
  • the paper described in the summary below, starting with an outline


The Work Group first heard from Erin O'Connell, who coordinates an ASMB effort to assess the status since 1993 of the Department's budget authority and spending for telehealth and telemedicine. She gave preliminary information on the direction of the project and promised a more complete report at a future meeting. She offered herself as a resource to the Work Group, and they in turn welcomed this new relationship with ASMB and offered the Committee as a resource.

They reviewed and slightly edited a revised charge and approved it unanimously, forwarding it to the Executive Subcommittee for action.

They then heard presentations on three types of uses of computer health records: patient records (Mr. Blair), population records (Dr. Friedman), and personal (AKA consumer) records (Dr. Deering). They were asked to listen for the common elements in all three types. They then discussed applications and issues related to these EHR systems, particularly confidentiality issues related to current online information processing. Attention was called to the lack of any NCVHS work at present on privacy issues in respect to online, consumer- and market-driven patient records.

The discussion period began with an attempt to identify the structure and core elements for a matrix on the NHII. However, members soon agreed that it would be sensible to consult the models developed by Australia, Canada, the U.K., and other entities before going forward with a matrix.

Much of the remainder of the day was spent considering ways for the NHII Work Group to maximize its impact. Dr. Detmer outlined his vision for the government's role in promoting health and the contributions NCHVS could make to that process. The group moved to the idea of laying out the potential uses of a national health information infrastructure, as a way of moving the private and public sectors toward a broader vision. They agreed to see that such a paper is written, drawing on the national models mentioned above. They welcomed Dr. Detmer?s suggestion that the Work Group produce an annual paper on the national health information infrastructure, tracking progress and identifying issues.

In planning the February meeting, members agreed to organize presentations to the full Committee on the three countries information models, with an overview by Dr. Detmer. They also discussed the importance of coordinating the work of the CPR and NHII Work Groups, with Mr. Blair playing a key role through his joint membership.

The second day of the meeting was devoted to a presentation by Dave Riley, consultant to the Department of Defense, on an information infrastructure he is developing for DOD Health Affairs. Following his presentation, Dr. Detmer commented on the applicability of the DOD models to the Committee's global-level interests in the NHII. He proposed that Mr. Riley sketch out a vision from that perspective. Mr. Riley suggested that he take the papers from the U.K., Canada and Australia and the DII master plan and integrate them into an NHII strategic plan. Dr. Detmer asked that he provide at least an outline of this plan for the next Work Group meeting. They agreed to revisit the population dimension another time.

Dr. Detmer encouraged the Work Group to firm up a work plan for the coming year as soon as possible.



After introductions and a review of the agenda, Dr. Detmer introduced Erin O'Connell of ASMB.


Ms. O'Connell coordinates an ASMB effort to assess the status of the Departments budget authority and spending for telehealth and telemedicine. The project is a high priority for the ASMB, with the goals of laying a foundation and providing direction for the future. Dr. Callahan initiated the project, some of the findings of which have been shared with MB. She

urged the Work Group to use her and work through her for needed data. The FY2000 request is for $101 million for a broad scope of telemedicine/ telehealth services. All HHS agencies are being surveyed about data from 1993 on.

Ms. O'Connell plans a more complete slide presentation in February; the purpose of this meeting is to introduce the project and determine the Work Group's interests. Dr. Deering noted that the ASMB research could supply much of the information needed for the Work Group's planned matrix.

After the brief description of the future presentation, Dr. Detmer expressed approval and asked in addition for information on regulatory barriers that should be brought to the attention of the administration. Dr. Deering asked for a delineation of the terms and categories being used (preferably in advance of the presentation) --e.g., clinical services, information management.

Dr. Detmer encouraged Ms. O'Connell and her colleagues to view the NCVHS Work Group as ?a port of call' for deliberations and testimony.

Dr. Steindel noted that the legal barriers across state lines are a major issue. Dr Deering added to the list reimbursement and financial impact, and queried about the relevant activities of HCFA, IHS, VA and FDA.

Ms. Greenberg hailed this first-ever interaction between NCVHS and ASMB staff, reflecting the benefits of working with the Data Council and having broader staffing. She echoed the point about the Committee being a resource. Dr. Deering acknowledged the role of Maureen Williams as the ?matchmaker.? Ms. Williams said she had been struck by the overlap between work in ASMB on electronic commerce and the charge of this Work Group, and wanted to help maximize their productivity and synergy.


The group affirmed its desire to retain the word ?National? in the Work Group's name. They then reviewed the text of a revised charge, discussed further edits, and made a motion to accept the charge as revised. The following day a print version incorporating the edits was presented, and the Work Group voted unanimously to accept the revised draft. The new language will be presented to the full Committee in February for approval. Dr. Detmer stressed framing the charge in terms of action, and having the matrix drive action. He also noted the importance of reporting to the full Committee. The revised charge is posted on the NCVHS Web site.


Dr. Detmer introduced the next series of presentations by explaining that they are intended as background for work on the matrix model. Dr. Deering urged the group to be listening for the common core elements of the three types of health records, which will form the basis for the matrix.

Patient Records: Mr. Blair

Mr. Blair provided a conceptual overview of the basic components of an electronic

health record (EHR) and its potential relationship to a health care information infrastructure.

He discussed these topics:

  • factors leading to the development of an EHR framework
  • a vision for EHR
  • major initiatives
  • essential elements of a global health record
  • what is needed to promote progress

EHR is a common element of various health care information infrastructures. A similar vision of EHR is emerging in the U.S. and the U.K., with attributes that are patient-focused and longitudinal. The primary user is the provider; secondary uses are clinical research, public health research, reimbursement, and other administrative uses.

The Patient Record Institute model divides the components of the health record into content (broadly defined), time frame (multi-episodic), and information representation. The various EHR functions include information capture, storage, processing, communication, and presentation. Anticipated EHR system characteristics include heightened security, usability, and reliability. Legal characteristics include retention, confidentiality, and authentication. Mr. Riley mentioned evidentiality as another legal characteristic. All of these functions and characteristics must operate in a dynamic medical environment.

The major implementations are taking place in Australia, Canada, the European Union, the U.K., and the U.S. Australia has a national health information model based on ISO standards for meta-data. (Dr. Braithwaite explained that the last term refers to looking at the characteristics of the data rather than the data themselves.)

The Canadian initiative, coordinated by the Canadian Institute for Health Information, seeks to create the framework for a national HII, emphasizing national databases to support protocols and guidelines. It is still at the proposal stage.

The E.U. approaches heath care from the top down. Projects that in the U.S. are dominated by vendors and consortia are publicly funded in the E.U.-- e.g., the Good European Health Record project. Key words are the Gamen Project and the Committee European du

Normalization (CEN). Dr. Detmer observed that these E.U.-funded projects were slow in getting started but are now bearing fruit and are complementary to developments in the U.S.

The U.K. has been proactive in setting forth its National Health Service Information Strategy. GPs are encouraged to use the Reed coding system, focusing on ambulatory environments.

Turning to the U.S., Mr. Blair observed that here there is a rich list of standard initiatives, but they fall short in terms of interoperability, clinical specificity, and security. However, they are implementable, not only in the U.S. but abroad. HL7 has been implemented in about 11 countries; SNOMED has been translated into 13 languages. Together, these systems can represent ?the first phase of an information infrastructure.? In addition, tool kits have been developed that will probably lead to standards for Internet use. Overall, he observed, each country or region is offering different things.

Mr. Blair focused on the elements of standards that might facilitate basic functions for EHRs. He hypothesized achieving the ability to share an EHR internationally to support global travel, and to provide care to underserved populations. Finally, he posed the question of what aspects, functions and characteristics would be necessary to achieve these goals, and what can be gleaned from the accomplishments around the world.

Dr. Detmer observed that the Work Group on Computer-Based Patient Records gets into the nuts and bolts of this task. Dr. Deering wondered what core elements can be inferred from the foregoing presentation.

Population Records: Dr. Friedman

Dr. Friedman noted the startling lack of a literature in the U.S. or abroad on the concept of a population health record. His own thinking on the subject defines ?health? broadly, in the W.H.O. sense of ?a state of complete physical, mental, and social well being? rather than just the absence of disease. ?Population-based data? pertains to a ?known population? that may be defined geopolitically, demographically (e.g., an ethnic group), as members of a health plan, or in other ways. In other words, it is not simply the aggregation of data on individuals. Thus data on a sample of the population is generalizable to the whole. This is the domain of health statistics.

Population records are derived from individual records, population area records, or a combination of the two. These data are commonly used for health policy and programs, for public health surveillance, and for interventions for prevention and control. Dr. Friedman noted that today's public health data systems are too past-focused, missing important health trends like the recent upsurge of alternative medicine use.

Population records need to be person-based, longitudinal, scalable, and minimally burdensome to data providers. Having longitudinal person-based records implies having personal identifiers, for which a legal and regulatory framework to guarantee security and confidentiality is a prerequisite.

Population health records would permit more integrated intervention. Ideally, there would be a unitary data system that could be used for various purposes. A useful model is something the Canadians are thinking about: sample-based pilot population health record systems that could have a variety of bases -- plan, geography, etc. The Framingham study is an example of one such system.

Mr. Riley said the Department of Defense is doing something of this nature with individual-level data for servicemen deployed in areas with risk of exposure to biological and chemical weapons. The system includes health risk appraisal evaluations, surveillance data, and environmental data.

The group discussed the similarities and differences between the population health record described by Dr. Friedman and instruments like NI-NHANES. Detailed ?snapshots? like HANES may still be needed but should be linked to other data sources.

Personal Records-- Demonstration of Online Initiatives: Dr. Deering

Dr. Deering noted that there is growing interest in personal medical record systems that enable consumers to keep a comprehensive record of individual and family health. Secondarily, they are a way to share this history with various providers. This is part of the broad movement toward consumer health activism, self-care, and empowerment. There is no charge to use systems currently on the Web; rather they are a way of attracting people to health sites that are gateways to commercial health information operations. From a policy perspective, these systems are interesting because they can afford portability and consistency, encompassing care from alternative as well as traditional providers.

There are three models for these systems, which Dr. Deering likened to Quicken (standard template, information kept at home), TurboTax (consumer plugs information into ?the other side'), and Visa (warehousing and clearinghouse services provided by an entity). The government patient record people are exploring the third model.

She gave examples of each. One of the most comprehensive of the first model comes from the AMA and is on its Web site. The AMA does not maintain it but merely provides the format and does the processing; the individual downloads it and maintains it at home. A system is being developed for Dr. Koop's Web site that involves online registration and will create a secure online medical record. Dr. Deering also described the following instruments and services:

  • The Health Tracker Corporation's system
  • Vital Works
  • HealthCentral .com
  • Kaiser Permanente Web service for its beneficiaries

Most of these ask for information on risk factors and facilitate simple health risk analysis online. Most are advertiser-supported.

Dr. Detmer said the second and third generation of these sites are more interactive and oriented to health management and illness management. At Dr. Deering's request, he demonstrated a model he helped to develop. Among other things, it allows patient and physician to interface both on data and on process of care. The patient controls the information flow. One feature allows the doctor to look at his/her entire patient population. Patients can also get administrative information on themselves from their plans.

Dr. Friedman noted the benefits of giving patients online access to provider records on them, e.g., test results and risk factors. Dr. Detmer commented that the biggest obstacle in general is getting physicians to enter clinical information. The Dutch and the British are way ahead of the U.S. in this regard, having substantially institutionalized the use of computer-based health records. In the absence of provider input, with a record entered entirely by the patient, what is available is the patient's version of ?their story,? which is itself significant.

Mr. Gellman called attention to the privacy risks of putting personal health information on the Web. Most of the proposed privacy legislation would not affect these sites because people enter the information voluntarily. He also discussed the implications of putting patients in control of their computerized records. It could undermine the ability of physicians, researchers and overseers to access and use records. He and Dr. Detmer agreed that its too early to foresee how such a capability will evolve. Dr. Steindel noted that in pathology, many people already control their own records, e.g., with storefront labs.

Dr. Harding asked where the financial interests are in these site and systems. Dr. Detmer said pharmaceutical companies are actively seeking out consumers, and this is a way to do so. Currently, any constraints on using the information volunteered by consumers are the result of self-regulation, and those instances are in the minority. He noted the wide range of comfort levels in the population about sharing personal information.

Dr. Deering said that those exploring the ?Visa? concept are thinking about what political, educational, marketing, and transactional process are needed as preconditions of putting information online. She noted that the driving motivation for that capability is not consumer empowerment but portability and consistency.

Mr. Riley said that legislation could at least require a full and open disclosure of the intent for using the information volunteered by consumers. This is in part an issue of

intellectual capital, meaning that individuals should be compensated in some way for the use of their information.


The group began with the intention of discussing the core elements for patient, population, and personal (AKA consumer) records, as a step toward developing a matrix on elements of the NHII. Their initial discussion centered on clarifying and reaching agreement on process, terminology, and structure. After reflecting on the morning's presentations, they agreed that the terms ?patient, population, and consumer? are most descriptive of the types and uses of health records. Documents using these terms must include definitions, as the terms are not self- explanatory. For cross-cutting categories, they agreed on ?functions, content, characteristics, and communications.? Related clarifications:

  • The variables, who controls and accesses the content, are subsets of function.
  • Security is a data characteristic.
  • Interoperability is a sub-bullet under communication.

Mr. Riley described his analytical work for DOD, leading to a request, which he accepted, to present his work the following day with reference to both patient records and the broader DOD record system he described early in the meeting (the Service Member Life Cycle Model).

After further discussion, the group agreed with Dr. Detmer that it was not productive

for them to go forward with the matrix exercise without reference to the information models

developed by Australia, Canada, the U.K., and other entities. He asked staff to review these

models looking for potential applications to the U.S. These reference models were mentioned:

  • Australian
  • DOD/VA initiative
  • HL7 Reference Information Model
  • British COSMOS clinical process model
  • Good European Health Records object model

Members turned to the question of where the Work Group could have the greatest impact. Dr. Detmer proposed that NCVHS has good reason to raise consciousness in policy circles about population health and population statistics, and the NHII Work Group is probably the best subgroup to do so. He wants to see the government enhance the health of the U.S. citizenry, not just services to it. This requires measuring the health of a population and getting the numerator and denominator to interface. It is also important to raise the notion among individuals that keeping one's health data up to date is a routine task of living, not unlike keeping one's driver's license up to date. The ultimate goal is a set of functionalities that allow synergistic activities to enhance health at the lowest cost and greatest effectiveness. This involves a link between interventions at the clinical and population levels.

Dr. Friedman observed that a Canadian goal is an information system that can isolate the impact of health care on health -- for him, ?a challenge that resonates.?

The group briefly discussed the commonalities among the three types/uses of records, mentioning medical events data, personal risk data, and patient demographics. Dr. Steindel postulated that having health records is a basic right and observed that the Work Group is looking for the NHII attributes that will help create an environment in which people can exercise that right.

Returning to the question of the Committee's impact, Ms. Haydock proposed that it seed a consortium for people already working in these areas. Dr. Detmer responded regretfully that now, nearly everyone is focused on their own track or application and has little interest in the linkage across domains. The group discussed whether these players could be induced or persuaded to take a broader view.

Dr. Friedman noted the value of laying out an ideal image of what a national health infrastructure information could do. Members cited the NHS, Australian and Canadian papers as good models. This led to the idea, which the group embraced, of developing a paper derived from those nations' documents, as a foundation for developing the Work Group's work plan. Staff was asked to look at the plans and models in relation to the Committee's vision document. Ms. Greenberg noted that the Work Group on Health Statistics for the Twenty-First Century has commissioned a paper from the Canadian Institute for Health Information that is also relevant.

They then planned their February activities, agreeing to organize presentations on the Australian, Canadian and British systems at the full Committee meeting. Members were asked to read the documents in advance.

Dr. Detmer and Mr. Blair discussed the relationship between the foci and activities of the CPR and NHII Work Groups. Mr. Blair stressed their synergy, and Dr. Detmer encouraged him to continue to serve on both groups in order to coordinate them and prevent redundancy. He noted that the groups could sometimes meet jointly.

Also related to the full Committee's work, Dr. Deering noted the lack of any work on privacy issues in respect to online, consumer- and market-driven patient records.

On another subject, she confirmed that the Work Group still wants an overview of commonalities among the three types of records, for use at the February meeting.

To help the Committee move forward on this issue, she proposed having someone speak on the market forces that might promote convergence, i.e., the business case for an integrated approach. Mr. Riley cited the POMA effort by Benchmarking Partners, which identified ?the value chain for the global health care environment.? The group discussed the

business trends toward more consumer control in the choice of health care plans and the resulting demand for more and better information. Dr. Detmer called attention to a paper he co-authored, ?Quality at a Crossroads,? published in the current Millbank Quarterly.

Dr. Deering and Ms. Greenberg urged Dr. Detmer to share his synthesis of these issues with the Committee. (Ms. Greenberg later said this ?valedictory? is on the agenda for the February full Committee meeting.) Dr. Deering observed that this could help reinforce the future orientation of the Committee, both in the way it thinks and the way it is perceived.

Dr. Detmer proposed that the Work Group do a paper covering

  • the international models and where they are pointing (based on or incorporating the document outlined above)
  • major drivers in the U.S.
  • recommendations for the private and public sectors

The Work Group could publish and post a concept paper every year, tracking progress and identifying issues. This idea received an enthusiastic response from members and staff. The first paper could be the three-nation analysis mentioned above, approached in relation to a vision for the NHII in the U.S. In addition, the Work Group could commission three or four papers a year on key topics.

The group deferred any work on the matrix. Dr. Detmer urged that a glossary be developed and disseminated. He and Ms. Greenberg discussed ideas for also developing links in the NCHVS Web site to other good Web sites for documents, organizations, and conferences. She said Committee staff is already working on a similar project, and it might be expanded. Staffing to screen possible candidates and keep the site up to date still needs to be worked out.

Finally, members briefly discussed developments on the G7 front. They also agreed on the merits of setting aside time in forthcoming meetings for briefings from non-members about activities the Work Group is tracking.

They then recessed until the following day.


After Dr. Deering summarized the previous day's proceedings, Dr. Detmer introduced the day's presenter.


Mr. Riley is chief architect for the next generation of DOD health care systems, now working as an independent consultant. He described his role in doing technology integration for DOD Health Affairs and in creating and implementing a vision for that department with business, information management and technical components. He then gave an overview of the DOD health information infrastructure (see transcript for details) and offered some thoughts on how it could be adapted for the purposes of this Work Group.

Mr. Riley's two deliverables are a health care interoperability paradigm and a strategy for data integration and migration. The latter task involves moving Health Affairs from 167 systems to 52 clinical systems, and ultimately to four integrated clinical systems. DOD's use of managed care makes it increasingly necessary to interface with the private sector, making interoperability a priority. Internally, it is developing an operational continuum from peacetime to full major conflict-- a new concept for DOD Health Affairs.

The health information infrastructure has a ?C9 definition?: ?the convergence of computers, communications and content to create collaborative communities to convey care. Each ?C? has several implications -- for example, one ?collaborative community? might be breast cancer patients brought together for an online support group. Mr. Riley said the ecosystem model is useful for understanding collaborative communities and their requirements and dynamics.

The need for architecture models necessitated a review of the various uses of this much- used term and a framework for evaluating architectures. To evaluate individual systems, think about interoperability, and create measurable outcomes. He developed an architectural scalability reference model and a levels of information system interoperability reference model.

Mr. Riley offered a structural analysis of information and the major forces operating at each level. For example, performance and functionality are dominant concerns at the application level; management of change and of complexity are dominant forces at the system architecture level; the enterprise level involves direction and control; and the highest level is global architecture, where the concern is how multiple enterprises can work together. The NC NCVHS NHII Work Group operates at that global level, where it deals with alliances and collaborations. It lacks the power to coerce but is able to lead by example.

In standardizing lower levels of the DOD architecture, Mr. Riley had to contend with clinical systems with clashing IT ?backbones.? The architecture scalability reference model permits development of a common infrastructure by clarifying the main concern of each player in DOD. The planners are trying to minimize unique data, and have mandated self-describing formats using XML.

The DII Master Plan, DOD's information infrastructure master plan, would be of greatest interest to the NHII Work Group and most adaptable to an ?NHII strategic plan.? Mr. Riley promised to provide a copy of the latest version at the next meeting.

Commenting on records, he noted that the basic purpose of the record is to preserve the content and context, and to allow for its retrieval. Collaboration to convey care requires not just data communication but shared understanding. ?Collaboration spaces? are n~needed to facilitate development of that shared understanding. The medical record is currently one of the greatest collaboration spaces for the provider and patient. (Dr. Detmer observed that ?this is where we start hitting the behavior wall.?) He described his work on an infrastructure for informal collaboration in clinical settings, which is now being beta tested at DOD Health Affairs. He promised to bring a description to the next NHII meeting.

Turning to the computer-based patient record, he outlined his phenomenological analysis, covering the areas to be monitored. (For example, the components of knowledge representation and reasoning include models for terminology, information, and knowledge.) To address semantic inconsistencies in various data models, he brought in people from IBM, 3M, Micromedics, PKC and Serius to discuss an optimal design for a lexicon. They formed the Clinical Lexicon Consortium and agreed to create a single information model to support the government. It incorporates SNOMED (40%), is based on the UMLS, and is still under R&D. The consortium has recently agreed to move to XML technology. Dr. Detmer commented that this activity is relevant to the Committee's work on standards. Mr. Riley observed that understanding that models operate at different levels of abstraction makes it easier to figure out how to harmonize them.

Other components of his phenomenology are temporal and spacial representation and reasoning, and a detailed analysis of security. He noted that the U.K. has done good work on sensitivity and confidentiality and is way ahead of the U.S. in this area. Dr. Detmer noted the relevance of these contributions to the Committee's work on security.

Dr. Detmer commented on the applicability of the DOD models to the Committee's global-level interests in the NHII. He proposed that Mr. Riley sketch out a vision from that perspective. Mr. Riley suggested that he take the papers from the U.K., Canada and Australia and the DII master plan and integrate them into an NHI strategic plan. Dr. Detmer asked that he provide at least an outline of this plan for the next Work Group meeting.

They agreed to revisit the population dimension another time.

Dr. Detmer remarked that the ultimate use of these systems is for optimization in the areas of quality, value, prevention, and education.

Dr. Deering recalled two related issues that arose earlier in the meeting, regarding 1) the principle of having the individual as the basis of the information infrastructure and 2) whether/how to distinguish between ?medical? and ?extra-plan health? dimensions. Mr. Riley

expounded on his concept of an ?c-health business ecosystem? but Dr. Deering's question was not directly addressed.

Finally, Dr. Detmer commented on the policy and values issues that underlie and precede technological solutions, and Mr. Riley offered a few more thoughts on the important role of this Work Group in influencing policy and setting boundaries. Dr. Deering observed that having a vision and strategy makes it possible to see what is missing in the work of other groups, and to offer reminders to round out the efforts.

Dr. Detmer expressed satisfaction with the Work Group's current strategy and work plan and urged the group to firm up its work plan for the coming year at the next meeting or soon thereafter. He then adjourned the meeting.

I hereby certify that, to the best of my knowledge,

the foregoing summary of minutes is accurate and complete.

/s/ John R. Lumpkin 8/23/99


Chair Date