The National Committee on Vital and Health Statistics was convened on Tuesday and Wednesday, March 3 and 4, in the Hubert H. Humphrey Building in Washington, D.C. The meeting was open to the public. Present:
Don E. Detmer, M.D., Chair
Jeffrey Blair
Simon P. Cohn, M.D.,
M.P.H.
Kathryn L. Coltin, M.P.H.
Kathleen Frawley, J.D., M.S.,
RRA
Daniel Friedman, Ph.D.
Kathleen Fyffe
Robert M.
Gellman, J.D.
Richard Harding, M.D.
Lisa Iezzoni, M.D., M.S.
Clement Joseph McDonald, M.D.
Paul Newacheck, Dr.P.H.
Barbara
Starfield, M.D.
George H. Van Amburg, M.P.H., M.S.
M. Elizabeth
Ward, M.N.
Hortensia Amaro, Ph.D.
John Lumpkin, M.D.
Vincent Mor, Ph.D.
Marjorie Greenberg, National Center for Health Statistics (NCHS),
Executive Secretary
James Scanlon, HHS, Executive Staff Director
Lynnette Araki, NCHS
J. Michael Fitzmaurice, Ph.D., Agency for
Health Care Policy and Research liaison
Bob Moore, Health Care
Financing Administration liaison
William Braithwaite, M.D., Ph.D.,
ASPE
Jackie Adler, NCHS
Teresa Cendrowska, ASTM
Helene Guilfoy, CHI
Carole Stephens, RN, Medic Computer Systems
Fred Decker, AHCA
Phyllis Doulaveris, NCHS
Chris Bergsten, AAHP
Jeanne
Gilliland, CDC
Pam Johnson, Amer. Soc. of Cataract & Refractive
Surgery
Jack Emery, AMA
Betsy Humphreys, National Library of
Medicine
Elizabeth Thomas, EDS
Lisa Fusco, Gordon & Barnett
Robert Beck, MultiState Assoc.
Michael Uelle, EDS
Mark
Epstein, Sc.D., NAHDO
Jim Klein, EDS
Carolyn Rimes, HCFA
Brenda
Gillum, CPSC
Karen Sealander, McDermott, Will & Emery
Krista
Robinson, EDS
Brian Thiel, PMA
Mike Tate, American Dental Assn.
Dianne Schneidman, American College of Surgeons
Chris Mancill,
HFMA
Paul Tang, Northwestern Memorial Hospital
Robert Bolan,
McLean Group
Diana Engel, Part B News
Joel Rosenblatt, CHDQ
Kathryn Foxhall, St. Anthony Publishing
Jeanne Scott, NDC Health
Information Systems
Mary Moien
Rich Searles, ARCA Systems
Barbara
Hetzler, NCHS
Dennis Melamet, Health Information Privacy Alert
Michael
Werdner, Arca Systems
Brenda Jackson, Integrated Medical Management
Sherry McKenzie, OASD
Paul Elstein, HCFA
Isaac Hautman, FDA
Melvyn Greberman, M.D., FDA
Col. Lynn Ray, Department of Defense
Arnold Milstein, M.D., Pacific Business Group
Janet Spinks,
AQUAS, Inc.
Diedra Abbott, College of American Pathologists
Henry
Heffernan, EDPNS
Suzanne Neimeyer, Aspen Publishers
Elizabeth
Morrow, Natl. Health Care Anti-Fraud Assn.
Chris Peterson, author
Dr. Detmer welcomed three new members: Kathleen Fyffe, Dan Friedman, Ph.D. and Paul Newacheck, Dr.P.H. He also welcomed Dr. Fitzmaurice as the new AHCPR liaison, and was asked to send a letter of appreciation of to Harvey Schwartz and David Garrison.
Mr. Scanlon reported on a new race and ethnicity data initiative: The Secretary has instructed agency heads to implement the new OMB policy on race and ethnicity in their agency data planning and collection cycles, and to try and collect useful supplemental data relevant to disparities between population groups. He also described planning and performance measurement activities in response to the Balanced Budget Amendment, and other work on survey integration and Healthy People 2010. He invited Committee advice about provider surveys.
Mr. Fanning reported on privacy activities on the Hill and work on a European Union directive limiting data exchange with countries without adequate protections.
Dr. Braithwaite reported that three NPRNs are due for publication very soon, and a Notice of Intent on individual health identifiers and two more NPRNs are due about six weeks after that. The hope is that an NPRN, reflecting new industry consensus, can be issued on the individual identifier by fall. The administrative simplification web site has a description of the complex clearance process for rules.
Dan Waldo and Liz Richter reported on HCFA's extensive work in compliance with the BBA, both developing data systems and preparing an extensive public information campaign. The new information campaign for consumers has five components: printed literature, media campaigns, an 800 number, face-to-face information to beneficiaries, and an Internet site (Medicare.gov). A knowledge base is being developed to support this campaign. HCFA is also developing a new model for data processing and information systems. These extraordinarily expensive systems and processes have not been fully funded by Congress.
The collection of encounter data for use in risk adjustment is the biggest data challenge over the next three years. Ms. Richter, who heads the workgroup deciding what non-inpatient hospital data to collect, described the work in that area. Unlike hospital inpatient data, this area was left to the discretion of the Secretary. To do risk adjustment, data are needed for all Medicare beneficiaries. For the first two years, only inpatient data will be used as a transition while HCFA's objectives and systems are developed. When the data can be competently collected, stored and used, HCFA will move to more comprehensive data collection, starting with physician and outpatient data.
Committee members expressed considerable interest in this project, together with some concern that it will eclipse other important HCFA data projects. They noted that both the Congressional timeframe and budget for this effort are inadequate, and the speakers welcomed the Committee's offers to support more realism in both areas. HCFA will begin "managing expectations" when it has determined what it can and cannot do. Mr. Gellman called attention to the privacy and fraud issues surrounding assigning people a risk adjustment number. Ms. Richter said the diagnostic groups and multiplier formulas would be published. The data systems will be consistent with HIPAA standards.
Dr. Detmer asked Dr. Starfield and Dr. Iezzoni to decide which of their Subcommittees would track this activity.
Richard Sorian, Deputy Director of the Commission, reported on the history and status of the Commission and the two documents it is producing. The Consumer Bill of Rights and Responsibilities was issued in November 1996 and quickly embraced by the President. It outlines consumer rights such as to information, choice, confidentiality, and appeal, as well as consumer responsibilities for such things as healthy living and self-care.
The Commission's 14-chapter final report is expected in mid-March. Among other things, it will recommend creation of two bodies -- one responsible for identifying national quality objectives and tracking progress on them; and the other responsible for advising on quality measurement. The report also urges group purchasers to be more active and coordinated in demanding quality services.
Dr. Detmer and others noted the National Committee's strong and long-standing interest in promoting quality and facilitating collaboration between the private and public sectors. The Committee discussed, revised, and ultimately approved a letter to Secretaries Shalala and Herman on the Consumer Bill of Rights and responsibilities, supporting the basic ideas and principles and commenting on a few areas. The letter also will be posted on the Committee's web site. They agreed to review the Commission's report upon its release.
The Committee heard presentations from Dr. Bart McCann of HCFA and Dr. Mark Segal of the American Medical Association. Dr. McCann reviewed the history of the codes and the guidelines, and discussed their importance to HCFA. He then addressed the controversy that has developed around the guidelines. Dr. Segal described the reaction of the medical community and the steps AMA and HCFA are taking in response. HCFA has extended the time for using the old guidelines to July 1998, and another extension has been requested. The AMA is spearheading a revision process in consultation with many medical organizations. The Committee was invited to attend an April meeting in Chicago for this purpose.
In the discussion session, several Committee members who are physicians described the alarm of their colleagues about possible severe penalties for inadvertent errors. The group discussed the misinformation circulating, and the panelists explained the reasons the guidelines are needed to distinguish levels and types of service. There was general agreement about the need to reduce the burden and improve physician education.
The Committee heard presentations from NCVHS member Kathryn Coltin, Dr. Paul Elstein of HCFA, James Tierney of NCQA, and Dr. Arnold Milstein of the Pacific Business Group on Health. Ms. Coltin gave an overview of the kinds of data issues encountered in administrative and survey data. Administrative data suffers in the areas of completeness, accuracy, timeliness, and consistency. A major problem with survey data for managed care organizations is low response rates and resulting response bias, because members are being over-surveyed. She enumerated some of the problems caused by these data inadequacies, including an inability to assess aspects of care, an undermining of the credibility of the data, a dependence on chart review, and an inability to target needed interventions.
Dr. Elstein noted that HCFA is part of the large collaborative effort that has produced HEDIS. He focused his report on a recent HCFA study of Medicare plans, using HEDIS measures. The research found that plans' EDI capabilities vary widely, that many had never done HEDIS before, and that many have difficulty getting good data. Also, many contracts are not set up to get good encounter data. Dr. Elstein also described other HCFA efforts and plans to assess the quality of care and the quality of data used to report care.
Mr. Tierney reported on NCQA's work to incorporate performance measurement into the accreditation standards process. New standards and guidelines for auditing HEDIS information will be piloted in the summer of 1998. A 1997 survey found a range of EDI capabilities and levels of data quality among plans. He outlined some of the specific weaknesses, such as not reporting clinically important conditions and not having information on public health services in the community. Backlogs of unentered data are a problem both Mr. Tierney and Ms. Coltin mentioned. Mr. Tierney said NCVHS can help by encouraging plans to improve their capture and use of available data and to create an environment that rewards data automation.
Dr. Milstein described a current health care system with so little information on quality that he likened it to an airplane allowed to fly with semi-opaque windshields. He challenged NCVHS to support motivated purchasers by advising the Secretary to leverage the "excellent multilateral thinking" that went into volume IV of HEDIS 3.0. He called attention to several means of improving quality, including standardized mandatory reporting of laboratory values and pharmacy data.
The discussion period focused on the possibility of rating plans for their information systems, applying HEDIS measures to non-managed care providers, and improving response rates of consumer surveys. The group also discussed confidentiality issues with the Health of Seniors survey and questions about the actionability of observed changes in SF36 measures. Dr. Milstein argued that making providers aware of their responsibility for risk-adjusted health status can be part of a far-reaching reengineering of the delivery of health care.
The group also discussed current and potential collaborations among the several quality-oriented national organizations, and the usefulness of using standardized state regulations to promote health care quality. It was noted that current payment policy creates incentives to record inaccurate diagnoses, undermining data quality. Establishing standard qualifiers for diagnoses was mentioned as a feasible change. The panelists also stressed the importance of access to accurate and complete laboratory and pharmacy data.
Asked what strategic steps would improve data quality, Ms. Coltin mentioned standardization, movement toward EDI, and feeding back information to providers so they see value in the data and are induced to provide better data. Dr. Milstein stressed the importance of putting "teeth" in the Committee's recommendations and the Department's regulations.
The Committee heard from the following panelists:
Dr. Tang reviewed the recommendations of the 1991 IOM report on the CPR, which led to the establishment of the Computer-based Patient Records Institute (CPRI). He outlined the major changes in health care organizations, policy, and information systems since 1991 and the consequences of the fact that data are still not available to do what is needed. He described the components of the CPR and noted that major missing pieces are decision support, capture of progress reports, and ubiquity. After noting the major impediments, he encouraged NCVHS to help the Department focus on the uniform health identifier, clinical terminology, confidentiality legislation, and cost sharing. The first three were identified as priorities at a recent conference organized by CPRI.
Mr. Waegemann said the four major functions of the medical record are shareability, comprehensiveness, interactivity, and serving as the nucleus of the health system. For these to happen, the record must have uniform information, structure, and rules of handling. He stressed that contemporary systems that claim to be electronic do not replace paper; in fact, they often increase the volume of paper. True electronic patient records must have a signature with inherent accountability, data integrity, continuous availability of data, and auditability. He outlined a five-level process leading to the electronic health record, and stressed that many changes are needed before patient records are freed from dependence on paper.
Mr. Lorton focused on the importance of showing industry how it will benefit from investment in information systems. Today, progress is undermined by an under investment in information and standards. He stressed that only the federal government can spearhead this movement.
Col. Ray and Dr. Kolodner described a new high-level partnership among the VA, DOD, IHS and LSU Medical Center aimed at sharing data on their shared populations. It was spurred by a Presidential initiative on Gulf War syndrome, but will extend beyond this population. Alaska natives are another shared population. A standard lexicon is being developed, along with a common data model with a set of business rules, event triggers, and communications protocols. The CPR under development will be vendor-independent. Dr. Kolodner identified the major impediments to the CPR and encouraged NCVHS to focus on defining the computer-based patient record and exploring potential collaborations with the G- CPR partnership. In addition to continuing its work on standards, confidentiality, and the unique patient identifier, he suggested that the Committee develop policies regarding data ownership and encourage consumer involvement through mechanisms beyond the Public Register.
In the discussion period, Dr. Richard Ferran of LSU described the potential integration of the G-CPR with data derived from care in the community. The panelists were asked to identify their top priority for government intervention to promote the CPR. Most focused on development of a standard lexicon for medical terminology. Other priorities mentioned were data ownership, interoperability, and individual identifiers.
NCVHS members discussed a desire to hear in the future from major system vendors for both hospital and outpatient care. They were encouraged to look for the rare "integrators" among vendors.
Mr. Gellman reported on a two-day hearing held in January on the identifiable record and data registries. Both are key concepts that need clarification for privacy legislation to be effective. The hearing did not result in recommendations, but it did draw attention to these key issues.
The group discussed the recent sale of pharmacy records by two companies for commercial use. The Subcommittee plans to address this issue next.
Dr. Iezzoni reported that the Subcommittee will have its third meeting on its Medicaid managed care project in April, after which it will develop its report and recommendations. She described a March 2 hearing at HCFA on its data gathering on home care and long-term care and rehabilitation hospitals, particularly as affected by the Balanced Budget Amendments. The BBA has imposed impossible data requirements and deadlines on HCFA that will also impose burdens on providers and patients. The Subcommittee is thinking of taking this on as its next big project. It will also hold a meeting on the islands and territories in July.
Dr. Starfield reported on progress by the Work Group on Population-Based Data. Dr. Friedman is helping plan a joint meeting with NCHS and CNSTAT. Dr. Newacheck is working on a plan for looking at national health surveys. The current work on attachments is proving to be very difficult because of the complexity of definitions.
Dr. Detmer noted that members are asked to notify him in advance if they are unable to attend meetings.
Dr. Detmer introduced and welcomed three new NCVHS members:
Dr. Detmer then announced that David Garrison has been reassigned and will no longer co-chair the Data Council, and Harvey Schwartz has also been reassigned and will be replaced by Dr. Fitzmaurice as AHCPR liaison. The Committee approved his sending a letter of appreciation on its behalf to each of these individuals.
He then informed the group that Dr. Lumpkin would chair an ad hoc work group to pull together a framework for Committee action on quality and national information policy.
He also reported that the report of the Committee's 1997 activities related to HIPAA has been submitted to the Vice President, as President of the Senate, and to other congressional leaders. He thanked the staff for working on it with him.
Mr. Scanlon said Mr. Garrison has been appointed Director of the Office for Civil Rights. He remains a member of the Data Council, which will now be co-chaired by Dr. Margaret Hamburg, the Assistant Secretary for Planning and Evaluation. Dr. David Satcher, recently approved as Surgeon General, is in the process of moving from Atlanta.
Race and ethnicity data policy. On November 1 the Secretary issued a policy requiring that virtually all HHS data systems and data collection include race and ethnicity data. Heretofore, some significant ones did not, and no clear policy existed. Agency heads have now been instructed to implement the new OMB policy in their agency data planning and collection cycles. In addition, the agencies are encouraged to expand data collection needed to improve research on disparities in health status and services needs. Subpopulation data collection is also encouraged, as well as other variables relevant to assessing and improving the health of minority populations. There will also be an effort to do more innovative analysis of existing data.
Government Performance and Results Act (GPRA). This law requires all federal agencies to submit a five-year strategic plan to Congress. The first round was submitted last fall, and they are to be updated every three years. Agencies also were required to develop annual performance plans with objectives and quantifiable performance goals. Next year, the agencies will begin reporting on their performance. These plans and objectives affect Congressional budget decisions.
Other projects. The Data Council's continuing projects include health data standards, the unique health identifier for individuals, classification systems and code sets (including work on ICD-10), and health information privacy. In addition, the Council will work on a conceptual framework for the national health information infrastructure.
The Data Council will review the status of survey integration over the next few months. The next steps are provider and employer surveys, seeking a more integrated look at care. The Department also has a work group on the monitoring and surveillance tools needed in the event of anti-tobacco legislation. The possibility of more user-friendly grant policies for states regarding integrated information systems is also being studied, as well as ways to tabulate the new OMB race/ethnicity policies. A data strategy to support Healthy People 2010 is being developed, and the Council is looking at data needs to support welfare reform. Finally, the Council is looking at what information systems are needed to support quality improvement within the Department's health programs.
Dr. Starfield asked about plans to supplement the handful of health objectives that focus on disparities across socioeconomic groups. Mr. Scanlon described the revision process for the objectives, and said a section of the 2010 objectives would focus on socioeconomic factors.
Asked by Mr. Van Amburg about the NCHS proposal for age adjustments, he said the recommendation from NCHS to use the projected year 2000 population distribution for age- adjusted mortality has been favorably received and is still under review.
Regarding the five-year plans, he said the Department submitted a single plan to Congress for all of its component parts, bringing all HHS activities around common goals. Most of HHS agencies also have their own strategic plans, which are compatible.
Asked about provider surveys, Mr. Scanlon said no changes are planned at present, but the area has received careful study within the government and conversations have begun with outside groups. It is a complex subject, and the Data Council welcomes input from the Committee.
The Secretary's recommendations were sent to Congress on September 11. The Senate Labor and Human Resources Committee has held three hearings on the topic. Senators Bennett and Jeffords are working on draft legislation, and Senators Kennedy and Lahey have introduced a bill (#1368). Department staff is providing technical assistance to the Committee. No immediate action is expected.
The Data Council is also looking into the potential impact in the U.S. of a data protection directive by the European Union. It requires as of October 1998 that member states have legal controls that effectively prevent the export of identifiable data to countries lacking adequate privacy protections. The adequacy of U.S. protections and the potential impact of these restrictions are being studied.
The Department's interagency drafting teams have put together six documents now being cleared for publication on the four identifiers required by HIPAA, on transactions, and on security. Three NPRMs have been cleared in the Department, and it is hoped they will be out in about two weeks. The Notice of Intent for the individual identifier and two others are expected a few weeks after the first batch.
Dr. Braithwaite said that he has posted on the administrative simplification web page an explanation of the three-level clearance process and the several stages between first publication and final enactment as a rule (after which most of the industry has two years to implement the standards).
A notice of intent is being used for the individual identifier because no industry consensus yet exists on what should be done on that. It is hoped that an NPRM on the individual identifier can be published sometime this fall, with claim attachments on about the same timeline. He expressed appreciation for the Committee's recent hearing on those attachments.
Also on the administrative simplification web page are instructions for signing up for an electronic mail notification system.
Mr. Waldo noted that the BBA posed a huge workload on HCFA in terms of regulatory activity and data collection at a time when the agency is also working on the millennium problem and other things. The data requirements call for both new data and more data from existing entities, as well as new data from new entities.
A major component is a new information campaign for consumers, which has five components: printed literature, media campaigns, an 800 number, face-to-face information to beneficiaries, and an Internet site (Medicare.gov).
The crux of the effort is a knowledge data base now being developed, fed by several data sources which Mr. Waldo named. There are demos of various programs with the VA, social HMOs and others. An extensive data infrastructure is needed to pull information together logically and get out to beneficiaries in a timely manner. HCFA is thus rethinking "the IT vision" for the way it processes data. Mr. Waldo showed a chart of the new conceptualization. He then reviewed the extraordinary cost of building and operating the new information infrastructure, and compared it to current appropriations, which are not adequate.
The collection of encounter data for use in risk adjustment is the biggest data challenge over the next three years. Ms. Richter, who heads the workgroup that is deciding what non- inpatient hospital data to collect, described the activity in that area. She explained that the BBA clearly mandated what inpatient encounter data to collect, but left the non-inpatient data to the discretion of the Secretary. Personal risk adjustment is needed to eliminate the incentives to not pick up sick people. For that, data are needed for all Medicare beneficiaries.
For the first two years, only inpatient data will be used as a transition, while HCFA makes sure it can collect the data efficiently, store it, and know what it wants to do with it. When the data can be competently collected, stored, and used, HCFA will move to more comprehensive data collection, starting with physician and outpatient data. The initial information will be used for rough estimates of the impact of risk adjusted payments. Data for July 1, 1997 to June 30, 1998 are due September 17. HCFA will then evaluate whether this time frame is feasible. The agency is working with the industry on categorizing edits and deciding which are reasonable to use.
Dr. Starfield commented that the data implications for person-based capitated prospective payments are tremendous, and of a different order than hospital based DRGs. This is of great interest to the Committee, and she expressed hope that HCFA would keep them in mind during the development process.
Dr. Iezzoni determined that they will not be using functional status information on the encounter forms, and commented that the time lines for this project are "impossible." For example, the home health systems that are a source of some of the data are only now being set up. She offered to help HCFA make the point to Congress that some of the requirements are not feasible. Mr. Waldo said they are now figuring out what they can and cannot do, and on that basis will "start managing expectations on the Hill." At that point, they would welcome the Committee's support. He said they would stay in close touch.
Dr. Cohn observed that this effort is "an antithesis of administrative simplification," and that the issues around UB-92, managed care, and cost are complex. Regarding outpatient data, he urged the HCFA teams to give the industry as much lead time as possible and to be very clear about what data they need to do risk adjustment. Ms. Richter said they want to find out how much data, especially physician data, the plans can "get their hands on," in view of such arrangements as subcontracts.
Mr. Gellman asked for details about how the risk adjustment will be carried out, and was told a multiplier that depends on diagnostic codes will be calculated for every individual. He remarked that 1) there is lots of opportunity for gaming in this system, and 2) attention is needed to the implications of a government-issued number on people's health status. He noted the potential for discrimination in hiring on this basis, among other worries. Ms. Richter later said that the 12 diagnosis groups and the multipliers will be published, while the rates for specific individuals will be handled discretely. Asked about the volume of the encounter data, she said they estimate 1.2 million hospital discharges and 12 million outpatient visits annually for managed care for a base of 6 million beneficiaries.
Dr. Iezzoni noted the number of other critical data projects in addition to this one on the capitated Medicare product, and Ms. Richter assured her that the others are getting adequate attention. Dr. Starfield observed that European countries use sociodemographic characteristics to calculate risk adjustment, rather than diagnostic factors. Ms. Richter said in response to another question that the team for this project is working closely with HCFA staff working on EDI standards, so as to be consistent with impending HIPAA standards. Dr. Detmer predicted that the government will have to offer more incentives to move to computerization, and Ms. Richter said HCFA's contract with plans will require the electronic submission of data. Dr. Newacheck determined that no severity adjustments will be made for diagnostic codes for this project, and said this may prove to be a problem. He cited a similar coding scheme developed by the National Association of Children's Hospitals and Related Institutions.
Dr. Cohn remarked on the inadequate financing for the development of information systems for this effort, and he proposed that the Committee speak in support of more adequate funding. Mr. Waldo welcomed this form of support. In view of the Committee's interest in risk adjustment, he suggested that it meet with HCFA's Office of Strategic Planning. Dr. Detmer suggested that Drs. Starfield and Iezzoni think about which of their Subcommittees should take on this concern.
The President created the Commission on March 26, 1997. Next week, less than a year later, it will have its final meeting and present its final report to the President. The Commission has 34 members -- Co-chairs Secretary Shalala and Secretary Herman (Labor), plus 32 people from the private sector. They were organized into four subcommittees. The idea was to develop a consensus opinion on key issues.
The Commission had open meetings, has a web page, and heard testimony from more than 150 individuals and organizations. Ultimately, 33 members reached consensus on, and one objected to, the Consumer Bill of Rights and Responsibilities. The document was presented in November 1997 and quickly embraced by the President, who has taken several actions on it and proposed others, including the possibility of legislation. "The bottom line" of the report is the belief that every American should have the rights laid out in the program. Another assumption is that to be entitled to the rights, consumers must honor their responsibilities, and these are also outlined.
Mr. Sorian reviewed the rights and recommendations in the document, including the right to information, access to emergency care, choice of plans and health care providers, respect and non-discrimination, and confidentiality of health care records.
Dr. Starfield, among others, praised the Bill of Rights and Responsibilities and the work of the Commission. She commented on four areas she felt were not sufficiently highlighted, including systematic disparities in quality of care across population groups and systematic gaps in evidence. Mr. Sorian said the forthcoming final report does address these.
He then discussed the 14-chapter report. To improve quality of care in the U.S., it offers a series of recommendations for a systematic national approach to correcting these basic problems: avoidable errors, overuse of services, under-utilization of services (particularly prevention), and geographic variation in services. The Commission recommends the creation of two entities: The first, The Advisory Council for Health Care Quality, would identify quality objectives and track progress in meeting them. It would be made up of private and public sector experts, and be publicly financed. The second, The Forum for Health Care Quality, Measurement and Reporting, would also have public and voluntary private participation, and would focus on quality measurement.
The Commission suggests an initial set of national aims for improvement, which the Advisory Council would subsequently be responsible for. The first six cover reducing underlying causes of illness, injury and disability; expanding research; assuring appropriate use; reducing errors; increasing patient participation; and addressing over-supply and under- supply of resources.
The report calls attention to the key role of group purchasers, and urges them to be more active and coordinated in demanding quality services. It also recommends specific things that can empower consumers, and it identifies specific vulnerable populations and encourages HCFA to develop a risk adjustment system that could be adopted in the private sector. It also discusses the role of existing quality oversight organizations and ways that their work might be better coordinated. It calls for more research in a number of areas, and it urges continuous leadership from the federal government. Finally, it focuses attention on the need to redress the rising number of uninsured people in this country.
Dr. Detmer commented that the Committee, which serves a "boundary role" between private and public sectors, is interested in how it might contribute to this effort. He thanked Mr. Sorian, and said the Committee would issue a response to the Commission's documents.
After lunch, the Committee reviewed a draft commentary on the Consumer Bill of Rights and Responsibilities, based on a review by Dr. Iezzoni and Mr. Gellman. Dr. Starfield offered some revisions to this draft, and the Committee discussed the issues raised by these members. The first concerned evidence-based medicine. They agreed on the need to recognize that evidence-based medicine is a future goal toward which progress will be slow, and to define the rules of evidence broadly enough to encompass the appropriate metrics. The second issue concerned the modifiers "acute," "chronic," and (particularly) "post-acute" for types of care. Members agreed that the notion of longitudinal care is pivotal, and that the term "post- acute" is not universally understood. They favored articulating the goal of extending clinically specific measures from the acute and primary care area to other settings such as nursing homes and home care.
Drs. Iezzoni, Starfield and Detmer were asked to draft a letter to Secretaries Shalala and Herman stating the Committee's comments on the bill of rights and responsibilities. There was general agreement that the statement should include an expression of support for the overall principles and ideas in the document. The letter also will be posted on the NCVHS web site.
The group then turned to the Commission's other recommendations. Mr. Blair wondered why the Commission did not recommend strengthening NCQA and AHCPR as the organizational framework for quality improvement, and he expressed concern about undermining these organizations. Other members noted the benefits of not tying the recommendations to existing bodies and the existing paradigm of quality. They agreed to defer consideration of a written response to the Commission's other recommendations until they have seen the Commission's full report.
Dr. McCann works on coding and payment policy. He gave some history of the E&M codes, a term introduced in 1992 for physicians' visits and consultations. The codes are very important to the Medicare program, accounting in 1996 for 40 percent of part B payments to physicians. However, of the 7,000 codes in the CPT, only about 90 pertain to E&M.
There are different categories of codes, and three to five levels in each category. Payments increase as the levels go up. The codes were published by CPT in 1992, and early research showed that people were interpreting the words very differently. When HCFA announced that it would use the CPT, all physicians and specialists had to use the codes. The basic structure of the E&M revolves around the three key components of history, exam, and medical decision making.
In 1992, HCFA, the AMA and the specialty societies mounted an extensive educational campaign for the proper use of the E&M codes. It has been successful in eliminating the wide variations in the use of the codes. In the same year, they began to develop documentation guidelines with specific definitions to assist physicians in choosing the appropriate codes. A revised version of the guidelines was published in summer 1997, and that version has stimulated a lot of controversy.
Dr. McCann explained that payers need to validate the site of service, medical necessity, appropriateness of care, and the accuracy of the reporting. HCFA has tried to strike a balance between satisfying these needs and not imposing too heavy a documentation burden. He briefly described the different types of history, exam, and medical decision making specified in the codes, which HCFA has tried to delineate and define in the guidelines. He acknowledged that despite the effort to give physicians some flexibility, they mainly feel anxiety because of the heightened vigilance around fraud and abuse. Overall, the response has been that the guidelines are too complex. He asserted that the complexity is needed to allow HCFA to distinguish between levels of service.
Dr. McDonald interjected that the documentation requires so much "stupid and irrelevant" intellectual activity that it not only is at the patient's expense, but turns all visits into complex ones. Moreover, it does not recognize "half of the things that are important to specific patients."
Dr. McCann explained that the 1996 Chief Financial Officer audit of Medicare claims (an activity required by law) found that 30 percent of the 3,000 audited claims were incorrect, representing $23.2 billion paid in error. Physicians are only one of six contributors to coding, but about half the errors identified resulted from insufficient or nonexistent documentation by providers. HCFA developed a corrective action plan in response to these findings, among other things increasing the number of random prepayment audits of E&M claims. These activities have focused a lot of attention on the guidelines.
HCFA, he said, recognizes the need for more education, and in October it extended permission for physicians to use the older guidelines if they preferred until July 1998. (For some specialties, the new ones are preferable.)
Dr. Segal said the AMA's work with HCFA on this project stems primarily from the need of physicians to have access to review criteria used by Medicare and other payers, and also from the AMA's need to define with precision the terms used in CPT. He stressed that there was an effort to educate doctors about the guidelines; but clearly it was insufficient and needs to be improved. When the AMA recognized last fall that a problem was emerging, it requested and was quickly granted a six-month extension. At the December AMA House of Delegates Meeting, the degree of concern about the overall burden became clear.
He stressed that a guiding principle of the CPT editorial panel has been to allow specialty societies to determine the content of their single-system examinations. However, it is clear that physicians have serious concerns about being unfairly penalized for inadvertent coding or documentation errors or omissions and severe fraud and abuse penalties. The AMA never intended these to be viewed as rigid rules, and physicians have never been regarded as a significant source of fraud and abuse.
The next version of the guidelines must be a function and byproduct of clinical care. The question is how much documentation for coding is necessary. To deal with this problem, the AMA is working closely with all elements of organized medicine to implement a three- pronged action plan involving advocacy, corrections to the guidelines, and education. The advocacy will demand that physicians be treated as ethical professionals, and that claims reviewers be required and empowered to exercise judgment, recognizing that code documentation has elements that are necessarily subjective and that physicians must be given the benefit of the doubt.
Various medical organizations have been asked to submit suggestions for improving the guidelines by March 15. There will be a meeting in Chicago on April 27, to which Dr. Segal invited the Committee to send representatives. The editorial panel will refine the meetings in February, May and August. An education campaign and teaching materials are also being developed. In addition, the AMA is talking with HCFA about a further extension of the grace period.
Dr. Detmer opened the topic for discussion, noting the Committee's support for importance of documentation and the relevance of the computer-based patient record to this issue.
Dr. Starfield raised the prospect of perverse incentives for upcoding, and Dr. McCann commented on Medicare's effort to find the balance between "too much" and "too little" documentation. He noted that Medicare needs to consistently discriminate between levels of service, and the days when doctors only needed to submit a code to Medicare are over.
Dr. McDonald reiterated his criticisms and asserted that a doctor-patient encounter "is like a fractal" that cannot be approximated with simple rules. He favored a simple time-based formula as a solution.
Dr. Cohn said he wants details on his patients, and he finds the guidelines fairly compatible with his records. He agreed with Dr. McDonald that in a clinic setting, however, it is not reasonable to require so much documentation. For other patients, he recommended more education of physicians on use of the guidelines.
Dr. Iezzoni said her colleagues are undercoding for fear of review by FBI agents and huge fines because of inadvertent errors. Dr. McCann said the reviews are done primarily by nursing staff, with support from physicians. Dr. Iezzoni said physicians are not being properly educated about the real process, and the misinformation is spurring "the groundswell of upsetness." Other Committee members corroborated her account with ones from their experience.
The group discussed the fact that some for-profit trainers are creating an atmosphere of fear and misinformation about the guidelines and penalties. Dr. Segal noted, however, that the $10,000 penalty is real. Dr. Harding asked how doctors could be made to feel that they are doing the right thing for their patients by improving documentation.
In response to a question, Dr. McCann said the use of a software template checklist is considered by HCFA to be adequate documentation. Dr. Fitzmaurice observed that technology is emerging that will eventually solve some of the problems, but Dr. Segal noted the danger of a proliferation of templates that deviate from the actual record. The goal, he said, is a system based on "equal pay for equal work" for physicians. Dr. McDonald predicted the rise of a documentation industry disconnected from the real nature of medicine, and he predicted that the more elaborate formulas will actually cost HCFA more money.
Dr. Detmer thanked the panelists and said the Committee may accept their invitation to attend the Chicago meeting.
Ms. Coltin gave an overview of data quality issues. The key types of data needed for quality measurement and improvement include demographic and enrollment data, diagnosis codes, procedure codes, pharmacy data, laboratory data, service dates, and identifiers. The data quality problems fall into two categories, pertaining to administrative data and survey data.
Administrative data have quality problems related to completeness, timeliness, accuracy, and consistency. The completeness issues include having no data at all and having missing data fields. The absence of encounter data results most often from carve-out arrangements, particularly for mental health and pharmacy benefits. Many contracts have no provisions requiring lab data, making it impossible to find out about abnormal test results. (She later explained that when carve-outs are done by the purchaser, the health plan has no right of access to data on services provided elsewhere.)
Like other data problems, missing data fields are often related to payment incentives. One source of problems is the bundling of procedure codes. Capitated medical groups often report no secondary diagnosis. Another reason for missing data fields is the lack of EDI capability of many MCOs.
Feeding data back to providers with case-mix adjusted utilization information creates an incentive for them to provide more complete information, in order to justify utilization rates.
Timeliness problems often stem from payment incentives as well as in-and-out-of-area issues, patients' use of non-contracted providers, and plans' limited EDI capability.
Ms. Coltin noted a recurring theme as she remarked that accuracy problems, too, stem from payment arrangements. In both fee-for-service and managed care, case mix adjustment can lead to code creep. Attempts to mask sensitive diagnoses also undermine accuracy, as does recording a rule-out as a diagnosis.
Because consistency problems are generally due to lack of standards, HIPAA standards are expected to make a big difference. There are inconsistencies in data definitions, code sets, and provider IDs as well as in patient IDs.
In a managed care setting, quality problems in survey data mainly concern response rates and response bias. The recent over-surveying of plan members by physician groups, plans, employers, business coalitions, the government, and pharmaceutical companies has resulted in very low response rates. (NCQA found an average of 40 percent, ranging from 17 to 65 percent.) Another reason members don't respond to surveys is their confidentiality concerns. These rates create problems of response bias. An analysis of non-respondents to a satisfaction survey showed them to be predominantly male, minority, healthy, and more satisfied.
The problems created by administrative data incompleteness include an inability to assess aspects of care, an undermining of the credibility of the data, a dependence on chart review with attendant higher risks for confidentiality, and an inability to target needed interventions. Poor timeliness makes it impossible to implement timely reminders. Accuracy problems lead to such problems as misreporting of quality measurements, misdirection of improvement interventions, and the inability to benchmark best practices. Survey data problems make it impossible to generalize to the universe from which the sample is drawn and likely that problems will be overstated or understated.
Dr. Detmer noted that the need for standards for standards is one dimension of this topic.
HCFA is part of the "massive collaboration" represented by HEDIS. Application of HEDIS to the Medicaid population resulted in Medicaid HEDIS, and in 1995 Medicare measures were added. HCFA's goals are plan-to-plan comparison, external review, and internal quality improvement. Recently all the Medicare contract markets were asked to report on four quality measures, and these and other data were reported to NCQA last summer. An independent organization is validating the data. It did a baseline survey of all 278 contract markets, and is doing a more in-depth look at 60 contracts using a sample that over-represents the larger ones.
Dr. Elstein said he has just received the draft report, and has been briefed informally. The research found that plans' EDI capabilities vary widely. Many had never done HEDIS before. Many have difficulty getting good data. And many contracts are not set up to get good encounter data. In the next year, HCFA will do more intensive validation, with on-site visits and review of a sample of medical records, and it is proposing a pilot of source-code information prior to submission of the data.
HCFA is also requiring Medicare plans to use the CAPS survey. In May, HCFA will conduct the survey Health of Seniors. The latter, which is built on the SF36, will establish a baseline for patients and then two years later look at outcomes. An effort will be made not to have the same patient in both surveys. These activities are planned for managed care plans, but HCFA wants to do similar things in fee for service, and a contract has been let to do performance measurement in that domain. The greatest difficulty there is finding the unit of accountability, but the pilot will see how to adapt performance measures and the two surveys to the non-managed care setting.
This year's round of HEDIS will use Medicare measures similar to those in 1997, but major changes are planned for HEDIS measures "down the road." Dr. Elstein stressed that the recent quality study was done on a trial basis, and that the plans have learned a lot from it. A summary of the data will be made available to the public soon.
NCQA has done more than 100 HEDIS audits since 1995. Two years ago, it convened a panel to work on standards and guidelines for auditing HEDIS information, and it has put together a methodology that will be piloted in the summer of 1998. NCQA is moving toward incorporating performance measurement into the accreditation standards process. A 1997 study of some 300 health plans found a striking variation in plan performance. There are problems with data quality, data collection, and interpretation, and all of these will be studied. The HEDIS audit methodology is an integrated program that looks at data capture mechanisms and how the data are manipulated to conform to HEDIS specifications.
Some of the weaknesses that have been identified are that claims do not reflect the clinically important patient conditions, and that public health services available in a given region are not reflected in health plan information systems. Also, there is a sizeable data entry backlog, and in many cases only limited historical information is kept on file. Claims processors often have too much discretion and inadequate oversight in keying data. Coding may be distorted because of payment incentives. Many plans have difficulty with the transition from legacy systems to more user-friendly systems that would make clinical data available for measurement. There is inadequate oversight of vendor data.
Mr. Tierney said NCVHS can help by encouraging plans to improve their capture and use of available data and to create an environment that rewards data automation. Automating data and locating the responsibility for data quality with the entity that reports, manipulates and reviews the data will contribute to resolving current quality problems. Federal standards are expected to contribute greatly, as well. NCQA will be providing guidance on data management practices.
Dr. Milstein is clinical advisor to "the quality-activated end of the purchaser spectrum," and he said his comments would represent the purchaser's perspective and advocate for specific improvements. After characterizing the current health care scene fairly negatively in terms of quality and information about quality, he challenged NCVHS to support motivated purchasers by advising the Secretary to leverage the "excellent multilateral thinking that went into volume IV of HEDIS 3.0, A Roadmap for Information Systems." The Secretary could make a big difference by exercising her authority to make that vision a reality.
Dr. Milstein also called attention to these means of quality improvement, among others: unique patient identifiers; standardized mandatory reporting of lab values and pharmacy data; extending a model like the UNOS (United Network for Organ Sharing) transplant outcomes reporting framework to other hospital treatments; and granting JCAHO and NCQA "deemed status" and specifying mandated data elements for quality of care measures. He noted that many measures that were considered by the NCQA Committee on Performance Measurement (CPM) had to be dropped because the necessary data elements were not mandated.
Dr. Milstein pointed out that it is within the Secretary's domain to use bigger financial incentives to build a business case for the accuracy and completeness of the data elements needed for quality measurement. He likened having poor quality or incomplete data to "allowing our commercial airline industry to operate with semi-opaque windshields." Finally, he stressed the importance of the Committee's role in advising the Secretary, and urged the Committee to "be bold and breathe some life into the concept of quality accountability." He promised the support of purchasers.
Dr. Starfield suggested rating plans on the quality of their information systems, and Mr. Tierney said he would pass on the idea. He noted that information standards will be part of the accreditation process starting in the year 2000, and Ms. Coltin pointed out that plans score zero for performance in the areas where they lack information capability.
Ms. Fyffe asked about applying HEDIS measures to the 55 percent of the market that is not managed care. Dr. Milstein responded that the measures are just as valid, but there is no locus of accountability in that setting. Mr. Tierney added that thought has been given to moving the unit of measurement to the provider level, but generally the denominator size is too small.
Dr. Newacheck noted the value of consumer surveys and asked what NCQA is doing to improve response rates. Mr. Tierny said they are trying to improve it, and Dr. Elstein said HCFA is working on a survey of non-respondents. Ms. Coltin said the CPM has approved a single gold standard member satisfaction survey and a consumer assessment of health plan survey so everyone will not use different ones, and also developed systems for cross-checking survey populations. The CPM also endorsed a plan to include telephone follow-up in the survey protocol. Dr. Milstein commented that HIPAA regulations could help by building a more robust administrative data set to measure quality. The group returned to this idea later.
Dr. Iezzoni expressed concern about the Health of Seniors survey in that mailed or over-the-telephone questions about people's functional status may make people uncomfortable and also lead them to give non-accurate responses. Dr. Elstein agreed, adding that the fact that plans will not know who is being surveyed may be insufficient assurance for the individuals. He invited the Committee's input on addressing this concern.
Dr. Iezzoni also expressed doubt that changes in an SF36 score produce anything that is actionable by health plans. Dr. Milstein said that although he preferred having more robust administrative data, he has seen the SF36 data used to help physicians realize that they are being held accountable for risk-adjusted change in functional status over time. This has led them to work on modifying health behavior and environmental factors affecting health status. Asked if payers will be willing to pay for this broader approach, he said that remains to be seen but at least delivery systems are starting to think about other inputs than just physician time. The idea is "to use this crisis in raised expectations and reduced money to force the long-overdue reengineering of what it means to deliver health care." This will involve telephone care, other professionals and paraprofessionals, and helping patients be effective self-managers.
Asked about NCQA's collaborative work with JCAHO and the Foundation for Accountability, Dr. Elstein said that HCFA supports and participates in the FAcct and also with the Joint Commission. These organizations are regarded as complementary in getting quality measures. Ms. Coltin said HCFA has been instrumental in stimulating joint efforts.
Ms. Ward asked whether state regulation might contribute positively, and the panelists supported this idea, provided the states had a standardized approach.
Dr. Iezzoni commented that her research has found huge disparities between the diagnoses on laboratory bills and those on the doctor's visit bills for the same person. This stems from payer policy only to pay for services for some diagnoses. She asked whether the quality measurement community has tried to work with big payers to remedy this. Ms. Coltin said that with little influence on payer decisions, they simply try to "work around" the data problems. Bob Mayes said HCFA's Office of Clinical Standards and Quality is well aware of the limitations of using administrative data for quality assessment. He later commented on the fundamental differences between "bounded" administrative data systems and clinical records. It will be difficult to reach consensus on standardizing clinical data, and he urged new efforts to define national health data.
Mr. Van Amburg asked Ms. Coltin to suggest two or three strategic steps to improve data quality, and to estimate the cost. Ms. Coltin mentioned standardization and movement toward EDI and feeding back information to providers so they see value in the data and start to provide better data. She referred to HCFA for an estimate of cost. Asked by Dr. Detmer what she encompassed in standardization, Ms. Coltin responded standard code sets, identifiers, data definitions, and rules for using codes and recording data elements. Dr. Elstein predicted that full implementation of the Balanced Budget Act consumer information on HEDIS will give plans an incentive to report better data.
Ms. Greenberg commented that physicians are getting a mixed message, with outpatient coding guidelines requiring them to code only to the known level of specificity and payment policy requiring a diagnosis as a precondition for payment. She agreed with Dr. Milstein that having more administrative data could have an impact.
In regard to the question about the biggest opportunity for incremental gain, Dr. Milstein said his experts tell him that accurate HCFA 1500 and UB-92 data and accurate and complete lab value data and pharmaceutical data would permit substantial increases in the ability to measure quality.
Ms. Coltin commented that a standardized computerized patient record would also make a huge difference. The data from the parts of her organization that use computerized patient records are far more accurate, complete and timely than the data from parts that do not.
Dr. Detmer commended the group that worked on volume IV of HEDIS, and he suggested the Committee plan to study it. He expressed frustration that "as the data mounts, quality remains an orphan," but also noted that quality research is an emerging science with some unrealized potential in creating a value-driven health care system. He then asked what unique patient identifier the panelists would recommend. Ms. Coltin and Mr. Tierney agreed that it should be a new identifier, "started from scratch," and assigned at birth. On the Social Security Number, Ms. Coltin said she didn't like it, and Dr. Elstein said major studies by the Social Security Administration have found big problems with it.
Dr. Starfield suggested as a relatively easy change having standard qualifiers for diagnoses, and Dr. Milstein said he would support that recommendation.
Dr. Cohn reiterated the importance of laboratory and pharmacy data as an adjunct to the claims record. He noted that the Committee is working on attachments as one way to make those data available. Ms. Coltin agreed, noting that with lab data the issue is access, while with pharmacy data the issue is quality because they then can easily become associated with the wrong family member.
Dr. Detmer invited Dr. Sondik, Director of the National Center for Health Statistics, to comment. Dr. Sondik endorsed the idea of "a scorecard on data," and said he hoped to see less process-oriented and more outcomes-oriented measures. He asked if there were progress in that direction, and Dr. Elstein said good outcomes measures are indeed being developed, although they won't be available for another couple of years. Asked for closing comments, Dr. Elstein said that "HCFA needs to hear a lot of the comments that were made around the table." He also noted the need for awareness of the tradeoff between confidentiality and quality.
Mr. Tierney said both process and outcome measures are important. The industry has been inventive in measuring quality with the tools given it, and it will do a better job of doing so if it is given better tools.
Dr. Milstein observed that outcomes can be measured directly or through processes that have been well correlated with outcomes. In closing, he noted that many past recommendations on what data should be provided have not yielded much effort by the health industry, and he urged the Committee and the Department to put some "teeth" into data completeness and accuracy.
Ms. Coltin noted the need to engage a wider variety of partners in the care process, including non-traditional care givers. She also observed that some information can only come from patients, and access to it needs to be ensured by protecting them from excessive intrusion and educating them about how the information will improve their health.
Dr. Detmer thanked the panelists for their leadership.
The Committee then reviewed a new draft of a letter to the Co-Chairs of the President's Commission, commenting on the Consumer Bill of Rights and Responsibilities. After making a few minor revisions, they approved the letter with one abstention.
Mr. Gellman reported on a two-day workshop held in January. The first day focused on the nature of the identifiable record, a pivotal concept in proposed legislation and a difficult one to define. He called attention to research by Latonia Sweeny of MIT, demonstrating the ability to link to personal records starting with a few non-unique identifiers. Such capabilities make it necessary for agencies and IRBs to constantly re-examine their information policies to guard against inadvertent disclosures. Also, legislation needs more than a simple word test to establish protections. The Subcommittee's goal was to find a solution to this problem, but this did not emerge from the hearings. They did serve the purpose of raising the visibility of the issue.
Mr. Blair remarked that concerns on the Hill about disclosure of genetic information have contributed to the fragmentation of legislation around confidentiality. Mr. Gellman agreed, noting the proliferation of narrow bills and confusion among Congress members about what to support.
The second day of the January hearing focused on data registries, which range from ones with public health value to ones that exist for commercial purposes. There is no comprehensive regulatory structure for registries, and proposed legislation has not addressed the issue of distinguishing between the good ones and the others. What is needed is either a statutory definition or a process for identifying the good registries.
Finally, Mr. Gellman circulated copies of recent stories about the sale of customer information to a third company by two pharmacy systems. He said the Subcommittee plans to investigate the use of pharmacy information and medical information for marketing purposes. Asked about the relevance to this concern of the bills already drafted, he said the Condit bill has an implicit but not explicit prohibition of such practices. Senators Bennett and Jeffords' bill probably would have authorized the practice.
Dr. Tang framed his presentation in terms of the 1991 report by an IOM Committee chaired by Dr. Detmer, calling for the adoption of computer based patient record (CPR) systems. He stressed that the goal is not to "electronify the paper record." The report's seven recommendations for developing and promoting the CPR included establishing a Computer- Based Patient Record Institute (CPRI), with several specific mandates to CPRI.
Since then, the world has become connected through Internet and other enabling technologies. Also, the policy and business landscapes of health care have changed, with mandates to cut costs, capitate, generate quality reports, and so on. However, the data for doing those things are not available. He outlined some of the consequences of not having the data and observed that the fears and concerns over billing compliance audits aired at yesterday's meeting are consequences of not having data. In addition, NCQA's road map for information systems requires information systems that do not yet exist. Indeed, existing information systems often get in the way. If the data were available, it would be possible to manage costs, manage disease, manage demand, and measure and improve the health care system.
Turning to the CPR, he described a system that begins with the clinical data repository and includes physician order entry, decision support, progress notes, and ubiquity (for authorized use), all of which permit and are used for care of people on a continuum. Major missing pieces in this process are decision support, capture of progress reports, and ubiquity. There is still much to be done in this developmental process.
Dr. Tang said the primary users of the CPR are physicians, nurses, and patients, and he showed a sample CPR system used in his clinic. Physicians love the system, and it also provides patients with a useful summary of their visits, including vital signs and a graphic summary of conditions such as blood pressure or sugars. He can use the same system, with confidentiality protections, to make "electronic house calls." He invited Committee members to conduct a site visit and see the system in action. The system is also used for analyzing aggregate data.
He then returned to the IOM report, to assess progress on its recommendations and goals. Although the adoption of the CPR is still a goal and data costs are not yet shared by data users, a public/private CPRI was established in 1992.
In 1996, CPRI organized a national summit on health information solutions, just months after HIPAA was enacted. More than 80 representatives of health care stakeholders reached consensus on three high-level priorities for furthering the development and use of CPRs: the universal health identifier, confidentiality and security regulations and laws, and terminology conferences. He noted that now may be the time for a standards acceleration project that CPRI proposed three years ago. CPRI has published seven well-received documents on confidentiality and security. It also has been active in getting the word out about CPR systems, another IOM recommendation. It held a terminology conference in November and plans another in 1998.
He highlighted "remaining impediments" or pressing needs in these areas:
Top concerns are the universal individual health identifier, clinical terminology, and cost- sharing models.
NCVHS can advance progress toward CPRs by continuing to conduct hearings and recommend solutions for standards, and by recommending public funding for accelerated development of terminology. It also should stay on track with confidentiality policies and help the Department find a role with respect to cost sharing.
Mr. Waegemann observed that historically patient records were kept for one caregiver. Today, in contrast, as many as 122 people need access to the medical record to document the care process and decision making, including management data and financial as well as clinical data. All of these components are part of the future electronic patient record. The four major functions of the medical record are share-ability, comprehensiveness, interactivity, and serving as the nucleus of the health system. For these to happen, the record must have uniform information, structure, and rules of medical record handling.
All of the so-called "electronic patient records" today need parallel paper back-up records as the legal archive and source of documentation, and thus do not qualify as such. True electronic patient records must have a signature with inherent accountability, data integrity, continuous availability of data, and auditability. In addition, uniform access and handling rules are needed. Such an electronic medical record would stand up in court because it is compliant with the principles of documentation. The MRI has been concerned that people are saying they are implementing CPRs when they actually have parallel paper systems. In fact, the paper record is getting bigger. Moreover, hospitals are still decentralized and only "empires" or departments are being computerized.
Mr. Waegemann then described a process with steps or levels on the way to the electronic medical record. It starts with automating sections of the record. This generally begins with document imaging, which some 22 percent of hospitals are doing. The next interim step is the computerized medical record, which requires a single data base that integrates legacy systems. Next is the electronic medical record or patient record. This is followed by electronic patient record systems and, finally, the electronic health record. The latter may reside with the patient and contains a much broader range of information, with input by the patient and traditional and non-traditional providers.
Currently, society is moving toward the electronic patient record (level 3). He stressed that it would not be possible under present conditions to get rid of all the paper records. Many changes are needed first in behavior, attitudes, technologies, systems, standards and legislation. He noted that the standards process is so complex precisely because the EMR is the nucleus of the entire health care information system. He likened the current stage to the beginning of the use of electricity in 1876. It will take time and leadership to get to "precise detailed plug and play standards."
He encouraged the Committee to work toward legislation that would permit providers to create electronic medical records as long as they comply with a set of requirements. Another goal is closer cooperation between ANSI HISP and NCVHS. Once the clinical inventory of clinical standards is finished, under Mr. Blair's leadership, the analysis stage should be undertaken, looking for gaps, inconsistencies, and duplications.
The HOST consortium was created in 1994 as "a neutral workplace for the industry." Mr. Lorton noted the impact of "two revolutions," in health care organization structure and in information. A basic problem is that information technology developments always outrun the ability of standards organizations to keep up with them. A major reason for this is that they are under-resourced because of organizations' failure to recognize the benefits of investing in information systems.
He gave some examples of the kind of creativity and progress that can be made with the proper motivation and cooperation, including an initiative around the idea of a master patient index and development of a CPR system for gastroenterologists. The challenge in such cases is to show industry how it will benefit, not just how the patient will benefit.
Finally, Mr. Lorton said the federal government is the only organization with the power to spearhead this movement, and the challenge is how to get it to serve as the leader in changing health care.
Col. Ray said the Government CPR project (G-CPR) is a collaborative data-sharing partnership among the Indian Health Service (IHS), the Department of Veterans Affairs (VA), the Department of Defense (DOD) and LSU Medical Center. The group plans to use existing standards and to work with NCVHS and others to create standards where none exist. Each partner has its own set of applications, and they are working on a standard lexicon. This requires a common data model with a set of business rules, event triggers, and communications protocols, all with adequate security and confidentiality protections.
The partners have a set of shared populations. A major focus is veterans of the Gulf War who may suffer Gulf War Syndrome. The G-CPR project was stimulated by the report of a Presidential Advisory Committee on the Gulf War Syndrome. The need to track this syndrome and prepare for possible future exposures has given a sense of urgency to this effort. Another shared patient population is Alaska natives.
Dr. Kolodner said the partnership is only 8-10 weeks old. He stressed that it will continue to work with the SDOs on standards, although its timetable is accelerated. A key issue for the CPR is that it be vendor-independent. The CPR system will permit the exchange of information among the partners. Major impediments to the CPR include differing definitions of the CPR; the complexity, length and cost of implementation; the lack of a common data model and a comprehensive clinical lexicon; and the need to balance confidentiality with secure and appropriate access. He also cited technology barriers and data ownership issues as impediments.
Dr. Kolodner's recommendations were framed in terms of addressing these and other impediments. He encouraged NCVHS to focus, among other things, on defining the computer-based patient record and exploring potential collaborations with the G-CPR partnership. He also recommended advancing clinical lexicon standards, the unique patient identifier, and strategies for balancing security and confidentiality. Finally, he suggested developing policies regarding data ownership and encouraging consumer involvement to determine expectations and concerns through mechanisms beyond the Public Register -- e.g., surveys, forums, and web-site discussion rooms. Noting that the G-CPR project represents an opportunity to serve as a test bed, he suggested that the Committee consider addressing content standards and coding standards "in chunks" rather than all at once.
Dr. Detmer thanked the panel for launching the Committee on its work on computer- based patient records, "the next order of business of the HIPAA mandate."
Mr. Scanlon asked about the status of hospital administration systems in respect to what is envisioned for the CPR. Dr. Tang said the current systems are mostly business oriented and need to be replaced with more clinically-oriented systems. The same is true for ambulatory care. Col. Ray observed that DOD is different in that it has not billed in the past and its systems are clinically based. Dr. Kolodner noted that the IHS is the most advanced in terms of moving its information systems into ambulatory care. Mr. Lorton commented that a "cultural problem" that must be addressed is that physicians are being asked to do something they did not "originally sign on" for.
Dr. Cohn was one of many Committee members to praise the panelists for their clear presentations. He commented on the significance and timeliness of the CPRI terminology conference, and then asked how the G-CPR project will be integrated with community data. Dr. Kolodner noted that more than 5 percent of VA's budget is spent on health care provided outside VAs, and it has gotten little information about that care. One reason for collaborating with LSU Medical Center is to figure out how to exchange information with a local institution.
Dr. Richard Ferrans of LSU described the personal health information carrier smart card that soldiers will carry. The goal is to capture data on reservists seen in civilian health systems and electronically transfer the information to DOD data bases. A survey capability is also needed, as well as the ability to determine the validity of data. He noted the importance of open standards and vendor-independent solutions for encouraging innovation.
Dr. Detmer predicted the evolution of three CPRs -- the personal record, the CPR as it relates to health delivery systems, and the computer-based population. This constellation was not anticipated, he said, when the discussions began at IOM in 1989.
Mr. Blair asked each panelist to identify a priority impediment that should be addressed with government assistance. Col. Ray and Dr. Kolodner both highlighted the clinical vocabulary or lexicon. Dr. Kolodner added that clarification of data ownership would lay the foundation for progress in other areas, including privacy and confidentiality. Mr. Lorton agreed about the lexicon, pointing out that it must be developed as something that can change rapidly in response to a changing clinical environment.
Mr. Waegemann said his priorities are documentation rules, followed by interoperability. He asserted that a lexicon would not solve all the problems. He also pointed out that several "niches of medical records" are already quite successful, such as home health care and some physicians' offices and nursing systems. The goal is to move from these homegrown systems to large-scale commercial systems.
Dr. Tang supported terminology as the top priority, and noted its relevance to NCVHS and HHS. Increasing attention to and funding for this priority would benefit the government itself. His other priority is the individual identifier.
At Dr. Detmer's request, Ms. Humphreys briefly described a new study by the Computer Science and Telecommunications Board funded by the National Library of Medicine, focusing on the health requirements for the next generation Internet and whether they conflict with other routine uses of the Internet.
Mr. Gellman noted that the G-CPR operates under the Privacy Act, and he asked Dr. Kolodner and colleagues to pay attention to their experience in this regard so they could one day report back to the Committee about it. Dr. Detmer noted that eventually a mandate such as this will be extended to the rest of the country.
Ms. Fyffe suggested hearing in the future from major hospital information system vendors to get a sense of the experience of people who are designing and implementing systems. Dr. Starfield added major outpatient vendors to the list. Mr. Lorton said HOST was considering asking major system vendors to talk on a panel about the transition at its July meeting. Mr. Mayes noted that three or more of the current panelists have built very large systems. Mr. Blair noted that what is needed is diverse perspective, including the post-acute care area. Col. Ray stressed the need to talk to "true integrators" that are able to bring all the parts together; this has not typically been a strength of vendors.
Ms. Greenberg took the opportunity to thank the Department of Veterans Affairs for supporting the Internet broadcast of NCVHS meetings, which many are finding useful.
The following topics were listed for the June agenda:
Dr. Iezzoni reported that the Subcommittee has held two hearings on Medicaid managed care, one in Washington, D.C. and one in Phoenix, AZ. It will have another site visit in Boston, MA in April.
A meeting on the islands and territories will be held in mid-July. The Office of Minority Health will pay for people to come in from the Pacific Islands. She thanked Mr. Scanlon and his staff for their help on this meeting, which is expected to make a difference in conveying the needs of people in these areas and ways to improve their information systems.
This Subcommittee and the data Subcommittee co-sponsored a March 2 meeting at HCFA on its data collections and systems for home health, rehab hospitals, and long-term care hospitals. She briefly described the immense requirements and impossibly short deadlines of the Balanced Budget Act, which requires "enormous new data gathering procedures." Dr. Detmer encouraged Mr. Mayes to put a summary of these activities on HCFA's web site. Dr. Iezzoni said her Subcommittee is thinking about data issues in this area as its next big project. She later added that she hoped Dr. Starfield's Subcommittee would remain involved in this issue of common concern.
The Congressional requirements will impose burdens not only on providers but also on patients. The Committee may be able to help HCFA think through how to respond to these requirements.
Mr. Mayes said that many HCFA staff members who are not familiar with the Committee's work found the March 2 hearing helpful, and benefitted from "hearing the outside world."
Dr. Starfield reported that the Work Group on Population-based Issues is now active, and Dr. Friedman is making good progress on plans for a joint meeting of NCVHS, NCHS and CNSTAT. Dr. Friedman added that a planning group for what has been called "the vision quest" is reaching agreement on goals, and he expects to have a work plan in a few weeks.
Dr. Starfield reported that Dr. Newacheck has helped develop a plan related to national health surveys. The Subcommittee's work on claims attachments "is proving to be a nightmare" because of the difficulty of defining attachments. She predicted that the recommendations will be framed as "a pathway to standards," and she referred to HCFA's proof of concept project.
Dr. Detmer reported that the Executive Subcommittee urges people to come to meetings and agreed to ask them to let Dr. Detmer know in advance if they are unable to attend. Future minutes will list absentees. The group discussed the difficulty staff has in getting calendar information from members. Dr. Detmer suggested a 48-hour turnaround for responding.
Dr. Harding observed that having meetings in excess of the expected 20 days per year is onerous for members. Dr. Detmer acknowledged that this deserves attention, and Ms. Greenberg noted the merits of piggy-backing more meetings to reduce travel time.
Dr. Detmer expressed appreciation to the three new members for their willingness to participate and the contributions they have already made. After thanking the staff, he adjourned the meeting.
I hereby certify that, to the best of my knowledge, the foregoing
summary of minutes is accurate and complete.
/s/ Don E. Detmer June 16, 1998
_________________________________________
Chair Date