Department of Health and Human Services

National Committee on Vital and Health Statistics

NATIONAL HEALTH INFORMATION INFRASTRUCTURE WORKGROUP

November 12, 2004

Hubert H. Humphrey Building
Washington, D.C.

Meeting Minutes

The NHII Workgroup of the National Committee on Vital and Health Statistics was convened on November 12, 2004 in Washington, D.C. The meeting was open to the public. Present:

Committee members

  • John R. Lumpkin, M.D., M.P.H., Chair
  • Jeffrey Blair, M.B.A. (by phone)
  • Richard K. Harding, M.D. (by phone)
  • John P. Houston, Esq.
  • Stanley M. Huff, M.D.
  • Robert W. Hungate
  • Kevin Vigilante

Staff and liaisons

  • Mary Jo Deering, NIH, Lead Staff to Workgroup
  • Marjorie Greenberg, NCHS/CDC, Executive Secretary
  • James Scanlon, ASPE, Executive Staff Director
  • J. Michael Fitzmaurice, Ph.D., AHRQ liaison
  • Steve Steindel, Ph.D., CDC liaison

Others

  • Debbie Jackson, NCHS
  • Bob Kambic, CMS
  • Kathleen Fyffe, ONCHIT
  • David Lansky, Markle Foundation
  • Bill Rollow, CMS
  • Michelle Williamson, NCHS
  • Marietta Squire, NCHS
  • David Harrington, MedicAlert Foundation
  • Janet Martino, MedicAlert
  • John Morgan, HealthRight
  • Carol Cronin, AARP consultant
  • Richard Marks, Patient Command
  • Bill Kanouse, Patient Command
  • Peter DeVault, Epic Systems
  • James Walker, Geisinger Health System
  • Scott McFarland, SimplyWell
  • Robert Faltune
  • Robin Kaye
  • Jay Burns, McKesson Corp.
  • Kristina Rollings, McKesson Corp.
  • Dan Rody, AHIMA
  • Laura Blum, JCAHO
  • Ernest Ludy, Ludy Family Foundation/EGL Investments
  • Cynthia Bauer, ODPHP
  • Eduardo Ortiz, VA
  • Susan Kanaan, consultant
  • Bob Kambic, CMS
  • Linda Fischetti, VA
  • Suzie Bebee, ASPE
  • R. Kaigh
  • R. Felton
  • Seth Rupp, Cerner
  • Matthew Hagopian, Disruptive Technologies
  • Dan Rode AHIMA
  • Sandy Williams, ODPHP
  • Kathryn Serkes, Assn. of Amer. Physicians & Surgeons
  • Grayce Warren Bolton, AARP
  • Jason DuBois, ACLA
  • Laura Vartain, Wexler & Walker
  • Zina Cary, National Health Council
  • Brian Baum, Duke University
  • Andrew Barbash, Holy Cross Hospital
  • Michael DeCarlo, Blue Cross/Blue Shield Assn.

Note: The transcript of this meeting is posted on the NCVHS Web site.

EXECUTIVE SUMMARY

The purpose of this hearing is to look into personal health records (PHRs) in the light of the personal health dimension of the National Health Information Infrastructure (NHII) envisioned by NCVHS. As it has monitored activities in this area, the NHII Workgroup has had concerns about interconnectivity, giving adequate attention to the public health arena, and ensuring that developments have optimum benefit for patients. These concerns are the context for the current focus on the PHR. The members of the first panel provided an overview of PHRs; those in the second described PHR tools that are independent of electronic health records (EHRs), and the third described PHRs that are derived from EHRs. The Workgroup concluded the meeting by planning the January 5-6 hearing, which will continue the exploration of PHRs. There were no other actions.

PANEL ONE: OVERVIEW

  • Kathleen Fyffe, ONCHIT
  • Bill Rollow, CMS
  • David Lansky, Markle Foundation and Connecting for Health

Ms. Fyffe discussed PHRs in the context of President Bush’s vision for HIT and the HHS Strategic Framework. She briefly described ONCHIT activities related to PHRs and offered suggestions for ways in which the NCVHS NHII Workgroup can support ONCHIT’s efforts in this area.

Dr. Rollow discussed PHRs in terms of quality, which has become an important part of the CMS agenda. PHRs are one of five “core elements” in the Quality Improvement Organization strategy, and the Medicare Beneficiary Portal that will be tested in Indiana starting in December is a step in the direction of PHRs. As it moves forward, CMS is interested in how it might stimulate innovation to help the evolution of PHRs have a positive impact on quality. Dr. Rollow proposed for discussion the idea that CMS identify a descriptive set of PHR attributes following consultation with Connecting for Health and others, with the prospect of eventually giving preference to PHRs that use them. He invited comments on whether creating such specifications is a good idea, and if so, how CMS might go about it.

Dr. Lansky focused his comments on the themes about PHRs that have emerged from the work of Connecting for Health, a consortium of more than 100 organizations from the private and public sectors. He also commented on the challenges the federal government faces in moving forward in the near future. He outlined nine “major uncertainties” identified by Connecting for Health, representing all that remains to be learned, and he expressed hope that the learning could happen collaboratively. He identified a number of possible governmental roles with respect to these uncertainties and highlighted policy priorities and strategies for 2005 and 2006.

In the discussion period following this panel, the group talked about staying abreast of all relevant activity in this area, the options for different types of records, the degree of consumer and provider interest, relevant standards activity, the importance of having near-term work on EHRs anticipate the evolution of PHRs, and the merits of eventually having a set of agreed-upon terms in this arena.

PANEL TWO:

Organizations and Vendors Offering PHRs That Are Not Derivatives of EHR Systems

  • James Walker, Geisinger Health Systems
  • Brian Baum, Health Record Network Program, Duke University
  • Scott McFarland, SimplyWell
  • David Harrington, MedicAlert Foundation

Dr. Walker briefly described the “patient EHR,” My Geisinger, for Geisinger, which serves about a half-million patients in 31 mostly-rural counties in central Pennsylvania. He reviewed research on the use of PHRs and also discussed the current barriers to implementing such tools nationally.

Health Record Network (HRN) was launched by Duke University. It is a not-for profit-organization with no institutional affiliation, aiming to create a national health utility to benefit all health consumers. HRN is targeting early adopters among consumers, creating awareness of the need to manage information, and providing a PHR as an “overlay network” that supports consumers and links into EHR systems. It is being piloted in the State of Wyoming, at Duke Heart Center, and possibly in Canada.

SimplyWell LLC is jointly owned by the Nebraska Medical Center and 625 physicians in private practice. The company now has a presence in 41 states, 49 by mid-2005. Participants are working age adults. The product is part of employee benefit structures, primarily for large, self-funded employers. Mr. McFarland described reimbursement policy issues that limit participation by others. SimplyWell provides yearly work-site health screenings, on the basis of which individual health action plans and electronic health records are developed. The major thrust of SimplyWell is encouraging behavior change; when participants receive episodic care, the electronic health record, which is interoperable, is “there for the clinicians to utilize.”

MedicAlert is a non-profit foundation, founded in 1956, that protects and saves members’ lives mainly by providing information about them in emergency situations. It has kept EHRs since 1985. Most communication is by telephone and the Internet. It has 4 million members world wide and currently fields 350,000 calls a year, including 1,000 “true emergencies.” Its primary function or purpose is “responding to emergency responders” and disseminating medical information as explicitly authorized by members. The member PHR is accessible via the Web, and a USB token application version is being developed. Mr. Harringon also commented on the importance of standards and interoperability and described his organization’s activities in this area.

In the discussion period, the group continued its exploration of interoperability issues, and questions about tensions between protecting patient information privacy and protecting patient health were introduced.

PANEL THREE:

Organizations and Vendors Offering PHRs That Are Derived from EHRs

  • Seth Rupp, Cerner Corporation
  • Peter DeVault, Epic
  • Kristina Rollings, McKesson Corp.

Cerner’s juvenile diabetes PHR initiative is branded as IQ Health. At present, it has 266 unique clients and 642 unique sites in the U.S. The most commonly used functionality is doctor-patient communication; in addition, patients can document self-care and manage a care plan, and providers can conduct asynchronous education and provide feedback. As barriers to interoperability, Mr. Rupp highlighted the lack of integration and common architecture, exacerbated by the proliferation of different Internet portal strategies and technologies. He also pointed to the challenges posed by varied levels of health literacy and Internet access, asserting a national PHR rollout will require a strategy to provide access for lower income populations.

Mr. DeVault introduced Epic’s MyChart application, which he described as an outgrowth of its EMR system. It is a tethered application, connected integrally to the EHR system and providing a “filtered view” of the patient record. The physician typically chooses what the patient will see, and in what form. Like the other panelists, Mr. DeVault described Epic’s approaches to patient authorization, noting that they vary with the business model. He offered thoughts on connectivity and interoperability, and on various ways that the PHR might evolve in relation to EHRs. He urged that the thinking about both the EHR and the PHR be treated as “the same set of problems.”

McKesson provides information solutions across the continuum of care and settings.

Its mobile technology and proactive disease management components are part of applications built “on top of” existing health care organization IT, with functions that pull data from those systems. Its three applications have a range of functions: the Horizon WP Patient Portal, for outpatients; the Horizon WP PatientVision, for inpatients; and the Telehealth Advisor, for home-based disease management. Each was developed collaboratively with a different hospital partner. As barriers to interoperability, Ms. Rollings identified provider unwillingness to share data, MPI resolution, the need to align vendors on a common approach, and the lack of final standards.

In the discussion period, the group further explored the paths to interoperability, mentioned the possible merits of having a predetermined set of data fields, discussed the differences and relationships between the EHR and the PHR, and noted the possibilities for using either locator or aggregator (repository) architecture. The HIMSS “Integrating the Healthcare Enterprise” demonstration project was cited as evidence that sharing data out of network is already technically possible; however, the remaining cultural, legal and political barriers were described as a “huge paradigm shift.”

DISCUSSION OF NEXT STEPS

The Workgroup’s final discussion focused on planning its January 5-6 hearing. Participants highlighted topics of special interest that had emerged from the day’s presentations. Summarizing this discussion, Dr. Lumpkin proposed the following sessions:

  1. preliminary findings on the federal role in PHRs
  2. consumer perspectives
  3. provider-based barriers — physicians and other caregivers
  4. the business case and other business aspects
  5. Workgroup synthesis and discussion of next steps

He noted that this will not be the only hearing on this subject, as there are many other topics to explore. The group also discussed the closely related standards activities (e.g., by HL7) yet to be explored, and the importance of cross-fertilization with the Subcommittee on Privacy and Confidentiality regarding PHR privacy issues. Dr. Lumpkin observed that the Workgroup would need to clarify what models it is talking about before it can address specific privacy questions, and Dr. Deering pointed out that this would need to happen before the Workgroup and Subcommittee co-host a February hearing on the subject.

Returning to earlier discussions about the relationship between PHRs and EHRs, Dr. Huff proposed as the best characterization that “much of the PHR builds on the same technology” as the EHR. He cited role-based authorization, audit trails, a database, and terminology standards as foundational elements that enable both the EHR and the PHR. He also cautioned about the possibility of losing population-based decision support capabilities through the fragmentation or inaccessibility of data.

Finally, in response to Dr. Rollow’s request for feedback (see above), the Workgroup agreed to encourage CMS to move ahead as planned.

DETAILED SUMMARY

CALL TO ORDER, WELCOME, INTRODUCTIONS, REVIEW OF AGENDA

Dr. Lumpkin called the meeting to order. Following introductions, he said the purpose of the meeting is to look into personal health records (PHRs) in terms of the personal health dimension of the national health information infrastructure (NHII). It has been nearly three years since NCVHS issued its report, Information for Health, and began to push for an NHII. Presentations on the uses of health information technology (HIT) at a 2002 IOM-sponsored conference on rapid advances in health care stimulated the Secretary’s interest and led to a surge of activity in this area. As it has monitored these activities, the NHII Workgroup has had concerns about interconnectivity, adequate attention to the public health arena, and ensuring that developments have optimum benefit for patients. These concerns are the context for the current focus on the PHR and the personal health dimension. A second hearing on the subject is planned for January 5-6.

PANEL ONE: Overview

  • Kathleen Fyffe, ONCHIT

Ms. Fyffe noted that on April 26, 2004, President Bush stated that medicine ought to use modern technologies to better share information, reduce medical errors, and reduce costs. He set as a goal that within ten years, everyone should have a personal electronic medical record, personally controlled, and federal government should take the lead to make this happen. The HHS Framework for Strategic Action issued in July 2004 includes personalizing care as a key goal. Ms. Fyffe noted that the universe of health information that consumers can access is overwhelming, and it needs to be relevant and customized to their needs. Encouraging the use of PHRs is a strategy in the HHS Framework.

The federal government uses information technology as a central tool and vehicle to disseminate health information and knowledge to consumers — for example, through the CDC Web site and the forthcoming CMS Medicare Beneficiary Portal. ONCHIT recently coordinated with CDC and private health plans to put links to the CDC flu Web site on the plans’ sites. It also has coordinated with PHR-related initiatives of the Connecting for Health Project and the American Medical Informatics Association. ONCHIT is also examining how PHR usage can be encouraged through consumer enrollment in health savings accounts.

Ms. Fyffe suggested that NCVHS could support these activities by looking at questions such as these:

  • The relationship between PHRs and electronic health records (EHRs)
  • The business case for PHRs
  • How consumers can be encouraged to use PHRs
  • The roles health care industry stakeholders (e.g., providers, employers, health plans, researchers) should play in PHRs
  • How providers can be encouraged to use PHRs and encourage their patients to use PHRs
  • How employers might encourage employees to use PHRs
  • Privacy issues related to personal health information
  • Bill Rollow, CMS

Dr. Rollow said his perspective on PHRs stems from his role as Director of the CMS Quality Improvement Group. Quality has become an important part of the CMS agenda, and IT adoption and use are seen as important means of achieving it. This is one of four strategies for promoting transformational change that are outlined in the 8th Scope of Work of the Quality Improvement Organizations (QIOs). Personal health records are one of five “core elements” in that strategy, along with e-prescribing, EHRs, and other elements.

The Beneficiary Portal that will be tested in Indiana starting in December is a step in the direction of PHRs, though the data are limited. It is seen as something to be built on with additional information, services and tools. As it moves forward, CMS is interested in how it might stimulate innovation to help PHRs develop in directions that can have a positive impact on quality. One area of interest is the functionalities PHRs might have. CMS envisions identifying a descriptive set and then giving preference to PHRs that use them — for example, linking to such PHRs from the Portal. In addition, physician offices could be encouraged and perhaps given financial incentives to contribute data to PHRs that meet the functionality specifications; and Chronic Care Improvement organizations could use these PHRs in disease management programs. Dr. Rollow mentioned several specific functionalities that might be part of the specifications, including a core set of clinical information, self-management tools, informed decision-making tools, communication capacities, and information on how to best use coverage, all with specific privacy and security specifications.

Finally, he stressed that he is presenting these ideas for discussion purposes, and he invited comments on whether creating such specifications is a good idea; if so, how CMS might go about “seeing a set of specifications come into being”; or whether an alternative process would be better. CMS has been talking with the Markle Foundation/Connecting for Health about helping to convene a process to look at 1) the business case to help stimulate PHR adoption and 2) a process for devising a set of technical specifications, functionalities, etc. as a “floor set” for PHRs.

  • David Lansky, Markle Foundation, Connecting for Health

[Slides]

Dr. Lansky focused his comments on the themes that have emerged from the work of Connecting for Health on the PHR and the challenges the federal government faces in moving forward in the near future. Connecting for Health is a consortium of more than 100 organizations from the private and public sectors. Several committees have worked on PHRs for more than 2 years, resulting in the July 2004 report, Connecting Americans to their Healthcare. The themes that have emerged include the fact of significant, pluralistic interest in this field and flourishing activity; major uncertainties about the path forward; and the need for interim policy development to pursue two goals: encouraging innovation while also supporting eventual interoperability.

Dr. Lansky outlined nine “major uncertainties” identified by Connecting for Health, representing all that remains to be learned. Noting his hope that “we can do that learning together,” he said the previous speakers had described pathways for narrowing the uncertainties in areas including the following:

  • “tethered” (i.e., connected to the EHR and other components of health care delivery) or not (i.e., free-standing software)
  • patient-controlled or not (the reality is that many early developments are controlled by the health enterprise, not the patient)
  • comprehensive or specialized approach (the first is optimal, but early uptake is greatest for specialized populations)
  • functional or structural approach (an array of functions, services and benefits, or a literal record or data structure)

Other uncertainties relate to the interaction with care redesign, the degree of physician support, links to other digital data, investment and financing models, and the role of government and other sectors. The government has a number of possible roles with respect to these uncertainties. Dr. Lansky described six possible areas for government activity: defining the PHR, setting data and interoperability standards, providing data, establishing rules and policies for PHR products, developing products and hosting patient clinical data, and offering clinical and prevention services such as reminders.

Finally, he highlighted several policy priorities and strategies for 2005 and 2006. The first is to encourage innovation in a way that is responsive to patient and community needs. The second is to lay the groundwork for interoperability — which he likened to creating a well manicured garden in which a thousand flowers can bloom. Another near-term policy strategy is to lay the groundwork for consolidation of data across time and space. The final two are to learn where the short-term value is and to set a foundation the long-term opportunity to redesign the health system.

In the same period, Connecting for Health will be developing attributes and descriptions of the PHR, identifying elemental core standards and articulating the fundamental basis of interoperability, clarifying the roles of various stakeholders, helping early implementers share findings, supporting a consistent signal to the public about benefits, rights, and their role in health, demonstrating the value of PHR functions, and disseminating early findings.

Discussion

Dr. Steindel noted the synergy among the activities of NCVHS, Connecting for Health, and other organizations and wondered what other organizations might be brought into play. He then asked the panelists to comment on the relative merits of continuous birth-to-death records versus a more life stage-oriented PHR.

Dr. Rollow commented on the importance of coordinating with other HIT/PHR efforts and awareness of their activities. He noted that HL7 has plans in the PHR area and CMS needs to understand how they connect.

Regarding the different PHR models, Dr. Lansky said both lifelong and life stage-oriented records are needed and the infrastructure needs to permit comprehensiveness across time and settings while also meeting individual needs at specific stages without being overwhelming. Connecting for Health research has found that patients are interested in “very specific functions and features” related to a particular stage of life, with comprehensive health information “dimmed in the sideline.” Ms. Fyffe agreed that both need to be available.

Noting that the PHR “ball is in play” now, Dr. Deering proposed that everyone involved in EHRs start providing for the eventuality of interoperability with PHRs, since things will move in that direction eventually. Dr. Rollow agreed. Dr. Deering noted the early-stage HL7 activity on the EHR DSTU, which will provide some of the raw materials for PHRs.

Asked about patient and practitioner adoption of PHRs, Dr. Lansky described the survey research findings described in the Connecting for Health report. People with a chronic illness who are between ages 45 and 65 show the greatest combination of interest and capacity to use the tools. Clinicians are most involved now in provider-sponsored consumer portals, in which the doctor is the gatekeeper to personal health data. He estimated that a third to a half of the doctors in these systems are comfortable making this available to patients.

Dr. Ortiz questioned the extent of consumer/patient interest in PHRs, citing recent Kaiser Permanente research to be published in the January/February 2005 JAMIA, as well as and other findings that show low patient interest in seeing their medical records and using care-related services such as asking the advice of clinicians. He cautioned against assuming that there will be large consumer demand. He also wondered whether a fully developed EHR might obviate the need for PHRs as a separate entity.

Dr. Lansky agreed about the importance of being “agnostic” until solid interest is proven. Anecdotally, Connecting for Health found that 15 percent of patients take up PHRs in response to an aggressive effort to offer them. The intensity and sustainability of use is not known, but a sustained 15 percent of the population would represent significant uptake. As for the second question, he said Connecting for Health agrees that the EHR and PHR will share common components. It stresses, however, that patients, not providers organizations and other institutions, will need to control some personal information and input and manage their own data and control access, and this is less likely in an EHR construct. Another consideration is what approach will best further the transformational objectives associated with a less provider- and institution-centric way of offering health care services and encouraging personal health management.

Dr. Lumpkin cautioned against closing the door on experimentation simply because the use case is not entirely clear. Noting that the term “personal health record” does not fully capture the functionalities and transformations that NCVHS had in mind with respect to the NHII personal health dimension, he talked about the potential use of such a tool to augment the information patients receive (and perhaps fail to understand) in clinical encounters. He asked Dr. Lansky what Connecting for Health had found about the importance of connectivity to the medical record for patients, and was told patients do regard this as important. Dr. Lansky added that given this interest, the viability of untethered models remains to be seen.

The question of providers’ concerns about being deluged with email from patients arose, and Dr. Ortiz said so far it has not emerged as a problem. Reimbursement does remain an issue, however, complicated by disparities between what doctors want to be paid for the interaction and what patients are willing to pay.

Dr. Huff, who co-chairs the HL7 vocabulary committee, observed that while it is important not to prematurely define the PHR, eventually it will be useful to have agreed-upon meanings for terms. He asked about the thinking about various roles with respect to standards. Dr. Rollow said that as with EHRs, both data and functionality standards will be needed for PHRs, aided by descriptive terminology. He predicted that the current mode of working for consensus on standards would be the way to move forward, rather than any rule-making by the government.

Dr. Huff predicted that after investing heavily to create a comprehensive EHR, institutions will not be willing to cede responsibility to anyone else to control their data; moreover, maintaining it in two places would be doubly expensive. He asked how the economics would work. Dr. Rollow said this issue is integral to all information exchange and a focus of current thinking. He described the notion of two versions of the record, a production version and an information-exchange or reporting version. Dr. Lansky said the Connecting for Health approach has the PHR residing in an integrating platform, along with other information sources, akin to the Quicken model.

PANEL TWO:

Organizations and Vendors Offering PHRs That Are Not Derivatives of EHR Systems

  • James Walker, M.D., Geisinger Health Systems: Review of Patient Uses of PHRs

[Slides]

Geisinger serves about a half-million patients in 31 mostly rural counties in central Pennsylvania. Its population is older, poor and less mobile than the U.S. average, but no less connected to the Internet. Geisinger completed its outpatient EHR in March 2002, and built a “patient EHR” as a view of the same electronic record, shared with external physicians. Geisinger’s goals, as part of the basic goal of transforming care, were to “delight patients” (who expressed great interest in it in focus groups and surveys), to empower patients, to improve practice efficiency, and to delight providers by “making the clerical work of health care disappear.”

Currently, MyGeisinger has 17,000 active users and uptake is accelerating; about 2,000 were added in the last month. Dr. Walker described the variations in patient interests and needs, e.g., children’s immunization records for young parents and proxy access for the adult caregivers of older relatives. The functionalities are in two categories: 1) viewing records (including problem list, allergies, medicines, histories, lab results and more) and 2) sending and receiving messages (including drug refills, appointment requests, referral requests and requests for medical advice). Patients are also sent reminders (reducing no-shows from nine to one percent).

Viewing lab results and seeking medical advice have been by far the greatest uses. In the latter area, Geisinger’s experience is consistent with research findings that the traffic is manageable and physicians find it acceptable, or even preferable to other forms of communication. Dr. Walker noted that Geisinger decided to write its own interpretations of all test results after finding all those developed by others “flatly inadequate” because of being at too high a reading level, frightening, and/or unhelpful. Drs. Hassol and Walker will have an article in the November/December 2004 JAMIA on patient and physician experiences with the patient EHR. In general, patients are asking for more information, and faster, and usability is critical.

Turning to the barriers preventing broad uptake of PHRs/patient EHRs, Dr. Walker observed that having the EHR in place made it far easier to add a patient-oriented module; so lack of EHRs designed to support PHRs is a barrier. Others are the lack of effectiveness evidence, the energetic and intellectual costs of change for physicians, financial reimbursement issues, and patient acceptance. He added that 30 percent of today’s patients want this tool, and Geisinger expects this number to grow, “even in a fairly disadvantaged population.” He reiterated the importance of having the PHR be usable for populations with diverse capacities and needs. Finally, he noted that much of the medical record today is unusable, and that legal complexities exist. Next steps involve demonstrating the potential benefits to both providers and patients.

  • Brian Baum, Health Record Network

[Slides]

The Health Record Network (HRN) was launched by Duke University. Citing the HHS Framework for Strategic Action, Mr. Baum observed that the vision exists for an information-rich, consumer-centric national future; the question is how to get there, which involves moving a huge industry. HRN’s strategy for achieving a fully integrated health information network involves developing standards, deploying systems, equipping providers, raising funding, providing incentives, creating local networks, and linking them. Each step carries an inherent risk, and it is difficult to gain momentum, especially after the dot-com crash reduced the general economy’s patience and support for change in the health care industry.

HRN sees its strategy as complementary with national strategies like the HHS Framework. After analyzing the options for moving the industry forward, HRN concluded that the strongest force was consumers, reasoning that even 10 percent of roughly 300 million people “is market force.” To engage consumers, they are segmenting the market and targeting early adopters—essentially the same groups identified by Geisinger and other sources. Next is creating awareness of the need to manage information and the benefits of PHRs, then providing an easily accessible, universal solution. HRN envisions providing a simple, valuable, consumer-controlled product, recognizing that the industry is not yet ready for interoperability. Mr. Baum showed a screen shot of the HRN personal health record. The initial record information includes family medical history, medications, allergies, and information on the clinical care team.

HRN is a not-for profit-organization with no institutional affiliation; it aims to create a national health utility to benefit all health consumers. It is “not in a competitive space” and can give rise to new business models. It is now assembling a Board of Directors and establishing relationships in the business, health and consumer communities. Mr. Baum stressed that HRN intends to complement the efforts of the NHII Workgroup and similar organizations building the clinical infrastructure. In terms of interoperability, the HRN is envisioned as an “overlay network” that supports consumers and links into EHR systems. At present, the State of Wyoming has approved a statewide pilot of HRN, and all patients of Duke Heart Center will have a chance to establish an HRN record that links to clinical and imaging data. In addition, HRN has been contacted by the Canadian government about conducting a parallel effort in Canada.

  • Scott McFarland, SimplyWell

[Slides]

SimplyWell LLC is jointly owned by the Nebraska Medical Center and 625 physicians in private practice in the central and high plains. Mr. McFarland said SimplyWell attributes its success, with “tremendous penetration in Nebraska, to the nature of SimplyWell’s ownership. The company now has a presence in 41 states, 49 by mid-2005. The product is part of employee benefit structures, primarily for large, self-funded employers. Participants are working age adults. SimplyWell provides yearly work-site health screenings including physical metrics and labs, on the basis of which individual health action plans and electronic health records are developed. The database is also a source of interested research subjects for University of Nebraska researchers, who can query the database.

SimplyWell is working for the elimination of avoidable medical claims, especially those related to lifestyle. Integrated disease management models are also being developed. Participants are given financial incentives to view their EHR, delivered through electronic cards. The company strongly encourages use of the nurse call center and access to online materials. Encouraging behavior change is a major thrust of the program. The model is being adopted, Mr. McFarland said, because it supports health care relationships continuously, not just in times of illness, and it “leverages the least utilized resource in U.S. health care delivery, the patient.” When participants do receive episodic care, the electronic health record is “there for the clinicians to utilize.” He noted that the record is interoperable.

SimplyWell is paid for by self-insured employers in the public and private sectors and physician practices (“physician owners”). Fully insured employers and small businesses are left out because group plans will not let them bill through the group plan, putatively because carriers are unwilling to invest in keeping enrollees healthy due to the high rate of turnover. Mr. McFarland called attention to House Bill 2501 in Pennsylvania, a disease management insurance tax credit policy that he hopes will become a model for other states, helping small businesses invest in preventive medicine and disease management. He also described SimplyWell’s consent authorization and control procedures, noting their finding that patients want to be in complete control of who sees their health information.

  • David Harrington, MedicAlert Foundation

[Slides]

MedicAlert is a non-profit foundation, founded in 1956, that protects and saves members’ lives mainly by providing information about them in emergency situations. It has kept EHRs since 1985. Most communication is by telephone and the Internet. It has 4 million members world wide and currently fields 350,000 calls a year, including 1,000 “true emergencies.” Its primary function or purpose is “responding to emergency responders” and disseminating medical information as explicitly authorized by members; it also notifies family members. MedicAlert members are in all age categories (e.g., 19 percent are younger than 19 and 30 percent are older than 64), distributed among the major serious chronic diseases. The business model is based on member subscriptions. The Foundation does outreach through partnerships with disease related organizations and provides sponsored memberships at the county clinic level.

The service enhances and speeds diagnostic assessment. The member PHR is accessible via the Web, and a USB token application version is being developed. The functionalities include patient medical history, digital images, family and physician contact information, and more. MedicAlert does advance directive management and drug utilization reviews and provides clinical trial protocol support and management that includes adverse event notification. The functionalities are developed based on “constant member interaction and feedback” through surveys and focus groups. MedicAlert polls physicians and other emergency response personnel as well as members to assess the value of various services and functions.

The Foundation is active in national standards development efforts. Mr. Harrington enumerated some of these activities and the specific standards used by MedicAlert, and said the Foundation is a “strong believer in interoperability.” The evolution of standards and the transition from message-based to services-based interfaces are major barriers; the HL7 EHR functional outline is a “great step forward.” He added that MedicAlert plans to “tether itself” to other parts of the health care industry and to implement interoperable interfaces with payers, providers, health plans and physician practice management systems and other information sources.

Finally, Mr. Harrington offered two comments in response to Dr. Ortiz’s earlier question about the demand for PHRs, first noting that making it possible to “actually save a life” must be factored into the answer. Second, he asserted that surveys about the extent of interest in using the PHR are “backward looking,” in contrast with a perspective based on emerging frameworks and contexts. He said he liked the idea of a public utility, which is the way MedicAlert thinks of itself.

Discussion

On interoperability, Dr. Ortiz noted that while everyone says they support it he wondered how much it is actually taking place between different organizations (e.g., those in this panel). Mr. McFarland said that SimplyWell has staff members who regularly do practice management consults with physicians across the U.S. to allow them access to the EHRs of their patients who are SimplyWell members. Duplicate tests are caught at the claims level and physicians are not paid for duplicative tests. Asked if the Nebraska-based participating physicians have an EHR, he said SimplyWell uses open source technology, built on a Linux platform. Its members are driving their physicians and systems to use this collection point for information. The idea is to have a consumer-based push to get providers to use the system interoperably. He advocated as a general approach starting with clean datasets for working-age populations, getting them used to using a PHR, and thereby generating a “consumerist push” toward provider participation and interoperability. He questioned the value of converting existing data on Medicare and Medicaid populations. He added that PHRs and EHRs did not start out as the focus of SimplyWell, but this dimension “unconsciously developed” as part of its preventive medicine services.

Continuing on this topic, Mr. Harrington stressed the critical importance of standards and cited the hard work going on in HL7 on the EHR. He added that it is more important to develop interoperability standards than data standards between disparate systems. He invited the large vendors on the afternoon panel to join in these efforts.

Mr. Blair asked about the panelists’ experience with conflicts between protecting patients’ privacy and the varied uses of information to protect their health. Mr. Harrington said MedicAlert has not encountered such conflicts, because of the nature of its work and its authorization mechanisms. Mr. McFarland said since SimplyWell draws blood all over the country, it encounters public reporting issues associated in particular with infectious diseases, requiring that they go through the appropriate channels while working to protect the patient. For Geisinger, Dr. Walker said physician access to patient health histories is often delayed because of the HIPAA requirement for specific authorization. He added that they expect to be able to use the patient EHR to enable patients to provide electronic authorization in real time, eliminating that delay.

Dr. Deering said a future hearing would address the related question of documentation for consent and authorization.

Dr. Lumpkin asked about automatic electronic updates to patient records from sources other than the patients. The responses were that MedicAlert is working on that; SimplyWell’s labwork results are sent electronically and interoperably; and HRN is building interoperability for such functions into its model.

PANEL THREE: Organizations and Vendors Offering PHRs Derived from EHRs

  • Seth Rupp, Cerner Corporation

Cerner’s juvenile diabetes PHR initiative is branded as IQ Health. At present, there are 266 unique clients and 642 unique sites in the U.S. Mr. Rupp identified five main categories of PHR users:

  • the healthy but concerned
  • the healthy but at risk
  • the chronically ill
  • proxies for health care (e.g., for “parents and peds”)
  • those seeking efficiency (e.g., for making appointments)

Cerner believes that the third category, those with chronic illness, are the most motivated to use PHRs as part of their daily health management regimen. After noting the multiple business models in the PHR space, Mr. Rupp described Cerner’s interoperable community model. There are four steps in the model: creating the EMR; community outreach to consumers, plans, employers, and others; structuring, storing and studying the evidence to create new knowledge; and implementing knowledge-driven care based on this learning.

As barriers to interoperability, he said the issues concern fundamental architecture and technology, e.g., lack of integration and common architecture. These are exacerbated by the proliferation of different Internet portal strategies and technologies that are creating new integration challenges. “Branding is something huge,” in addition to security concerns. He noted that for PHRs, integration barriers vary based on the scale of implementation. Cerner maintains all its data in one common place, with users accessing the data through different interfaces. It uses standardized medical terminology and vocabularies. Only clinicians with “the appropriate relationship” to the patient or explicit special patient authorization are allowed access to patient data; proxies must also be given explicit authorization.

The most commonly used functionality is communication between patients and their doctors’ offices. Patients can also document self-care and manage a care plan established with providers, and providers can conduct asynchronous education and provide feedback. Other functionalities include a specialized home diary and a historical view. Cerner has given every juvenile diabetic in the U.S. a personal IQHealth, providing a PHR, a diabetes management center, and messaging capabilities. Diabetes was chosen because it’s very interactive, it’s a family disease, and it’s very data intensive.

Mr. Rupp noted the challenges associated with reaching out to users with varying degrees of access to care and levels of health literacy, and said that as PHRs are rolled out nationally there must be a strategy to give lower income populations Internet access. IQHealth content is written at a “common level” of literacy and education, and non-English language support is provided. Some outreach is done through the Juvenile Diabetes Foundation. The primary interest and partnership to date is from pediatric hospitals. Providers can offer connectivity with the clinical EMR.

  • Peter DeVault, Epic

Epic is primarily an enterprise EHR system with adjunct services such as scheduling billing, and so on. Its MyChart application, a PHR, is “a layer of services offered to consumers on top of that set of applications.” It enables two types of functionality, access (to information and services) and communication (to providers, customer services and payors). Essentially, MyChart strengthens relationships between consumers and these entities. It is a tethered application, connected integrally to the EHR system and providing a “filtered view” of the patient record. The physician typically chooses what the patient will see, and in what form. (Mr. DeVault noted that as such, this “doesn’t really solve the problem,” described by President Bush when he called for a PHR in 10 years, of having to cart around notebooks full of medical records to various providers. Solving this would require 100 percent of the patient’s cumulative medical record.) However, it does allow for a shared interpretation of medical record data, and the concept of a shared health record is central to the Epic notion of the PHR.

Mr. DeVault described Epic’s approaches to patient authorization , noting that they vary with the business model. Epic customers use a wide variety of business models, with subscription as the most common. Currently, 156,000 people log on to Epic at least once a year, and 35 to 40 percent log in on any given month. The largest demographic segment is women in their 40s. Epic provides customers with tools to reach special populations, for example, with disease management tools. The company is developing several functionalities targeted to specific diseases, starting with cancer.

Mr. DeVault said he views MyChart’s development as an outgrowth of Epic’s EHR as a model for what is likely to happen in other places. Services were developed in response to its customers’ (health care organizations’) requests, themselves responding to consumer interests and also seeking competitive advantage. Opening up connectivity through the PHR allows patients to take advantage of the information resources on the Internet.

He envisioned an evolution in which the “tether” between the PHR and the EHR weakens and the PHR begins to orbit, first its “parent” EHR system and then several EHR systems, after which it would develop the features of stand-along systems and the ability to connect to multiple data sources. Based on this analogy, the PHR can be thought of as “a special case of the EHR system. He cautioned against “trying to solve the same set of problems twice” and urged that the thinking about both the EHR and the PHR be treated as “the same set of problems.” This way, the problem of interoperability can be solved for both EHRs and PHRs. PHRS may be a special case of an EHR, or they may a functional subset of an EHR system; in either case, the two need to interoperate.

  • Kristina Rollings, McKesson Corp.

McKesson provides information solutions across the continuum of care and settings. It developed its patient-centered care strategy as a way to leverage health care organizations’ existing information technology and investments and to create a portable interface for patients. Ms. Rollings described their formative market research on what health care organizations want, what consumers are doing online in other industry markets, and consumer trends regarding self-service functions. It also surveyed its customer base to determine what functions patients are requesting and consulted strategic advisory councils, patient focus groups and research by Pew and other sources. The result is the company’s patient-centered care strategy, with mobile technology and proactive disease management as important components. The applications are built “on top of” existing health care organization IT, with functions that pull data from those systems.

There are three applications, each with a broad range of functions: the Horizon WP Patient Portal, for outpatients; the Horizon WP PatientVision, for inpatients; and the Telehealth Advisor, for home-based disease management. Each was developed collaboratively with a different hospital partner.

To handle user consent, authorization and control, these systems use three user types—guest, registered and connected—with different requirements and levels of service. Ms. Rollings also described the standards used by McKesson. At this point, they believe they’re “interoperable within the McKesson source system background,” and they are working to become interoperable with other vendors in order to cross multiple health systems and fit within multiple PHR models. As barriers to interoperability, she identified provider unwillingness to share data, MPI resolution, the need to align vendors on a common approach, and the lack of final standards.

Discussion

Asked by Dr. Lumpkin about the status of Epic installations with respect to interoperability, Mr. DeVault said they are developing a Care Everywhere platform, which Kaiser users will be the first to use. To accomplish this, all Kaiser regions have agreed to standardize some vocabulary and build their systems in a particular way. Other “less tightly collaborative communities of Epic users” will eventually be able to share data without sharing procedure master files and other sets of vocabularies.

Dr. Kambic challenged the proposition that the EHR and the PHR are the same thing and stressed in particular the need to substitute consumer-friendly terminology for medical jargon. He referred to the PHR attributes listed in the latest Connecting for Health report, noting some (e.g., consumer control of access) that are incompatible with EHR attributes. He urged the present panelists, whom he called “the big players,” to think about these unique attributes and not just equate EHR and PHR functionalities. Mr. Rupp agreed that the two are distinct. He depicted a set of multiple functions, views and terminologies—possibly on a continuum—that include the standalone PHR, the standalone electronic medical record (EMR), and the patient view into the EMR as well as the EHR. The value of the EMR is connectivity to the physician. Mr. DeVault agreed with Dr. Kambic that the PHR is distinct, but he asserted that “it’s largely a subset or superset of functionality of an EHR system, and EHR system technology will enable PHR systems.” He added that the need to translate raw health record information into something useful to patients “argues for the tethered model where the PHR is an outgrowth of the EHR,” with its beneficial presence of a provider who can decide how to present and interpret data. Ms. Rollings agreed and described McKesson’s applications as a “toolset” that allows data in the patient health record to come from multiple sources including the EHR and patient entry.

In response to a question from Dr. Deering, Ms. Rollings said she thought a predefined set of data fields would be helpful in overcoming providers’ resistance to sharing data; Mr. DeVault agreed.

A question from Mr. Hungate prompted panelists to describe the ways in which their tools enable patients to record and track functional status and related matters, as distinct from health care.

Dr. Steindel observed that the system being described by the various participants is essentially one system with a continuum of attributes, for use in a variety of settings across a patient’s lifetime.

Dr. Deering commented on the differences between locator architecture and aggregator (repository) architecture and cautioned against assuming that information will be held in the latter model when a “finding tool” might be developed that can execute particular tasks for specific needs. Dr. Steindel noted that current systems being tested all use the repository model; the other has not been tested. Mr. DeVault noted the related questions about ownership of data. Both he and Ms. Rollings said that their systems use a locator model, pulling together data on an as-needed basis. Other comments, including one later from Dr. Huff, suggested that experimental efforts with the locator model were forthcoming.

The panelists were asked to realistically predict the likelihood of being able to share data outside their systems once the barriers of provider resistance to data-sharing and interoperability hurdles are resolved, and what would need to happen to move in that direction. Stating that “It’s happening right now,” Mr. DeVault described the HIMSS “Integrating the Healthcare Enterprise” demonstration project, which is being conducted using existing technology and standards, “prov[ing] that it can be done” technically once the cultural, legal and political barriers are eliminated. Mr. Rupp said that “the common denominator today is HL7,” and different business entities and corporations are already sharing data. He predicted that initiatives such as RHIOs or LHIIs will “force the interoperability” at a broader level. Dr. DeVault added that the demand for EHR systems to be interoperable with other local or regional systems will force vendors to develop those systems. Ms. Rollings noted that the technical piece is the easiest one to solve; the harder ones have to do with workflow in the health care industry, where information-sharing represents a “huge paradigm shift,” as well as for patients, who have never had this kind of access.

Public Comment

From the audience, Robin Kaye expressed concern that a “system of networks” communicating and transferring patient information would in effect be a “national patient information database,” and she pointed out that some people do not want their medical information put in an electronic format or made accessible interoperatively without their consent or shared with third parties. She said she was assured at a 1997 NCVHS meeting that there were no plans to create a national patient information database, and questioned the compatibility between the President’s stated goal of everyone having electronic records and HIPAA provisions that allow patients to ask for restrictions. Finally, she stressed that in her experience, “you can’t always presume it’s a wonderful thing to share information.”

Dr. Lumpkin stated that the Committee had not developed advice for the Department on this matter, and the Subcommittee on Privacy and Confidentiality plans hearings on the issues raised by Ms. Kaye.

DISCUSSION OF NEXT STEPS

The Workgroup’s discussion focused on planning the January 5-6 hearing. Dr. Lumpkin highlighted two topics of interest that had emerged from the preceding panels:

  • patients’ views of the PHR and what they want from it
  • the antitrust issues related to sharing patient information

The idea of exploring consumer/patient perspectives and experiences generated considerable support from participants. There was interest in hearing both from patient/consumer advocacy organizations and from individuals currently using PHRs. Although privacy issues were raised in relation to the patient perspective, the group decided to address them separately in conjunction with the Subcommittee on Privacy and Confidentiality.

Dr. Kambic suggested categorizing and segmenting the continuum of PHR tools and approaches. Dr. Deering noted the existence of “pre-PHR tools” that start out on paper and are translated into non-interoperable electronic records; she wondered if the Committee would want to learn about these models. Dr. Lumpkin expressed interest in exploring the interface between the PHR and telehealth.

Returning to earlier discussions about the relationship between PHRs and EHRs, Dr. Huff proposed as the best characterization that “much of the PHR builds on the same technology” as the EHR. He cited role-based authorization, audit trails, a database, and terminology standards as foundational elements that enable both the EHR and the PHR. On the subject of the various models, he noted that Utah has been funded to do a locator model of the EHR and it also has elements of the aggregation model; he noted that there is not necessarily a “clean” distinction between models. To his comments on several technical issues to be resolved, Dr. Lumpkin said those would be best addressed by the Subcommittee on Standards and Security. Finally, Dr. Huff cautioned about the possibility of losing population-based decision support capabilities through the fragmentation or inaccessibility of data.

Dr. Deering noted that the HHS Office of Disease Prevention and Health Promotion has commissioned a “mini whitepaper” from David Lansky and colleagues on the policy issues for the federal government in getting involved in PHRs, and she suggested that the Workgroup might want a report on its preliminary findings in January. She also suggested exploring the emerging “ancillary industry of content” for translating medical knowledge into general information, education and decision support for consumers. She noted its potential to become a trusted, licensed source for interpreting test results for patients, perhaps thereby making providers more friendly toward PHRs. Dr. Lumpkin agreed about looking at this arena but cautioned against the notion of relieving personal physicians of an interpretive role with their patients.

Ms. Fischetti suggested hearing from non-physician health care professionals who contribute to the medical record about their fears, which may be different from those of physicians. Another professional group of interest is those in charge of the business of health care. Dr. Huff suggested learning about not just antitrust issues but more broadly, the business case and business issues related to sharing clinical data with competitors. He wondered if this is something the government will have to “fund forever,” and if not, what sustaining business model would cause people to exchange and share data.

In summary, Dr. Lumpkin proposed the following sessions for the January hearing:

  • preliminary findings on the federal role in PHRs
  • consumer perspectives
  • provider-based barriers — physicians and other caregivers
  • the business case and other business aspects
  • Workgroup synthesis and discussion of next steps

He noted that this will not be the only hearing on this subject, as there are many other topics to explore—e.g., the interrelationship with telehealth. The group also discussed the closely related standards activities (e.g., by HL7) that have yet to be explored. The group agreed on the proposed general structure and on approaches for filling out the panels. They discussed the importance of cross-fertilization with the Subcommittee on Privacy and Confidentiality on PHR privacy issues. Dr. Lumpkin observed that different models present different privacy issues, and the Workgroup would need to clarify what models it is talking about before it can address specific questions about privacy. Dr. Deering noted that this would need to happen before the Workgroup and Privacy Subcommittee co-host a February hearing on the subject.

In response to a question about the intended end result of the Committee’s deliberations, Dr. Lumpkin noted the rapid Departmental activity in this area, under the direction of the Office of the National Coordinator. The Committee’s intention is to enhance and contribute to that process as needed. He noted that the Committee is available to assist CMS and ONCHIT on request—for example, by holding hearings on specific questions.

Dr. Kambic reminded the group of Dr. Rollow’s request for feedback on proposed CMS activities in this area—i.e., asking Connecting for Health to gather experts and outline a business case and technical specifications; and then perhaps in Fall, 2005, taking these to an SDO and asking it to outline technical standards that could be translated into interoperable PHRs. Mr. Blair observed that the traditional SDOs are already working together on interoperability; he cited the HIMSS demonstration project in which several members of Panel Three are involved. The challenge, he said, is to ensure that the work on interoperability is inclusive and broad enough to accommodate new ideas and new models, such as those represented in Panel Two. Dr. Lumpkin noted the long timeframe CMS envisions for developing guidelines, allowing the market to mature, and the Workgroup agreed to encourage CMS to move ahead as planned.

Dr. Lumpkin then adjourned the meeting.

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

s

_____________________________________________________________________

Chair Date