Department of Health and Human Services
NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS
Subcommittee on Standards and Security
March 25-26, 2003
Washington, D.C.
_________________________________________________________________
-Minutes-
The Subcommittee on Standards and Security of the National Committee on Vital and Health Statistics (NCVHS) held hearings on March 25-26, 2003, at the Hurbert H. Humphreys Building in Washington, D.C. The meeting was open to the public. Present:
Others
EXECUTIVE SUMMARY
March 25-26, 2003
The Subcommittee on Standards and Security held hearings March 25-26, 2003 in an ongoing process focused on HIPAA administrative simplification. The Committee is the main public advisory committee to the Department of Health and Human Services (HHS) on national health information policy. During the two days, the Subcommittee continued their work on PMRI terminologies and heard nine reports and updates regarding efforts and ways to improve and stabilize the HIPAA process; contingency planning related to the October implementation of the administrative and financial transactions rule; and the status of Consolidated Healthcare Informatics’ (CHI) recommendations for terminology use in the federal government, the ASCA database, and the ICD-10 cost/impact study.
Dr. Cohn noted the Secretary had announced the first set of uniform standards for the electronic exchange of clinical data to be used within federal healthcare agencies. He said these standards, based on the Committee’s previous recommendations and CHI’s work, reinforced the importance of the Subcommittee’s work and the integrity of their process.
Review of Agenda
· Jeff Blair
Mr. Blair explained that the Subcommittee set forth in February 2002 the first set of recommendations for PMRI message format standards. Members began studying these terminologies in August and by December they agreed on a set of questions for PMRI terminology developers aimed at capturing the criteria for selection. That morning, Dr. Sujansky would present his compilation and analysis of the developers’ responses and issues he’d identified. In addition to determining a core set of PMRI terminologies, the Subcommittee had to decide whether to consider other domain terminologies and tighter agreement about infrastructure and mapping between the core and other terminologies. That day, members would also consider areas where more input was needed from developers and users. The Subcommittee also would consider policies and practices to address the federal agencies and professional associations interested in the information collected.
Brief Review of Scope and Criteria for Selection of PMRI Terminologies
· Walter Sujansky
Dr. Sujansky said the focus was on selecting a core terminology group that: covered domains of interest, was technically sound and internally consistent, and met requirements for terminology standards for PMRI. Members sought to identify key related but separate terminologies that were widely used and they viewed their relationships with the core terminology group essential.
Review of Questionnaire Response Data
· Walter Sujansky
Dr. Sujansky categorized the terminologies in 10 categories. Responses were numerically scored in keeping with four essential technical criteria. Terminologies meeting them were scored against the desired organizational process criteria. Results were analyzed to determine information still needed and next steps. Dr. Sujansky reported that all four terminologies in the general category met the essential technical criteria and were a rich, representative set that could be included in the core. These terminologies scored consistently well against all quantitative measures including essential and desired technical and organizational process criteria. Dr. Sujansky noted that further analysis was required to resolve a number of issues.
Discussion – Triage Based on Scope
Walter Sujansky
Dentistry, alternative medicine, and nursing were viewed in scope for the initial core terminology set. Members considered a pragmatic approach: defining a manageable number of key categories for terminologies, using these as drivers to identify domain coverage supporting those purposes, and considering the context in which terminologies would be used and reconciled. They considered what constituted the core and noted the needs of necessary specialized terminologies augmented it. Members discussed leveraging data for the development of evidence-based best practices and looking at use cases in terms of a patient medical record’s primary use and systems the terminology drove. The Subcommittee noted mapping was complex and varied and required further analysis. Another issue was overlap. Members focused on interoperability (i.e., sharing) and comparability. They affirmed expanding beyond existing Health Level Seven (HL-7) messages and including UMLS as a bridge to multiple terminologies.
Review of Issues and Actions to Address Them
· Jeff Blair
Mr. Blair noted two additional issues: The Subcommittee agreed that Dr. Sujansky would pursue the missing or incomplete information in the data including licensing cost information, uncertainty about the concept orientation of some terminologies, and determining which nursing terminologies were in SNOMED CT. Members will decide on technical criteria, and then look at the governance structure, finding mechanism, and market acceptance. Terminologies that didn’t make it into the report will be assessed individually. Proprietary information in the questionnaire will be excluded from distribution. Developers will be given a chance to respond to the analysis. Members solicited issues to explore with users of these terminologies and identified three groups (vendors using the terminologies as part of their application software; terminology servers providing support functions; and end-users) as testifiers for May 21. The Subcommittee will brief the full Committee about their draft recommendations at the June meeting.
Day Two
Ms. Trudel said publication of the National Provider Identifier Final Rule was anticipated this summer. HL-7 had requested that the Subcommittee assess a new XML-based standard before deciding on the attachments proposed rule. She noted development of a series of 10 technical information papers providing guidance to providers. Mr. Scanlon said the Secretary’s announcement was a relatively terse statement that the Federal Health Care Enterprise (FHCE) would adopt a suite of about a dozen standards (mostly the Committee’s recommendations) to exchange clinical data. The announcement focused on standards convergence and that FHCE would orient future systems development using this framework. The Secretary hoped the industry followed suit.
Update on Workgroup for Electronic Data Interchange (WEDI) Industry Activities
· Jim Schuping, M.D., Chief Staff Executive, WEDI
· Steve Lazarus, M.D., President, Bounty Information Group; immediate past Chair, Workgroup for Electronic Data Interchange
· Ed Jones, M.D., Chair, Workgroup for Electronic Data Interchange; Commissioner, Electronic Health Care Network Accreditation Commission; Private Consultant
Mr. Schuping reported that WEDI and the task group on the regulatory process, comprised of a good cross section of the industry, were formulating strategies and ideas for renewing the regulatory process and a contingency plan.
Noting that regulatory change came about in numerous ways, Mr. Lazarus presented an overview of discussions about the pass-through deliberatory process. Mr. Lazarus said the group was considering recommending holding to less than four-or-five opportunities for public input as a way to shorten the pre- and regulatory process.
Mr. Jones reviewed the contingency planning group’s efforts. Contingencies from the health plan (payer) side stemmed from concern that providers wouldn’t be compliant and the burden would be on the health plans. Providers worried about a large cash-flow problem. Estimates of non-compliance on October 16 ranged from five to 50 percent. Several WEDI analyses preliminarily indicated that even a five-percent non-compliance rate could trigger a significant cash-flow problem. WEDI sought to identify five-to-ten contingencies likely to be faced on the health plan and provider sides. WEDI also focused on testing issues and vendors.
· Jared Adair, Consolidated Healthcare Informatics Initiative
Ms. Adair presented an overview of the Secretary’s announcement regarding the standards CHI supported for the federal government. Ms. Adair explained that CHI had adopted the first five standards in a series and identified a portfolio of 24 domains that needed to be addressed. CHI had workgroups for the next six standards. Teams worked on six domains. Ms. Adair noted preliminary discussions about setting up a forum with the Subcommittee members and the private sector in mid-April.
· Doug Godesky, Consolidated Healthcare Informatics Initiative
Mr. Godesky presented the project management perspective. CHI’s target portfolio included the 24 messaging standards, domains and vocabulary domains the project defined to OMB. Six teams were deployed in the field with about 30 subject matter experts.
Presentation of ASCA Statistics
· J. Michael Fitzmaurice, AHRQ
Dr. Fitzmaurice reported on the covered entities that opted to extend mandatory HIPAA coverage for their transactions and code transmissions for a year by establishing a compliance plan that started testing April 15 and submitting information to CMS about their process. All 548,644 submissions were ultimately in electronic form (30,000-50,000 were converted from paper). Identifying information was deleted. Dr. Fitzmaurice said CMS addressed the Congressional requirement to look at effective solutions to compliance problems: while the plans didn’t contain effective solutions, they detailed the problems and, by inference, looked at factors influencing those who finished first.
Presentation of ASCA Statistics
Dr. Steindel presented an overview of information in the ASCA database. Ninety percent of those commenting were providers; six-and-a-half percent were health plans. Dr. Steindel noted multiple reasons for delays: the main reasons involved waiting; money was cited by about 17 percent of the respondents. More than a third said they didn’t know the estimated cost. Some 36 percent of those who gave cost ranges estimated under $10,000. Twenty percent considered themselves in the $10,000-$100,000 range. Dr. Steindel defined the process steps: awareness, operational, and testing. Nearly half the respondents hadn’t completed any phase when they asked for the extension. About 30 percent had completed the awareness phase.
Status of ICD-10 Cost/Impact Study
Ms. Pickett reported that a contract had been signed and she emphasized the intent to put this back on track and expedite the process. She said the Subcommittee probably could review results at the May Subcommittee meeting and bring it to the full Committee in June.
Public Testimony
· Tony Martinez, Martinez, Bass and Associates
Mr. Martinez reported that the preliminary registration data for the demonstration project authorized by the Secretary was compelling. The Foundation for Integrative Health Care (FIHC) and Alternative Link received over 8,223 raw registrations and additional faxes and hard copy registrations. Registered entities included over 146 insurance corporations and conventional health plans, 43 HMOs, 12 state Medicaid programs, 129 governmental agencies, 35 integrative delivery networks, 131 hospital systems, 48 PPOs and specialty PPOs, over 58 IPAs, 16 leading medical schools, 961 academics, 24 third-party administrators, and 111 application developers and service providers.
Dr. Cohn noted the next hearing was May 21-22. The first day was devoted to issues related to PMRI terminologies. The second day was originally reserved for ICD-10, but that final report wasn’t in and the agenda for then and the April meeting remained open.
DAY ONE
Dr. Cohn noted the Secretary had announced the first set of uniform standards for the electronic exchange of clinical data to be used within federal healthcare agencies. He said these standards, based on the Committee’s previous recommendations and CHI’s work, reinforced the importance of the Subcommittee’s work and the integrity of their process. Dr. Cohn said that day they’d continue to work on PMRI terminologies.
Review of Agenda
· Jeff Blair
Mr. Blair provided background and context for the day's activity. The Health Insurance Portability and Accountability Act passed in August 1996 charged the Committee to study issues related to the adoption of uniform data standards for patient medical record information (PMRI), the electronic exchange of that information, and reporting to the Secretary by August 2000. That report (available at ncvhs.hhs.gov) set forth guiding principles and the framework used to select and make recommendations for PMRI standards. It also formed the foundation for the first set of recommendations for PMRI message format standards set forth in February 2002. The Subcommittee began studying, evaluating, selecting, and recommending PMRI terminologies last August and by December members agreed on a set of questions for PMRI terminology developers aimed at capturing the criteria for selection. Questionnaires went out January 6 with a February 14 deadline for responses. On May 22 the Subcommittee will listen to user critiques and comments. Any further expert industry testimony needed for clarification will be sought in June. The Subcommittee intends to submit the first draft of their recommendations to the Secretary in July, refine it in August, and bring the recommendations to the full Committee for approval by September.
Responses continued to come, even after the deadline. Some 38 of 41 questionnaires were been analyzed. That morning, Dr. Sujansky would present his compilation and analysis and the issues he’d identified. In addition to determining a core set of PMRI terminologies that came close to meeting the criteria, the Subcommittee still had to decide whether to consider other domain terminologies and tighter agreement regarding the infrastructure and mapping between the core and other terminologies. That day members also wanted to consider areas where additional information was needed from the PMRI terminology developers and users of the terminologies that would testify on May 21. Mr. Blair said the Subcommittee also needed to agree on policies and practices to address the federal agencies and other professional associations that expressed interest in the information collected.
Brief Review of Scope and Criteria for Selection of PMRI Terminologies
· Walter Sujansky
Dr. Sujansky reviewed the scope and criteria document the Subcommittee had prepared as the context for their analysis. The Subcommittee had focused on selecting a core terminology group: a set of terminologies covering domains of interest, technically sound and internally consistent that met the requirements for terminology standards for PMRI. The Subcommittee’s goal was to identify a group of terminologies that would meet those goals. Dr. Sujansky noted they’d also discussed the need to identify important related terminologies in widespread use. He noted that a perspective on the relationship between the core terminology group and related terminologies was essential in facilitating adoption of the standards.
Review of Questionnaire Response Data
· Walter Sujansky
Dr. Sujansky explained that the questionnaire had over 100 questions. Sixty questions were unconditional; affirmative answers to others led to a series of questions. About 60 questionnaires were sent to representatives of 45 terminology developers. And 38 of the 45 developers responded within the timeframe.
Dr. Sujansky categorized the terminologies in 10 categories: (1) general (i.e., MEPS, MedDRA, NCI Thesaurus, SNOMED-CT) with data across many domains; (2) laboratory (HSBT, LOINC, and International Society for Blood Transfusions); (3) drugs (NDC, NDFRT, NDDF Plus, and RX Norm); (4) nursing (nine terminologies); (5) alternative medicine (only ABC codes); (6) dental (four terminologies from the American Dental Association); (7) devices (Universal Medical Device Nomenclature System); (8) messaging terminologies; (9) statistical abstraction and administrative (medical records, clinical purposes, and clinical data analysis); and (10) miscellaneous.
Participants noted that certain codes integrated other code sets and discussed the extent to which overlap of content could be a problem.
Dr. Sujansky noted that three vocabulary sources (UMLS, ICFDH and Multum Lexicon) responded too late for the analysis. Ms. Humphreys suggested, in regard to the Unified Medical Language System (UMLS) Metathesaurus, which contains most of the individual vocabularies being evaluated by the Subcommittee, the chief issue was whether anything in the Metathesaurus might diminish the value of the core terminology set’s components. Dr. McDonald advocated including other large vocabularies. Dr. Fitzmaurice commented on the absence of HCPCS (responsible for 40-50 percent of the spending in the United States), but questioned expending the time and resources to fit them in. The Subcommittee decided to hold that possibility throughout their discussions.
Ms. Humphreys noted that only two of the hundred vocabularies in the UMLS Metathesaurus were appropriately considered the National Library of Medicine’s (NLM) own terminologies: RxNorm and the Medical Subject Headings (MeSH). Reiterating that the purpose of the Metathesaurus was to map among those hundred terminologies, Ms. Humphreys said questions devised for a single terminology didn’t fit. Asked if creating alternative hierarchies could make this less complicated, Dr. Sujansky said it might, but they’d need to discuss that specifically.
Dr. Sujansky reported that nearly half the responses came from terminologies maintained by professional societies and trade groups. Eight respondents were part of the U.S. government. The rest were split fairly evenly between academic, non-profit, private, and for-profit private entities. Transcribing the data onto a spreadsheet and assessing it in terms of completeness quantitatively evaluated it. Dr. Sujansky numerically scored responses in keeping with four essential technical criteria previously identified. Terminologies that didn’t meet the criteria were excluded. Terminologies meeting them were also scored against the desired organizational process criteria. The results were analyzed to determine what information might still be needed and the next steps.
The 38 respondents answered a median of 88 percent of the 60 unconditional questions. Most follow-up questions were answered, though important information was omitted in some areas, especially regarding licensing terms. Many answers were vague. Dr. Sujansky reported they couldn’t determine the accuracy of responses, but could detect internal inconsistencies and disparities with how a particular terminology was otherwise regarded.
Dr. Sujansky said their next step was to apply numerical scoring and categorize the terminologies across the host of criteria. Selection criteria amenable to numeric scoring and questions related to those criteria had been identified and, wherever possible, a three-level scoring metric for the criteria was developed and applied to each terminology.
Concept orientation was the first set of criteria applied. Dr. Sujansky noted that concept permanence, non-ambiguity and explicit version IDs (so parties exchanging data knew they were using the same set of terms) were also important.
The essential criteria were first applied to terminologies that met them: MEDCIN, MedDRA, SNOMED-CT, NCI Thesaurus, LOINC, NDDF, RxNorm, SNODENT, UMDNS codes and HL7 Version 3 codes. The second group consisted of terminologies that failed to meet essential technical criteria with a high degree of confidence. Dr. Sujansky noted this judgment was more subjective and difficult. He reported that most exclusions failed to meet the concept permanence criterion. The third category was indeterminate terminologies based primarily on the concept of orientation. Instead of a distinction between concepts and terms, Dr. Sujansky required a clear definition of a concept, without redundancy or ambiguity.
Dr. Steindel expressed concern that they didn’t continue to analyze the 13 excluded terminologies and pool responses to subsequent questions. He said the profession would benefit from a breakdown of the information on those other terminologies. Mr. Blair replied that they needed to strike a balance between staying on schedule as much as possible and the flexibility needed to reach precise definitions. The Subcommittee hadn’t reached consensus on the definition of a core, and was going forward with the analysis to get closer to a definition. Next, they’d define domains and other terminologies that required mapping.
Dr. Sujansky noted that in the August 2000 report the Committee delineated essential and desired criteria for the core terminology group. Ideally, they’d have rich data regarding the features of all the terminologies, but Dr. Sujansky said that the process, with its limited time and resources, demanded they move along the terminology selection process.
Dr. Sujansky noted that all four terminologies in the general category area met the essential technical criteria and were a rich and representative set of terminologies that could be included in a core. He went on to apply the desired, non-essential technical criteria (comprehensive domain coverage, meaningless identifiers, multi-hierarchies, non-redundancy, formal concept definitions, infrastructure tools for collaborative development, changed sets, mapping to other terminologies, and local customization) and score 23 terminologies that might meet them. Seven of the nine were amenable to scoring.
Dr. Sujansky emphasized that context had to be applied to domain coverage. Adequacy varied from terminology to terminology, depending on whether the domain area was included in the core terminology group. Dr. Sujansky pointed out the need to consider the level of abstraction a terminology might include and whether it was appropriate to PMRI.
Regarding level of abstraction, NDDF Plus and RxNorm differed by a factor of about 12. However, NDDF included all the NDC codes and other levels of abstraction. Dr. Sujansky said NDC might be the wrong level of abstraction for PMRI; RxNorm, with only 40,000 concepts, might include all the important concepts needed for medical records.
Dr. Sujansky described the aggregate scoring for the desired essential and nonessential technical criteria and the organizational and process criteria. Dr. McDonald pressed the importance of distinguishing the nominals and suggested a category for no cost. And he said derivative works that involved using intellectual property in a program (e.g., a vocabulary) would be wanted without constraint, but a derivative that referred to a third variant of the code was more subtle. One allowed that in open source, but Dr. McDonald asked where the standard was if everyone could make derivatives. While it was desirable to be able to incorporate into software freely, he contended that there was a subtle line about holding back on derivatives of the code system itself. Ms. Burke-Bebee noted there was a degree of acceptance in the marketplace for terminologies developed by non-government, non-professional association entities. If their value was great, she said they shouldn’t be disqualified because they charged for using their terminology server or a licence fee per-site.
Dr. McDonald contended they needed to push as hard as possible for a cost low enough that there wouldn’t be financial constraints, inflation, or other traps. Dr. Sujansky noted that significant licensing costs and intellectual property restrictions were associated with the Universal Medical Device Nomenclature System. Dr. Cohn said the underlying issue was that developing and maintaining good terminologies had a cost. He noted a couple financing methodologies. Government could provide grants or other means to create and maintain a code set. LOINC was maintained this way. Dr. Cohn said that any recommendation that involved major licensing costs needed to involve the government in underwriting the costs of the terminology’s maintenance. Dr. Huff added that they didn’t want licensing done in a way that gave a company a government-mandated monopoly.
Discussing drug terminologies, Dr. Sujansky reported that NDF was a proprietary terminology and entailed significant costs. RxNorm scored well in the licensing costs and IP restriction area. NDF also scored fairly well and was likely to pass the essential criteria, but wasn’t included at this time for a technical reason. Dr. Sujansky reminded everyone that RxNorm and the NDF were interrelated.
Dr. Sujansky reported it was unclear whether SNODENT was a part of SNOMED, in terms of licensing. He suggested that if the dental domain was crucial, they might want to look into it.
Dr. Sujansky noted that licensing costs and restrictions were the criteria that caused terminologies to receive a low score in this category of characteristics; few terminologies had requirements for third-party platforms and tools, though some required use of a terminology server. The important point was whether it was possible for people to build their own system to work with the terminology, instead of having to use a tool from the terminology developer of a third party.
Dr. Sujansky also noted that many terminologies that didn’t meet the essential criteria scored well in terms of licensing and IP restrictions. They were maintained as parts of messaging code sets and were already included in those other resources and less costly. Dr. Sujansky said at this point they shouldn’t exclude terminologies based on licensing costs. He said the other aspect of the desired organizational process criteria, responsiveness to constituents, could be scored by looking at the criteria, the update frequency, varied sources of update requests, and availability of training. The update frequency depended on the terminology type.
In discussing details of the scoring criteria, Dr. Sujansky showed the results for “responsive to constituents.” Almost all terminologies that met the essential criteria did well in terms of “update frequency” and “varied sources of update requests”. Only SNODENT and the HL7 Version 3 codes were updated less frequently than annually. Dr. Sujansky said the distinguishing criterion in this category was availability of training.
Dr. Sujansky noted that much of the organizational and process criteria couldn’t be evaluated quantitatively in a straightforward way. He said it was difficult to decide what was a good funding mechanism for the terminology developers. Dr. Sujansky said those responses had to be evaluated subjectively for a smaller set before they could determine which terminologies had long-term viability. Another criterion, governance structure, involved trade offs between formal balloting-based and consensus-based governance structures. There was consensus about the updating of terminologies. Dr. Sujansky said further evaluation and discussion was needed. He noted they’d received a lot of data about market acceptance (e.g., the number of licensees and users, who they were, how many were government organizations), but he said he didn’t know whether having a large number of end-user licensees and no vendors was better than having eight-of-the-ten top vendors and no end-user or government licensees.
Noting questions about licensing costs and IP, Dr. Sujansky offered two totals, excluding that area from one set. SNOMED received the highest scores and the NCI Thesaurus also scored well. MEDCIN scored 17 out of 26. Dr. Sujansky said the question was whether the $25-per-site license was justifiable if SNOMED was selected. He noted LOINC also scored well and was viable in the lab domain. Members noted that the processes LOINC used in other domains were similar to those for the lab domains.
Noting that some terms in MedDRA were entry terms and synonyms for primary concepts, Ms. Humphreys suggested that, instead of a concept orientation, MedDRA consisted of preferred terms with bunches of terms pointing to them, only some of which were synonyms. She said she would verify this by checking on MedDRA data in the Metathesaurus.
Dr. Steindel noted a few responses from vendors indicated different points than Dr. Sujansky presented for the terminology, but he doubted that those changes would affect the aggregate. Ms. Humphreys noted that being as accurate as possible increased credibility. Dr. Sujansky said all the data was available and offered to discuss whatever the Committee questioned.
Dr. Sujansky gauged the completeness and accuracy of the data sufficient to score the terminologies against much of the selection criteria. He noted that the same set of terminologies scored consistently well against all quantitative measures including essential and desired technical and organizational process criteria. This set of terminologies covered much of the clinical domain required for PMRI standards. Overlapping gaps might exist in certain domain areas in terms of diagnoses, symptoms, findings, observations, procedures, and nursing observations. Noting there were no other sources for a comprehensive terminology for device representation, Dr. Sujansky said gaps might also be found if the Subcommittee determined that the UMDNS’s licensing provisions were exclusionary.
Dr. Sujansky noted that further analysis was required to resolve a number of issues. Did SNOMED-CT subsume other general terminologies? Would MedDRA, MEDCIN, or NCI Thesaurus be needed along with SNOMED-CT? Did SNOMED include nursing observations? Could overlapping terminologies be combined in the core terminology group to support a PMRI terminology standard, and if so, how could they be combined? And how could overlaps be accommodated in the terminologies group?
Discussion – Triage Based on Scope
Walter Sujansky
The Subcommittee agreed that dentistry was in scope for the initial core terminology set. Ms. Greenberg pointed out that none of the nursing classifications made the top category. Noting that alternative medicine’s only submission hadn’t met the essential criteria, Ms. Greenberg said she wanted to ensure it received adequate coverage in the core. She also noted that the whole range of functionality and functional status had to be considered.
Ms. Humphreys commented on the potential usefulness of including ICF, which is likely to include concepts that are not well covered in other terminologies being considered. Noting UMLS’s uniqueness, she questioned the value of incorporating the UMLS responses to the questionnaire in a table with many of its constituent vocabularies. On the other hand, even though it is not intended as a clinical terminology, sections of NLM’s medical subject headings (which would score high on the technical criteria, process, and licensing) might fill in gaps in available clinical terminologies in some areas (e.g., investigational drugs).
Mr. Blair agreed that they needed to cover dentistry, alternative medicine, and nursing. He suggested defining the core narrowly and stating that other domains referred to it. Another way would be defining the core broadly in terms of terminologies that met the technical criteria of the core and terminologies that had to work to be recognized as core members. He noted they could also do a combination.
Dr. Steindel asked what constituted the core. A few terminologies came out high in all criteria, but were deficient in specific areas. Ms. Humphreys said perhaps they could move along candidates that covered much of what they were interested in and then focus on the gaps and how to fill them. She noted that specialized vocabularies might already be available to fill some of the gaps.
Dr. Sujansky said the data on mapping was complex and varied among the terminologies and its analysis required further examination. Dr. Cohn observed that the issue of mappings wasn’t about its existence, but about its quality for the intended purpose.
Dr. Zubeldia suggested they could refine the product by having the terminology developers who’d responded review their evaluation and provide feedback.
Dr. Huff said the Subcommittee needed to consider in their strategy that answering questions about overlap required more context of use. He pointed out that much of the apparent overlap between NDFRT, NDDF Plus, and RxNorm wasn’t real: NDFRT and NDDF Plus were dispensed drugs and RxNorm addressed clinical drugs named at the generic level.
Dr. Huff also considered the strategy of implementation. One idea was that core terminologies became part of the Metathesaurus and components participating in messages and in the EHR would be identified out of the Metathesaurus’ total concepts. Another possibility was identifying them as core terminologies, even though people buying that terminology or getting it from a vendor would implement them. Dr. Huff said the overlap would be more troublesome because they wouldn’t be in a single, consistent core and wouldn’t make the correlation between the various terminologies. Their only options were to include or exclude a terminology. He noted they might include a lot of things they didn’t want in order to get these eight important concepts covered in a given domain area.
Ms. Humphreys suggested that, in addition to designating a core set of clinical terminologies, it might also be possible to designate subsets of other vocabularies as part of the designated clinical subset in the UMLS. For example, some concepts only present in the Medical Subject Headings might to tagged as part of the clinical subset if this would help to fill in gaps.
Dr. Huff advocated not thinking of this as static. He pointed out that there could be correctable deficiencies that were immutable. In terms of scope, some things were defined as terminologies because they were intended to report statistics and aggregate findings, rather than primary findings. Dr. Huff emphasized the need to be sure of boundaries and scopes.
Noting that a number of classifications already were accepted as part of the initial HIPAA process, Dr. Cohn suggested that the larger issue was overlap. Dr. Huff said that unless they were specific on where terminologies would be used, they wouldn’t achieve their goal of communicating and sharing this information across healthcare enterprises and government agencies.
Dr. McDonald remarked that scope changed and mutated over time. Everything was a fractal and it might not be good to make such hard lines between them. He noted, too, that historically vocabularies had been word/phrase dictionaries, but they hadn’t yet developed all the concepts to be passed around as identifiable. Ms. Humphreys added that it probably wasn’t possible to establish a principle of how to integrate any vocabulary. You had to know the structure and coverage of the specific vocabularies that you wanted in a coherent set. Then you could develop the principles for integrating them.
Recalling earlier discussions about requirements and use cases, Dr. Sujansky encouraged the Subcommittee to articulate the goals of the PMRI terminology standards they recommended. He said the right model for integrating terminologies depended on what they integrated. For example, if a third party did all the integration, developers of the source terminologies would be uninvolved and the task would be more difficult. Mutual decisions had to be made in order to get a tight, functional, usable integration of multiple terminologies in a core terminology group. Either that, or synonyms could be added that made it easier to map to the designation codes.
Dr. McDonald cautioned that difficult details to be implemented would come up with the mapping process, as they had in privacy. RXNorm probably would want to be freely available internationally. If RxNorm was to be integrated with SNOMED, that might preclude free use of RxNorm internationally. Ms. Humphreys said that the way RxNorm was currently structured the pieces came from sources that could be distributed internationally.
Mr. Blair and Dr. Huff suggested approaching this process pragmatically: defining a manageable number of major categories or purposes for terminologies, using them as drivers to identify the domain coverage that supported those purposes, and refining this further to consider the context in which the terminologies would be used and reconciled. In order to determine the core terminologies, Dr. Huff said they needed to be explicit about the purpose they had for each. The big picture was that they would be used to share patient-specific information between interested and appropriate parties. The drill down was to enable public health data to be passed from healthcare providers to CDC and for CMS to determine the quality or appropriateness of patient care based on these items. They wanted to send data to disease-specific registries so data contribution would be more cost-effective.
Dr. Steindel suggested that they look at their use cases with respect to the primary use of a patient medical record, and the systems the terminology would drive. The primary use cases were internal to the medical institution. If they could cast the information outside the institution, so much the better, but Dr. Steindel said that shouldn’t be a principal driver. Dr. McDonald added other items (e.g., ordering a diagnostic study, medication, or prescription; sending results of diagnostic studies and clinical measurements) necessary within and across institutions, office practice, and service providers. Dr. Sujansky noted that they were talking about exchange of data and delivery of care; electronic exchange of PMRI was an explicit aspect and driver of standardization.
Dr. Huff pointed out that if they shared drug prescription information at this level they could address the set of drug codes needed to enable a particular behavior. There’d be ambiguities about what they wanted with drug codes until they could do this. Considering whether this was broad enough to include ambulatory, alternative health, and long-term care, Dr. McDonald responded that the issue was collecting data, writing orders, providing services, and sending reports. What differed were the specific codes used.
Members reaffirmed the Committee’s vision voiced in their August 2000 report that clinically specific standards for PMRI would bring them closer to an environment where they could capture data once at the point of care for patient care purposes, so derivatives of that information were available for reimbursement, public health, and clinical research.
Dr. Cohn added that the issue of leveraging data for the development of evidence-based best practices also needed to be reflected in their endpoint. He noted they’d talked about two things: interoperability (i.e., sharing) and comparability. Noting Dr. McDonald pointed out that many things weren’t captured, Dr. Huff said he heard that they’d like the terminologies to represent the results of the consultations.
Dr. McDonald commented that they were enumerating messages that would be sent. Practitioners had to do consults, write notes, label items, receive results, and place orders. There were fields for this, essential vocabularies, and values for the results represented as a different set of codes.
Dr. Huff included consultations, goals, and plans on the list. He questioned whether they were trying to choose code sets that represented aggregate classes. He said he didn’t see it within scope to create a data element indicating the number of patients that had adverse drug events during their hospitalization. Dr. Steindel said the PMRI terminology should determine who was on beta blockers, but not the percentage. They sought to derive this from PMRI information, and the base terminology had to contain it. The base terminology also needed to convey information used to derive public health statistics. Dr. Huff said hopefully it would be the same terminology public health used to reduce burden. But he added that the requirement for public health shouldn’t be a deciding factor.
Dr. Huff said another part of this was creating expectations. He emphasized that these terminologies should also be used in clinical trials and in sharing that data. Rather than continuing to create ad hoc terminologies for each clinical trial, they wanted a unified way to tell people about adverse drug events.
Dr. Steindel said that they were also looking at both a difference between standards (which could be more than one group of core sets of core terminologies) and specialized terminologies that could augment the set.
Dr. Steindel noted that the needs of necessary specialized terminologies (e.g., the ICO tooth number designation) augmented the core. If they kept the core four terminologies they would have basic sets of terminologies with little overlap. And if they expanded the core with the specialized terminologies, they’d have a core of up to 20. Dr. Steindel noted another option was to make the specialized terminologies essential parts of the PMRI record. Asked about overlap between specialized and core terminologies, Dr. Steindel pointed out that NCBI terminology demonstrated that mappings identified overlap. He said the Metathesaurus would enable them to identify areas with overlap and extension. Three likely terminologies to have overlap with organisms were MESH, SNOMED and NCBI taxonomy. In many cases, all three had the same organism. UMLS CUIs could identify areas with similar concepts. NCBI terminology might have an organism only used for research that wasn’t listed in other terminologies, but the code for a terminology-designated part of the standard core was considered the primary code. Dr. Steindel said they were relying on UMLS to interchange various code sets; ideally, people would use the designated primary terminology to avoid the extra step.
Dr. Huff advocated including all the terminologies as core, and using frequency of use to prioritize what was included first. He suggested that they could first do the 17,000 most important and frequently used organisms in clinical medicine, but the goal would be to eventually have all 100,000 organisms in the core. Ultimately, all were pertinent in some domain for the consistent, reusable, and comparable exchange of patient data.
Dr. Steindel noted that frequently used nursing terminologies critical for any type of reporting system were in the specialized domain. But he pointed out that, using the criteria they’d established, they wouldn’t think of them as core terminologies. Dr. Sujansky observed that they’d raised an issue of domain coverage versus the technical criteria they’d decided also were required or important for these terminologies.
Noting they’d questioned earlier if there was a way to use this report to alter terminologies so they fit better into the technical criteria, Dr. Steindel suggested that many of the nursing terminologies were examples of that. The January 2003 release of SNOMED included nursing intervention concepts from NIC, OMAHA, NANDA, and the Georgetown Home Health Care Classification and they were working on adding the preoperative nursing data set and NOC.
Dr. Sujansky said they first needed to see how much of the content was already in SNOWMED and how they could move along terminologies that had unique content, but didn’t fulfill the technical requirements. Dr. Sujansky said it would be useful to identify use cases, scenarios or specific requirements for standardizing PMRI terminologies; the question was if they should do it then.
Dr. Huff distinguished between what they were presently doing and terminologies like ICD-9 used for reporting aggregate public statistics, noting they weren’t seeking terminologies for indexing medical literature or reporting the classical use of classifications. Dr. McDonald reiterated the importance of looking at concrete realities and testing the vocabulary. Noting they wanted to avoid missing something extremely important, Dr. Sujansky agreed that they needed these cases and a statement of exactly what criteria was supported. Dr. Huff said he would fill in more detail, looking at clinical trials reporting and disease-specific registries in HL7 and DICOM and using them as a test.
Mr. Blair acknowledged there were pragmatic benefits from using HL7 messages as a driver for semantic clinical specificity of terminologies, but he expressed concern that expanded views of patient care (e.g., alternative medicine and other ambulatory functions) wouldn’t be covered. Dr. Huff noted that HL7 messages were fully capable of sending alternative medical observations or orders and that anything shared had to have a combination of syntactic and semantic content and involved a format of exchange as well as the codes and semantics of data involved in that exchange. Dr. McDonald noted that the federal government specified that they would use HL7 and DICOM. Members affirmed expanding this beyond existing HL7 messages and including UMLS as the bridge to the multiple terminologies in an implementation strategy. Dr. Huff will prepare a working draft.
Review of Issues and Actions to Address Them
· Jeff Blair
Mr. Blair noted two additional issues: missing information and how to get additional information for clarification on open-ended issues. Dr. Sujansky said they needed to pose questions about licensing and the 10 terminologies identified as viable candidates. They also needed to pursue issues of concept orientation; nine terminologies in the intermediate group would be ruled in or out based solely on concept orientation. Dr. Sujansky said he would pursue these. Ms. Humphreys added that they needed clarity about the nursing terminologies that fell into the lower two categories and whether they were incorporated in SNOMED CT.
Asked about market penetration and acceptance, Dr. Sujansky noted they had data about that in the questionnaire but hadn’t yet incorporated it into the analysis. He explained there was a sequencing issue and it took less time to decide on technical criteria first. Once they had a final short list, they’d look at the governance structure, finding mechanism, and market acceptance in a more subjective way.
Dr. Bickford reported that Deb Konicek, who was integrally involved in the incorporation of nursing terminology within SNOMED, had indicated that they had a table that mapped the nursing terminologies to each other. Dr. Sujansky asked if SNOMED had yet incorporated content under its authority and editorial process. Ms. Bickford said they’d incorporated diagnoses for NANDA, OMAHA and Home Health. Her understanding was that all the nursing terminologies they’d been able to secure would be incorporated in SNOMED by January 2004.
Dr. Sujansky said he would learn more about the drug area and include those topics in the revised report.
Mr. Blair noted that they’d set aside May 21 so terminology users could help them identify strengths, weaknesses, limitations and constraints. They planned to verify and validate information received from the terminology developers and to understand other issues related to use and abuse. Mr. Blair solicited issues the Subcommittee wanted to explore with users of these terminologies.
Noting that they were using vocabularies in the business case, Dr. Fitzmaurice suggested that the users give them a paragraph on the business case and how this saved them money or provided additional value to user terminologies. He also suggested asking in which areas they currently used terminologies, their sense of the next three fruitful areas, and about information they wanted but couldn’t get (e.g., medical knowledge they’d obtain if their system could speak the same language as NLM).
Ms. Burke-Bebee suggested that they pursue vendors doing work with servers, databases, and implementing the terminologies that way. And she remarked that sharing the document with the terminology developers that had responded would further pique their interest in participating.
Mr. Blair said he was inclined to stick with the decision made in December to not invite terminology developers to testify because, if they did, they’d have to invite all 38 developers for equity. He said it would be more appropriate for Dr. Sujansky to get clarifications directly. Noting any inaccurate statements in the report could lessen the credibility of the report, Dr. Huff suggested giving developers a chance to respond and correct as appropriate, but not necessarily doing everything they said. Mr. Blair agreed. Pointing out that if they distributed the data, they’d have to ask permission, Ms. Humphreys said they could do both at once.
Noting that waiting for the participants’ response would be beyond the scope of work they’d envisioned, Mr. Blair said they’d have to re-negotiate and expand the consulting agreement. Dr. Fitzmaurice said they didn’t have enough information to decide now. It could sizably expand the scope and they were likely to miss the September deadline. But the Committee agreed that it was important, and thus worth it.
Ms. Greenberg noted that responders’ feedback in the past had been low and the document was available to anyone. She suggested that Mr. Blair and the staff could review responses. Any needing to be addressed further could be passed on to Dr. Sujansky. Dr. Sujansky expressed concern about protracted discussions about subjective issues and explaining principles of terminology merits to 10 or 20 different people. Dr. Huff said it was Dr. Sujansky’s determination whether he’d misunderstood and how he scored that category. Ms. Humphreys doubted that Dr. Sujansky would be inundated.
Dr. Zubeldia suggested summoning the draft, without the spreadsheet, explaining that it was a work in progress. Vendors could submit supplemental information in writing up until the deadline. He suggested making it unidirectional. Dr. Sujansky asked that the input be sent to the NCVH staff for compiling. He reminded the Subcommittee that this might delay the entire process past September. Ms. Greenberg said they could see what they had at the deadline and reassess how that might change the scope or time.
The Subcommittee agreed that comments concerning disagreements and explanations about scoring had to be in writing. Dr. Sujansky said he would address directly issues where additional information was needed, but that was all he would address directly with the respondents at that time. The Subcommittee concluded that Walter would get questions answered as possible, redo the draft for April 17, and subsequently send the draft to the contractors to be returned with comments by mid-May.
Mr. Blair revisited the crafting of an agenda and testifiers for May 21. The Subcommittee agreed that three groups of users of the terminologies would provide the most useful testimony: vendors using the terminologies as part of their application software offerings; terminology servers (e.g., Apalon, Health Language Center) providing enablers, browsers, or other support functions for the use of the terminologies; and end-users (e.g., providers). They would target users who had experience with these terminologies and could critique their strengths and weaknesses. Dr. Fitzmaurice added hospital CIOs, pointing out that they made hard decisions about the worth of putting vocabulary into their systems and might have insights into the trade offs.
Mr. Blair asked for names of entities, organizations, and individuals that would be good testifiers. Ms. Humphreys suggested Dave Riley on behalf of the Department of Defense (DoD) as a user of MEDCIN. She noted that Dr. McDonald knew many LOINC users. Pointing out that if the draft of the use case was available soon it might help in deciding whom to invite, Dr. Sujansky asked how long it would take Dr. Huff to produce it. Dr. Huff said that he’d compile his list in about a week.
Noting it was highly probable that LOINC would be part of the recommendations, Mr. Blair encouraged everyone to tolerate the fact that they were out-of-synch with lab LOINC. The alternative would be to grossly accelerate the terminology selection process. Remarking that there was credibility and integrity to their process and that people felt it had value, the Subcommittee chose to retain the integrity of their selection process that brought them to September, and tolerate being out of synch with the previous selection lab LOINC by CHI.
The Subcommittee discussed concept-oriention. Ms. Humphreys remarked that it wasn’t simply having multiple names for concepts that made a terminology concept-oriented. Ms. Greenberg cautioned that if an essential criterion put a terminology in or out of consideration, they should be able to explain it clearly.
Dr. Sujansky said a problem was that definition required an exhaustive examination of the terminology. The terminologies were coded concepts with possible multiple synonymous text representations and hierarchical or definitional relationships to other coded concepts. He said they could make clear determinations for some. But anything that wasn’t classified elsewhere was vague and not concept-oriented: without examining every code, one couldn’t conclude with full confidence that there weren’t redundant and vague concepts. Ms. Greenberg asked how he’d affirmed some terminologies without examining the code. Dr. Sujansky said he’d decided based on their statement that they represented synonyms explicitly and distinguished concepts and terms. Some had said intent was important: when they found redundant concepts, they remedied them. Their intent was to represent concepts that had the same meaning, explicitly representing that clearly and making different terms for the same concept. Terminology that didn’t distinguish concepts and terms couldn’t do this. The intent had to be concept-oriented. Code sets from four messaging terminologies were in the running: ISO (device communication), DICOM, NCPDP, and HL7 (which wasn’t concept-oriented).
Asked if there was any reason to include the code sets from any of those messaging standards other than HL7 among the core terminology group, Dr. Zubeldia said they weren’t portable from one messaging system to another, so there was no reason to pair them. He added that he’d also listed X12, UBX MAL, and UN EDIFACT, but they had code sets that were unique to their standard.
Ms. Humphreys suggested that it was a problem that these smaller code sets were primarily available embedded lists in the implementation guides for message standards, including HIPAA. They needed to be converted to a reasonable controlled vocabulary form in order to be used efficiently in health information systems. Dr. Sujansky agreed and asked if there was a chance that the complete set of codes used in DICOM would be included in the core terminology. Dr. McDonald asked what the complete code set was. Dr. Sujansky didn’t know. There were about 400 radiology-related items; a handful crossed over into other categories.
Ms. Humphreys predicted that eventually they’d have to determine which of the adopted messaging systems’ codes sets had to go into the UMLS in order to share and map. Dr. Sujansky asked whether the whole terminology would be included or if necessary elements would be placed in another terminology that was part of the core. He considered the latter more likely. Ms. Humphreys said they might instead have an HL7 X terminology that included everything that wasn’t in the other. Dr. Zubeldia said, even if they selected the dosage forms of HL7, those code sets eventually had to be mapped: under HIPAA, the pharmacy message had to be sent in NCPDP and the 835 for payment also would have a different list.
Asking for guidance, Dr. Sujansky agreed that the immediate short-term question was how much time to spend investigating whether DICOM’s codes were concept-oriented. He said he wouldn’t pursue anything that didn’t meet the essential technical criteria. Dr. Cohn pointed out that, although they wouldn’t select them as core, they still might be essential.
Dr. McDonald acknowledged mapping issues, but he said nearly always the problem was slight model differences. He noted this wasn’t arbitrary: there were different sets of use cases. Members commented that the codes within the message standards needed work, but someone other than Dr. Sujansky needed to do it.
Dr. Sujansky decided he wouldn’t look at terminologies they’d already excluded because of concept permanence, but would determine the concept orientation of those in the intermediate category.
Ms. Greenberg said she was still uncomfortable with an essential criterion that the Subcommittee couldn’t define or determine. Ms. Humphreys said the issue was whether there was planned redundancy; often the tip off came through examining how alternative names were treated and what they were.
Members agreed to exclude terminologies that didn’t meet essential technical criteria because of concept permanence. The Subcommittee deferred investigation of the messaging terminologies. Ms. Humphreys will check into MedDRA and help Dr. Sujansky with ICF and ICPC. She suggested that they annotate some of the nursing terminologies they verified were incorporated into SNOMED CT.
Dr. Sujansky summarized that the Subcommittee had agreed that he would pursue the missing or incomplete information in the data they presently had, including licensing cost information, uncertainty about the concept orientation of a number of terminologies, and determining which nursing terminologies were in SNOMED CT. Subsequently, he would revise the draft by April 17. The version would be distributed to the terminology developers who’d responded. Their written comments were to be returned to the staff. Initial assessment would be made and the information forwarded to Dr. Sujansky. A subsequent version of the draft report would be prepared at a point determined by the volume of comments received.
The three terminologies that didn’t make it into the report (ICF, UMLS, Multum) will be assessed individually. The staff will put ICF’s response onto the spreadsheet and apply the scoring criteria, and Dr. Sujansky will review and adopt it as if the original process had been followed. Ms. Humphreys said that if the staff would put Multum’s response onto a spreadsheet and send it to NLM, they’d code it. Members noted that UMLS was really an interlingua of other terminologies and didn’t need to be put into the matrix, but had to be looked at structurally.
Ms. Humphreys noted they might want to look at MeSH after they had identified the core clinical set. It was conceivable that MeSH might fill some of the gaps identified in a subsequent gap analysis. They might also describe MeSH separately for the drug area.
Dr. Sujansky proposed that the questionnaire and responses be made available. Dr. Fitzmaurice cautioned that in that case they might not get any feedback from the vendors they were inviting to testify.
Dr. Huff expanded his list of issues too time consuming to cover in this day’s session to include data exchange, information sharing within the scope of the project, and anything else excluded from the scope. Mr. Blair said the list would serve as a tool to help identify the terminology domains and whether they were in scope.
Members discussed how and if responses to the questionnaire should be distributed. With the Freedom of Information Act, documents submitted to the Committee became public domain and some respondents had indicated that they considered their answers about market acceptance proprietary. Members clarified that proprietary information would be excluded from distribution. Spreadsheets will be distributed only to Committee members and immediate staff. Because this was a federal advisory committee and their materials were in public domain, members only felt it necessary to e-mail terminology developers who’d indicated they were responding with proprietary information on the questionnaire.
Dr. Cohn said the Subcommittee would brief the full Committee at the June meeting about their draft recommendations and directions. Members decided to announce this as step one. The Subcommittee would continue to take steps to narrow the possibilities until they came up with solid recommendations.
Day Two
Ms. Trudel reported that nothing was published since the last meeting. The National Provider Identifier Final Rule was nearly through the Department review; publication was anticipated this summer. She said the national health plan identifier would probably be proposed at least a month after the provider.
HL-7 had requested that the Subcommittee assess a new XML-based standard before deciding on the attachment’s proposed rule. A formal request will be submitted.
Ms. Trudel noted the development of a series of 10 technical information papers providing guidance to providers. Subjects include: how to determine if one was a covered entity, key dates, how to gauge when one’s software vendor or billing service was ready, what to expect from health plans, what one had to know about the transactions, testing, trading partner agreements, and enforcement. The first information paper was published. The rest roll out over the next couple of weeks. All ten will be posted on the CMS Web site and available as reprints.
Asked if changes to the claims attachment affected the timing of the notice of proposed rule making (NPRM), Ms. Trudel explained that depended on the new standard and when it was completed and balloted. HL-7 expected to be done in June. The Subcommittee would see if changes were needed and move the regulation forward. Ms. Trudel predicted it would be at least a few months before the Subcommittee had a document ready. Members agreed to see if HL-7 could brief them in June, so it could be brought to that full Committee meeting. Ms. Trudel said she didn’t know if X-12 and HL-7 would approve the collaborative effort in June, but she suggested the Subcommittee consider that when they looked at how their part of the standard fit in with the enveloping X-12 component.
Mr. Blair wondered how to interpret the Secretary’s announcement the previous Friday in Detroit that referenced the standards adopted by CHI and related it to the National Health Information Infrastructure (NHII), a concept set forth by the Committee. Noting the standards appeared consistent with what the Subcommittee recommended, but that the announcement didn’t explain whether this was the HHS guidance and incentives for the message format standards, Mr. Blair asked if it was a parallel phenomenon. Mr. Scanlon explained that it was a relatively terse announcement that CHI (i.e., HHS, DoD, the Department of Veterans Affairs (VA), and other healthcare agencies) would adopt a suite of about a dozen standards (e.g., HL-7, IEEE, NCPDP scripts, LAB LOINC, and DICOM) to exchange clinical data. These were mostly the standards the Committee recommended: the messaging standards (part of the first PMRI set) and LAB LOINC. The announcement hadn’t focused on incentives, but on standards convergence and that FHCE would orient future systems development using this framework. Mr. Scanlon said the Secretary hoped the industry followed suit, using this leadership and direction with their investments. He noted that VA, DoD, and the Indian Health Service already used these standards to varying degrees and full incorporation and implementation were planned in their systems development work.
Mr. Blair observed that the Subcommittee’s recommendations had indicated that these would be HIPAA PMRI standards as well as guidance to the industry. Mr. Scanlon said the vehicle the Department and federal agencies used was an example and exhortation to the industry. He noted that this strategy had been used many times in health insurance and other technology areas where the standard wasn’t regulated or mandated, but was demonstrated or enforced (e.g., consensus, purchasing power) and was in line with what the Committee recommended. After some discussion about whether the Committee, which serves in an advisory capacity to CHI, should have been recognized, Ms. Greenberg concluded that it might be best when those adopting Committee recommendations thought of them as their own.
Update on Workgroup for Electronic Data Interchange (WEDI) Industry Activities
· Jim Schuping, M.D., Chief Staff Executive, WEDI
Mr. Schuping reported that WEDI was still formulating strategies and ideas for renewing the regulatory process and a contingency plan. The task group on the regulatory process was comprised of a good cross section of the industry: CMS, the Office for Civil Rights (OCR), the standards groups, providers, payers, and clearinghouses.
Responding to Dr. Zubeldia’s reference to WEDI “picking off” one of his attorneys, Mr. Schuping clarified that WEDI had invited Mr. Marks to participate on a number of occasions when they needed his expertise. He said WEDI was aware that Claredi had retained Mr. Marks to do a study.
Update on Workgroup for Electronic Data Interchange (WEDI) Industry Activities
· Steve Lazarus, M.D., President, Bounty Information Group; immediate past Chair, Workgroup for Electronic Data Interchange
Mr. Lazarus presented an overview of discussions about the pass-through deliberatory process. The group continued to work on the issue of regulatory reform or process change. Mr. Lazarus pointed out that the group looked beyond what WEDI normally considered regulatory process. He noted regulatory change came about in numerous ways: e.g., a question or issue submitted to the designated standards maintenance organizations (DSMOs) regarding the transactions, an action within an X-12 and/or NCPDP work group, or WEDI raising a business issue with another organization. A process initiated in such ways didn’t wind up in the formal regulatory process until the organizations balloted or the DSMOs recommended a change to CMS. The issue driving these discussions was that it could take three years or longer from when an issue was raised to have a final rule in the federal register, and another six-months-to-two-years to implement. Mr. Lazarus said the group was considering recommending holding to less than four-or-five opportunities for public input as a way to shorten the pre- and regulatory process. Mr. Lazarus emphasized that the major underlying thrust was to look at the whole process, not only what happened once the federal government began to evaluate a candidate for a regulation change.
Mr. Lazarus noted there was consensus that maintenance issues might be dealt with more easily than major changes and new regulations. WEDI looked for maintenance issues that could be totally eliminated from the process or required only one review. Mr. Lazarus said he thought there also was consensus on other possibilities the group wasn’t ready to report on. He noted they’d mostly dealt with transaction-related concerns, where they’d had the most experience. Mr. Lazarus encouraged the Subcommittee to consider whether their recommendations applied to security, privacy and other regulatory processes under HIPAA. The group will report to the WEDI Board; Mr. Lazarus said he anticipated the Board would make a recommendation by early summer.
Discussion
Dr. Zubeldia asked if they’d looked into the regulatory, X-12, and standards development processes when they culled the number and length of public review records. Mr. Lazarus said they’d looked at the end-to-end process and the number of times there’d been a review. The Committee will make recommendations about what pieces might be shortened or eliminated due to redundancy.
Dr. Cohn’s hierarchy of wishes was that the recommendations didn’t require legislative or major regulatory changes; instead, normal business ways should make the process more responsive to industry needs in a timely manner. Dr. Cohn asked about any pieces they’d established as preferences as they moved forward. Mr. Lazarus said presently they could only report on the recommendation to minimize the regulatory part of the process. Maintenance seemed the most logical candidate and was where they’d focused. Mr. Lazarus pointed out that this didn’t mean that it could be done and that not having a crisp definition of maintenance was part of the issue.
Mr. Lazarus clarified that there was no fixed time frame. Their focus had been on the transactions because that was what they’d worked with most and understood. He remarked that opportunities for more formal public input seemed to slow the process, rather than streamline it. He noted they didn’t have as many opportunities in the security and privacy arena, because they didn’t have the standards organizations or a DEMO process in place. Mr. Lazarus said they hoped to have recommendations on these issues by early summer.
Building on Dr. Zubeldia’s question, Mr. Schuping said he’d asked the Department for clarification about possible recommendations without legislative activity. The feedback was that it was better for them to formulate and submit their recommendations. The Department counsel would then tell them whether they could be done without Congressional or other actions. Mr. Schuping said they didn’t yet have a time line for the task group on regulatory process. They felt an urgency to get their report on the contingency planning work through the Board and to the Department by late April, so the Secretary had an opportunity to make recommendations to the industry by June 1, giving the industry a realistic chance to respond.
Update on Workgroup for Electronic Data Interchange (WEDI) Industry Activities
· Ed Jones, M.D., Chair, Workgroup for Electronic Data Interchange; Commissioner, Electronic Health Care Network Accreditation Commission; Private Consultant
Mr. Jones reviewed the contingency planning group’s efforts, noting that although it hadn’t been on the EMAC Commission’s meeting agenda, the issue of contingency and compliance by October 16 became a two-hour conversation. WEDI’s 31 board members represented the government, standards organizations, providers, and payers. Mr. Jones emphasized that everyone who spoke indicated that he or she would be compliant, but expressed concern about trading partners and whether the transactions would flow.
Contingencies from the health plan (payer) side stemmed from concern that providers submitting electronically with non-standard formats or content wouldn’t be compliant and the burden would be on the health plans. Providers worried about a large cash-flow problem; one recourse was to revert to paper. Health plans worried about being buried under an avalanche of paper.
Mr. Jones noted estimates of non-compliance on October 16 ranged from five to 50 percent. Several WEDI analyses preliminarily indicated that even a five-percent non-compliance rate could trigger a significant cash-flow problem.
Mr. Jones noted that Richard Marks was helping WEDI consider legal alternatives. He said WEDI sought to identify five-to-ten contingencies likely to be faced on the health plan and provider sides. WEDI also focused on testing issues and vendors. Once identified, WEDI wanted to determine solutions and make recommendations to the Committee and the Secretary in three areas: what industry could do within the existing rules, what needed to be done that would be consistent with the Secretary’s authorities, and what might be done with additional statutory authority.
Mr. Jones emphasized that WEDI’s role was to take a particular corporate legal approach to compliance and deal with business associates and trading partners. WEDI strove for recommendations based on consensus that made sense. Mr. Jones stressed that a high priority for the WEDI board was getting recommendations forward no later than June 1, so contingency issues could be addressed.
Discussion
Mr. Schuping reported receiving a number of electronic transactions. Some met the standards; many didn’t. Given the brief time left, he said they were uncertain whether to send them back or come up with a trading partner agreement. He noted that other federal laws or state constraints on claims payment time frames might be affected. He also noted a number of TPA’s handling a large proportion of transactions in the healthcare industry were late getting into things and that this might be a problem on the health plan side. Mr. Schuping said the best service WEDI could bring was identifying the broader context of problems and suggesting areas to address.
Mr. Jones suggested that the Committee recommend that the clearinghouses assist the providers with compliance, particularly in the business associate context. The industry was large, dynamic, and fragmented and the government might have more impact on the provider side. Mr. Lazarus added that a clearinghouse might be a viable option for a health plan or TPA that wasn’t ready. The clearinghouse could take the standard transaction and convert it into a proprietary or paper format. It wouldn’t help that entity get the transactions in place, but it would keep the cash flow and payment process going. On the other hand, because changes in the required variables had to be provided by the provider in the claim, using a clearinghouse for claims transactions might be more of an option for the payer side.
Dr. Fitzmaurice asked if the providers realized that after October 16 the Administrative Simplification Act required all but small providers to submit claims to the Medicare program electronically, which then qualified them as a HIPAA- covered entity. Mr. Lazarus said that WEDI, CMS, and others sent out that message but couldn’t tell how many wouldn’t be ready, though they tried gauging that through who was testing.
Mr. Lazarus said most the providers had required upgrades for their practice management or hospital patient accounting systems to capture new data elements and/or reformat, sometimes using the clearinghouse to format the data. He said they didn’t yet have enough information to know if they were getting rapid responses to their requests to vendors. Mr. Lazarus remarked that having an April date for testing had made people aware of the need to act now, rather than wait until October.
Noting that many trading partners said they’d be ready, but their trading partners wouldn’t, Dr. Zubeldia asked if they could quantify the size of the problem and some statistics. Could they give some idea of the new data elements that Mr. Lazarus mentioned weren’t present and how many transactions were affected by the need to send them? Mr. Jones said that an EMAC commissioner who was also a WEDI Board member would provide an ongoing analysis of his large company’s client base in terms of implementation and compliance. The latest survey showed that about 90 percent of their trading partners were either compliant or testing the transactions. That number had grown fairly significantly over the last couple of months. He said they’d come up with a measure if that didn’t disclose any proprietary information, yet gave a snapshot of what was going on. The data elements were a categorization issue. Mr. Jones said they could identify those.
Mr. Lazarus said WEDI and the Council for Affordable Quality Health Care (CAQH) had launched a Web site that provided health plans an opportunity to voluntarily post their dates of testing and production for each of the transactions. Some 25 health plans had put up their information for the X-12 and transactions. Three submitted their data for the NCPDP transaction. Mr. Lazarus said WEDI and CAQH were about to launch a similar effort on clearinghouse readiness.
Despite a greater capacity to transmit the X-12 and transactions, particularly the claims, Mr. Scanlon asked to what extent not having all the data items caused a large proportion of non-compliant transactions. Noting that some health plans didn’t require those new data items to adjudicate a claim, Mr. Lazarus said the question was whether it could be processed, even though it wasn’t a complete data set. Dr. Zubeldia asked if WEDI could explore this concept. Noting that some clearinghouses put a provider into production if 95 percent of the claims were clean, Dr. Zubeldia questioned if HIPAA set the bar too high.
Dr. Cohn asked what WEDI was doing to get everyone moving forward, particularly in regard to contingency planning. Mr. Schuping said WEDI, through its SNIP initiative, rapidly developed about 30 regional affiliates charged with delivering information to local communities. Affiliates were conducting one-day education programs. In December, WEDI also completed a joint pilot program with CMS, conducting programs in Nebraska and South Dakota. In addition, a number of white papers and other reports published on the WEDI and WEDI SNIP Web sites received thousands of hits each month. Documents could be downloaded for free and disseminated to the communities or republished. In collaboration with OCR, WEDI also conducted a one-day privacy workshop for the faculty of all department staff in the Chicago area. Over 1,200 people participated on a Sunday. A videotape on the OCR program will further disseminate that information.
Mr. Jones said the WEDI Foundation strove to convey information about HIPAA to the African American, Hispanic, rural, and under served communities. He added that the AMA would market a book, HIPAA Plain and Simple, to the provider community. Dr. Cohn requested that the Committee receive a copy.
· Jared Adair, Consolidated Healthcare Informatics Initiative
Ms. Adair presented an overview of the Secretary’s announcement regarding the standards CHI supported for the federal government, noting other speakers and the audience complimented the Secretary on his leadership role. She also said it was significant that CHI acknowledged a leadership role in standards.
Ms. Adair explained that CHI had adopted the first five standards in a series and identified a portfolio of 24 domains that needed to be addressed. CHI’s intent and responsibility was to identify additions to both. Ms. Adair noted that CHI had workgroups for the next six standards. The workgroups were at various stages and could end up with a range of results. CHI hoped to find and, working with the Committee, present a mature standard that met all the needs. Ms. Adair noted, however, that CHI might find that there was either no standard for what was needed, none mature enough to adopt, or that all immediate possibilities lacked essentials. Ms. Adair said CHI would report such findings to the Committee and, hopefully, inspire industry to mature adoptable standards.
Ms. Adair reported that teams were working on six domains: medications, interventions/procedures, demographics, immunizations, lab result content, and the clinical encounter. As an OMB-sponsored eGov initiative, CHI had a responsibility to post their progress on their Web site (eGov.gov).
Ms. Adair emphasized the importance of working together, the private sector’s confidence in CHI, and how much of everyone’s needs they could meet. Ms. Adair noted that preliminary discussions with Mr. Blair, Ms. Greenberg, and Dr. Cohn about setting up a forum with Subcommittee members and the private sector in mid-April. She emphasized the importance of not waiting until the Committee next met at the end of May.
Noting LOINC as an example, Ms. Adair pointed out that the work of each group wouldn’t likely be one-size-fits-all. She explained that CHI deployed a group to look at a domain, and because of other work they’d done, basically knew the definition of that domain, what needed to be included, and where to draw lines around it. Participating in so many meetings and activities, they also understood the universe of standards in the field. Ms. Adair said CHI felt comfortable moving ahead with their analysis and participating with the Subcommittee in a public forum to explain what they’d seen, the issues, and their preliminary recommendation. Noting a team in another domain might not have a clear definition or feel they understood this niche in the universe of potential standards, Ms. Adair said they could ask for help (e.g., request the Subcommittee host an activity that allowed CHI to get information on the front end, work on it, and hopefully make preliminary recommendations). Ms. Adair thanked the Subcommittee for the morning’s research-and-analytic discussion, noting it helped her realize that she hadn’t thought through every scenario and those were the two extremes. Ms. Adair envisioned that CHI would see if there were other scenarios they could present in a forum Mr. Blair and Dr. Cohn would host, so everyone could see if they worked. Acknowledging that others might have different opinions, Ms. Adair emphasized the importance of getting people to understand that the recommendations made sense.
Discussion
Dr. Cohn spoke for the Subcommittee, stating they would do everything they could to move the process forward. Ms. Adair said CHI wanted to understand the spectrum and people’s opinions during their recommendation and adoption process. While they were adopting for the federal government, Ms. Adair said CHI was happy that their Council was reflective of people who provided care, paid for it, and used the data, because they formed a cross section for looking at true needs.
Dr. Huff seconded the idea that the most communication possible between those choosing standards and CHI was good. All the standards organizations were considering what terminology should be adopted for use in those standards and potential was there for them to choose differently than CHI. Dr. Huff said he strongly supported this opportunity through the Committee for an open dialogue between CHI and the standards groups.
Ms. Adair explained that due to their own business drivers CHI was establishing priorities within domains where they deployed teams. She noted the possibility that CHI might ask the Subcommittee who was working on the same things if there was a way for them to reprioritize in order to be of even greater assistance. Ms. Adair said CHI recognized that this might change some of the Subcommittee’s plans, but she emphasized that at least having these conversations would be helpful, supportive, and consistent with what Dr. Huff just said.
Dr. Huff reflected that another level needed to be considered as they put these standards into use. He noted, for example, that drug codes were used in more than one context. One set of codes might be needed for drug ordering and others for talking about what a patient was allergic to or tracking what was dispensed. Dr. Huff said presently the big work was deciding what part of a terminology should be used specifically in a given part of a message. He pointed out that they couldn’t address that until they talked about the message structure and exactly where they’d use it. Dr. Huff emphasized that numerous complementary activities were underway and it would behoove the standards organizations and CHI to stay tightly coupled.
Ms. Bebee said one area that they could work on right away was preparing for the May meeting. Dr. Fitzmaurice and she had been putting together questions for the three panels and would circulate that to the group for input.
· Doug Odesky, Consolidated Healthcare Informatics Initiative
Mr. Odesky presented the project management perspective. As an OMB sponsored initiative for eGovernment, an expectation of the President’s Management Agenda was that a national initiative was highly structured with specific milestones and goals. Mr. Odesky said CHI worked hard the first year to develop a target portfolio. The targets were the 24 messaging standards, domains and vocabulary domains the project defined to OMB. Targets could change, but wouldn’t likely shrink. And the portfolio of eGovernment projects had a defined start and end point.
Six teams were deployed in the field with about 30 subject matter experts. Each team had a lead and an expert helped harmonize the teams’ work. As Ms. Adair had described, solutions would come more readily for some teams, and Mr. Odesky expected uneven results in subsequent stages.
Discussion
Ms. Greenberg clarified that these teams were a cooperative activity of HHS, DoD, VA, and other agencies. Dr. Steindel said CHI was attempting to pull in USDA, EPA and all units of the federal government. Mr. Odesky added that the process ensured that each team included an HHS, DoD and VA member. The government structure required that the team defined their domain and the range of topics addressed, then researched and made recommendations to the Council. Mr. Odesky noted that the entire Federal Health Enterprise was represented at the table. Each could vet the recommendation in their agency and bring it back for the Council decision on government-wide adoption. Dr. Steindel added that because they recognized that the federal government might not have expertise in all areas, workgroups could consult with outside experts.
Noting that the government and private sectors had a number of cross-cutting strategic issues (e.g., patient safety), Dr. Cohn asked if other processes were used to ensure those issues were addressed or part of a team charge. Dr. Steindel said CHI recognized patient safety as an important charge in two broad realms: medication-related patient safety and other areas. They’d charged the medication group to look at patient safety for medication errors. Dr. Steindel said they planned to look at public health reporting to be sure they received reports of patient safety. The devices group was going to look at another area of patient safety. Those groups hadn’t yet been formulated.
Mr. Odesky outlined the need for a strategy to form this portfolio of 24 domains, as defined within the CHI project. He emphasized that they needed to go through the governance and adoption process for standards that had many clear, strong choices and an outcome within reach. He said they needed to wrestle with how the government would stimulate solutions in order to close the gaps. Mr. Odesky said they needed domains to deal with gaps and where they projected hard and even no choices. He said they also needed models for how government would deal with more domains that faced these challenges long term.
Recalling the press announcement, Mr. Blair asked how the standards meshed with activities within HHS on NHII. Mr. Scanlon said standards were necessary, but not sufficient. HHS was trying to place these activities in the context of a full NHII vision so people understood where privacy, standards, research, demos and leadership, convening-type activities, and work with NCVHS fit in. All of this was conceived as contributors to and elements of the NHII.
Presentation of ASCA Statistics
· J. Michael Fitzmaurice, AHRQ
Dr. Fitzmaurice explained that ASCA gave covered entities an option to extend mandatory HIPAA coverage for their transactions and code transmissions for one year. To do so, entities had to establish a compliance plan that started testing on April 15 and submit information about their process. CMS had been charged with setting up the process (they produced a mandatory model application form available on their Web site) and giving the Committee a sample compliance plan for analysis. Dr. Fitzmaurice noted they were dealing with a whole population. All 548,644 submissions were ultimately in electronic form (30,000-50,000 were converted from paper submissions). In accordance with ASCA, identifying information was deleted, protecting the confidentiality of the covered entities. Dr. Fitzmaurice noted 94 percent of the respondents were providers. Health plans represented six-and-a-half percent; clearinghouses, one percent.
Congress charged NCVS with publishing and disseminating reports to the public containing effective solutions and analyses to compliance problems identified. The assumption was that those finishing first had better processes. Problems linked to inabilities in meeting the October 16, 2002 deadline were also studied. The average number of problems reported was three.
Dr. Fitzmaurice said they’d addressed the Congressional requirement to look at effective solutions to compliance problems, but he cautioned that the plans didn’t contain effective solutions. However, they detailed the problems and, by inference, looked at factors influencing those who finished first.
Presentation of ASCA Statistics
Dr. Steindel presented an overview of information in the ASCA database. Ninety percent of those commenting were providers. One percent were clearinghouses; about six-and-a-half percent were health plans. A few responders considered themselves multiple types: clearinghouses, health plans, and providers.
Some 44 percent of the providers indicated they were physicians or group practices. A third of the respondents indicated they were “Other”. Eight-and-a-half percent were dentists, three-and-a-half percent were pharmacies, three percent were nursing homes, two-and-a-half percent were durable medical equipment suppliers, two percent were hospitals, one-and-a-half percent were home health agencies, and half-a-percent were hospices.
Respondents were asked where they were in the process and, if requesting an extension, why, participants chose from a list of 12 reasons. Dr. Steindel noted there were multiple reasons for delays. The median number of reasons selected was three. 90 percent of the respondents gave between two and seven reasons. A few gave all 12. A large number gave only one reason.
Dr. Steindel reviewed the rank order of reasons for delays of the total population and breakdowns of reasons for delays by the various entity and provider types. Noting the observation might be useful for targeted interventions, Dr. Steindel pointed out that the main reasons for delay were waiting reasons. First was implementation time: two-thirds of the population felt they didn’t have enough implementation time. Nearly half the population was waiting on their vendor with no details provided. About 45 percent of the population waited for standards information. And Forty one percent felt they needed more testing time. Over a third of the population sought clarification on the standards. A third waited on their clearinghouse.
Money was cited as a reason by about 17 percent of the population; hardware by 16 percent; staff by 12 percent, data requirements by nine percent. Other reasons were cited by eight percent. Dr. Steindel noted they’d collected text for the other reasons, but it hadn’t yet been analyzed. About seven percent of the population cited code set implementations.
Dr. Steindel observed that some reasons for delays (e.g., waiting on clearinghouses or vendors) depended on others and appeared beyond the respondents’ control. Other reasons could be controlled and, in many cases, had been since the extension process began. Dr. Steindel noted they’d heard that morning about educational and other clarification materials, concerning standards and implementation information. He noted the length of calendar time for implementation and testing seemed to be a major reason for seeking extensions. He suggested that some of the time delays for clearinghouses and vendors might be due to waiting for final rule publication. Dr. Steindel noted that the final terminology standard didn’t come out until February 20, after the ASCA registration process. He emphasized that a sizable percentage of the major reasons for delay weren’t internal.
More than a third of the respondents said they didn’t know the estimated cost. Some 36 percent of those who gave cost ranges estimated their cost would be under $10,000. Twenty percent considered themselves in the $10,000-$100,000 range. About 5.1 percent designated between $100,000 and $500,000. Under two percent fell between $500.000-$1,000,000. And 2.4 percent said the cost would be over $1,000,000. Dr. Steindel said it would be interesting to look at the provided breakdown of the estimated cost by provider and entity.
Dr. Steindel noted the application defined the process steps. The first phase was awareness. The second was operational. And the final phase was testing. Nearly half the respondents hadn’t completed any phase when they asked for the extension. About 30 percent had completed Phase One. Only 3.4 percent of the population had completed testing or Phase Three.
Dr. Steindel noted the questionnaire asked the month and year a phase was begun and completed. In calculating the number of days people predicted the process took, he estimated that phases began and ended the first of the month. The maximum number of days people indicated for Phase One was 335; the median number of days was 61. The maximum number of days indicated for the operational phase was just over 3,200; the median number of days was 184. The maximum number of days was just over 3,100 for testing; the median number of days was 153. Dr. Steindel emphasized the number in the five-to-95-percent range: 0 - 273 days for awareness, 0 - 550 for operational, 0 - 489 for testing. The median number of days indicated from the start of the awareness process to the end of the testing process was 304. The number in the five-to-95-percent range was 0 - 669 days. The maximum number of days was just under 3,000 days for the whole process. Dr. Steindel emphasized that not all respondents gave numbers for every phase and half the population said they hadn’t even begun.
About 42 percent of the respondents hadn’t reviewed the current process. Half said they’d just begun the review. About seven-and-a-half percent said they’d completed their review. Dr. Steindel reported that 54 percent hadn’t started computer software development, 43 percent hadn’t completed it, and three-and-a-half percent had completed it. Some 55 percent hadn’t started the computer training for the staff, 43 percent had begun but not completed, and two-and-a-half percent said they’d completed training.
Noting that ASCA was asked to provide examples of best practices, Dr. Steindel pointed out that no questions had been asked specifically about practices. To help satisfy this requirement, he explained that they’d looked for attributes in the groups completing all the phases, though they couldn’t say that completing the phases meant best practices were used. He noted they’d selected it as the one group that they might make meaningful statements about. A logistic regression analysis was done against key factors that could contribute to completed testing: entity type, estimated cost, number of delay reasons, provider type, use of a contractor, and geographic location (i.e., CMS regions the entity was located in). They’d found that: clearinghouses had completed Phase Three more than others, those implementing within the lower ranges of the cost estimates had a better chance of completing Phase Three than those responding that they didn’t know, and dentists had the best chance of completing Phase Three.
Discussion
Dr. Zubeldia remarked that the provider table showed that dentists, with a total of 35,000, represented 8.5 percent of the providers. He said about 17 percent of the dentists filed claims electronically and reported they’d be ready. He said the numbers were skewed because most dentists, many of whom filed electronically, didn’t file for the ASCA extension. Pointing out that durable medical equipment suppliers, hospitals, and pharmacies also showed a high odds ratio for completing Phase Three, Dr. Steindel said many discussions could be made from this same point of view. Dr. Fitzmaurice said those groups employed people to focus on business practices whereas many smaller providers concentrated more on patient activity and left billing to a clerk or billing service.
Dr. Steindel observed that those who’d used a contractor had a greater chance of completing Phase Three than those undecided about contractor use. He said a couple of CMS regions showed a significant odds ratio of completing Phase Three. Dr. Steindel was unsure what this meant, but thought, in part, it might be due to locations of education programs or significant groups of entities that had completed Phase Three.
Dr. Fitzmaurice reminded everyone that they first wanted to satisfy the Congressional requirement and look for guidance from the Committee as to what additional items might be needed. Secondly, he asked what other analysis could be accomplished.
Mr. Blair asked if they were able to look at a provider sector (e.g., hospitals) where a fairly good number reported they’d gone through Phase Three and were ready. He said it’d be instructive for those still going through the process to know their time frames for testing and training. Members suggested they do the analysis as days it took by entity type for each phase of the process, determine whether it fell within the Subcommittee’s six-month boundary for testing, and how long it took to do each phase.
Dr. Zubeldia said he was trying to resolve a paradox: they were hearing that clearinghouses were three times more likely than providers to get done with Phase Three, yet 33 percent of providers were waiting for their clearinghouse. Dr. Steindel suggested the reason might be the limited number of clearinghouses and that providers were waiting on people handling higher volumes of transactions. Mr. Scanlon said the basic issue was not having good control numbers. They had estimates of the total number of covered entities, which they’d published in their impact analysis, but Mr. Scanlon said no one knew exact numbers and they were unable to make causal inferences. He suggested they compare the number of people requesting an extension in each major category of covered entities with the total numbers published in their regulations. He noted that 8,000 hospitals appeared to have filed, which was more than 75 percent and, virtually, might be all of them, but they hadn’t received anywhere near half of the 700,000 physicians practicing in the United States.
Dr. Steindel reported he’d done a calculation on respondents who’d indicated the date they would finish testing; 3.9 percent gave a day beyond October 31. Dr. Zubeldia noted that some people indicated ten years beyond that date.
Dr. Zubeldia moved on to what the Committee could do to assist the industry in complying with the rules without spending ten years in testing. Before the addenda were published, 46 percent of the respondents waited for clarification on the standards or standards information. Dr. Zubeldia noted that the types of interventions the Committee could recommend, based on these data, was limited. Dr. Cohn said their prime agenda was to get a final report and post it on their Web site. Second, he thought education was valuable. He suggested posting the short informational and educational documents the Subcommittee proposed developing with CMS and WEDI’s white papers on the NCVHS Web site or creating a links to the CMS and WEDI sites. He requested that everyone review the draft documents over the next ten days and send changes and anything that needed to be explored to Dr. Fitzmaurice and Dr. Steindel so they could get the document posted on the Web site.
Dr. Cohn asked what kind of approval process the draft documents should go through. Ms. Greenberg said the Subcommittee should review it first. Dr. Cohn gave them a week for review. Ms. Greenberg suggested they then circulate it to the full Committee, giving them a short time as well. A formal approval process wasn’t necessary, and the Subcommittee planned to do this by e-mail. They will then post it on the NCVHS Web site along with suggestions and links. The Subcommittee would tell the full Committee what they were doing and ask for suggestions for links and useful materials.
Dr. Cohn noted the document didn’t have recommendations. However, he said if there were things that needed to be further analyzed or errors in the report, they needed to tend to them.
Ms. Greenberg didn’t think it would be fruitful for the Committee to develop materials when CMS and WEDI already were developing them. She proposed that the members review their documents looking for gaps, as the Committee did with privacy. She asked them to look for audiences or topics that hadn’t been addressed and consider them as suggestions to CMS or WEDI.
Ms. Trudel suggested adding a paragraph to the section about analysis discussing how some reasons weren’t amenable to the Committee’s solutions. Instead, they primarily spoke to outreach and a need for materials. The Committee felt that highlighting several resources that spoke to these shortcomings and providing links to particular websites was the best way to meet the intent of the law. She agreed to supply the Web sites and draft language for the paragraph.
Status of ICD-10 Cost/Impact Study
Donna Pickett, NCVHS
Ms. Pickett reported that a contract had been signed and background information (testimonies provided to the Subcommittee on ICD-10-CM and ICD-10-PCS, letters, transcripts from Web sites) had been forwarded to RAND. She emphasized the intent to put this back on track and expedite the process. Ms. Pickett said she expected to hear the next day about RAND’s availability for the kick-off meeting and would contact Dr. Cohn about setting up their conference call. Ms. Pickett and he agreed that the Subcommittee probably could review results in the May Subcommittee meeting and bring it to the full Committee in June. Based on what they’d seen, they anticipated it would be reviewed in August and go to a September full Committee meeting. Ms. Pickett said another reason to expedite getting documentation was to assist in identifying how the timeframe should be amended, since obviously it was different than what was originally proposed.
Public Testimony
· Tony Martinez, Martinez, Bass and Associates
Mr. Martinez reported that the preliminary registration data for the demonstration project authorized by the Secretary was compelling. FIHC and Alternative Link received over 8,223 raw registrations and additional faxes and hard copy registrations. Registered entities included over 146 insurance corporations and conventional health plans, 43 HMOs, 12 state Medicaid programs, 129 governmental agencies, 35 integrative delivery networks, 131 hospital systems, 48 PPOs and specialty PPOs, over 58 IPAs, 16 leading medical schools, 961 academics, 24 third-party administrators, and 111 application developers and service providers.
Mr. Martinez expressed Alternative Link’s and FIHC’s pleasure with this wide-breadth of interest in the ABC codes, noting it sent a message about the need to look at how the HIPAA process, coding, and the full range of healthcare transactions were being developed. Mr. Martinez said it was interesting to hear that Alternative Link and FIHC encountered the same challenges that the Committee reported regarding entities in terms of compliance with HIPAA. Alternative Link and FIHC were working to get everybody who should be involved with the project on board and registered. Mr. Martinez said they looked forward to sharing their findings with the Committee and assisting with the overall process.
Discussion
Mr. Martinez clarified that about 8,300 people registered on-line. They were still tabulating their fax and mail registrations. The waiver was for the demonstration project. They would have all those people and data over the next two years.
Dr. Cohn noted the next hearing was May 21-22. The first day was devoted to issues related to PMRI terminologies. The second day was originally reserved for ICD-10, but that final report wasn’t in and the agenda for then and the April meeting remained open. Mrs. Squire will e-mail later this week querying on hearing dates for August, October, and December. Dr. Cohn adjourned the meeting at 12:32 p.m.
I hereby certify that, to the best of my knowledge, the foregoing
summary of minutes is accurate and complete.
/s/ 3/3/04
_________________________________________________
Chair Date