The Subcommittee on Health Data Needs, Standards and Security of the National Committee on Vital and Health Statistics was convened on Wednesday afternoon, November 5, in the Hubert H. Humphrey Building in Washington, D.C. The meeting was open to the public. Present:
Barbara Starfield, M.D., M.P.H., Subcommittee Chair
John R.
Lumpkin, M.D., M.P.H., Clinical and Administrative Standards Work Group
Chair
Jeffrey Blair
Simon P. Cohn, M.D., M.P.H.
Kathryn L.
Coltin, M.P.H.
Kathleen Frawley, J.D.
Clement McDonald, M.D.
Marjorie Greenberg, Executive Secretary, National Center for Health
Statistics (NCHS)
James Scanlon, Executive Staff Director, DHHS/ASPE
Bob Moore, HCFA liaison
Judy Ball, Ph.D., ASPE
Gail Janes,
Ph.D., CDC
David Smith, HCFA
(Note: Documents pertaining to this meeting can be found on the NCVHS web page at HTTP://ASPE.OS.DHHS.GOV/NCVHS/)
The Subcommittee met in break-out session on November 5. They were briefed on the first annual report to Congress, on work on claims attachments and on data standards ownership and maintenance, and on the status of the Department's proposed rulemaking. In addition to discussing these topics, they discussed the computerized patient record and plans for advising the Department on the unique personal identifier and holding several hearings.
The Subcommittee was briefed by David Smith of HCFA on the draft of the Committee's first report to Congress on implementation of administrative simplification, in conformance with a charge in HIPAA to report annually. The report was prepared by Mr. Smith and Ms. Ball, based on an outline approved by the Subcommittee at its September meeting. It reviews the Department's implementation strategy and highlights special concerns of the Committee. Mr. Smith asked for comments from members by e-mail by November 26.
Members were generally positive about the draft, and offered a few suggestions. Mr. Blair proposed eliciting from major vendors statements about their intent to use the Committee's recommendations and the anticipated impact. Others thought this would make more sense in another year, after the rules have been out. Ms. Greenberg suggested mentioning the Subcommittee on Population Specific Issues and its work on administrative simplification at the June 3-4, 1997 hearing. The Subcommittee agreed with this, but suggested that information on general Committee activities be published separately in the annual report to the HHS Committee Management Office. Both this and the annual report to Congress on HIPAA implementation will be posted on the NCVHS Web site, but only the latter will be sent to Congress. Dr. Lumpkin suggested deferring further discussion of ways to apprise Congress and others of the full scope of the Committee's activities to an Executive Subcommittee meeting.
The Subcommittee then passed a motion adopting in principle the draft annual report on administrative simplification and authorizing the Chair of the NCVHS to review and give final approval to the final draft, which will reflect today's discussion and future comments from members.
Steven Barr of HCFA briefed the Subcommittee on an effort to automate attachments. He leads a group whose charter is to try and decipher the meaning of the phrase "automated transactions" in HIPAA and the distinctions between attachments and claims. He noted that many surveys have basically failed in efforts to determine the status of attachments, because of poor response rates. The general approach of the resulting reports has been to recommend automating attachments.
The working group started by separating the "front-end question" from the contractor to the provider and the response from the provider. It established an ANSI X12 277 transaction as an optional electronic vehicle for the request from the payor. Then in 1995 it started work on automating the responses. The passage of HIPAA accelerated this process. The team developed a method for using the ANSI X12 275 transaction as the vehicle for electronic responses from providers to payors.
Dr. McDonald has worked on this project as a member of HL7 and has helped use HL7 terminology to bring back clinical and laboratory information. Last week, the 275 identifier data maintenance was completed and sent to Technical Assessment Group at X12. The team is working with HL7 to develop a unique messaging language that can be used to define attachments. Mr. Barr explained that questions from contractors to providers can be asked in at least 30 ways. As a result, much of the initial communication between them is about clarifying the request. A messaging standard will simplify that stage. A LOINC code identifies the message being requested. A separate code list will be developed for the attachments that will be automated. The idea is to put the coding information outside the control of the standard group so it can be updated rapidly.
Dr. McDonald offered a few more explanatory comments and described several typical payor requests and the great complexity and heterogeneity of the situation at present. He said the structure being developed is a spreadsheet that defines named attachments that can be easily adapted. The goal is to be able to map between "the lumpy granularity of the current attachment definitions" and the specificity of a pharmacy or lab system. Many attachments are open-ended questions, and part of the effort is to categorize these so they can be automated. The many variants would be made explicit, and their number will be reduced.
Dr. Starfield asked how much effort on this was warranted in view of the fact that "this is a way station toward the computerized patient record." Dr. McDonald responded that if this aspect is not straightened out, the medical record will "solidify in the wrong way."
Dr. Cohn noted that this transaction has been used as an envelope, the contents of which have depended on idiosyncratic definitions worked out between trading partners. He asked if the new framework would be extensible. Dr. McDonald said the envelope would send HL7 messages. He added that the process has three-stages: to partition things that were not in the bill, to make formal labels for the contents, and to "explode them into what may be many clinical labels." This provides a mechanism for electronic transport.
Dr. Lumpkin wondered whether it made sense to concentrate initially on automating the front-end request from the payor and to defer looking at the response. Dr. McDonald said it is better to look at both directions at once. Mr. Barr said the Department is holding a meeting in Chicago in mid-November of some 70 providers and payors, with the goal of organizing a group to work on standardizing the messaging from the contractor to the provider. In response to a question from Mr. Blair, he said this was being done independent of the work of HL7 and ASC X12, but in interaction with them. In addition, five Medicare contractors and five commercial contractors have agreed to test the proposed standard messages over a three-to-six month period beginning around February 1998. The target date for issuing an NPRM is August 1998.
Dr. Starfield turned the discussion to the Subcommittee's role and next steps. Ms. Greenberg noted that the team has developed 19 questions that the Committee might use in the forthcoming hearings on attachments. Mr. Barr asked the Committee for guidance and suggested that it use the questions to elicit views from stakeholders. He noted that attention has not yet focused on attachments, and that the Department and industry will not only be busy with the first nine regulations but will have to deal with attachments, which are "twice as complicated and laborious as those nine combined."
Mr. Blair asked if an 80-20 rule could be applied to discern a reasonable number of questions that would cover 80 percent of the situations and be a goal for the first year. Mr. Barr said the Department realizes this will be an iterative process. Dr. Lumpkin and Dr. McDonald again exchanged views on whether to carve out and standardize the payor-to-provider request. Dr. McDonald said the problem is making the transaction "crisp enough" that it can be done in either direction. Mr. Barr added that the Department needs input on the 275 (provider response) by March, and the 277 (payor request) is already in place (although not mandated).
Dr. Lumpkin stressed the need to structure the hearing and give people something to react to. He also noted that the Committee's role seems to be to determine what the payors want. Ms. Greenberg asked what the Committee wants to accomplish at the hearing.
Mr. Barr noted that some people raise questions about the need for attachments. For example, the NUBC is concerned that automating them simply makes it more convenient to ask more questions. The underlying issue is whether attachment information is needed in the presence of standardized data sets. Mr. Moore said HCFA has convened a work group to look at the question of what payors want to know.
Returning to the 80-20 question raised by Mr. Blair, Dr. McDonald said there are 28 data sets or batteries, which are basically clinical. Ten to twenty attachments provide 80 percent of the information. Mr. Barr said his team needs guidance on where to go with claims attachments. It plans to use the aforementioned 28 to start with. The paramount question for a hearing, he said, is what is an attachment and what is a claim?
Dr. Lumpkin proposed that the Subcommittee state its views on this question as something for people to respond to. Dr. McDonald added that the Committee also should articulate the goal of reducing attachments. Mr. Moore observed that the reason for attachments is that payors don't trust providers, and in fact, this atmosphere is creating a trend toward even more attachments. What is needed is a vehicle by which payors feel comfortable about their payments.
Ms. Greenberg noted that "the HIPAA mantra" during the last year has been to keep the bill simple and use attachments as a vehicle for other desired information. Accordingly, many groups including public health ones are waiting to use attachments for data that have been removed from claims. She stressed that public health people need to be invited to the hearing as well, or to a subsequent one.
Dr. McDonald asserted that attachments should specify things that support claims. He asked for the Committee's perspective on encouraging splitting out clinical data from the claim into an attachment. The group deferred addressing this question.
The hearing will be held February 9-10 in Washington, D.C. The Subcommittee agreed to provide advance materials to participants in the form of background documents and a set of questions. The 19 questions will be circulated to members for comments and additions. Lead staff will coordinate invitations.
Finally, Mr. Moore informed the Subcommittee that the current Data Council work group will be reconstituted as an implementation team focusing on attachments. Mr. Barr introduced Maria Ward and Derwin Day, both from HCFA, who are also on the team.
Ms. Ball introduced Bobbie Redding of HCFA to report on discussions at X12 and other meetings. Ms. Redding said the general view is that data content maintenance should remain in the private sector. Content has to date been governed by NUCC, NUBC, ADA and NCPDP. Most have managed both paper and electronic formats, and in many people's opinion they have the best expertise in the country on data content. It is thought that the people who understand claims and remittance advice could manage the data content of all HIPAA transactions exchanged between plans and providers. She noted that health care premium payment and enrollment are in a different category. She asserted that the Committee should welcome the work of these organizations, which would probably be willing to expand their membership if so advised.
Ms. Redding added that a special X12-N task group is managing the coordination of implementation guides. It is thought that the committees with expertise on data content and the work group for the implementation guides could together make a coordination body. This would assure that like structures would be used in all transactions whenever data content changes. She and Dr. McDonald then discussed the difficulties of keeping up with changing codes and the best structure for handling them. Dr. McDonald advocated exploring a way to externalize the codes and possibly have them in a publicly available code system or repository. Mr. Mays commented that agreement is possible at the concept level, even though data elements and codes are contextually specific. The data dictionary is intended to build a meta model on top of the elements, because it is not possible to get agreement on the definitions and use of each element.
Dr. Cohn stated that the activity to date on the data dictionary is not adequate. What is needed is both a data model that indicates the definition and the permitted values, and a process to reduce unnecessary variation. He asserted that it is the Committee's responsibility to work on this. He supported Dr. McDonald's comments, and noted that in addition to the code and definition issues around data there is the issue of syntax. Ms. Greenberg agreed with Drs. Cohn and McDonald about the need to standardize those things that are "not standardized for no good reason."
Ms. Redding suggested returning to the topic of the decision-making process, and noted the desirability of a consensus process involving payors and providers who understand the data structures. Dr. Lumpkin suggested focusing on the data content issues that have the greatest impact on providers. Dr. Cohn reminded his colleagues that the purpose of this discussion is to plan hearings at which these issues will be discussed. He said the hearings need to address, among other things, how the process has worked so far and how the participants think it should work.
It was noted that the organizations that create the content of the transactions do not have the same open consensus building process required by ANSI and favored by HIPAA. This person [unidentified] wondered whether the organizations should be encouraged to implement an ANSI-style consensus process. Ms. Redding pointed out that while the ANSI process has merit, it can also lead to multiple solutions and people can "buy votes."
Dr. Lumpkin questioned whether it might be confusing to conduct hearings on how to change the standards just as the rules are being released. The group agreed with his suggestion to discuss this issue again at the March meeting and to decide on a hearing date at that time.
Ms. Karen Trudel, Co-Chair of the Data Council's Health Data Standards Committee, reported that although no NPRMs have been published, most of the work on them is completed. She noted that "it is a massive undertaking."
Mr. Scanlon reviewed the Data Council's discussion of the unique identifier issue, the Committee's recommendation that none be recommended until privacy legislation is enacted, and various options available to the Secretary. One option is to publish a Notice of Intent, initiating a participatory process that would presumably lead to an NPRM, but with more input along the way from stakeholders. Such a notice would probably be published after the NPRMs. The Data Council hopes the Committee will hold hearings following publication of the Notice of Intent. It wants to have a better sense of the costs and scale of implementation of various identifier options. He agreed with Dr. Starfield that in addition, it needs to understand the current baseline and what States are using now. He cited a recent Social Security Administration study of the costs of creating a new social security card, which gives an idea of what a health identifier might cost.
Dr. Lumpkin observed that the Committee has a unique opportunity to help the Department by holding hearings. It should move beyond the work the Committee has already done on this issue, with the goal of identifying one or two identifiers to recommend to the Secretary. Ms. Trudel noted that the Identifiers Implementation Team has started work on the types of information that would go into a preliminary document from the Department on this topic. It is desirable to hold the hearings after publication of this document, to focus the discussion. The group discussed timing issues at some length.
Subcommittee members agreed to hold the hearing in Chicago in May, and to coordinate with the other Subcommittees. The hearing will probably be hosted by the Executive Subcommittee on behalf of the entire Committee. If the Notice of Intent has not been published in time, the hearings will be postponed. Another hearing on the unique identifier will probably be held in Washington.
The Subcommittee agreed to recommend that the March meeting of the full Committee include a panel on the computerized patient record, to bring everyone up to speed. In addition, the Subcommittee will study an IOM report that was recently re-published with an update.
The Subcommittee confirmed Ms. Greenberg's recommendation that the 60-day comment period following publication of the NPRMs not be compressed.
Finally, the group briefly discussed the core data element recommendations. Ms. Greenberg reported that the staff did a crosswalk between the data dictionary and the core data elements and found that the vast majority of those recommended by the Committee as ready are included. As it was impossible to find the elements in the dictionary itself, it was necessary to go into the implementation guide to find them.
Ms. Greenberg called special attention to the fact that external cause of injury is not included in the professional claim. In addition, these elements were not found:
These missing elements have been identified to HCFA, and the Data Council has discussed whether they should be added prior to release of the NPRMs. Another option is to try and get them into the appropriate standards over the next year. On the elements on which the Committee recommended work (e.g., functional status), Ms. Greenberg noted that the Data Council has not initiated a process for advancing them.
Dr. Starfield commented that she is getting phone calls saying that the Committee is not doing its job and that information such as birth weight and social security number are being lost. Mr. Moore said that of the 42 items recommended by the Committee (including several identified as not ready), 35 have been included in the standards and others are under consideration. Adding some items is "an administrative nightmare," and some may be revisited in a few months once more definitive code sets are identified.
Ms. Greenberg pointed out that the added value of the Committee's recommendations lies in the ones not already in the uniform claim. She expressed particular concern about external cause of injury because of the past effort to reach agreement with NUBC on this and because violence and injury are recognized as major public health issues. Dr. Starfield observed that the Committee needs to be able to report to its constituency about the status of and prospects for its core data elements. Ms. Greenberg acknowledged that a great deal has been accomplished in a very short time; the point is to make sure that there is a plan for accomplishing more.
Dr. Cohn observed that the aforementioned issues point again to the need for a useful data dictionary, as well as to issues of ownership and maintenance.
Ms. Greenberg reported that several agencies are planning a workshop for public health professionals, to raise their awareness about HIPAA. This is part of the larger public education and outreach effort.
Dr. Starfield then adjourned the meeting.
I hereby certify that, to the best of my knowledge, the
foregoing summary of minutes is accurate and complete.
/s/ Barbara Starfield June 10, 1998
________________________________
Chair Date