The Subcommittee on Health Data Needs, Standards and Security of the National Committee on Vital and Health Statistics was convened on Monday and Tuesday, February 10 and 11, in the Hubert H. Humphrey Building in Washington, D.C. The meeting was open to the public. Those present included:
John R. Lumpkin, M.D., M.P.H., K2 Standards Work Group Chair
Simon P. Cohn, M.D., M.P.H., F.A.C.P.
Kathryn L. Coltin, M.P.H.
Lisa Iezzoni, M.D., M.S.
Don Detmer, M.D.
Robert Gellman, J.D.
Barbara Starfield. M.D., Subcommittee Chairperson
George Van Amburg
Marjorie Greenberg, Acting Executive Secretary, NCHS
James Scanlon, Executive Staff Director, DHHS
William Braithwaite, M.D., Ph.D., DHHS
Bob Moore, Health Care Financing Administration liaison
Lynnette Araki, NCHS
Gail Janes, PhD, CDC
Jackie Adler, NCHS
Donna Pickett, NCHS
Gary Beatty, Mayo Foundation
Jeffrey Snow, Liberty Mutual
Louis Stanislao, Options Health Care Inc.
Pat Hamby, HBO & Co.
Charles Meyer, HBO & Co.
Tim Sparapani, Hall & Dorr
Dan Staniec, NCPDP
Denny Blaszcynski, AFEHCT
Doug Fielding, MegaWest Systems, Inc.
Kepa Zubeldia, Envoy-NEIC
Alfred Buck, PhD, JCAHO
Patrick Whitaker, CDC
Daniel Orenstein, Powers, Pyles
Bob Beckley, PDX
Michael Kremer, NCHS
Sara Hambert, Eclipsys
Linda McNish, PhD, NCHS
Lyndalee Korn, TRW Systems Integration Group
Susan Abernathy, CDC
Donald Stockford, Dept of Veterans Affairs
James Tierney, NCQA
Roy Bussewitz, NACDS
Robert Iadicicco, OPM
Jean Narcisi, ADA
Jeff Blair, IBM
Gary Friend, IMS America
Kathleen Strouse, Aspen Systems
Chris Bergsten, AAHP
Harvey Schwartz, PhD, AHCPR
Doug Long, IMS America
Ellen McClecky, BNA
Paul Wilson, IMS-America
Tina Tran, Clinical Lab. Mgt. Assn.
John Kelly, MD, HBO & Co.
Stan Edinger, Ph.D., AHCPR
Susan Seare, Medicode, Inc.
Paul Placek, PhD, NCHS
Larry Watkins, Medic Computer Systems, Inc.
Jane Harman, PhD, NCHS
David Carlson, Enterprise Systems, Inc.
Henry Heffernan, EDPNS
Gary Berg-Cross, Ph.D., SAIC
Suzon Landrek, MBS
Gary Knaus, Context Software Systems, Inc.
Kimberly Dyer, HCFA
Brian Riewerts, HFMA
Jeffrey Cooper, Coopersoft
Virginia Huth, OMB
Jan Lovorn, Cylink
Rich Ankney, Certico
Douglas Mann, Battelle
Rick Peters, ASTM
Bob Beckley, PDX
Ed Hammond, PhD, Duke University
Steve Bass, Washington Publishing Co.
Gary Friend, IMS America
Mark McDougall, Health Level Seven
Phong Ngo, SAIC
Teresa Cengrowska, ASTM
Joseph Pressley, HCFA
(Note: Documents pertaining to this meeting can be found on the NCVHS web page at HTTP://ASPE.OS.DHHS.GOV/NCVHS/)
The NCVHS Subcommittee on Data Needs, Standards and Security met on February 10-11 to hear presentations on the Department's survey integration activities, and to hear and discuss the testimony of three panels on the administrative simplification provisions of the Health Insurance Portability and Accountability Act of 1996 (HIPAA, Kennedy-Kassebaum, or K2).
Dr. Alfred Buck, Joint Commission on Accreditation of Healthcare Organizations; James Tierney, National Committee on Quality Assurance; Randy Desonia, National Governors Association, Center for Best Practices; and Bob Merold, IMS.
All the speakers in this panel emphasized the importance of clinical data and of using this stage of administrative simplification to lay the groundwork for improving it. Their testimony stimulated a discussion that continued throughout the meeting on whether, or how, to make the initial round of administrative simplification useful for clinical purposes. Varied views were expressed about the nature and relevance of administrative data, but there was agreement about the problems with a narrow billing perspective. One theme that continued throughout the meeting was how to make administrative data useful in a capitated environment.
Subcommittee concerns about proliferation of performance measures and the danger that they will confound efforts to simplify health care and reduce administrative costs generated a recommendation for a national consensus process to agree on indicators and measures. Mr. Desonia's participation called attention to state activities, and led to a Subcommittee decision to solicit input from model states and to consider a hearing on state standardization activities.
In the course of the two-day meeting, there was considerable discussion of the ANSI X12 standards, beginning with this first panel. The testimony combined positive comments about the standards and cautionary tales about the difficulty people are having implementing them. Many panelists recommended a combination of implementation guides and training to remedy this problem.
Concerns about coding and classification were another major theme of the meeting. The first panel stressed the issue of accuracy in the use of code sets.
There was considerable discussion of the Committee's core data elements recommendations. The first panel stressed its importance in providing a context and reference point for the entire standardization process.
These and other panelists were urged to submit follow-up comments in writing.
Benjamin Curtis, The Health Information Network Connection; Denny Blaszcynski, IBM Global Network, Value-Added Network; Kepa Zubeldia, Envoy NEIC, clearinghouse spokesperson; Jeanne Scott, CIS Technologies
This panel generated the idea of "standards police" to see that standards are used, and agreed with previous panelists' suggestions about implementation guides and training. It was suggested that EHNAC could serve as a compliance verification body.
The clearinghouse and VAN members of AFEHCT from this panel returned the following day to present these recommendations: 1) that the Secretary adopt the NCPDP claim for pharmacy use and the ANSI X12 standards for all other transactions, and 2) that the Secretary endorse and recommend that payors use commercial off-the-shelf EDI translator products and discourage the use of one-time custom translation programs.
David Carlson, Enterprise Systems (ESI); Pat Hamby and Chuck Meyer, HBO & Co.; Bob Beckley, PDX, Inc.; Gary Knaus, Context Software Systems, Inc.; Doug Fielding, MegaWest Systems; Larry Watkins, Medic Computer Systems
The Subcommittee discussed the ANSI X12 837 transaction standard further with this panel, as well as code set issues. One theme continued that the Subcommittee heard repeatedly: an appeal by pharmacy sector representatives for a continued mandate to use NCPDP for pharmacy use and the ASC X12 standards for all other transactions.
These panelists expressed a concern also voiced by others, that standardization could involve a premature move into the clinical area that impedes development of the computer- based patient record.
The Subcommittee discussed the focus of the April hearings on coding and classifications and who should testify. They agreed to a focus on medical classification, acknowledging the need to address broader issues in subsequent work. Another subject of discussion was how to most effectively participate in or liaison with the HHS implementation teams, in preparation for the Subcommittee's June recommendations to the full Committee. They heard status reports from Ms. Pickett and Mr. Mayes, co-chairs of the Coding/Classification and Infrastructure implementation teams, respectively. Mr. Mayes said the latter group is thinking in terms of a phased approach in which standards are introduced gradually and the field knows well in advance what to expect.
The group reviewed its liaisons to the HHS teams, with these results:
Infrastructure, cross-cutting issues - Dr. Cohn
Systems security - Ms. Frawley
Coding and classification - Dr. Iezzoni, Dr. Cohn
Identifiers - Dr. Lumpkin
Claims - Ms. Coltin
Eligibility - Mr. Van Amburg
Jeff Blair, Cochairman of the Technical Coordinating Committee of ANSI HISB, presented an inventory of healthcare informatics standards. This inventory was prepared to assist the Department in selecting standards in accordance with the HIPAA mandate.
Jean Narcisi, American Dental Association; Ed Hammond, Duke University; Chair, HL7; Dan Staniec, NCPDP; Lee Barrett, ASC X12
Following their presentations, Dr. Braithwaite asked the panelists to comment on how to put teeth into the standards. There were a variety of suggestions. Attention focused on payors' need to detect fraud as an impediment to implementation of standards, and it was suggested that a consensual process be facilitated to identify the information needed for that purpose. The idea of testing and accreditation was supported, with each SDO representative saying its organization could perform these functions for its standards.
Following some discussion, the Subcommittee heard from two additional panelists: Phong Ngo, of X3L8 Secretariat Information Technology, and Rick Peters, of ASTM and ANSI HISB. Mr. Ngo described the 11179 and in particular its approach to variability. Dr. Peters' area of responsibility with both of his groups is security and confidentiality. His chief message was that most issues in this area are well handled by technologies that are available outside the health care industry in other parts of the computer industry.
Returning to the general discussion, the group discussed Mr. Scanlon's question about where the boundaries are between standards and implementation guides, and also how standardized the data items and code sets are within the standard. Dr. Lumpkin observed that the testimony in this meeting suggests a high degree of comfort in the industry with adopting all of the existing standards, and he asked if this were indeed the case. These questions generated a discussion of how to fit the various standards together, and whether they do fit. The panelists generally asserted that the pieces do fit together and that it is clear which to use in which instance. They stressed that the industry, through the SDOs, is working out common problems by means of such efforts as building a common data dictionary and creating a common business model.
Mr. Van Amburg asked members for comments on recommendations distributed in November on community health assessment, the result of hearings by the Subcommittee on State and Community Health Statistics. The recommendations may be submitted to the full Committee in March. In this regard, the Subcommittee took note of the continuing interest in state efforts around administrative simplification and the close links between this issue and public health and community assessment.
The Subcommittee passed a motion approving a revised draft of its charge, adding consumer groups to item #2.
The group then reviewed with Mr. Scanlon the areas in which the Department could use the Subcommittee's advice. Although there was interest in defining an overarching vision or set of questions to guide other decisions, members agreed that given their time constraints and numerous commitments, they would focus on the area in which the Department has asked for advice, namely, data collection from providers of care.
The Subcommittee focused on two issues in regard to administrative simplification: what use will be made of the Committee's core data recommendations, and how to approach the transaction standards as the first stage of a process that will create an environment hospitable to the development of the computer-based patient record. Members were asked for a sense of what the Subcommittee might report at the March NCVHS meeting in regard to the K2 project. They expressed comfort with the National Provider Identifier, now that it has been separated from the location index.
The meeting was called to order by Dr. John Lumpkin, Chair of the Subcommittee's Work Group on Data Standards. Following introductions, he noted the tight time frame under which the Committee is operating, the need to stay focussed on the goal of administrative simplification and its intended clinical benefits, and the Committee's continuing concern with the issue of variability.
Each of the panels for this meeting was asked to comment on the following on behalf of their organizations: 1) their expectations for administrative simplification under HIPAA, 2) their degree of satisfaction with the Department's process, 3) current problems with the transactions identified for standardization, and 4) how the goal of administrative simplification can best be achieved.
Dr. Lumpkin welcomed the first panelist.
Dr. Buck said he would focus on the first and third questions. Regarding expectations, JCAHO hopes for improved processing and transmission of healthcare information and facilitation of the integration of data from clinical, administrative and financial domains. He offered four concerns or recommendations:
· National accrediting entities, when appropriate, should be referenced as standard- setting organizations. (JCAHO has been a pioneer in developing and promulgating consensus-based national standards.)
· Clinically oriented data elements need more comprehensive treatment in legislation and oversight activities, to facilitate integration with financial and administrative data elements.
· External comparative metrics that provide high-level organization comparisons should be stringently defined. In this regard, he suggested formation of a task force of national accrediting bodies to address the issue of report cards.
· National accrediting bodies should be designated in new or clarified legislation as health care quality oversight organizations. Failing that (if concepts and terms are carried over from P.L. 104-191), they should be included in the definition of data clearinghouse.
After giving some background on NCQA, Mr. Tierney stressed the fact that few if any health plans in the U.S. are satisfied with the extent and quality of the clinical information in their databases. They draw such data primarily from claims databases "because they are available, not because they are ideal." Data linkages within plans and among contracted providers are nonexistent or difficult to implement, and the lack thereof increases the need for manual data collection for performance measurement, adding needlessly to the complexity.
What is needed is not just information systems but an information framework. To this end, NCQA has commissioned an Information Systems Work Group for HEDIS 3.0 and charged it with developing a road map for the managed care industry as to what upgrades and information systems would provide the greatest increase in performance measurement capacity.
NCQA hopes that the administrative simplification requirements will move the managed care industry toward standardization in the structure, content, definition, coding and transfer of clinical data.
Mr. Desonia set his comments in the context of the devolution of responsibility for programmatic decision-making to states and their resultant need for more timely and focused information on health care activities. His Center provides technical assistance to governors, their staff and key state agency policy makers. He said his remarks did not represent official NGA policy.
He encouraged the Subcommittee to keep three trends in mind as it develops its recommendations. The first is the devolution of programmatic decision making to states, which are becoming increasingly interested in state-specific information. He urged the Subcommittee to learn from state EDI efforts, such as that of New York.
The second trend is toward performance outcome measurement systems, in which states such as Oregon and Minnesota are developing outcomes-based systems for Medicaid and other programs. This trend raises questions about how to transform categorical data sets into an information system that can measure a program's effect on a client.
The third trend is state agency restructuring, generally along more functional lines and with a consolidation of information systems into one major agency. This trend highlights the fragmentation of data systems and the fact that as constituted, they provide no useful information for decision making.
Mr. Desonia urged the Subcommittee in developing its recommendations to work with governors and their staffs, who offer a unique cross-agency policy and information investment perspective. NGA would welcome the opportunity to facilitate such an interaction. He added that he has never seen the kind of gubernatorial demand for better information systems that he is seeing now, and this is a wonderful opportunity for state and federal government to work together.
IMS is the largest health information company in the world, creating information databases for use in promoting health care efficiency and effectiveness. Operating in 83 countries, it works primarily on tracking pharmaceutical products and also is taking an active role in patient-centered information. Mr. Merold endorsed the comments of the JCAHO and NCQA representatives, with whom IMS is in ongoing dialogue. He noted IMS's extensive experience in protecting privacy, including its use of encryption technology, maintenance of firewalls, and operation under EU guidelines.
IMS uses data for oversight and management. It supports standardization because poor quality and the absence of standards contribute to time consumption and waste, negatively affecting both costs and the quality of patient care. He cited the difficulties of tracking patient compliance as an example, noting the difference this information could make in outcomes and costs. Variability is "the single biggest headache we have in pushing the information to its optimal end."
He noted that the National Committee can play a role in "setting the bar" for the standardization process and accelerating progress. In doing so, he recommended that it make use of existing and emergent standards. He joined the previous speakers in calling for greater emphasis on clinical information alongside administrative information.
Responding to these presentations, Dr. Iezzoni hailed the emphasis on clinical data but noted the Committee's charge to begin with the simplification of administrative data. She asked for comments on whether despite this emphasis, the addition of a few clinical elements might be justified in the short run. Dr. Lumpkin noted Kennedy-Kassebaum's reference to attachments as one context for broadening the purview to clinical data.
Dr. Buck characterized the separation of billing and clinical data as one that is "neither real... nor constructive," and he suggested that claims forms are an important link to clinical data. He outlined what he sees as three levels of application: the "data element technical piece;" an intermediary level addressed "partly, but significantly" by accrediting agencies; and the national framework. The latter, he asserted, is not addressed in Kennedy-Kassebaum. Mr. Merold agreed with these comments, and added that claims information today is inadequate for "95 percent" of the broader applications of data.
Dr. Cohn commented on the proliferation of performance measures and the danger that they will confound efforts to simplify health care and administrative costs. He asked the panelists for their thoughts on simplifying or standardizing performance measures. Dr. Buck responded that the de facto standards, the UB92 and the HCFA 1500, are inadequate for the kind of assessment system the country needs. What is needed, he said, is a consensus process to settle on a set of performance measures. Mr. Tierney added that the work of NCVHS is "the beginning steps to where we want to go in performance measurement," a process that is burdensome now because of the absence of the right data elements. He agreed with Dr. Buck that financial data are not what are needed for this purpose.
Mr. Desonia urged the Committee to approach this as an evolving issue, rather than trying to solve it all at once. He noted that CDC is trying to focus the Healthy People 2000 process and Oregon is reassessing its indicators. Citing California, Mr. Merold added that "a few 800 pound gorillas aren't waiting for national standards."
Dr. Lumpkin questioned Dr. Buck and Mr. Tierney about how their organizations use data, and then asked them whether emerging standards such as X12N promote or mitigate against their organizations' work. Dr. Buck said the standards are compatible and would not cause a problem, provided they do not reduce clinical content. Mr. Tierney said the standards would enhance the accreditation and performance measurement process. Mr. Desonia said the standards would not necessarily cause problems for states; he noted the difficulties states have when their hospitals are on the border and they have to share information with other states.
Dr. Lumpkin then asked for comments on coding and coding sets, such as the need to use both the CPT-4 and the ICD-9. Mr. Merold said latitude and variability result in reduced value in the information collected "at the other end," and he called this one of the most critical areas for his organization. A narrower code set on provider classification of diseases and procedures would be preferable. He agreed with Mr. Tierney that the issue is less a matter of what data sets exist than of their correct and accurate use. Asked to comment on the prospect of payers' telling providers to use only sections of codes, Mr. Merold said that kind of restriction would be a problem.
Dr. Buck agreed that accuracy is a serious problem; he noted that the industry currently regards as acceptable a 10 percent error rate in coding, and he added that the level of refinement is inadequate in both ICD-9 and CPT. Finally, he said he was particularly concerned that the financial people would fail to get involved in the integration process and instead turn to financial figures, mistakenly regarding them as objective or hard. This, he said, would be a tragedy.
Dr. Cohn asked the panelists to advise the Committee on how to make sure that what is done in the short run for administrative simplification and standardization promotes, rather than impedes, progress in the future. Dr. Buck first highlighted the Committee's leadership in providing consensus forums, and then stressed the need to support the larger NCVHS agenda and particularly its work on core data elements. He called the latter "one of the biggest foundation blocks" for the entire process, and something that is critical to the interests of everyone present.
Mr. Merold agreed, and noted the importance of anticipating future capabilities. For example, much of the billing administrative infrastructure will soon be obsolete; progress will take place in data warehousing, decision support, and clinical records, and it is key to "surround the billing system with a number of additional elements."
Asked to estimate how far standardization of the administrative component moves things toward general data standardization, the panelists suggested a range of one-third to one- half. Mr. Tierney added that the process is beginning where it should, with the low-hanging fruit. Mr. Desonia cautioned that there may be disappointment among those who mistakenly expect this effort to achieve 100 percent of what is needed. He urged an effort to sensitize policy makers to the fact that this will be a long and incremental process.
Dr. Buck said he thought Dr. Cohn's question was critical; he noted that we are on the threshold of technological breakthroughs that will go a long way to solving issues of data- based linkage and the construction of meaningful measures. He noted that the Committee's work on confidentiality is another important area of activity.
Returning to the subject of clinical data, Dr. Iezzoni noted some confusion about whether the focus is filling in encounter data or lumping clinical data onto attachments. She and Dr. Lumpkin agreed on the need to be clear about this, with Dr. Lumpkin noting that one question is simply how much the Subcommittee can realistically take on.
Ms. Coltin noted two common threads among the panelists' testimonies: the inadequacy of standard billing transactions for performance measurements, and the need to look to the future. She observed that in the future, capitated payments will probably change the purpose of the transaction to such things as oversight and performance monitoring, and in this light she asked for comments on what is missing from the transaction data set. Dr. Lumpkin added that written follow-up comments on this subject will be welcome, as well.
Dr. Buck observed that for internal management purposes, many organizations already link clinical activity to their best assessment (however imperfect) of resource use, and this trend will continue. Regarding missing information, he first cited the universal patient ID, which would permit analysis of such things as continuity of care and episodes of care, and following patients between plans. He also expressed concern about the widespread misunderstanding of what billing data have to say about costs, not only by politicians and the public but even by some HMO managers.
Mr. Desonia noted the need for better information on the chronically ill population, whose care is particularly costly and will be increasingly so. Mr. Merold commented that in a capitated environment, the quality of administrative records is very poor because there is no incentive to maintain it. Only clinical information gives understanding of key questions. He added that the greatest opportunity for cost reduction and simplification lies in the use of clinical information to identify which procedures produce the best outcomes for patients.
Mr. Tierney predicted that once patient records are automated, some forms of standardization may permit direct access to databases rather than the transfer of data, bypassing the need for a central repository and thus for the encounter form used today.
Mr. Scanlon asked for comments on state activities that are noteworthy as good models or things to be avoided, and Mr. Desonia offered to help identify specific states. Dr. Lumpkin then thanked the panel and, following a break, introduced the next one.
Mr. Curtis and the three presenters who followed him are all officers of the Association for Electronic Health Care Transactions (AFEHCT). He spoke on behalf of both THINC and AFEHCT. THINC is an electronic health information network that facilitates the electronic exchange of business and patient-care data through a private electronic intranet system. AFEHCT is a trade association of vendors and suppliers to the health care EDI industry.
Mr. Blaszcynski said he would speak on behalf of the members of the value-added network (VAN) community, which connects trading partners doing electronic commerce in industries that include the health care industry. VAN members wholeheartedly support the adoption and mandatory use of the standards identified in HIPAA. Although the industry has moved toward electronic media, disparate players with different agendas and reporting requirements have introduced cost and complexity into the communication system. The health care industry lags behind others in EDI because of a lack of investment in communications infrastructure and the complexity imposed by differing proprietary requirements. HIPAA is "the jump start the industry needs." Even with standardization, the value-added network industry will still have work because of a continuing need for technical and administrative support for handling health care data.
Mr. Blaszcynski noted the potential dangers of an "overzealous desire" to protect data masked in the language of privacy and confidentiality issues. When his industry handles information, it does not know what type of data exist in its messages, and it maintains a high degree of security. He pointed out that it is impossible to prevent someone from maliciously obtaining and misusing data; but health care encounter data are always vulnerable to that, from the time they are collected. Reasonable security for movement and handling should be expected, plus strong penalties and disincentives for unauthorized access and misuse.
To address issues and help define the electronic commerce solutions the Secretary will recommend to Congress, AFEHCT members suggest a series of workshops with third party service vendors, who can be expected to continue to play an intermediary role in the health care system. Similar meetings have been held with HCFA, and they were very useful. AFEHCT offers to help organize and participate in these activities.
Dr. Zubeldia said his comments would offer the perspective of clearinghouses, which should not be confused with VANS and billing services. He described the complexity of 15 separate business relationships among the six players in health care networks, calling it collectively "an explosion of connectivity." Each relationship involves multiple data standards. Clearinghouses promote administrative simplification by providing a means for all the players to connect. They have been laying the infrastructure for administrative simplification for 15 years, acting as a trading partner.
One of the issues affecting clearinghouses is privacy and confidentiality. The encryption methods used for banking and other purposes may be inadequate for health care information, which has to be protected for at least 50 years. The Electronic Health Care Network Accreditation Commission (EHNAC) reviews how clearinghouses maintain security. Clearinghouses often serve as the firewall between providers, payers, and the rest of the industry.
Another issue is regulation, with each state regulating clearinghouses in different ways. State use of EHNAC could establish the uniform regulation of clearinghouses. EHNAC has organized a clearinghouse network commission, and clearinghouses have been involved in the EDI and standards movements. EHNAC accreditation is seen as a way to solve problems. Clearinghouses also want to participate in the standardization and administrative simplification process.
Clearinghouses see HIPAA as an investment in the future that will pay off. Their concerns center on inadequate contractual relationships, digital signatures and encryption, the fact that EDI standards are not mandatory, the need for standards for data as well as transaction sets, and the multiple state regulatory requirements.
Dr. Zubeldia concluded with these suggestions for the Committee: that the Committee increase clearinghouse involvement; that standard EDI trading partner agreements specifically represent the role of clearinghouses; standards for data content requirements and for compliance verification; and accreditation of clearinghouses through the EHNAC process.
Ms. Scott said she would represent pharmacy perspectives. She noted that the work on administrative simplification legislation began in the industry at least six years ago. The pharmacy area has a concern that the NCPDP, which is working smoothly and seamlessly for 99 percent of pharmacy transactions, will be overturned in favor of another standard. Her organization asks the Subcommittee to recognize and endorse the NCPDP communications standard. Apart from this, her industry supports the legislation. Other lesser concerns include the cost of the impending changes and the need for better standardization of physician and hospital outpatient areas.
Finally, Ms. Scott observed that administrative simplification can best be achieved by focusing attention on the areas where standardization has been most difficult to accomplish, while accepting what has already been done.
Discussion of these presentations began with a query from Mr. Moore about whether clearinghouses foresee benefits from the development of a single standard for transaction sets. Dr. Zubeldia responded that a single standard would indeed help. The group agreed that the law intends for users to have only one option once a standard is selected, and that this is desirable.
Dr. Lumpkin raised the question of how the groups represented are affected by, and view, capitation. Mr. Curtis said that even with capitation, basic information will still be collected on encounters. Dr. Zubeldia said that the X12 standard would help encounter reporting because encounter information can easily be added to an X12 claim. Additional clinical information -- e.g., prescriptions given, lab work ordered, follow-up plan -- would be very appropriate.
Ms. Scott commented on a prevailing misconception that capitation results in a lack of interest in data, while the reality is that reporting will continue on services and resource utilization. Most value-added companies and clearinghouses see such information services as important. She noted that some of the data such as Dr. Zubeldia mentioned are already being picked up via other mechanisms.
Dr. Lumpkin asked the panelists for specific recommendations as to what standards to adopt and what to avoid; to a question from Dr. Zubeldia, he said he meant for the standard to be "an only child, not an adopted child " -- i.e., as the only standard. After an initial reiteration of pharmacy's desire to continue using the NCPDP, the question generated a wide- ranging discussion of a variety of problems with the 837.
Mr. Curtis advised deciding what standard(s) to recommend based on what is already in use by the dominant players in the industry. This, he noted, would mean that HCFA's choice will become the de facto standard, adding that his organization's recent meetings with HCFA have afforded a chance to apprise HCFA of the potential impact of its decisions on the industry.
Dr. Zubeldia said his organization has foregone proprietary formats and gives new submitters the choice of the NSF or the 837. Their priority is to have claims come to them as 837, rather than the five percent that do now, so they no longer have to translate them. He believes the X12 is more useful than the flat file standards, but he also wonders how providers will be convinced to send X12 transactions rather than flat files. Mr. Moore agreed that the issue is providers' capability to submit the 837. Elaborating on these comments, Dr. Zubeldia and other panelists tied the issue to software, noting that someone will have to bear the cost of conversion to a single standard.
Dr. Braithwaite and others observed that "something is wrong here" when HCFA is paying contractors to accept the 837 while, as Dr. Zubeldia indicates, they are telling others in the industry they cannot do so. To Dr. Braithwaite's statement that the new law directs payers to accept the standard transaction if someone sends it to them, Ms. Scott pointed out that HCFA is already telling them to accept it, and they are not doing so. Dr. Zubeldia reiterated his concern about "this adopted child we are creating"; he asked what "standards police" exist to make sure the standard is being used.
Mr. Moore recalled a previous panelist's statement that it took a year to implement the 837, and he expressed frustration that the standard identified as the best is not making possible the communication everyone says they want. As solutions to these problems, Dr. Zubeldia and other panelists suggested an implementation guide, training, and a standard translation put out by HCFA. Mr. Blaszcynski asserted that a standard will go only part of the way toward administrative simplification if it does not come with a standard implementation guide; the two must go hand in hand. He also said, in answer to Dr. Lumpkin's earlier question, that payers and providers want standards for eligibility, enrollment, remittance advice, and claims submission. He stressed the strong support for standards in his industry and its confidence that its services will continue to be needed.
Dr. Zubeldia returned to his comment about a standards police, urging consideration of a body responsible for reference implementation and compliance verification, to promote the use of the standards. He agreed with Dr. Braithwaite that EHNAC could perform that role. Mr. Curtis added that the legislation provides the clear incentive for it to perform that role and for others to recognize its authority.
Dr. Lumpkin thanked the panelists, and urged them to submit further comments in writing.
Ms. Hamby said that the underlying themes of her comments are uniformity of data, interpretation, and implementation.
Health identifiers for providers, health plans and individuals are desperately needed, and their implementation will greatly reduce expenses. As for transaction standards, Mr. Fielding cautioned that even HCFA has not succeeded in getting its contractors to standardize their practice. If HIPAA fails, the future of EDI in the medical industry will be limited. Nevertheless, he is "ecstatic" about the passage of HIPAA and hopes it will be successful. His company has no problems with the transactions specified under HIPAA and is less concerned about which standard is chosen than that something is implemented to standardize transactions.
As a cautionary tale, Mr. Fielding recounted his company's discouraging experience with trying repeatedly to get payers to implement the 837, ending with a decision not to attempt any more 837s until the HIPAA standards are enforced. His company "wholeheartedly supports the 837," but believes it will be difficult to get payers to understand and effectively implement standards for claims transactions. He stressed that vendors are not the problem when it comes to implementing the 837, and he endorsed Dr. Zubeldia's suggestion of a "standards police" to promote compliance and provide a grievance process. Finally, he called attention to the important telecommunications aspect of sending messages, which HIPAA has not addressed.
Dr. Lumpkin began the discussion with these panelists by asking for further comments on the 837. The panelists stressed the need for implementation guides that apply nationally. Asked if there are instances of successful implementation of the 837, they said there are, and cited examples. Mr. Watkins said it takes about three months to bring up a new vendor with the X12 850 transaction set, which is somewhat comparable. Dr. Zubeldia pointed out that it takes two to three months to implement the NSF or the UB92.
Dr. Iezzoni asked the panelists to comment on their experiences with CPT and other code sets. Ms. Hamby said the AMA has changed its fee structure, and use of CPT has become very expensive. Mr. Knaus added that the costs are passed on to the end user, but the annual paperwork required is onerous as well. In response to a follow-up question, the panelists said there are no comparable problems with other code sets. Asked about the HCPCS, Mr. Carlson said the key is having a universal product identifier. Whatever system is used, there should be a single source and it should be a "living source" because of the rate of change in health care.
Dr. Cohn referred to disease management and other future uses of data, and asked how the panelists view the various standards options in this light. Mr. Beckley said the key there, as well, is code sets, which must be the same. He noted that pharmacy is starting to develop codes to communicate additional information to payers, such as about duplicate therapies; these need to be standardized, too. Mr. Carlson observed that software was originally built on the model of paper-based standards, leading to a multiple additions and modifications. He contrasted this approach with object-oriented technology, which is a blueprint for looking at information and modeling business problems and that permits change and extension. Mr. Knaus noted that while the transaction sets will help improve aspects of information and its clinical use, they will not improve diagnosis, supply, or procedure codes. He stressed that front-end coding is key to making those accurate and useful.
Dr. Lumpkin asked for comments on the links between billing and clinical information in these companies' operations. Mr. Carlson noted that HL7 was launched years ago to focus on clinical information needs within hospitals; Mr. Meyer stressed the role of clinical repositories in making possible the "virtual longitudinal patient record" and disease management activities using both clinical and financial data. Mr. Beckley said pharmacy has had the financial piece in place for five years and spends most of its time focusing on clinical information.
Mr. Scanlon asked about the panelists' experience with security standards, and Dr. Blair noted that Rick Peter has put together guidelines for data security from the CPRI, ASTM standards, and work within HL7. Mr. Meyer cautioned against mandating a specific scenario now, because the field of cryptography is constantly expanding. Mr. Carlson noted that the technical task of preventing the reading of data has been solved; what is still up in the air is "who should get it and what parts of it should they see and who keeps track of who you gave it to."
Dr. Cohn returned to the 837, noting that the general message he has heard is that it is immature. Noting that the implementation deadline is 1999, he asked the panelists to indicate what would need to happen for them to be comfortable with it as a single standard. He later urged the panelists to submit written comments on this and other topics.
Mr. Watkins stressed that the problems with the 837 are not in the standard itself but in the lack of education as to its use. Moreover, there are fewer implementation problems than there used to be; furthermore, the 837 is easier to bring up than the NSF, which has "a lot of ambiguity in it." Mr. Beckley said the 837 does not meet the pharmacy industry's needs because it is batch oriented; if forced to use it, pharmacy would lose all the clinical benefits of the NCPDP. Ms. Hamby said the important thing is to use a single and consistent standard, regardless of the choice, and to have adequate education for users.
Mr. Watkins pointed out that the NSF and its predecessors are a processing format, and what is needed is a communication standard that permits back-and-forth translation without affecting the processing. The 837 can better handle health care claims because of its compression and the intent behind its development. Mr. Moore said HCFA would move to the Medicare Transaction System when the standards are implemented in the year 2000, and it will no longer use processing formats after that.
He expressed HCFA's concern that some providers would be unable to support the 837 and would go to paper if that were mandated. Mr. Knaus agreed that some providers would have difficulty and would have to get new technology. Mr. Carlson suggested defining what is needed and putting it out for a bid. Mr. Beckley commented that pharmacy's experience suggests that the transition period is a long one. Mr. Meyer added that organizations are starting to plan for 1998, implying that it would be helpful to know where the standardization process is headed well in advance.
Dr. Lumpkin thanked the panelists for meeting with the Subcommittee, and he asked them to submit written comments on what standards are mature enough to be put in place.
Dr. Starfield then turned the group's attention to the Subcommittee's broader charge concerning health data needs, data collection efforts, data utility, and security. She welcomed Ed Hunter and Ross Arnett for a report on the HHS survey integration initiative.
The HHS survey integration effort began with analysis of the various surveys conducted by the Department and a decision to focus on the major, recurring surveys. The goals of integration are to streamline and rationalize the surveys, fill critical gaps, improve analytic capabilities, and reduce overlap and duplication. The first step taken was integration of the household surveys, by drawing the National Medical Expenditure Survey (NMES), renamed the Medical Expenditure Panel Survey (MEPS), from the National Health Interview Survey (NHIS) sample. This increased the analytic capacity of the data and saved $6 to $8 million in sampling costs.
The next phase in survey integration will somehow focus on the three nursing home surveys. A big issue under consideration is how to rethink and rationalize collection of information from and about providers of care. NCHS also plans to use the National Immunization Survey as a launching point and sampling frame for another telephone survey asking NHIS-type questions. This project is called SLAITS (State and Local Integrated Telephone Survey). SLAITs can be targeted to geographic areas and tailored by states, and adjustments can be made for households that do not have telephones.
The benefits of survey integration to date include coordination of efforts, filling of some gaps, the continuous collection of health insurance, health care use, and expenditure information, and new efficiencies in sampling and data collection. The Department now shares questions and aims for consistency in questions among surveys, where appropriate.
HHS also is aware of the potential risks and downsides of integration, however, such as combining surveys into "a nice appetizing target" for budget cuts. Mr. Arnett noted that linking the surveys links their fates as well. It is also possible to over-consolidate and develop a single survey that is too big. Another downside is that the number of vehicles for supplements is now reduced, increasing the competition for supplements.
The survey integration effort extends beyond the Department and HHS is now looking at other employer surveys conducted by the federal government. An interdepartmental working group has been proposed to look at integration possibilities in this area.
Mr. Hunter called attention to an editorial on survey integration in the magazine of the American Statistical Association. He noted that the integration process has coincided with development of the HHS Data Council, and it operates in that broader context. Integration represents a change in the way the Department thinks about data collection; unlike in the past, the information policy process routinely asks whether a survey is the right mechanism for obtaining particular information. There is also exploration of whether existing surveys and questionnaire sources can be used to obtain the desired information. He noted that interagency collaboration has proved very difficult, but the financial and analytical payoffs are worth the effort. Unanticipated technical and conceptual problems have surfaced as well, and he enumerated some of them. In addition, even with the consensus on the need for more up-front collaboration, questions remain about how to balance agency prerogatives and operating imperatives with the need to generate information with applicability and usefulness beyond that agency.
Mr. Hunter stressed that the Department invites the Committee's help in keeping the surveys in the context of other data sources, and also in looking for ways to integrate, link, and fill gaps. It also would like help in setting priorities, and welcomes pressure to continue the cross-agency collaborative planning process. He noted that the survey integration effort would not have occurred without external pressure from OMB and others. He asked the Committee to "keep reminding the Department that there are things we haven't quite finished yet." An immediate activity on which the Department welcomes advice is forthcoming integration efforts in the health care provider area.
The Data Council, ASPE and NCHS have launched a "conceptual framework" project to ask basic questions such as what information is needed about providers, what questions HHS should be asking, and what the overall policy concerns are. HHS will be looking at these questions over the next year in workshops and other settings. NCVHS can have a role here as well.
Ms. Coltin asked about the possibility of standardizing survey data elements in a way that would permit linkage to transaction data. Mr. Arnett said this has been considered but is further down the road. He noted that an electronic patient record would obviate the need for the expenditure survey, and there is potential for collaboration with HCFA's work in the provider area. Mr. Scanlon added that in the past, population based surveys have sometimes been linked with Medicare claims history. In the future, the National Provider ID and file could be a sampling frame for studies.
Dr. Iezzoni noted that realizing the potential for linkages and simply analyzing the data being collected both require resources, and they seem to be in short supply. Mr. Hunter agreed, and said the Department is already unable to take full advantage of what it has collected; this is one reason why it puts out public use tapes. Any new resources are for very targeted initiatives. While there is no interest in collecting information that can be gotten through other means, sometimes surveys are the only way to get information. Mr. Arnett added that as data are combined, they become more identifiable and more precautions are needed to protect confidentiality. The Department is considering having data centers where researchers can work with the information, but this too takes resources.
Dr. Lumpkin hailed the Department's survey integration efforts and predicted that the country will have a better focus on the importance of data in the future; in the meantime, it is important to continue to collect the data, to maintain a longitudinal capability. Noting that some states are investing in improved data, he urged the Department to make full use of state data and capacities and to look for ways to link and connect these sources. Mr. Arnett agreed, but cautioned that this can be very complicated because every state has its own requirements and demands. Dr. Lumpkin suggested exploring with national associations of state officials (e.g., ASTHO) the possibility of model standards that could be offered to states to achieve greater uniformity.
Dr. Cohn asked that the Committee receive a briefing soon on the Department's thinking about the objectives and outcomes of each of the surveys, so it can help think about alternate data sources, and Mr. Hunter said this would be done.
Asked by Dr. Starfield about linking AHCPR's Consumer Satisfaction survey to other surveys, Mr. Arnett said that because the MEPS satisfaction questions go into less depth, the questions are instead coordinated so they can be cross-walked. To a follow-up question about the utility of the provider surveys and other sources of information on providers, Mr. Hunter said the aforementioned conceptual framework project will look at this question. Dr. Starfield observed that it is increasingly hard to interpret the data from NMES because providers work in so many different kinds of organizations. Mr. Arnett added that ASPE and AHCPR are working with the AMA on a physician survey.
In conclusion, Mr. Arnett stressed the importance of the Committee's role in being an "honest broker" and helping the Department address and balance turf issues as it goes forward with survey integration. Dr. Starfield said this is the Committee's intent.
The Subcommittee urged the Department to make full use of state data and capacities and to look for ways to link and connect these and other sources. Dr. Lumpkin suggested exploring possible model standards with national associations of state officials (e.g., ASTHO). The Subcommittee asked for a briefing on the Department's thinking about the objectives and outcomes of each of the surveys, so it can help think about alternate data sources.
A Subcommittee discussion of K2 work group issues and other business began at the end of the first day of the meeting and resumed the next morning.
The first topic was coding and classification issues and how to effectively establish liaison with the HHS implementation teams. Dr. Lumpkin observed that three issues are involved: what to do about competing sets, how much granularity to recommend, and whether to recommend requiring uniformity. Dr. Iezzoni raised the issue of definitions, which may vary even within the Committee. She recommended confining the April hearing to clinical classification coding sets. Dr. Cohn noted that the code sets having to do with transaction standards are only part of the picture, and he urged keeping the focus broader than medical classification in order to better promote comparability. Dr. Starfield suggested matrixing these issues to standard-setting efforts. Ms. Greenberg said the implementation teams would be analyzing similarities and differences among the coding sets. Mr. Moore added that information from the SDOs will serve as a starting point.
Returning to the hearings, Dr. Iezzoni agreed with Dr. Cohn about the larger picture, but recommended confining the hearings to the narrower subject. Regarding the types of people who should be invited to testify, she suggested those who produce the classification systems, those who work outside the hospital environment, and users who can speak to the strengths and weaknesses of various code sets. Others agreed with her categories. Specific names will be supplied to staff, who will issue the invitations and announcements.
It was agreed that discussion is also needed to establish a conceptual framework on coding and classifications, and that this would be scheduled for the full Committee meeting in March.
Dr. Cohn offered several other items as ongoing issues for the Subcommittee's attention: claims attachments, state examples of standards implementation, and personal identifiers. The last is to be reviewed at the March NCVHS meeting. Dr. Lumpkin suggested this overall plan: identifiers and the coding/classification conceptual framework will be addressed at the March full meeting; the hearing on coding/classification will be held in April; and the Subcommittee will look at state examples and attachments in June or July. Mr. Moore mentioned the tight deadlines for recommendations to Congress, and it was noted that the adoption of standards for claims attachments can be deferred. It was agreed that information on state practices would be solicited by mail, and then a decision would be made about whether to have a hearing on this topic. Names were suggested for the March conceptual discussion.
The following issues were identified as topics to be addressed by a panel on coding and classification:
· Clarification of the terms "classification system" and "coding system," which are often used interchangeably.
· The purpose of each system.
· Ways to reconcile apparently irreconcilable systems, as well as where the most serious problems exist.
· The panel of users might address gaps, while those who produce the codes could address overlaps and responsiveness of the various systems to change.
· Question for developers about their code sets' applicability to administrative transactions.
· How users deal with the multiplicity of codes.
Returning to the question of the Subcommittee's relationship to the HHS implementation teams, the group discussed what role is realistic for the Subcommittee. Dr. Lumpkin suggested having the liaisons follow each team's progress and products, watching for anything that should be shared with the rest of the work group. The goal is for the Committee to be ready to make decisions in June. Ms. Coltin asked for information on the teams' work plans.
Ms. Pickett reviewed the work and planning of the Coding and Classification team, which is still gearing up for its "massive task." The team has long-term plans relating to the computer based patient record (CPR), and short-term plans on administrative transactions. It welcomes the Subcommittee's suggestions. The group discussed some of the details of the work plan with Ms. Pickett and Dr. Braithwaite, and then the meeting recessed for the evening.
The morning discussion continued the focus on gearing up for June recommendations and relating efforts to the Department's process. The group agreed on the need to clarify the short-term goals and their relation to the longer-term objectives related to the computer-based patient record.
One axis of this discussion was the relationship between clinical and billing data and whether any clinical facets are feasible in the near term. The group agreed with Ms. Coltin that the generic term "administrative data" is more accurate than "billing data" for the short- term focus, partly because of the extent of capitation today. Several people stressed the overlaps between administrative and clinical data, as well as the clinical uses of administrative data.
Dr. Lumpkin shared his view of the K2 work group's role as interfacing with the HHS teams and "bouncing" its impressions of their work off the public, thus providing broader input into the process. He said his chief concern is that the Committee not recommend anything in June that will ultimately impede progress and innovation. Dr. Iezzoni said the June purview is so narrow that it is unlikely to close any doors; the question for her is whether the administrative data sets should be expanded slightly to include clinically relevant data.
Bob Mayes, co-Chair of the Infrastructure team, said the implementation teams are thinking in terms of "a phased approach" in which abrupt and disruptive change is avoided and the intention is to indicate the direction the Department plans to head and to introduce changes gradually.
Ms. Greenberg observed that the Committee's core data element recommendations help provide the context for this transition process; she noted that although they are basically administrative data sets, they have clinical utility. Mr. Scanlon added that the first set of standards are seen as a platform on which the later standards for clinical data can be built.
Dr. Lumpkin expressed his view that however useful additional data can be for various purposes, it is important to keep data reporting requirements simple and as inexpensive as possible until the CPR makes it easy to collect a lot of information. Every prospective additional data element must be considered in the light of the cost of adding it.
Mr. Mayes reported that the infrastructure team has several priorities: developing a standard set of evaluation criteria for all the standards, developing a template for the regulatory proposals, and creating a master data dictionary. The NCVHS core data elements will be included in the latter. Ms. Greenberg noted that the Committee's compendium compiling the relevant elements in other data sets in relation to the core data elements should be helpful to the Department's analysis. Mr. Mayes will be provided with a copy.
Asked if the teams were looking at federal programs outside HHS (e.g., AFDC, WIC) and at state efforts, Mr. Mayes said the teams have members from HCFA and some comments from states through the Medicaid Bureau. He agreed that this point needs to be stressed to the teams. CHAMPUS and the VA are also represented on the teams. Mr. Moore suggested inviting people from the welfare side to participate.
Dr. Zubeldia read a statement prepared by the panelists from the previous day who are AFEHCT members representing clearinghouses and VANs. They recommend 1) that the Secretary adopt the NCPDP claim for pharmacy use and the ASC X12 standards for all other transactions, and 2) that the Secretary endorse and recommend that payors use commercial off- the-shelf EDI translator products and discourage the use of one-time custom translation programs.
Dr. Lumpkin invited the group to submit written background and supporting material for its recommendation.
Jeff Blair, Cochairman of the Technical Coordinating Committee of ANSI HISB, presented an inventory of healthcare informatics standards. This inventory included standards that were developed by both ANSI and non-ANSI Standards Development Organizations (SDOs) and was prepared to assist the Department in selecting standards in accordance with the HIPAA mandate.
The templates used by each SDO to create a profile of their standard are included in the appendix. There is one template for transaction standards and another template for frameworks, architectures, and models. In all, the report contains an executive summary, transaction standards, supporting standards, appendices, and templates.
Mr. Blair acknowledged the work of many SDO and ANSI HISB members on the inventory, particularly that of Jean Narcisi who received and compiled the templates.
Dr. Lumpkin thanked Mr. Blair and ANSI HISB for making this major contribution and for providing the Committee with an excellent resource.
Like previous panels, presenters structured their remarks in response to four questions, on 1) their organizations' expectations for administrative simplification under HIPAA, 2) their degree of satisfaction with the Department's process, 3) current problems with the transactions identified for standardization, and 4) how the goal of administrative simplification can best be achieved.
Ms. Narcisi described the ADA's projects with other SDOs to prepare the dental profession for future electronic administrative and clinical applications, based on a vision of seamless data transfer throughout all facets of health care. To this end, in coordination with other SDOs the ADA has established a Computer-based Oral Health Record (COHR), the baseline version of which (#1.0) was released in 1996, referred to as the ADA 1000. It could be used to guide the development of databases for patient information in computer-based patient records and other applications.
3. The ADA believes that most of the transactions specified under HIPAA are appropriate, but it notes that most dental transactions are now transmitted using proprietary formats. It therefore encourages a migration process toward X-12 formats. One exception to its support for the designated transaction standards is health claims attachments, which are not needed for dental claims. Attachments could be eliminated if claims transactions contained uniform content and payors were required to standardize their processing methods. Thus, standardization of claims attachment information should be limited to a well-defined set of possible attachments and eventually eliminated.
The ADA is concerned that the legislation will reach beyond its core focus and seek premature standardization of clinical transactions for the computer-based patient record and supporting code sets. CPR coding issues should not drive the standards for the claims transactions, and adoption of standards for clinical transactions and supporting code sets should be deferred until preliminary work is completed. Finally, the ADA believes that the ANSI HISB should be assigned responsibility for coordinating the computer-based patient record and other clinical standards. It urges that uniform implementation of the standards be required.
Dr. Hammond noted that about 90 percent of vendors in the clinical information system market offer the HL-7 interface. He described the organization's links to other organizations and its international work.
NCPDP is the only ANSI-accredited SDO for the pharmacy services sector. It is a forum in which the pharmacy industry develops standards, using a consensus process, and it represents "the voice of pharmacy" in the standards development arena. NCPDP's version 3.2 has been submitted to ANSI for approval.
1. HIPAA's main impact will be on the health care industry as a whole. Its cost impact on pharmacy will be minimal because pharmacy transactions are already standardized and automated. However, it will have beneficial impacts on this sector by facilitating connections to the rest of health care, providing identifiers that will improve clinical services, enhancing security, and improving the quality of health care.
Mr. Staniec noted that 15 years ago, pharmacy had the same problems other sectors are having now, with a plethora of paper forms where it now has one standard form and process. It recommends the continued use of its electronic claim format between pharmacy providers, insurance carriers, third-party administrators and other parties.
2. NCPDP is pleased with the Department's process.
3. NCPDP has had a successful and evolving standard for nine years.
4. In addition to its previous recommendations, NCPDP recommends attention to interoperability and to standardization at the data content level, including data dictionaries.
Mr. Barrett explained that ANSI does not develop standards; rather, it accredits other organizations to develop standards and fosters a formal consensus-building process. ASC X12 operates under the same principles as other SDOs. The standards development process can take one to three years. X12 is a coalition of more than 800 members, and it works with many other organizations, including HCFA. Key areas of current activity are building a common data dictionary and more common data content among the standards. X12N is not just "a group of technologists" but has public and private sector involvement from many constituencies. It is developing business models for transactions. Mr. Barrett also noted the international scope of its work, which contributes to international standard development.
X12 has been working with other SDOs to resolve outstanding issues, and will hold a meeting in March to discuss the 837 and NSF and try and reach a consensus for a recommendation to NCVHS. It will also try and reach a consensus about pharmacy transactions. It is working with HCFA and others to meet HIPAA time lines and to support the Committee's work.
Asked by Dr. Iezzoni about his organization's thinking about clinical content, he said it is consulting with NUCC, NUBC, AMA, ADA and other constituencies. He agreed that it would be a good idea to have more discussions with end users such as NCQA and JCAHO. Mr. Blair said the ANSI HISB has invited NCQA representatives to collaborate on correlating HEDIS data requirements to the rest of the standards represented by ANSI HISB.
Dr. Starfield noted the value of a table provided by a panelist, listing the HIPAA transaction types against the accredited equivalent. She asked for another column of "other equivalents," and Mr. Moore said Department staff would provide this. Dr. Hammond said tables are difficult because different groups define standards in different ways. Mr. Blair noted that in the inventory, each template indicates the readiness of the standard.
Dr. Braithwaite asked the panelists to comment on how to put teeth into the standards. Mr. Blair suggested possibly requiring whoever chooses to vary from the standard to absorb the cost of the tailoring. Ms. Narcisi said the ADA is exploring how to handle payors' concerns about fraud detection. She noted that payors have admitted that it is cheaper to pay fraudulent dental claims than to process the attachments needed to detect them, and this has resulted in changes in practices with claims. The ADA is trying to develop an accreditation process whereby practice management vendors must meet certain guidelines and pass various tests.
Dr. Hammond urged an effort to find out what information payors need and then reach a compromise on accommodating these needs as simply as possible, leaving aside all the "baggage that has rolled along for years." He added that convincing the industry that the standards are stable and that there will be a return on their investment will also help with compliance. He expressed optimism that with proper testing and certification, payors will cooperate.
Regarding teeth, Mr. Staniec observed that pharmacy standards are voluntary and widely supported. He attributed this to the fact that the standards are driven by business needs and developed through a consensual process. He agreed with the idea of giving teeth to HISB. Mr. Barrett added that education is another key to payor participation, along with testing and accreditation. He supported Dr. Hammond's suggestion of a coalition to try to reach consensus around unresolved issues.
To pin down the idea of testing and accreditation, Dr. Lumpkin asked each SDO representative whether his/her organization is the appropriate entity for testing and accrediting its standards. The ADA, HL7, NCPDP and X12 representatives all answered affirmatively. Dr. Braithwaite concluded that it is therefore not necessary to have a "standards police."
Mr. Beckley (PDX) disagreed, pointing out that SDOs have no power; the power resides with the payor, who can refuse to pay the claim. Dr. Lumpkin countered that HIPAA provisions include penalty clauses with fines for failure to follow the standard. Dr. Hammond pointed out that because of the inevitable ambiguity in the standards, some lack of compliance will be due to confusion. He urged tolerance and a grace period before punitive measures are applied. Dr. Lumpkin observed that the standard enforcement should include testing and accreditation. Mr. Beatty noted that codes are a factor here as well, and something over which the SDOs do not have control. The payors also need to look at the coding, and standards and compliancy checks are needed for the codes.
Mr. Meyer (HBO) pointed out that market factors create incentives for the software industry to demonstrate certification and compliance with standards; there is no such pressure on intermediaries, which are monopolistic. Dr. Lumpkin countered that the political process is such that if the existing penalty clause does not result in compliance, Congress will impose more stringent penalties. He urged that the industry be given the benefit of the doubt, and that efforts be made to address payors' legitimate concerns about fraud. Mr. Meyer responded that payors should beef up their internal audit processes for this purpose, rather than "penalizing the entire market."
The discussion session was set aside for two additional presentations.
This committee deals with data specification, among other things. Mr. Ngo stressed the distinctions between data integration and data sharing and the benefits of the latter, given the differing interests of various parties. The group has set up rules for registration of data elements to create a forum and a process in which a registry leads to negotiation to combine data with others with similar interests. He described the 11179, which includes a framework for how things fit together, a classification of concepts, basic attributes for data elements, rules and guidelines for formulation of data definitions, naming and identification principles, and registration of data elements. The hope for this process is that by sharing information, people will eventually work together and ultimately create a set of integrated data.
Dr. Peters' area of responsibility with both groups is security and confidentiality. His chief message was that most issues in this area are well handled by technologies that are available outside the health care industry in other parts of the computer industry. He identified five areas within health care that require special attention because of differences from the general industry: multiple signatures; confidentiality issues relating to auditing and quality assurance; general professional disciplines regarding confidentiality; and the exchange of information among providers (an area in which American practices contrast with those in Europe).
In addition to noting that most of the work in this area can be adapted from the computer industry, these organizations encourage the Committee to be "relatively stringent" in the security and confidentiality components of its recommendations.
Returning to a general discussion with the panel, Mr. Scanlon asked where the boundaries are between standards and implementation guides, and also how standardized the data items and code sets are within the standard.
Mr. Blair said the HISB inventory indicates the level of specificity for each component in the template. The SDOs that are ANSI accredited are fully standardized. Mr. Staniec noted the distinction between ANSI accredited organizations and ANSI standards, using the NCPDP as an example. The former means that the organization's process for approving standards is approved. HIPAA says the Secretary is to adopt standards created in an ANSI-approved SDO, and there are not many of those.
Mr. Barrett said that some standards are not as specific as would be desirable and include optional or conditional elements. The implementation guides are more specific, to minimize ambiguity. There is a commitment to a consensus process across the industry for developing implementation guides, and this is a major focus of X12.
Dr. Iezzoni noted a reference in the ANSI inventory to the 1,000 internal codes maintained by, and 350 external codes referenced by, ASC X12. She remarked at the high number and asked if there were an effort to "come down to 1." Mr. Barrett replied that the implementation guides specify the code sets to minimize variation. In response to a follow up question, he said there would continue to be flexibility in the diagnostic code sets accepted. He asserted that it would be difficult to achieve a single code set, adding that the organization has not "specifically grappled with that." Ms. Narcisi said the ADA has selected code sets, and sometimes codes, for specific areas. Implementation guides are looking at one application for the health claim. Mr. Staniec observed that pharmacy uses one code set; the remaining issue is educating pharmacists on its use.
Dr. Hammond commented that the problem is less the existence of many code sets than the limitations of all of them for representing data. An additional issue, even if there were a single coding set, concerns the grouping of data rather than its atomic representation. The "ideal world" would be the derivation of higher level sets from a basic atomic set, to meet the needs of both groups. He then commented on issues related to implementation guides and their varied meanings, and the need for conformance documents that remove all ambiguity from broader standards. He predicted that "tool sets" such as this would be developed by the private sector.
Mr. Ngo said the 11179 treats ambiguity by classifying code sets as incomplete, recorded, certified, or standardized. The latter classification means there is no ambiguity.
Dr. Zubeldia observed that there are three kinds of elements in code sets: those with an internal code, those with internal code sets specific to X12, and external code sets with a pointer to the organization that maintains it. Implementation guides can identify what code set has to be used. The problem is with elements that don't have a code set, such as prior authorization, which varies from state to state.
Dr. Lumpkin turned the attention to the Subcommittee's mandated tasks. He noted that the testimony in this meeting suggests a high degree of comfort in the industry with adopting all of the existing standards. He asked whether this reflects the reality, and invited written responses to this question. Using the metaphor of creating an object with inlaid wooden pieces in reference to the Committee's forthcoming recommendations, he suggested that the pieces have to be shaved to fit the space, and the question is where to do the shaving.
Dr. Hammond suggested, instead, the image of laying the pieces side by side and then filling in the rest of the outline. He asserted that it is easy to determine which standard to use for which purpose -- a point that other panelists later supported -- and that the pieces do, in fact, fit together. Another point that other panelists reinforced is that the industry, through the SDOs, is working out common problems through such efforts as building the common data dictionary and creating a common business model.
Ms. Narcisi noted that for dental transactions, only one standard is relevant; the "migration" is necessary because of the large number of proprietary formats in use. She noted that the SDOs are working together on clinical applications for the future.
Mr. Blair agreed that there is little problem of overlap in the next 18 months, because the standards complement each other. The issues to be resolved concern the future, as standards are created for computer-based patient records; that is where the convergence takes place and where the work is needed. He stressed that the SDOs share a vision and have begun to work together. Once costs are more under control, the next large effort will be around measuring and managing quality.
Mr. Moore asked how the Payer ID and Provider ID would affect the SDOs represented, and all responded that the IDs would not create problems. Dr. Hammond commented that although it will be costly initially, having a single number will make possible many things that have been impossible. The crucial thing is to demonstrate that the numbers are truly unique, that they will stick, and that they are ubiquitous.
Returning to the subject of code sets, Dr. Peters stated that vendors' biggest investment is in code sets; mapping is "not a big deal." He noted that the best code sets are proprietary ones. Dr. Hammond stated that the HL7 reference information model includes many contributions from vendors, and he expects the same kind of cooperation for the vocabulary effort.
Dr. Braithwaite asked whether the SDOs were relying on revenue from implementation guides, and how this could be reconciled with the requirement to provide low cost, easy distribution methods for HHS standards. Both Mr. Beatty and Dr. Hammond said their organizations' standards and implementation information are available on the Web. Asked if implementation guides are specific to different industry sectors, Mr. Beatty said X12 has 13 or 14 implementation guides.
Dr. Zubeldia observed that X12 has a patient information transaction that was assigned to carry an HL7 transaction. This could take care of attachments and leave HL7 for what it was designed to do. The general principle is having each standard apply to what it was designed for, leaving unanswered only the question of whether there is enough time to migrate to the desired combination of standards.
The group confirmed that its hearing on coding and classification will take place April 15-16.
Mr. Van Amburg reminded the group about recommendations distributed in November on community health assessment, the result of hearings by the Subcommittee on State and Community Health Statistics. Some Subcommittee members have already commented, and he asked for comments from the others within a week so that the recommendations can be submitted to the full Committee in March.
Dr. Starfield noted the continuing interest in state efforts around administrative simplification, and Mr. Van Amburg agreed that this issue is closely linked with public health and community assessment. The group agreed to ask several exemplary states for written materials and then to decide whether it wants to hold a hearing on this topic in late May or early June. Subcommittee staffer Dr. Gail Janes will develop questions and identify the states with Mr. Van Amburg.
The Subcommittee then passed a motion approving a revised draft of its charge, adding consumer groups to item #2.
Members then reviewed the previous day's presentation and requests by Mr. Hunter and Mr. Arnett, with Mr. Scanlon calling special attention to the issues surrounding provider data collection. Dr. Starfield added these items to the list of issues, based on the previous day's presentation:
· helping keep surveys in the context of other data sources
· thoughts about redesign to fill gaps
· setting priorities across survey areas
· keeping the pressure on for interagency collaboration
· specific tasks such as survey linkage and provider data
Dr. Lumpkin appealed for an overarching vision about what the major questions are, noting that clarity about this would lead to clarity about priorities. Such a vision might also make the Department's data activities more comprehensible to Congress. Other frameworks that might be helpful are the pyramid developed by Dr. Iezzoni and Ms. Coltin, and the Data Council's research initiative work. Mr. Scanlon said there is no "grand vision for surveys," although a conceptual model is reflected in the survey integration plan, having to do with broad population-based interest. He reiterated that the biggest challenge now is in the provider area. He also outlined several initiatives in the public health data area, on which the Committee's advice would be welcome:
· an electronic registry of state health information integration efforts
· conceptual work on a strategy for public health infrastructure data
· performance measurement and community health -- IOM panel
· conceptual work related to performance partnership grants
· Healthy People 2000 activities
· strategies for developing State level data.
Dr. Starfield and others also commented on CDC's interest in using managed care data for public health purposes. Dr. Janes and Ms. Coltin described a CDC project on low birth weight using administrative and public data sets.
Dr. Iezzoni agreed with Dr. Lumpkin about the need for a broad vision, but she pointed out the Committee's onerous other commitments and urged that in this area it take guidance from the Department about where its advice is most needed. This comment returned the focus to the provider surveys. Dr. Starfield asked that all Subcommittee members be supplied with brief descriptions of relevant surveys. This activity will be discussed at the March Subcommittee meeting.
Turning to the Subcommittee's role in regard to administrative simplification, Dr. Starfield noted that one question is what use will be made of the Committee's core data recommendations. Dr. Lumpkin said another is how to approach the transaction standards as stage one of a process that will create an environment hospitable to the development of the computer-based patient record. He raised a question about HCFA's clinical record abstractions, in response to which Mr. Mayes said the core data set that HCFA abstracts is very comparable to the NCVHS core data elements. HCFA has thought about discontinuing this data collection. The current set is primarily fee-for-service inpatients, but HCFA is starting some ambulatory care projects.
Dr. Lumpkin suggested using the information to evaluate the utility of the core data set. Dr. Starfield noted that a key facet of the core data elements is the ability to go across settings, and Mr. Mayes said HCFA is reorganizing so that it can look at quality of care across settings and in terms of episodes of care. It still has to decide whether the project under discussion is worth the abstraction time. He welcomed suggestions from the Subcommittee. Regarding adding clinical data to claims to monitor quality of care, he added that those "on the quality side of the house" have opposed it because the information is not put there for or by the clinician.
Ms. Greenberg commented that to get the core data elements recommended by the Committee, a computer-based patient record is not needed, because most are already in administrative data sets although not in a useful form. She added that no one is undertaking any of the research agenda recommended by the Committee. The question is whether the Committee wants to go beyond the data recommended by NUCC; recommendations of this nature are within its mandate.
Ms. Coltin observed that the Committee's failure to resolve which elements belonged to the encounter and which to enrollment provided an "escape route" for those who prefer not to deal with the recommendations. Dr. Starfield noted that the enrollment and eligibility implementation team could properly deal with some elements.
Dr. Iezzoni asked whether the Subcommittee should be pushing to include clinical elements in the 837, and Mr. Mayes cautioned against using this vehicle since the transactions may not be around for long. It would be wiser, he said, to focus on supporting standards and to lay the groundwork for capturing this information during the process of care. Ms. Coltin expressed frustration that the 837 developers view their task in terms of claims rather than "equivalent encounter information." The latter orientation would permit questions about the uses of the information in a managed care context, but the development process has taken the narrower focus. Dr. Lumpkin suggested that it might be better to start from the beginning instead of trying to convert billing transactions into an encounter document, but she disagreed, saying it is an additive process in which financial data are very germane to setting capitation rates and other kinds of analysis.
Dr. Starfield asked how the Department planned to use the core data elements in its work on standards, and Mr. Mayes said they would be included in the master edit. Mr. Moore referred to the compendium prepared by the NCVHS contractor, comparing the data elements. X12 has also done an analysis.
Mr. Scanlon asked for a sense of what the Subcommittee might report at the March NCVHS meeting in regard to the K2 project. Dr. Starfield asked members to send their comments to Mr. Scanlon, who will compile them for Dr. Detmer for possible referral to the full Committee. Regarding identifiers, Mr. Moore described the Department's process of developing regulations, noting that the process is very deadline-driven and involves a built-in time frame for public review and comment. Payer ID draft regulations will be ready by early March.
In response to a query, Subcommittee members said they are comfortable with the National Provider Identifier, now that it has been separated from the location index.
Dr. Starfield then adjourned the meeting.
I hereby certify that, to the best of my knowledge,
the foregoing summary of minutes is accurate and complete.
/s/ Barbara Starfield December 2, 1997
_________________________________________________________
Chair Date