The Subcommittee on Health Data Needs, Standards, and Security of the National Committee on Vital and Health Statistics was convened on June 15, 1998, in the Hubert H. Humphrey Building in Washington, D.C. The meeting was open to the public.
Present:
John R. Lumpkin, M.D., Chair
Jeffrey S. Blair, M.B.A.
Simon P.
Cohn, M.D.
Kathryn L. Coltin, M.P.H.
Kathleen A. Frawley, J.D.,
M.S., RRA
Daniel Friedman, Ph.D.
Kathleen Fyffe, M.H.A.
Robert
M. Gellman, J.D.
Clement Joseph McDonald, M.D.
James Scanlon, H.H.S., Executive Staff Director, National Committee
Marjorie S. Greenberg, NCHS, Executive Secretary
Judy Ball,
Ph.D., ASPE
William Braithwaite, M.D., Ph.D., ASPE
Karen Trudel,
HCFA
Stanley Nachimson, HCFA
Steve Barr, HCFA
Gary Beatty, Mayo
Foundation
George Beeler, Health Level 7
John Burke, HCFA
Mary
Emerson, HCFA
Milana Gianini, Alternate Link
Betsy Humphreys,
National Library of Medicine (NLM)
Rob Kolodner, M.D., VA and GCPR
project
Tim McNamara, Multim
Dr. Douglas Mann, Battelle Memorial
Institute
Bob Mayes, HCFA
Barbara Redding, HCFA
Frank
Serrao, TSI International
Christine Stahlecker, Empire Blue
Cross/Blue Shield
Maria Ward, Blue Cross/Blue Shield of Illinois
Dr. Kepa Zubeldia, Envoy Corporation
The Subcommittee voted unanimously to amend a section of the Subcommittee's charge and to recommend adoption by the full Committee of the amended charge.
Dr. Cohn and Mr. Blair reviewed the possible scope of the workgroup, including uniform data standards, data security, data quality, and the advisability of developing model, uniform legislation among the states, and the development of a lexicon. Dr. Rob Kolodner of the Department of Veterans Affairs described how the scope under discussion relates to the scope of the standardization project underway in GCPR, and he gave an update on this project.
The Subcommittee agreed on the importance of developing a vision statement on the CPR/EMR and what the Committee expects it to do in various settings. Dr. Kolodner agreed to participate on the workgroup. The group discussed each of the possible elements of the work plan, deciding which to include and which to leave for other groups. The Subcommittee approved a charter and a charge for the work group, subject to modification and subsequent final acceptance. It must still agree on a vision, taking the executive summary of the IOM document as a starting point for this effort.
Dr. Braithwaite explained the highlights of this NPRM and draft NCVHS comments. Much of the Subcommittee's discussion focused on "pay to" organizations. It was suggested that the Subcommittee recommend that in addition to provider and payer identifiers, another classification of identifier be established for entities (such as the rest of health care, clearinghouses, claims processors, PPAs, PPOs, "pay to" providers, and other entities) that do not fall within the categories of payer and provider. The consensus was to recommend that since some of the current EDI routing function is being diminished by the NPI, an identifier should be added for EDI transaction routing.
Members agreed on several comments that will be communicated by letter. Consensus was expressed on the appropriateness of urging HHS to consider having the process for evaluating proposals for new standards rest with a private sector organization with public sector involvement. Members recommended modifying the provision to urge HHS and OMB to follow the intent of the PRA and HIPAA by not applying the Paperwork Reduction Act to the adoption of these national standards.
Dr. Braithwaite also presented draft comments on the NPRM on electronic transactions and code sets. Subcommittee members concurred with the recommendation to incorporate the implementation schedule into the final rules. They agreed in principle with the concept of replacing HCFA HCPCS codes with universal product numbers over time. The sense of the group was to let stand the recommendations to eliminate local codes and to establish an expedited process on a national basis to establish new codes, and for HHS to provide the resources to establish and support the necessary infrastructure. The Subcommittee felt it appropriate to (1) add a recommendation it had made in the past concerning support of standard coding guidelines, and (2) break out the following and combine it with the issue of the unified coding guidelines: "In addition, we reiterate our willingness to work with HHS and other organizations toward the development of a unified framework for coding procedures." Members agreed that long-term care should not be treated with special leniency.
Ms. Humphreys reviewed the current issues surrounding procedure classification systems, some of the attributes of an ideal system, and various ideas for moving toward solutions. She noted the thinking about mapping that emerged from the November CPRI meeting, and she described some of the efforts underway at NLM in this area. NLM is focusing on two important aspects of developing a good system for automated or semi-automated mapping between levels. The first is identifying what is necessary to make mapping work in a semi-automated fashion, and the second is determining the appropriate characteristics of both the detailed vocabularies and the codes. The plan is to secure buy-in from vocabulary and code developers.
The Subcommittee met with the following individuals representing HCFA, vendors, and SDOs: Steve Barr, Gary Beatty, Barbara Redding, George Beeler, Christine Stahlecker, Frank Serrao, and Kepa Zubeldia. The Subcommittee was briefed on the status of HCFA's proof of concept project. They discussed at length with the panel whether to recommend requiring HL7 messages in the X12 275 claims attachment, or whether ASCII should also be permitted. The preponderance of support was for the former option. The Subcommittee voted, with one abstention, to recommend to the full Committee that only HL7 messages be allowed in the X12 275 claims attachment.
Mr. Barr then discussed the results of a stop-work order that has interrupted plans to test claims attachments standards and delayed preparation of implementation guides for the ANSI 277 and 275. Members voted unanimously to advise the Secretary that funding for HIPAA implementation is a good idea, and to recommend that funding be included in the budget and supported. Members agreed with Dr. Lumpkin that more discussion within the Subcommittee is necessary before reporting to the full Committee, given that the standard in question has not yet been tested.
The Subcommittee discussed the first hearing on unique individual identifiers to be held July 20-21, 1998, in Chicago. A contingency plan is in place if the Notice of Intent (NOI) is not out in time: A white paper will present the issues. The NOI itself is a more preliminary process than the NPRM. It will include unbiased questions, background material, and proposals important to the discussion on unique identifiers for individuals. The NOI will go out for a 60-day comment period, the comments will be analyzed, and then a decision would be made on whether to issue an NPRM that would make a proposal for an identifier.
Subcommittee members were asked to review the draft questions that will be asked at the July hearing, and also to identify individuals to invite. There was a lengthy discussion of possible groups and individuals to invite.
Dr. Lumpkin opened the meeting, which was broadcast on the Internet, and requested that attendees introduce themselves.
Dr. Lumpkin reported that under the recommended reorganization of the Subcommittees of the NCVHS, the new name of this group would be the Subcommittee on Standards and Security. He then read the proposed charge. Following suggestions by Drs. McDonald and Cohn, the Subcommittee voted unanimously to amend a section of the Subcommittee's charge and to recommend adoption by the full Committee of the amended charge:
(4) Based upon consultation and analysis, make recommendations to the full Committee on the implementation of the provisions relating to health data standards outlined in Public Law 104-191, with special attention to the standards relating to administrative transactions and electronic data interchange standards, unique health identifiers for plans, providers, employers and individuals, code sets and classification systems and electronic signatures and uniform data standards for patient medical record information and electronic exchange of such information, including computer-based patient record systems.
Under the assumption that a Work Group on Computer-Based Patient Records of the Subcommittee on Standards and Security will be constituted as part of the Committee reorganization, Dr. Cohn reported that he and Mr. Blair had been asked to lead the group, and that Mr. Blair has developed a first draft of a computer-based record workplan. Dr. Cohn acknowledged that administrative simplification provisions, especially administrative and financial transaction standards, have not yet been developed.
Dr. Cohn invited Subcommittee members to join the work group, the immediate objectives of which are to discuss and possibly achieve consensus on a charter for that work group and the scope of its deliverables. Dr. Cohn read the draft charge for the proposed work group. The Subcommittee voted unanimously to accept Dr. McDonald's suggestion to add "other interested parties" to the list of targets in section (4) for outreach, liaison, etc.
Discussions ensued on the distinctions between a "legislative proposal" and "writing legislation," on the possible limitations of stating "patient medical record information" and then "electronic exchange of such information," on the distinctions between paper records and computerized records, and on recognizing the broad variety of providers in the field. To clarify the issue of limitations on types of information to be exchanged electronically, Dr. McDonald suggested the following language in section (4): ". . . adoption of uniform data standards for patient medical record information and related information, and the electronic exchange of such information." The Subcommittee unanimously approved this change and authorized further wordsmithing by Dr. Cohn and Mr. Blair before presentation of the recommendation to the full Committee.
Dr. Cohn identified several concerns about the wide spectrum of possible interpretations of the language regarding uniform data standards and electronic exchange. Issues relating to electronic exchange include messaging that takes place between systems, standardization of data for benchmarking and other sorts of comparative information, management of data in coordination of care at multiple sites, and mobility of information to follow patients' changes in providers. Each of these issues has its own implications for scope and deliverables.
Mr. Blair presented the suggested scope of deliverables to be included in the electronic medical record (EMR) workplan. As background, Mr. Blair noted that the deadline is August 2000 to develop the recommendations and proposed legislation in this area.
The first major area of focus is uniform data standards, including message format standards and patient clinical information standards. This will include data sets, data definition, data models, data elements, and code sets. An inventory of clinical information standards was expected to be distributed, upon which evaluation must be conducted if the Committee chooses to pursue this area of focus.
A second area might be data security, as distinguished from privacy and confidentiality. Data security is the information system's implementation to ensure privacy and confidentiality. By September 1998, it is hoped, ANSI HISB will deliver an inventory of data security standards and guidelines.
A third area for consideration relates to data quality -- the usefulness of the information to be transferred and interpreted once in the information infrastructure. Several organizations involved in data quality might be called upon for guidance.
A fourth area is the advisability of developing model, uniform legislation among the states. Uniform legislation would facilitate development of uniform systems by vendors. Other themes to examine would be privacy and confidentiality, record retention, electronic signatures and their authentication.
A fifth, major focus is the development of a standard lexicon of clinically specific data dictionaries whose code sets are updated on a regular basis and whose associated level of confidence, accuracy and reliability can support continuous availability, especially in acute care areas. A sixth area relates to the ability of the electronic system to retain what are meant to be lifetime records.
Mr. Blair noted that Mr. Bob Mays and Ms. Fyffe had contributed written supplements regarding these approaches. Mr. Mayes questioned whether the resources or time were available to take on such an ambitious scope of work. A time line has been developed to suggest the amount of time required to address the issues.
Dr. Cohn suggested that the members review the six items and discuss whether or not they considered them to be in scope, with the goal to set priorities for the items. Mr. Blair added that, following this discussion, the items that remain on the 's “wish list” would be considered in terms of available resources.
Mr. Bob Mays expressed his concern with the amount of detail to be approached in the proposing of standards, but stated that he believed the Subcommittee was on the right track in its deliberation plan. Ms. Coltin suggested that the scope should be guided by what the Subcommittee felt must be done within the four-year time line dictated by the legislation.
Ms. Frawley agreed with Mr. Blair that section 5(3)?clinically specific data set and codes?is the most critical item in the workplan. Certain other items in the workplan are less important, she noted, because others are known to be taking on responsibility for those tasks. Mr. Blair noted that data security issues are the purview of the . Ms. Frawley stated her opinion that when the security standards are published, they will be less of an issue in terms of the 's workplan because they will have to be adopted, and they will influence the patient medical record systems.
Dr. McDonald cautioned that an arbitrary, but probably inappropriate, distinction is often made between data being transferred and data in a medical record. In order to avoid "separate threads in separate work areas," the Subcommittee should be aware that any element in the medical record might be transmitted.
In response to a query, Rob Kolodner described how the scope under discussion relates to the scope of the standardization project underway in GCPR. The original charge was to develop a common medical record; that charge was interpreted differently, including the provision that the same software must be used. The GCPR group decided that it would take longer to develop a common medical record than to focus on the computer-based patient record (CPR)?the data as distinguished from the system. The partnership has not yet wrestled with the structure of the medical record versus the information in the record.
The project will first define "trigger events," events that cause the need to exchange information among the partners, identify the particular data information needed about the patient or set of patients, and then determine how to standardize that subset.
In recognition that the project cannot do it all, they are taking an approach they call "chunking," that is, "drilling down a little bit and following a path, knowing that it will expand" and that there will be some issues they will not have dealt with yet. They are focusing primarily on the data representation model.
They expect to put in place security policies and standards for information exchange that include methods for tracing the exchanges. Dr. Kolodner noted a concern of Dr. McDonald's that information be available for access in an emergency, yet not abused; the issue is within the scope of what must be addressed, but the solution has not yet been determined.
Communication message formatting is also included in the scope of work, and it is possible that the project will address identifying the "problem list as an entity with certain data." Mr. Kolodner closed his presentation with an acknowledgment that they have focused most of their attention on vocabulary issues as an initial step.
In developing a shared vision of what constitutes an EMR, Dr. Lumpkin suggested the image of describing an elephant from different viewpoints. He stated that to avoid long-term problems, it would be beneficial to develop a statement of vision for an EMR and what the Committee expects it to be able to do in a variety of settings.
Mr. Blair stated that his vision was shaped by the book The Computer-Based Patient Record: An Essential Technology for Health Care. Dr. Cohn concurred, stating that the book would be useful in the effort to finalize scope and deliverables. Dr. Cohn also stated that his vision involved uniform data standards for patient medical record information and the exchange, as opposed to looking at all aspects of what a CPR should be.
Mr. Mayes concurred in the importance of articulating a vision, and stated that one benefit the Committee could offer would be to identify areas that must be focused on to develop national legislation to enable that vision. He noted that the vision enabling might look different from state to state and group to group in terms of its physical implementation. Dr. Cohn acknowledged the importance of agreeing upon a vision statement early in the process for validation.
Mr. Kolodner suggested that principles be defined to guide determination of the scope. In response to an invitation by Mr. Blair, Mr. Kolodner agreed to collaborate with the work group. Dr. McDonald concurred with the usefulness of principles to guide the scope, particularly enabling principles rather than prescriptive principles.
Mr. Blair stated his view of the work group as "a catalyst, a facilitator, a sponsor, a forum . . . to facilitate what is going on out there in the private sector."
Dr. Cohn noted his intention to review GCPR's work in terms of principles and other information, and initiated discussion on the various suggested elements of scope in the workplan. The Subcommittee agreed to include in the scope guidelines and standards for administrative and clinical message formats.
A discussion on patient clinical information (data sets, data definitions, data elements, data models and code sets) touched on (1) classification of data elements as required/not required, (2) a forum for various interests that are developing data models (i.e., Health Level 7 [HL7]), and (3) the pitfalls inherent in using various terms. Mr. Mayes noted the potential usefulness of defining processes for determining data content (i.e., who should decide what is in an ER record, who would review it, should it be reviewed?). Dr. Cohn responded to an earlier question that data sets should properly be identified as required or optional.
Dr. Cohn proposed that data sets be deleted as a specified element; his position was supported by Ms. Coltin. Dr. Lumpkin pointed out that the readability and specificity of records, especially over time, is an important consideration for this work group, perhaps more so than the issue of data sets. Mr. Kolodner stated that the readability issue is important only in situations where one provider's records must be accessed by another provider.
Ms. Greenberg agreed that the Subcommittee need not define the data set for a CPR. Mr. Mayes noted that the challenge of semantic management is heightened when you must link data sets together and manage those semantics in a coordinated, structured fashion; this management is necessary to make sense to the various users and to permit reasonable high-level views as well. Mr. Blair agreed that definition of data sets should be left to a standards organization, but a role for the Subcommittee would be to determine whether these organizations need facilitation to achieve interoperability in a timely manner. Ms. Coltin commented that functionality must be provided to permit a hard-copy print of a patient's medical record.
The discussion proceeded to guidelines and standards for data security. Ms. Frawley reiterated her view that this area is not a high priority for the work group because of the 's limited resources and because an NPRM will be issued on security standards. Dr. Braithwaite suggested that the issue is of limited importance, however, as in a case where the application of security standards to the EPR is problematic.
Committee members concurred that formulation of guidelines and standards for consistent documentation of patient medical record information, including policies and practices, content, quality and technologies (e.g., transcription, scanning, optical character recognition, voice input), is out of scope because of its ambitious nature and because it would be difficult for the work group to make a difference in this area.
The next topic under discussion was model legislation to promote consistent state laws- regulations for the acceptance, authentication, sharing ownership and retention of patient medical information and electronic form across state boundaries. Mr. Gellman pointed out semantic difficulties with the terms "model legislation" and "ownership," and the potential for overlap with privacy and preemption issues. Dr. Lumpkin suggested that such barriers be considered as part of the deliberations in this area, and that appropriate mechanisms, such as rule making, may be used to overcome jurisdictional conflicts.
In order to avoid important aspects of the discussion falling through the cracks, Mr. Blair suggested that input on difficulties should be solicited from vendors who are trying to create systems to be installed in interstate, integrated delivery systems.
Dr. Cohn proposed either asking the Subcommittee on Privacy and Confidentiality to review the issues in this area and to advise the work group on if and how it should be approached, or maintaining the issue as a low-level priority. Dr. Lumpkin expressed his opinion that the item should be retained as a "side dish" on the work group's menu. He reminded the Subcommittee that the members had adopted the principle of floor exemption for privacy and confidentiality, that states would have the right to have more restrictive privacy covenants. By definition, the Committee endorses a system in which laws are different from state to state, as they currently are, and in which mechanisms are in place for moving medical information across the boundaries. Ms. Frawley concurred with keeping the item on the list, but stated that AHIMA developed model legislative language four years ago (which, she noted in response to a subsequent question, covers authentication by such means as electronic signatures), and that many states have already promulgated regulations that accept an electronic record. In addition, she noted, the retention issue is already addressed by federal and state statutes and regulations; ownership is worthwhile to keep on the list, but as a low priority.
The Subcommittee next discussed recommendations to enhance the coordination and maintenance of both administrative and clinically specific data codes that can support the demands of a computer environment that requires continuous availability, greater interoperability, more timely updates and greater clinical specificity. To carry through the metaphor established in the previous discussion, the Subcommittee members considered this item to be the "main course" on the menu.
The next item under consideration was a recommendation to promote development of guidelines and standards for EMR storage. Despite Dr. Cohn's preference for considering permanence in media as out of scope, Dr. Lumpkin championed making a statement to the effect that institutions must take whatever steps are necessary to ensure permanence of the record.
The discussion moved on to EMR processing (i.e., standards for decision support rules, medical knowledge couplers). Dr. Braithwaite suggested that the actual processing would not be of interest to the Committee, but rather the ability to produce and communicate the information from the clinical record through those processes so that appropriate online, point-of-service decision-making can be conducted. Mr. Blair suggested that this may be an area in which to develop a policy or principle. Mr. Kolodner stated that if the goal is to further and facilitate technologies, change should not be impeded by rigidity, and that remaining on a high level makes sense here. GCPR is focusing on the services that allow exchange of data, but this has not yet been addressed by the project. Standards or guidelines for reasonable care might be useful in malpractice litigation. Mr. Blair suggested that more input to the work group would be beneficial in the areas of processing functions and services (e.g., decision support, standards, knowledge representation problem, knowledge couplers, object-oriented standards) to decide whether or not these matters are in scope.
Mr. Kolodner updated the Subcommittee on the recent activities of GCPR. In order to advance public health and individual care, GCPR is focusing on those individuals whom all its members share in common?primarily those who are or were in the military. Project participants are working together on the issue of orders: whether an order was given, as opposed to a prescription or a lab test and its results. It is evident that the scope is at a lower level than that of the work group and the Subcommittee, but the scope is still wide. The plan is for both clinicians and key administrative people to work together on identifying trigger events (as described above). Within that exercise, priorities will be set as to the starting point, which could be as small as a subset of laboratory data or could involve something that is well defined but more complex, such as military discharge physicals. Because the patients of the participants in the GCPR project are mobile in far- reaching organizations, barriers are expected to be encountered.
The project has decided that a key principle will be the use of open architectures and standards; that is, the set of standards for how clinical information is to be exchanged must be in the public domain. Project members believe that their combined size is sufficient to generate adequate clout to put this into practice. It would be appropriate for the work group to discuss the possibility of serving as an advisory group to the GCPR effort. In addition, GCPR has discussed with HL7 the possibility of its being a work group or special interest group to HL7, populated by GCPR staffs or front-line people, in an open process and on a mutually complementary time line. This operational design has as its goal the setting of standards with appropriate input from appropriate stakeholders. Care of the patient is the starting set for GCPR's activity?and for any public health activity?rather than diverse business processes. Dr. Cohn stated that the work group could happily serve in an advisory capacity to the GCPR.
In response to a question, Mr. Kolodner stated that if the GCPR project succeeds in doing only a government CPR, it will have failed, because it is critical to reach all providers who care for their patients. GCPR currently seeks a vendor to provide a new "utility" service-to provide bidirectional routing and switching via the Internet, as well as the critical new function of translation-that will bring information together in a stream to an entity that needs the information, but does not control the wires down to the organizational entities. GCPR expects to work with HL7 on the data model and data representation.
Mr. Kolodner clarified that GCPR will not be a developer of an application per se, but will come up with the data and the network infrastructure, and will look to vendors who can begin to use standards and interface with it. Several applications are expected to be part of that effort certain audit-trail functions may need interfaces, but primarily it will not be the application where the provider interacts with the data. He noted that in addition to HL7, GCPR is also working with Kaiser and Mayo.
Dr. Cohn suggested that the topic of public domain be addressed in the future. He then summarized the meeting thus far: The work group approved a charter and a charge, subject to modification and subsequent final acceptance. The work group must still agree on a vision; the executive summary of the IOM document will be a starting point for this effort. It was clear that a major revision of the scope document will be necessary, and collaboration with Dr. Cohn and Mr. Blair was urged. Plans were made, if time permits, to discuss some of these remaining issues at the Subcommittee breakout session the following day.
Subcommittee members and members of the full Committee will be contacted individually to ascertain their interest in serving on the work group.
Dr. Braithwaite introduced HCFA-45P, "Draft NCVHS Responses to Issues Raised in Notice of Proposed Rule Making (NPRM) on the National Standard Health Care Identifier." The NPRM recommends an 8-character alphanumeric national provider identifier (NPI). After discussion last year, the Committee recommended a 10-digit numeric identifier, and this position is reiterated in the draft comments. In the interest of operational efficiency, the Committee now recommends that the Department also evaluate a slightly longer numeric identifier.
Dr. Braithwaite interrupted his point-by-point discussion to present general comments regarding the NPRM. Two principles are proposed. The first is that the entities to be enumerated should be defined clearly and explicitly, and it should be made clear for what purpose they are to be enumerated. Moreover, only two types of health care providers should be recognized, human beings and institutions that deliver health care services. In the NPRM draft, the NPI covers groups of providers, organizations, and bookkeeping ("pay to") organizations that are not health care providers per se. The Committee suggests that these designations be simplified so that only identifiable entities that actually deliver health care are enumerated as health care providers. "Pay to" organizations that are not health care providers would use their unique IRS taxpayer identification numbers. Ways to identify other provider groups would be developed.
The second principle is that the information used for enumeration be simplified. Only the essential information for uniquely identifying individual health care providers should be used, and the function of enumerating health care providers should be separated out from all other functions proposed for the data items in the NPRM.
In discussion, the members demonstrated general consensus in support of these principles. But some members raised questions concerning how a home health care provider, a school-based clinic, or providers in an emergency room would be enumerated.
Ms. Mary Emerson, co-chair of the NPI work group, stated that while the proposal in the draft comments is viable, other possible proposals also have been put forth. Another proposal was that any provider-related entity that must be identified on a claim or in other transactions should get an NPI.
Ms. Emerson asserted that the draft comments should establish a clear, overarching vision; it would be a simple matter to determine who gets an NPI and under what rules. Under this plan, however, some entities would not have standard identifiers. Dr. Braithwaite pointed out that a specific location in the standard transaction is available in which to designate a contract, thus obviating the need for "pay to" organizations to have NPIs. Further discussion pointed to the drawbacks and confusion that might arise under a system whereby either providers that are difficult to "pigeon hole" would be denied an NPI, or every contract or carve-out of a health provider is given an NPI. A complicating factor is that HCFA must issue NPIs under its existing budget. (Mr. Gellman requested a copy of the legal opinion on this issue.)
Ms. Fyffe complimented Dr. Braithwaite on the clarity of the language in (a) and (b). Dr. Kepa Zubeldia noted that the mandate in the law is to have a unique identifier for provider, for payer, for employer, and for individual. Dr. Zubeldia suggested that the Committee recommend that in addition to provider and payer identifiers, another classification of identifier be established for those entities?such as the rest of health care, clearinghouses, claims processors, PPAs, PPOs, "pay to" providers, and other entities?that do not fall within the categories of payer and provider.
Dr. Lumpkin focused the discussion on section (a). The Subcommittee consensus was to recommend that since some of the current EDI routing function is being diminished by NPI, an identifier should be added for EDI transaction routing?similar to the old OID, but with an updated name.
Concerning section (b), Ms. Greenberg commented that the distinction between enrollment and enumeration is useful. She questioned, however, the advisability of incorporating the provision to eliminate collecting information for other purposes, such as research. This would be the only national provider file that would include all possible plans, and would have potential as a sampling frame for national provider studies by HCFA, NCHS, and others; its value would be enhanced if enumeration were not the only element. The Subcommittee agreed to delete the phrase "such as research," and to modify the last sentence to recommend that the Department of Health and Human Services (HHS) conduct a cost/benefit analysis, because there are important uses of the "other information."
Dr. Braithwaite led the discussion to other areas in the document. Consensus was reached to recommend reconsideration of the cost/benefit of the single enumeration function versus using all the federal programs, based on the fact that the cost of running the registry under these recommendations would be significantly less because there would be fewer providers to enumerate and less data to use, compared to the NPRM proposal.
In response to a question, Ms. Emerson stated that the database is sufficiently robust to recognize two medical functions of an individual; it allows for multiple iterations of a classification. She also noted that an existing robust taxonomy is being used for the NPI.
Consensus was expressed on the appropriateness of the draft comment (on page 9) that urged HHS to consider having the process for evaluating proposals for new standards rest with a private sector organization with public sector involvement.
Members of the Subcommittee recommended modifying the provision to urge HHS or OMB to follow the intent of the PRA and HIPAA by not applying the Paperwork Reduction Act to the development of these national standards by recommending a cost/benefit analysis (page 8, "Detailed Information").
In response to a question, Dr. Braithwaite stated that no code in the identifier will distinguish types of providers, that the identifiers will appear as random numbers.
Dr. Braithwaite introduced HCFA-149-P and draft NCVHS responses to the NPRM on the standards for electronic transactions. Subcommittee members concurred with the recommendation to incorporate the implementation schedule (page 2) into the final rules.
Subcommittee members agreed in principle with the concept of replacing HCFA HCPCS codes with universal product numbers over time.
The sense of the Subcommittee was to let stand the NCVHS recommendations to eliminate local codes and to establish an expedited process on a national basis to establish new codes, and for HHS to provide the resources to establish and support the necessary infrastructure.
Members felt it appropriate to (1) add a recommendation it had made in the past concerning support of standard coding guidelines, and (2) to break out the following (page 3, under local codes) and combine it with the issue of the unified coding guidelines: "In addition, we reiterate our willingness to work with HHS and other organizations toward the development of a unified framework for coding procedures." Ms. Fyffe offered to work with Dr. Braithwaite to soften the language of the first sentence in the local codes section.
The Subcommittee recommended that the language concerning the enrollment transaction be modified to say "under stringent safeguard for legitimate purposes," to encompass both research and business. Members agreed that long-term care should not be treated with special leniency.
In response to a question, John Burke commented that the Office of Management and Budget is soliciting comment on whether or not to require use of a standard in information collection involving a third party.
Upon the introduction of a topic-problems associated with the NDC nomenclature system?that had not been considered previously by the Subcommittee, Dr. Lumpkin suggested that the individual who raised the issue submit his views as a comment to the NPRM.
After a lively discussion on the two options presented in the NPRM relating to new and revised ANSI standards, it was agreed that the Subcommittee would finalize the language of their recommendation during the lunch break. The issues included due process and openness of membership in private data maintenance bodies, whether or not the National Uniform Billing Committee (NUBC) and the National Uniform Claim Committee (NUCC) should continue to exist even if they cannot meet ANSI standards, the appropriateness of the language regarding ANSI accreditation, and the desirability of a large degree of interconnectedness.
Mr. Gary Beatty of the Mayo Foundation and co-chair of the Health Care Task Group within X12N stated that their management group has focused on a number of issues regarding data content committees. Some organizations do not wish to be ANSI accredited or to adhere to the rules and time frames associated with the ANSI standards process. It is important that the committees operate in a coordinated system. Dr. Douglas Mann of the Battelle Memorial Institute stated the need for a mechanism that both provides for timely review and recognizes the changing nature of standards development.
Dr. McDonald asserted that it may be too early in the process to make decisions on code sets, except that there should be one set for each field in the database and that it may be expected to have multiple organizations develop the standards. Dr. McDonald agreed to work on language to express the sentiment of the Subcommittee.
Dr. Zubeldia pointed out that the data content extends beyond code sets to requirements, required/optional data elements, and the constitution of the set for an eligibility transaction, for instance, or a referral. Since some transactions are strictly under the domain of X12, he suggested that X12 be included in the formula.
Mr. Mayes stated that a framework exists for the discussion of medical registries in a structured, focused manner, ISO 11179 and the ANSI X3285 on MediData registries. The idea of registering value domains is gaining attention in health care and beyond.
Ms. Humphreys, National Library of Medicine, stated that one view of a framework for procedure classification would be a coordinated, interrelated set of controlled vocabularies and procedure codes that together would support detailed clinical documentation, statistical reporting, billing, quality assessment, research, and other functions that would need coded procedure data. The framework would also provide a mechanism for automated conversion of the vocabulary used in clinical documentation to whatever level of codes was needed for statistics, billing, or other purposes.
A number of problems have been identified with current procedure coding systems that should be addressed in a new framework for procedure classification: inadequate coverage of nonphysician procedures, duplication of codes for a procedure done in different settings, difficulty in relating hospital and physician data for the same episode of care because the procedure coding systems used for the two components are difficult to relate, inadequate detail provided by current systems developed for statistical and billing purposes, and production and distribution complaints.
The ideal procedure-coding system of the future would be maintained in a timely fashion through an open process with full collaboration of a range of licensed providers. A high standard of openness could be proposed that does not approach the ANSI level; the interpretation of "timely" can vary widely. The need for published, uniform prices or a published pricing algorithm for the network use of the electronic versions of these systems is also evident. The ideal system would also be low cost or no cost.
No consensus in the Subcommittee or in the field has been reached on the ideal system to use as a starting point for the procedure-coding system of the future. Suggestions include starting with a revised CPT, or ICD-10-PCS, or SNOMED International.
The structure of the framework itself has been raised by some individuals: Ought there be a single system developed by a single responsible entity that addresses all levels, or ought there be a set of interlocking, mutually exclusive systems addressing different levels of this procedure? The framework notion, which presumes a set of related vocabularies, is politically attractive. The technical feasibility, practicality, and efficacy of using a set of related codes and of relying on automated procedures to convert data captured at a more detailed clinical, granular level into a higher level of aggregation for statistical or billing purposes, can be tested using current coding systems. Participants at a recent CPRI meeting expressed interest in an official, workable mapping between, for example, SNOMED International and billing and statistical systems such as CPT and ICD-9-CM.
NLM is focusing on two important aspects of developing a good system for automated or semi-automated mapping between these levels. The first is identifying what is necessary to make mapping work in a semi-automated fashion, and the second is determining the appropriate characteristics of both the detailed vocabularies and the codes needed to support accurate automated mapping.
NLM believes that three elements are necessary to support automated mapping: a basic data map between a detailed clinical vocabulary used in patient records and the potentially related diagnosis codes or billing procedures, a set of machinable decision rules that identify the data elements that must be consulted to select among alternative codes and what to do with what values in those data elements, and a clinical information system capable of using the rules and the map. Although none of the three elements exists to date, it is known how to do the basic data map; the process is described well in a Read System Technical Report issued in January 1998 in Great Britain. There is also a place to put these maps; a partial map already exists in the Unified Medical Language System (UMLS) Metathesaurus, and the rest of the map could be created and distributed there.
The Arden Syntax Standard, now under the aegis of HL7, may be the applicable standard for defining and exchanging the decision rules, but R&D is required to define and test both the format and content of the rules. An international standard for format and content would be desirable. R&D is also needed whereby clinical system developers conduct iterative testing of the map and rules to see if they can be used in a clinical information system to achieve automatic conversion from the more detailed level to a higher aggregation needed for either billing or statistical purposes.
NLM's plan is to secure buy-in from vocabulary and code developers; the College of American Pathologists and the American Medical Association (AMA) are considering NLM’s proposal. Standards organizations and other interested parties, including clinical system developers, would be involved. A review would be conducted to determine the functions to be performed and the data required at each of the three levels: data map, decision-rule level, and clinical information system. The data maps would then be put in the UMLS Metathesaurus, and NLM would fund R&D work involving other organizations to develop and test iteratively the format and content of the rules, and the implementation of the use of the maps and rules in the clinical systems.
Ms. Humphreys asserted that this plan would provide useful input into how to chunk data in CPRs and what characteristics and degree of conceptual interrelationship there must be between a detailed clinical vocabulary and a higher-level coding system to make automated mapping a reality. Dr. Cohn opined that mapping may allow for a reasonable framework, and he raised the issue of overlap. Ms. Humphreys asserted that the mapping study may reveal areas offering the greatest opportunity to share resources or to use the same number of resources to accomplish more. She acknowledged that intellectual property issues have not yet been addressed and that the testing of current systems will be incomplete insofar as some existing systems have inadequate coverage of some procedures undertaken currently.
In response to a question, Ms. Humphreys stated that from a technical standpoint, the use of a common set of tools would be the easiest way to ensure compatibility and lack of duplication. Melinna Giannini of Alternative Link, noting that one of the key needs of EDI is to package a broad range of different structures of codes into a messaging format that will work for all parties, suggested that the messaging format initially could be standardized with UMLS and decoded on the other end.
Ms. Humphreys stated that start-up of the project is contingent upon agreement with NLM's potential collaborators, but considerable discussion also must be conducted with the range of interested players to develop a draft proposal on how to instantiate the rules, to evaluate progress, and then to apply the process in a clinical system.
In response to a question concerning clinically specific code sets, Ms. Humphreys assured the Subcommittee that data would not be discarded; it would remain as part of the patient record system (the clinical information system). The goal is to take detailed data and semi-automatically convert it to acceptable billing and statistical codes, which would then probably also become part of the official record. It is not clear at this point how much data to save over what period of time.
Mr. Steve Barr, HCFA, introduced the topics under discussion, the effect of the stoppage of funding for continuation of the Proof of Concept Project and the issue of whether to place ASCII or formatted data into the binary identification segment of the 275 transaction that will transmit information from provider to payer. A number of speakers presented their views.
Mr. Gary Beatty stated that he represents one of the providers that hopes to participate on the pilot project for claim attachments to support the billing process. The Mayo Foundation submits around 80 percent of its claims electronically; greater use of electronic submissions is impeded by the inclusion of attachments information.
Mr. Beatty stated that despite the HCFA stop-work order, Mayo is moving forward with the process to at least set the groundwork for a timely restart. He expressed concern that the Department of Health and Human Services, under the Health Insurance Portability and Accountability Act process, is continually asked to do more with fewer resources, and recommended that they be appropriately funded to ensure that the quality of services they render to the health care industry is "what is expected."
On the subject of codified information, Mr. Beatty recommended that information that can be codified within the 275 transaction be so codified, and that all other information should seek another standard?such as HL7, DICOM, or others?to be codified in appropriate fashion, and, as a last resort, that text format be used.
Ms. Barbara Redding reported on guidance concerning two options: HL7-structured information and either HL7 or ASCII text?provider's choice. The Proof of Concept project team strongly favors LOINC, logical observation identifiers names and codes, because they are more specific than the X12 region codes. NUCC favors options for the provider because it would be more user friendly; advice from NUBC has not been received.
Mr. George Beeler, from Health Level 7, reported that HL7 has participated with X12N for about a year and a half; their activities are complementary. The focus has been to bring HL7's expertise in clinical processes to bear on the claims attachment challenge that X12 and the HIPAA legislation have presented. The quality issue regarding attachments is that they be consistent, that each element is unambiguously identified, and that there be a mechanism to ensure data integrity for appropriate data elements to capture such things as unit of measure, date, and time.
The system under consideration by the Joint Committee is a table-driven model that permits no uncertainty as to what is being transferred. This model also is expected to permit the use of existing and/or rapidly developed software and to be sufficiently flexible to add new responses by defining new sets of codes within the table structure.
Mr. Beeler described the conceptual framework of the model. Each claim attachment or attachment element has its own identifier and an element code. One can define an attachment element that is a simple ASCII text message, and the only "energy barrier" to its use is that it be assigned a unique code to indicate that it is a text for some purpose or that it is carrying a form of response.
Ms. Christine Stahlecker, co-chair of the X12 Insurance , Health Care Task Group, Patient Information Work Group, stated that the current task of the work group is to complete the draft of the claims attachment standard so that HHS can consider it for recommendation to the Secretary for adoption under HIPAA. She presented to the Subcommittee a brief list of definitions and background information on ASCII.
Ms. Stahlecker asserted that although ASCII text can work as an alternative to HL7 messages, it should not be considered as an alternative to HL7. She stated that pursuing the ASCII alternative at this time increases the risk to success of the claims attachment standard recommendation. Ms. Stahlecker summarized concerns with the free-format ASCII text:
What is the benefit of burdening this standard with multiple methods of expressing text, when the HL7 standard message can express both text and codified data?
· Standards developers have not addressed the implications of free-format text as claims attachments.
Identification of the version of ASCII text being used may be required, and it has not been evaluated.
Free-format text in other ANSI standards has not been considered.
Translation processing can be variable; standards for the process are not addressed by HIPAA and presumably would need to be defined in the implementation guide.
§ The publication industry-experts in handling text-is moving toward more structure for data expression, not less. SGML and HTML are becoming de facto standards, among many, for document exchange and processing. XML is being evaluated to express HL7 and X12 standards.
Ms. Stahlecker questioned the benefit of building a base of ASCII installations during the start-up of the Claims Attachment Standard when the industry is moving in the opposite direction.
Ms. Stahlecker acknowledged that advocates of the ASCII text assert:
· The majority of providers do not store medical records electronically. Those that do, use transcription services and may be able to forward text data electronically.
§ Small providers may not be able to create HL7 standard messages.
§ Additional cost and learning curve will be incurred for providers to implement another standard.
Ms. Stahlecker asserted that ASCII text was not perceived as the solution among the members of the work groups. The general opinion was that text data from transcription services can be included in the HL7 standard message; small providers typically use software or service vendors who would offer HL7 capability, and the cost and learning curve for any standard is to be expected.
The benefit derives from having one standard way of expressing data, and this recommendation conforms with other HIPAA standards with the goal to be precise and unambiguous. Medicare will need to assess the position on standards in the free PC software.
Within the work group, progress towards consensus has been made, but "commitment is still elusive." A May 1998 vote approved the use of only HL7 messages in the X12 275 claims attachment recommendation, and the draft implementation guide describing the X12 and HL7 components was completed. The HL7 component is now proceeding through the SDO approval process. No dissent with this position was expressed at the June 1998 X12 working session at which representatives of NUBC, NUCC, and the ADA were present. The X12 275 additional information to support a health care claim or encounter, published as a draft implementation guide for preliminary review, can be found on the Washington Publishing website under "Guides and Development." Comments are to be accepted by the work group until August 2 and will be addressed in the final draft, expected to be completed in October.
Mr. Frank Serrao of TSI International, a data transformation and EDI translator company, pointed out that with adoption of the HL7 standard, small providers may be limited in their choices of products in the market.
Dr. Zubeldia of Envoy Corporation, stating that he was not representing an official position, noted that it makes no difference from the clearinghouse perspective which model is selected. The attachment content will be treated as a binary large object. Dr. Zubeldia suggested other implications for the issue. He raised the question of the attachment being a PowerPoint presentation or a dental x- ray: Although one can put binary objects inside the HL7 inside the 275, the point of doing so is not clear.
Perhaps, he said, there should be three different types of attachments that can be codified in X12 in the 275 itself. The certificate of medical necessity, for example, if it should go out of a claim, can be easily codified in X12. In order to be codified in HL7, it may have to go through "contortions." The second level of attachment could be an attachment easily codified in HL7 or DICOM, or easily codified as a binary file, PowerPoint file, or something else that could go into the binary segment. These two types are preferable "because they are codified and can be processed automatically by a computer in order to help in the adjudication of the claim, which is the bottom line." Other attachments might not be able to be codified and processed by a computer; those could be in ASCII text.
Mr. Barr queried Mr. Serrao whether, if HL7 is chosen, there would be demand for other companies to enter the marketplace. Mr. Serrao replied that demand will create a market, but timeliness is one issue and the structure of their product is another.
Mr. Barr explained that in the NPRM for transactions, it was mentioned that use of HL7 was to be considered for attachments. In his view, the majority of small providers will engage a service to do their billing in the future, leaving only a small minority to do their own programming work. Mr. Braithwaite added that some small providers can support the HL7 message inside the X12 message for attachments, preferably with a phase-in period.
Mr. Beeler commented that only 1 of 70 HL7 messages is under discussion, and that "optionality" would not be part of this process. Ms. Stahlecker stated that direct submissions from small providers are a small percentage of their submitter population, but that a small vendor system may need to be addressed.
Ms. Stahlecker noted that Software Technology's DataGate, which operates on a variety of platforms and handles both X12 and HL7 messages, was to be used by New York Hospital as part of Empire's Proof of Concept test. She suggested the advantage of getting the "low-hanging fruit" those who are investing in electronic capabilities should have a standard developed for them so they might begin to use it. She recommended permitting Empire to continue to test the submission of electronic attachments and medical records to prove the bin segment in the 275 is able to process that.
In further support of HL7, Mr. Beatty noted that they use a common, commercial-off-the-shelf translator for their X12 transactions that also supports the HL7 standard, in the PC environment. In addition, he predicted, innovative solutions will come forward as more electronic commerce becomes available across the Internet in a secure fashion.
Mr. Barr stated that the market has been tested and evaluated in terms of translator types; in the test, five payer locations were required to use a different translator. Before stoppage of the work, five translator companies were in place that were going to use HL7 data within X12N 275 standard messages as a test of that feasibility.
In response to a question, Ms. Stahlecker stated that DataGate supports HL7 and is capable of taking the 275 with the HL7 content, but that it is not the only product that can do that.
In response to a question concerning payers, insurers, HMOs and others who will be receiving data, Dr. Zubeldia said the focus properly should be on providers who will be sending the information.
Mr. Barr reiterated Ms. Stahlecker's point that at a recent conference in a discussion by 200 people, there were no questions or dissent concerning use of HL7.
In response to a question regarding how to achieve the goal of increasing automation in the health industry, Mr. Barr stated that the work group believes the use of HL7 only would be more appropriate. As a point of clarification, Dr. Lumpkin noted that the purpose of HIPAA is to decrease complexity and to simplify, not merely to increase automation.
Discussion continued regarding the two options, including consideration of the desirable degree of flexibility, inconsistency of multiple options with other proposed HIPAA transactions, drawbacks related to the nature of ASCII text and the possible necessity of reformatting, the possibility of transmission via fax or e-mail, and what constitutes simplification.
Ms. Stahlecker noted that providers are not required to bill electronically, and that the true EDI benefit occurs when the processing on the receiver side of transactions can be automated, which drives the quest for a more structured definition of the claims attachment.
Dr. McDonald commented that it appears to some that some providers reject anything that makes it easier to get attachments, which may be perceived by some as harassment rather than information gathering. Restructuring questions in a standardized format might eliminate that perception, it was suggested.
Ms. Maria Ward, Blue Cross/Blue Shield of Illinois, clarified that the work group to which Ms. Stahlecker referred was the Proof of Concept Team, composed of three "Blues," United HealthCare and Exact Medicare Services. It was the vote of those payers, with one dissenting vote, to move forward with HL7 only. The issue of options for translator companies is not seen as a problem for the payer community.
Mr. Serrao suggested that in order to enable small providers, they need a low-end EDI product, such as the one TSI developed for Wal-Mart, with HL7 embedded. Dr. Lumpkin clarified that the more important discussion is focused on providers who use small vendors, and in summary noted that the large vendors are expected to make the necessary changes, but in terms of small vendors, the discussion is on incremental increases and what they must do to differentiate between sending ASCII versus HL7.
Dr. Lumpkin suggested that the committee address in the future ways to discern fraudulent claims and nonfraudulent claims on the basis of evidence-based rationale.
The Subcommittee voted, with one abstention, to recommend to the full committee Option 1, which would require HL7 within X12, 275 bin.
Mr. Barr introduced a discussion on funding and HCFA's stop-work order. He described the work that had been accomplished up to that point, the work that has since continued on a voluntary basis, the collaborative efforts that are no longer enthusiastically supported by the involved parties and the cessation of considerable good will. He explained the expected delays in completion of the implementation guides for both the ANSI 277 and the 275, the halt in testing for the 275 transaction, and the necessity to rewrite the NPRM to address 275 transactions only predicated on work being done by X12 without the Proof of Concept Team.
Mr. Barr asserted that one of the greatest drawbacks of the stop-work order is that the team is now precluded from following through on its original plans to do testing for 277 and test the entire feasibility of the entire phase of using LOINC codes and using 275 transaction and the bin segment and using a new transaction that had never been developed. Ms. Trudel explained that funding for the work of the team was not approved in a fiscal 1998 supplemental appropriation bill, but funding for the claims attachments has been included in the fiscal 1999 and 2000 budget requests.
Following this presentation, the Subcommittee voted unanimously to advise the Secretary that funding for HIPAA implementation is a good idea and that funding should be included in the budget and supported. Dr. Lumpkin suggested that rather than moving forward with an untested standard, more discussion within the Subcommittee is necessary before forwarding a recommendation to the full Committee.
In response to a question, Ms. Trudel enumerated other short-term impacts of limited funding related to some development activities for the National Provider System and for the system that would assign payer identification numbers. The impact on the final timetable and deliverables for those systems is expected to be small.
The Subcommittee discussed the first hearing on unique individual identifiers to be held July 20-21, 1998, in Chicago. A contingency plan was in place if the Notice of Intent (NOI) were not out in time: A white paper would present the issues.
Ms. Emerson summarized the NOI process and explained its differences from the NPRM. The NOI, a more preliminary process than the NPRM, includes unbiased questions, background material, and proposals important to the discussion on unique identifiers for individuals. The NOI will go out for a 60-day comment period, the comments will be analyzed, and then a decision would be made on whether to issue an NPRM that would make a proposal for an identifier.
Dr. McDonald described two events in the last 18 months related to this issue: (1) the high- visibility theft of identities facilitated by agencies that revealed too much information when a Social Security number was given, and (2) advances in biometrics. These events suggest the possibility for new solutions. Mr. Blair suggested that a master person index might be a useful model. Dr. Lumpkin noted that three hearings are planned after the NOI comes out.
Subcommittee members were asked to review the questions presented to them that will be asked at the July hearing, and also to identify individuals to invite. Dr. Braithwaite described the planned structure of the hearing, which would include six panels: providers, plans, SDOs, diverse physicians from the CPRI, public health and research, consumers and privacy advocates, and suggested addition of a panel on use of technology in identifying individuals.
Dr. Lumpkin suggested that the Chicago hearing feature Chicago-based groups, including the AMA, and individuals from the HIV and disability communities. Ms. Humphreys suggested inviting those involved with the Gulf War syndrome, but it was decided that the attendant issues were clinical rather than identification issues. Dr. Lumpkin suggested that there may be advocacy groups for getting information on a particular disease of interest that find barriers to securing aggregated data.
Dr. McDonald suggested organizations that have used the universal health identifier, such as veterans and the military, Canada and Germany. Dr. Lumpkin suggested that this category would be appropriate for a Washington-based meeting; plans for two panels emerged for September?biometrics and international experience.
Ms. Greenberg suggested inviting physicians from such organizations as the American Diabetes Association, which have a strong interest in coordination of care for people who have chronic conditions, as well as continuity of care. Dr. Lumpkin suggested advocates for homeless individuals; Ms. Greenberg noted that HRSA has programs for homeless individuals, which would be a good source, and Ms. Humphreys suggested contacting Dr. Barnett's lab at Massachusetts General Hospital regarding its patient record system used in treating homeless persons. Dr. Lumpkin noted that the integrated maternal and child health system uses the unique identifier. Ms. Humphreys stated that the Western Governors Association, which is implementing the health passport, smart ID, has a very strong interest in technology in health care and telemedicine.
Mr. Scanlon noted that Minnesota is one state that has used identifiers. Ms. Frawley suggested finding out through NAHDO. Ms. Humphreys mentioned Dan Maloney of the VA, who has been following smart card developments for G8 efforts. It was acknowledged that Mr. Gellman submitted a two-page list of potential witnesses.
Dr. Cohn expressed satisfaction with the questions, and suggested that after the July hearing that a smaller group of questions might be appropriate.
With the request that members review the questions before discussion resumed on them at the breakout session the following day, Dr. Lumpkin concluded the meeting.
I hereby certify that to the best of my knowledge, the foregoing summary of minutes is accurate and complete.
/s/ John R. Lumpkin 10/27/98
Chair Date