Hubert H. Humphrey Building
Room 303A-339A
200
Independence Avenue, S.W.
Washington, D.C.
Barbara Starfield, Chair
Dave Smith
Steven Barr
Karen Trudel
Jim
Scanlon
Marjorie Greenberg
Clement J. McDonald
Kathleen Fawley
Jeff
Blair
John R. Lumpkin
Robert Moore
Kathryn L. Coltin
Simon P.
Cohn
Report on First Annual Report to Congress: Dave Smith
Report on Claims Attachments: Steven Barr
Report on Ownership and Maintenance of Data Standards
Report by Dr. Karen Trudel
Report on Notice of Intent on the Unique Identifier by Jim Scanlon
Discussion on the Computerized Patient Record
Report on Core Data Elements: Marjorie Greenberg
DR. STARFIELD: The Subcommittee on Health Data Needs, Standards and Security is now in session. I think every member of the subcommittee is here that is going to be here.
We have a lot of things to talk about, and I suspect that some of you may even add more to the agenda. So what we will do now is review the agenda; it is in Tab B of the book. Could you look it over and tell us what could be deleted or what needs to be added?
One of the things that -- just to start the discussion -- that isn't easily identifiable is a plan of action for what we do with the unique identifier, the individual identifier. We need to start some discussion on developing a plan of action for that. I wondered whether you thought there ought to be a special place for it.
I know, Jim, you are going to report on that, but that is a half an hour report on -- do you think that is enough? Do you want to move it up, maybe, because it is so key maybe to some of the other things?
PARTICIPANT: We can move it up. Do you want the panel first, or do you want to go right ahead?
DR. STARFIELD: The panel? I'll tell you what, why don't we discuss the annual report first, because I think you are ready to do that, David. John is suggesting we do it after the 2:30 session.
PARTICIPANT: Add it into the ownership and maintenance of data standard plan for hearings, and maybe we can roll it in there and push some of this stuff back.
DR. STARFIELD: Is that okay with everybody, to stick it in there? Okay.
So let's hear about the draft of the first annual report to Congress. David?
DR. SMITH: Good. Thank you, Dr. Starfield. My name is Dave Smith. I work with the Health Care Financing Administration in the Office of Information Services and Information Technology Investment Management Group. That is a mouthful.
It is an honor for me to be able to address the Subcommittee on Health Data Needs, Standards and Security. I appreciate this opportunity.
The topic I am going to discuss today is the draft of the first annual report to Congress on the status of the implementation of the administrative simplification provisions of the Health Insurance Portability and Accountability Act of 1996, HIPAA.
Section 263 of HIPAA amended Section 306K of the Public Health Service Act, by adding at the end of that section the following language pertaining to the National Committee's annual reporting requirement. Not later than one year after the data of enactment of the Health Insurance Portability and Accountability Act of 1996, and annually thereafter, the committee shall submit to the Congress and make public a report regarding the implementation of Part C of Title 11 of the Social Security Act.
Such report shall address the following subjects to the extent that the committee determines appropriate. A, the extent to which persons are required to comply with Part C of Title 11 of the Social Security Act or cooperating or implementing the standards under such part. B, the extent to which such entities are meeting security standards adopted under such part and the types of penalties assessed for non-compliance with such standards. C, whether the federal and state governments are receiving information of sufficient quality to meet the responsibilities under such part. D, any problems that exist with respect to implementation of such part, and E, the extent to which timetables under such part are being met.
At its September meeting, the National Committee approved an outline for the annual report, and agreed that the first report would provide an overview and commentary on the implementation strategy, followed by the Department of Health and Human Services over the past year, and would highlight issues of special concern to the committee, such as privacy and security.
I prepared the draft that you have in front of you based on the outline that was approved at the September meeting, with just a few minor modifications. But I didn't exclude anything that was indicated as being of interest to the National Committee. Judy Ball also provided considerable assistance in the development of the report. The draft incudes comments I received from the Health Data Standards Committee. If the subcommittee agrees, I would appreciate any comments from the subcommittee that the members may have by Wednesday, November 26.
I would like now to go over briefly the outline that I used to prepare the report.
First of all, there is an executive summary, which talks about basically what the report is all about, and summarizes the content. I have an introduction, which includes the background, purpose of the report, the content of the report, the requirements of the statute, which include the requirements for standards, the timetables that have to be met, and the expanded responsibilities of the NCVHS.
In the next section is the implementation process that was actually put in place to get these standards implemented. Part of that is the Department of Health and Human Services implementation strategy, including the establishment of the Health Data Standards Committee, which is part of the Data Council, and the implementation teams, and the use of the Data Council as the overall coordinating arm for this effort in the department.
Then I discuss the guiding principles for the selection of the standards and the private sector contributions to selecting the standards, discuss the results of the NCVHS hearings, and the NCVHS liaison with the Department of Health and Human Services, and finally, I listed the recommendations that were made by the NCVHS to the department relative to various standards that are under consideration.
In the next section, I talked about the progress to date on the identifiers, transaction statements and data content, security, claims attachments, privacy, and the implementation plan and communications strategy that is being developed.
The next section talks about the special privacy and security concerns that were expressed by the National Committee, including the need for federal privacy legislation, linkage of the individual identifier to privacy protections, the fact that privacy protections need to be two-tiered, and the security concerns of the National Committee.
The next to last section deals with implementation issues, discussing how to identify and resolve implementation issues, identifying the need for new standards, measuring standards implementation status.
The final section was a brief conclusion to the report.
I would be happy now to take any questions or comments you may have at this time on the report itself.
DR. STARFIELD: You all should have that document. David, I have a question on the privacy protections that need to be two-tiered, on page 24. The second one-sentence paragraph of that talks about the potential for legal requirement that health information be used only for the purposes for which it was collected.
That strikes me as rather vague. You got it?
DR. SMITH: No, I don't.
DR. STARFIELD: Page 24.
DR. SMITH: It is under C, and it is the --
DR. STARFIELD: Under C, second paragraph.
DR. SMITH: Okay.
DR. STARFIELD: That says a mouthful. It basically says no linkage of data. Do we really mean that? Did we say that?
DR. SMITH: That was in the information that was provided in the recommendations made by the National Committee on privacy. That is where I got that information from, or that particular language. It was actually in the report itself that was provided to the Secretary of HHS.
DR. STARFIELD: I suppose that could be interpreted very broadly, as to what you mean by purposes for which it was collected. But it also could be interpreted very narrowly. To me, that leads to a tremendous ambiguity.
DR. MOORE: Actually, Barbara, that concept was both in the NCVHS recommendation to the department and the department's recommendation to the Congress. It is the notion that in essence, the identifying information should only be used for health purposes and something recently related to the original function.
DR. STARFIELD: Oh, but that is identifiable information. This doesn't say identifiable information here.
DR. MOORE: We should clear that up.
DR. STARFIELD: I'm sorry, I think that is an important qualification.
DR. SMITH: Thank you, I appreciate that comment.
DR. LUMPKIN: Just one comment on the security section. There is really no mention of the vendor community. I think we should put them in. I think we heard some testimony related to -- even though in the writing of the software, they disable a lot of the security mechanisms that are built into the operating systems, but there is no interest on the part of the purchasers to do that. But the vendors are going to need to be involved in any solution to data security.
DR. SMITH: I'll include that. Thank you.
DR. STARFIELD: Kathy.
DR. COLTIN: I had a comment on page 13. You list the dates of the committee and subcommittee meetings, and it seems to me that it is less important to perhaps know the specific dates than to know the topics that were addressed, and the stakeholders from whom we heard.
So I thought it might be helpful to really just have what were the topics We had hearings on coding. We had hearings on security, et cetera. Then the variety of stakeholder groups that were represented in those hearings, to assure them that we tried to cover many viewpoints.
MS. GREENBERG: On that point, I think it is actually here in the agenda book, the fact sheets that we prepared, that the public education group did. We did identify the topic. You can just take it from there. I can get it to you in electronic form.
DR. SMITH: I appreciate it, thank you.
DR. SWARTZ: Could that be made as an addendum or an attachment to the report, rather than part of the body of the report, so that it doesn't get so -- because if you start listing all of the people who were in attendance, and the issues and the topics that were identified with all the different meetings.
DR. SMITH: Well, in fact, the detailed agenda approach --
MS. GREENBERG: I was thinking of just the broad headings, providers, payors, vendors. You know what I mean? That it can be done fairly concisely.
DR. SMITH: Yes. Thank you.
DR. BLAIR: I wish I had had a chance to have read it through completely; I had to skim through it. I thought it was a tremendous compilation of everything that we had done in the committee, as it was reported and documented.
I'm not sure what is exactly appropriate here, but do you think it would be of value, or do other committee members feel it would be of value for just a little bit of effort to be done to maybe contact some of the major health care information system vendors, to indicate their intent to take the recommendations that are coming out of the committee, which of course then winds up being adopted as standards by the Department of Health and Human Services? By taking those things and integrating into their systems, and if they have a positive feeling for the way that it has been done, or a positive feeling for the fact that it is going to help improve their efficiency and effectiveness, I would think that that -- from the standpoint of Congress, I think that could really pin down the fact that we are doing work that the private sector finds of value.
DR. SMITH: Would it be appropriate to do that prior to the actual regulations being published?
DR. BLAIR: At this point, since it happens to be a report on what we have done so far this year, what I was thinking of is just their expectations. A lot of them have testified and a lot of them have been monitoring what we have recommended. It has been published on the Internet and in magazines and stuff. So anyway, I thought it might be an addition that may be of value.
DR. SMITH: I'd like to hear from the rest of the subcommittee about that, if that is a good idea.
DR. BALL: Would you have a suggestion for the list of vendors that should be contacted?
DR. BLAIR: I could provide a quick list, sure. Do you want me to just rattle some off? Just the major vendors in the --
DR. LUMPKIN: I think it is something that we may want to do perhaps for our next one, after the proposed rule has been out and has been fairly well discussed, and we know what is going to happen. So I would look at maybe doing this a year from now, as well as holding a hearing on that particular subject towards the fall, in preparation for our annual report.
I'm just a little bit concerned that we are going to be asking people -- they are going to feel like we are asking them for a commitment on doing something that they don't even know what that something is yet, officially.
DR. MOORE: And it involves more than vendors, too, obviously. But Congress has already said, this is what we believe needs to be done, go forth and do it. I don't think we need to convince them anymore. I think now we just have to get it done.
Then I think as part of the annual reports, we could talk about how well it has been done, what the feelings are in the community, and what some of the barriers are.
Dave, I wanted to ask, is there any major requirement that was in the legislation that we have not addressed for our annual report, that we have had to leave out because it is the first report?
DR. SMITH: Nothing has been implemented yet. Most of the requirement is to do reporting of what the status is of the implementation, and if there are any bars to implementation, and we would have to indicate what those are, and anything that is causing it not to happen in a timely fashion, if there has been adequate cooperation on the part of the community or in federal and state governments.
DR. LUMPKIN: I wonder if the drafter of this bill would have taken that into account.
DR. STARFIELD: I guess it is important to get across the notion that we are on schedule, I think. I don't remember whether it says that, but I think we are on schedule, is that right?
DR. MOORE: The committee is on schedule, that's for sure. I think you finessed this very nicely, Dave, in terms of the NPRMs we were in there to change. You did identify a couple -- even at this preliminary stage, you did identify a couple of issues that would be potential implementation issues, or potential implementation barriers, at least two or three, I think.
DR. SMITH: One of the things that I do want to mention, in the event that the subcommittee members didn't realize what I was doing. There are many cases where I am saying that something was published, but it hasn't happened yet.
The reason for that is that we are going to wait until the actual date of getting this report out, and then incorporate whatever has been published as of that date. So it will be up to the minute, if that is in December, or whenever that is. Right now, it is redlined in the information that you have.
DR. STARFIELD: Marjorie.
MS. GREENBERG: I sent some e-mail comments to Dave and Judy, and I think they have done an excellent job. I feel particularly in their debt, having written annual reports for this committee more years than I would like to remember, I know it is a big job, and I am appreciative.
I just was going to mention one of them here, just to see if anybody has any problem with it. I thought that even though this is a HIPAA related annual report, I thought that we should mention the Subcommittee on Population Specific Issues, because they have actually been part of the HIPAA process, including sponsoring the June 3 and 4 meeting, and mention that the purpose of that meeting in San Francisco, in addition to getting out and getting views from another part of the country, was to hear particularly about the impact that there might be on specific populations.
I assume nobody has any problem with that. Particularly that June 3 and 4 hearing, because it was very oriented towards the administrative simplification. I don't know whether we would want to include anything in this report about that subcommittee's particular -- well, they worked on the racial and ethnic standard, and of course that is part of the standard process also, and I don't know whether that belongs in here or not.
Also, their next year's agenda that they are working on now is looking at data needs and requirements for Medicaid managed care, again looking at the standardization aspects of that. I don't know whether that is going a little out of sync or not.
But I think that by including some of that, it does show that the committee is obviously trying to meet all of its HIPAA specific requirements, but looks at standards in a broad sense, and that the application of these standards is also -- has many different -- will take many different forms.
DR. STARFIELD: Marjorie, could you clarify? There is a separate annual report.
MS. GREENBERG: Well, let me mention that. There really is not anymore a separate annual report. You haven't even noticed that you didn't have to write anything?
The new charter for the National Committee, going on two years old now, so we are going to have to renew it, so it is going to have to be rechartered -- you can't hear me? I'm sorry. When the committee was rechartered, the requirement for the old annual report as we loved and hated it was removed.
There is a departmental requirement for an annual report to the committee management office that is primarily financial, et cetera, but does include -- and it will be passed out to you tomorrow -- it does include what is really the equivalent of the executive summary of the old annual report, where it is a bullet on all the committees' and subcommittees' accomplishments.
Actually, in talking with Don and Jim the other day, we thought it would be good when this report is finalized and we put it up on the website, that we also put up that executive summary to capture the full range of activities. And of course, many of these activities in this report will also be in those bullets. We have copies of that.
Then the question is -- we could also send that to Congress. We always used to send our annual report to Congress, actually, too.
Then just something for you to think about is whether that is sufficient, or whether -- I am not myself in favor, probably for selfish reasons, of going back to the old annual report. Yet, it does concern me somewhat. I think we can discuss this at the executive subcommittee meeting. But from 1949 on, there has been annual reports of the National Committee, and all of a sudden, 1995 is the last one. Then there will be this annual report to Congress on HIPAA.
As we start talking about quality today, et cetera, I don't know how all that is going to be captured I think when we talked about the recharter and separate reports on those issues.
So I think it is just something to think about in the future. But no, there is not really -- other than this executive summary that is done for the committee management office, there is not the other annual report anymore.
DR. STARFIELD: Your suggestion is really a way to get around the fact that we don't have a full annual report. Maybe you can put that in the executive summary. Otherwise, you are going to have to cull through everything to look for everything the committee has done with relationship to HIPAA, not only what is going on in the population, but the specific subcommittee.
DR. LUMPKIN: I agree with not discussing that here. I think it is something perhaps that we may want to do with the executive committee, and have a recommendation to come back. But certainly, there are enough important things that are done that aren't related to HIPAA, and ought to be, that a report is in order. And whether or not we attach those in appendix to the HIPAA report, oh, by the way, we also did this, so that is the justification for sending it to Congress.
But I think it is something that we need to look at, perhaps not for this year, but something that we would plan on doing for next year as we look at a more robust HIPAA report also.
MS. GREENBERG: On the short question, does anybody have a problem with adding in at least the relevant stuff about the Subcommittee on Population Specific Issues to this report? Okay.
DR. MOORE: I was going to make the point that I think the report to Congress is a little different than the other reports on HIPAA. I think we really ought to keep this fairly clean, other than the additions we just talked about, and get that going, and then I think we can talk about a summary of the committee's activities overall, and how we might want to disseminate that, to report to the Secretary.
MS. GREENBERG: Yes, that is what it was.
DR. STARFIELD: Now, I assume you want some action from us, that we can report back to the committee tomorrow on the fact that we approved this?
DR. SMITH: Yes, we can talk about that, I guess. But I would like to have comments by the 26th of November. Evidently there is some discussion going to be happening tomorrow about the full committee's action.
DR. STARFIELD: Well, no, I don't know.
DR. MOORE: I think Don wanted to -- first of all, Dave should get all the comments, in terms of whatever we have. But Don wanted to bring this up tomorrow in terms of -- if the full committee agrees, and you would want to report this out, Barbara, obviously -- if the full committee agrees in principle or in concept with the report, subject to the comments that others would support, that he may want to refer to the executive committee.
DR. STARFIELD: John, you had a comment?
PARTICIPANT: I'll refer to my distinguished colleague with the J.D. degree.
DR. FAWLEY: Oh, yes, right. Thanks. Dave, just to comment on page 13, on some of the dates of the hearings. We did have the hearings in California in June, and I just wanted to make sure. And we've got dates listed for hearings September 8 and 9, and actually, that was a full committee meeting. We didn't have any hearings in conjunction with that meeting.
MS. GREENBERG: The 8th, the subcommittee had a marathon breakout session. We heard on the unique identifier. I thought that one might qualify. The meeting on the 9th with the privacy I didn't think qualified as a hearing.
DR. FAWLEY: No. I guess I am thinking of hearings where we actually held formal hearings. I was just concerned about --
DR. STARFIELD: So we just have to clear up some of the date issues.
DR. LUMPKIN: I would like to move that we recommend to the full committee that we adopt in principle the annual report, and that we authorize the Chair to review the final draft for final approval.
DR. STARFIELD: So moved.
PARTICIPANT: Second.
DR. STARFIELD: Discussion?
DR. SMITH: Can I say one more thing? I have asked for comments, but I didn't tell you how to get them to me. My e-mail address is dsmith1 @ HCFA.gov.
DR. STARFIELD: You are the first HCFA person I have known to have a number. But I guess with a name like Smith, --
DR. SMITH: Yes, because there are so many D. Smiths, right.
DR. STARFIELD: Would you have three or four comments? I think three specific ones about changes to be made.
DR. SMITH: Yes, just cleaning up the dates, and the comments you mentioned earlier about the Subcommittee on Population Specific Issues, so I have those.
DR. STARFIELD: And you got that individually identifiable --
DR. SMITH: Excuse me?
DR. STARFIELD: The one on page 24.
DR. SMITH: Right, the one on page 24, right.
DR. BALL: We will consolidate our notes. We will probably get them all that way.
DR. LUMPKIN: Your due date was November 26?
DR. SMITH: The 26th, that's correct.
DR. STARFIELD: But we don't have to write you again about things --
DR. SMITH: Not about the ones that were discussed today, unless you have something more in amplification that you think is appropriate; by the way, I thought about such-and-such that I didn't think of during the hearing.
DR. STARFIELD: We have a motion on the table. It has been seconded. All in favor?
(Chorus of Ayes.)
DR. STARFIELD: Opposed? Abstentions? Thank you very much.
DR. SMITH: Thank you.
DR. STARFIELD: We will move on to the next item, which is the report from Steven Barr on claims attachments.
DR. BARR: Thank you, Dr. Starfield. I want to speak about the process we are doing right now in trying to automate attachments.
When HIPAA legislation was passed, there was clearly defined the identification of certain standards that had to be implemented: eligibility, claims status and all the transactions related to that. But there was a one-liner in there that came after that, which said automated transactions, which was the charter of the group that I am working with, which is to try to decipher what the word automated transactions was, two, to decipher what the word attachments really means, and take it away from claims.
We have gone through and made presentations. We have been told that people think that the HIPAA legislation identifying the first nine are the mountain of all transactions, then the attachment is the Mt. St. Helens of all transactions.
One of the problems we came across was, there had been survey after survey -- WETI has done a survey, the NUCC has done a survey, HCFA has done numerous surveys, and who else knows who else has done surveys, all trying to determine the status of attachments in this country -- commercial, Medicare, whatever it is.
Whenever they would try to do these surveys, one of the problems is, one, there is a lack of response from the people who are requested to participate, and two, the terminology of what an attachment becomes is vague, depending on who you ask the question to. Everybody who has done this eventually gets the documents together and keeps a running file and makes a report, and then they just go away quietly and say, we recommend the automated attachments.
One of the charters that we did was, we have tried to take that step to the final limit. The group that I am working with has done a two-pronged attack to this. One, we have tried to automate the front-end question, which is a question that comes from the contractor to the provider. We have done that electronically via a ANSE X12 and 277 transaction, which allows for an electronic request for information coming from the contractor to the provider.
Now, that doesn't mean you have to do it electronically, but now we have a vehicle available that will allow for this to be done.
We proved this out in a proof of concept study that HCFA chartered in 1995, and proved successful. While we were working on that, our next phase was, we thought we would try to slowly walk into automating the responses back from the provider to the contractor, and then HIPAA legislation got passed, which really escalated the process.
We in turn sent our study and our survey out to 378 insurance companies, and got a total of 25 or 30 responses back. But as unlucky as it is, I was also the NUCC person who did an attachment survey, so I had my old documentation, plus I had the new documentation, and I had two desks full of attachments.
What we have done is tried to develop a method where you can now respond back to the request that comes in via the 277, or by a paper request electronically. We have taken the ANSE X12 275 transaction as the vehicle that will respond back to the 277.
Now, the 275 is a patient record which allows for anything to be sent back, whether it is binary information which includes X-rays, EKGs, specific documentation or little bits of information.
In that 275, we have also worked with Clem McDonald at HL7, and Clem has been a part of our special interest group at HL7, to work together and to use HL7 terminology and orders and results terminology as a mechanism to bring back both clinical and laboratory information, and codified information back also.
So we really expanded the realm of what we are really doing in regards to automating attachments. We are going beyond the normal scope of what our charter was. People have said that is beyond comprehension, but we have gone that route.
Last week, we finally got the 275 identifier data maintenance done, and that data maintenance has been sent to Technical Assessment Group at X12, which means it is now in the hopper, so we can utilize the transaction for the standard that we proposed to utilize. We are working closely with HL7, and trying to develop a unique messaging language, is probably the word I should use, to define attachments.
I have to explain this to you, because you may get a little bit lost, because we got lost when Clement explained it to us.
In today's environment, there is a two-pronged problem in today's environment. One, the question that comes from the contractor to the provider can be asked at least 30 different ways at a minimum. You can ask for the same request, depending on what contractor you are, a different way. So therefore, the provider, who is at the mercy of the contractor, has to make a rough determination and hope he makes the right one about what is being requested of him. So there is that vagueness on the front end about what the message should be on the front end.
Then there is that message that comes back from the provider to the contractor: I hope I got this right, because this is my answer to your question.
I am going to take the back end first. What we have done is defined a messaging standard, where in today's environment -- I give an example. If you have a question regarding today, a contractor would say, what is the patient's weight, which is a generic terminology. We have taken that level back down to a more finite level and said, patient's weight is really a vague terminology, but do you really want patient's weight? Do you want patient's weight estimated, patient's weight measured, patient's weight stated?
We have now defined that weight into more definitive terminology, and defined those terminology requests into what is commonly referred to as a LOIN code, which is an acronym for logical observation identified name codes. This LOIN code is a five-position field which identifies the message that is being requested.
So for example, 1012-3 would be patient's weight estimated. So the contractor we hope would then take the patient's weight question, 1012-3, patient's weight estimated, send that request back, and then the provider would know exactly what they are looking for.
Now, in the earlier session, one of the things mentioned is, if you can visualize this, there would be a separate code list for whatever attachments we plan to automate on the first pass.
So the ambulance attachment list would have a whole group of identifying information per ambulance. It would be out of a normal standard, so that if somebody wanted to add to it or modify it, it would not be involved in the process of updating it; it would just add to the bottom. It allows for you to put this coding information outside of a control of the standard group. So you can update it as you need it.
If you need to refer to it, all you have to do is be a pointer, to say I now want to report to the ambulance code list, hold down the ambulance code list, and identify what all the codes are and what all the messaging is, in case you don't understand it. Those would be identified by LOIN coding.
Clem is doing this for approximately 28 -- Clem got involved in this, and really thought there were not that many attachments. So we got into it, and we got on the phone one day and he said, I've never seen so many attachments in my whole life. I said, Clem, you have only seen the tip of the iceberg.
That is the back-end messaging, and how we plan to do that. We also have the capability embedded in the HL7 structure, and have the mechanism to transport it, once we get approval. We can bring in the DICOM standard, which allows for mammograms, EKGs, X-rays and all that information electronically via the HL7 and the 275. So we are going on to have that capability.
Again, that is not an attachment per se. We can also bring in op notes.
DR. STARFIELD: Bring in what?
DR. BARR: Operative notes, the whole report. We don't want to bring in the whole op report, but if you have a need for it, we have a mechanism, a way to bring it in. We are looking for you not to request that, just request a finite piece of information. But in case you need it, we have a mechanism to bring that whole record in.
DR. MC DONALD: You did a great job on a very complex subject. The perspective of a naive clinician looking at this problem or seeing this problem, he sent me a stack, and I actually read through about 500 of them. An attachment is literally anything they can staple onto something, so it is just about anything on earth.
But I think it is important to categorize them. There are a number of them that are just poorly transmitted bills, and please send the ICD-9 code, which isn't really -- it is almost an error message. There is another kind that is an identifiable part of the chart. They actually have a thing that says, send me the op note. In fact, that is very, very common. Colleagues of mine who are in surgery say they just send it routinely, just so they don't get that message. Discharge summaries, cardiac echoes requested. There is an attachment that is related to end stage renal disease. They want to know the last creatinine, the last hemoglobin, the dose, and I think the last time it was changed, or something like that.
So what these typically are -- and then there are a number of them that have been formalized as what I would call flat records or data sets. How they do that traditionally, as I understand this business, you take the printed form and just start reading off the fields. You make this list, and that is the set.
That is the generic prototype. The generic prototype you think of as a list of elements or a list of variables. The structure we are talking about is a spreadsheet that defines the named attachments that can be easily adapted.
But it is not just a list of data elements, because in the billing term, it is often very vague, and there may be three different hemoglobins that you can pick out of a lab system. You may say, I won't accept any of these, so the list would include exploded.
The real goal is to get the granularity, be able to map between this lumpy granularity of the current attachment definitions to the specific that might be recorded in a pharmacy system or a lab system, so you can yank that out automatically.
Now, having said that, there is an awful lot in this attachment which no one has said well anywhere. I went through four of them and asked them to explain 50 variables. One of them was called CC and another one called PA. That is all there was, those two letters. And a bunch of other ones called date, maybe two of them called date. Which date?
So there is a lot of -- I'm sure somewhere in somebody's data system, they have got that better defined. But what I got in the spreadsheets are completely unfathomable. That is part of the work of this, is getting these things fathomed.
Now, there is a third class of attachments that is just an open-ended question, which is not likely to be a variable stored in the chart. Some easy ones were like, why did you do these two tests in these patients three days after they died? You can understand why you would ask that question. It was probably a date mis-recorded somewhere about something.
There are billions of those. I think billion may not be too far off, because every insurance company says them differently, and they find different ways of saying them.
Now, they have codified 480 specific questions or attachments. I refer to attachments more often as a battery of these things. So if we can get them organized and defined, I think maybe we can start to automate them. But there is a ton of work to be done.
DR. STARFIELD: The question is, how much of this work do we want to do? This is all a way station towards the computerized patient record, is it not?
DR. MC DONALD: It is very important. Because it is a way station, this is that (word lost) universe. You remember the first three picoseconds of the universe, we had anti-matter, matter, and it all solidified one way? If we don't get this right, the medical record is going to solidify in the wrong way. We are going to end up with all these vague date and PS and TA things, which we will never know what the heck they are.
DR. COHN: I am remembering the history of this particular transaction. I do remember when it was originally developed. It was just an envelope, and the only way you could make any sense out of it was that you had to have quote-unquote wiling trading partners who knew what CC meant or TA meant, or get some agreement on that. Obviously, you are trying to generalize it into something that really handles multiple pieces.
DR. MC DONALD: You've got two different pieces. TA and CC had nothing in 275 and it wasn't part of an implementation guide.
DR. COHN: No, no, I'm saying that this history of this standard was that there was never anything well enough to find on it, at least to my understanding.
DR. MC DONALD: Right, it was a way to send stuff.
DR. COHN: It was a way to send stuff, but all the information in it had to be defined between two partners.
DR. MC DONALD: I think it really wasn't used much.
DR. COHN: Right, that is why it wasn't used much. Now, I guess I like Clem's approach. Clem has a long history of biting off handlable pieces. Obviously, what you are trying to do is to structure the unstructured, hopefully not unstructurable, but unstructured.
I guess I am wondering, is this going to be an extensible framework? Or is this going to be something that falls apart almost as we get started. I think for some of the questions you were asking, it was absolutely fine. But I can see this as being the envelope that sends the medical record the way you are going, because everybody will be asking everything from one note --
DR. MC DONALD: What it is going to send inside it as HL7 messages, so it could indeed contain a lot of stuff. So inside the envelope there is going to be an HL7. If you asked for electrolytes -- I haven't seen one like that -- as one would get them in an HL7 message, because that is something the lab system can generate, wrap into this HL7 wrapper and send.
But I maybe mis-stated. Many of these attachments are well-structured documents, except that there are 40 different variants of those structures. The elements, at least when you boil them together to say this looks like the average attachment, need a lot more elaboration. I maybe overstated that. I think it is a matter of going back to the folks who contributed them and be clear what these things really are.
When we find out what they really are, some of them are things that are already in the bill, so they don't need to be in the attachment part. So we have a lot of three-stage processes. One is to partition out those things that weren't already in the bill when it came across, and then to make formal labels for those things that go in there, and then explode those into what may be many clinical labels.
DR. BARR: We don't want to repeat something that is already in existence in the current data set that is already coming in on the 837. We don't want to do that. We are not the vehicle that will promulgate the claim. That is not our intent.
All we are doing is in reality having a mechanism that, if there is data that is transported that would be needed, they do have the mechanism to transport it electronically, or by paper in today's environment.
What Clem was mentioning is that there are various -- for example, we talk in an ambulance attachment per se. That is currently in the 837, and there are data elements in the 837. When I did my sorting of the ambulance attachments through the two surveys that I had, I found 56 different versions of an ambulance attachment. I don't mean just 56 headers; there were 56 different versions, with each one having a different data element somewhere along the line.
Now, the question you ask is, why did somebody have to ask a unique question that was not in contractor A, but contractor B needed it? There is all this information that has to be identified somewhere along the line. The laborious part is trying to define all the data elements that we come across.
DR. COHN: Right, and I was just trying to make sure that you were coming up with a model that -- once again, you are chipping off little pieces. I just wanted to make sure that you don't do something that is not extensible beyond these little chips. I am still a little uncertain about what you're doing, but certainly it needs to be chipped off somewhere.
DR. MC DONALD: The alternatives are to take the raw ill-defined attachments and make people do chart reviews, and fill them in. So what we are really trying to do is make these more formally defined variables, many of which are, like hemoglobin. They ask for hemoglobin, they ask for creatinine, they ask for last dose of wheaten dose. Then recognize that in the clinical record, there may be more than one variant of these. Then they have to decide which one of those they really want, or any of them. If they say they want to know something about anemia, we would like to turn that into a question about hemoglobin, just because that is something that we can easily find in the system.
So there is like three dimensions of this negotiation to come up with something. Let's say they are all defined. The idea would be, there would be a catalog of questions or variables in which if two attachments were used to say the same thing they would be stated the same. But there is also this need to have some logic.
It is a spreadsheet, this transform is a spreadsheet. They have a label in X12 that says ambulance attachment. Then we have a spreadsheet which can be extended or varied in two directions. One is to say one of these three would work, and then the second time is that over time they may want to add or take away. But we wouldn't have 57 variants of the attachments. So when they said attachment ambulance, it would mean this current spreadsheet.
Now, next month they may change it, or maybe yearly, whatever the frequency was. Then the hope is, when they are asking for weight, they wouldn't have to do a chart extraction. You could go to the clinical system and pull out the body weight. But then you do have these variants, estimated, stated or measured. So those variants would be explicit.
DR. STARFIELD: Or a weight 10 years ago, the last one on the chart.
DR. LUMPKIN: Perhaps you can help me, because as I am listening to the discussion, it becomes clear to me that what we talk about with an automated attachment is really two different processes. One is an automated request for additional information, parentheses, attachment. Then the second is the response to that request.
I differentiate those two, because if our goal is administrative simplification and you have so many different variants of that, then a real service would be provided by just clarifying and defining an automated request for attachment. It would not necessarily have to link an automated system. There could be a paper response. If they request an op report, and what they want is an op report of this procedure and that is now in a structured mechanism, then the clinician will know what to send in. That is a service.
So I would feel much more comfortable if we focused in on the request and then looked at subsequently the response, which could be automated.
DR. MC DONALD: Automated or not, if you define the request in what I would call insurance lingo, it is hard for the physician to know for sure what they want. I think you've got to look at both sides while you are thinking about saying it, so the vocabulary is at least transformable.
DR. BARR: One of the things that we are going to do, to answer your question, we are having a meeting in Chicago next Thursday and Friday. We are bringing in possibly 70 payors and providers and the people that are working with the proof of concept project. We are just going to bring in with the intent to organize some kind of structure or group or committee to do exactly what you were saying. We hope to get a group together that will start standardizing the messaging done from the contractor to the provider throughout the entire country.
We realize that that is a really big problem, that the standardized messaging on the front end is just as important as the variable of the messages on the back end. So we are having a two-day session in Chicago next week. If any of you want to come, you are more than welcome to come. Clem has an invitation in his back pocket.
DR. BLAIR: Is this done along with HL7 and ASC?
DR. BARR: No, it is a HCFA sponsored meeting, where we are going to have payors and providers that we selected. We just wanted to have the audience targeted to a small enough work group, so it would not be a two-day presentation, it would be two days of interaction, and working together to try to come up with some goal.
What I really visualize is that out of this two- day meeting, we can get together some payors and some providers that would formalize some kind of work group and go away and actually start trying to standardize the messaging from the contractor to the provider that you were mentioning, Dr. Lumpkin, because we feel that is a really big project.
DR. BLAIR: I'm just wondering if this is going to be independent of the work that HL7 and ASC is doing, or are they going to be a part of this group.
DR. BARR: We look at it as if this group would start on its own course, and we would have interaction with them. But this attachment group would not be the directive of what is done in that group.
DR. STARFIELD: Given that we agree that this is an important process of the work of this committee over the next two years, as we move into the computerized patient record, can you give us the game plan? I guess you started doing that with telling us about next week.
We all have these questions for the National Committee claims attachments. We have this piece of paper. You want some reaction on that, too, right, Marjorie?
MS. GREENBERG: These were questions that were developed by the team that suggest the full committee might want to include, or hope that they would include, I guess, in the hearings that you are going to be planning.
DR. STARFIELD: So maybe that is the next step, to talk about hearings. Maybe you could handle that, and talk about how you want input into these claims attachments. I've got several, I'm sure many people do.
DR. MC DONALD: You are right, about the stuff getting across, but we have to get down a little deeper in the questions, so that it is crystal clear what they mean by the questions, whether or not you automate them, because the poor guy looking at them, perhaps the first date he sticks in there is the wrong date.
But the second part though is, when we are talking about things we find in the chart, I think the world is going to actually work out pretty well. Everybody knows what those things are. It is these open-ended questions which would not be found in the chart, which would be a wondrous service, if they could ask the questions one way when they are asking about one thing.
DR. LUMPKIN: But by standardizing, what we get to is that many of the -- I suspect, just based upon my experience in the world and knowing how clinical practice goes, some event occurred in which one piece of information wasn't available, whether related to fraud or some other thing, and the company decides that we need to have this piece of paper, so let's have it every time we get the sheets so we don't have to ask for it. Or if we do ask for it, at least we're getting the information, even though we may not need all of it.
So each company means something different, because of their past experience. Once you have a structured query mechanism, when they determine what their need is, they will fit their need within that structured query language. The clinician will know, if I get this question and I send this response, that will cover their needs.
DR. MC DONALD: If what? This is not a query language, it is English. These are what these questions are. They say, please explain why this patient had two bypass surgeries in the last two months. That is literally what it is. No one is proposing to make up a query language, but that is a common kind of question. The threshold for certain procedures in the table is, don't do more than X of them in six months, and if you hit that threshold, you get kicked out.
Now, what makes it tricky is, everybody has different tables. They don't always want to tell what they are for maybe good reasons or bad reasons. Then on top of it all, the physicians think this is all just a game to make the payment late. I don't think that is true, but it really is a complicated beast.
The other side of it is, the physician will send the whole chart every time. Then that drowns the payors, but then they feel they are getting even.
DR. STARFIELD: But you can look at this kind of question, explain to me why I should pay twice, or explain to me so I can understand about the quality of care. You could look at it either way.
DR. MC DONALD: In fairness, they do have -- there are maybe 100 of these stated as a more standard form, but there hasn't been agreement yet, that the payor industry would agree to comply or stay within some constrained number of these. If their goal is to make the physicians upset, well, then it was easier to do it in a limited number of questions.
The good thing that will come out of all this is that there is likely to be a reduced number of variants of how these questions are asked, and ideally some formalism in submission of the question to a question list, and you don't ask another one just to put a verb in instead of a noun and get away with a new question.
So I think the net result, just reading it will be easier, because you can train your staff on 400 questions not on 10 million.
DR. STARFIELD: I suppose there is no way you are going to get around this need for some kind of attachments, even when we have a computerized patient record, because there will always be questions.
DR. MC DONALD: That is the further confusion; the question is called an attachment.
DR. BARR: Going to your question about, do we have a plan, I think we were referring to -- what we are planning to do is, you go to page three, I think it is page three.
DR. STARFIELD: It is something that was handed out, the attachment proof of concept.
DR. BARR: Page two gives you who are participating and their roles and participation, what the payor will do and what the provider will do.
At the present time, we have five Medicare contractors and five commercial contractors that are coming to the table and have agreed to do the testing for this. The commercial payors are paying this out of their own pocket, the Medicare portion is being paid out of Medicare, but for the first time, commercial payors are coming to the table and have said they will do the testing.
This will probably be the first standard that will be tested before the reg comes out. We are having a three-month to six-month test period, hopefully which will begin in February of 1998, where we are actually going to start testing the work we are doing with the X12 275 and the HL7 structure at the five Medicare contract locations and the five commercial contractor locations.
My target date for issuing an NPRM is going to be August of 1998. I should have some kind of data, and I should have a little bit clearer path than anybody else, because I will have something to go by that has already been written and proven.
Going back to the questions that we asked, we are really interested, because in the work group that we have right now, there is a lot of controversy even amongst the work group membership today, whether we are going down the right path or not.
This group has made a lot of decisions that are going to affect a lot of people and a lot of cost and a lot of time, and have done this to the best of their ability and the best of their knowledge, with very little input from anybody else, and have tried to make the decisions based upon our knowledge of what the commercial insurance companies do, what we found out from the NUCC, and what we found out from amongst our own people.
We are looking for this committee to give us some kind of guidance, whether we are going right or wrong, and what they really want us to do. One of the reasons why we gave those questions to you is as a base for you to go out, whether you want to modify them or not, and ask for the additional hearings.
People have been really aware of what is going on in HIPAA regs, in the first nine that are coming out. Nobody knows about attachments. We have gone through and made presentations at all the standards groups we possibly can. We just had one at the NUBC, we've done one at the NUCC, we've done the Blue Cross/Blue Shield caucus, we've got one scheduled at WEDI. But nobody is even aware of attachments coming back down the road.
If you think about it, '98 is going to be here in two months. We are going to be swamped with figuring out what is going on with the first nine regs, and here we come down the road with something that is twice as complicated and twice as laborious as those nine combined. We are really in a bind, because we are making a lot of assumptions in our little vacuum.
DR. BLAIR: This may be as much to Clem as it is to Steve. Is there any possibility to discern out of the studies you have been doing and the work groups that you have had that we could maybe apply an 80-20 rule, where there are 100 questions that will cover 80 percent of the situations, and at least within this next year we could achieve that? The open-ended questions, gosh, that could be five years or something.
DR. BARR: What we visualize is, we realize this is not going to be a one-shot project. We know that if we can just get the mechanisms in place to do the transaction, utilizing HL7 and 275, that we are going to have to go into this in greater depth as the years go on.
This is not going to be like the current regs, saying here it is, this is the standard, and this would be the promulgated standard forever. We know that we are going to have to phase a lot of this work in as we go along.
We know that for example the X-rays, EKGs and all that mammogram information is related to DICOM standards. We have had interface with DICOM at the HL7 meeting in San Francisco, where we asked them if they wanted to participate with us in this study. They have gone back to their own work group and we are waiting for a response to come back from them. We haven't heard a response yet. The next HL7 meeting is in January.
So we have people on our work group that are willing to test the DICOM standards for transaction electronically, and we have a hospital based facility that is willing to test it. We have a provider that is going to test it, and we have a contractor willing to test it. But we don't have approval from DICOM to utilize their standards. So we are in a holding pattern with all those standards right now. If we don't get agreement by January, January may be too late already for us to test. So that may have to go back to the next phase.
DR. MC DONALD: You asked a little more narrow question. I think in fact, they picked out -- I thought it was 10, I heard 28 attachments. You know the old joke about, you say the number and they laugh, 33, 45? That is kind of like what these attachments are. You say the number, and underneath it there is all this other stuff that you know is there. So we've got 28 of these jokes that we have to work with.
DR. LUMPKIN: I think you and I are the only ones that know that joke.
DR. MC DONALD: But there is a narrowing down, and the questions themselves are in a separate category. That is a matter of -- next week, could we come to agreement on 50 or 80 or whatever of those questions and how they should be stated.
DR. LUMPKIN: I question though whether or not
we are trying to bite off more than we can chew, by looking for an interactive standard.
There are two components. There is payor and provider, and then there is the standard back from provider to payor. The ability to standardize from payor to provider seems to be a relatively low hanging apple.
Now, we may have to --
DR. MC DONALD: I think it's the same problem.
DR. LUMPKIN: No, I think it is a different problem, because once you get into sending stuff back, you have to have a standardized format for an X-ray, you have to -- how do you pull stuff out?
DR. MC DONALD: I don't mean to make it too narrow, but the problem is not the back and forth; the problem is making it crisp enough that you can do either direction.
You've got to look at them, you've got to spend a day digging through these things to understand.
DR. BARR: I could send you the same file I sent Clem.
DR. MC DONALD: The problem is, no human being could figure out on average what these things are. You want to define it in terms of the practice notes, because they are delivering it.
The other side of it is, there are some parts of the world -- payors would be happy to get it all coming in electronically, but that just increases their power to ask attachment questions against the little provider, who could spend all day answering those questions.
So there is also -- if we don't keep their ultimate interests in mind, I think the providers are going to the hospitals to start, because they have some electronic data. They could return the discharge summaries and the out notes and the ESRD reports and the ambulance reports with a little bit of work in the industry, because they are sitting in computers in their places, and saving some time. But if all they do is key in, reading someone's screen, they might rather have it be more laborious on the other side, too. So we have got to make sure that we have fairness on both sides.
DR. STARFIELD: How do you want comments back on these 19 questions?
DR. BARR: You can send them back to me.
DR. STARFIELD: We'll be here all afternoon if we discuss them, so how do you want them back?
DR. BARR: You can send them back to me.
DR. LUMPKIN: I think they are now our questions.
DR. BARR: They are your questions to deal with. We just gave you as a guideline 19 questions that we focused on that we thought you might want to ask the people who are going to be attending the next hearing session. They are just basically guideline questions for you to utilize whichever way you would like to. These are questions that we thought were important to the work group.
DR. LUMPKIN: The reason why I am trying to clarify this question, because I am not trying to answer the question, I am trying to more clearly identify what we hope to get out of a hearing, and to ask people about attachments at a hearing without some structure to what we are looking at I think would be not the best use of anybody's time.
So if in fact we can more closely segment what it is that we want to know about, if in fact we are looking at X12 277, and maybe I don't understand it, and hopefully you can send me the dates of the meeting in Chicago and I'll try to rearrange my schedule to attend.
But that is the reason why I am trying to determine if we need to have hearings on both ways or if the 275 is not mature enough, we don't know what DICOM's feeling is on this whole issue, do we maybe do 277 and then some other time do 275 or whatever we want to see for the return of the response to the 277 request.
DR. BARR: The only problem I visualize with that is, your next meeting is in March of 1998. By March of 1998, we will already have been going down the road, started testing on the 275 response back in some capacity.
So we have already gone down the road. If you delay it any longer down the road, we will already have made a lot of decisions that we really would like to have your input to, rather than wait.
The 277 request is not a mandate that it be done by HIPAA. It is just a mechanism that was already in place. This group created it because we wanted to have an electronic means to ask information electronically. It is not a HIPAA mandate. No one has to use it. They can still request the information by paper. That is their choice, as to what they really want to do.
DR. LUMPKIN: But my point is that our charge under HIPAA is administrative simplification. What I am hearing from Clem is that the simplification is not how you send the information into the payor. It is trying to understand what the heck they want. So in response, since you are not really clear what they want, you send them everything you've got.
DR. BARR: The first one isn't in our responsibility. It is 275 that is the one that is actually in our responsibility, so that is the other tangle on all this.
DICOM actually can be sent as a wrapper inside of HL7, so there is already a way to do that. We already worked out how to do it with HL7, and the truth is, images are technically even harder to send, because they are these big bulky things and the telephone lines don't work off of them. So I don't think that is going to be a horrible thing if we delay the images.
DR. COHN: I think part of the discussion and the difficulties of this discussion are that we are in here talking about the project, we are talking about what is going to be done.
Normally, when this committee has approached these issues of transaction standards and other needs, we have asked the public and vendors in the industry what they need first of all, and then connected that with what is being done. This is obviously happening backwards, in the sense that all these things are happening right now. We are being asked to make some judgments about the value and validity of what is happening without any public input, other than HL7 and X12, which are obviously very valid input mechanisms.
But I am still left with the question of really, what does industry need in all of this stuff, as opposed to what it is we are going to produce.
MS. GREENBERG: I just wanted to make a comment about the scheduling. The next full committee meeting is in March.
Now, this subcommittee is planning to have a meeting either in January or February, which was tentatively to address the issue of claims attachments. I guess there is a possibility of having more than one subcommittee between now and March, although we have had difficulty even establishing the date for the one.
So anyway, it will be either by January -- it is looking more like February from what we found out in polling people. So that is something that really has got to be determined.
But the more basic question is, what do you want to accomplish at the hearing? But I just wanted to remind the subcommittee that we still have to pick a date for this.
DR. MC DONALD: They have started this before the legislation even came along, so there was some momentum, and they are taking advantage of it.
I have to praise them for trying to want to test something before we declare a truth and right, regardless of what people say in public meetings. Steve, you might mention that the UUNC meeting, there was a sentiment that was expressed about attachments.
DR. BARR: You made a presentation at the NUBC meeting yesterday on attachments. The concern was that it goes down to, policy one is, is there a need for attachments per se at all. With the unified standard data set that is coming through, supposedly that data set which will be maintained in the 837, which will allow for almost all information to come in or claim to be processed, in reality what more information do you need in regards to attachments, and are we just promulgating the claim itself by having all this attachment information.
The NUBC was concerned that if we automate attachments, we are making it much more convenient to ask a lot more questions a lot more efficiently, and are killing the messenger. They are concerned about that need, because their question is really related to a policy issue: in reality, do you really need any attachment information in today's environment with the standardized data set.
MS. GREENBERG: Are they assuming that the data set will be considerably expanded to include information that now is being -- people are asking questions about?
DR. BARR: No.
DR. MOORE: First of all, getting somewhat into Simon's question, we recognize and the legislation recognizes that attachments wasn't going to be an easy issue to answer. That is why they gave us another year to do it.
We also recognize that we had a late start on the other transactions. We are suffering right now. We are not going to make the October deadline that was submitted.
DR. MC DONALD: You mean November?
DR. MOORE: We keep saying October has got 63 days in it this year. But we have convened a group of major payors and other community people, which are the same people ironically that come to this committee to testify. We are trying to wrestle with some of the technical issues, and this is a work group, not the implementation team yet, to try to get some of this groundwork laid as to what is an attachment, how many of them are there, what the hell are we asking for if we are all payors.
I know some in HCFA hate it when I say that we are going to put information out of the bill and into an attachment and not take the things out of an attachment and make the bill bigger, because the bigger you make the claim, the more often it changes every time you change what you ask for.
Then we are back in that position of, everybody has got to change the version of the claim, because we are changing it because HCFA wants something, one state wants something, so we try to say, what are the attachments. Are they statistical? Some states have said to us, our legislation wants statistics, are we going to look at a statistical page for states, and states could get together and decide -- we even asked CDC if they would lead that effort to pull that together with public health, say what is it you want from providers that are attachments.
There are a lot of things that go into this, even beyond what you probably looked at. Attachments can be almost anything, it can even be something or public health, it could be something for a payor, it could be a lot of things. But no one is getting together to put some of these issues on the table and try to flesh out what we are talking about. That is the work that Steve is leading. It is mainly comprised of people from the private sector, not people from the public sector.
DR. MC DONALD: The clinical things we call attachments, those are the clinical things that get attached.
The bill itself goes on for miles. The billing record itself or the claim record has at least 30 distinct variables, like, amino acid dose is in there, that are ad hoc and irregular. One of the thoughtful parts of the group thought we ought to get those out and make them regular, so they are not limited to the use that people thought up one day, because you might want some other ones.
Then on top of that, there isn't a place in the claim for the whole discharge summary. That is clearly, when it is asked for it has got to be one of the top five of the op note.
DR. BLAIR: Since I am partially reiterating the question that I had before -- and I thought, Clem, you would wind up saying that there might be 28 questions that would cover the predominance --
DR. MC DONALD: There are 28 data sets or batteries. You might think of them as clinical.
DR. BLAIR: If you focused on those, is that going to pretty much cover 80 percent of the questions or whatever?
DR. MC DONALD: Probably 10 attachments are the top -- or 10 or 20 are the top 80 percent of the line.
DR. BLAIR: I feel like Steve is asking us as a committee for a little guidance. I think the kind of guidance you are asking for is, what is the scope of defining health claim attachments to be standardized by this committee a little more than a year from now, February of '99. Maybe what is realistic is for us to say that the scope is some number that could be done within this next year, that would be able to handle those 28 different data sets. Is that getting close to what is realistic?
DR. STARFIELD: But Steve asked us, should we be bothering with claims attachments at all. Isn't that what you asked?
DR. BARR: No, we think you have to give us some direction on where we should go with claims attachments. But regarding the 28 that you mentioned, they are the 28 that were identified throughout all the surveys that were done, whether it was the NUCC, HCFA's, whatever it is, as the top 28 used attachments.
So we figured we would use utilize and identify those 28, just those 28 to start. We will have taken -- when you take away the op report that is also a big player and you put those down, you will have approximately 80 percent of all the attachments that are being asked today automated.
Now, that leaves a lot of information from the Medicaid bureaus, the states. We have got a request from CDC that wants to transmit information through us. NCVHS wants to transmit information through our transactions. SSA came to us the other day and wants to put disability information through our transaction. People realize that there is a transaction and a mechanism to bring data, clinical or whatever kind of data it is, electronically now, to say stop what you're doing now and do this, and I need this.
Once this gets out, you're going to have to have a cave and try to hide.
DR. LUMPKIN: What I would suggest is, in our hearing we focus on a day that we would deal with responses to three proposals. One is that we would perhaps ask Clem and Steve to work together on -- maybe Steve can help Clem write a paper, which would be Clem's paper, so it doesn't have to go through any clearances.
DR. MC DONALD: We have one. We can send you the --
DR. LUMPKIN: Fine. But what I am proposing is that the hearing would be about those two or three documents. It would be about 277 and 275. We would structure panels that would give us some general responses to those documents, and then a panel that would give us some general responses -- or maybe to all three, but essentially we would focus in on those three batches of things.
One is our general position on transactions, which would be those two or three documents maybe pulled together into one, and our intention of supporting the approach to go with 275 an 277.
DR. STARFIELD: Who would we have at the hearing?
DR. LUMPKIN: I think we would ask providers, we would ask individual providers, hospitals. We would probably want the clearinghouses there, and the payors.
DR. BARR: I think one of the paramount questions that is asked in that paper that we gave you. We are tugged at the sleeve every time we make this decision, and we have gone around and around as a group, is the $64,000 question: what is an attachment and what is a claim.
We as a group wrestle with this problem every time we have a teleconference, every week we wrestle with this same situation. Depending on which day it is depending on which direction we go.
DR. LUMPKIN: My inclination would not be to ask that question, because we will get 28 different answers, or 35 different answers. I think I am more inclined to put our foot in it some way, and say, this is what our opinion is, and let them have at it. At least we give them a target to go at.
DR. MC DONALD: I think we have got to categorize somehow the name, because otherwise -- we have to define the scope, and we have to be clear that there are a couple of variations in attachments. One of them is certification information. It is not about the patient at all, it is not billing data. So there are some funny things in here, and I would really like to get the committee to at least give some thought to -- you keep saying 275, 275 doesn't really do it without HL7s, so there is this other critter underneath it. But that gives it the structure that is pretty well worked out for the clinical stuff.
I would like to at least consider whether this whole issue about merging and blurring clinical data into claims data, in a way that it doesn't mesh with what you store in clinical systems, number one, and number two, is actually uglier in sin. You have variables for drugs and not the other 1500 drugs that might be there.
I think there could be a lot of value in isolating these, because the claim would stay more fixed, and there is more variable structure that would be available to send additional clinical data.
So we at least ought to grapple with that to sort out this data, because it is all mished together in a way that is really hard for anybody to figure out.
DR. STARFIELD: Well, we still haven't got a focus for the hearings. Do you want to make a proposal for the hearing?
DR. MC DONALD: One of the things that we should say is what can we do about reducing attachments, period, not just hiding them in the claim, but just not having them. That is what they came up with. It is a big nuisance to both sides, and if you can figure out a way to do the work with fewer of them, that would be good.
DR. BARR: That would settle a lot of problems. All we are doing is taking paper and making an electronic -- when in reality, we would like to eliminate the paper and the electronic.
DR. MOORE: The reason we have the attachments is because we the payors don't trust the providers. As we hear more horror tales with fraud and abuse and the IG and the Department of Justice folks around, we are going to be driven, we HCFA, to do more attachments. Before we pay for something, someone is going to say, why aren't you checking this? If we don't check it, we will get beat up again.
So the question that Clem is saying, I don't think attachments should be part of the bill, it should be something different, because it will change. And as conditions change, some will go away and others will rise up. So we need a vehicle by which payors feel more comfortable when we pay for health care, that we are paying for something reasonable to the right person for the right reason.
MS. GREENBERG: Obviously, you are going to have to put boundaries around this, and I think that is what the team or the work group is asking for some guidance on, too. But I think if you are talking -- and I understand your not wanting to totally open this up.
But I do think you have to understand that, as Steve indicated, there are groups -- and in particular, some of them were just at the X12 meeting out in wherever that was, Los Angeles, okay. And public health is among them, including components of CDC, et cetera, that have been told, don't try to put that into the bill. This will go into an attachment.
That has really been the message, the HIPAA mantra, during this whole year, about clinical or other types of information. Don't try to smoosh that into the bill, let's keep the bill simple. Although I must say that I don't consider the current bill simple when I look at all the elements that are in it. But nonetheless, let's not make it worse than it is. And these would go into attachments.
So if you are saying -- and there are state public health agencies, et cetera, who are going to X12 and trying to participate in this. So I think they need to be invited to the hearing as well, and I think you at least need to hear what these peoples' needs or expectations are. And maybe there will be a phase two. Maybe phase one is to just try to deal with what is going on between the payors and the providers, and maybe in doing that you will provide for some of these other people.
But I think that you need to know that that has been a subtext that has been going on this past year, as people anticipate the attachments issue.
DR. LUMPKIN: Well, again, I am not really uncomfortable with everyone trying to hang their at on the attachments. I think that is a burden that has to be negotiated as a contractual agreement, as in any data collection.
My concern is that, you have 25 or 30 different payors who are all trying to fight fraud and abuse from the same set of small percentage of vendors, yet they all come up with different questions, and the questions are all seeking the same information, and this poor provider doesn't know what to give them, so they give them the whole chart.
I think we can deal with some, recognizing that there may be a lot of people who hang their hat on the attachment. If there is a public health attachment for reportable diseases, that makes some sense, a request for attachment. But I don't think that there needs to be -- I think there needs to be some sort of standardization of the inquiry.
So if we can perhaps focus in on the bit that we can maybe have some influence on, that may be what we can focus our hearing around in the early part of next year.
DR. STARFIELD: John, why don't you lead us on a discussion about the hearing, the date and what we need to do to prepare for that?
DR. LUMPKIN: I would propose, not having checked with the budget folks or anything like that, just kind of looking, it looks like the largest number of people in the committee can be present on the 2nd and 3rd of February.
Now, rumor has it that X12-N is meeting in Anaheim. So my question would be whether or not it would be workable if we were to have our hearing in Anaheim.
DR. BARR: Can I ask a question? That week in Anaheim is crucial to this group for doing the proof of concept because, one, we have to -- we just got the data maintenance one for the 275 patient record. We have not even started to write the implementation guide at all for this 275. That will be a crucial week for us to start writing and fine tuning that implementation guide, so we can meet the deadline to be available for the NPRM as it comes out.
DR. LUMPKIN: So you are saying that it won't do any good to have us meet there with you?
DR. BARR: Either after or before, but not the same exact week. That's all I'm saying.
MS. GREENBERG: Also, the 4th and the 5th and ANSE HSB is meeting.
DR. MC DONALD: Can we come back to scope? I think attachment can't be anything you have ever thought of, like it can't be clothes or it can't be gasoline. I think mentioning immunization and all that stuff is a completely confusing thing. So I think attachments ought to at least specify things that support claims. They are blurry enough with that, but they are not any kind of clinical or lab or statistical data. There are messages around to submit other things different ways.
DR. MOORE: As Marjorie said, there are a lot of other organizations out there that have latched onto the idea that an attachment is one way of getting data that they need.
Now, some of the states have said to us in meetings that they are looking for statistical data that has nothing to do with the claim or the payment of the claim by the provider. Yet, they are saying put it on the --
MS. GREENBERG: This isn't data that they are currently getting on the claim? Because I thought one of the issues was, like on the UB-92, some states are currently getting information for that, and once that is standardized they won't have those elements on there, unless they can all agree on the few that will be added. So then they thought they could get it through an attachment.
DR. BARR: Correct, and we are running across that same problem in the state Medicaid. Each state Medicaid has got the same problem. There is a lot of state Medicaid mandated data that is specific to a state, that will not go in the bill, and we are trying to work with the state Medicaid agency to try to define a subset of all specific data for Medicaid that we can put in through an attachment.
DR. MC DONALD: Those are things that support the claims.
DR. BARR: Right. As you can see, this is the reason why this has been a project that no one has ever wanted to touch.
DR. COHN: It has been a number of weeks since anybody has sent in any of their availability. Wouldn't it make sense to ask people if those dates are still right for all of them, before you identify --
(Simultaneous discussion.)
DR. COHN: I was going to suggest the 9th and 10th in Washington, of February.
DR. STARFIELD: The population subcommittee meets on the 10th, so actually, that provides some potential for overlap, if we want a few hours for overlap.
MS. GREENBERG: The population subcommittee is planning a site visit on the 9th and 10th, very possibly to Arizona.
DR. BARR: At the X12 meeting, they usually break up on Thursday. Our work would probably be done through Wednesday, if you wanted to meet Thursday and Friday of that week of the X12 meeting.
(Discussion off the record regarding meeting dates.)
DR. MC DONALD: Can I come back to the question of what the committee's perspective is on encouraging splitting out clinical data from the claim into an attachment? There may be more things, but I can say it more clearly that way. Because that is mostly what keeps changing, that they want to get these additional clinical variables.
That has been an argument going back and forth. I think actually there is maybe -- is there a balance of view on that now, on the federal committees, whether to lump them or split them?
DR. STARFIELD: You mean, just split up the clinical data from the other administrative --
DR. MC DONALD: Yes, and make it more regular and more like it would be in some other context.
DR. LUMPKIN: Can you give me an example of what you mean?
DR. MC DONALD: If you look at 837, it is about this thick printed out. It is not just a claim when you read through it. It has variables in it like how much does it charge and what is the ICD-9 code, what you usually think of as a claim. But it also has variables like, what is the creatinine, what is the hemoglobin, what is the amino acid dose you gave, and it goes on. There's probably 15 specific drugs or doses that are selected out for inclusion in it.
I think there is something in there like creatinine kinase IMB, but not troponin. It is not sustainable, they are just flat record fields: give me this variable. And they are there for reasons, but it would be a lot easier if one generalized that and abstracted it out into a general attachment structure that could map to a lab system or a pharmacy system.
DR. LUMPKIN: Can I perhaps refine your question? Because really, if we're talking about doing something along those lines, we're going to have the NPRM out on the street. We've already got 837. Our window to alter that is probably past.
But if in fact we were to design the attachments so that that information would now be available through the attachment at the next remake we could just delete that stuff out of 837. It is much easier to delete stuff out of a currently existing standard than to add stuff to --
DR. MC DONALD: Let's talk to the folks who are wrestling with those issues. I wasn't proposing to tear off the current ones, but rather ignore them after awhile.
DR. LUMPKIN: Within that context, then I think that is a valid discussion.
DR. STARFIELD: I think we have five minutes at most to finish this up. You want some questions. We need some questions for the hearing, right? This is a good list to start with, not necessarily to stop with. Do you want reactions to this list?
DR. LUMPKIN: Could we just submit additional questions?
DR. STARFIELD: Were you suggesting that these documents that might be prepared for discussion would incorporate some of these questions?
DR. LUMPKIN: I think to the extent that we put up straw figurines to let people strike at, our best guess, I think we would get a better hearing than if we just ask open-ended questions. So I think that is why the documents would be useful.
DR. STARFIELD: You are proposing that Clem and company write a background document?
DR. LUMPKIN: No, actually -- well, they may have been, but I'd rather not -- there are a number of documents, two or three, anyway, that both give examples, some real life examples in some sense, and we could collect together a package of that and certainly distribute it to this group. I'm not sure whether it should first be looked at, just pick out what you want to send outside. But this would help to give you more background on what the issues are.
DR. MOORE: That can be used as the basis for doing the invitations to all of those, like we did before when we had the hearings, where we sent out the invitations to those, once we decided what represents the best to bring to the committee to do the presentation. Ask the questions that are there, give examples, so that you can focus the issues and get to some conclusions. So we can work on that.
DR. STARFIELD: How do you want to go about generating the questions? We know for hearings, we need very specific -- can we do e-mail to you, or to what?
DR. MC DONALD: Could you send out a request with these questions in it to add to? That would be the easiest response. Then we could add to them or make suggestions.
DR. BLAIR: Clem, when you were winding up saying 837 is going to wind up having a lot of these --
DR. MC DONALD: It has them. It has had them. It has always had them.
DR. BLAIR: Were you trying to indicate that many of the things that are showing up in terms of claims for health claim attachments in fact are going to be taken care of by 837 when it is widely implemented, and that part of our problem will be solved with that? Is that what you're saying?
DR. MC DONALD: No, there's two things. That was said, that that has everything solved, in another discussion. But what I was saying was that we -- and this is not my idea, I am actually echoing some of the things that were discussed at these X12 meetings and in HCFA and various groups. We ought to clean up the conceptualization of this problem, and things that are claims should belong to the claims, and things that are clinical should belong to clinical, and then the clinical ones can be defined in terms of the clinical system, and they would be the attachments.
Now, there's more to attachments than this. The point would be to clean out these arbitrarily blessed variables that are no longer necessarily the right ones to measure the things you're trying to measure, but they are now fixed in as data set elements. If we get this attachment strategy arranged, we can define what is the current best set of variables for the particular purpose.
See, ESRD is all based on, they want to make sure you're not using too much erythropoetin, and there are some algorithms for how much you should use. So you can figure out the variables. Some of them are related to heart attacks, and they want to see proof you had a heart attack. Well, now troponin is going to replace CKMB. So we can do that more graciously if we do it in a structure to which clinical systems are accustomed, like a lab system is structured.
If we get them out of the claim, then the claim won't have to be changed every two weeks. This other structure will be a general structure which would accommodate a lot of these things.
DR. LUMPKIN: I see.
DR. STARFIELD: I think we have a game plan. Are you happy with it? Okay, can we move on to another topic?
DR. MOORE: I just want to make one comment. I think that right now, this group is working as a work group. It is not one of the implementation teams that is on the standards committee. So we will probably have to form another team, working toward the attachments. That becomes a standard work group. We have people already ready to move in to be the chairs and co-chairs with a supporting cast of thousands.
DR. STARFIELD: My understanding is that there was going to be an implementation --
DR. MOORE: Yes. We just need to let the committee know. But right now, we have been telling everyone there are six implementation teams. This looks like there will be a seventh one being added that will be focusing on attachments.
DR. BARR: We could not wait for the time line. As Bob said, we were behind the schedule. We would have waited until everything got out the door in the first NPR and then started to do it. We would never have met the time line that was established for one year.
So we have gone down our path, and we hope we're --
DR. MOORE: We'll stop Steve. We'll hold him right to the reins and not let him get away.
MS. GREENBERG: You might want to introduce Phyllis, who is here also, and anyone else who is here.
DR. BARR: I can introduce the rest of the team. Maria Ward, from HCSA, and Derwin Day from HCSA. They are going to sponsor the meeting in Chicago.
DR. STARFIELD: But they are not actually going to be on the team, right?
DR. BARR: No. They are on the proof of concept team --
DR. STARFIELD: They are on the work group.
DR. BARR: -- who are developing the standards and everything like that, the worker bees. There is Phyllis, who is the co-chair of the implementation team and is also a member of the proof of concept team. There is Andrea Wallace, who is on the proof of concept team. Behind her is Julie Meeham and Mary Ann Busman, who are on the proof of concept team.
DR. MC DONALD: Where is Gail?
DR. BARR: Gail, she vacated when it got too controversial. She left. But Gail Carter, who is also here from Anthem.
DR. STARFIELD: Okay, thank you very much.
MS. GREENBERG: So can we just summarize? We have agreed that the subcommittee is going to have a hearing on the 9th and 10th of February, and --
DR. STARFIELD: We'll get together some background documents and we'll get some questions.
MS. GREENBERG: And I guess lead staff will coordinate however these invitations are going to go out, or what have you. We may need to have a conference call or something of that nature.
PARTICIPANT: Where will the meeting be on the 9th and 10th?
MS. GREENBERG: In Washington.
DR. STARFIELD: The next item on the agenda is ownership and maintenance of data standards, and for hearings, with a to be announced leader of the discussion. Marjorie?
PARTICIPANT: Who is Mr. TBA?
MS. GREENBERG: I do not know. Judy?
DR. BALL: If you have TBA, you're looking at an earlier version of the agenda. Bobbie Redding is TBA. She is here to report on discussions at the X12 meetings about ownership and maintenance.
DR. REDDING: And other meetings as well. The view -- and I think it is a good view -- is that the data content maintenance should continue to be in the private sector as it has been for many years. I think that some of the questions have to do with who is best to do this, what could they do best, and how best to approach it.
Currently, the de facto standards that are being used pre-HIPAA, the content is fairly much governed by the National Uniform Claims Committee, which is sponsored by the American Medical Association, the National Uniform Billing Committee, which is sponsored by the American Hospital Association, and that is an institutional claim, the American Dental Association, and the National Council for Prescription Drug Programs, which has established retail drug claims.
Most of these committees have to manage paper and electronic format concurrently, because many health insurers require equal data sets on both paper and electronic.
These committees are the home of the best expertise in the country on data content, in my opinion and in the opinion of many other people.
Most of the committee deal with claims and remittance advice data primarily. But I think that you could make a pretty good case that the people that understand claims and remittance advice which are very intensively health care provider transactions, could manage the data content of all the HIPAA transactions that are routinely exchanged between health plans and health care providers.
There are others that do not fall into that category, for example, health care premium payment, which is typically health plan to employer. I'm not at all sure where the right home would be for that particular transaction. Enrollment is another one, the 820. That is not typically a health care provider transaction.
I think that the committee should welcome this work. I think that they would be glad to expand their membership if that would be wise, and it probably would be. This world is much bigger than the world they come from. Most of them have been in business since the '70s.
So that's all I have to say.
PARTICIPANT: When one talks about the transactions, I think it is crystal clear. When one talks about the code lists, these are going to cross -- ideally, the code lists would cross standards.
DR. REDDING: They do and they will. There is another piece to this that I did not mention. X12-N has established a special task group that is managing the coordination of the implementation guides. The actual guides are owned by the work groups.
We were thinking -- and this is all set in very dry sand, but we were thinking that perhaps the committees where their expertise is data content, the work group, whose expertise is the implementation guides, which is where the code sets live, by the way, except for the medical coding, would together make a coordination body similar to this special task group that is currently doing this kind of work, anyway.
So whenever data content changed, you would always be assured that like structures would be used in all of the transactions where that particular data was changed. Indeed, they do own the code sets that are administrative in nature, which are in the implementation guides.
PARTICIPANT: That didn't address the question I was going after.
DR. REDDING: I'm sorry.
PARTICIPANT: Crossing standards, the NCPDP and HL7 and X12, they have a field called gender in all of them.
DR. REDDING: Yes.
PARTICIPANT: And I'll bet you -- I haven't looked at it -- every one of them is different.
DR. REDDING: I can assure you that there can't be more than two. I haven't checked the NCPDP, but I can assure you, for X12 they are identical.
PARTICIPANT: Within X12. I'm just saying, across these areas --
PARTICIPANT: With two answers for gender, or --
DR. REDDING: I don't know.
PARTICIPANT: No, actually, there's three in all of them, three plus. Then there is race and ethnicity and it goes on and on and on.
I really think I would like to open -- well, there are two other issues. The standards groups -- I'm most familiar with HL7 -- they can't put out a new standard every week.
DR. REDDING: Nobody should put out a new standard every week.
PARTICIPANT: But the codes come up faster than that. If someone says there is a new bacteria, Legionella, overnight they are reporting Legionella. So there is some argument for trying to dissect the code sets, or at least large classes of them, out of the message standards and having them be -- I would also argue for having them think about it, where they aren't expensive, to encourage their wider use.
DR. REDDING: Are you talking about diagnosis and procedure coding? What exactly do you have in mind here?
PARTICIPANT: Diagnosis and procedure are a paradigm of the issue.
DR. REDDING: They are separate.
PARTICIPANT: They are separate, but there are issues there, too. But I'm saying gender, I'm saying sex, I'm saying action codes. There are things that cross these standards that are stupidly specialized by each group. Let me retract that, not stupidly, but that by nature, they end up with different things in them when they kind of meant the same thing. Some sort of an overview listing of these things would be very helpful.
Let me take another one, the one I do know about. X12 has kind of a unified -- they are very good about their codes, but they don't unify them with other code systems, like the code for grams isn't related to the scientific code for grams, and the pig iron code might be okay in that one. But there are things like that. They don't look outside themselves, and they should, because of the nature of this beast as we start using this data for many different things.
So I would at least like to explore the possibility of having a way to have the codes be externalized in various ways, maybe in various bodies, depending on who is most appropriate for them, and at least have a twinkle of a thought that they might end up in some publicly available code system or repository.
DR. REDDING: They kind of are now, although what you say about redundancy is probably true. But we have built a national data dictionary for all of the data sets except for retail drug. We have consolidated a lot of data names in a lot of coding systems.
DR. MC DONALD: But that's you, who?
DR. REDDING: We, HCFA, with X12.
DR. STARFIELD: Bob, do you have a comment?
DR. MAYS: Yes, this is Bob Mays. And actually, we are even going to be expanding that work, because you're absolutely right, data elements -- and I would include codes in that -- are actually very contextually specific, and there are good reasons. Gender was the perfect example. I have used it with other people. We tend to think of two or three codes for gender. EPA has nine codes for gender, because they deal with organisms that don't just fall into our concept of male and female.
And yet, at the gender level, at the concept level, you can have agreement. I can agree with EPA on what the definition of gender is from a concept. I don't have to agree with them in every instance of the implementation of that. We should try to say one is male across all the things.
So starting with the master dictionary, which is really just a compilation of the data elements across the transactions, where we would like to go then is more towards the registry concept as exemplified in say the ISO-11-179 or the ANSE X3-258 kind of standard, which begins to build a little bit more formal meta model on top of just the elements, that we can then ultimately begin to discuss this.
You are not going to be able to go element by element, particularly as you broaden out to these different groups. It is one thing if you are dealing with NUCC and one or two or three of the transactions. When you try to pull in a lot of groups, you can't just take a list.
In one of the hearings here, we started out with just the NUCC, and we started to try to go through A to Z. It is not going to work. You're not going to be able to go through every element and get complete consensus, that every element is going to be defined in all instances that it might be used in exactly the same way. I don't really think that it needs to be. I think it can be split out and looked at as different levels.
DR. COHN: I just want to make a comment about that. I know Bob Mays knows I have some strong opinions on the data dictionary activities. Currently, what has happened is not adequate.
I actually want to speak in some support of what Clem is talking about. There probably needs to be both things going on. One is a data model, which indicates the definition as well as the permitted values in a way that they can be looked at and understood what relates to what needs to occur. But then there needs to be a process to reduce unnecessary variation, which I think is what Clem is talking about.
There are some places where you really do want to have differences or expansions of values, permitted values. There are other places where you look at between two different systems, when one said A through E and the other one said one through five, and they don't relate in any way, and there is no reason for that.
I think it is really the responsibility of the committee to work on that, but we need to be able to have a document that allows us to do that. So I think clearly, we need to be working on that.
Now, I am actually going to further complicate Clem's model of the issues around data, by adding one additional element. You talked about the codes as being a problematic issue, and then said the other stuff is pretty clear.
I tend to think of it as three. One is codes, the other one is the data definitions, as opposed to like adding a new data element. Then there is the syntax. I think the syntax is pretty clear. I am a lot less clear on how it is that -- how it should be determined in a national standard when new data elements get included. And certainly I agree with you about the issues around code sets, that there needs to be some other process for that. I think both of those are a little less clear to me.
DR. STARFIELD: There is lots of history on how things get added, how data elements get added, for example, to the ICD. We have a regular process for doing that.
DR. MC DONALD: I was actually speaking for that kind of a process that was independent. I didn't really mean to imply that the data elements were easy. I was saying that that was pretty well dealt with by the message transactions and the process, the six month cycle or whatever the cycle is that works for that kind of stuff. It doesn't work as well for registration of codes.
The issue about the codes, you said it should be zero, one, two. ISO said it should be M, F and U. So it is not just -- if you do zero, one and three, what you would really like to have is a catalog of these things that maybe one question uses these three and another question uses two more, but when you use the same ones, they meant the same thing.
DR. REDDING: Well, the catalogs are not as beautifully arranged as what you envision, but they exist. They are on the website, and I'll give you all the website address.
DR. STARFIELD: What is this?
DR. REDDING: All of the information, the data dictionary, all of the data definitions, all the code sets are displayed on the Washington Publishing website right now.
DR. COHN: And that includes the NCPDP?
DR. STARFIELD: That is the X12.
DR. REDDING: No, it does not. The NCPDP dictionary is a separate issue. They sell standards, and you can certainly see it. But we have made no attempt to coordinate NCPDP and X12-N and implementations to date.
DR. MC DONALD: That's my point.
DR. REDDING: However, we did do the other seven or eight, and I think that is quite an accomplishment. One at a time.
DR. MC DONALD: That's wonderful, that is good. I wasn't trying to diminish that.
DR. REDDING: There are difficulties with aligning and online interactive transaction, where you have eligibility information, claims information and remittance advice information all in conversational mode in one transaction. That is a little different than the X12-N batch transactions, and it is just a little bit more of a challenge.
DR. MC DONALD: I wasn't saying we need to coordinate the transactions, but when you have fields with similar code sets, is what I was talking about. I know you have gender in NCPDP. I don't know how they code it. I know you have it in X12. There's a lot of them like that.
DR. REDDING: No, there's two. There is NCPDP and there's X-12.
DR. MC DONALD: There's three. There is HL7 -- there's four, there is DICOM and Edeffect.
MS. GREENBERG: I don't think we're going to solve this today, but at the same time, I want to support what Clem and Simon are saying.
Just as an example, CDC is trying to standardize its data for surveillance systems, et cetera. Many of them, if not all of them, collect a basic core of information and they all collect it differently. In doing so, they are laying out what are the standards.
Yes, HL7 and X12, and if they've got NCPDP and a few others, also, they are all different for the same things, such as marital status, where I don't think organisms have marital status. But I don't know. I actually think that is a little far-fetched.
For most purposes, dealing with human beings in health care, I think you could get agreement on gender. There are some issues, I know, that some programs have, if they are collecting information on fetuses, et cetera. But this is absolutely true.
I differentiate between when things aren't standardized for a good reason and when they are not standardized for no good reason; I know that is double negatives. But I do think on our long-range plate, that is something we should --
DR. REDDING: No one questions that, but I would be interested in comments you might have about the decision making process. One of the things that are attractive about the committees that have done this in the past is that they have a combination of payors and providers, both of whom understand the data structures pretty well. One knows what they want, and the other one knows whether or not they have it to give. And believe me, I have seen insurers go to committees and ask for information that providers simply do not have.
So I think doing this in a consensus process is a good idea.
DR. LUMPKIN: I think I have concerns about coding issues. But as I'm trying to think about it, there are front end and back end coding issues that have some impact upon simplification of life and the medical environment.
We have talked a lot about male and female and you and whatever else is there. But on the front end, it doesn't really matter. The clinician, the person filling out the record, is going to put in some standard code, whatever is in their software. That is going to be mapped to whatever transaction standard.
DR. MC DONALD: That's costly. That is what you'd like to avoid.
DR. LUMPKIN: But no matter what front end they have, there is going to be some translation, depending upon whether they are going to continue to use their legacy system.
There are other coding issues which we have talked about, having to do a lot with procedure codes and these other things, where the front end has to have the code, until the point where we get to natural medical language or something along those lines, which again is going to be some sort of standardized input.
So I just think that when we focus in on those data content issues, we need to focus in the ones which have the biggest impact upon whether or not the system is complex for the provider, in trying to have to put in the same data three different ways, because those can be resolved by mapping, even though it may be a little bit more expensive. CPU time at least at the provider setting is fairly cheap, because most of the effort is spent inputting the data rather than translating the data on their units.
So I don't know if you see the point, but I wonder if --
DR. MC DONALD: I agree with the point. I agree with the point.
DR. COHN: I think we've gone in some ways a little far afield from our discussion, which I understand is really called plans for hearings related to this issue, as opposed to solving it in the moment.
I guess as I conceptualize the issue, I think there are a number of things that somehow need to be handled, some of which can be done in hearings, some of which need to be done by our colleagues in HHS.
One thing that needs to be done by HHS is providing a data dictionary and framework of data in a way where they and we can all look at it to identify unnecessary variation around code sets and all of that.
Now, there is however a real place for hearings in all of this. I personally think that the issues around content, ownership and maintenance and all of this is a lot more of an issue than has been described so far in this session. I think that there is an assumption that people are playing easily and well in the private sector on all of these issues, and it would be something where I would believe that a lot more when those that are players come to the committee and in the public state that this is how they are working together, and they all agree that this is how they are working together, and that they are all satisfied with that, just because I think we're moving to a different world now.
It used to be that maybe there were work content champions and transactions champions, but there also were separate standards. Now we are really merging many of those standards together, UV-92, 837 and others. So I think it would be very helpful to hear from the affected parties how they think it should all work, and then we can make a determination, if that is what we want to recommend.
DR. STARFIELD: Bill, you wanted to make a comment, and then I'm going to ask John to wrap it up.
PARTICIPANT: The question I had was, the HIPAA law was very specific about instructing the department to pay attention to the industry consensus based standards under ANSE accredited standards developing organizations. But in fact, as Bobbie has described, the organizations which create the content of these transactions, which then the X12 people figure out how to get from one place to the other, don't have the same open consensus building due process as required by ANSE accredited organizations.
I am wondering, both from those of you who participated in the ANSE accredited process and those of you who work with the NUCC, NUBC, et cetera, whether these groups ought to, or at least would consider, implementing an ANSE style due process to make everybody feel comfortable that in fact they were participating in the overall industry consensus building process, and not running off and doing something under their own control.
DR. STARFIELD: Why don't we get a response from Clem, then John?
DR. MC DONALD: I don't know. I was just wondering whether that was a comment or a question.
PARTICIPANT: Well, it is a question, --
DR. MC DONALD: It's a good comment.
PARTICIPANT: -- if that is something that those groups ought to -- we ought to be encouraging those groups to be looking at as a way to get into the consensus world a little more than they are now.
DR. REDDING: I have mixed emotions. I've worked with all three bodies, and on X12 you get my vote. At the NUCC, I can't buy a vote.
PARTICIPANT: You can't?
DR. REDDING: No, you can't.
PARTICIPANT: No, that wasn't a question. It was a startled response.
DR. MC DONALD: Obviously, they don't meet in Chicago.
DR. REDDING: Oh, right. So I have mixed emotions about it. I know the ANSE process has lots of merit. But on the other hand, that is also how X12 wound up with five different homes for every claim data element, so that you could have an infinite number of formats, all of which met the standard.
A lot of people can buy votes, and people come and go. I actually saw it happen. An individual came and asked for a particular data element, and everybody said, no, no, that is a suffix on a CPT code. You don't need it again, and the individual was voted down and went away. The next meeting, they added it. The right people were in the room.
DR. MC DONALD: There is a different world when we have a marketplace and the costs are set by what people will pay. But this is a legislative requirement. Obviously, there is some discussion about what the costs are going to be once this early subsidization with Washington Publishing has gone away, and how long that lasts.
That is actually part of the reason I would like to see the codes out of there, but that's another story.
DR. LUMPKIN: I actually have a proposal for how we handle this issue, because we are looking for implementation in the year 2000 and 2001 for the current standards for which the NPRMs are going to be out sometime soon.
I'm just wondering, if it may be a little too confusing without having even gotten those adopted and in place, if we conduct hearings on how we're going to change them.
DR. MC DONALD: No, I was just simply asking what is the charge. Is it going to be $20 million, five million, a million dollars? It has never been mentioned.
DR. LUMPKIN: This was actually trying to summarize our discussion, rather than to answer your question. Maybe we need to think about this a little bit more, discuss it again at our March meeting, and then perhaps schedule hearings based upon where we see things sitting at that particular point, rather than trying to look for hearings any time sooner than that.
DR. REDDING: Sounds like a winner to me.
DR. STARFIELD: That's right.
DR. BLAIR: The only thing I'm concerned about is whether or not we really answered Steve's question to us. Have we given you the guidance in this meeting that you need in the time frame that you need it?
DR. STARFIELD: Are we going back to the last discussion item, not this one?
DR. LUMPKIN: This discussion item was the maintenance and ownership --
DR. BLAIR: I'm sorry, you can tell where my mind is.
DR. STARFIELD: I think it is actually time, if everybody accepts what John proposes, I think it is time to move on.
DR. BLAIR: I'm not sure I understood it, because I was listening to an answer to my question.
DR. LUMPKIN: I think we're going to put this discussion on the work group agenda for March full committee meeting.
DR. BLAIR: Which?
DR. LUMPKIN: The maintenance, and then let's discuss a little bit, and then perhaps at that particular time work on scheduling the content of a hearing, probably later on in that calendar year.
MS. GREENBERG: I think it would be good, probably in the next month or whatever, if we think there is going to be some kind of a hearing or meeting in April, that we select those dates as well. So I guess we'll query people.
DR. LUMPKIN: I would suspect that we have enough to do, that we ought to pick a date in April anyway. Then we can pick a subject. I think that may still be a little bit early for maintenance issues, but there are other things on our plate.
DR. STARFIELD: Okay, let's get on to those other issues. Karen, if you're ready, perhaps we can go to your item and then we can come back and discuss the unique identifier altogether, the proposed rulemaking. Karen?
DR. TRUDEL: Well, as Bill and I'm sure Bob have already told you, as of today, October 36th, we still have not published any notice of proposed rulemaking. I know Bill mentioned that the good news is that the work is primarily done. We are doing a lot of formatting and doing all the documents that need to go with the regulations.
I gather that he mentioned the status in some of the issues on the specific regulations. Did you discuss the individual identifier?
PARTICIPANT: In very vague terms.
DR. TRUDEL: And I assume you want to do that next?
DR. STARFIELD: That's the next topic.
DR. TRUDEL: Okay. Let me just quickly say what are some of the complexities that are involved in this. We have needed to coordinate with a number of other federal agencies, including the Department of Labor and the Department of the Treasury. It is a massive undertaking. Each package is well over 100 pages long, and each package contains common language, so that every time one is edited, the others all have to be edited to match.
I have given up making estimates as to when the first regulation will reach the Federal Register.
DR. STARFIELD: He gave us, too.
DR. TRUDEL: It is a credibility thing. What we anticipate is, when the first one finally does get all the way through the process, we'll have a much better idea of when the subsequent ones will be published.
Aside from that, I just think that there are very few policy issues and decisions that need to be made. Some of them are very minor. They have to do with fine-tuning language so that it meets the intent of the statute, and we are handling those things very quickly. The implementation teams have done a really incredible job.
Any questions?
DR. STARFIELD: Any questions? Thank you very much. I think we'll next turn to Jim to talk about the department's plans for the unique identifier, and that will take us into discussion that will consume us for the next hour, unless we get done early. We only have 15 minutes to discuss the computerized patient record.
DR. SCANLON: Let's see how we go.
DR. STARFIELD: We'll see how we go, yes.
DR. SCANLON: Thank you, Barbara. Well, if there was consensus on the identifier standard and most of the identifiers, there clearly was not consensus on what the unique identifier should be or whether there should be one that the Secretary should adopt.
The Data Council took this under discussion in October and tried to sort this all out. In fact, the National Committee itself recommended the Secretary not adopt the unique identifier for individuals until after privacy legislation is in place. Again, that is a somewhat unpredictable piece in itself, though it looks like there is at least some progress on the Hill.
At any rate, the leadership of the department and the Secretary has to agree to this, obviously, and she is thinking it over. But the leadership in the department thinks that this is clearly a different sort of a standard, and it has implications why beyond transaction standards per se, and even the other identifiers. In fact, it raises the issue of a national issue, social issue.
I think clearly there is a desire -- and again, I can't guarantee this will come true, but there is a desire to proceed a little more cautiously, in a little more participatory manner and a little more developmental manner in this area.
A couple of options are under consideration, and let me share them with you, because I think they involve the NCHS holding hearings in virtually every option. The options have to do both with procedure, how does one proceed to move the debate forward, and number two, what are the substantive options as well.
In one option, and I think the idea is to fall back to a more participatory, early stage notice of intent. This is not a full-blown notice of proposed rulemaking, where one says we have analyzed these and this is the proposal we're making, and this is the impact and we invite your comment. This is a notice to the concerned communities that this is the requirement in the law. These are the positive things that have been mentioned, these are the limitations, these are the concerns, these are some of the possible options, we invite your comment and any information you have relating to the feasibility, privacy concerns, cost, feasibility issues and so on.
That is an early stage of potential rulemaking, but it is a much more participatory stage. That would probably have to be coupled with some hearings in terms of what the affected communities -- again, the provider community as well as provider, payor and so on, groups we have heard from before, but we haven't really asked them to focus on this issue.
I think we would almost want to ask them to focus on what the potential candidates might be with federal privacy legislation and what they might be without federal privacy legislation.
I think at any rate, such a notice would be part of an overall plan, in terms of how do we get from the notice of intent to resolving the issue in some way, probably months later.
I think in general, the timing of this notice would be after probably the last of the notices of proposed rulemaking. I think that makes sense for a number of reasons. It would basically move the whole timetable back a little bit, but actually not that far. A notice of intent like this would have to be subsequently followed by a notice of proposed rules and then ultimately by a final rule, if that were the case. But that doesn't necessarily have to add a lt more time, given the sensitivity and the level of this sort of an issue.
But I think the department at the Data Council meetings was hoping that the National Committee -- and Kathleen attended the meeting, and heard the discussion as well, the department was hoping that the National Committee would be willing to hold hearings, whatever timing we can arrange, whatever would be the optimal timing, probably after a notice, so people would be able to respond to the various options and the various concerns raised in such a notice, an then proceed from there.
DR. STARFIELD: I think this whole issue of the unique identifier is a unifying issue for this subcommittee, because the people who are interested in community data sets are so involved in this discussion about a unique identifier, at least for health purposes. So maybe now for the first time, the whole subcommittee will come together and we'll look for input from new members as well. The privacy subcommittee is also going to discuss this tomorrow morning, is that right?
DR. SCANLON: Yes.
DR. STARFIELD: So there is going to be a need for more coordination.
DR. SCANLON: I might add that clearly, this National Committee had a quite serious analysis of all the options. My consultant to the department did a fair amount of analytical work as well. Yet, in terms of the skill of this sort of an undertaking, I don't think we feel we have a good sense of the costs or the scale of implementation might be. And of course, we can get colleagues at the Social Security Administration and others to look at analogies, but not really this thing specifically.
So there's a lot of information I think people feel they don't have in terms of baseline, what is being used now, what the level of difficulty would be or what the privacy concerns are and how you have addressed them.
DR. STARFIELD: There are some states that are using it; you need an identifier.
DR. SCANLON: Exactly. Probably most medical records have an identifying number of some kind, and it may often be the social security number. But I don't think there is a good understanding of the baseline. There is not a good understanding of the cost estimates, a lot of difficulty issues. And just the sheer scale.
The Social Security Administration sent a report to Congress, indicating what the various costs might be if they had to in essence create a social security card, and basically re-authenticate and reassign the cards by having people visit either the local Social Security office. I think the estimates were $10 billion.
We're not talking about the same level of authentication or cards necessarily, but the scale of the effort I think is comparable, 277 million Americans with additions per year.
So that is about as far as we have gotten in the department. We are hoping for a decision back from the Secretary soon to proceed, but this is the sort of thinking that is going on.
DR. STARFIELD: John?
DR. LUMPKIN: I think we have a unique opportunity to provide assistance to the department by conducting hearings. So my question would be, obviously the character of the hearing would not be -- I don't think we would want to cover the ground we have already covered, that we have a recommendation that is in place.
It would be along the lines that should the Secretary choose to adopt the unique identifier, this is the one that we would recommend, if we get to that point. Maybe we might provide clarity and recommend two, and hopefully one of them gets knocked out of the box by somebody else. But obviously, the only way to do that would be to conduct hearings and delve into it in more depth.
If we agree that that is a course of action, then the next issue becomes when would we conduct those hearings. Looking at the calendar, it looks like we have kind of been holding it the 12th and 13th of January, anyway. I wonder if that is too soon.
DR. LUMPKIN: I'm not sure we would have a notice of intent. The desire might be to have a notice to get the -- Bill and I were thinking through just the general schedule, and I think we were looking more at April, weren't we, Bill? I know April is a popular month. Possibly March?
DR. STARFIELD: Do we need to wait for the intent?
DR. LUMPKIN: I think it would make the hearings more meaningful, because there would be something to focus on, in terms of what to evaluate.
DR. STARFIELD: Karen.
DR. TRUDEL: I think that probably would be very beneficial. The identifiers implementation team has begun some preliminary work on the types of information that would go into any kind of a document that we would put forth to lay out the options and the work that is done to date. I think if we did not have that kind of information to frame the debate, it would be a lot more unfocused.
DR. STARFIELD: Okay, we're talking about April then.
DR. LUMPKIN: I just have a problem with April or May, because it is the last two months of our legislative session, and the governor prefers if I don't leave the state. I can slip away, but I don't think it would be a good plan.
DR. SCANLON: Maybe we can come to Chicago.
PARTICIPANT: We could do it in Chicago.
DR. LUMPKIN: It usually isn't that hazardous. It just, who cares -- when our legislature really cares about public health anyway, so usually it is pretty easy then. But we could do it in Chicago in April.
DR. STARFIELD: Why don't we do it in Chicago? I think it is going to be pretty hard to get anything before that.
DR. LUMPKIN: Okay, that will work for me. It is not Arizona.
MS. GREENBERG: There might be an advantage to not having it in Washington, I don't know.
DR. SCANLON: Well, there are advantages and disadvantages. I think we would like to make it easy for people to come, from the privacy community and other communities. I know you have a number of provider associations in Chicago, of course. I don't know that the privacy community -- they've got offices here in Washington. I guess we would want to make it easy for anyone to come who wanted to come.
MS. GREENBERG: That's more central.
DR. STARFIELD: Do you think this is something that you would want to have a joint thing on with the privacy subcommittee?
DR. FAWLEY: Well, Bob was talking about that, about having a joint hearing and not having the two subcommittees going off on divergent paths at all. So that is what the discussion was. I told him I would report back tomorrow morning, but the idea would be to have a joint strategy there.
I can say, the special populations committee just ended, and we spoke very briefly. We did not recommend to be doing by subcommittee. We thought hearings made sense. It has got to be the whole committee. We need to define what it is we need to do. I said I would bring that back to you.
DR. LUMPKIN: So that would be hearings before the full committee then.
MS. GREENBERG: Well, but it could be a separate date. It doesn't have to be at the time of the full committee meeting. Do you have a sense, Kathleen, having worked with a lot of people around this issue, whether -- location? Do you have a sense on location?
DR. FAWLEY: I certainly think having it in Chicago would be good, just because of the fact that there are a lot of people from the West Coast and the Western states, and it is difficult for them to get to Washington. Maybe what we would wind up having to do is maybe have some days in one location and another day here in Washington, to accommodate all of the groups.
I think the biggest problem that we have is that it is just not focused enough. You say unique identifier and people really don't know -- my concern about picking a date is, are we sure the notice of intent will be published by April. I don't know. Most of the HHS staff are going to be working on finishing the NPRMs. You figure that is going to go through the beginning of the year, a lot of that process. That doesn't put a lot of time in to getting a notice of intent out and published in the Federal Register and getting people time to start thinking about testifying.
DR. COHN: I don't know what the position of the subcommittee is on all this, but Kathleen raises a question about timing. If I were the Secretary of HHS or another -- someone who worked for them, I would be probably recommending that the notice not come out until after the whole process with the other transactions had been completed, only so as not to sully the water. I think even a notice of intent is going to cause quite a stir in this area, and I think it will get confusing to people about whether it is a notice of intent or a notice of proposed rules about the other six proposed rules. There may be some value -- once again, we have no control over this. This is not within our jurisdiction, but if I were someone trying to manage this process, I would try to keep things as simple as possible. Maybe we should seek guidance from HHS as to when they think this is really going to come out.
DR. SCANLON: You're thinking, Simon, that not only should the unique identifier for individuals notice follow the proposed rules for the standards, but even wait until they are finalized?
DR. COHN: I'm just thinking in my own mind that there is a lot of sensory input out there, and sending out six notices of proposed rulemaking and one notice of intent, it would be very easy, if I were someone who wasn't as well versed in all of this, to get a little bit confused about what was what and what the intent was of one or another.
The other rules might be inadvertently involved in the maelstrom that was really not intended.
DR. STARFIELD: John.
DR. LUMPKIN: It is going to be really hard for us to predict. One could also say that if you wait too late into the year, you're going to run into the Congressional elections and you don't want to draw up a rule like this right before then, either.
So my proposal would be to schedule a time in April. We will know at our March meeting whether or not the rule will be out. It is much easier to cancel a hearing with all of our schedules than it is to schedule one. I suspect that if we set an April hearing, we'll discuss this tomorrow whether or not it is a full committee hearing based upon logistics. I'm just concerned whether or not we want to have a subcommittee hearing, so we can assure a quorum and invite the whole committee.
MS. GREENBERG: Or we can do it under the executive subcommittee. That is what we did officially, and then a quorum is more manageable.
DR. LUMPKIN: Right, just on that issue, and then we can schedule it. Then that will give us a time, if nothing is happening, our subcommittee may choose to schedule something else for that particular time slot and location.
DR. STARFIELD: So you don't want any dates now? We'll deal with that tomorrow?
DR. LUMPKIN: Let's.
MS. GREENBERG: I think we may just -- let's see what the privacy committee thinks. We may just need to query everybody. Although we could do it tomorrow, see if we can get something tomorrow. But we may need to do it by e-mail also. There are some new members.
I just wanted to follow up on what Kathleen said. One of the things that was discussed at the Data Council, because she was saying we might have more than one hearing, was just looking possibly from the hearing point of view at the model that was done with the race and ethnicity, and have regional hearings. So we may actually need to have more than one hearing, and then having them in different places would make sense. So having the first one in Chicago would be perfectly fine.
DR. STARFIELD: And we get to go to Arizona or Anaheim.
MS. GREENBERG: You sure it won't snow in Chicago?
DR. LUMPKIN: I can give you the with the same degree of certainty that the NPRM will be that it won't snow in Chicago. High degree of probability.
DR. STARFIELD: We want to make sure that the new members get all the background on the unique identifier, that document that the contractor produced. We are actually pretty far along on it, in terms of thinking about options.
MS. GREENBERG: Did we end up putting that on the website, or did we --
DR. LUMPKIN: No, because we don't have the final version.
MS. GREENBERG: Oh, okay.
DR. STARFIELD: That is also the two new members that aren't here today, that need to be brought up to snuff on that.
MS. GREENBERG: Yes, sure.
DR. STARFIELD: Maybe we can get on to the computerized patient record. We have not initiated any discussions on the computerized patient record, except as it has come up in the course of discussions, is that correct? And it is not something we have to deal with right now. However, we have to plan how we are going to think about it.
DR. LUMPKIN: My suggestion would be, if I may, I would propose that perhaps we should start off and maybe look at the March meeting to have one panel, so that all the committee can begin to get up to speed on the conceptual background material, what we need to understand about a computerized patient record. I think a lot of us talk about it, but we're not sure how that is different than a word processor, even though we all know it is a lot more complicated.
DR. STARFIELD: The IOM came out with a report, but that is about six years old now, right? Wasn't it in 1991?
DR. COHN: It was actually just re-published.
DR. FAWLEY: It was just re-published last week.
DR. LUMPKIN: With an update.
DR. STARFIELD: Was it changed? It was changed.
DR. FAWLEY: It was updated.
MS. GREENBERG: We should get that. Can you do that?
DR. LUMPKIN: Yes, we'll get a copy.
DR. STARFIELD: Okay, so that's the first step, is for everybody to read that document.
MS. GREENBERG: There is a meeting next week on vocabulary issues that several members I think are participating in.
DR. COHN: I guess I should probably make this announcement again tomorrow, but next week, for those of you who perhaps remember, is the national terminology conference that is being partially sponsored, or in collaboration I guess with the National Committee.
DR. STARFIELD: Planned, planned collaboration.
DR. COHN: Yes, coming up with the right words here. NCHS has also been involved, HCFA, many of the other organizations here today. Certainly, if any of you want to know more about it, feel free to talk to me afterwards. But hopefully, many of the members of the committee can attend, because this should be a very interesting discussion of the things that may move into the computerized patient record discussions.
DR. STARFIELD: Okay, so that is the first step. Is there any second step that we want to take at this point? We planned to have part of our meeting devoted to this in March.
MS. GREENBERG: You are suggesting the full committee?
DR. COHN: Full committee.
DR. STARFIELD: Now, can I ask that we will not plan to add on a third day to the March meeting? You've got hearings in February and then in April. Is that enough, do you think?
MS. GREENBERG: I think unless the executive subcommittee hears from -- this maybe will not be totally decided until the executive subcommittee meeting with the subcommittee chairs, which is November 18 and 19. But I think unless there is a need identified by the subcommittees, it would probably be the two days, the 3rd and the 4th.
DR. STARFIELD: I also intend to start talking about the national data sets, but we will do that at the next meeting when we have the full complement of new members here.
MS. GREENBERG: Speaking of timing, because I did get a question, maybe we should just clear the air on this. No matter when the NPRMs are published, or whether it snows in Chicago -- no, when they are published, there still will be a 60-day comment period, because there was some concern that in order to meet the time frame, that might be compressed, and I said no. But I thought maybe we should just be official on that.
DR. LUMPKIN: No, the process will work its way in full when they are published. It will be 60 days, and that of course is when I am presuming we will have a large amount of comment and a large amount of analysis to do based on the comments.
DR. TRUDEL: Our regulations team, which has a lot of experience in this area in HCFA, is estimating that we will get upwards of 10,000 comments per package. That is not counting the individual identifier, which is, the sky is the limit.
DR. LUMPKIN: And we are looking for volunteers from this committee to help with it.
DR. SCANLON: If we get 10 volunteers, that is only a thousand per person.
DR. STARFIELD: Okay, on the computerized patient record. I would like to come back for a few minutes to the -- Marjorie, are you -- ?
MS. GREENBERG: No, fine.
DR. STARFIELD: Okay, to the core data elements. You may recall at the last meeting, we asked staff to do a crosswalk between the data dictionary and the core data elements. It turns out that there are probably only five core data elements that are not in there, at least, those that we thought were ready for implementation.
Marjorie, could you amplify on that?
MS. GREENBERG: Yes. This is a difficult area to be real definitive. I won't be any more definitive than anybody else has been about these other subjects. There are a few analyses that have been done. I think it is safe to say that the majority of elements, certainly the majority of the elements that were in the committee's core data elements report and recommended, and certainly the vast majority of those that were ready for prime time, that didn't require additional work and definitional work, et cetera, are included in these translations. Of course, that is not surprising, because it worked a lot with the billing committees, who currently maintain content. Also, they are just basic things that everybody has to collect.
Now, we have identified -- we did some work at NCHS, actually looking at the implementation guides. I would say it is almost impossible to figure this out from the data dictionaries. We are not talking about even that the same categories are used as were recommended by the committee. We are just talking about the data element, pretty much meaning the same thing being present, because as we discussed a little earlier, HL7 and X12 and everybody uses something different on some very basic elements, like marital status or even gender.
But just to find out whether the elements are even there is hard to do from the data dictionary. On some of them, you have to go into the implementation guide, or you find the element maybe in the data dictionary and then you go into the implementation guide, and just see what really is meant there.
In doing that, in the last few days we identified another element. I have raised a question about it, but I don't know what the definitive answer is. But external cause of injury, although it is included in the implementation guide for the institutional bill, and as we know it was added to the UB-92, so that form locator identified as being a department, it has an odd identification. But actually, that form locator that is on the UB-92 is in the implementation guide for the institutional claim, for the external cause of injury.
It turns out that even though the National Uniform Claims Committee in their report to the department and to the committee said that they were including external cause of injury in their data set as one of the elements in the core data elements that they did include that wasn't already included, we could not find it anywhere in the implementation guide. So that would be --
DR. MOORE: There are some issues that -- we are still looking at the code sets and the elements that go into this. The work is not done. As far as code sets, both X12 and those individual work groups, whether it is enrollment or any one of the claims transactions, et cetera, I think there is still activity going on.
I know just a couple of weeks ago, the team came back and said they were looking at the weight that is identified in the claim, and they are making a change and worked with X12 to include newborns as part of that, when the claim is for newborns, to identify it.
So there is an effort to cross-walk what came out of the core data set to include it in these transactions, and to get the different code sets. What I think that there are some differences on are those areas like functional status, or things that this committee even decided that it was not ready yet, that those are not in there. Do they belong in some kind of attachment or do they belong in a claim? I think we need to wrestle with that. Are they a claim element or are they an attachment element?
DR. STARFIELD: Well, I think we do have to wrestle with it.
MS. GREENBERG: These are the ones that we identified, that don't seem to be there, but as you said, with newborn birth weight, there is a weight variable, so it is possible that could be worked in.
The other ones are years of schooling or education, which would be in -- if collected, probably was recommended for an enrollment data set, not for a claim, or not for the encounter data set. So that is the enrollment. And there is a separate implementation guide for that.
In fact, as the NPRM is currently written, the part on the enrollment and eligibility actually poses the issue of whether some additional socioeconomic or demographic elements should be included in that transaction. There are some who maintain that it should and others who say that it shouldn't. So without specifying particular elements, there is some discussion of that even in the NPRM.
The others are, on the institutional claim, the qualifier for secondary diagnoses. That has been implemented. That is being used in at least two states that we know of, New York and California, and that was pretty much ready for use, if it was chosen to be included. That is whether or not the other diagnoses other than the principal -- this is in-patient, whether the onset was before admission, after admission or unknown. So that one doesn't appear to be anywhere in that transaction.
We also have the schooling, we've got the newborn birth weight, but it could possibly be included. We've got the qualifier. I've got this written down somewhere.
DR. STARFIELD: It is the location of the encounter.
MS. GREENBERG: Right, it is the location of the encounter if it is different than the provider's billing address, again, something that should be available and I think was -- so it would be ready to implement, and can be used for access issues. And the disposition of the patient on an out-patient basis.
This has been in the -- and it would be in the institutional claim if it were used for an out-patient visit, or it would be on the professional claim.
DR. MOORE: If it were an institutional out-patient visit, it would be there. If it were a physician service, it wouldn't be there, or a DME claim or a lab claim or something of that nature, which is a 1500.
MS. GREENBERG: This is something that has been collected in ambulatory care surveys, and it has been in the uniform ambulatory concept for years. It was never implemented into the HCFA 1500. I think as the committee recommended, it was pretty basic, even alive or dead. Out-patient has advanced to the extent that it is not quite like it used to be. You could have that possibility.
DR. STARFIELD: The relationship between principles of care and secondary care is a big issue now in the health care system, in managed care. You need to have that disposition.
MS. GREENBERG: So those are the ones that were identified. Some of them are apparently -- we have identified them to the team or to HCFA, and there has been some discussion about this at the Data Council as to whether some of them should be added prior to -- nobody wants to slow down getting these NPRMs out, which as we have heard, there are enough other things to slow them down.
I think even people have talked about working through the standards process to try to get these into the appropriate transaction standards over the next year or whatever the process is, and then since these aren't going to be implemented until the year 2000 anyway, whether they might actually be available, or at least the following year or something; there are different ways of approaching this.
There are also the elements that weren't identified as being ready for prime time. We have had some concern that although the important ones -- and although the committee recommended to the Data Council that some work be done on them, obviously there has been a lot of work -- it is not that people in the department have had down time, there is no doubt about it, but there is no process really for advancing these, such as functional status and some of the others that were recommended. We couldn't go forward now to X12 if we wanted to with a consensus recommendation to include it, because they needed more work. At least there probably is work going on in various places just related to these, but there is not any kind of coordinated effort going on, trying to advance those.
So I think those are as I see it the outstanding issues on those elements.
DR. STARFIELD: My concern is, I'm beginning to get phone calls saying, you guys aren't doing your job. All of a sudden we are losing items of information, like birth weight and social security number. It is hard to know whether they are there or not there or what they are there for and what they are not there for. So we really should I think be pretty clear at least about core data elements.
DR. MOORE: At my last count, there were 42 core data elements that this group put forward.
DR. STARFIELD: Ready for implementation? No, not ready.
DR. MOORE: No, 42 total.
MS. GREENBERG: Forty-two total. They weren't all ready for implementation.
DR. MOORE: And there were 35 of them included in those standards, and there is consideration of others being added, like the low birth weight to be added, because there is a place for weight.
I know that the committee made the decision that if it is a mother's claim, the low birth weight of the baby should be put on the claim, but I'm not sure how that would be done. I'm not sure how we would go about telling providers to include that, and how we would capture it back.
It is like sometimes in HCFA, our policy people tell us to do something, but it is an administrative nightmare to get it done in an operating setting. It is far more difficult. Things like the type of encounter. We do have the type of service, as well as the place of service. Some of these things that you have I think are derived data. When you look at more than one piece of information, you get an answer that fills in some of this. We do the same thing when we make information out of data. Two pieces of information form the third. You don't need to carry the third to make it a separate data element.
So I think maybe some of these things can be revisited after we get the -- I would say probably spring of next year, like February. Once we get more definitive code sets identified and get through this initial period, working with X12 and other bodies as to what are the code sets for each of the data elements and get it out there, then we can come back and revisit it in a more -- fashion. But I think we are meeting 85 percent, my estimate, of what the committee asked for in the different codes.
MS. GREENBERG: I started out by saying that the majority are there.
DR. MOORE: I can't give you all.
MS. GREENBERG: But in fact, I think the added value that the committee -- because there wasn't that much value probably, relatively, in the committee recommending things that were already in the uniform claim. They were there, they appreciated your support for them, but I think the added value presumably was in some of the additional areas that were recommended. For the most part, they are not included. I think there are reasons why they aren't. I think there are approaches to trying to rectify that.
I am concerned about this latest thing with the external cause of injury, because I think if they come back and say, you can always include that, it is one of the ICD-9 CM codes. That was the battle that we fought over a period of years with NUBC, and it was agreed to add a separate element. In fact, if you look at the NUCC's report, it does acknowledge that it would be a separate element, and now it doesn't seem to be there.
So that kind of concerns me, because it is a very basic public health issue, and violence and injury as we know has been pretty much accepted as a major national issue.
But I'm just reporting what our analysis has shown.
DR. STARFIELD: But I think at least we have to be clear within the committee as to what is there, what is not there, what you have to derive, but it is derivable, so that we can answer questions from the constituency that is the public. We have to really know what is possible to get out of it. So if we are working toward that, I think that's good.
DR. MOORE: I guess I was mistaken, because when I worked with you all on the initial development of these, I didn't know they came from someone outside. I thought some of these were laid on the table by the work group, as like to have as well as necessary. I don't know where you are getting the pressure from outside, Barbara.
DR. STARFIELD: I'm getting calls from people, saying they have a great concern that we are removing things that are on the UB-92, and they are just not there anymore.
DR. MOORE: I've not heard anyone -- if there is any complaint, I think some of the complaints are that we are requiring more in the 837 than we ever required in the UB-92. That is the reverse of it.
MS. GREENBERG: I think it is the state elements that may be of some concern about that.
DR. MOORE: It is the statistical elements.
MS. GREENBERG: One of that is the qualifier for secondary diagnoses. Now it is being captured by New York State on the UB-92 as one of their optional elements.
DR. MOORE: And some of the states are requiring things as we speak. Kathy just leaned over and said that Massachusetts is requiring a code for what, language?
DR. COLTIN: It's part of our contracts with our state Medicaid agencies, that we have to provide enrollment literature. I think it is going to be in the consumer bill of rights thing, too, to people in their primary language. So if you're telling them what their benefits are and how do use the health care system and so forth, that you do it in the language that they are able to read.
So in the enrollment, we need to collect what that language is, so that we then know what to send out.
MS. GREENBERG: That was one of the recommendations that wasn't included in the core data elements, but language was one of them.
This morning, we talked about quality, and how these things come together. Lisa has testified to the Quality Commission, and has said on many occasions that adding that qualifier for the secondary diagnoses for hospital data would really be a major improvement in being able to use hospital level data for severity and other kinds of risk adjustment, et cetera.
So I think it is not irrelevant. At the same time, I think we have accomplished an enormous amount in a short period of time, so I would hope we have a plan. We look at some of these and say, we recommended them two years ago, but we don't really think we need them anymore, or that we're not going to push for them. But where they aren't there and they aren't being collected and they currently can't be derived, then I think it would be good to have some kind of a plan.
DR. MOORE: I guess I tend to think that we do have a plan. When it is identified -- and I don't think that when we did the core data elements before, we had this HIPAA legislation, that the entire community was as well aware of what the committee had asked for. I think that since we have -- in the past year, this committee has had a greater impact on that community, and a lot of this information has gotten out there, and we have worked with the community.
So I think that we do know the areas that we have gaps and where those gaps are, and that there is a genuine attempt to bring that into the full -- if you will, in one of the elements. That is why it was considered, like low birth weight of the baby to be added, because the weight was already there. But when we started this, when we looked at the UB-92 as it is constituted on paper, we didn't see weight on it. But it is in the 837, and we were surprised when we found it there ourselves.
DR. COHN: This discussion is very interesting, first of all.
MS. GREENBERG: It wasn't intended to be a debate.
DR. COHN: And it clearly speaks first of all of some of the complaints we have had about the data dictionary work that had been done up until now, only in the sense that you're looking for data elements and you can't tell whether they are there or not. Obviously, the point of a data dictionary is to make these questions easy to get the answers to. That is really why we have spoken of a data model and these other approaches, so we can figure these questions out.
Now, the other piece of it that I just want to bring to everyone's attention in this committee is the fact that we are obviously a little late in terms of discussions about what should be in the current or proposed data standards. But this question, if there are things that aren't in there, really fall under a previous discussion that we had around ownership and maintenance of data standards. It really brings once again to fore this primary issue of, well, if it isn't there, how in the heck are we or anyone else in the world going to get things included or changed? It really speaks to the fact that this really is an important issue.
DR. STARFIELD: I think we do have a game plan. Are there any other things that we need to discuss?
MS. GREENBERG: Just to say that the Public Health Service obviously -- several agencies are interested in this, and have raised this, and are thinking in terms of having a workshop and bringing in public health people, and maybe others, to try to involve people more and get people more aware of what the processes are. I think Lisa said in this Quality Commission, they didn't have a clue about HIPAA, they had never heard of it. So do we want to load everything into a claim, or we don't want to load.
I think people do need to be aware of the process. If we are going to rely on the private sector processes with public sector participation has we have in the past, it is going to be important to -- as we address issues like how do you get content changed and added and all of that, that more people be aware of this. They are going to have to probably participate in the same processes.
This is going to be the process, then people who haven't participated in the past are going to need to understand it better, and to come together and try to prioritize what is important. This is a hard thing to get, as everybody knows. But I think it will be easier for many different constituencies to maybe advance five elements than to be advancing a hundred different elements.
So I think this is a large public education and outreach activity that obviously we haven't really been able to even touch the surface on. But as we are thinking about this internally, we very much welcome input from the committee.
DR. MC DONALD: You sound like you're saying people have these other kind of goals, and they are going to try to use the claims mechanism to pick them up. Is that what I'm hearing?
MS. GREENBERG: Not necessarily. This is called the claim or encounter. We are talking about information that has multiple uses.
DR. MC DONALD: No, I understand that. What I am just concerned about is the idea -- well, two things. To distinguish between a standard that gives you a place to put it, I think it requires that people fill it out. There are two different issues there. The minimum you might want is to have it there. If you make everybody fill it out, you're going to get a huge backlash.
MS. GREENBERG: True. This is the case with external cause on the UB-92. It wasn't actually required for anybody by any payor. But the fact that it was there allowed it to begin to be collected in a standardized way. A number of states then did -- that were already trying to get it into their databases, were able to include it in their databases, because it was much easier for the hospitals to just put it on the UB-92 than to have to submit it separately.
So I think requirement is another issue, though I don't really know how this is all working out with the 837, because I think we're not going to allow individual arrangements anymore between providers. The idea I guess is, if you submit everything on there, you have met the requirements.
Now, some payors won't use everything. But I guess if you haven't submitted everything on there, you haven't met the requirements.
DR. MC DONALD: No, not on 837.
MS. GREENBERG: Everything that is relevant. I don't know how this works.
DR. MC DONALD: But the other side of that was that I think there are other ways to get standard messages created than trying to pile them onto the claim. That is what I want to clarify. If people are going to do immunization reporting or they are going to do other -- or statistic reporting, they really need a separate activity to think that through, and design it and publish it. The claim won't be able to carry the weight of all the goals.
MS. GREENBERG: I think that is what people need to think through. It is just like with our surveys at NCHS. Does it make the most sense -- can you collect this through a health interview survey? No. You ask people, they don't know the answer. You have got to use the appropriate mechanism.
So are there elements that people want that can appropriately be included on an enrollment form or on an enrollment transaction or on an encounter transaction, or no, you need some other type of thing. Those are the types of issues that people have to be thinking about. I agree with you.
DR. STARFIELD: I guess Steve Jenks made that point, what problems do you want to solve, and then what is the best way to solve them. I think that is a good way to look at it.
I take it that nobody would be upset if we adjourned? Don't need a motion for that.
(Whereupon, the meeting was adjourned at 4:56 p.m.)