[This Transcript is Unedited]



Meeting of:

The Subcommittee on Standards and Security

December 10, 2003

Hubert H. Humphrey Building Room 705A 200 Independence Avenue, SW Washington, DC 20201

Proceedings By: CASET Associates, Ltd. 10201 Lee Highway, 160 Fairfax, VA 22030 (703) 352-0091


P R O C E E D I N G S [8:30 a.m.]

Agenda Item: Call to Order – Dr. Cohn DR. COHN: Okay, would everyone please be seated? We’re going to get started. Good morning, would everyone please be seated so we can get started this morning? I want to call this meeting to order, this is the second day of hearings of the Subcommittee on Standards and Security of the National Committee on Vital and Health Statistics. The committee is the main public advisory committee to the U.S. Department of Health and Human Services on national health information policy. I apologize for the gravelly voice but it will get better as we go on in the morning. I am Simon Cohn, chairman of the subcommittee, national director for health information policy for Kaiser Permanente. I want to welcome fellow subcommittee members, HHS staff and others here in person. I also want to welcome those listening in on the internet and obviously want to as I always do remind everyone to speak clearly and into the microphone so the people on the internet can hear us. Actually today we have an abbreviated session even though we’re starting a little bit earlier we expect to adjourn about 12:30. This morning we begin with a HIPAA update, this will be followed by an update on the claims attachment standard, and we’ll probably evolve into a discussion around the Notice of Proposed Rule Making, or planned Notice of Proposed Rule Making. We’ll also be hearing about a pilot related to that standard and then they’ll be a time for various public reactions and discussions by the subcommittee on the claims attachment standard. Additionally later on this morning, as many of you know the recently approved Medicare reform legislation calls on the Secretary to adopt standards for E-Prescribing, and the NCVHS has been directed to developed such standards recommendations. We will be discussing a draft project plan on that effort late this morning and we will obviously defer refining that over successive subcommittee meetings. Obviously I want to thank Jeff Blair our vice chair for developing that work plan. I want to emphasize as I always do this is an open session, we’re always happy to take public comment on any of the issues coming before us either from those in attendance as long as the comments are relevant to the topic we’re describing. Obviously for those on the internet or in the public at large we’re more then happy to take emails or letters on any of the topics that we’re considering. With that let’s have introductions around the room, actually around the table and then around the room. For those on the subcommittee if there are issues coming before us for which you need to publicly recuse yourself please so state in your introduction. With that, Jeff, would you like to start out? MR. BLAIR: Yes, thank you. Jeff Blair, vice president, Medical Records Institute, vice chair of the Subcommittee on Standards and Security and I’m a member of AMIA, ASTM, HL7, and HIMSS, and there’s nothing that I’m aware of that I need to recuse myself from. DR. STEINDEL: Steve Steindel, Centers for Disease Control and Prevention, staff to the subcommittee and liaison to the full committee. MR. SCANLON: Good morning, I’m Jim Scanlon, I’m executive staff director for the full committee and I’m with the Office of Science and Data Policy here in HHS. DR. HUFF: I’m Stan Huff, a member of the committee, I’m with Intermountain Health Care and the University of Utah in Salt Lake City. I need to recuse myself if we talk about any issues related to HL7 or to LOINC or any business related to 3M. DR. FITZMAURICE: I’m Michael Fitzmaurice, I’m senior science advisor for information technology to the director to the Agency for Healthcare Research and Quality. I’m staff to the subcommittee and liaison to the full committee. MS. GRAHAM: Gail Graham, Department of Veterans Affairs, staff to the subcommittee. MS. RYDER: Suzan Ryder, I work for Empire Medicare Services, a division of Empire Blue Cross/Blue Shield. MR. RISHEL: I’m Wes Rishel, the firm that pays me is Gartner but most of the time I spend seems to be with my board appointments to HL7, HIMSS, WEDI, and the EL(?) Initiative. MS. AULD: Vivian Auld, National Library of Medicine, staff to the subcommittee. MS. GREENBERG: Marjorie Greenberg, National Center for Health Statistics, CDC, and executive secretary to the committee. MS. FRIEDMAN: Maria Friedman, Centers for Medicare and Medicaid Services, staff to the subcommittee. MS. TRUDEL: Karen Trudel, Centers for Medicare and Medicaid Services. MS. SQUIRE: Marietta Squire, CDC, NCHS, and staff to the subcommittee. MS. GUILFOY(?): Helene Guilfoy, Guilfoy Consulting. MS. GRECCO(?): Kim Grecco, Hogan and Hartson. MS. LESH: Kathy Lesh, the Kevar Company. MR. KILE: Frank Kile, the American Dental Association. MS. BRADFORD: Alicia Bradford, CMS. MS. BUSHMAN(?): Mary Lynn Bushman, Empire Medicare Services. DR. COHN: I understand that there’s actually two individuals who have called in via our line, could you please introduce yourselves? DR. ZUBELDIA: Kepa Zubeldia with Claredi Corporation, member of the committee and subcommittee, and as far as conflicts I may be participating in the pilot that Sue Ryder will present later, so that could potentially be a conflict. DR. COHN: Is there any other persons calling in? Well, Kepa, good morning, thank you for joining us, I realize you’re in San Diego, probably a good choice, but we’re happy you can join us for this morning. I think we’re going to be leading off with Karen Trudel, who’s going to be presenting an update on HIPAA and that probably fits the tone for the day. Agenda Item: HIPAA Update – Ms. Trudel MS. TRUDEL: Well, I’ll be brief, perhaps that will set the tone, too. I’m going to talk about three things, one is the status of regulations, another is just shortly an overview of what I’m personally hearing about industry readiness, and then third a little bit about enforcement and the enforcement process, complaint that we’re getting to date, etc. With respect to the regulations we are still on target to publish the National Provider ID Final Rule on December 20, well, now, it won’t be December 26th because the federal government will be closed, in any case, late December is the publication for the National Provider Identifier Final Rule. We are beginning the review process on the claims attachment proposed rule with the CDA approach that Sue and Wes will be talking about a little later. And the plan identifier also continues to be in the review process. We are beginning to establish a strategy on proposed rule for ICD-10 with special attention going to the notion of the impact analysis to make sure that we can use the information that we got from the RAND study, augment that with other information and really come up with something that will be very solid, that the industry will be able to comment on hopefully very concretely. We’re also working on the next modifications regulation, which will include as far as I know some updates that have to do with responses to the Secretary’s Committee on Regulatory Reform, some suggestions were made there for improving definitions and clarifying some information about direct data entry, etc., so we’re working on that as well. Moving on to readiness, the latest Medicare statistics that I’ve seen show that the percentage of HIPAA compliant claims in production is now around 50 percent, the carriers, the fiscal intermediaries are working very intensively with providers and clearinghouses to get their testing done. We continue to see an upward trend and there are a number of problems as you would imagine with something this complex. We’ve been meeting periodically with representatives of various health plans and it appears that we’re all hearing the same kinds of things, so this is just a complicated process that we need to work our way through, and of course the contingency gives us an opportunity to do that without harming cash flow or operations. Moving on to enforcement, we have received just a handful of complaints, I think at the last time I checked it was maybe about ten that were valid complaints that had to do with transactions and code sets, and our staff is working through that process, getting back to the people who filed the complaints, back to the people against whom the complaints were filed, trying to work through either coming up with corrective action plans or providing technical assistance to the covered entities, so that process is ongoing and at this point the workload is quite manageable. I think that’s about all I have to report, I’ll be glad to take questions. DR. COHN: I think Mike and then Jeff. DR. FITZMAURICE: Karen, I think you mentioned that you’ve met with health plans and have heard all kinds of things. Were the kinds of things you heard the same areas of difficulty that the Medicare program as a health plan itself is facing and what were some of those kinds of things? MS. TRUDEL: It’s just a matter of working through problems one by one, it seems that various submitters are having, I mean the problems are just all over the board, you can’t say it’s just ten data elements or it’s one thing, it seems to be just a complete mixed bag and that’s what we’re seeing, too. DR. FITZMAURICE: Can you say that it’s more professional providers, that is physicians rather then hospitals? Can you say that it’s more vendor issues and not having the right software rather then inability for clearinghouses to get the formats right? MS. TRUDEL: I would hesitate to make any kind of generalizations at all. DR. FITZMAURICE: Is there somebody from the Medicare program who could talk about the issues from the standpoint of the Medicare health plan as compared with other health plans? MS. TRUDEL: Yeah, we could certainly get someone from the Office of Information Services to come and do that at the next meeting if you’d like. DR. FITZMAURICE: A second question, that is HIPAA was put out to improve efficiency through the use of electronic transactions, and a lot of people in the industry have told me that they would like to see an internet policy come out that would allow them to submit claims using the internet, feeling that that would be one of the most cost efficient methods they could have for transmitting claims. For the past year or more that issue has surfaced and is there a plan policy that would permit contractors and the people who send their claims into the Medicare contractors to use the internet for such purposes and what is the biggest obstacle for that? MS. TRUDEL: Again, I’m sorry Mike, I can’t really answer that, that’s in the purview of our chief information officer and I’m not familiar with what’s going on in that area right now. And again, we can arrange to have someone come and present to that if that’s something that the subcommittee is interested in. DR. FITZMAURICE: Thank you, Karen. DR. COHN: Jeff? MR. BLAIR: Karen, I read something in the Trade Press and it indicated that there was a percentage figure for compliance of, in other words the number of claims that are received that comply with the HIPAA financial transactions that HHS has, and I don’t want to quote what that percentage is, I just want to ask you is there a threshold that has been set forth and when we reach that threshold that is when HHS will then wind up saying okay, we’ll give the industry a certain amount of more time? Or is that still being something where you’re seeing how things go and leaving all options open? MS. TRUDEL: I guess the easy answer is we haven’t set a specific percentage like we get up to 85 percent, that triggers all these different actions. What we’re really looking at is how our progress is going and what the curve looks like, so I think that’s really probably the more important way to look at it. And the other thing I want to stress is that we won’t be pulling the plug on the overnight anyway, we will make sure that we give providers and submitters ample notice that as of a certain date the contingency would be lifted. MR. BLAIR: Thank you. DR. COHN: Stan and then Kepa. DR. HUFF: A quick question. I’ve been recommending the RAND study as reading for friends and relatives, but I had a hard time finding an electronic copy on the web. Has it been posted somewhere? Where is that? MS. TRUDEL: Yeah. MS. GREENBERG: It certainly will be posted on the web. In order to, as I think I mentioned at a previous meeting, we need to bring the contractual relationship to RAND to a close because we had extended the work and extended the time period, and so we are waiting for the final electronic version. There’s paperwork going back and forth, so until that’s concluded we’re not going to have the final electronic version to post on the web and I expect it very, very soon but we just don’t have the final version at this point. DR. COHN: So in other words you’re seeking a version that doesn’t have draft in light print across every page and things like this. MS. GREENBERG: Well, if we don’t get this, it’s not RAND’s fault, I mean it’s a question of just getting the paperwork through both the government and RAND but we should have it very soon. DR. COHN: Mike on this point or a different point? DR. FITZMAURICE: It’s a follow-up to what Stan asked. DR. COHN: Okay. DR. FITZMAURICE: In the last meeting of the full meeting of NCVHS the NCVHS approved the ASCA report and I had the same experience trying to refer someone to the NCVHS website for it. Did I miss it or is it not there yet? MS. GREENBERG: I can’t check the website at this moment but I will, you provided it to, we’ll make sure that it gets posted, it should be posted and in fact we have the annual HIPAA report references that it’s posted. So there shouldn’t be anything delaying that, I apologize — DR. COHN: Kepa? DR. ZUBELDIA: Thank you. First I would like to everybody to speak directly and very close to the microphone, it’s very, very hard to hear. And because of that I missed Karen’s number, whether you are at 50 or 60 percent. MS. TRUDEL: It’s at about 50 right now Kepa. DR. ZUBELDIA: Karen, are you also tracking the number of submitters and the number of providers that are sending electronic claims or CMS is only tracking the number of claims? MS. TRUDEL: We’re tracking the number of claims and the number of submitters, not the number of providers. DR. ZUBELDIA: Do you know approximately what percentage of submitters are sending the HIPAA transactions? MS. TRUDEL: Not off the top of my head but we get a weekly report on that, we are looking at it. DR. ZUBELDIA: And how about the other transactions, how is that going? MS. TRUDEL: Those are somewhat lower, we’re especially having issues with the coordination and benefits transaction, there is a need to coordinate and make sure that both the sender and the receiver understand exactly what is in the transaction, what kinds of editing is being done, there are problems with gap filling when we get a transaction that’s not HIPAA and we cross it over, there’s a need to fill some gaps in order to make a compliant transaction, and the rules for that are causing some consternation but we’re working that through basically trading partner by trading partner and I understand that we’re making some really good progress there. DR. ZUBELDIA: Thank you. DR. COHN: Michael? DR. FITZMAURICE: Are there any coordination of benefit claims coming through electronically and is there difficulty with those in particular compared with the other transactions? MS. TRUDEL: From Medicare perspective the coordination of benefit transactions, are you talking about MSP claims? Medicare’s Secondary Payer? DR. FITZMAURICE: That’s part of it, yes, the other part is Medicare Primary Payer and then going to other insurance companies. MS. TRUDEL: Right, the crossing over at the back end is what’s the problem is and it’s just making sure that everyone understands what rules we followed to build what we thought was a compliant transaction. But yeah, we are sending crossover files in production. DR. COHN: Helene, please introduce yourself. MS. GUILFOY: Karen, are those rules of what you’re requiring to be sent forward in COB published somewhere? Because the only thing that the providers have had to work with is an IG, which is different then what the Medicare contractors are requiring. In other words the Medicare contractors are not requiring a fully formatted fully data content claim as the primary claim going to Medicare, yet the payers on the COB side are yet requiring that. So if a provider submits a fully compliant fully content rich claim to Medicare then it can be sent forward but if a provider is submitting what the Medicare contractors are requesting then you don’t have the information to submit a compliant claim forward to the secondary payer. MS. TRUDEL: We are building compliant claims for crossover, even when they come in on paper. MS. GUILFOY: I’m not talking about paper, I’m talking about electronics. MS. TRUDEL: Or electronic in old formats, it doesn’t matter, we’re gap filling or expanding, we’re doing whatever we need to do to build a compliant crossover claim even though the incoming claim from the provider was not. So there aren’t any requirements on the provider in this matter, this is something that — MS. GUILFOY: I guess my question is how can you back fill when the information is just not available in the electronic format. We can talk about this later but — MS. TRUDEL: I think it’s also an issue that’s outside of my purview and more in the purview of the EDI folks at CMS. MS. GUILFOY: So is there somebody that — MS. TRUDEL: Yeah, I’ll give you a contact. MS. GUILFOY: Super. Thank you. DR. COHN: We may have to schedule a session with the CIO or his representative to talk about some of these issues. Other questions before we move on? Karen, thank you very much, and I’m pleased that progress is being made and we’re moving forward. Now obviously the next part of the discussions really focus on claims attachment, the purpose here is obviously to be briefed on exactly what’s happening. The NCVHS as part of its responsibilities needs to respond to any Notice of Proposed Rule Making, which will eventually be produced about claims attachments, so it’s important obviously that we be informed but this also has a secondary value of obviously informing the greater community at large. So Wes, we’re obviously happy to have you start the discussion and my understanding is then Suzan you’ll be continuing on and talking about the pilot, I don’t know if it’s underway or being planned but I think we’ll find out from you which is which. So Wes, thank you, and by the way thank you for coming all the way from California. Agenda Item: Claims Attachment Update – Mr. Rishel MR. RISHEL: Kepa, can you hear me okay? DR. ZUBELDIA: Can you speak a little bit closer? MR. RISHEL: If you hear a bumping sound it’s me against the microphone. I am here today with several hats on, it’s fortunate that I have so many heads to hold the hats. Speaking as a representative of HL7, also as the primary author of this approach that we’re describing today. I have several goals to accomplish in the time we have available today, clearly to inform the subcommittee on the revised approach and how it relates to the prior approach to claims attachments, this was done in 1998, and some related policy issues associated with this approach. The second is to solicit the support of the NCVHS with the Secretary in order to gain experience and maintain the momentum on attachments, and I’ll talk more specifically about that as we go forward. Finally, what I’m going to be describing here is a dual level approach, that is it can be very simple or it can be more complex, but provide more structured data, and this parallelism of this duality of this approach is the primary point that I want to bring home today during the meeting, I’d like to suggest that the considerations that lead to this approach for claims attachments are also generally relevant to the NHII and I’d like to have the committee consider this dual approach moving forward in other contexts. We’ll start with a brief introduction, talk about the underlying standard, which is the HL7 clinical document architecture, talk about how that’s applied to attachments, and then finally give the timeline and summarize what I’m trying to accomplish today. The introductory material is purposefully light because you’ve heard this before as a subcommittee, however, in the handout there are slides that provide more detail on the history of the approach to claims attachments. Just in general I sort of, I used to, this slide used to say big issues and addressable issues, but since the security reg came out and sort of preempted the term addressable I’ve changed it to manageable issues. Clearly there’s a lot of concern and stress in the industry around claims attachments. And there are payers who want no ability to make the attachment predictable to the provider, they want to be able to ask whatever they want, and providers who want to requirements for attachments. Nothing in our proposal deals with these issues in any way whatsoever, those are policy issues that the Gods on the high mountains deal with those issues. The issues that we are addressing are predictable content for the attachments, providing some help for the payer problem of providers sending the entire chart as opposed to sending data specific to the claim, dealing with the issues that’s been discussed considerably in the industry lately, which is the advantages of a very unstructured approach that can get started quickly versus the advantage of a more structure approach that can create more efficiencies in the form of auto adjudication. The coding system for attachment questions, syntax, and how this might apply to other communication models besides simply batch EDI transactions. I’ll take questions at any time. Just a comment, whatever approach we use here is likely to also be used to provide supporting information for pre-cert or pre-authorization requests, the thinking is running that way. In the past, and as documented at the end of the presentation, there was a substantial industry outreach performed in I guess 1998 or 1997 to determine the most frequently used attachments. One of the conclusions was to, let me restate that, the conclusion was it was necessary to determine the most frequently used attachments and that work went on. There is consideration given to where HL7 messages already exist, the need to standardize the questions that the payers ask on an industry consensus basis, because it’s only when the content is predictable that the providers can improve their processes and collect the necessary information up front and so forth. Form attachment workgroups and use LOINC codes rather then another code set that I’m going to let someone else tell you, but the concern was that the existing code set was not granular enough to be specific on the questions that we’re asking and the LOINC committee volunteered to take the identified questions and identified attachment types and assign LOINC codes to them. Specifically LOINC was not in any sense deciding what were the valid attachments or what were the valid questions on the attachments, they’re simply doing an informatics job of taking those concepts and assigning LOINC codes to them. The existing standards and recommendations that comprise the old approach, and this information pretty well carries through to the new approach, provide for two work flows, one is unsolicited where a provider submits a claim with supporting documentation, this happens in the paper world where they literally staple the attachment to the claim and send it in and the metaphor is an electronic staple combining the two together. If the information is sufficient to decide whether to pay or not then the payer will pay the claim, otherwise they’ll deny or otherwise get back to the provider saying they need more information. The second workflow is where a claim comes in and there’s a need for more information in order to pay and in that case there’s a request that goes back to the provider in an EDI manner, the provider assembles the supporting documentation, sends it in using the same transaction that would have been stapled to a claim in the prior workflow, and then there’s an adjudication determination based on that information. The choreography between the 277 and the 275, the request and the response, are that the 277 asks for attachments for some of the questions off an attachment, so a 277 might ask for the ambulance attachment, which is a list of probably a dozen questions, or it might only ask what was the reason why the ambulance bypassed the nearest provider for this run, which is one of the specific questions that’s in the last. And LOINC codes are sent out in the 277 to either identify the attachment or the questions within the attachment according to the specifics. The balance of the choreography is the 275 sends back the components, which represents all of the questions that the entire attachment was asked, or some of the questions if they were asked specifically, and the components can be broken down into parts, so in order to get structured data where it says physician name and address they’ll be a code to identify the physician name and a code to identify the physician address and so forth so there’s a hierarchy of attachment components and then the lowest level in the hierarch is parts of a single component. All of these are identified by LOINC codes again. The original proposal was prepared in the period 1997, 1999, under considerable pressure to get it out in time for the NPRM, which was due, sorry, I can’t avoid saying that, and wouldn’t if I could, in that proposal the 275 contained an imbedded HL7 version two results message, which is a very generalized message, it can contain anything from a lab result to an op note to an image depending on how it’s used. The HL7 syntax is similar to X-12 but different, there is limited support for free text and images of text, in other words you take a page from a chart, copy it and send it essentially at fax quality imbedded in the transactions. There was six attachment types identified and the NPRM didn’t forth come. There have been some changes in the technology, changes in the industry, we saw the delay as an opportunity to move forward and to deal with some issues that we’ve become increasingly aware of, both through informal dialogue on attachments and through the experience we’ve had implementing other HIPAA transactions. In the new proposal we take out that version two HL7 message and put in an XML document instead, but the content in the current proposal is the same as the content that was proposed in 1999, that is, if I can remember all six dwarfs here, lab, generalized transcribed reports, any kind of a transcribed report, medications, rehabilitation plans and I think there’s eight or nine specific kinds of rehabilitative services, minor differences in the format for those, emergency department reports and ambulance. The committee has made considerable work since then on home health and durable medical equipment but our current understanding, the plan is not to include those in the original NPRM but once the mechanism is established adding additional attachment types is technically a very straightforward job. There are policy issues about whether that needs an NPRM or not but the technical part of the job either to define it or to implement it is very straightforward once you’ve built the basic machinery for attachments. This issue of structured data is a fundamental problem we have in health care, which is is it better to get the information there more, but have it be verbal information, either free text or a fax like image of some text that was either written or typed, or is it better to require more implementation to get structured data in front of the adjudicating system so there can be additional efficiencies associated with adjudication. One of the biggest concerns about attachments in general is the lack of an electronic medical record system in many provider organizations, which means that the data certainly isn’t available in a structured form, ready to be sent to the payer in a way that they could auto adjudicate. At the same time all payers that I’ve talked to have only, at best only limited plans and usually no plans to auto adjudicate based on a claims attachment, so their job will be electronically to do what happens on paper, which is get that information in front of a person who will make the auto adjudication decision. Nonetheless, some health plans are telling us that for specific attachment types the adjudication decision is very simple and sort of the most dramatic example of that is attachments for periodontal procedures, generally they only look at pocket depth, one number they look at and they make the decision. So while the general problem of auto adjudicating claims attachments is extraordinarily difficult there are specific problems where it could be extraordinarily easy. And so an approach that said let’s take the simple route and make, sort of bury a fax inside a 275 transaction or just use fax or whatever the simplest route might be, does get to immediate ROI by saving people paper, postage, and process issues but it does not, it sort of shuts the door on any kind of improvements that would come further down the path and what we’re proposing is an approach to have your cake and eat it, too, in this area. Syntax, the legacy syntax is X-12 and HL7 version two are what we used in 1998 when we were doing this because of the requirement that it be an ANSI standard and they’re fine syntaxes, they’ve worked for many years in many situations, they do require a certain level of sophistication to deal with them and the cost of software mappers that can deal with the syntax is for the most part starts at $30,000 dollars and runs up to a couple hundred thousand dollars. And they have ways to work with textual information, free text, but they’re a little bit awkward in that regard. Back in 1998 XML was sort of at the peak of the hype cycle, we were hearing about XML enabled Coke machines and everything in the world, it was going to do away with standards and so forth. As a technology it survived the wild expectations followed by the trough of disillusionment that invariably follows and now it’s beginning to rise towards the plateau of productivity, so if you go out now and you look at a new application that has a need to transfer information 99 percent of the time it’s going to be using XML. Why? Because it’s a very flexible structure, because it’s easy to understand for an analyst to look at the XML and understand what it says, and because there’s a lot of free or inexpensive software that makes the programmers more productive using XML then using a legacy syntax or an ad hoc format for a data file. One of the important features of XML to this proposal is a related standard from W3C which is called the XML style sheet language, SSL. What it is is a standard way of describing a mapping, so right now if one payer is using See Beyond(?) to, and I’m not picking on any individual vendor, this is just about vendors, See Beyond to map an HL7 version two transaction and they wanted to move to Mercatur, or they wanted to share their intellectual property with another one that was using Mercatur, they’d have to redo all of that because the language for the mapping is proprietary to the vendor. In the case of SXL the language for the mapping is itself standard and can be freely shared across people using different formalities. And in fact it’s built into the browser you’re using today on your laptop, many of the pages you see now are actually coming out in XML and being translated into a human readable form before you look at them on the browser. XML is good in that it’s equally at home with structured data and text, and as I said it’s currently the syntax of choice for new endeavors. I’ve already anticipated what this slide was going to say but it just points out that style sheets are the way, are a standard way to take the information associated, presented in XML, which is on the left here, and create a human readable form which is shown bottom right. The HL7 clinical document architecture is the next topic, it is a specification that builds on top of XML and so XML says there will be tags. The CDA specification says the tags will be a document type tag and another kind of tag and another kind of tag that allow people to write programs to create XML in one place and read it in somewhere else and understand how to do with it. Its objective was to standardize clinical documents recognizing that most clinical documents today represent textual information rather then structured information. The important thing about it, however, is it allows that textual information to be supplemented with coded structured information as well. As far as we know, we know that this was the first ANSI standard in health care to be based on XML, it was approved in the year 2000, and as far as we know it was the first ANSI standard on any topic that was done in XML. The way it organizes the tags in the document is broadly into two parts, a header, which contains structured information identifying the patient, identifying the author of the document, identifying the status of the document in terms of is it a draft or signed or amended, identifying the kind of document it is, just the kind of things that you would want in a computer to manage the document, whether or not you’re processing the document or just saving it as text to show to somebody later on. The body, the other half of the specifications is where the clinical content goes, and it can be organized more or less according to whether you want to break your material down into sections and sub-sections, whether you want it to be displayed in a tabular form or a textual form and so forth, and the CDA at the current standardized level doesn’t say much else about that, it just gives you the names of tags you’ll use to describe your document. DR. FITZMAURICE: Excuse me, Wes? Are you saying that for the CDA header there are a set of tags that give you the metadata about what’s inside the document? And that those tag names can only be used by somebody following these standards for that particular metadata? MR. RISHEL: No, metadata is a term that gets applied at different levels, one person saying anything you can do I can do meta, but in this case both, the metadata is the tag structure — DR. FITZMAURICE: In other words it’s the formatting that is the location, not the name of the tag that counts for the CDA header. MR. RISHEL: In the CDA header each bit of information might be the identity of the patient, is tagged with a known tag and therefore it’s accessible by computers that want to know where to file this document or things like that. In the body there’s more freedom in terms of how you arrangement the information and the tag, the fundamental of the CDA is that it was designed to provide common processing documents that were primarily there to put text in front of a person’s face, but it did allow that to be supplemented with coded information, it does allow, and it’s usual I’ve anticipated my next slide. The CDA can include images and if you were to take this ugly drawing of a finger here and replace it with a dermatologists picture that he took with his $100 dollar digital camera you’d sort of get the flow of it, I just wasn’t willing to generate the rash in order to take the picture. And it’s very much like a web page in terms of how they’re included except that in a web page you might get the text from over here and the image from over there, somewhere different. In the case of the CDA the normal procedure is to include them together in a package so that the document can be rendered without reference to the web or anything like that. Now we’ll describe how we take that basic CDA standard and supplement it with additional requirements for attachments. The fundamental flow as I’ve described is essentially the same, this is just a graphic that describes how the X-12 transactions coordinate with the HL7 transactions or CDA documents and how that works with LOINC. Probably the important thing to notice that there are two levels in X-12, that is X-12 defines a transaction set and then separately publishes an implementation guide that says how to use that transaction set for a specific business purpose. In the same way HL7 publishes the CDA and then prepared a separate implementation guide that says how to use it for attachments. But we go a step further, after that basic implementation guide we have several booklets that describe how to describe specific attachments, so there’s a booklet for an ambulance attachment, there’s a booklet for rehab attachments, and those are there in order to decide the content rather then the way that content is organized, so the ambulance attachments says what are the components, the questions you can ask about an ambulance run, what are the answer parts associated with those components, and what are the LOINC codes that identify those questions and answers and the document itself. The attachment specs give predictable content, as we said there are the six basic ones that I’ve described, in addition work is going on in DMERC in home health and there is a process available that’s been on the shelf since ’99 for continued outreach to determine future attachment types, and as we say once the basic principles are established the business of creating a new booklet for a new attachment type is an analysts job, that is it’s a job of finding out or achieving consensus on what the questions and answer parts ought to be and then it’s just a matter of turning the crank with LOINC and with HL7 to produce the booklet, there’s not a lot of new thinking about the informatics associated with it at that point. The key characteristic of this new approach that gets to the issues that I’ve been describing is that each of those booklets describes two variants logically. One is called the human decision variant, and one is called the computer decision variant. The human decision variant matches what we described as the prevailing workflow, which is to say the information is available on paper that could be scanned, or in text form in a file, or could be typed by a typist into a system, but it’s not highly coded and highly structured. And the health plan, that’s consistent with what the health plan wants to do, they want to put that information in front of a person rather then auto adjudicate it. So in that approach it’s low impact on health plans, that is they literally could create a starting implementation that would take a style sheet provided by HL7, apply it to this XML CDA document, and look at that on a web browser and that would do it without any having to invest in a special mapper for the HL7 part of the transaction and so forth. Very easy to do and for those health plans that are skeptical, I mean the thing about HIPAA is the health plan must, the provider may, and therefore some health plans are concerned that they’re going to have to spend a lot of money implementing attachments and very few providers will come to the part. Well, because this is particularly easy to implement on the provider side and doesn’t required structured data for the human decision variant, therefore more providers will come, so that is so the investment on the health plan side will be paid off with more usage, but the investment is also lower on the health plan side because they can make the basic implementation using tools that for the most part are already available. The computer decision variant makes use of that ability in the CDA of supplementing the text with code, with structured information, so if a health plans wants to auto adjudicate and it gets to the computer decision variant then it can use a more complex style sheet to extract the information in computer readable form and send that into their adjudication system. And it uses, the human decision variant only uses one LOINC code to say what kind of an attachment is this. The computer decision variant uses LOINC codes to identify the components and the answers in the same way we described before. Mike? DR. FITZMAURICE: Wes, for this to work does the provider have to invest the time to learn the appropriate LOINC codes when the provider sends that information into the health plan? MR. RISHEL: For the human decision variant to work they have to know a couple dozen LOINC codes for the attachments. For the computer decision variant to work they have to know a couple hundred LOINC codes plus the thousands of LOINC codes that describe individual lab tests. So the general thought is if they’re going to use the computer decision variant they’re going to have structured data spinning somewhere in the organization and they will have either used LOINC or they will have used codes that are mappable to LOINC for that. For the less sophisticated organizations you’d have to have some code to identify what kinds of attachment it is, the fact that it’s a five digit number instead of a two letter code or something is not that significant in terms of the complexity. MR. BLAIR: Could I ask a question? In reality maybe the burden wouldn’t be quite so great on a provider because if they’re dealing with claim attachments that have ambulance codes all the time then maybe that number collapses and maybe their opportunity to be able to go to the computer readable variant might not be quite as heavy. Is my presumption do you think valid or am I just doing wishful thinking? MR. RISHEL: Well, I’m in favor of wishful thinking if it gets us somewhere. It’s very valid, I think that certain of the attachments are really claim extensions and the ambulance would be an example of that, that is they’re asking for essentially administrative data, you’re not going to go to the patient’s medical record and pull this data out to answer the attachment requirement anyway. By giving predictable content it allows the people who have a billing system for their ambulance to collect the data up front during the run and they have it available and so forth. So for those kinds of attachments, again, the fact that use a five digit number instead of a two digit code is we think an acceptable level of complexity. DR. COHN: Wes, can I ask a question? Let me just make sure that I understand and I apologize looking through to see your slides to see if you’re going to get into this further, but I guess I’m wondering how much confusion there is in the industry around how all of this works. I mean for example do some people think that just by using an XML standard that they can therefore avoid use of any of the LOINC codes? I mean is that something, I mean I’m looking at a letter that we received from a large trade organization that I would at least wonder by reading it whether they think that use of the XML will mean that the providers that they represent would have to use LOINC. And so I guess I’m wondering do you think that there’s some confusion there as opposed to really being two levels of complexity? I guess I’m also wondering just to make sure I’m hearing right that somebody who gets a request as a 277 request but they’re actually would be a LOINC in that, which would be there sort of a question and that’s the proposal there? MR. RISHEL: That’s where you get to the couple of hundred LOINC codes total for the attachments outside of lab, where there are, Stan can give us the number, but there are more LOINC codes. I would be surprised if there wasn’t confusion in the industry. There was a time when I talked about the hype cycle around XML, where there was a popular belief that the XML standard itself with tags that looked like human words describing the data meant you didn’t really need standards because the computer could figure out, it said person there, right, that’s a person. We’re past that for the most part now, people understand that XML is an enabling standard and you need standards above it to say what the tags are. But I certainly have heard that just use XML as if it were the magic bullet, but it’s not. Did I answer your question okay? DR. COHN: I think what you’re saying is yes, there’s confusion out there. MR. RISHEL: Yeah, there is. Not to beat the horse I’m going to beat later more but I think one way to deal with the confusion would be to get an NPRM out there and let the industry reply to it, that is right now a lot of the confusion is that we can only get the word out so fast about this approach. DR. COHN: Sure. I think Jeff has one question and then we’ll let you get back — MR. BLAIR: I just didn’t know whether you preferred if we save our questions until you’re — MR. RISHEL: I prefer the interaction. MR. BLAIR: Then can you entertain another question? Wes, could you give us some feeling for the convergence of the different XML protocols, the degree to which there really is a single XML standard? Are we close, what is the status of that? MR. RISHEL: I’m going to answer that question twice, the first time talking about XML itself, the second time talking about application standards that are built based on XML. For XML itself there is a family of standards, the base standard is XML, there’s SXL, there’s others, there’s continued growth towards more highly specific general horizontal IT world standards in that area, but the fundamental XML is reasonably stable at this point. It is, if you’re doing something new, whether it’s defining an interchange for a program or defining a standard most everybody now starts with XML unless there’s a very strong case why they wouldn’t, so XML itself is table. MR. BLAIR: But the thrust of my question and I don’t remember the different variations of XML standards, there was like two or three several years ago, and I just didn’t know if there’s been convergence of those. Is there essentially one that’s become predominant now? MR. RISHEL: You’re talking about sort of the general XML itself as opposed to how you apply it to application? MR. BLAIR: Yeah. MR. RISHEL: Yes, it’s stable, I mean it’s stable and it’s well understood. What you probably heard about was some controversy over the schema language associated with XML, the XML schema language works out to be a camel in the sense of being a horse designed by a committee, and there was a time when there was resistance to implementing that and there were other schema languages written by individuals of note that might have been easier to implement and equally powerful and so forth, but that, we’re more or less beyond that, that is SXL schema language is the way you describe the tags for how you use XML in an application. We certainly have grappled with that in HL7, we found that our SXL schemas were difficult to implement using the tools because they were fairly green in terms of their use of it. Fortunately this particular schema that we rely on for attachments was a very simple and straightforward schema and all the tools seem to work okay on it so we were lucky in that regard. Steve? DR. STEINDEL: Wes, I’m concerned a little bit about the statement about the number of LOINC codes people would actually have to know. First, in the original set of claims attachments except for lab and clinical reports, yes it was a couple of hundred LOINC codes, and as you noted labs still remains the same as whatever the lab LOINC area is, but the clinical report portion of the claims attachment in the previous version listed all the clinical LOINC codes as well. Are you dropping those now? MR. RISHEL: No, I misspoke. Within the clinical report attachment booklet there are perhaps in the revised version a couple of hundred clinical reports identified. But in addition for certain reports like an ultrasound for a newborn there might have been a hundred additional defined for the sub parts of that. In our current approach a provider could provide to that detail but they don’t need to so they only need to know the top level LOINC codes that describe the document overall. DR. STEINDEL: Now, this comes to the second, thank you for clarifying that, this comes to the second part of my question. When you pose this as a question and answer where the payer is asking the question that I would like to see, could it be that the payer would be asking the question that I would like to see the serum sodium result in which case the provider would actually have to know that LOINC code because they’re asking a very specific question instead of asking for laboratory test results, which would be one LOINC code. MR. RISHEL: Indeed, in the case of lab there is a need for a LOINC to English translation in order for the provider to know what is asked. DR. STEINDEL: But analogously that could exist for the clinical as well. MR. RISHEL: I guess it’s possible, it’s hypothetically possible that a payer would ask for only some specific measurement out of a procedure or so forth, I don’t consider that to be a very likely possibility. DR. STEINDEL: I’m probably agreeing with you but I just wanted to — MR. RISHEL: I appreciate that, I did overlook that in my discussion. Mike? DR. FITZMAURICE: But isn’t it likely that vendors will simply plug in modules to their administrative record systems if not their clinical record systems that you get a request in from the health plan and it pops up saying I would like to have the sodium serum level without even seeing a LOINC code, it would just be imbedded in there, the physician might never have to know the LOINC code, but just see the request in English. MR. RISHEL: That’s the wonderful world of the future. I think that there’s a lot of concern about the steps to get there in the future, the sense that even many hospitals and certainly most practices just don’t have a record system of that sophistication available right now in order to build their responses, and I think what we’re trying to do now is create a dual track approach that allows for evolution, I’m going to describe that evolution in a little more detail in a second. So the important thing to recognize is that each of these attachments can be sent in a way that’s very simple and straightforward or it can be sent in more structured. Kepa? DR. ZUBELDIA: Before you go on I have a question. The attachment as it’s described it was before sent inside, the LOINC codes are sent inside the HL7, whether it’s version two or CDA, which is sent inside an X-12 message. With the inclusion of a scanned image or pictures that HL7 message can be very large, perhaps in excess of one megabyte, and that would be sent as one X-12 segment. We’re currently running into a situation were certain translators are programmed to require specific delimiter and they have all kinds of artificial restrictions. Could you talk a little bit about the benefits that this X-12 envelope brings to the attachment? Because I suspect it doesn’t bring any benefit once we go to the CDA and not only doesn’t bring any benefit it’s going to get in the way of attachments being handled properly, especially if you have to go through double translation because there’s two syntaxes imbedded in each other. And what are the possibilities of completely removing the X-12 envelope and you sending an XML message since there is a header to the XML message that apparently contains everything that we need? MR. RISHEL: I’m going to answer that at three levels, the first level has to do with where there is a benefit associated with the 275, second level talks about similar information flows without the 275, and the third level talks about web based DDE sort of approaches to doing attachment requirement. The first level answer is for those providers that have sophisticated software that generates their own 837s now and reads 835s, there is an advantage to having the 275 as an envelope and that advantage is is they can use the same communications mechanism that they have used in the past assuming that their translators support the bin segment. This is the first I had heard of those restriction on the bin segment with certain mappers although anything is possible. We did go through several of the major mapper vendors, specifically See Beyond, Mercatur, and Cybase(?), which has the paper free product, and talked to them at length about this and also Microsoft with the BizTalks(?) server, and they were all able to support the bin segment, or claimed availability let’s say. In addition, the structure of the 275 is parallel to the 837 in terms of the relationship to line items in the 837 and the positioning of the bin segment inside the 275 provides useful information to the payer in relating the attachment to the specific service that it relates to. So there are specific data items that would be required in the header if we used it as a replacement for the 275. That’s conceivable to do that, it’s simply not in the document we balloted at this time. I do want to add a bit of information to what you said about images. We are recommending the CCITT group for format for page images, which runs about 50K per page, it’s the one that’s sent between fax machines and so if there was a 200 page document submitted that would get up to your megabyte limitation, did I do the arithmetic right? Yeah, I think so. But for most places where a page or two are copies out of a chart into it, it would be more like 100K rather then a megabyte. I don’t think I did that arithmetic right, 20 pages would get you the megabyte. But nonetheless I think that a megabyte would be a long attachment except for those providers that decide to fax the whole record in. Now the other two levels of answer to your question, Kepa, I’m going to address further in the presentation, so I’ll go forward then if it’s okay. This is sort of a graphical illustration of what I’ve been describing. Payers can accept any of four formats out of these two variants, there are four options for how to create an attachment. One is just a little bit of XML to identify the header information that I described along with a scanned image. And that we see providers being able to generate that with very simple programs. During the April HL7 meeting we asked a couple of vendors to come in and do demonstrations of our new approach. Microsoft came in with $199 dollars software package called InfoPath, it’s part of Microsoft Office version 2003, and built literally in a matter of days a user interface to collect the information for an ambulance form and out of that package they formatted both the XML and the 275 necessary to support it. Not that we would expect very many providers to be that sophisticated in using InfoPath, but these hundreds of vendors to low end providers, actually I think it’s about 2,000 including the Mrs. Adams’ Storm Door and Computer Record Company, and that level of sophistication is within their scope even the low end vendors, so we think it would be easy for vendors to generate that. One of the options there is to scan and include the scanned image and that was demonstrated. We had another vendor, we didn’t want to have just, any time you do anything with Microsoft there’s always a knee-jerk reaction against it so we had Sun come in as well and they, a particular company that used Sun’s software demonstrated a web based approach to filling out attachments and sending in scanned images, I’ll talk more about that in a little while. Probably the important thing to note from this slide is that suppose some provider decided for some reason they wanted to use the human decision variant rather then the computer decision variant for some specific attachment. The way the style sheet works the payer wouldn’t have to do anything different with that variant then it would if it received the human decision variant, that is they can process it and extract out the text information and put it in front of the person, so there’s no need for a payer who doesn’t want to use auto adjudication to go to additional expense to process attachments that happen to come in in the computer decision variant. I think that’s critical towards this incremental approach that I’m going to be describing. Moving on, I know I’m running over, the other vendor in Cleveland who described a workflow that could work well with low end practices, smaller practices, that either have no automated billing system or they have one and they’ve bought the upgrades for five years, I was talking about the sort of the levels in lack of support in a lot of practices around the world. Finally there are those that do EDI but they do it by sending a print image to a clearinghouse so they really don’t have any software in-house to deal with X-12 transactions. There are many different variants but the one we talked about was when you get to the point that the health plan decides they want additional information they send an email in the clear that is unencrypted to the practice that says there is a request for information, click on this URL. It doesn’t identify a patient, there’s no protected health information in the email. The provider, using their email system, would click on the URL, would have to log in to the payer’s website, therefore create the authentication and the encrypted link that’s required to deal with protected health information, and then would be shown a work list of claims for which additional information was requested. They would click on a claim and they would see both the LOINC code and the explanation of what the LOINC is in English there, and they would have the ability to either type a reply or to upload a scanned document that was responsive to the reply. So with a PC, web browser, email, and a $60 dollar scanner they’d be in a position to bypass paper in both directions, that is the request came electronically and the response went electronically. And the way this demo worked, at the end of that dialogue, what the payer ended up with was the XML form of the attachment in the same way that it might have come in from a large provider through batch EDI, so they didn’t need a special processing program to deal with the DDE input that’s any different then they would for the web input. We think this is one of the critical things about this particular approach is that because XML is sort of the underlying language of the web it gives us an ability to go to the high end where they really want, they really want automated processes, they’re dealing with thousands of claims a day, they want to have work lists and all the things that come from using the computer down to the low end where a lot of physicians practice where they need all the help they can get to be able to make use of electronic transactions. The question, I talked earlier about the fact that a health plan that wanted to use human decision making didn’t need to do anything special if some provider decided to send in the computer decision variant of the transaction. Why would that ever happen? Well, there’s always some practice somewhere that wants to be elaborate, but a more likely scenario is the one I’m about to describe. Let’s suppose that a specific payer, dealing with a specific kind of attachment, decides it would like to auto adjudicate. Under the proposal that we’re describing they’re obligated to accept the human decision variant, which they can’t auto adjudicate, but we believe that if they made it known to the provider community that attachments that came in in the computer decision variant would be adjudicated faster, that would provide an incentive for the providers to begin to create the human decision variant, the computer decision variant. Furthermore, they wouldn’t have to do it for all attachment types, only for the specific attachment type that they’re talking about. However, once they begin to send that to the specific payer who’s ready to auto adjudicate they’re going to want to send that same format to the other payers who aren’t ready to auto adjudicate and that’s okay because our approach allows the style sheet to extract the human decision material and present that to them. So what we’re proposing here, and this is a policy issue, it’s not, the technology supports it, it takes a policy to make it happen, is that we could start out with the low end business of saving people paper and postage, but the standards supports getting to the higher order efficiencies and of the process of going there is not the typical one size fits all HIPAA NPRM approach, it’s driven by incentives on both the payer and provider side, that is the payer’s incentive to auto adjudicate is to save staff time, and to improve their statistics for prompt payment and a bunch of things like that. The provider is able to respond specifically according to their capabilities to an incentive by getting paid faster. So this is what we think is the basic strength of this approach to get the early results early but not right off the ability to go to higher order efficiencies as we go forward with auto adjudication. Here’s a timeline for this, those of you hear Gartner presentations know that when we make predictions we normally put a probability after it. I stopped putting probabilities on predictions for when NPRMs will come out. But we would hope that, we have finalized the ballot, we’re just about to publish the document, we would hope that an NPRM would come out in 2004, mandatory compliance could come in 2006. We also hope, and other speakers are going to speak about the possibility of pilots going on starting immediately if possible that will a, validate the approach and in fact may even lead to some voluntary adoption of this before the mandated date — DR. COHN: Wes, just to look at your slides, just a question, is that assuming that they’ll be a final rule that comes out later in 2004? Or are you assuming that a rule will come out mandating mandatory compliance two years starting in fall of 2006? Just curiosity. MR. RISHEL: I was assuming a relatively short processing time and I recognize that that’s an ambitious — DR. COHN: I mean Karen Trudel isn’t here so we can’t ask her to comment, I was just curious based on our histories of going from NPRMs to full rules. MR. RISHEL: Frankly I’m 100 percent focused on that NPRM right now. DR. COHN: I do understand, and I did notice the question mark regarding the — MR. RISHEL: But you’re right, the latency time between the time the NPRM comment period closes and the time that the final rule comes out makes this fall 2006 date wildly optimistic, but we do what we can. And I actually do believe, in contravention to all the experience so far with HIPAA, I do believe that there is the possibility for voluntary adoption of this in advance of the mandatory requirement, at least for some payers and some providers. DR. COHN: I agree with you, there seems to be a significant business case for why people should do this, which I think is our pilot project discussion. So obviously finish up and move into that. MR. RISHEL: So I’m going to wrap up now. I’ve described to you the approach and the similarities and differences to the prior approach and the fact that there’s a policy issue around this human decision/computer decision variant that I described. I’m asking for the support of the committee in terms of two things, both of which might be implemented through a letter to the Secretary. One is to support demo projects such as the one we’re going to describe, but more of them, particularly demo projects that go beyond just Medicare contractors, to the commercial world and Medicaid’s and so forth. And the second is help us avoid any backsliding on the intended date of May 2004 for the NPRM. There’s a new bill that passed recently that might take some people’s time, all we can do is hope to create some awareness that this is important within the Department. And then, as I said before, my sense is that when we talk about clinical information among providers through the National Health Information Infrastructure many of the same issues involved, are involved, there’s a substantial difference between the capabilities or even the presence of electronic medical record systems in different provider organizations, but most of them have basic internet access. So if we look at a dual approach, that is supports scanned or verbal information to start with, and has a growing requirement towards more structured information over time, but a common standard that support that as opposed as the need for a revolutionary introduction of a different standard to go to structure, I think that would make NHII get early results early and still get to the higher order of patient safety and evidence based medicine requirements as provider electronic health record systems become more prevalent in the industry. So thank you very much for your time and excellent questions. DR. COHN: Okay, well, Wes thank you, and obviously I think we’ll probably as we talk through the morning we’ll get to your issues about letters and things like that. MR. RISHEL: Great. DR. COHN: Questions before we move on to Suzan? And then just to sort of recognize that we’ll be having discussion and interaction after the break. Suzan, why don’t we let you lead off on the pilot. Agenda Item: Claims Attachment Pilot – Ms. Ryder MS. RYDER: Good morning and I hope you all can hear me, I’ll try and speak directly into the microphone. As you are probably aware my name again is Suzan Ryder and I work for Empire Medicare Services. Empire Medicare Services is a division of Empire Blue Cross and Blue Shield. We are a fiscal intermediary for over 1600 Medicare institutional providers that cover in four states. As a carrier we represent over 86,000 medical professional providers, and that covers a 16 county area in New York as well as the state of New Jersey. So my perspective this morning is primarily based on Medicare. Everything I’m going to discuss, both our past pilot and our proposed pilot, deals with Medicare processing. In 2000, actually starting in October of 1999 and extending in September 2000 Empire Medicare Services was funded by CMS to do a proof of concept on the electronic attachments. Our goal here was obviously to evaluate the draft 275 additional information to support a health care claim or encounter in the 277 health care claim request for additional information. At that time, and sorry to keep bringing this up, there was no NPRM and we utilized the draft electronic attachment guides. At that time we utilized for both the 275 and the 277 version 4020. Our goal, our objective, was to evaluate those implementation guides for completeness, clarity, accuracy, and ability to meet our business requirements. Completeness was measured by did the transaction contain every data element that was currently used by Empire Medicare Services. Clarity, was there an interpretation issues, interpretation issues between our team members as we began development of our interface files, our crosswalks, and our mapping structures. And later we looked for interpretation issues between Empire Medicare Services and our one and only provider who participated in the pilot. As far as the business need we wanted to evaluate could EMS or Medicare Services create all the necessary questions via the standard coding structure. As always with every project we were confronted with the same constraints that every payer and provider is, time, money, and resources. So what we determined to do was focus on the 277, the actual request for additional information. We also limited, instead of using all six of the proposed attachment types that Wes reviewed earlier we selected two. We selected the rehabilitative services for institutional and ambulance for medical. We based our selection on those attachment types that presented the best opportunity for specific detailed requests that could be codified. We also based our selection on requests that did not require in our estimation full medical record documentation. During the pilot we were successful in creating the 277 from our existing paper process. Today the environment produces paper requests based on qualifying criteria that you’re all aware of such as specific diagnosis codes or procedure codes, or we may also select based on a provider ID. We were successful in creating an interface file that included all the data elements in the implementation guide. This was not an interface file, this was specific to a line of business or specific to Medicare. The good news about doing this is this interface file then could be share with any other payer or provider in the health care industry. The medical policies part of the Medicare adjudication system, as with any adjudication system, are coded as reason codes, so within the application system our task was to take each and every reason code that applied to the selected attachment types, and actually map those to a LOINC code. We were successful in doing that. The maps, however, were based on Medicare reason codes and while those maps could be shared and utilized by contractors they didn’t necessarily pertain to other insurers. Provider results. As I mentioned we had one institutional provider who actually piloted with us during this time, they actually received our 277 and were able to translate it. During the pilot what we did was we had a parallel production process. They continued to receive their normal paper requests for additional attachments at the same time we produced the 277, so what they were able to do was translate the 277 and then compare it to the paper that they had received. The 277, this provider actually did an estimate for us and what they estimated was just implementing the 277 for all six attachment types would save them approximately $8700 dollars per year. One of the key lessons learned here, though, was the provider could not produce the 275. It was learned that the provider, like many of their peers, while they used X-12 transactions and were very adapt at using LOINC they actually all medical information was scanned and placed in image databases. This prevented the provider from translating the data into the proposed 275 codified structure. That information was presented to X-12 and HL7 and I think we were instrumental in that pilot in getting us to where Wes shows us that we are today, allowing for adoption of inclusion of imaged or text information. Now I’ll just review some of the recommendation that were the outcome of our pilot. Additional LOINC codes for therapies, it was discovered that the LOINC codes that we used were not specific to the type of therapy. Specifically within a therapy you can have physical therapy, occupational, cardiac, psychiatric, skilled nursing, alcohol/substance abuse, medical, social services, and respiratory, numerous therapies. And what we found was we needed to be able to have a specific LOINC tied to each one of those therapies. Again, this was brought to the national attention and now LOINC, additional LOINC codes are established specific to each therapy type. We also recommended that, to CMS, that review of their medical policies would be needed to prevent the instance where one medical policy or reason code could be mapped to multiple LOINC codes. This one to many situation was specific to our Medicare processing but we believe this would be a similar situation that other payers would be confronted with. Also we suggested that we needed to review the processing systems to prevent multiple requests for one claim. Requests today are generated based on HCPCS codes and revenue codes, which can result in multiple 277s going out for a claim. We felt that the Medicare processing systems also needed to be reviewed to prevent multiple requests at the claim level and line level. It was determined during the pilot that the Medicare claim processing systems produce reason codes at the claim level and then duplicate ones for the line level, so again, unless we’re very careful we start sending out multiple 277s for one claim. Testing also showed that the system currently could not support the production of paper or electronic requests based upon the provider identifier, and obviously going forward we want the ability for a provider to tell us when they wanted to receive electronic versus paper, so we would need the capability to create both. Key to the pilot was once again the determination that without the ability to exchange non-codified information providers would be severely limited in their use of the claim attachment standards. And that brings us to a proposal that Empire has just recently submitted to CMS. We are suggesting that similar to our proof of concept pilot that was completed in 2000 that we be funded to continue on with a working model. The difference obviously between proof of concept and the working model is at the end of 2000 we did not implement what we had developed. In this proposal we’re hoping that once we can successfully exchange the transactions that we can move to a production mode and become an early implementer of the transaction sets. We’ve also heard from our provider community that their excited about doing this but they do want to be able to also move to a production. And again, the proposal starts up by obviously we would be evaluating the 275 and the 277. Again, without the Notice of Proposed Rule Making what we’re going to propose is that we start with the draft electronic attachment guides, only those would now be version 4050. It’s my understanding that version 4050 will be up for vote potentially in the February X-12 meeting and may be final before the pilot standards. What has changed since 2000? Well, as Wes has really taken us through so accurately this morning, we are using XML, we now have CDA clinical text based or imaged information that can be exchanged using standard structure and semantic. Again, as we heard this morning, using a standard 275 with a CDA we can exchange between covered entities and clinical documents can be encoded in XML and can be rendered into human readable documents. Our proposal is to take the first steps, we’re not looking for computer to computer immediately, our processes today as I’ll take you through are very, very manual, potentially all human intervention, but we’re looking for the first steps to take us towards computer to computer. Again, I think for all covered entities, payers, providers, vendors, clearinghouses, we now have extensive experience in the use of the HIPAA named X-12 transactions and implementation guides, and since 2000 we have gotten better at using XML and the XML EDI tools. Specifically what we’re trying to do in this proposed pilot is prove the exchange of the 275 and the CDA. Potentially what we need to do is have a provider take a 277 from us and now match it to their medical record information. The 275 would be created from an application system using application data. Similar to what Wes presented we expect that in that 275 will be our identifier information, our enveloping information, that the providers that we are dealing with now have systems in place that create 837s, that can read 835s, so they’re in a position to actually create this 275 envelope. Additionally, the CDA would be created from their medical record information such as imaged data. This flow chart is our really a high level overview of what we expect in the provider community, and again I would just emphasize that the reason for the pilot is this is our assumption as to how this will work. There are many unknowns once we get into this flow. But the working model shows that we’re proposing to use all our existing networks, existing telecommunication networks, protocols and processes. We are not proposing to use the internet, so we’re looking at a mainframe to mainframe transfer, mainframe to mainframe transfer of the 277 into the provider’s facility. At that point it becomes more questionable as to how the provider will actually translate and then utilize the 277. As we saw in the proof of concept pilot the provider actually created a hard copy and then sent that off to their examiners to pull the imaged information. We expect that in many cases that’s what they’ll do here, too, they’ll actually create a human readable form of the 277 and that the examiner will actually be searching the imaged database and pulling up the information to return the 275 CDA to us. From the payers perspective the proof of successful exchange will be our ability to receive the 275 CDA again using existing telecom network and transport processes. The data contained within the CDA would then be used by our server application to display the imaged data that was received. Again, to kind of go through from a payer perspective in a little bit more detail, the transaction would be received again via our existing telecom into our mainframe application where we would validate the 275 envelope information, which is very similar to what we would do with an 837, ensure that you’re an acceptable sender, that we have you on file, and that we do expect this transmission. Next, an interface file would be created to update the suspended claim preventing it from denying after 45 days. The process today is that in the paper world after we determine that we need additional information we suspend the claim, we send out the request to the provider, and that claim, actually the provider has 45 days to respond to that request or the claim is denied. And I would also mention that the full claim is denied even though the medical record request could be at the line or the service level. So it’s important that as we receive the 275 that we stop that clock so that the claim does not deny, and also we would use that interface file to notify our medical review examiner of the attachments received. A second interface file would be created with the CDA data and with the claim identifying information, information such as the claim number, the subscriber ID, and the receipt date, and that would be transferred to our server application. Here’s another unknown, we have not tried to take in a transaction into our mainframe and then poured it down to our server applications. The service application would determine the type of data that we received, was it imaged, was it a text data, it would then compress the information and store the images with the identifying information in an SQL database. That database then would be used by our medical review examiner to query and then pull up the information to match against the claim. One of the things that we did already talk about this morning is another portion of this pilot at this point would be how large are the files that we’re receiving in, are we able to compress them, how much space are we going to need both in as we’re transporting it into the mainframe and translating it as well as saving it to an SQL database. The goal here is to be able to provide some evaluation back to CMS. We would evaluate reductions in payer operational costs, I don’t think I need to go into a whole lot of detail about the costs from a payer perspective in handling and controlling paper attachment information, controlling the paper that goes to the provider, asking for the additional information, as well as controlling the information back in. I would also mention that both the proof of concept pilot and this proposal are based on solicited requests. We are not looking for unsolicited claim attachment information to come in, we’re looking for a provider to wait for the 277 to be received before they send us information. However, in the real world, in the paper world today, we do receive attachments that are unsolicited. We do not utilize those, we save them, we control them, we store them, we don’t use them, we don’t pull them up to look at the claim that could come in the future, or the claim that may have already adjudicated. This will give us a potential we believe using the SQL database to maybe relook that and look towards the future of accepting unsolicited attachments. We would also evaluate what we believe one of the areas that presents the biggest potential for savings to the payer are reductions in claim denial due to attachment data not being received. Specifically there are significant costs associated with provider requests for reconsideration of claims that were denied because we didn’t receive the information. And again, this goes back to 45 days and if we don’t receive the information we deny the claim. At that time the provider is required to submit a reconsideration or appeal for us to relook the claim, submit the information that we had previously requested, and that could result in either a full reversal where we would pay the entire claim, or more likely a partial reversal where depending on the information we may only deny one or two services. And just to show you what our experience is in FY 2003, fiscal year 2003, on the institutional side for Empire Medicare Services we received over 17,000 requests for reconsiderations or appeals and over 50 percent of those resulted in a reversal. On the provider side we would also be able to evaluate reduction in their operational costs, again, they have very similar costs, handling of the information both inbound and outbound. I would also mention on the part a side, the institutional side, providers have access to an online application where they can go and see what claims have been suspended for additional information. However, they do not have the capability on that online system to enter in responses to those requests, so they must then send us the paper attachments. And we even see providers who will express mail those attachments to ensure that they are received by us within the 45 day time limit. Also we would be able to evaluate the delay in payment. Now what I focused on was the delay in payment when the claim is actually denied for lack of information and how long it takes us to reprocess, to process the reconsideration and actually adjust the claim. However, what I didn’t put in here which is also a substantial amount, is the delay in payment while we’re waiting for that 45 days or the timeframe that it takes to mail the request and receive the mail attachment. What I wanted to show you was some of the statistics we show. From May through September of 2003 this shows you some of the requests for additional information and how many denials were received in that period. Now each of these providers that I’m showing statistics for have already come forward to volunteer to be part of this working model pilot. You can see here that we run the gambit from nine percent of denials up to 50 percent. But let me show you some of our best case/worst case scenarios. First providers, 614 requests, not one of those which were denied. This was the provider who participated by the way in the proof of concept, this provider has a very adapt process of monitoring and ensuring that those requests are dealt with quickly. Now I have another worst case scenario, I have a provider who also volunteered to be part of the pilot, and that’s no wonder to me that he would volunteer. He had a 99 percent denial rate. Now what this picture doesn’t show you is out of how many attachments denied were reversed, and that would be something that we would be looking to do as part of the evaluation of our pilot. In addition to cost concerns or evaluations we would also be as part of this pilot evaluating the impact of HIPAA privacy, specifically the standard to address minimum necessary information. Specifically some of the questions that we would want to address here is were the LOINC codes used in the 277, limit the information that can be requested by the payer, can full medical record documentation still be requested by the payer. On the provider side will the provide be able to limit the information returned via the image or does the image indeed result in a full medical record. Can the provider limit their response by matching the 277 request to their image database? An example of this is our 277 would refer to a specific claim service line and its associated dates. The provider normally would use those dates to go and query against their database to match, however, many times the supporting information may have been image scanned or stored prior to the dates of service, so there is another unknown, we’re not certain how the provider is going to be able to match that 277 and may again resort back to the human review of those 277s coming in. This is just, this slide just recaps everything that I just went through and what we would be evaluating. In addition though I would just add that the standard interface file would again be developed and would be available not only to Medicare contractors but to the health are industry. Our goal again would be to utilize or create an interface file and the crosswalks that would utilize all the data elements in the guides. In the proposal we know that we will be up against the same constraints that we were in the original pilot and what our focus here is on producing the 275. We know from the proof of concept pilot that we can produce a 277 accurately and deliver it to the provider community, we don’t know if we can process the 275 back inbound. What we would be doing is building off of our proof of concept, building up from the 277 that was created prior. We would now have to update obviously to use all six attachment types. And most importantly in our proof of concept we only had one provider who volunteered. Now some of that reason was because of a limitation of funding, that provider was not funded to participate with us and unfortunately because it was a proof of concept we didn’t even move what he developed into a production mode. That same provider has come forward and volunteered to be part of this pilot but their real interest is that we move into a production environment, even before legislation. So what’s important to us is to include more providers, we don’t want just one institutional provider, we want a number of providers and we want to do both institutional and professional. We believe that this pilot, or it is our hope, that this will position not only Empire but CMS and the participating providers to respond to the NPRM, or based on timing, to the final rule with actual experience in using the transaction. This would include the evaluation of version 4050 as well as the LOINC codes. What we’ve seen already and we’ve only done a limited survey of our providers, we’ve already seen enthusiastic response from providers to participate in this pilot. We’ve received responses from a clearinghouse as well as vendors. I think from Wes’ earlier presentation we heard that many providers are sending their images to a clearinghouse and that may be why this clearinghouse is very anxious to participate. Vendors we’re hearing have started to develop software that can address electronic claim attachments. We’ve also been approached by third party certification vendors to help us in validating what we create. We’ve been approached by translator vendors who are willing to assist us in creating the maps and the interfaces. And consultants. I think one of the key things in our success are expert advisors. We have advisors from HL7, ASIC, as well as the 275 and 277 workgroup. Empire Medicare Services for years now has been very active in both participating with HL7 and with X-12. I kind of went through that real fast. Are there questions? DR. ZUBELDIA: I have some. DR. COHN: Kepa, why don’t you lead off? DR. ZUBELDIA: Sue, in working with the clearinghouses, we do have at least one clearinghouse very interested in the vendor, do you expect any problems to that or what do you think would be different from going point to point with the providers? MS. RYDER: What would be different going point to point with a provider as opposed to a clearinghouse or vendor? What we’ve said to the clearinghouse as well as the vendors who have approached us is we need them to get a sponsor provider, it cannot be at their say, we want to know which provider is actually going to work with them to create the information. Having said that, I think it’s going to be very close to doing point to point with a provider. DR. ZUBELDIA: Do you have or are you working on the ability to receive scanned image or only the structured information? MS. RYDER: Our real goal is, from our proof of concept, is to receive scanned imaged information. The provider that participated in that pilot and has volunteered for this one, that’s primarily where they’re going to keep all of their information so that’s a real goal, is to go after the imaged information and bring it down to our servers, and then present it bask as Wes has said, back to web application internal to Medicare. DR. ZUBELDIA: So you’re going to be using the SXL timesheet to save information? MS. RYDER: We’re going to try. DR. COHN: Suzan, let me ask a question, I’m just trying to figure out how all this works. Now you’re talking about it sounds like a fairly significant amount of work, maybe I’m missing this, it may have just been a lot of over slides, but when you have that many slides it usually involves a fair amount of work. Now let’s see, we’ve got an NPRM that Wes would really like out in May, a final rule that somehow starts going into the process begins to come out, you’re trying to deal with at the same time a pilot to provide input to things. How does the timeframe all work on this one? I guess I’m a little concerned that you may either make a discovery of one sort or another that might effect the final rule that might otherwise come out. I mean if I were somehow organizing all this I’d say geez, that pilot sounds goods, let’s do a pilot and then we’ll see what the results of it are and we’ll use that to inform an NPRM, but does that put you off into Wes’, beyond Wes’ retirement for a final rule or whatever. Do either you or Wes want to comment? MS. RYDER: I’ll start and then I’m sure Wes may also want to comment. I think what we did in 2000 with that proof of concept was got a good feel for what a 277, how to use LOINC codes, what a 275 might look like. So could we do this at the same time that the NPRM potentially is being produced and even put out to public comment? I think we could. I don’t think, and maybe I’m being overly optimistic, which I tend to be, I don’t believe we’re going to see show stoppers, I don’t believe we’re going to get into a situation where we’re going to say we cannot go forward. We know the 277 works, we feel very strongly that the 275 will work from the proof of concept and just working with people like Wes, so what we’re expecting from this is more ability to show providers and payers the potential for using these transactions. MR. RISHEL: A couple points. I think that it would be difficult to issue a final rule until the evaluation of a pilot had gone on. Would it better to hold up the NPRM until a final rule, you can argue that both ways. I certainly, if I knew that there were pilots going on at the level of quality that this proposed pilot is that covered Medicare, Medicaid and commercial insurers, I would certainly as if my preference made a difference, my preference would be to wait and have the NPRM wait then. As it stands I would be reluctant to give a reason to delay the NPRM until we were certain. Once it gets delayed when does it get scheduled again. DR. COHN: Sure, I think we’ve already seen that once. Okay, as I said all these ideas are good, I’m just trying to figure out sort of how timeframes work and it causes a little confusion. MR. RISHEL: At the risk of making an obvious comment, there is, no matter what the outcome of the 2004 elections there’s a latency period in 2005 because of changes within the Administration or change of Administration, and we’d like to at least get the NPRM out before that latency period. DR. COHN: Other comments or questions before we take a break? We’ll obviously be discussing this more after the break. Yes? MR. EMORY: My name is Jack Emory, I’m with the American Medical Association. You’re going to hear from Jean Narcisi later but let me give you my reaction to what I just heard Suzan say. They have a proof of concept pilot with one provider. That’s an institutional provider, it’s not a physician provider. I’m very much concerned about that, and while I hear your pilot and you want to include other institutional providers later on, or professional providers, that’s fine. But I’m really concerned about making recommendations to anyone based on one providers’ experience with this pilot or this concept. I also heard about your sort of limitation of saying it’s going to be mainframe to mainframe. A physician’s mainframe to a carrier mainframe, to a carrier, what about a private payer? It’s still a mainframe to a mainframe, which is really sort of pushing physicians into dealing with billing services and clearinghouses, and many of them do. But when you push them into doing that you’re raising their costs and that’s got to be taken into consideration and I don’t see that and I’m concerned about it. I encourage the pilots as you’re talking about pilots to talk about something other then mainframe to mainframe, something that doesn’t force the physician community or other small provider to move to a more costly way of doing business. I’m also, I see a lot of advisors to your project and to your pilots about dealing with X-12 and other groups of that sort that frankly don’t have an awful lot of individual provider input. We get provider input through the NUCC and the NUBC but it’s very much after the fact and that process, the development of standards process is a process that’s dominated by consultants, by payers, by vendors, not by providers. And I’m concerned about that, that those kinds of recommendations don’t give the provider community the opportunity to be involved in this process. I want to encourage the committee, I want to encourage the people working on HIPAA standards to move forward with attachments because I think attachments are a way of making the submission of additional data and the payment of data, accurate payment, good, worthwhile, we ought to move forward with this. But I am concerned about making decisions based on very limited information from the professional community and from the provider community based on one pilot. DR. ZUBELDIA: Can I ask Jack a question? Jack, I’ve been involved in the standard for about 14 years now and it’s an open process and everybody has the opportunity to participate, but the fact is providers are not participating. What can be done or what can the NCVHS do, or the AMA do, to encourage more provider participation in the development of the standards? MR. EMORY: Fair question, Kepa, and I think we frankly are recognizing that we need to be more involved then we have been in the past. Early on I don’t think most of us understood the importance of HIPAA and the role that the standards setting organizations are playing in the development of standards. We are certainly have reassessed that and understand that we need to be more involved and intend to be. But I do think that the people who are working on standards, there needs to be a balance so that it’s not the vendor, it’s not the consultant, it’s not just the payer, and that there is an equal amount of footing in the decision process for the non-consultants, non-vendor, non-payer, give us the provider community an equal opportunity, not after the fact as is the problem right now. And it is a problem, Kepa, I mean you’ve got a big organization that’s spending an awful lot of time and energy and effort in developing standards but it is done for the most part without a lot of input from the provider community and we’ve got to somehow come to grips with that and figure out how we can increase the provider input in this process earlier. DR. COHN: Wes, and then I think Suzan had a comment also. MS. RYDER: I just wanted to, couple points, the one institutional provider in the proof of concept was because we really couldn’t get participation and we did, and obviously that was a known, I mean providers have the same constraints we have. Additionally our focus on mainframe to mainframe is doing just what you said, it’s building upon the exchange of the 837 so that we have those processes in place. And the other thing that limits us to some extent is obviously that as a Medicare contractor we’re not allowed to use the internet for that type of exchange. I think the internet is where we have real potential to get involvement from the smaller practice. And finally, thank you for the additional information about advisors, we work in New York and Connecticut with hospital association and obviously on the professional side we work with a lot of medical societies, and so maybe we could solicit participation from them as an advisor to us as well. DR. COHN: Wes, did you have a comment? MR. RISHEL: Just a couple points. At the risk of belaboring the obvious one of the benefits of this revised approach is to decrease the level of sophistication required in payer’s systems, so the concern about mainframe to mainframe is not, certainly all our intent, and one of the main goals of doing this was to be able to find technological approaches that didn’t require a mainframe on either side but certainly not on the provider side. Speaking as, for 20 more days, the chair of HL7 I’d like to say that we have a strong provider representation in HL7 meetings but it’s mainly from institutional providers. We certainly, there’s a sort of a level of effort that’s required to participate in the debate that requires an investment of more time then many individual providers are willing to make, but we certainly would welcome ways of getting provider involvement, either individually, and we have some individual providers, or through associations. We offer ways, for example, to participate on specific issues without having to be a member of the organization and so forth, so we’re open to whatever we can do increase the informed involvement of individual providers, but I emphasize the informed, we don’t want it, we want to know what to do to get the job done right. DR. COHN: I guess I’m reminded that I think in our PMRI report from the year 2000 we actually observed the issue that Jack was just describing, and I think tried to make a number of recommendations. There is issues as I think we’ve all described of the fact that the standards groups typically meet for a week at a time four times a year and so there is just a major time requirement for all of this. And as one who routinely had to look through one of the HL7 standards, which I had sort of not looked through previously which should have been relatively straightforward and simple, I was actually sort of reminded that sometimes all of this doesn’t come out in the most user friendly language either, even on topics that should be relatively user friendly. That being said we obviously all need to sort of think about that, we’re probably not going to solve it today but we do need to sort of reflect on if there are things we can suggest as we move into 2004 to make sure there’s appropriate representation on all of this. Now any final questions? We’re going to be discussing this after the break but I do think we should allow people a break at this point and then we’ll be hearing some additional reactions and just sort of a general discussion on what we ought to do on next steps for a couple minutes. Now we are running a couple of minutes late, let’s try to take a ten minute break, we’ll come back shortly before 11:00. [Brief break.] DR. COHN: Okay, well why don’t we get started with the next session, obviously this is called reactions and subcommittee discussions on claims attachments, and welcome Jean Narcisi, are you representing the AMA today or the NUCC? MS. NARCISI: I am representing the AMA. DR. COHN: Okay, great. Well, welcome and let’s let you make your presentation. Agenda Item: Reactions and Subcommittee Discussion – Ms. Narcisi MS. NARCISI: Thank you very much Dr. Cohn. As Dr. Cohn mentioned my name is Jean Narcisi and I am the director of electronic medical systems of the American Medical Association. It is my pleasure to appear today on behalf of the AMA before the Subcommittee on Standards and Security of the National Committee on Vital and Health Statistics. I would like to thank you for the opportunity to testify. The AMA defines a claim as the submission of information by a physician or insured individual to a third party payer using a standardized format, for example a CMS 1500 paper claim form or an electronic format, sufficient to establish that covered health care services were provided. The claim includes a request for payment or reimbursement to the physician or insured individuals. The AMA believes that an attachment would include information either requested by the payer based on pre-payment or post-payment follow-up or provided by the physician at the time of submission of the claim. Attachments consist of information presumably not available on the initial claim or claim format that provides further supporting details on the claim. Attachment information should include information that tends to be highly situational in nature, for example, requested only for specific types of services, and cannot be readily accommodated in a standardized paper or electronic claims format. Attachments should accommodate of paper and electronic technologies and should allow for structured data elements, images, and free form text as appropriate. In addition, physicians should have the ability to submit attachments at the time of a claims submission and should not have to wait for a payer query or specific attachment requirement to explain the unusual circumstances associated with the pattern of treatment. Physicians are currently asked for a variety of supporting information to adjudicate a claim, this can include information that is contained in the patient medical record, such as operative notes, test results, or information to substantiate the level of service provided. Other attachment data includes additional administrative information dealing with patient eligibility, for example copies of drivers licenses, Social Security cards, Medicaid cards, etc. Many physicians are also asked to verify information of the patient’s relationship with the insured and other information to assist in coordination of benefits. Often the COF information is not readily available to the physician. In addition, some payers, especially Medicare and Medicaid, request attachment information using special forms to reflect patient consent, Medical Necessity of Ordered Durable Medical Equipment, the Advanced Beneficiary Notice, and the cost and duration of use of drugs and supplies. The methods by which this information is requested depend on the specific payer, physicians relationship with the payer, and their technical capabilities. Submissions can sometimes be made electronically if a payer has an electronic form in place. All too often, however, even when the claim is submitted electronically the physician must send in the attachment information manually. In some instances, and also contributing to cost and inefficiency, rare requests can be satisfied via the telephone. Frequently based on expectations of payer requirements for specific written requirements physicians submit additional information with each claim of a certain type so as not to delay reimbursement to the payment or payment to the physician. Lack of standardization across earth(?) and payers is a serious problem. Searching the patient’s medical record or administrative file well after the original claim is submitted to gather the additional information and place it in the format required by the payer creates undue burdens on physicians. The AMA believes that in some instances these attachments and document requirements involve payers’ legitimate needs for sufficient information to assess coverage or justification for specific types of services or to meet contractual or regulatory requirements. All too often, unfortunately, many physicians have concluded that these requests are intended to delay payment of claims or to provide a basis for unwarranted denial. Also, they also reflect a desire to pressure physicians into billing at a lower level of service. The AMA has found that requests do vary considerably across payers. This causes substantial cost and delay for physicians. This variation includes when attachments are requested, in what format they should be submitted and the availability of electronic submission. The AMA believes that considerable strides can be made on the attachments issue. The AMA also believes that standardization in an electronic exchange of attachment information would reduce the workload for both the requester and respondent to attachments and that this could ultimately result in cost savings. In order to work the standardization must apply to all payers, standardization in the circumstances when attachments will be required will lead to an increased burden on physicians. The current HIPAA implementation guides for the claims transactions are very complex and include many confusing statements regarding the requirement of certain data items. For example, the requirement for a physician to report certain items in a claim is conditional on a specific situation. Many of the situations are not clearly defined in the implementation guide. This means that some payers will require a particular data item and others will not. Therefore, the AMA is concerned that the payer or governmental response to the HIPAA standardization of the claim will be an expanded use of attachment requests to circumvent claims standardization. Administrative simplification will not be achieved for physicians if each payer requests differing amounts of additional material as an attachment. The AMA believes that standardization of formats and electronic exchange should reduce the cost of preparing and submitting attachments and enable physicians and the health care system to realize the full benefits of electronic data interchange in administrative simplification. At the same time, as indicated above, accomplishing such partial standardization without standardizing when attachments are required, could make matters far worse. Given the current extent of physician use of electronic medical records, the state of standardization of such systems, and the lack of standardized links between clinical and administrative systems, the cost of obtaining attachments information from existing electronic or paper medical records and then placing it into a standardized electronic format could be prohibitive. Fundamentally the AMA believes that the HIPAA mandate for claims standardization as well as electronic claims formats that were intended to be designed to be less constrained then the paper formats provides a conceptual, regulatory, and technical framework to reduce or eliminate much attachment use. Although the HIPAA transactions recently implemented provide a standardized format to transmit a claim the data requirements of each payer are far from being standardized. Obviously contractual provisions vary across and within specific companies resulting in different data requirements. Nevertheless, the number of companion documents developed by the payers makes it difficult for the physicians to determine exactly what data is necessary to process the claim for each payer. The AMA suggests that the current system for submitting claims and other transactions in the HIPAA format should be fully operational by all payers and physicians that submit electronic transactions before the claims attachment standards are adopted as HIPAA standards. As I stated previously it is the position of the AMA that the health care industry should standardize when attachments are required and not just how they are submitted. Guidelines also need to be established regarding the type and the number of requests for information that should be permissible from payers to physicians. The attachment standard has been designed so that a payer can send an electronic request or several request for additional information. Physicians should know up front what additional clinical information will be required for specific services so they can either submit it with the claim or when the attachment data or images are available. The AMA also believes that there should be an organization or a group other then those responsible for developing the messaging transactions responsible for developing the type of request for information that should be permissible from payers to physicians as well as what additional clinical information will be required for specific services of claims attachments transactions. This organization should be representative of all parties effected by health care electronic data interchange, for example, providers, payers, standard development organizations, and regulatory agencies. Based on their structure in current and anticipated responsibilities the National Uniform Claim Committee and the National Uniform Billing Committee are appropriate to assume this task. The NUCC has an official operating protocol that provides full due process, open meetings, and the ability for non-members to generate agenda items. Fundamentally data content for business processes around claims submission and claims payment should be maintained through committees like the NUCC and the NUBC that focus on formal, balanced representation of key parties, using a consensus approach to decision making. Maintenance of the data content and the related business functions is a policy related activity, therefore, it should be conducted through the kind of public/private partnership that these two committees exemplify. Claims and attachment standards should not be viewed as primarily technical communications standards. In addition the AMA believes that standard implementation guidelines for code sets are essential for uniform national application of the code sets. If standard guidelines for medical code sets were adopted many attachments would be eliminated. If health plans and physicians are permitted to implement and interpret medical data code sets as they see fit the purpose of administrative simplification will not be achieved. An important part of administrative simplification, a reduced regulatory hassle certainly includes the simplification of instructions for the coding of health care services. The overwhelming amount of paperwork to which physicians are subject would be significantly reduced if coding guidelines were standardized within electronic transactions. The AMA believes that the CPT guidelines and instructions should be specified as a national standard for implement CPT codes. The instructions and guidelines contained in the CPT book are subject to the same rigorous editorial process used to develop CPT codes. The CPT Editorial Panel and CPT advisors consider CPT section guidelines, specific code level instructions and definitions, and the application of modifiers at the same time the language for CPT code descriptors are developed. As proper use of CPT codes is based on all the associated material contained in the CPT book, for example, simple, intermediate, and complex repair are defined in the book prior to the actual report codes, so that users understand the circumstances for reporting each. Also, coding conventions such as add-on codes are explained in the guidelines. The use of codes and descriptors apart from this information limits the functionality of CPT in its uniform application and contributes to improper coding interpretations, which are counter to the purpose of having national standard code sets. Therefore, the AMA strongly encourages the subcommittee and the NCVHS to recommend that the CPT guidelines and instructions for applying the codes also be included as a national standard. As stated previously the AMA believes that standardization of formats in an electronic exchange should reduce the cost of preparing and submitting attachments, and enable physicians and the health care system to realize the full benefits of electronic data interchange in administrative simplification. However, until standardization is achieved regarding when attachments are required and not just how they are submitted, the use of the attachment standards should remain optional and based on trading partner agreements between physicians and payers. Thank you for this opportunity to present the views of the American Medical Association. I would be pleased to respond to any questions that you might have. DR. COHN: Well, Jean, thank you very much. I guess based on your testimony I do need to publicly comment that I’m a member of the CPT Editorial Panel and therefore would recuse myself from any discussion of that part of your testimony. In addition, even though I’m technically not required to recuse myself I am a member of the NUCC, so I don’t think I can comment on your testimony at all but I’ll be happy to facilitate the conversation. Obviously, is there any other public comment or discussion, relationship, from the public or from the audience in terms of reactions to the claims attachments? MS. GREENBERG: You might just want to note that we do have a letter from the American Hospital Association. DR. COHN: Yes, also on this, and I’m sure they’ll be a lot more discussions as we come down the way on all this. DR. ZUBELDIA: Could I get Jean’s and Sue Ryder’s testimony sent to me via email? DR. COHN: Marietta is saying yes, and obviously anything else we have we’ll be happy to send you a copy of electronically. Are there any questions for Jean or for others? Michael? DR. FITZMAURICE: As we’ve changed the attachment from what it was before now moving toward the clinical data architecture that may put a burden on physicians to learn some LOINC codes in order to send their stuff electronically. I presume that the advantage for the physician is the physician by sending it electronically will get paid faster. But it still is a burden on physicians to codify the information or put it in the right place when sent in. Has the AMA looked at the burden of this and do you have any suggestions for how the burden might be minimized if this comes to pass? MS. NARCISI: Well, no we have not looked at it, however, through the presentation that Wes and Suzan just gave I do understand that there are a number of LOINC codes that would have to be known by the provider community if a request came in. So what we’re saying is let’s look at what really requires an attachment, what types of services do physicians perform that require attachments in addition to reporting it on the claim. If we could define those things I think that would help the physicians know up front, even before they would get a request, that they are going to have to have some type of justification sent in, either at the time the claim is sent in or when that data is available. Sometimes the reports are not available at the time the claim is sent in, or the image may not be available at the time the claim is sent in. DR. FITZMAURICE: I heard Empire Blue Cross talking about a pilot they were doing to try to learn more about how the claims process would work with the electronic claims attachment, do you know of any similar pilots on the provider side to see what advantages there might be for the provider? What would happen to the cost and timeliness of claims submission and the attachment submission if that standard were to come to pass? MS. NARCISI: No, I have not heard of any types of pilots that have been going on in the provider community. I think most of the physicians anyway were busy trying to gear up for HIPAA and the privacy standards and then also again in the transaction standards. DR. COHN: Jeff? And then Kepa. MR. BLAIR: Jean, you wound up indicating that additional clarification would be very helpful to providers if there was an understanding beforehand as to what circumstances would require an attachment and the NUBC and NUCC are in a good position to, well, this is the area where I wanted to understand it better. Were you thinking that NUBC and NUCC would essentially be in a position to perform the negotiations with the payers where there would be a standardized set of agreements as to what circumstances would require claims and then you would wind up coming to the NCVHS or HHS and looking for federal sanction, endorsement, approval of that? My thought is NUBC and NUCC could do that negotiation probably with at least some of the payers but you don’t, at least to my knowledge you don’t have the ability to enforce it, you don’t have the authorization to enforce it. Would you then be looking for an entity to provide the enforcement? Is that what you were thinking? MS. NARCISI: Well, Jeff, and I will clarify, although I am representing the AMA here today I do chair the National Uniform Claim Committee, so I am intimately involved in the whole process. My testimony today is representative of physician’s view, however, I do believe that the NUCC and the NUBC could take a look at some of the proposed formats, for instance the clinical reports are one of them that we would be, the physicians especially, would be most concerned about, having a request come to the physician saying I need this, I need this, I need this, and then being able to respond, now they would probably have to respond with paper instead of an electronic response. However, the ambulance, the emergency department, and the rehabilitation services, maybe those are standards that are ready to be released, maybe they’re ready to be out there if they have been tested, I noticed in the Empire testimony they looked at ambulance and rehabilitation services, maybe those are the two that you focus on first. However, the other ones, the medications, the lab results, and the clinical reports, are the ones that I think need a little bit more analysis and I believe that the NUCC and the NUBC represents the constituencies that could pull this type of information together. Now the NUCC and the NUBC of course can’t make any final decision on these are going to be the types of services that will require an attachment and this won’t, because I think payers based on contractual agreements will vary. However, maybe it has to be looked at like the claim was looked at, that there’s a maximum data set, maybe every payer will have to accept an electronic attachment whether they need it or not for a certain service. But I think that’s what’s I’m really trying to stress that there’s too much unknown on what these services are — MR. BLAIR: Let me refine my question then. Are you proposing then that if the NUBC and the NUCC during the next six months, 12 months, 18 months, takes on that task that your recommendations you would want to be included in an NPRM for claim attachments, is that what you’re leading towards? MS. NARCISI: Well, either we would be recognized in the NPRM or our recommendations would come forward to the NCVHS, and you would make the recommendation. But we would probably have to somehow be included because we would probably do some type of focus groups and analysis to determine which ones. I also mentioned the CPT guidelines because I also believe that that will eliminate some of the paperwork that’s flowing back and forth from the physician’s offices. DR. COHN: Kepa, you had a question? DR. ZUBELDIA: Yes. Jean the X-12 implementation guides have situational requirements and even though the situational requirements are somewhat ambiguous at some points, is it feasible to achieve what you’re trying to convey here with situational requirements for attachments and say this attachment may be requested when, and then if this condition is not met the attachment cannot be requested, or this attachment is always required when, and explain the condition and if that condition is not met the attachment is not required and not used? Is that feasible or is it too complex to do it that way? MS. NARCISI: Kepa, I believe that would be one of the solutions, however, I think we need to do a better job in our notes that would define what the situation is. I think there are too many instances within the implementation guides for the claims, for instance a good one would be provider taxonomy, well, it’s not required unless you know that it’s required for adjudication of a claim, so the provider has to know up front what payers want that provider taxonomy codes and what providers do not. So I think that would be an example, I think we need to put our hands around this a little bit better and define what the conditions would be. DR. ZUBELDIA: Would you also comment on the dependency of attachments on HL7, or independent of HL7, what’s the situation there and how, what’s the penetration of HL7 in physician systems and in the systems that can generate these attachments? Are the practice management systems capable of generating HL7 structures? Do they really depend on HL7 or not anymore with CDA? I don’t know, Wes may want to comment on this, too. MS. NARCISI: Kepa, I don’t have any type of statistics about physicians use of HL7 within their offices, however, I do believe that it is probably on the lower end. I don’t think most physician’s offices are even using X-12, I think they’re sending their information, their data off to a billing service or a clearinghouse and then it is sent. DR. COHN: We’re here and Wes is actually coming to a microphone also. MR. RISHEL: Kepa, I think there are no statistics but it’s generally believed that the use of HL7 in physician’s offices borders on zero, they’re lucky if they have one system much less two that need to intercommunicate, and it’s not only the use of HL7, it’s having the data available in a structured form at all that is the underlying issue, and that is the underlying reason why this revision limits the use of structure to the kind of administrative data you need to manage the transaction and the policy proposal is for an incentive based approach to adding structure over time that would give physicians the freedom to decide to invest in the systems if they saw the payoff in terms of more prompt payment or fewer denials, but not mandate that it happen by a data certain. DR. ZUBELDIA: So this kind of leads me to believe that there is going to be two very different and perhaps not really convergent implementations of attachment, one the structured or unstructured computer to computer transmission, I won’t say mainframe, computer to computer that Sue Ryder was talking about, where certain providers have the capability in their HIS system or perhaps practice management system to send either structured attachment or an image of the attachment. And perhaps the bulk of the attachments will be more like what Wes described, the payer sends an email with an URL to a web page that says here’s a request for attachment come to my non-standard payer specific web page and send me your attachment. And that will require the transmission over the internet of possibly an image or an SXL style sheet type of CDA document. Could you comment on that? Both Jean and Wes? MS. NARCISI: Kepa, this is Jean. I believe that the ability to use the internet for sending additional information would be beneficial, however, I really see it as a major problem, that there isn’t any kind of maybe, I wouldn’t want to call it a total standard but anything on the payer end that would be similar, otherwise you’re going to have every payer with a different type of web page having different requirements in the way that it’s sent asking different questions, so I think that you need to standardize, if that’s going to be the way that additional information is sent then there needs to be some kind of structure around that, otherwise it’s going to be totally non-proprietary and all over the board again. MR. RISHEL: Kepa, I think there’s probably three gradations ultimately as opposed to two. First off I would be surprised if in the two years following claims attachments becoming mandatory there was any requirement for structured submission to any payer. I think that their drive towards the efficiency improvements associated with auto adjudication have many constraints that are based on their internal IT structures and their organizational structures that are more profound then simply the issues of dealing with the data formats. When they do come I think they will come in a very specific manner, that is this specific claim type there is the opportunity to auto adjudicate whereas for many claim, and certainly anything that looks at a dictated note or anything like that we’re a long ways away from that. So I believe that for many years the volume of transactions will be of the form of text that was collected for specific claim types, such as ambulance or text that was copies from a chart. Now the submission means needs to vary according to a different set of issues, and that has to do with the largely of the quality and type of the vendor system that’s available in the physician’s office or institutional department, certainly the experience we get talking to physicians, and this is the we, the Gartner we at this point, is that they like being able to use a website when they have nothing else available and when the payer represents a preponderance of their business. They would choose to use paper rather then learn the other six websites that it might take to get another ten percent of their business, but they’re willing to learn a couple of websites if that represents the preponderance of their business. One of the areas that is potential in this approach would be because there’s an XML document definition and an SXL style sheet, there would be the potential to create a consensus version of how you ask the questions on the website and so forth and create, I would say identical capabilities, it’s very difficult to standardize a user interface, but at least a lot of commonality from one payer to the other. The main concern I think that I have is to meet the need of getting the information flowing electronically as quickly as possible and for a diversity of providers, namely those that have systems and need computer flows in order to manage the volumes of their claims on the one side, and those that have rather poor system support at this time. But yet to do it in such a way that we don’t need another law or another regulation to begin to use auto adjudication when that comes around. DR. COHN: Wes, thank you. Now, Marjorie you wanted to ask a question and then I wanted to try to sort of frame this a little bit and move us into something where there’s actually some action around here. Please. MS. GREENBERG: I guess in addition to my general passion for standards one of the things that had appealed to me as a consumer and potential patient about the original claims attachment standard was the fact that it was only necessary to send the specific information that was requested as opposed to what I think is a fairly common practice, is maybe sending the whole report or the whole chart, or what have you. I do understand how this, on the other hand, what’s been discussed here about the lack of capacity at this point by many providers to have that kind of pinpointed response and so that you’re allowing for this more text based or in fact to continue to be able to just send this XML version of the complete document. And a statement here in the AHA testimony, which I realize there’s nobody here from AHA to respond, but I wanted to ask Jean something in relationship to it and it says, it talks about the XML standard, it says offers a more efficient way for provides to submit attachment information and the rule must state clearly, this is their recommendation, that when a provider uses the XML standard to forward to the health plan an entire document or record containing the information the health plan is looking for the provider is compliant with the privacy rules minimum necessary requirements. So what I guess I wanted to ask whether AMA also would take that position, and then also if you have any thoughts on perhaps just work that you’re proposing by the NUBC and NUCC might help address this problem of the value of just really providing the information that’s requested as opposed to the entire record or report from the point of view of essentially sending the information that is minimally necessary. MS. NARCISI: Well, we do agree with that statement that there could be some confidentiality issues with being able to send the information in XML format, that’s fine, but how much of it is actually the minimum amount needed. But I do believe that the NUCC and the NUBC could really kind of work through these types of situations that would define what portion of a report would be necessary and what services, simply would require an attachment more often. I think we could really work those through and I don’t think there’s been enough time spent on that. I think that the technical solution that has been proposed through X-12 and the CDA, the new CDA from HL7 is excellent but I think that there are just a lot of other business issues that we need to kind of work through before we implement this kind of a recommendation. MR. RISHEL: Could I just make a quick comment about that? The technology supports restricting the image to relevant information, so it’s a policy issue as opposed to a technology issue. DR. COHN: Carol, did you have a comment? DR. BICKFORD: Carol Bickford, American Nurses Association. One of the first rules of an information system solution is carefully looking at your existing business processes and determining how the world should better be run in the future state rather then as is. What I’m hearing is a significant automation of a bad, bad, bad scenario, which raises significant cost to our health care system, which is supposedly is going down the tubes right now. So I’m raising the question why do we have attachments? Why are we using them? Who makes the decisions? And my sense is it just gets filed, in which case why? Does anybody pay attention that material? And indeed, could we have a better process at the front end that includes the minimum data set sort of content? Are we looking at better business practices simplification, appropriate decision making, some whole process reviewed, not just the payers, not just the clinicians, health care consumer is part of this, how does it all fit together? Are we just automating a bad process? DR. COHN: Carol, I think I will take that as a statement as opposed to a question. I want to move the conversation a little bit and I guess, and maybe this is sort of following Carol’s comment, but I think one has to be aware that obviously claims attachments do exist today, I mean there are requests for medical record, requests for additional information that occur with great frequency depending on the payer and depending on the provider. If Dr. McDonald were here, and I know he did Gilman’s(?) work a number of year’s ago trying to get payers and providers and others to agree on sort of the rules around all of this stuff, I think he came to the realization, and unfortunately he’s in London at this point so I have to potentially paraphrase for him, but I think he came to the conclusion that this was still early to come up with business rules. There seemed to be no agreement, no willingness by the industry on any side to come up with agreements on all of this. Certainly I think Jean, whatever organization you happen to be representing, I think any of them, as a committee we would be more then happy for any organization that can forge industry wide consensus on any of this issue, these issues, to have those consensuses come before the NCVHS for public discussion and I think that there would probably ways in some sort of future versions of rules for them to find a place. But of course the big issue is the industry wide consensus, which as I said has been elusive so far in this area. Anyone else you know that can do that, please have them come forward and I think we would be happy to hear it. Now, having said all of that, I am still struck that there’s actually, this is an area where perhaps perfection is the enemy of the good, and I would like to have obviously everything absolutely sewn up on claims attachments, all the rules in place and all of this, and yet I think that there’s value even without that, the moving forward with the issue of claims attachments. I guess the question I have for the committee is, and this is maybe along the lines of what Wes Rishel was recommending earlier, what I’m hearing is actually there’s a need for additional pilots, and additional demonstrations, and additional tests, and evaluations of all of that because I think we’re all talking more at a level where we really haven’t had a chance to see how this is all going to play out and I think I’m hearing that maybe the committee ought to be supportive of that idea. And so that’s really a question for the committee. The other thing which I think Wes commented on is maybe a letter to the Secretary encouraging that everything go forward, that there actually eventually there be an NPRM, that there be these demonstrations, potentially the government could help fund but there needs to be private funding and capabilities also so that we can begin to get some experience in this area and help make this a reality. Now I guess I’m, I don’t have a letter in front of me, I’m actually wondering if this is generally things that the subcommittee thinks are good things, that we should then develop a letter and in January review a finished letter on something like this, and so I’m just sort of once again positing, sort of set up a letter and action item here. And I think we actually have Helene who wants to make a comment, hopefully on this topic, and then Jeff I know wants to make a comment and others. MS. GUILFOY: This is Helene Guilfoy, and of course my question is on the topic, but it might be considered as a statement as Carol’s was. Mike asked earlier, I think it was you Mike, that asked about internet, when Medicare would be allowing internet submissions, and Empire, who is willing to do this pilot study has said that because they’re a Medicare contractor they’re restricted in using Medicare submissions, which then causes them to have to do computer to computer, which will exclude generally the physician groups from doing anything, it will also exclude any other small provider from doing any of the participating in any of the pilot programs. Since Medicare is the larger payer in the country perhaps a recommendation from the committee to move Medicare along in permitting internet transmissions, at least to facilitate these pilots, because then we could include many more submitters, providers, in the pilot program if we could use the internet to submit these requests forward. So that’s my, hopefully was not a statement but rather was a suggestion for the committee to think about. DR. COHN: Helene, thank you. Comments from the subcommittee on the various things we’re talking about? Jeff? MR. BLAIR: My first comment is to commend the progress that’s been made within HL7, I think that the claim attachments that are using XML and CDA provide much more flexibility and I really think they’re a dramatic improvement over what had to be pulled together five years, I think it was five years ago. So I was impressed and I also was very pleased to see that Empire Blue Cross is providing a pilot. DR. COHN: Jeff, actually I think they’re proposing a pilot. MR. BLAIR: Thank you, proposing a pilot. And with the testimony that Jean has made, with the comments that Kepa has made, and Helene and Carol has made, I in general tend to support an NPRM process beginning as soon as possible. However, it sounds like there may be important information that may be obtained from a pilot and especially to serve the professional side as well as the institutional side, and if there’s some way to get more providers included, both institutional and physicians, I think that that becomes very important, because we’re leading towards a mandate and heaven knows with the experience we’ve had with other HIPAA regulations and mandates I think we need more information. And I think that Jean Narcisi’s comment that maybe we also need to get better information as to when an attachment is necessary and to try to standardize those I think was also compelling. So my thinking, I would really like to see the AMA if possible try to move quickly with NUBC and, let’s see, NUBC and NUCC, to try to work together to make those recommendations maybe in parallel with the pilot and to encourage maybe group practices and physician offices, to encourage them to become part of this pilot, and to try to move that forward as quickly as possible. And my thought is that I’m a little bit concerned if an NPRM begins to be coming out before we have the results of that pilot. DR. COHN: Wes? MR. RISHEL: I hope that the subcommittee will take into account the need for funded pilots and the funding mechanisms associated with creating the pilots, in particular we have this weird interaction between CMS internet policy and the ability to mount a pilot in the internet area, so if the time involved in working on CMS internet policy precludes Medicare contractors being used for a pilot with physician’s offices for the low end approach then there needs to be a funding mechanism to fund a commercial payer to operate that pilot instead of a Medicare contractor, or Medicaid or some other mechanism — too much the importance of the pilots but at the same time I’m very reluctant, the NPRM can be delayed at the last minute, I mean it doesn’t it have to be delayed in announced basis, so I’d just like to see us not create a reason to stop work on the NPRM at this point. DR. COHN: And actually let me make a couple of comments also. Jeff I sort of find myself tightening up with your term mandate, I mean be aware that when we’re talking about claims attachments really the only mandate is mandating payers to be capable of accepting claims attachments. And already in the industry payers ask provides every day, be it institutional or individual physicians, for additional information to be sent as part of a claim, so that could continue into the foreseeable future be it on a paper basis. So really what we’re talking about is getting at least the receiving end of the community computerized so that they’re capable of accepting it assuming that an institution or a provider would want to send it. And I think that there’s, I mean given that this is already going on there’s I think ample reason for us to move forward to at least try to computerize that recognizing that there probably is some administrative simplification and cost savings associated with that. The reason question I think which you’re bringing up and I do agree is how do you make sure that the learnings from the pilots are reflected in a final rule, and I think we maybe need to make clear that, I mean maybe an NPRM should not wait for a pilot but maybe a final rule should wait for pilot results to make sure that the final rule doesn’t come out a little differently, and maybe that’s more the issue. MR. BLAIR: No, no, I tend to want to move things forward, too, and that may be a good approach. I think what we want to do is try to encourage the expansion of a pilot to make sure that it includes more then institutional providers, and that it includes more institutional providers and that it yields good meaningful results to help so that the final rule in fact is acceptable to the industry as a whole. DR. ZUBELDIA: I’m raising my hand but you can’t see it. There’s two things that I think are important, one is to make sure that the participants in the pilot do comment with their feedback as part of the regulatory process. I don’t know if that would be as part of the open period of the NPRM or, and I don’t know this, perhaps there is another way, for additional comment to be solicited from the participants of the pilot and be considered as part of the process of writing the final rule. And perhaps there could be the NPRM period and perhaps through the gathering the feedback from the pilot through NCVHS there could be additional comments or additional guidance provided to the Secretary by the NCVHS before the final rule is written to consider the feedback from the pilot, I think that’s important. But like Wes said I would hate to delay the NPRM again just to get the results of the pilot, I think the NPRM needs to go out and then the feedback from the pilot incorporated into the final rule. The other thing is that if you run a pilot with one of anything it’s probably going to not be the same as the reality and it’s probably going to always succeed, whether it’s one payer or one provider or one vendor, and I think the vendors need to be involved in the pilot, both payer vendors and provider vendors. My inclination would be to run a pilot with more then one, more then one payer, more then one vendor, more then one provider and to try to find a community where the attachments provide enough of a critical mass so there is at least a few interested parties participating in the pilot. There are some communities in the country that are already providing this kind of brewing environment for critical mass, perhaps in Boston, perhaps in Utah, perhaps in Hawaii, where they have very active communities working on additional transactions and working on these kinds of things. But I think it’s important we get more then one involved in the pilot. DR. COHN: And Kepa, I think you need to disclose your state of residence. DR. ZUBELDIA: I live in Utah. But Utah has the Utah Health Information Network that has been involved and I have no affiliation with UHIN, but UHIN has been involved in developing transactions before HIPAA existed and using the X-12 transactions. Hawaii has a similar collaborative, Hawaii Health Information Network I believe it is, or Hawaii Collaborative, where there are several payers in Hawaii and practically all of the providers that collaborate in that network. Boston has a similar environment, Minnesota has a similar environment, there are environments that lend themselves to this sort of piloting even if it’s unfunded. DR. COHN: Now Stan you’ve been, what I’m trying to get a sense is about whether we should go forward with a letter or that we begin to produce and all of this stuff, and I’ve heard from Jeff that I’m hearing generally it sounds like he’s supportive of the idea of these concepts, I think Kepa it sounds to me like you’re also in agreement with sort of the key issues of recommending multiple pilots, recommending going forward with an NPRM and not waiting. Stan, comments? DR. HUFF: That’s my same feeling, I would like to see us go forward in parallel with preparation of an NPRM, as well as strong support for pilots. And coupled with that again, as Jeff has said, I’m convinced from Jean’s discussion that the more we could provide direction about when claims attachments are needed the better this will be accepted by the industry. One question I have in regard to that and the form for that, you talked in your testimony, Jean, about the NUCC and you note some provisions of openness in that, but does NUCC follow ANSI’s guidelines for an open consensus process? MS. NARCISI: Yes, we do. DR. HUFF: So you have elected, openly elected officers, you have to justify negative ballots, you have to — MS. NARCISI: No, we are not ANSI accredited but we do follow the, I guess let me back up. We don’t follow the ANSI accredited openness policies, we do follow the guidelines that were prescribed from HHS to be considered a designated standard maintenance organization. DR. HUFF: But not the ANSI open consensus standard guidelines. MS. NARCISI: No, we are not an ANSI accredited. DR. HUFF: But anyway, I’m in general, I favor going forward with an NPRM and with the pilots and I would encourage a form where more, if possible, and I’m like you, in talking with Clem it seemed like it might not be possible to reach consensus about when claims attachments are needed but we ought to try that, I mean it certainly would be a good thing if it could be achieved. DR. ZUBELDIA: I have a question for Sue Ryder. DR. COHN: Well, actually Jeff has a comment, Kepa, I’d ask you to make your question brief and then we need to sort of move on. MR. BLAIR: My comment is that in addition to indicating that we support an NPRM going forward I have a level of comfort that I’d like to express that I’d like to see in the letter, that the pilot site go forward as quickly as possible and that the, not only the AMA and I would expand it to the ACP if they’re available and other physician groups do everything that they can to work with Empire Blue Cross to try to get physicians involved, physician offices involved to participate in the pilot. I think that if you could expand the institutional ones as well I would say extending Kepa’s comment that if you only have one then you may not get the diversity to flesh out issues and problems. I have a comfort level that if you get, if you could target or try to get four institutional providers and four professional providers that once you get four or more my comfort level begins to increase, that you will begin to flesh out issues that will be helpful. I don’t know if you could put that in the letter but I would like to see that as a target at least, an attempt to try to do that. DR. COHN: Kepa? DR. ZUBELDIA: This is for Sue Ryder. I’m aware that New Jersey and New York, they both have a collaborative work on HIPAA. Sue could you recruit some of the other payers and more providers in New Jersey to participate in the pilot even if there is no funding from CMS? MS. RYDER: Absolutely. We did do an outreach to New Jersey Shore, which is a WEDI-SNIP regional SNIP associated and did get some response from them but we could do additional outreach there as well. DR. COHN: Well, what I guess what I’d like to do is to wrap-up this discussion and I think what I’m hearing from everyone, and I guess I just need either a nod of heads or Kepa, I know you’d put your hand through the microphone but I need to hear you say yes, is that Maria and I will sort of work on putting together a letter for our January subcommittee meeting to go to the full committee late January and hopefully go to the Secretary afterwards a, sort of encouraging the Secretary to sort of move resolutely forward with getting a claims attachment NPRM out, and also I think supporting, recognizing the importance and supporting the idea of either both government and other types of multiple pilots to be testing the claims attachment standard out and it’s really a time feasible not holding up the NPRM for this but more that the results should feed into the final rule. And I’m sure we’ll come up with other beautiful language that goes along with that but I think I was hearing from people, feeling that we ought to have potentially multiple pilots because I think we need multiple payers and certainly multiple providers both institutional and others being involved with this. Now I don’t think part of the letter is really the discussion of getting industry consensus on business uses but we can consider that, certainly any work done in that area would be desirable but once again I’m not sure that’s really part of the letter but we can sort of think about whether or not there’s some sentence about that about the desirability of increased industry consensus on the value, on sort of the use cases and situations. Is there anything else that we need to have in this letter? Susie? MS. BEBEE: I just have a question, I didn’t hear all of Karen’s HIPAA update but will writing a letter to recommend that the NPRM go out ASAP, what’s the status of it being written? PARTICIPANT: It’s in review. DR. COHN: It won’t be out by late January I don’t think. Other comments on this one? Kepa are you in agreement with what we’re sort of saying? DR. ZUBELDIA: Yes, I am. DR. COHN: Other agreement? Okay, well why don’t we do that as an action. I really want to thank the testifiers, obviously for coming to talk to us about this. This will not be the last time we talk about claims attachments but obviously hopefully maybe we’re going to help move things forward a little bit as a result of this conversation, and I guess we’ll be hopeful that we’ll see a Notice of Proposed Rule Making out shortly and maybe there’s some ways, if this turns out to be something of obvious business benefit maybe this gets implemented in advance of the final rule being, or not so much the final rule coming out but of the two year window, which would be I think a wonderful reinforcement of the business value of these sorts of initiatives. Well, thank you. Now, everyone else, we are running 35 minutes late on our agenda. In the remaining time I think we want to talk for a couple of minutes about the E-Prescribing piece and obviously this is going to be a very abbreviated discussion and then I wanted to reflect a little bit about what appears to be happening or suggestions for what it looks like we’ll be doing in January at our next subcommittee meeting. Agenda Item: E-Prescribing Planning – Dr. Cohn Now Jeff, you’ve obviously passed out copies of a draft project plan — MR. BLAIR: Preliminary. DR. COHN: A draft preliminary draft draft draft project plan. Now let me just ask I think sort of even almost before we start talking about that, and recognizing that we may in January actually need a briefing from HHS sort of talking about timeframes and all of this, but do we, I mean almost fundamental to all of this is what sort of timeframes are we talking about beyond which that which is mentioned in the legislation. And as I say that I’m looking sort at Jim Scanlon to see if he has any insights or otherwise, or if you don’t that’s fine but maybe by January we could have a better on that. Jim do you have any thoughts on this? MR. SCANLON: Well, let me just say a couple words. The President passed, the President signed the Medicare Prescription Drug Modernization and Improvement Act on Monday and there are a lot of jobs and a lot of responsibilities and a lot of deadlines for a lot of people obviously in that bill, the bulk of them fall to the Medicare program, but not only, there’s plenty assignments to a number of other people. Within HHS we have our folks looking at all the various requirements, what goes into effect immediately, and upon enactment, what are the deadlines, who’s responsible, what’s the best implementation mechanism, so that work is going on within HHS. Maybe for the January meeting we’ll try to give the committee at least an update on where that is. With respect to the E-Prescribing specifically that will be part of the overall plan but that, the provisions relating to standards for E-Prescribing have a designated series of steps, let me just mention because it’s actually, we’re looking at this already but let me start sort of immediately, the NCVHS is to develop recommendations to HHS for standards relating to E-Prescribing and we obviously will go into that later. The Secretary, and there’s no date for the NCVHS recommendations but we have to work backwards obviously, the Secretary of HHS is to develop or adopt an initial set of standards for E-Prescribing no later then September 1st, 2005. But let me continue because this is not the usual adoption. These will be used in a pilot study that would take place during the year beginning January 1st, 2006, and ending that January 1st, 2007. So it is a year long pilot study and evaluation that would be required. We’ll have to have our lawyers, and I think Karen and I have talked a little bit about this, when the Secretary develops or identifies those initial standards only to be used in demonstration projects, that’s probably not an NPRM situation, but it could be, we’ll have to see. And if it is then it’s going to be a very tough deadline to make, to have final, but normally for demonstration projects and tests that’s not normally a regulatory situation. But we’ll have to see, we’ll check with our lawyers. Having completed, let’s assume all goes well, having completed the pilot studies as of that January 2007 then an evaluation and a report to Congress is due no later then that following April 2007. And then HHS is then to adopt final standards no later then April 1st, 2008, and then the following year April 1st, 2009, the standards would be required. Again, this is the timeline outlined in the statute itself. So we’ll have to look, these are just deadlines, I think we’ll have to look at what’s required at each stage but you see the long, this long process that’s being scheduled here. So for the NCVHS to make those recommendations so that the Secretary would have at least an initial set of recommendations to announce or adopt or develop by September 1st, 2005 we have to work backwards, so we’re really talking about the next year, 2004, and then the better part of 2005. DR. COHN: I’m sure Carol will probably have a comment on this also. I guess to me the fundamental issue in all of this looking at any project plan is, and this issue as you described as whether or not there’s an actual regulatory process that involves an NPRM — MR. SCANLON: For the pilots. DR. COHN: Because obviously that has probably more impact on I think the great work that Jeff has done putting together a draft project plan then almost any other item. And so I was just trying to get a sense of how quickly we can get some resolution on that. Karen, do you have any comments about this? MS. TRUDEL: No, I think Jim pretty much covered all the high spots and we’re doing the same kind of analysis at CMS and don’t have a project plan with dates at this point. MR. SCANLON: We will try, Simon, to get a reading on what would be required, at least whether the lawyers are looking at an NPRM type situation or more of a study demo type situation for that first set of standards. DR. COHN: Well, certainly the reading of the law certainly would lead one to think that they were talking about demonstrations, but once again this is, I’m neither a lawyer nor do I play one on TV, so obviously we have to wait for I think that sort of guidance. Jeff, do you want to spend, I realize we’re sort of running out, I’m actually trying to make sure that everybody can get out at the time, but do you want to talk at all, actually not all, why don’t you talk for five or ten minutes about the early draft, early, early draft of the plan. MR. BLAIR: I’ll try to keep it to five or less and that way some people can maybe give us some feedback. I think all of the subcommittee members already have a copy of the preliminary, preliminary, preliminary draft and Jim, in addition to the guidance that you could give us on the timeframes, not only am I not a lawyer but I don’t see very well, and with the help of my assistant as I wound up pouring through the conference report trying to identify all of the requirements that Congress has directed to the NCVHS and how we go through this process you see that in the preliminary, in this preliminary work plan, and we are endeavoring to try to track through those references to other sections both in this legislation and other legislation. If it’s possible as you go through your stuff for you to provide some help or guidance to us to make sure that we don’t leave out a requirement in what we need to do beyond what the subcommittee can identify itself in reading the law that would really help us a lot. MR. SCANLON: Absolutely, Jeff. We’re actually, all of the HHS agencies are now looking through the whole entire text of the bill and there’s more then just CMS, you’re right, FDA is effected, CDC is effected, obviously the NCVHS is effected in their task, so I’ll think we’ll have a better sense of all of the various requirements. MR. BLAIR: Let me just for the rest of the subcommittee so that you do understand, it’s not just the prescription part and the reimbursement part, it refers to tracking eligibility, eligibility in terms of the patient, eligibility in terms of what drugs they’re responsible for, coordinating this with the formularies of the health plans, and referencing it to the voluntary portion as opposed to the Medicare portion of the overall legislation. So that’s what I’m referring to when I’m saying that there’s a lot of references to other portions of the bill and to other bills and that we coordinate to make sure that this is compliant with other standards such as the HIPAA standards, so that’s the maze that we’re trying to make sure that we don’t overlook a requirement as we go through. So that’s one piece. The other piece is that the conference report was more specific then I have seen in identifying the stakeholders that NCVHS should consult with for testimony, that is listed there in the work plan right up front, however, while it was a good start it didn’t include vendors of E-Prescribing services, of E-Prescribing software, it didn’t include entities like RXHUB(?), it didn’t include demonstration projects, HIMSS is starting one, I think there may be others that would be important testimony for us to receive. We’ve received other areas from our chair, in short what I would ask from the subcommittee members or anyone in the room or that hears us on the internet that has a chance to look through this work plan please advise us of any other stakeholders that we should be listening to or anything else that’s left out of the work plan. In this case Simon would you mind if I just go ahead and give my email address so that folks could send that to me and I could make sure the subcommittee gets another version of this as soon as I can pull this stuff together. Is that alright? DR. COHN: I guess everybody knows your email on the subcommittee, were you intending to broadcast your email on the internet? MS. GREENBERG: Everything could be sent to our staff here. MR. BLAIR: That’s fine, that’s fine. And then that I would indicate that with the information we have now we appear to be looking at somewhere between six and eight days of testimony with just the information we have right now, and that is the reason why the timeframe would be so tight. With that I would ask for any comments, suggestions, guidance in the remaining five minutes that we have. DR. COHN: Any comments from any of the subcommittee members at this point? DR. HUFF: Just looking at your preliminary plan, Jeff, the one thing I noticed is that it was your, I guess your informed opinion that we may need to hire a consultant to assist in this and your schedule at least for doing that was sooner rather then later, and I wonder if we need to decide that sooner rather then later so that we can get that done if we feel that’s necessary. MR. BLAIR: Thank you, Stan. MS. GREENBERG: We do have the capacity to do that but of course we need guidance on — DR. HUFF: It was a question, I mean I haven’t done this as many times as you guys yet, so I don’t know, do we think we need a consultant? DR. COHN: Yes, I mean normally, I think the expectation, I think the committee itself without augmenting staff can write two, three, four, even five page letters, but when we get much beyond that, and I think something like this probably would require a report to be attached to it, for that sort of work I think you need dedicated resources that this becomes their day job or at least part of their day job as opposed to their after hours job. So I think really the bigger issue is, it’s hard for me to imagine we wouldn’t want one, someone to assist, and one of the questions that everyone should think if they have particular names that we can consider in the evaluation. Marjorie? MS. GREENBERG: I would say two since I think the work plan suggests a similar process of identifying testifiers and then having a questionnaire filled out and then an analysis of the questionnaire, that’s very labor intensive and that’s something that, we’ve used a consultant before with the PMRI, I mean I think the good news from the point of view of just timing would be if an NPRM process wasn’t needed but the fact that it’s associated with that is that if you don’t go through an NPRM process even though it is for piloting you really want to make sure that we get all the input we can through the hearing process to guide the Department’s selection of even pilot standards. So if I recall we had a pretty aggressive meeting schedule before this happened so I would concur that a knowledgeable consultant who could hit the ground running, clearly that’s the important thing, would be quite useful. DR. COHN: And obviously we shouldn’t we come up with names in the open session but if other people have ideas they should email them to Jeff or myself or Marjorie or Jim and we can sort of look at it. Michael? DR. FITZMAURICE: One of the things that I was struck by and maybe I just missed it is that there is a standard for some E-Prescribing, NCPDP has a script standard and I didn’t see much reference to that in the committee report or in the law itself, and yet I think that’s one place that we would to rely on advice for in the committee is people who are already doing it and already working with pharmacies. So we should probably make that clear up front that we would want to rely on them for advice. MR. SCANLON: There actually is some mention like you remember in the statute relating to the standard for the card actually and it does refer to the NCPDP work and other things like that, so this is going to take a while to sort out. DR. COHN: Okay, good. Well, I mean recognizing that we’re beginning to wind down and I am trying very hard to make sure that everybody we do adjourn on time. I want to sort of transition us, we will have more time to discuss about this in that we have a certain amount of offline work. I did want to talk a little bit about the hearing dates in January as well as remind everyone that the third day of those, there’s actually a third day of those meetings now, which is the full committee as I understand. Agenda Item: Planning for Next Meeting – Subcommittee MS. GREENBERG: Excuse me, what was that? DR. COHN: We were talking about the January meeting and I was referencing that I, at least as I understand, I’m not sure I’ve seen anything in writing about a full committee meeting — MS. GREENBERG: Definitely on the 29th. DR. COHN: Okay, I guess it’s just I’ve not seen an announcement of it. MS. GREENBERG: But you will. And it will be up on the web. I’m sorry that it hasn’t been up there, this is all just kind of moving in real time here. But I think it was definitely agreed at the Executive Subcommittee meeting on the suite of meetings, which is on the 27th, half day on the 28th for this subcommittee? DR. COHN: Well, let me just talk about what we’re doing. I mean basically I think we’re, I mean once again we have a full day on the 29th for the full committee, 27th and 28th — MS. GREENBERG: Probably something maybe like 9:00 to 3:00 but in any event — DR. COHN: In any event, on the 27th I think what’s happening for our subcommittee is probably the morning of the 27th other CHI business since I believe we still have some final rules, final things to come forward on. In the afternoon we’re going to have a session on the security rule, which John Paul will be sort of taking the lead on, we’re working with him to put together the testimony and all of that. MS. GREENBERG: That’s not officially in conjunction with the Privacy Subcommittee? DR. COHN: No. MS. GREENBERG: They of course will be invited to participate. DR. COHN: Now on the 28th currently there is a discussion, at least one panel that we intended to have on dentistry and SNODENT issues, recognizing that was something that came up when we were talking about the PMRI and we’ve been, there have been issues related to all of that that we need to sort of follow-up on. We’ve also I think received notice from the head of the DSMOs, who is Frank Percorni(?) this year I believe, about coming to give us the year end report from the DSMOs and recommendations sort of for what needs to be happening for the standards going forward. Now beyond that I was actually hoping that we could reserve, and once again I don’t know if we have the whole afternoon or we have part of the afternoon or what exactly is happening but clearly we’re going to want to have a briefing around the E-Prescribing piece, I mean just to at least get some clarification on timing and expectations, and I’d like to have some time for us to really sit down and talk through the project plan. Certainly if we have until middle of next year to reach a conclusion we probably don’t have to start adding additional hearing dates at this moment but certainly if we have a more aggressive schedule all should be prepared to bring your calendars along. Now Marjorie has a question and then Jeff has a question. MS. GREENBERG: Well, not a question, just to mention that on the 28th the Populations Subcommittee is meeting in the morning, the Quality Workgroup is meeting in the afternoon, and we are hoping to have an orientation for new members from 4:00 to 6:00 on the 28th depending upon the availability of new members but that would probably involve participation by the members of the Executive Subcommittee, which involves the chair of this subcommittee. And that evening on the 28th we have dinner to thank the retiring members and welcome the new members, we’re planning a dinner, so I encourage everyone to participate, family and friends of the committee, and then a full day on the 29th, so that’s what we’re currently planning. DR. COHN: Okay, so I think what we’re describing is consistent with that, we’ll just make sure to adjourn by 4:00 p.m. on the 28th. MS. GREENBERG: No later. DR. COHN: No later then, okay. MS. GREENBERG: We’re going to take over the entire Humphrey Building for all these meetings. MR. BLAIR: Jim, I always do this, this is typical Jeff Blair, but if you are able to get a judgment from the attorneys as to whether or not there is an NPRM and if there is going to have to be an NPRM done, how soon we have to get our recommendations completed, if there’s any possibility that you could get that information to us prior to the January meeting where we could respond and fold that information into another version of the work plan where people could read it before hand and we could wind up in the actual meeting working on the issues and refinements of the work plan as opposed to just getting that information in January. MR. SCANLON: We’ll try to have at least that initial reading, Jeff, on what’s required and the overall timeline by then, for the January meeting, maybe before, in the coming weeks I think. MS. GREENBERG: Is there any anticipated need for a conference call of this subcommittee in January? DR. COHN: Jeff, you can’t see me but I’m saying no, I don’t think, I think late January will be soon enough for us to reconvene, unless there’s some other business that any of the subcommittee members have that we’re not aware of. I guess I should also mention on the two days of our meetings we will be considering that letter that we will have drafted but that’s once again something that could be handled that day. MS. GREENBERG: And the CHI letters as well. DR. COHN: That goes along with the CHI presentations. Okay, well, is there any other business coming before the subcommittee? Well, I want to thank you all for your hard work, happy holidays, the meeting is adjourned. [Whereupon at 12:30 p.m. the meeting was adjourned.]