[This Transcript is Unedited]
Dept. of Health & Human Services
Hubert Humphrey Bldg., Room 705A
Washington, DC 20020
DR. COHN: Please be seated. We are going to get started here in just a second.
Good morning. I want to call this meeting to order. This is the first day of two days of hearings of the Subcommittee on Standards and Security of the National Committee of Vital and Health Statistics. The committee is the main public advisory committee to HHS on national health information policy.
I am Simon Cohn, a physician and chairman of the subcommittee. I am the National Director for National Health Information Policy for Kaiser Permanente. I want to welcome fellow subcommittee members, HHS staff and others here in person, and I want to welcome those listening on the internet. I want to remind everyone to speak clearly and into the microphone if you would so the people on the internet can hear.
Today is the third of several hearings planned for 2002 on code set issues. The focus today and tomorrow is on HIPAA medical data code sets and whether there's a need for the possible expansion to the list of codes or a need for replacement. Today we are specifically focused on diagnosis code sets. Tomorrow we will be further discussing issues that have been identified from previous medical data code set hearings.
We want to thank Donna Pickett, Betsy Humphreys, Vivian Ault, Pat Brooks and Karen Trudel for their help putting these sessions together.
With that, let's have introductions around the table and then around the room. For those on the national committee, I would ask as part of your introduction, if you would note if there are any issues coming before the subcommittee today for which you need to recuse yourselves. With that, Jeff, would you like to start with the introductions?
MR. BLAIR: Good morning. I'm Jeff Blair. I am Vice President of the Medical Records Institute. I'm Vice Chair of the Subcommittee on Standards and Security and there's nothing that I can think of where I need to recuse myself or where there's a conflict of interest.
MR. AUGUSTINE: Brady Augustine, Corporate Director for Special Projects, Gambrel Health Care, member of the committee, member of the subcommittee. I have no conflicts.
DR. FITZMAURICE: Michael Fitzmaurice, Senior Science Advisor for Information Technology at the Agency for Health Care Research and Quality. I'm liaison to the committee and staff to the subcommittee.
DR. YASNOFF: Bill Yasnoff, Associate Director for Science, Public Health Practice Program Office, Centers for Disease Control and Prevention, liaison to the subcommittee.
MS. HUMPHREYS: Betsy Humphreys, National Library of Medicine and staff to the subcommittee.
MS. BEBE: Sissy Bebe, NCHS staff to the subcommittee.
MS. GREENBERG: Marjorie Greenberg, NCHS, CDC and executive secretary to the committee.
MS. TRUDEL: Karen Trudel, Centers for Medicare and Medicaid Service and staff to the subcommittee.
MS. SQUIRE: Marietta Squire, staff to the subcommittee.
MS. USABA: Virginia Usaba, Georgetown University.
MS. ARSWAGA: Patricia Arswaga, Blue Cross and Blue Shield Association.
MS. PICKETT: Donna Pickett, NCHS, CDC.
MS. FISHER: Elizabeth Fisher, NCHS, CDC.
MR. BURGLAND: David Burgland, NCHS, CDC.
MS. GLODGETT: Holly Glodgett, ANISIS.
MS. STEINBECK: Barbara Steinbeck, 3-M Health Information Systems.
MS. GRANT: Thelma Grant, 3-M Health Information Systems.
MS. FRASIER: Chris Frasier, American Academy of Professional Coders.
MS. WILLIAMSON: Michelle Williamson, NCHS, CDC.
MR. STEINDELL: Steve Steindell, Centers for Disease Control and Prevention.
MS. CHAPPEL: Jody Chappel, American Academy of Audiology.
MS. AULD: Vivian Auld, National Library of Medicine.
MR. MARTIN: Peter Martin, Acupuncture and Oriental Medicine Alliance.
MS. JACKSON: Debby Jackson, National Center for Health Statistics, CDC.
MS. BLOOM: Amy Bloom, NCHS.
MS. NERHEWAL: Nancy Nerhewal, College of American Pathologists.
DR. REGIER: Darrell Regier, American Psychiatric Association and its research arm, the American Psychiatric Institute for Research and Education.
DR. Mirin: Steve Mirin, American Psychiatric Association.
DR. MULDOON: John Muldoon, NACRI, the National Association of Children's Hospitals.
DR. KRAKAUER: Henry Krakauer, Office of Health Policy.
DR. HANBY: Pat Hanby, McKesson Corporation.
MS. SMITH: Louise Smith, McKesson Corporation.
DR. SERKOWSKI: Art Serkowski, Center for Devices and Radiological Health.
DR. POWERS: Laura Powers, American Academy of Neurology.
MS. LEON-CHISEN: Nelly Leon-Chisen, American Hospital Association.
DR. KLAUS: Linda Klaus, American Health Information Management Association.
MS. PROPHET: Sue Prophet, American Health Information Management Association.
MS. BONNELL: Claudia Bonnell, Blue Cross and Blue Shield Association.
DR. SONDIK: Ed Sondik, National Center for Health Statistics.
MS. GIANINNI: Laurina Gianinni, alternative link.
MS. FREIBERG: Elizabeth Freiberg, independent consultant.
MR. GILFOY: Wayne Gilfoy, Gilfoy Consulting.
DR. ZUBELDIA: Kepa Zubeldia, member of the committee, subcommittee.
DR. COHN: Kepa, thank you for being here. I actually want to thank you for your willingness to lead these sessions. You should indicate, obviously, whether there's anything you have to recuse yourself for regarding these sessions. I, for myself, would say this is one session where I have no conflicts of interest which is a pleasure. Kepa, do you have any beginning comments you would like to make on this session.
DR. ZUBELDIA: No.
MS. TRUDEL: I just want to announce to anyone who's not aware of it that late Friday afternoon three HIPAA regulations went on display at the office of the Federal Register. Two of them are, the subcommittee is extremely familiar with. One has to do with the DSMO changes, the addenda, and the other has to do with the national drug code and some other changes that relate to the retail pharmacy trans actions. All of those are proposals with a 30 day comment period and also the final rule for the employer identifier that would adopt the national, the IRS's employer identification number. So that's pretty good news.
DR. COHN: Karen, we want to congratulate you and the department for getting these regs out. Obviously, the timing is also good for the subcommittee and since obviously part of our role is to develop a letter for the full committee in response to these notices of proposed rule making, and it sounds that the timing will be such that we can develop this and be able to vet it with the full committee at the June full meetings.
MS. TRUDEL: They will be published on Friday. Thirty day comment period.
DR. ZUBELDIA: This meeting is somewhat of a continuation of the meeting we had April 9 looking at the code sets and the applicability of ICD-10. We looked at the procedure codes; we are going to be looking at the diagnosis codes this time. And then tomorrow we will be looking some more at the procedure code issues.
We have a series of people explaining to us what is it we need to do. Kind of a funny way to put it, but we are trying to figure out what is the best, most appropriate way to address these coding issues. They are very controversial coding issues and there is tremendous impact potentially on the entire industry on the decisions, the recommendations from this committee. So I want to invite Dr. Sondik to come and talk to us about the possible replacement of ICD-9 with ICD-10-CM.
Agenda Item: Possible replacement of ICD-9-CM, Vol. 1 & 2 (diagnosis) with ICD-10-CM - Dr. Ed Sondik
DR. SONDIK: Thank you very much. It's a pleasure to be here.
I have a prepared statement which we will hand out later this morning if that's okay.
I'm the Director of the National Center for Health Statistics, a part of the Centers for Disease Control and Prevention. It's a pleasure to be here this morning to talk about the implementation of ICD-10-CM as a replacement for the diagnosis portion of ICD-9-CM.
Our mission at the national center, as I think you all know quite well is to provide statistical information to guide actions and policies to improve the health of the American people. Our surveys and data systems provide the fundamental public health and health policy statistics that meet the needs of a wide range of users.
Now, the international classification of diseases, the ICD, and in fact, a variety of classifications are really at the core of what we do at NCHS. The ICD was developed and is now maintained collaboratively between the World Health Organization and 10 international centers. NCHS houses the WHO collaborating center for the family of international classifications for North America. That includes the US and Canada which has responsibility for the implementation of the ICD and the newest member of the WHO family, the ICF, international classification of disability, functioning and health, actually functioning, disability and health, to get it in the right order.
The ICD is the core of the family of WHO classifications which includes ICF, ICD-02 for oncology and various specialty adaptations of the ICD. In 1989, ICD-10 was adopted by the 43rd World Health Assembly. In the United States, we implemented ICD-10 a decade later beginning in 1999.
Since 1979, now that's 23 years ago, the US has used ICD-9-CM which was adapted from ICD-9 through extensive modifications including the use of fifth digit subclassifications. The structure of clinical modification revisions permits statistics in ICD-9-CM, to be collapsed back to ICD-9 for comparability between mortality and morbidity statistics. ICD-9-CM has been adapted and adopted by the federal government and the private sector for a number of purposes including statistical reporting, data collection, quality of care analysis, resource utilization, research and reimbursement.
Since 1985, the ICD-9-CM, has been updated on an annual basis to accommodate changes in medical technology and the need to provide greater specificity in classifying diagnosis and external causes of injury. Requests for modification are handled through the ICD-9-CM coordination and maintenance committee. This committee, which is co-chaired by NCHS and the centers for Medicare and Medicaid services, was formed to provide a public forum to discuss possible updates and revisions to ICD-9-CM. The committee discusses a variety of topics including the need to update changes due to changes in medical technology, the need for greater specificity in classifying diagnoses as well as the need to correct inaccuracies in classification.
No official changes are made before being brought before this committee. Copies of the topics and summaries of the meetings appear on our web site.
While the ICD-9-CM has evolved considerably in these last 23 years, it is rapidly losing its capacity to incorporate additional specificity, whether for newly identified diagnoses or over advances in medicine. There are few remaining areas in ICD-9-CM for expansions that are needed to enhance the classification details.
This is one of the reasons that your committee, the NCVHS, in 1993, recommended that the department immediately commit resources to assess the applicability of the international standard classification of diseases and related health problems which was ICD-10 for coding morbidity, to identify problem areas and to make modifications and necessary and to develop implementation plans.
In September of 1994, NCHS awarded a contract to the Center for Health Policy Studies to evaluate ICD-10 as a statistical classification for morbidity reporting in the US, specifically emphasizing comparisons with the ICD-9-CM. The initial purpose of this comprehensive evaluation was to, one, verify whether ICD-10 was a significant enough improvement over ICD-9 to warrant its implementation for morbidity reporting in the US and, number two, to develop recommendations to improve ICD-10 and to correct any problems identified during the course of the evaluation.
ICD-10, as published by WHO, has greatly expanded since it was first implemented for mortality reporting. ICD-10 includes significant improvements in coding primary care encounters, external causes of injury, mental disorders, neoplasms and preventive health as well as advances in medicine in medical technology that have occurred between the two revisions.
There are a number of other advances as well that have been brought into ICD-10 which includes new histologic types and more space to accommodate expansions in future.
Continuing the chronology, in 1996, in October, WHO had authorized the publication of ICD-10 in 37 languages with 28 countries having implemented ICD-10 for mortality or morbidity applications. The remainder of the countries were expected to implement ICD-10 before the year 2000. With the exception of the United States and Italy, both of which use ICD-9-CM, every other G-7 nation as well as Australia, has moved to the use of ICD-10 or a clinical modification for their morbidity data.
A technical advisory panel, convened by the NCHS contractor, concluded that there were compelling reasons for recommending an improved clinical modification version of ICD-10. In fact, that panel strongly recommended that NCHS proceed right away with implementation of a revised version, a prototype of which had been developed by the contractor. They stated that ICD-10-CM represents a significant improvement in the clinical specificity, the ease of use and access ability over ICD-10 and ICD-9-CM. Hence, we make the strongest possible recommendation that the ICD-CM's tabular list and alphabetic index be adopted and implemented as the standard US classification as soon as possible.
From the time of that recommendation, now many years ago, there's been further work on enhancements to ICD-10-CM, and this was largely carried out through the review of the coordination and maintenance committee minutes, collaborations with many surgical, medical specialty groups and as a result of comments received following the three month open comment period held in 1998.
We believe the resulting ICD-10-CM provides further additional classification detail and includes logical restructuring in the relative placement of certain entities. Further, ICD-10-CM remedies many cumbersome classification dilemmas that had impaired ICD-9-CM, such as a major expansion in the chapter dealing with factors influencing health status, and contact with health services and the musculo skeletal chapter.
Based on this work to date, we are confident that ICD-10-CM is well-positioned to replace ICD-9-CM and will provide the necessary data for statistical reporting as well as administrative transactions.
Now, it is absolutely true that some in the health care sector have expressed the concern that the cost of implementation might outweigh the advantages gained from ICD-10-CM, including the expense of retraining and converting computerized information reporting systems based on ICD-9.
The structure and conventions in ICD-10-CM are similar to those in ICD-9-CM. Therefore anyone who is familiar with ICD-9-CM will be able to move to ICD-10-CM without a major retraining effort. It is true that the alphanumeric structure of ICD-10 will require software modifications and there is no way to avoid the cost of developing new edits, lists of valid codes and other support packages.
However, the benefits of the added specificity, the improvements in clinical accuracy and consistency and the enhanced ease of coding should offset, as we can make these trade-offs, the conversion costs.
The availability of educational materials and training programs, and the HIPAA-mandated 24-month lead time before implementation will further allow for a smoother transition for the implementation of ICD-10-CM. It will facilitate integration of ICD changes with other system changes and new system development and allow the cost of implementation to be spread over several years.
NCHS will make available crosswalks between the two versions, ICD-9-CM and ICD-10-CM, and has plans to offer ICD-10-CM electronic formats as well as to produce this classification in book form. Although some have stated they would prefer to continue to update ICD-9-CM rather than implement ICD-10-CM, such a decision will make national and international comparisons and aggregations difficult.
This is particularly true because parts of ICD-10 have already been implemented in some fields. The neoplasm chapter of ICD-10 is comparable with ICDO-2 which has been used in cancer registries since 1995 and chapter five on mental health disorders is comparable to DSM-IV.
As I previously noted, the US has already implemented ICD-10 in 1999 for mortality reporting which has introduced challenges for states as they endeavor to maintain comparability between mortality statistics now captured in ICD-10 and morbidity statistics captured in ICD-9-CM.
We believe that the continued use of ICD-9-CM, in administrative transactions beyond the year 2004 would be ill-advised for the following reasons.
First, the ability to update ICD-9-CM beyond the year 2002 is very limited and may eventually lead to a loss of important data and second, for federal agencies and other organizations who are already engaged in international collaborative efforts or use mortality and morbidity data together, the continued use of ICD-9-CM would have negative program impacts.
Many in the US and the international community have awaited the implementation of ICD-10-CM and if questioned why our implementation has lagged behind that in other countries. We are considered a leader in both mortality and morbidity classification, and several countries have based their clinical modifications of ICD-10 on US work.
In 1999, the Institute of Medicine's Committee on Injury Prevention and Control recommended that a high priority be directed at insuring the uniform and reliable coding of both the external cause, the nature of injury using the ICD on all health systems data, particularly on hospital and emergency departments' discharge records. The committee noted that special efforts should be directed at training to insure the optimal use of the 10th revision of the ICD.
We believe we have an international obligation through treaty with the WHO to implement ICD-10 in the US. We've accomplished the transition to ICD-10 for mortality and must prepare to do the same morbidity applications.
Others in the health care sector have cited the need to conduct cost benefit analysis or a return on investment analysis. Surely as I have mentioned, efforts should be made to minimize the costs through appropriate tools and cross blocks. But quantifying the benefits to public health and health care of better information can indeed be challenging.
To improve our quality of life and our health status, it's critical that our public health and epidemiological data bases, which are used extensively throughout the US and the world, be based on the most current internationally recognized versions of ICD. As leaders in the world public health community, we really should expect nothing else.
The creation, review and revision of health policy relies on the availability of accurate and timely health care data generated by both providers and payers and by statistical surveys and other health research efforts.
As I mentioned earlier, classification is really at the core of what we do in producing health statistics. Our information systems need the most current classifications available. The ICD-9-CM has served this role in the US, as I mentioned, for 23 years. We believe it's time to move forward and implement ICD-10-CM. We need the improvement it affords for better health care research, patient safety and other quality of care improvements.
We cannot afford to continue to collect data using outdated classification systems at a time when health care increasingly calls for and expects better health care data to support the health care system.
I don't think it's a question of whether, but when, we should implement ICD-10-CM. How soon we should make the transition and how to best make that transition. We believe that implementation should occur as soon as possible with a target date no later than October in 2005.
Thank you very much for the opportunity to present this perspective and we look forward to a continuing and productive relationship with the committee in working with you to improve our classifications. Thank you.
DR. ZUBELDIA: Thank you very much. Excellent testimony.
I would like to start questions with a clarification. You said that chapter five of ICD-10 is comparable with DSM-4 and we heard a few months ago that DSM-4 and ICD-9 were practically identical with some differences in the description of the codes and when to apply the codes. Isn't that right?
MS. TRUDEL: ICD-10-CM and chapter five are almost, they are almost identical.
DR. ZUBELDIA: Not nine? Thank you.
DR. COHN: Thank you very much for your testimony. I guess I had a couple of questions. One is, I just wanted to make sure that I was understanding your dates that you had identified. You are actually advocating at this point for -- and this is supported by HHS -- for a 2004 roll-out which means that you are advocating for getting in this into the rule making process immediately and having a final rule completed before the year end. Is that it?
DR. SONDIK: Implemented by 2008, by October 2005. That will give us some latitude.
DR. COHN: Okay, the other question is -- Thank you for that clarification. It didn't seem like there was enough time for us to do things that needed to be don.
The other question, and I guess I'm just thinking ahead to a notice of proposed rule making which, at least from ones that I've seen historically, typically has issues, identifies benefits and costs and usually tries to quantitate a time them a little bit, at least from my understanding. Now, I guess I'm sort of struggling a little bit with how you see pursuing that. I mean, if I have been one of the people who has been asking for some idea of both costs as well as final benefits. Do you have a sense of the costs at this point? Do you have a sense of how we've approached the final benefits piece?
DR. SONDIK: Well, I think the costs, one can, there are ways to certainly approach that. The benefits, I think, are the more difficult part, and I think the benefits, probably, I think, to address them in some comprehensive way is difficult with this because what we're doing is providing a different language, and with that different language we may be able to determine things, understand better qualities of care. To me the great benefit here comes from what we will see as from the benefits of research that we are able to do by using this new classification.
At the same time, because the classification is more specific in terms of morbid conditions, diagnoses, it may well be that we are better able to understand care, the resources that go into, so maybe the best way to approach this would be to look at a number of examples, and to, as best as one can, quantify the benefits that would come from those examples. So it might be, for example, looking at a situation with a classification in which we get a more precise diagnosis and then tending to that, looking at the resources that go with that diagnosis as would have been appended to or been able to collect data on the earlier diagnosis.
So, for example, we might have a diagnosis that would be in, according to ICD-9-CM, all encompassing, let's say, and would include a variety of different conditions within it. With ICD-10-CM, we may be able to parse that out and understand better those conditions and the resources that are attendant to those and that could lead to improvements in the allocation of resources of care, understanding the costs of care, understanding the resources that go into addressing those particular conditions.
I think the ability to share our data internationally, for example, with Canada, would be, which has very similar practice patterns, if not health care system, certainly practice patterns and how diseases are addressed, I think that would be enormously useful and with other developed countries as well, and as I noted in the testimony, this will become an increasing problem if we don't make this type of change.
It's hard to quantify the benefits from increased knowledge, but I'm sure that we could identify some examples that would help to understand just what these benefits would be.
I also think there's a matter of a value judgment involved in the sense of whether or not the information that we are using is as current as it could be. If we can identify diagnoses that are new, that are not included or readily included in ICD-9-CM, but are included in ICD-10-CM, I think there's a value in and of itself in being assured that, in fact, we are tracking our health care system and the processes within our health care system as accurately as we can. I think, in part, that's a value judgment.
I mean, it says that should we be using something that's based on ICD-8? We would say no. Clearly that's out of -8. We need something that is as relevant and up to date as it can be. We have a very complex system as we all know. In order to understand it, to understand the kind of care that people are getting, who is getting care and of what types, who isn't getting care, the issues related to the quality of care, we need a system of data collection and appropriate classifications and attendant research that's able to give us the answers that we need and to monitor and improve our health care system.
MS. HUMPHREYS: I just want to comment as one who has attempted to map more current clinically specific terminologies to ICD-9-CM, that it doesn't work very well because if you have an updated current clinically specific terminology and you want to map it to the classification systems and diagnosis systems, it's extremely difficult to do because of the outdated organization within ICD-9-CM in some areas.
If the committee and NCVHS and the subcommittee is looking forward to the time where we will capture clinically specific information and in some cases we will be able to generate accurately higher level aggregations for statistical and billing purposes, that is a major problem with ICD-9-CM, in my opinion, having been personally involved with trying to make a current view of oncology, for example, or even breakdowns in various types of diseases, map up accurately to a view of the world which is, as Ed pointed out, going on 30 years old.
DR. SONDIK: Another way to quantify this, and I must say, given my background, I'm very sensitive to the need to quantify things and put it into particularly dollar terms. I think it's very difficult here, but I do think laying out, laying ICD-9-CM next to ICD-10-CM and looking at the differences would provide, I think, a considerable amount of information on why one needs to move forward from ICD-9-CM.
If, in fact, the modifications were relatively minor, and it seemed that ICD-9-CM was a framework that could accommodate the changes that we would expect in the future, then it's quite conceivable that it wouldn't be worthwhile moving ahead but I think you'll hear in at least in a portion of the testimony that the differences are so significant at this point and likely to grow that we need a framework that is more attuned to medicine today and the way medicine will be over the next five to ten years.
So looking at those differences, I think literally, looking at them laid out, I think could be very useful.
DR. FITZMAURICE: I resonate with a couple of things you said, Ed. One is the needing to understand our health system better, and the second one is the value of research being not quite known.
Our health system is about $1.3 trillion and in between 13 and a half and 14 percent of our gross domestic product. That's a large sector of our economy and we do need to understand that better.
How much are we spending on human genome research? I don't know the answer to that question. I don't have the dollars in front of me, but for whatever we're spending on it, we think it's worthwhile. We think it's going to help us understand the diseases we have now and in the future. As we map out the human genome, we've to relate it to something. If we are relating it to diagnoses that are coded in ICD-9, it's less specific. If it's related to diagnoses coded in ICD-10, it gives us even more power to the money we are spending on the human genome and looking at that relationship.
Also, as we accumulate five, ten, 20 years of ICD-10 data, we'll be able to comb that for greater specificity in the diagnoses, go back to tissue samples and blood samples that are collected and learn more about what we are finding out with the human genome and its relationship to diseases. The payoff will be much less, but a lot of the research investment we are making if all these things continue to be coded in ICD-9.
MR. AUGUSTINE: Let me just say that being a biostatistician, it's always good to have more accurate data because I can draw better inferences. So I'm a huge supporter of this, but being in the private sector I have concerns not only about cost and lead time, but also about the departments within HHS or CMS or the contractor's ability to meet these demands as well, and what input do you have as far as that concerns?
DR. SONDIK: Well, we stand ready to support this to the extent that we, whatever extent we need to go. We think it's extremely important. We work with CMS, and my understanding is that CMS is very much in favor of this as well, but I'm not here speaking for CMR. I'm speaking for NCHS and the CDC.
I would think we need to be more specific about what it will actually, what it will take to implement this, as I said in the testimony, we stand ready to support this with a variety of different tools.
We think that one portion of this, which is the coding and the training would be something that would not require a great deal of re-training on the part of individuals because the systems are relatively the same but there is no question that computer work and edits and so forth would have to be done. And, again, NCHS would be certainly there, together with our sister agencies to provide the help that we can with that.
DR. ZUBELDIA: Thank you very much for the testimony and answering all of our questions. Let's move onto Donna Pickett. Donna is going to give us another view of ICD-10-CM, and maybe this will clarify some of the differences between the two systems.
Agenda Item: Overview of ICD-10-CM - Donna Pickett, NCHS
MS. PICKETT: I'm going to fast forward through some of the slides since some of that information was actually covered in DR. Sondik's remarks.
What I would like to share with you this morning are a couple of things. One, Dr. Sondik mention that we are one of the last G-7 nations not to have implemented ICD-10 for morbidity or clinical modification of it. Of the countries that have actually implemented it, they are actually in some instances using it for reimbursement and/or case mix systems so that it is not an impossible task because we have the experience of other countries to look at to see that it's actually been accomplished so I just wanted to make note of that.
Expanded detail. As an example of the detail that is now in 10 and 10-CM, and you will see another slide later on this, currently in ICD-9-CM, we have one three-digit code, category rather, for diabetes, and then there is expansion at the fourth and fifth digit level. If you look at the comparison with ICD-10, and that's the WHO version, you'll see that that 250 code thousand is represented by five new codes, again within ICD-9-CM, we cannot make the expansions that are in ICD-10, and later you will see a slide where we've actually accomplished greater specification and enhancement in ICD-10-CM beyond that which WHO has given us in ICD-10.
There's been some restructuring of codes, and again, this is to provide additional room for expansion and more expansion also occurred in the external cause of injury chapter where you had the external cause of injury, E-800 through E-999. That now in ICD-10 translates to V-00 through Y-98, again a great deal of expansion there upon which we've actually added much more enhancements in 10-CM.
And in creating ICD-10-CM, we've actually worked with a number of groups at various phases. We've worked very closely with the medical specialty surgical societies, professional organizations and other federal agencies so that the work that we have managed to accomplish with the creation of ICD-10-CM, was not done in a vacuum. We have reached out to the many partners who've worked with us over the years, those who have participated in the ICD-9-CM coordination and maintenance committee process and those who had problems with ICD-9-CM, but realized that changes could not be accomplished within ICD-9-CM, but this certainly was an opportunity to make changes in ICD-10-CM.
Now, what have we done in ICD-10-CM? We've added a sixth character, we've added laterality, some of the more complex structures of ICD, we have managed to modify as we did in ICD-9-CM to make this a truly diagnosis classification, we've removed any relationships with procedures and procedure codes from the classification. We've added trimester information to the OB codes, we've revised the diabetes codes to be consistent with the American Diabetes Association new categorization of diabetes; we've expanded the injury codes, post-operative complications, alcohol and substance abuse codes; and we've added code extensions for injuries and external causes of injury.
Now, earlier I showed you this comparison between diabetes and ICD-9-CM with one three-digit category with expansion at the fourth and fifth digit level and then the ICD-10 by WHO which expanded that further. If you look at what we have in ICD-10-CM, we have harmonized the categories in ICD-10-CM with the American Diabetes Association new categorization. I shouldn't say new because that categorization was actually implemented in 1977 so it isn't exactly new, but again, it's a concept that could not be retrofitted into ICD-9-CM, but we had an opportunity to harmonize in ICD-10-CM.
As many of you know who are familiar with ICD-9-CM, under open wounds, you have a number of categories that are categorized, the group there. You have got laceration, with and without foreign body, puncture wound, with and without foreign body and bite. All of these things go to open wound. Similarly, in superficial wounds, you have a number of things that are actually grouped under superficial wounds, trying to relate that information to outcomes research or any additional detailed analysis, by having these all grouped together pretty much renders that analysis nearly impossible unless you go back to source document, the medical record and for those who are working in research and only working with data that has been already abstracted, you don't have the ability to go back to that information.
The previous slide was just showing you the expansions that have been made at the fourth and fifth digit and sixth digit character to show that we have expansion for laterality, for body site, and to break out things that are specific to, under superficial, whether it's an abrasion or a hematoma or along those lines and the same thing for open wound. We also have expansion for post-operative complications which has been considered very important for the work that's going on in medical errors and patient safety.
Currently in ICD-9-CM, you have codes that are in the complications chapter, and you have one code that doesn't provide you necessarily body system information or the exact type of complication unless you pair that with another code. In ICD-10-CM, we have codes within the body system chapter that specifically identify the different types of complications and you have two examples in your handout and here on the screen.
There has also been combination codes created for poisoning and external cause of injury. Currently in ICD-9-CM, it requires two codes to identify the poisoning as well as the intent or whether it was accidental or self-inflicted. In ICD-10-CM, we've actually merged those two concepts so you get all that information in one code without having to look for two codes.
With the injury and external cause chapter we've actually created extensions and this comes at the request of many in the injury community and elsewhere who have struggled to use the injury codes to do a number of things. One, to identify types of injuries but also to look at health care services utilization for repeat encounters for fractures or open wounds.
However, in ICD-9-CM, you can't keep using the code in, in fact, the fracture is healing. There is a code for healing fracture but it's in the after-care section. You don't have site specificity associated with that so there's no way to actually go back and try to aggregate the data to see what your true costs are. Some have actually violated the coding guidelines by saying if the patient is coming back for this particular, let's say, fracture, for instance, just as a check-up, please use the fracture code again when, in fact, it is not a new fracture, it's a healing fracture, so by adding extensions to denote initial encounter, subsequent encounter, you will have that information as part of the code.
And similar extensions have been added to the fractures to denote whether it's an initial encounter for a closed fracture, an open fracture, or whether it's an encounter for routine healing.
Replacement of occurrence codes which are in ICD-9-CM, what we've actually done is to expand the code beyond the five digits and actually have details so where you see an X at the end of a code, for instance, the Y-920-X, that means there is additional detail so that there is further delineation of home which is included in the WHO version, the driveway your home, the pool of your home or actually rooms within the home and, again, wherever you see an X it means that additional expansions have occurred to provide greater detail.
A new concept that was added to ICD-10 by WHO is activity: What the person was engaged in when they were injured. Now, what WHO did was provide that as a fourth and/or fifth digit character. However, having done that, it meant that there could be no future expansion and many that worked with us indicated that they needed the detail at that particular level so we've taken activity and actually created a new three-digit sub category, a three-digit category with sub categories and fifth digit codes for expansion so again, where ever you see an X that means that there is additional detail with that particular code.
The status, we've incorporated the public comments from the posting that we did back in 1998. We finalized the tabular list revisions. We have a revised index and a crosswalk. We will be posting ICD-10-CM shortly on the NCHS web site and we do have an alpha version of the data base and work will continue on that this year. We will be working on revised guidelines. What we are doing now in terms of revising guidelines is to move to the new format that we've actually created under ICD-9-CM.
This year, the cooperating parties agreed that it was time to reformat the guidelines to make them easier to use and we are just about finished with that exercise and so once that has stabilized we will take the 10-CM rules that are applicable to 10-CM and move them into the existing coding guidelines. Again with additional detail, we needed additional guidelines to make the use of the classification adequate and accurate and consistent over time.
What are the disadvantages of remaining with ICD-9-CM? As Dr. Sondik pointed out, the classification is 20-plus years old. We have absolutely almost no room to further modify the classification. Some of the examples that I just showed you for diabetes, the modifications we've made for injuries, we have no ability to include any of those things in ICD-9-CM. We are pretty much out of space, and some of the things that are in 10-CM were by request of the people that have worked with and are users of the ICD-9-CM so it's very difficult to keep going back to these groups and saying, you know, great idea, but we can't accommodate your proposal because there's no room in 9-CM but those things are included in 10-CM already and in some instances, some of the proposals that have come to us were actually already in 10-CM and now we are struggling to figure out how to place it into 9-CM and sometimes not being able to do it completely.
We are losing data. There are a lot of things that are happening in the health care system, as we well know, and 9-CM cannot capture that information, and again, as Dr. Sondik indicated, the states are already using ICD-10 for mortality and now have a bit of a challenge trying to compare their 9-CM collected data for morbidity with their ICD-10 mortality data.
We believe that the number of expansions that we have made to ICD-10-CM, are relevant for epidemiological research and decision support services and we think that that data will also be able to be used in a number of other settings, including grouping, case mix, reimbursement, and patient safety and surveillance and prevention activities.
We've harmonized parts of ICD-10-CM with other classifications. Chapter five of ICD-10-CM, has been harmonized with DSM-4. Chapter two of ICD-10-CM, is already in use by the cancer registries by virtue of their use of ICDO-2, and we've actually harmonized with the nursing classification, approximately 90 to 95 percent of what is described by way of diagnosis in their classifications.
These issues have just been covered by the panel so I will move on. And for those that are interested, that is the web site address for NCHS and at some point, as I indicated, we will be posting the ICD-10-CM classification, the tabular and the alphabetic index on our web site.
DR. ZUBELDIA: Thank you very much. This was very, very informative, and your face shows it. Let me ask you about some of this. The kitchen sink effect that I see here, and it is really concerning, in classification you have added all these extensions including place of occurrence and activity, I'm really concerned now that I see this that this place of occurrence is unrelated to a diagnostic classification. It's just a place of occurrence, and that probably should be conveyed in some other way.
Activity is probably not best conveyed this way, but maybe with a different coding system. Are you trying to create a coding system that will satisfy all the needs or is this just for diagnostics?
MS. PICKETT: Let me clarify. The place of occurrence, external cause of injury code, it's already in ICD-9-CM. It already exists. This was a code that was created by WHO, and it is in 9-CM. What we did in ICD-9-CM was to expand it to give it its own unique three-digit category and then further expansion, but the code already exists and we've moved that into ICD-10-CM because the code exists in ICD-10 as published by the World Health Organization.
The additional specificity that we've added to the code comes at the request of the users of the classification to be able to have more information about the place of occurrence other than saying home, where in the home so that they don't go back to the clinical record. But, again, it's not a new concept. It's already in ICD-9, ICD-9-CM, as well as ICD-10.
For the activity, that is also a concept that is included in ICD-10. All we've done in ICD-10-CM, is to expand the specificity under that particular category so we are only expanding on top of what WHO has already put in place and is already being used internationally. So those codes already exist.
DR. COHN: I just wanted a clarification. I actually talked to a number of my consultants within my organization when I was trying to find out what was going on with ICD-10-CM, and many of them sort of commented that it was not available currently because it was not on the web site. Apparently it was, as I understood, being worked on for the last year or two.
And so, are we being requested to consider something that isn't finished? Has this been reviewed publicly, the complete version? What is it that everybody saw last time that was up on the web site versus what you are proposing now?
MS. PICKETT: What they saw on the web site in 1998 was a draft version which we invited public comment on primarily to see if, in fact, we had included the right amount of detail, whether some of the expansions that we had made were on target or whether we needed to go further and from those comments, I think about 48 percent of the comments that came in during the public comment period, were actually focused on the injury and external cause of injury chapters but based on those public comments, we did go back and provide additional work on the classification to expand it out. We worked with a few other groups, the American Academy of Orthopedic Surgeons and some others that saw on opportunity to provide additional detail that was not in the draft that we had posted on the web site.
The version we are about to post on the web site is what we are calling a near final draft. We are posting it, again, not for content changes, but to have people take a look at it one more time and just familiarize themselves with the changes that have been made, but what we will be doing is for some of the more recent inclusions that we've made to 10-CM, we will be doing another testing of the classification just to make sure that the alphabetic index is complete enough so that the concepts that we've included actually can be found in the index because you start with the index and then go to the tabular and to make sure that the documentation that most of the HIM professionals would be dealing with, actually would allow them to use the codes as we've created them based on the detail that we've added.
So it's a near final draft. I would not say it's final final, because we did want to do some additional testing, but again, for many of the changes that we've made, for instance, the diabetes and some of the cardiovascular changes, those changes have already been adopted or adapted by other countries so they are already in use. But again, some of the fine tuning that we've done, we've made some other modifications to the OB chapter. We want to make sure that these things are useful, but we believe they are and again, some of these things have already been implemented but again, we just want to make sure of that before saying that this is final.
MR. AUGUSTINE: I would like to revisit the issue that Kepa was talking about, and I'm not familiar with the place of occurrence and activities codes. Now, since they don't actually describe it as disease state or condition or injury, is there direction given that those shouldn't be entered by themselves, that they are kind of add-on codes?
MS. PICKETT: The external cause of injury codes are always used in combination with an injury code or another code. I mean, that is the way that the guidelines are written currently in ICD-9-CM and so it would be applicable to the guidelines in ICD-10-CM. What needs to be added to the guidelines for ICD-10-CM would be the additional instruction on the application of the activity code.
DR. FITZMAURICE: For patient safety, one of the important bits of information you like to have for a patient in the hospital is, is it known whether the patient had this condition on admission to the hospital or was it acquired in the hospital after admission? Is there any part of the code in ICD-10 or ICD-9 that would relate that information or should this be just an separate variable if it's desired to be collected for patient safety?
MS. PICKETT: Based on some of the discussions we've had with the NUBC and how that might find its way into reporting format, that information at this time would be deemed to be separate from the classification itself.
DR. FITZMAURICE: Thank you.
MS. GREENBERG: If I could just add something. It is actually going to be in the health care service reporting guide. It is in the next version of the 837 and hopefully will actually be in potentially in the next version of the HIPAA implementation guide but it's now in the 837 standard but it's a separate element.
DR. FITZMAURICE: Thank you, Marjorie.
DR. YASNOFF: You and Dr. Sondik mentioned this 1994 study, and I'm wondering if in that study any methodologies were used or identified or any data was developed that would be helpful in supporting the current recommendation for conversion.
MS. PICKETT: I would have to go back and lack at that report. It's been a while since we've looked at that -- 1994 -- so I could follow up with the subcommittee.
And just as one aside, staff just pointed out to me that there is a code in ICD-10-CM for nosocomial infection that could be used in combination with the disease entity or the condition so that there is a way within the classification itself to make that link.
DR. FITZMAURICE: But not for something like a broken arm or broken leg. I mean, I'm not complaining. Something is better than nothing. Thank you, Donna.
MR. BLAIR: I imagine that you considered this as you were developing the latest version of ICD-10-CM for diagnostic codes, and that would be what you might have anticipated during the next 10 or 15 years. Our understanding of human genome and DNA, what effect that would have on diagnosis would have an exponential increase in the need for certain sections of ICD-10-CM.
MS. PICKETT: Yes, that possibility certainly looms large and there is some unused codes in ICD-10-CM so at some point if we needed to consider expansions to 10-CM, there is room to do that.
DR. ZUBELDIA: Thank you. Let's change the schedule a little bit. Let's have the break and let's see if we can limit the break to about 10 minutes and then ask Elizabeth Fisher to give us the summary of the code set hearings. So we will reconvene at 10:25.
(Brief recess.)
DR. ZUBELDIA: Let's get started and see if we can catch up time. We now have Elizabeth Fisher with NCHS to present the summary of the 1997 NCVHS hearings on this topic.
Agenda Item: Summary of May 17-18, 1997, NCVHS hearings on code sets - Elizabeth Fisher, NCHS
MS. FISHER: Thank you. I guess my job is to sort of transform you back in time a little bit to summarize the meeting that took place about five years ago, the meeting of the Subcommittee on Health Data Needs, Standards and Security, to give you just again a summary of what happened at that meeting. There's still the testimonies and the transcript are available out on the web site but they are about 300 pages long, so I'm hopefully saving some time of having to read through all those.
The purpose of the meeting was to receive input from users of classification systems, code sets, vocabularies, nomenclatures, those were the terminologies everyone used at the meeting, and to discuss the implementation of the HIPAA standards for code sets. Over the course of the two days, there was testimony considered for both diagnoses and procedure code sets, but my summary today is going to concentrate just on the diagnosis code sets.
There were several testifying panels representing all different aspects of classification use which included professional health care providers, health care facilities, special data users, providers and health care plans, public health and research, federal agency data users and developers of coding and classification systems.
I provided a slide here which probably looks like alphabet soup but it's just a list of all the presenting organizations over the course of those two days.
The areas considered at that meeting were a series of questions that were posed to the presenters to provide input on what they were currently using and describe how they were using the code sets that they used to recommend an initial standard to use, given the time frame of the HIPAA regulations, to recommend whether migration to ICD-10-CM, was possible by year 2000, how best to achieve the goal of simplification and retain medically meaningful information, to provide input on to maintain the code sets and what resource implications they saw would be needed in their institutions and their organizations in changing to a new classification of code sets.
Additionally, there were a number of other questions asked which -- I'm not necessarily going to summarize as much as the previous questions -- but additional input was requested on whether one classification system could serve most users, whether the same disease classification could be used for administrative transactions as well as statistical reporting; whether the current systems in use represented their disciplines well; any problems that they had encountered with linking coded data; the implications of these code sets and systems with the computerized patient records; the market acceptance for recommended systems; how the national committee could assist the US in developing international systems; whether single procedure classifications should be considered and requiring universal acceptance of code sets.
So the first area of consideration was what code set or what classification system was currently in use, and in reviewing all of the testimonies, it seemed that the main one in use for diagnosis reporting was the ICD-9-CM for a variety of uses, which I have listed some of them here which were disease indexing, statistical reporting, provider billing and reimbursement, disease staging, performance indicator measurement, longitudinal research and HETUS measures.
Additional systems which were mentioned in testimonies included the DSM for mental health disorders, SNOMAD, the LOIN Codes, the read codes, the national drug codes and the nursing diagnosis classifications.
Some of the strengths listed for the ICD during those two days of testimony included that it's a well-known classification used by many people, many organizations, including hospitals, physicians, other providers of care, payers and data system users. In addition they felt that training and education methods were already in place for the ICD. There was international recognition, comparability, and that you can map the US clinical modification version back to the WHO version. Data systems are set up to accept it. It's an open system with availability for -- it's in the public domain -- and the comfort level was high with the ICD.
There are also weaknesses listed for the ICD, that it was not specific enough for all data users needs. It did not provide enough clinical detail; it couldn't classify severity of illness, in particular preventive medicine, home health and nursing users were specific to needing more detail in the ICD to be able to use it.
They also felt, in some cases, that it was a lengthy time frame involved. It updates the system into add new codes to the classification and also that ICD-9-CM at that time was identified as running out of space to add new codes and that the use of the ICD was not consistent in use by all users, and what seemed to commonly come out was the interpretation by payers.
So when asked what to recommend as the initial standard for implementing the HIPAA regulations, most said to continue to use the ICD-9-CM for diagnosis reporting with a plan to migrate to ICD-10-CM. Some said that a combination of using ICD either 9- or 10-CM with other systems -- I would use on example of DSM but they mentioned to use it in combination with other classification systems.
And some said use the ICD with crosswalks and links to other systems such as SNOMAD and the reed codes.
When asked whether they thought it was possible to implement ICD-10-CM by year 2000, again this was 1997, most felt that was too ambitious of a timetable, many were already dealing with Y-2K issues and then their own systems enveloped. Adding that was just going to be just too ambitious to try and take on. That they also felt the need to evaluate ICD-10-CM because some had not fully been able to look at the entire classification, and this was before it had been posted on the web site for open comment.
They suggested, of course, that whatever standard was implemented that thorough field testing be done and a cost benefit analysis be completed. But many said not to delay implementation of ICD-10-CM too long citing the reasons I have already given -- the lack of space to expand and comparability internationally.
When asked how, what was the best way to achieve administrative simplification while still attaining medically meaningful information, some suggestions included to have just a single classification in use while again others said combining classification systems to meet all users needs was a better approach.
Many people talked about having a single set of guidelines to be adhered to by all, and I have a separate slide discussing that. Again, the concept of having thorough testing and a cost benefit analysis before implementation of a new standard was brought up, and also the use of modifiers to enhance data collection.
In terms of the common set of guidelines, this seemed to be a common theme throughout a lot of the testimony. What was brought out in that regard was to have one set of guidelines for all health care encounters instead of in-patient versus out-patient, to have uniform adherence to those coding rules by those assigning the codes as well as by users of data and in particular, again what was mentioned was to require all payers to abide by one set of rules published by an official source and that all payers must use and accept all the codes in the code set and its guidelines without modification.
When asked for input on how to maintain code sets, while most felt that having a code set in the public domain without copyright restrictions was the best way, and to use the current coordination maintenance committee process with the annual regular updates that took place. Also, to allow users to voluntarily input, provide input to the updated classification including physicians, professional groups and other users of the classification to make sure that maintaining the code set was responsive to all needs of the health care community, all users of the classification, to have it readily available to the public and that the maintenance process should be government funded.
When asked about resource implications of implementing new code sets, of course the common themes that we've already heard about were brought out in terms of training all of the users and the various types of users, the coders, physicians and data users, the comprehensive system programming and conversions that would need to take place; that there would certainly be a proficiency level and learning curve involved for a time before people became fully comfortable with a new code set; the need to have crosswalks and reporting to compare data across the different systems; have resources to remap rules and tables and existing systems; and, again, to always weigh the benefits of change against the resource implications that would be required.
So, I have summary conclusions. I don't know if that's really the proper term. Maybe it's just again the common themes that were brought out during those two days of testimony that there are many different classification and data needs by the different users. For reporting diagnoses, it seemed to be a common theme to continue to use the ICD using the ICD-9-CM for now and then moving eventually to ICD-10-CM following its complete development and testing.
To have the code set in the public domain with no copyrights and to allow open input and access to the maintenance process; to have uniform coding guidelines, reporting requirements, and code acceptance by all of the users; to have a cost benefit analysis done prior to moving to ICD-10-CM and to possibly have ICD linked to other systems.
Additionally, this is a few other recommendations and feedback, common in the testimony were to make sure that any change must be compatible with all the payment systems in place right now -- the DRGs, the APCs, etc.; That one system may not serve all data users' purposes and we should always consider all the users rather than just hospitals in selecting and implementing code sets; to use existing proven coding systems and to keep considering the impact of these systems as we migrate to a computerized patient record.
Any questions, I would be happy to answer them.
DR. ZUBELDIA: Thank you, Elizabeth. It's great to be reminded what happened five years ago.
I have a question. Was there any discussion on the testing, development and testing? It says that one of the conclusions of the testimony of important diagnosis continued to use ICD-9 now then move to ICD-10 following the development and testing.
MS. FISHER: I think what I read in all the testimonies in the transcript was, again, this was before the classification was placed out on the web site for open comment so a lot of people were even very unfamiliar with the clinical modification changes that had already taken place and they wanted to make sure that before we migrated to ICD-10-CM that there be some open comment as well as some testing of the classification. They just felt unfamiliar with it because they hadn't fully seen it.
DR. ZUBELDIA: Was there any discussion on how that was to be conducted? Was the government to run a pilot somewhere for morbidity or for --
MS. FISHER: There really wasn't any discussion in the testimonies that I recall as to methodology or anything like that.
DR. FITZMAURICE: I have got to fess up that Bill Yasnoff leaned over to me and said this sounds like what we are hearing today. Was anybody here at that hearing and I said I believe I was at that hearing.
MS. FISHER: I remember.
DR. FITZMAURICE: And it seems like it was yesterday that things haven't changed since this hearing very much. We still have a job ahead of us.
MS. HUMPHREYS: I think that's the one fear. There's a lot of these people who have had an opportunity to review and then attempt to apply ICD-10-CM since 1997.
DR. ZUBELDIA: Okay. Thank you very much for the reminder.
Let's bring in the first panel. We have a series of providers who are going to talk to us about the possible replacement of ICD-9 with ICD-10.
This is a series of the panels that we have during this meeting on the users of the diagnosis classifications. So we are going to start with Dr. Powers. And I would proceed with the panel from our left to our right.
DR. POWERS: Thank you for the opportunity to present today. I'm Laura Powers. I'm a neurologist, a board certified neurologist. I represent the American Academy of Neurology.
I am a practicing physician. I have been in practice for almost 20 years at this point. I was around when we started using ICD-9-CM codes for office billing and it was in the early 1990's that I became interested in this and also that I started working from the Academy with NCHS in order to improve neurologic diagnoses in ICD-9-CM.
The American Academy of Neurology represents 18,040 members and 13,652 of those people are board certified neurologists. Others are researchers and affiliated persons and non-board certified neurologists. That may not be a large number but the diseases that we treat including stroke, multiple sclerosis, Alzheimer's disease and epilepsy, have a rather significant effect on US health care dollar expenditures.
Often regarded as mysterious and previously regarded as having very limited treatment, neurologic diseases are increasingly understood and treatable. There has been a huge explosion in neurologic treatment in the last 10 to 15 years. Finalizing the identification of the human genome has been a boon to the field of neurology.
With its expansion of knowledge our diagnostic classifications have changed and at this point in time ICD-9-CM no longer keeps up with the changes in neurologic classification. There are several limitations to ICD-9-CM and I feel that I'm just repeating what other persons have already said and what other people will say today.
The first is that there are a limited number of codes. In the last 10 years that I have been working with NCHS in order to improve neurologic diagnoses, the number of codes has dwindled. The number of fifth digits is almost non-existent anymore and we are having to put codes in different places in order to place in new commonly used diagnoses.
Some of our diagnoses are scattered. As an example, sleep disorders are in the symptom chapter at this point. In ICD-10-CM, they are back in the neurology chapter. Tension headache is a problem for us and I'm sure the psychiatrists may not agree with me, but in ICD-9-CM, it is coded as 307.81 with a psychalgiss and basically what this does is it results in a denial of access of tension headache patients to neurologists who are the most common persons to treat those. We see tension headache as a muscle tension headache, that it is an extension of, to some extent, vascular headaches and the whole headache spectrum, that it is not a psychiatric diagnosis, and it has resulted in a denial of access to patients, a denial of access for patients to neurologists.
In ICD-10, tension headache is back with the other headache diagnoses in the neurology chapter.
Once again, older classifications are a huge problem for us. Mitochondrial diseases are essentially non-existent in ICD-9-CM. We know that there are many diseases previously diagnosed otherwise that we now know are of mitochondrial origin.
Essential tremor in ICD-9-CM is listed as essential in other specified forms of tremor. Here, again, we have an access issue in that there are many carriers who feel that's too non-specific a code to use whereas approximately ten percent of the population has essential tremor.
The epilepsy codes in ICD-10-CM structurally are not that much different that ICD-9 but the language involved is so much more similar to the language that we, as neurologists and other physicians use in describing these diagnoses.
In ICD-9-CM, transverse myelitis is given a fairly non-specific code as 323.9 which implies an infectious etiology whereas most of the transverse myelitis patients that we see are non-infectious. In ICD-10 that's fixed. It's placed with the demyelinating diseases.
These are only a few examples. I can already tell you that in our own extraction of the neurologic codes from ICD-9-CM that we provide to neurologists, we performed a crosswalk with stroke codes. We looked at several stroke classifications and decided that the World Health Organization and the ICD-10-NA, neurologic adaptation, for research coding for stroke was actually the most appropriate so we are very happy with proceeding to ICD-10-CM for stroke diagnoses.
And then there's the hassle factor. As a physician in the office, we tell physicians, we teach physicians they should code themselves because only a physician knows some of the details as to whether or not a headache is intractable or not intractable, whether epilepsy is intractable or not intractable, whether it's a dominant hemisphere stroke or a non-dominant hemisphere stroke.
On the other hand, there are so many nuances to knowing how to use each code, not only do you have to know the general guidelines of coding, but you have to know what has come down from coding clinic, about how to use specific diagnoses and you have to know some other information about the codes and it's very difficult for someone who hasn't had full coding training to code these codes in the office. ICD-10 is so much more similar to the language that we use when we are describing diagnoses on an everyday basis but that will eliminate some of these problems.
As a result of having difficulty choosing a code, less than accurate coding exists, and I have had several persons who do research tell me that using ICD-9-CM for research purposes, if you are really relying on the data that comes from that, it's not reliable because people out there are not coding correctly.
We have certain requests for the plans for implementation, of course, with timing, the timing is built in. I understand from what was mention this morning that HIPAA says a year between, to implementation, from the time that it's decided that ICD-10-CM will be placed. As neurologists, we did this with CPT-4. ICD-10-CM is not a different concept. It's the same concept, same general structure. A little bit of a change from just a numeric system to an alphanumeric system but we are smart enough to make that change. It's so much more familiar to us because the diagnoses are much more similar in language to the we use that it won't be difficult for us to adapt.
And also to reflect what has already been said this morning about rules, we would greatly appreciate if there was some way to have the rules binding to all payers who use ICD-10-CM. It is so difficult for us to know which insurance Will let us use which codes under which circumstances, which Medicare carrier will let us use which codes under which circumstances even though it should be uniform and it would be best if everyone followed the guidelines and then if coding clinic made changes, everyone was bound to those changes.
In summary then, the American Academy of Neurology supports proceeding with the implementation of ICD-10-CM as a national classification system binding to all third party payers and with rules binding also to all payers. We request and understand that is built in an adequate transition period during which we can train our members and as a means to challenge arbitrary computer edits which will inevitably occur early in the implementation of this system.
DR. ZUBELDIA: Thank you. Dr. Mirin?
Agenda Item: Steve Mirin, American Psychiatric Association
DR. MIRIN: Thank you, Mr. Chairman and members of the committee. I'm Steve Mirin, Medical Director of the APA, American Psychiatric Association, representing 38,000 psychiatrists in the United States and overseas. Accompanying me today is Dr. Darrell Regier, formerly of the National Institute of Mental Health and one of the world's experts on the classification of mental disorders.
In April, I had the opportunity to address the subcommittee to advocate for the designation of DSM-4 as the official code set for mental disorders under HIPAA and I appreciate the opportunity to continue the discussion today.
As you know, DSM-4 is used by psychiatrists, physician and all mental health professionals to describe patient diagnoses with sufficient specificity so that clear communication among these professionals is possible for both clinical management and recordkeeping. DSM-4 is also the official mental disorder code set for Medicaid in all states, the federal employee's health benefit program, the Food and Drug Administration, the National Institutes of Health and most private insurance companies.
And the failure to designate DSM-4 code descriptors under HIPAA will mean all of these practitioners, agencies and institutions will have to abandon their use of the diagnostic criteria in favor of ICD-9-CM which, as I mentioned a month ago or more, essentially reflects medical science as it was in the late 1960's.
Having DSM-4 designated as the descriptor code set and delegating responsibility to APA, its developer for convening scientific experts needed to periodically update these descriptors would address this problem. ICD-9-CM was developed in this country and published in 1977 to add additional specificity to other non-psychiatric sections and the overall structure and the lack of operationalized definitions, particularly in the mental disorders sections, rendered it inadequate for use by mental health professionals and the glossary of definitions in ICD-9-CM was actually developed for ICD-8 which was parallel to DSM-2 which was the operational code set when I was a resident in psychiatry in 1967 through 1971.
In 1980, concerns about specificity and subtyping provided the 9th edition of ICD. Provided in the 9th edition were not reflective of contemporary scientific thought and so APA proceeded with the third major revision of the DSM which departed from the ICD-9 format.
Since that time, the DSM has become the de facto system of classification of mental disorders for all mental health professionals practicing in the US as well as for health care facilities and federal and state government agencies. It's overwhelming acceptance is due to its reliable and clear definitions of mental disorders, its reliance on descriptions of observable signs and symptoms and the use of a documented evidence base derived from a comprehensive review of empirical literature as well as input from a broad array of professionals and health care organizations.
There have been subsequent revisions and today DSM-4 and its operational criteria and definitions are coupled with ICD-9-CM code numbers and that enables the insurance industry and government agencies and institutions to apply these criteria in determining diagnoses, the need for treatment, the appropriateness of the treatments provided.
The seamless merging of DSM and ICD-9-CM worked so well that many mental health professionals are only dimly aware of the existence of ICD-9-CM and assume that the diagnostic codes themselves were developed by APA which we know is not the case.
But the continued use of ICD-9-CM does pose some serious problems. I'll just mention a couple of them.
We have the issue of archaic terminology in ICD-9-CM which at times carries negative connotations for patients. For example in ICD-9-CM, the word psychosis is used in its 1960's sense as an indicator of severity as opposed to a designation that indicates the presence of some rather specific signs and symptoms and this can have very adverse effects for patients with respect to their future insurability.
The limitations in the numeric coding system have already been touched upon. Interestingly, in psychiatry a number of DSM disorders share the same ICD-9-CM code. In fact, for code 292.89, there are 29 different substance use disorders designated under that one code so that poses a profound problem, particularly with respect to research on the incidence and prevalence of these disorders.
There are in ICD-9-CM a number of categories that are not recognized as mental disorders in the United States, including, and these are primarily because they appeared in ICD-9 which was constructed in the early 1970's, but they have no empirical foundation. A couple of examples include something called misery and unhappiness disorder and academic under achievement disorder. I think those are rather common but probably do not meet the medical necessity criteria we operate under today.
Now, the continued use of archaic subtypes and wording forces medical coders to make diagnostic distinctions that have really no basis in current science and lack any precise definition. And leaving the interpretation up to the coders' discretion results in potentially poor reliability. This is more than just a hypothetical concern since mental health professionals and institution use DSM-4 for diagnosis, recordkeeping, and billing.
Using the codes provided in DSM-4 as surrogates for the ICD-9-CM categories, but the coders working in acute care hospitals continue to use ICD-9CM in order to determine the most appropriate diagnostic code to add to the medical record and to billing statements and this is really inconsistent with the HIPAA mandate of adopting single uniform coding systems for appropriate specialties and even the adoption of ICD-10-CM will not address this problem inherent in both ICD-9 and ICD-10-CM, namely the lack of diagnostic criteria. Only DSM-4 fulfills this need.
Now, tension headache notwithstanding, I want to express our strongest possible support for replacing ICD-9-CM with ICD-10-CM. It provides greater freedom for adding new categories and subtypes. It reflects the most current usage of medical terminology. It was developed by WHO with over a decade of research and support from the NIMH and with the close cooperation of APA as we developed the parallel but more extensive DSM-4.
During the time that this was taking place, it was expected that ICD-10-CM would be adopted for use in this country for morbidity as well as mortality reporting but as you know, ICD-9-CM has remained the standard for Medicare due to concerns voiced at the time by HCFA now CMS over the cost to re-program computers and accommodate the conversion to the alphanumeric system of ICD-10.
In recognition of the need to use a classification system that recognizes modern day terminology, has sufficient room for expansion to include new codes, and reflect ongoing medical discovery, most of the developed world has moved on, and the United States has continued to lag behind many other countries and hold back some countries that look to us for leadership in health care.
So, in summary, I think we have five recommendations. For over 50 years, APA has provided national an international leadership in the development of scientifically based diagnostic criteria for mental and addictive disorders for both DSM and the ICD. The US National Committee on Vital and Health Statistics has provided similar leadership for other sections of the ICD. The ICD-9-CM system is 25 years old, archaic for most disorders and even more dated for mental disorders with terminology and glossary definitions created in the mid-60's for ICD-8 and DSM-2.
For mental disorders, the introduction of ICD-10 will allow the replacement of this system with one that reflects DSM-4 compatible names for mental disorders. It will allow the adoption of an alphanumeric code set. It will provide space for identifying for scientifically reliable and valid disorder definitions, facilitate documentation of prevalence and treatment rates for mental disorders, and the public health burden caused by emerging new disorders including substance use disorders.
ICD-10-CM, however, has no operationalized definitions or diagnostic criteria and only an introductory statement referring the reader to the DSM-4 for definitions of psychiatric terms and conditions. And it's questionable what effect, if any, this informal reference will have with respect to compliance with HIPAA regulations and since a major goal of HIPAA is administrative simplification of diagnostic reporting, continued use of two separate code sets for men toll disorders seems inconsistent with that goal.
Furthermore, HIPAA requires code sets to be maintained pursuant to a formal regulatory review and updating process which the APA can do and continued reliance on the current informal dependance of both ICD-9-CM and ICD-10-CM on DSM-4 will no longer be legally permissible.
So to avoid needless destruction of the mental health service delivery in this country, we request that the NCVHS approve adoption of DSM-4 criteria as the official descriptors for ICD-9-CM, designation of the APA as the DSMO for these criteria, accountable to the Secretary in the same manner that AMA and ADA are at the present time; approval and implementation of ICD-9-CM for HIPAA reported requirements at the earliest possible date.
Thank you for the opportunity to present these recommendation for improving the administration of mental health services in the United States.
DR. ZUBELDIA: Thank you. On your last recommendation, I assume you meant ICD-10 for HIPAA.
DR. POWERS: Yes, ICD-10-CM.
DR. ZUBELDIA: Dr. Strunk?
Agenda Item: Al Strunk, American Association of Obstetricians and Gynecologists
DR. STRUNK: Good morning. My name is Al Strunk. I'm an obstetrician gynecologist and Vice President for Fellowship Activities with the American College of Obstetricians and Gynecologists, commonly known as ACOG.
I think I might advance your agenda by first giving what I hope will be a succinct statement and then if anyone is interested in talking about 50 or 60 specific examples in which ICD-10 better serves us that ICD-9, we can do that.
I'm pleased to be here today to represent ACOG, an organization of 44,000 physicians specializing in health care for women. The college is grateful for the opportunity to offer comments on the option of ICD-10-CM as the standard code set for reporting diagnoses.
Accurate diagnosis coding is, of course, critical to improving the quality of health care, in forming decisions about public health policy and facilitating smooth processing of health insurance claims.
ACOG believes that ICD-10-CM represents a significant improvement over ICD-9-CM for these purposes. We support the adoption of ICD-10-CM as the standard code set.
As we all know, the World Health Organization introduced ICD-10 in 1993. Many other countries use ICD-10-based systems for reporting both mortality and morbidity statistics. Our continued reliance on an ICD-9-based system for morbidity statistics inhibits, we believe, the free-flow of data that's essential to international medical research and disease surveillance. United States recently began using ICD-10 for mortality statistics. Establishing congruence between morbidity and mortality reporting would enhance tracking of outcomes and identification of effective interventions.
It's been almost three decades since ICD-9-CM was introduced. During that time, our medical knowledge has expanded greatly as has the range of treatments offered.
Obstetric care provides several examples of this rapid pace of change. Developments in sonography have revolutionized the examination of the fetus. The ability to make genetic diagnoses has changed dramatically in the last 30 years and will change exponentially over the next 10 years. Our understanding and clinical use of maternal and fetal physiology and the management of labor has improved significantly.
Knowledge and classification of the important maternal conditions such as diabetes mellitus and hypertensive diseases of pregnancy has also changed.
ICD-9-CM is woefully inadequate to capture these changes. NCHS staff and the coordination and maintenance committee have been very responsive to ACOG recommendations for changing the obstetric chapter of ICD-9-CM to meet current needs. However, the ability to change and expand ICD-9-CM has been stretched to its limit and any further expansion will be difficult. ICD-10-CM offers greatly enhanced capabilities for identifying and tracking important obstetric conditions and interventions. These data are needed for high priority public health efforts to reduce adverse pregnancy outcomes.
I would like to highlight four specific aspects of ICD-10-CM that will facilitate reporting the issues of growing importance in obstetric care.
First, the addition of a fifth digit to report the trimester of pregnancy should prove valuable in efforts to monitor the provision of prenatal care and the occurrence of complications. ICD-9-CM does not allow for identification of trimesters.
ICD-10-CM also responds to the growing prevalence of multiple pregnancy by permitting identification of individual fetuses in a multiple gestation. This will make it possible for the first time to link a specific condition to a specific fetus.
In addition, ICD-10-CM includes expanded codes for reporting abnormal findings on antenatal screening. The range of screening tests offered or recommended to pregnant women continues to expand. These screening tests generate additional episodes of care which require appropriate diagnosis codes. None are now available.
Finally, ICD-10-CM will include a category for care provided to the fetus due to abnormalities.
There are several challenges and potential disadvantages to adopting ICD-10-CM. Physicians, hospitals, coders, payers, will need to learn a new system. Also the six-digit ICD-10-CM codes will dictate that computer systems and software be reconfigured. The cost associated with training and computer changes will not be trivial. In an environment of declining reimbursement and increasing regulatory demands, physicians, hospitals, and health care providers should not be expected to bear the entire burden of these costs alone. Planning for implementation of ICD-10-CM must address cost issues.
On balance, ACOG believes that the benefits offered by ICD-10-CM are substantial enough to justify the costs and disruption its adoption will inevitably entail. I appreciate having the opportunity the present ACOG's views today. I especially want to note the college's appreciation for the manner in which NCHS collaborates with the physician community. We are eager to work with NCHS in developing plans for educating physicians about ICD-10-CM and addressing barriers to its implementation and I would be happy to respond to your questions if any.
DR. ZUBELDIA: Thank you. Ms. Warner?
DR. WARNER: Thank you for the opportunity to speak today. I'm Dr. Margaret Warner from the Office of Analysis, Epidemiology and Health Promotion at the National Center for Health Statistics and CDC.
I prepared written comments but I also wanted just to reiterate what the rest of the panelists have spoken about and that is that the classification framework really needs to keep up with the knowledge area in a subject. I think injury prevention and control, which is my area -- I'm an injury epidemiologist -- has really advanced with the science in the last 20 years. I'll just use motor vehicle related injury and death as an example of that, but you can also find parallels in burn injury research and other areas of injury.
You think about the last 20 years where we've come with motor vehicle related injury research, we now have air bag and car seats and rear brake lights, the third rear brake lights. Car seats, in particular, was informed by data collected using ICD-9 for fatalities and that is when they looked at the types of fatalities by age and they looked at infants, they noticed that they were way over-represented and they realized that was because infants were flying across the car and through the windshield. And now we have car seats and now with air bags, we are noticing a decrease in the number of fatalities and an increase in the number of lower limb fractures, and hence our morbidity data needs to keep up with the -- the classification for morbidity needs to keep up with that trend and we need to have more classifications for fractures, particularly in the lower limbs.
This is also true in our understanding of head injuries and in spinal cord injuries. So injury prevention as with obstetrics and mental health disorders, we need to keep up our classification system and these are my written comments.
Chapter 19 of ICD-10-CM, entitled injury, poisoning and certain other consequences of external causes and chapter 20 entitled external causes of morbidity and mortality data is a final result of extensive consultation and review with experts in the field of injury prevention and control. Emergency room physicians, epidemiologists, statisticians and others from both the public and private sector were among the many professionals consulted on the design of the classification system.
I'm going to comment on improvements to the surveillance of health data and injury and poisoning provided with ICD-10-CM.
Overall the adoption of ICD-10-CM would be an improvement over continuing with ICD-9-CM in order to maintain compatibility with US mortality data which is using ICD-10 to code fatalities and because the extension codes greater detail and more future expansions will allow for far better data describing both the external causes and the nature of injury using the classification system.
The classification system also would provide details that makes it more appropriate for a wider range of users, for instance, with workers' compensation potentially. It is important to have comparable codes for data sets within the United States to facilitate data analysis. For injury prevention and control, scientific communities the incompatibility between the structures of ICD-9-CM and ICD-10, makes analysis of data more difficult, more time consuming and the results less precise.
As of 1999, the US has been coding mortality data using ICD-10. ICD-10 and ICD-10-CM are completely compatible, but on many levels, ICD-9-CM and ICD-10 are incompatible. For instance, ICD-9-CM is organized by the type of injury and ICD-10 is organized by the body region. The ICD-9-CM uses external cause codes E-800 to E-999 and the nature of injury codes 800 to 999. And ICD-10 uses external cause codes B, W, X and Y and the nature of injury codes S, M, T.
And in addition, and probably most important, there is not a one to one relationship between the codes in ICD-9-CM and the codes in ICD-10. It is important for the United States to compare our injury and poisoning morbidity data between countries in order to better understand the causes and circumstances of injuries within our own country.
As has already been mentioned, many other country, including Australia and in the European community, are already using ICD-10 or a version of it which will make the US data not comparable and outdated. It will become very cumbersome to compare between the countries which is already cumbersome using the same classification system and different applications of the rules.
Many injury codes in ICD-10-CM have far greater specificity than the codes in ICD-9-CM. For example, capturing data in injury and poisonings and toxic effects is included in the final, will be greatly enhanced using ICD-10-CM. The intent of the poisoning or the toxic effects is included as a final character in the diagnosis codes in ICD-10-CM. The external cause codes in ICD-9-CM are not specific to the agent involved. ICD-10-CM allows room for expansion of these codes at the fifth character that is not possible with the ICD-9-CM because, as we've learned it is running out of room.
It's essential to include room for expansioning on the poisoning code because the number of biological and medicinal substances coming into use is rapidly increasing as we all know.
Classification systems need to be modified in order to accommodate emerging public health needs. ICD-9-CM has simply run out of room and thus many specific codes needed for injury control are not available. ICD-10-CM is designed to accommodate expansion, thus eliminating the situation that ICD-9-CM currently faces.
Injury codes in ICD-10-CM enable the user to differentiate between an initial and subsequent visit for the same injury as well as the sequelae of injury and Donna Pickett pointed this out in her presentation.
This will permit more detailed analysis of people injured rather than the current situation which allows only analysis of health care utilization patterns. For example, the ICD-9-CM coding guidelines require that a fracture be coded only once at its initial episode. Thereafter, any further care the fracture code is coded as after care and information on the type of fracture is last during the course of healing.
Neither the guidelines nor the codes in ICD-9-CM can be changed to correct this and as there's no additional digits available in the classification system.
With the extensions that apply to all injury codes, the episode of care is identified and the fracture code is used continuously throughout the treatment. Data on fractures as well as for other injuries will be far superior with ICD-10-CM because of the ability to differentiate between initial and subsequent visits for injury.
In ICD-10-CM, the level of detail available to code the injuries is greatly improved compared to ICD-9-CM. This is true for the nature of injury codes such as fracture and traumatic spinal cord injury and for external causes such as the transportation related injuries. Again, I believe that is keeping up with our current knowledge engine in injury prevention and control.
Another example of the level of detail provided is the addition of laterality to all ICD-10-CM codes. This enables the identification of the exact site of an injury. Laterality is not included in ICD-9-CM and there are insufficient digits at the code level to allow for this expansion.
For fractures, an example of the important detail provided is the distinction between degrees of open fractures that are provided in ICD-10-CM.
In summary, ICD-10-CM will provide many improvements to the collection of injury morbidity data and the surveillance of injury and poisoning in the United States and will allow for comparability with mortality data in this country and morbidity data in others. Thank you.
DR. ZUBELDIA: Thank you very much to the entire panel for their testimony. One more? Oh, yes.
Agenda Item: John Muldoon, National Association of Children's Hospitals
DR. MULDOON: Thank you. John Muldoon from the National Association of Children's Hospitals and related institutions, an organization that represents approximately 160 member hospitals. We had testified five years ago when, in essence, today we present the same recommendation that the United States move forward with ICD-10-CM.
I'll say a few words about NACHRI's activities and interests in this area and then some summary remarks of the importance but I will attempt to be very summary level and not repetitive of the themes and the specifics that you have already heard.
NACHRI works with ICD-9-CM-based information in a number of ways with children's hospitals around the country. One way is in development of ICD-9-based classification systems, both in-patient and population-based systems. We have for many years run comparative reporting programs for children's hospitals and we do a variety of policy analyses, and in all of these activities we find limitation with the specificity of the information until the ICD-9 codes and thus our interest to see the word and the United States move on to ICD-10-CM.
Many of the limitations are generic, they are not pediatric-specific. They have been talked about throughout this session. The complication codes are extremely frustrating to lose information on what kind of complication you are talking about. There are some specific complication codes, but there are a lot of generic codes. The injuries and the aftercare where you lose sight of what is the injury or the status of it being a new injury or previous injury; the mental health codes being very, very out dated. They are amongst the most prevalent of conditions which children interact with the health care system -- the language, the terminology, the definitions as described earlier are very out of date. I haven't ever come across someone in the children's hospital world and mental health field who worked with ICD-9-CM diagnosis code. The neurologic conditions are also important in childhood.
Cancer is a lot less common, of course in children but also without the histological information, it doesn't give you very much. The title often tells you nothing about the particular cancer. Neuroblastoma is a good example, whether it's in the adrenal gland or thorax, it is somewhere you don't know what kind of cancer you are talking about with ICD-9-CM.
Of course, areas of specific interest are that have pediatric in particular are the congenital anomaly codes and the perinatal codes. We've always felt they are particularly underdeveloped with respect to their specificity. That might be because some of their conditions are relatively low volume, at least relative to the prevalence of certain adult health conditions, but for whatever reason, quite a few of them lack a great deal in specificity.
I'll just provide a few examples. When you look at the congenital anomaly codes, in a number of instances, there isn't the distinction made between atresia, stenosis, insufficiency, hypoplasia and other malformations, particularly true with the heart conditions, some of the renal conditions. The distinction is not made also whether it's bilateral or unilateral. That makes a pretty big difference.
Often a number of very different conditions are in the same code. This is most common with the not elsewhere classified codes. One of my personal favorites is 759-89, the other congenital anomalies affecting multiple systems and there is approximately 150 different terms that are indexed there, and most of them are quite infrequent, very rare, so I'm not suggesting that they all need, that they are on separate codes, but when you look at ICD-10-CM, they break them out into about a half dozen syndrome types that are predominantly affected in certain body systems so that kind of approach we think strikes a reasonable balance.
Cranial-facial anomalies, they all come in the same package, 756-0 and having to break out there, and the list goes on. In the prenatal diagnoses, infections are very important in the newborn period and there's not a lot of specificity amongst the pneumonia diagnoses or sepsis. The fact is in any breakout, we've just gotten newborn sepsis broken out from perinatal infection, NEC, one of our other favorite codes, but not the specific organism.
You can go on and on with certain of the hemorrhage codes and pulmonary codes and noxious substances and whether there's a specific effect, the syndrome on the newborn and then it winds up with one of my other favorites, 7709.8, other conditions not elsewhere classified and there you have muscle tone disorders thrown together with cardiac failure and renal failures so that's really a tremendous conclusion so we very much support the need and the movement to go from ICD-9-CM to ICD-10-CM and I thank you for the opportunity to present our thoughts.
DR. ZUBELDIA: Thank you. Sorry I hadn't seen you on the sidelines there.
This is been great testimony from actual users of the codes. I have a question for Dr. Mirin on the DSM-4. From your recommendations, it looks like whether ICD-9 or ICD-10, DSM-4 is what you want under HIPAA, is that correct?
DR. Mirin: As far as the descriptors are concerned, yes. The DSM-4 at present when you look at ICD-10-CM in the mental disorders section it says "see DSM-4" so the codes essentially are the ICD codes but the descriptors are the descriptors from DSM-4.
DR. ZUBELDIA: So if the ICD-10 including guidelines was left under HIPAA, that would satisfy your need to see the DSM-4 for descriptors.
DR. Mirin: That is correct.
DR. COHN: I have a bunch of questions but maybe I'll just follow-up on your DSM question. The other scenario, as you know, that we have been wrestling with, as I'm reviewing your testimony today, it really sounds like what you really need, I think, in the near future is some sort of a letter indicating that HHS acknowledges the crosswalk between ICD-9 and DSM-4 and that it's a valid crosswalk, at least as I'm reading your testimony. Is that correct?
DR. Mirin: Well, for as long as ICD-9-CM is in use, and beyond the implementation date for HIPAA, yes, it would be enormously beneficial to the field to have that crosswalk acknowledged.
DR. COHN: And then for ICD-10, the question becomes whether an acknowledgment of the crosswalk or the actual reference to the descriptors.
DR. Mirin: Yes, that's okay.
DR. COHN: Now, let me just ask a general question. There's actually a couple of things I wanted us to dwell on. I'm hearing from a number of you, and actually in some of the written testimony from people that weren't here, now, Steve, you are talking a lot about definitions and the lack of definitions in ICD-9 as I understand, has some definitions but doesn't have others -- I think I'm correct on that -- and I guess we expect that to be something that happens in ICD-10 also. I guess I almost need Marjorie to correct me if I'm wrong on this one, but how important do you think having definitions in a diagnostic coding system are so that people know what is what and what isn't what?
DR. Mirin: I think that in a field like ours in which there is such enormous difference between, say, earlier iterations and the DSM and the current DSM and, I predict, between the fifth edition of the DSM and DSM-4. In an emerging field, being able to be as specific as possible with respect to diagnostic criteria I think is crucial particularly from a public health standpoint to understand what the true prevalence is of these disorders, not only in the United States but across all countries. So our goal in developing the DSM and its subsequent iterations is to end up with a single worldwide classification of mental illness so that researchers in Bangladesh and in Brooklyn will be looking at the same patients.
DR. COHN: You are saying that it will cause common definitions as well as common titles.
DR. Mirin: That is right.
DR. COHN: Okay, now Mel Ingobar(?) From CMS also commented about issues around, I think it was chronic renal failure, the fact that there weren't clear definitions in the ICD about that. What is the status of definitions of terms in ICD-9 and what's planned for ICD-10? Marjorie, can you comment on that? Is it done?
MS. GREENBERG: Donna, I think, in response to, Donna Pickett, in response to the last hearing has been working on a document in that regard. Do you want at the --
DR. COHN: Of definitions of terms in ICD?
MS. GREENBERG: No, about --
DR. COHN: Oh, I was just referring to definitions of terms in ICD. Are they there, are they not? Are they some there and some aren't, not at all?
MS. GREENBERG: The purpose and the use of the disease classification is to assign a code based on the narrative in the medical record. It is not to diagnose patients. Just as in the case of the APA, they have actually incorporated it into a classification. In other cases, in other specialties, in order to assign that narrative, there's a whole literature which leads you to that. Classification itself is not considered a diagnostic tool.
DR. COHN: Sure, but that doesn't include definition of what that is. Actually I'm just referencing a couple of different presentations.
MS. GREENBERG: The greater -- on the other hand -- I mean, the greater the detail in the classification such as the considerably enhanced detail intensium over 9-CM I think makes it easier to select obviously the correct code and it differentiates to a greater extent but that's not the same as actually providing diagnostic criteria, but I would really like to defer to Donna or to Amy Bloom here, in fact, I just would like to point out that my entire staff has developed 10-CM and I have heard in Australia they had 28 people or something but we work harder and smarter perhaps, but Amy Bloom, David Berglund, Beth Fisher and Donna Pickett, under Donna's leadership were really the crew so that when people have mentioned they appreciated collaborations that are within NCHS, it's really this team that they have been collaborating with so I just wanted to acknowledge them.
Amy Bloom has probably been with us longer than any of us or of anyone, so Amy, would you like to comment on this particular issue that Dr. Cohn is raising?
MS. BLOOM: As a statistical classification we specifically chose not include any definitions and official coding guidelines require that if a physician states a specific diagnosis the coder is required to assign a code based on the physician's documentation. We do try to keep coders from having to confirm a diagnosis. They are supposed to review a medical record thoroughly but they are not really supposed to determine whether or not a patient's lab results or test results actually confirm a diagnosis and, in fact, at the APA's request, the glossary of mental disorders has been removed from the ICD-10-CM.
The glossary of mental disorders that's currently in 9-CM is not maintained. It has not been updated in the 23 years. It's simply an appendix to the ICD that we've kept on the official version of the classification but we haven't been maintaining it at all just as we haven't maintained the list of industrial accidents and such. There's just a number of appendices in the ICD that have remained on as our commitment to the ICD but we haven't maintained them.
With 10-CM, we did remove the glossary of mental disorders that was in the ICD-10 at the request of the APA, and then that was the reference to refer to the DSM was added so that people would know the definitions for mental disorders should be found in the DSM.
We do have some definitions. We do define an acute MI as being a four week time period. The coder is supposed to continue to code an MI for four weeks after it's been diagnosed so if there is a readmission or such you should continue to code it. We do have specific definitions for trimester as was requested by ACOG so that when a physician goes to select a particular trimester for the pregnant patient, they know exactly what the trimester definition is. That's in the beginning of the chapter. We do have definitions for premature ruptured membrane, preterm, postterm, that sort of thing, but otherwise, I'm trying to think. We don't really have -- we have instructions.
What we have is instructions on how to properly assign a diagnosis code as far as sequencing of paired codes, combination codes and such as that but not what does pneumonia mean, what does a heart attack mean. We don't have any definitions for renal failure and that sort of thing and we did not have plans to do that.
DR. COHN: Okay, thank you for the clarification. I have one other question, if that's okay. This item may require Marjorie to call in some of her troops also, but Laura, I was listening to you and I was listening to Albert and others about how out of date ICD-9 has gotten over the last 30 years and some of it has to do with the fact that there isn't room. And, Laura, you were talking about that.
But I heard as much maybe that our view of the world had changed and that, for example, headaches are no longer a psychiatric illness. We can certainly see this in diabetes where it isn't so much space but it's just that we think of diabetes completely differently and obviously, one can't help but wonder as you move into the world of genomics and all of this other stuff coming up in the next couple of years. I know neurology is going to be very impacted by that, as well as certainly OB-GYN, that we may suddenly be thinking about things completely differently again. Am I off on this one? That is, I think, what I was hearing.
DR. POWERS: Absolutely correct. Using as an example, there's a wonderful web site from Washington University at St. Louis for neuromuscular disease and when the web site first went up it used more traditional classifications of neuromuscular disease. It has completely changed over the last several years with more genetic classifications and then direct links to OMIM(?) For many of those because we're learning so much more about the genetic classifications of neurologic disease.
DR. COHN: Let me ask a question then, maybe once again to Marjorie if it's okay. Now, we've obviously talked and I'm obviously convinced that the new coding system, ICD-10, has more space for additions, but there's this whole issue of view of the world changing such as what happened with diabetes where it isn't a question of just putting some extra numbers in but you really have to completely re-order things. What plan, let me ask, whatever changed ICD-10 is going to be very expensive, and I don't think anybody is going to be very happy in eight years hearing about another moving to ICD-11 or-12 or whatever. I mean, this is --
MS. GREENBERG: I think one very important distinction between ICD-9 and ICD-10, and this is, I'm not talking about the clinical modifications, but is that ICD-9 was never changed. ICD-9-CM we did develop, begin in 1985 to update it on an annual basis but it was always, it had to be based on the, you know, the core ICD-9 at the three digit level but there was no process for changing ICD-9 so ICD-9-CM was obviously always tied to ICD-9 which had been first published in 1979.
ICD-10, when ICD-10 was implemented and approved by the World Health Assembly, it was agreed that there should be an updating process associated with ICD-10 so that it could and, really using as a model the updating process that had been used in United States for ICD-9-CM so that it would be possible to make changes, corrections, but changes to try to keep up with medical science. For example there never actually were codes in ICD-9 for HIV-AIDS. We developed some here, in the US, but because there was no mechanism for changing ICD-9.
So there is a process now that is carried out by the collaborating centers as Dr. Sondik mentioned, the ICD is maintained by the WHO in conjunction with 10 collaborating centers. The one for North America is housed at NCHS.
And it's possible that whole sections of 10, in fact, could be updated, perhaps even a whole chapter, I mean, the mental health chapter. That's a possibility, without going through an entire, to an entirely new classification and then, of course, 10-CM would fit into that.
I mean, we have, if you were going to completely change the concept as we have sort of with diabetes, you wouldn't just want to do it in 10-CM. I mean you would want to see it done probably in 10 as well and there is not process for doing that. Now, if we are still on 9-CM and they make changes to 10, it really won't help us any in our clinical modification, but the process is there, it is currently being carried out. It has not reached its full fruition, but we have committees set up, international committees set up, both on mortality data and on mortality and morbidity data as well for modifying the ICD-10 and a number of modifications have been made.
I think the potential to make the kind of changes you are mentioning within the context of ICD-10 exists. Now, I can't say that, you know, for all of our lifetime. I would certainly say our working lifetime I think it will work. For all of our lifetime if we, according to NCHS life expectancy tables, this could hopefully be quite long. There wouldn't be a need for, say, a completely new ICD-11. But I think we are not tied to a classification. ICD, as we were with ICD-9 that could not be updated, modified, and even rather dramatically.
For example, just one more example, the oncology area. Really even in ICD-10, requires some modification, the lymphomas, etc., That are reflected in ICD-03 and we are currently looking at how we can migrate that aspect of ICD-10 to the latest, to ICD-03 and then we would piggyback on that, on 10-CM.
So I think that the responsibility to keep up is much greater.
DR. KOLODNER: I would like to thank the presenters. There were some excellent comments and first of all, Marjorie, the description that you have of ICD-10, sometimes in some of the newer models or other things is called the evergreening process so instead of having something where you had to revise it and come up with another revision, the evergreening process was built in or at least trying to. We'll see how well it works. Hopefully we'll extend it at least 23 years.
One of the things that I heard as kind of a common theme across all the presenters was first of all the fact that as end users, and the people who have to do the coding in the ambulatory care settings, usually without a coder to then process it, where the vast majority of the codes are assigned, that there's really a distortion trying to fit the current clinical understanding into the old codes and that although most of us have learned those in our training and kind of can stuff it in some place that changing to a new system that's much more consistent with the clinical care and understanding and concepts will actually not be as difficult for the end users because it is actually, it aligns with how we're practicing and I wanted to see if I understood that correctly since the real, in terms of the number of codes assigned, the real burden is going to be on the users in the ambulatory care settings, often in the clinics and offices.
DR. Mirin: If I may lead off, that is the way life is. The clinicians essentially assign the codes in ambulatory settings and also in private psychiatric hospitals, in acute care general hospitals, it's a somewhat different situation, but there will be utter confusion over all the folks who deliver mental health care and are accustomed to essentially entering the diagnostic code from DSM-4 into their notes. Their notes often conform to what they understand to be the diagnostic code and vice versa and so when even in an acute care setting, when somebody goes through and reads the chart, they are going to come out hopefully with the conclusion about a diagnosis, a DSM-4 diagnosis that fits the clinical record.
So then in the absence of having that available to them, once HIPAA is implemented, they are going to be searching through the ICD-9-CM codes to try to find something that essentially fits their understanding of the patient, and in some instances that's going to be extraordinarily difficult because it's going to, particularly for our young people, where we are trying very hard to adopt a totally evidence-based approach to diagnosis and to treatment, to throe them back into a world that's 25 to 30 years old in which there was a whole different theoretical understanding of many of these illnesses, completely devoid of the genetics determinants of illness and the biochemistry and the psychopharmacology will be a disservice ultimately to the patients.
DR. POWERS: I agree. The ICD-10-CM is more our language and its true that coders do choose the codes in many offices and in many hospitals and when the physician is able to use their own language instead of an antiquated language and there's very, it's much less stressful for us.
DR. STRUNK: We've been fortunate to have very good collaboration and communication between NCHS staff and our committee on coding and nomenclature and our own staff and while nothing is absolutely seamless, it will, for example, make it a lot easier to talk about, for example infections of the genital-urinary tract in pregnancy as not being just one category but involving either the kidney or the bladder or the urethra or other discreet areas.
Likewise, when we talk about abnormal findings on an ante-natal screen of a mother which currently does not include any kind of a breakdown in ICD-10-CM will have hematologic findings, biochemical findings, ultrasonographic findings, chromosomal genetic findings, and so on, so we do not really anticipate a great deal of difficulty in making this transition and to the contrary, we really see a tremendous benefit in the added specificity that this will allow us.
DR. WARNER: For injury prevention and control, I think that the classification exists but not to the great degree because people just didn't think it was important to classify the types and levels of fractures like they do now just because people understand more that you can prevent injuries and that you can heal them and our science has grown. In addition for brain injuries, traumatic brain injuries, where you can't see the brain, you know, there's no crush, now we have sonograms and different methods to look at and diagnosis those types of brain injuries.
DR. MULDOON: In the classification system development work that we do, I happen to spend quite a bit of the time translating what the ICD-9 codes really are to the physician members of the team. I anticipate spending a lot less time with ICD-10-CM explaining what the code really contains so ditto to everything that's been said.
DR. ZUBELDIA: Dr. Powers, you mentioned the work rules for the payers that they ought to follow some rules, maybe even the same rules. What do you think the impact will be by changing from ICD-9 to ICD-10? Could it be a catalyst for all the payers to follow the same rules or will it be an opportunity to invent their own rules?
DR. POWERS: I'm not sure who legislates, how the third party payers behave. It could be either one. I think it's an opportunity for us to go to a single rule system, and I would hope that that would happen. I'm not sure if there's any way to force the insurance companies to do that, but that would be so helpful to the entire medical community if everyone followed the same rules.
MS. HUMPHREYS: I just want to point out that the rule that's been published does, for ICD-9-CM diagnosis, give the coding rules and the guidelines and requires everyone to use them. I don't know what's going to happen, but as I understand it, the private payers with the implementation of the HIPAA rule that we've already passed, I mean, I'm in favor personally of moving to 10-CM, but I think this thing actually happens in advance of 10-CM. It happens as soon as 9-CM is the HIPAA, once the transaction rule goes into effect.
DR. COHN: Well, Laura, this is one I'm a little confused about and maybe, what I'm sensing, your issue is not the insurers don't accept the claim, they just don't pay on it.
DR. POWERS: Or they will tell us that you cant use this code. Period.
DR. COHN: Okay, well that solves that.
DR. POWERS: Yes, you cannot use this code.
MS. HUMPHREYS: I believe that as I understand it, and I think I understand this part of it, with the implementation of the transaction rule for ICD-9-CM, for which there was a universally accepted or shall we say an official national coding set, that is part of the standard and everyone, it says very explicitly, is required, they are not required to change their rules for what they would pay for but they are required to accept all valid codes and not to reject your claim because you used some code which they didn't want you to use if it is a standard code.
DR. ZUBELDIA: There is a fine distinction between rejecting a claim and accepting a claim and not paying for it, and HIPAA actually requires them to accept the claim. It doesn't require them to pay for it.
MS. HUMPHREYS: Right, well, no, that obviously is the billing arrangement that, you know, that's the business rule that they have in place.
DR. ZUBELDIA: And I think that when the payers say you cannot use this code, they don't mean you can't use it. Sure you can, but you won't get paid for it.
MR. HUMPHREYS: No, I think that in fact, based on the testimony we heard, we heard that under the current thing it is the latter. I mean, you send in a claim and maybe there are four or five codes on it and this particular payer is supposed to pay for one or two of the procedures and, in fact, we heard testimony that there were cases that the whole claim would come back for you to delete the code which was a standard code but this particular payer didn't even want to see it in their system.
So actually I think that the requirement that everyone recognize and not reject a claim based upon the presence of a standard code will a some cases smooth out the payment process even if the payer doesn't intend to pay for that particular code.
DR. BLAIR: As a clarification, are you talking about a specific code or a code set that they are rejecting?
MS. HUMPHREYS: No, they were usually rejecting specific codes and, of course, there was the issue of this has to do with the, also the implementation dates because in many cases, a new edition or the new updates would come out and a particular payer would lag behind other payers in terms of accepting valid additions to, say, ICD-9-CM, and therefore people who would now be coding diseases in their local environments according to the currently valid set of codes would find that for some month period, many claims were being rejected because that particular payer hadn't actually updated the system. I'm not the authority on this. I mean, obviously Donna and others are, but I recall this and that this really was a major issue so I just want to comment that that particular issue does not need to, theoretically does not wait on the move to ICD-10-CM.
However, for all the reasons you have described, movement to 10-CM is very desirable if we want data that reflects what people actually know about diseases.
DR. ZUBELDIA: Is that the issue you were referring to?
DR. POWERS: Yes, that is correct, and it also happened to us when we were coding for certain procedures or laboratory tests. Some insurance companies were requiring us to use rule-out codes which we don't use with, officially.
DR. ZUBELDIA: Have you looked at DRG's? And I know that maybe I should be addressing this to the hospitals but how, all the DRG groupers are going to have to change and what the impact would be. That's a hospital question.
DR. FITZMAURICE: I want to ask Steve a question about the DSM codes. We've asked other coding developers questions about the costs of producing it, the charges for the code books, the fact that somebody can pull up an ICD-9-CM and use your codes and developing codes and maintaining codes is not free so I'm not going at it from the standpoint that all coding systems ought to be free or all of them should be supported by the government or all should be supported in the private sector.
Just questions of information. Do you charge for DSM-4 code books? Generally, what is the charge? Do you have a pricing structure for electronic acquisition of your codes? That is, here's a disk and you can use it in a vendor system. Is there a charge for having your code used in ICD-9-CM?
DR. Mirin: I would make a distinction between the codes themselves which are, in fact, the ICD-9-CM codes. Those are available. What we develop is a set of descriptors in the form of a text book which clinicians use so that they understand what they are talking about when they use a particular code so the codes themselves are free and available to anyone. What we've copyrighted is the text that goes along with the codes and that's essentially what we provide for the field.
DR. FITZMAURICE: Is there a charge for providing the -- do definitions go along with the codes?
DR. Mirin: The definitions that go along with the codes are part of the textbook in essence. It's called DSM-4-R at the present time, and there's a significant cost that APA incurs in the ongoing development of every year we put aside a very substantial amount of money so that we can work on the next iteration of the DSM.
DR. FITZMAURICE: Is this out of the goodness of your heart or do the charges for the book generally cover the costs of maintaining and updating the coding system and definitions?
DR. Mirin: I would say that it's basically our obligation to the field to continue to update the classification of disease that essentially is the hallmark of the field. We certainly derive some revenue from selling this text although interestingly enough, the vast majority of purchasers of this particular text book are actually at the college level where people learn about the classification of mental disorders. It's not psychiatrists per se.
DR. FITZMAURICE: So the costs of maintaining this, part of it comes out of the revenue from the books, part of it comes out of, say the dues that a member of the association would pay as part of their professional responsibilities and obligation?
DR. Mirin: That is right.
DR. FITZMAURICE: Thank you.
DR. ZUBELDIA: Dr. Cohn and then we'll break for lunch.
DR. COHN: This is just sort of perhaps a summary question. Recognizing that it's going to take some time for the industry to transition, how much lead time do you think the health care industry needs to move to something like ICD-10 diagnosis?
DR. Mirin: Since the DSM-4 is totally compatible with ICD-10-CM, we could do it probably by this afternoon.
(Laughter.)
DR. POWERS: Just, once again, comparing it perhaps unfairly to CPT-4, this is not a new concept. It's a matter of just learning different numbers and the categories and the language is actually improved so the -- I don't know how long it will take the health care industry to change their computers, but from the physician standpoint, we can learn fairly quickly, and I would say that the mandated one year, if that's true that that's a previous mandate, two years, that's plenty of time for us to educate our members.
DR. COHN: Okay, so within two years you would be able to get all your systems ready, everybody in your offices trained in all of this and be comfortable with that.
DR. POWERS: We did it faster than that for CPT-4 I believe.
DR. STRUNK: And we would be able to do it within that time frame. Our greater concern is what is going to happen in other elements of the industry with respect to their software and their computers.
DR. WARNER: I guess I don't feel I'm qualified to answer this question.
DR. MULDOON: Conceptually it's not difficult. I think our speakers this afternoon will speak more specifically to some of the software and related implementation issues. I would defer to their knowledge.
DR. ZUBELDIA: Thank you. Thank you very much for your insightful testimony. We are going to recess for 55 minutes and we will reconvene at 1:00.
(Whereupon the meeting recessed for lunch at 12:05 p.m., to be reconvened at 1:00 p.m.)
DR. ZUBELDIA: We are reconvening this panel on the possible replacement of the ICD-9 diagnosis with ICD-10. This morning we had a panel of users of ICD-9 diagnosis codes or ICD-10 diagnosis codes, and we are going to continue in the afternoon, this is a continuation of the first panel, the last panel from this morning, and we will be hearing from other users of the code sets, and from a vendor. We will start the panel with Sue Prophet and then we will move on this action.
Agenda Item: Sue Prophet, American Health Information Management Association
MS. PROPHET: Thank you. Chairman Cohn, members of the NCVHS Standards and Securities Subcommittee, ladies and gentlemen, good afternoon.
I am Sue Prophet, Director of Coding Policy and Compliance for the American Health Information Management Association. On behalf of the association, thank you for allowing us the opportunity to provide input on replacing the ICD-9-CM diagnosis coding system with ICD-10-CM.
As you know from our previous testimonies, AHIMA is a professional non-profit association of more than 41,000 members who manage patient information in the form of health records and data bases and provider health plan, government and private organizations. AHIMA recognizes the significant importance of coding and coding standards and consistency of health care diagnostic and procedural coding is one of our association's key issues. Attachment one of our written testimony is AHIMA's statement on consistency of health care diagnostic and procedural coding.
Our testimony today is directed at the replacement of ICD-9-CM diagnosis codes with ICD-10-CM. Obviously, in the interest of time, my comments this afternoon are an abbreviation of our written testimony which you all have.
I'll cover AHIMA's perspective on the status of ICD-9-CM diagnosis codes and whether they should be replaced, the proposed ICD-10-CM, lessons learned from the experience of other countries that have implemented ICD-10, and implementation issues that will arise if ICD-10-CM is adopted and specifically on training of health care professionals in this coding system, whether they are professional coders or others whose tasks include using this system, and why the Department of Health and Human Services Secretary's decision concerning ICD-10-CM should not be made in a void.
Regarding replacement of the ICD-9-CM diagnosis coding system with ICD-10-CM, AHIMA believes that replacement of the 9-CM diagnosis codes is absolutely necessary and 10-CM should be adopted as the replacement. You have all already heard today from other presenters regarding how obsolete 9-CM is, how out dated the terminology is and inconsistent with current medical knowledge so I won't belabor the point of how bad off 9-CM is.
Since many countries have already implemented ICD-10 or are in the process of doing so, continued use of 9-CM causes significant difficulty in comparing international data used in public health, research and development, and the study of issues related to quality, safety and efficacy of medical care.
As you have heard today, our ability to compare mortality and morbidity data within the US was compromised as ICD-10 has been used in the US for mortality statistics since 1999. The need for greater coding accuracy and specificity has heightened considerably since the implementation of 9-CM. ICD was primarily used in the hospital in-patient settings for indexing purposes at the time 9-CM was implemented. Therefore this system is woefully inadequate for meeting the needs of non-inpatient hospital settings or serving the many other purposes for which ICD data are currently used.
9-CM codes were not used for reimbursement purposes until well after the system's development and implementation. Once prospective payment systems came into existence, the concerns of data quality, coding education, and medical record documentation received new emphasis.
The consequences of inaccurate claims data in a fee-for-service environment had not been nearly as critical. Today, several prospective payment systems for different health care settings have been implemented which depend heavily on 9-CM diagnosis codes, and many non-PPS payment methodologies require complete, accurate, and detailed coding in order to negotiate or calculate appropriate reimbursement rates, determine coverage, and establish medical necessity. Daily, health care entities including provider, payers, researchers and other users face problems caused by continued use of an obsolete diagnosis coding systems, and some of these problems are described in further detail in our written testimony.
AHIMA fully supports the adoption of 10-CM as the replacement for the 9-CM diagnosis codes. We have been involved with the US implementation of ICD-10 since 1987, when we established an internal task force to look at issues related to implementation of ICD-10 and interestingly enough, our conclusion from that task force raises many of the points and many of the discussion items we are having here today. Not a lot has changed.
We were represented on the NCHS technical advisory panel, and submitted many extensive comments regarding the draft of 10-CM when it came out.
As you have heard, 10-CM represents a significant improvement over both 9-CM and ICD-10. It incorporates much greater specificity and clinical detail which will result in major improvements in the quality and usefulness of the data for all of the purposes that we have outlined in our written testimony. It provides much better information for non-hospital, in-patient encounters, clinical decision making, and outcomes research.
The new alphanumeric structure allows considerable space for future revision without disruption of the numbering system, much more than is possible with 9-CM and notable improvements in the content and format are described in detail in our written testimony.
A new code set today must meet the requirements for adoption and maintenance set forth by HIPAA and 10-CM meets these criteria. Attachment three of our written testimony outlines the HIPAA criteria. AHIMA does not believe that reimbursement considerations should drive code set revisions; however, good, specific coded data should be use and can determine and support appropriate reimbursement. The payment computation system and not the coding system should define the payment allowing for cost service or location data or cost analysis, research and so forth.
The level of specificity in 10-CM will provide payers, policy makers and providers with more detailed information for establishing appropriate reimbursement rates, evaluating and improving the quality of patient care, improving efficiencies in health care delivery, reducing health care costs and effectively monitoring resource and service utilization.
We believe that final regulations promulgating adoption of 10-CM as the replacement for the ICD-9-CM diagnosis coding system should be issued no later than 2003 with an implementation date no later than 2005 in order to avoid further delay in implementing this code set.
The US needs to move forward as quickly as possible to implement a much-improved diagnosis coding system. Our country needs to move forward in order to fully benefit from the code set enhancements that are necessary for high quality data and ultimately high quality patient care.
HIPAA implementation is often raised as the reason for holding back on the decision to replace the 9-CM coding system. With the current US diagnosis coding system in crisis and the rest of the world completing implementation of ICD-10 or a modification, the US delayed implementation and we are daily losing domestic and international data comparability and delaying realization of the many benefits of using this much improved coding system.
A decision to implement 10-CM cannot be delayed until all of the current HIPAA rules are fully implemented, nor can the health care industry afford to hold off the reimbursement and other considerations that are impacted with any coding system delay. The definition of the ICD-10-CM coding standard should include the system rules and guidelines in a process allowing public input to codes that maintenance should be continued with the implementation of 10-CM.
An advisory group comprised of representatives of stakeholders should be established to provide input into the maintenance and guideline development processes.
In February, we indicated that maintenance is a key issue for any complete and flexible medical coding system. AHIMA has espoused six principles for medical code set maintenance processes which are described in attachment four.
The existing official coding guidelines will need to be revised for use with 10-CM. The current development, maintenance and approval processes for developing the 9-CM official coding guidelines must continue for 10-CM and these guidelines as well as the rules and conventions must be included in the 10-CM code set standard. This will insure proper and consistent applications by all users.
A standard effective date for 10-CM implementation should be established by which all affected payers and providers would be required to abide. A formal implementation plan encompassing all effected entities should be established once adoption of 10-CM is a certainty. Such a formal implementation plan, established well in advance of the scheduled implementation date, within the requirements of HIPAA, will insure a successful transition to 10-CM.
It is imperative that the NCVHS and the HHS Secretary with industry consultation initiate such a plan immediately given the time constraints required for proposed and final regulations and the various activities necessary to prepare for implementation. Delay in such decision making and commitment could leave the US with a hopelessly outdated system unable to provide the country with meaningful data necessary for all the purposes outlined in our written testimony and an inability to compare global or morbidity and mortality data.
To learn from the ICD-10 implementation experience of those who have gone before us, AHIMA surveyed several countries that have already implemented ICD-10 or a modification for morbidity purposes. A number of countries that responded expressed consternation regarding the failure of the US to have adopted 10-CM by now, noting the detrimental impact on international health care data comparability.
Like the US, both Australia and Canada developed a modification of ICD-10 for use in their countries -- ICD-10-AM and ICD-10-CA, respectively. To varying degrees, both countries use ICD for reimbursement purposes. A major impetus behind their decision to adopt ICD-10 was the need for a more up-to-date classification for a more clinically credible case mix system, an improved applicability in non-hospital settings. Canada was the first country to move to an entirely electronic product for ICD-10 codes.
One significant challenge reported by both countries was a lack of vendor readiness despite ongoing communication from the start of implementation planning.
Development of 10-CM educational materials should commence as soon as the coding system is finalized and publicly available. It is extremely important for the strategic progression of 10-CM implementation to be determined as early as possible. There are many more stakeholders today that when 9-CM was implemented. And the range of health care settings using 9-CM diagnosis codes has expanded considerably.
It would be unwise to neglect the magnitude of the training issue; however it's important the recognize that since 10-CM retains the traditional ICD format and many of the same conventions, it will not be difficult to learn for those users who have been trained in 9-CM diagnosis coding.
Implementation of any new coding system would require training for coders, the clinicians responsible for documentation, and a growing number of data users throughout the health care industry. The range of users and settings for which training programs have to be designed and provided is much wider for 10-CM and 10-PCS. Many of the categories of data users are listed in our written testimony.
Work force shortages and variable skill levels among those responsible for coding in diverse health care settings also mean there are many on the-the-job trained coders for whom re-training will be more demanding and coding professionals must understand how to properly use and interpret data derived from both the 9-CM and 10-CM diagnosis codes. Experienced coding professionals will primarily need to be educated on changes in structure, disease classification, definitions and guidelines.
The greater degree of clinical specificity in 10-CM will require coding professionals to have a better understanding of medical terminology and human anatomy. Because there is greater specificity in 10-CM, clinicians will have to change behaviors in documentation so that the more specific appropriate code can be selected, a goal which is also consistent with the industry's goals to eliminate medical error.
In Australia, two-day training workshops for experienced coding professionals were conducted. In Canada, education was provided in three phases, including self-learning packages and practical application in a computer laboratory. All of the training programs in Canada utilized coding software instead of hard copy code books. Both countries offered post-implementation refresher workshops to coding professionals. Special educational sessions were provided to data users.
A crosswalk between 9-CM and 10-CM should be developed and it's our understanding that NCHS will be developing or has developed such a crosswalk. Information systems personnel will need to understand the logic and hierarchical structure of 10-CM and data users will need to understand the definition and composition of categories in the classifications.
A comparability study to assist users in discriminating between real changes and utilization by diagnosis and those changes that are artifacts of coding system changes should be conducted and it is our understanding that NCHS plans to conduct such a comparability study.
During the transition period the quality of data will undoubtedly suffer as implementation of any new coding system involves a learning curve. The average learning curve in Australia and Canada was four to six months for coding professionals to feel comfortable with the new coding system. Both countries reported that coding professionals found ICD-10 no more or less difficult to learn than ICD-9. Productivity by both those reporting the codes and those processing the claims will undoubtedly be temporarily affected during the transition period until proficiency and confidence in using the new system are achieved.
Caution should be use when conducting longitudinal data analysis diagnoses may be classified differently in the two systems or code definitions may have changed, making it easy to misinterpret data. Users will need to understand the changes in trends or case volume may simply be a reflection of changes in data reporting.
We believe the interactive, internet-based training would be an excellent methodology available today for training both coding professionals, end users of coded data on 10-CM, relatively quickly and specifically.
While the transition poses significant challenges, there are, fortunately, many training vehicles available and appropriate for this effort. Today's training methods have significantly improved. While face-to-face seminars remain an effective training vehicle, new technologies offer probably superior alternatives. Audio seminars can be delivered at low cost to a large audience and web-based training offers accessible, flexible and highly effective training opportunities.
Given the variety of educational media available and the number of coding products and services vendors, we anticipate numerous vendors will offer a variety of competitively priced educational programs for 10-CM in formats to suit everyone's needs and preferred learning styles.
For entry level and advanced coding programs and for the coding component of health information management academic programs, we anticipate the 10-CM training would be incorporated into the existing ICD coding curricula. We also expect a period of time when students enrolled in coding programs must learn to use both 9-CM and 10-CM diagnosis codes.
AHIMA has a significant history and is committed to the training of the industry in health information management and education will be made accessible through the internet and our strong network of components, state associations and coding professionals.
Because the scope of health information management includes coding, we are certainly prepared to also address appropriate coding documentation, management, including development and coordination of implementation strategies and data analysis issues related to this change.
Coding leadership, professional development and coding consistency are central, strategic and mission issues for AHIMA so you can certainly count on our support.
One last issue to keep in mind with training is that while the educational materials can be developed quickly after 10-CM is finalized and released, the training itself should not be conducted so early in advance of implementation that effectiveness of the training is diminished.
Ideally, while 10-CM and 10-PCS should be adopted at the same time in order to limit the transition period when disruption of data quality, coding accuracy, coding productivity and prompt reimbursement of claims will occur, we acknowledge that the systems are so distinctly different that it would be possible to implement them at separate times, and consideration must be given as to the benefits and drawbacks of implementing the two systems at different times versus simultaneously. This is a question that would need to be addressed in the context of the environment and in order to insure that there are appropriate timetables and strategic planning developed to accommodate the needs of the industry and not just one coding system.
AHIMA realizes that there will be resistance to changes in the status quo, but it must be recognized that while there are significant cost implications associated with adopting 10-CM as a standard code set, there are also significant costs associated with maintaining the status quo.
We are paying a very high price for having delayed this long and the cost continues to increase. We are already at least a decade behind in implementing new ICD modifications and catching up is more costly than staying current. Further delay will not reduce the direct costs and we cannot ignore the indirect costs any longer.
10-CM has the capacity to grow as medical science grows, and it could serve our health care diagnostic data needs for many, many years to come. We believe that the benefits of the improved data resulting from 10-CM are well worth the cost and the difficulty encountered during the transition period.
We also believe that we are incurring significant cost by utilizing a hopelessly outdated and limited system and that ultimately reductions in cost will be realized as a result of the availability of better data.
A number of benefits that we expect as a result of implementing an improved coding system are outlined in our written testimony.
Benefits noted by Australia and Canada as a result of implementing ICD-10 include greater detail and specificity, a more clinically credible classification system resulting in significantly improved data and a system that is easier to use and understand by coding professionals and other users; a much-improved coding system structure and code presentation; and the ability to conduct international comparisons.
In conclusion, AHIMA and its national network of coding professionals are uniquely capable of assisting in the research on the best implementation strategy. Our coding professionals have the training and expertise to quickly learn and utilize a new coding system and limit the learning curve. AHIMA is also uniquely capable of taking the lead in national workforce retraining and user education for both new procedure and diagnosis coding systems.
Finally, Mr. Chairman, AHIMA is pleased to note that we have been asked by Mr. Steven Beal, senor vice president and chief financial officer of the Federation of American Hospitals to indicate to you and the full subcommittee that the federation has endorsed and recommends the content and approach of our testimony today.
Thank you for the opportunity to present our views regarding replacement of 9-CM diagnosis codes with ICD-10-CM and the environment that must be considered and part of this decision. AHIMA is deeply committed to working with the Department of Health and Human Services, with NCHS and other health care industry groups to advance coding practice and improve our nation's health care data through adoption of new code set standards.
Today I am accompanied by Linda Cause, AHIMA's Chief Executive Officer, and Dan Rowady, AHIMA's Vice President for Policy and Government Relations, and we are prepared to answer any questions or concerns the subcommittee might have today, during the two days of testimony, or in the future. Thank you.
Agenda Item: Chris Frazier, American Academy of Professional Coders
MS. FRAZIER: My name is Chris Frazier and I am from the American Academy of Professional Coders and this is the first time the AAPC has been at one of the meetings to give testimony, and to tell you the truth, I arrived with more questions, and the questions have changed while listening to the other discussions, I have more questions than I do have recommendations.
But to begin with, from this morning's discussion, it appears that there is agreement to the superiority of I-10 over I-9 as well as agreement of the concerns of cost in terms of training coders and also the cost of software.
From these discussions -- and I still agree with my testimony that we do need more information about education and training materials -- but it did bring up more questions this morning about implementation.
First, from what I understand, and I want these issues clarified because I have to go home and give information to my organization so as I understand, there will be a 24-month implementation period at the publication of the NPRM in the Federal Register. Is that set in HIPAA and so that's, say, if the final rule came out in 2003, implementation would be in 2005 which would give us a two-year lead time. I don't know if you can answer that specifically, but I am just looking for information that I can take back to the coders without misleading them or giving them any misinformation because we do represent about 28,000 people.
MR. BLAIR: Kepa, I know that this is not our normal procedure, but we might give consideration to giving her answers to her questions immediately because --
DR. ZUBELDIA: If the questions are as simple as that one, yes.
MR. BLAIR: Because it may also lead to other follow-on questions which may give us some useful information.
DR. COHN: Actually, Kepa, I would to offer you the opportunity to answer this simple question and we will see if we agree with your answer.
DR. ZUBELDIA: After publication of the final rule there is 60 days for Congressional review, followed by 24 months until the final rule becomes mandated. They actually have 26 months.
MS. GREENBERG: That is the original rule but that's not the revision. This would be a revision to the rule.
DR. COHN: Actually, if I can comment, even before Karen, I think this may be an issue that we may want to recommend in a letter because it sounds like it could either be considered a code set update or it could actually be something that goes to a full final rule. Betsy is looking at me.
MS. HUMPHREYS: My feeling is that when, certainly the intent of everything we've done with this rule, we do not regard, I mean, ICD-9-CM is updated every year. We do not regard the change from ICD-9-CM to ICD-10-CM as just continuing to update 9-CM. I mean it as a new code set I think, from the point of view of the way we've been interpreting the rule.
MS. TRUDEL: It is not an initial standard.
MS. HUMPHREYS: No. It's a revision to the standard.
MS. TRUDEL: That should require two years from the initial standards so when the first transactions in code set rule is published, when the first security rule is published, there is a 24-month implementation time clock. On any subsequent ones, it is 180 days or more, and I think Simon is right. If there's a general consensus that two years is what the industry needs, then that's what ought to be recommended to the Secretary and it is possible to place a future implementation date on a regulation so that's perfectly doable.
MR. BLAIR: But don't confuse the final rule with the NPRM. The NPRM is the notice of proposed rule making that would come out first and then there would be public comment on that before the final rule would come out.
MS. FRAZIER: Okay, thank you. My next question. From my understanding this morning, the NCHS will publish the complete version of I-10-CM on its web site, and I believe this was answered when I was speaking to someone earlier. And this will be -- when do you anticipate publishing this, I'm sorry, near final draft, and does the near final draft include the incorporation of public comments? I believe you did say it, from the draft version that was taken off the site I believe, oh, two, three years ago and so it incorporates those comments, and that will be available on the NCHS site and anticipated time?
MS. GREENBERG: Was your question whether the --
MS. FRAZIER: Yes, I'm wondering if once that near draft version is up on the NCHS site, will we be able to use that at the AAPC for developing materials for our coders? Is that near enough a final that we can take that and start educating our coders on the system?
MS. GREENBERG: The guidelines will be posted.
MS. FRAZIER: I'm sorry for deviating from how you usually do this, but it's nice that this has raised so many questions.
DR. COHN: Other people will have these same questions.
MS. PICKETT: The ICD-10-CM will be posted shortly after the entire 10-CM is posted, the guidelines will be posted, too. At some point we will also be posting training material that is specific to that and from there generally what the groups that we have collaborated with have indicated is that they would take that material and revise it to accommodate their constituencies. I think you heard Dr. Powers from the neurological association mention that they are prepared to work with their constituency in the training but it would be based on the training materials and such that we have prepared as a baseline, recognizing the fact that other constituencies may have slightly different needs in terms of the details of their particular professional group.
MS. FRAZIER: Okay, and Ingenix published a book, I-10 Made Easy, and it was kind of, just gave information about the I-10-CM. Would it be premature for the AAPC at this point to publish information, say, in our newsletter comparable to I-10 Made Easy that would give an overview of the I-10-CM for our membership?
MS. PICKETT: First, it would be difficult for me to respond to that. What Ingenix has done is their property so I can't speak to what is in it or because it was not developed by NCHS and I'm only really able to reflect on what NCHS has done and not what a private publisher has accomplished.
MS. FRAZIER: I'm sorry, I'm just wondering when we can get the information out because we get so many calls from our membership about the both the procedural and the diagnostic coding.
As you had mention before, the training and education should not be so difficult because of the similarities between I-10 and I-9 and so, but you are going to, there will be training materials available.
I thank you for this information that we heard this morning because I can take this, as I said, home to my membership but I'm still concerned with a definitive time line for the training and education of coders. The discussion this morning raises several questions so I feel like, again, I'm turning the discussion back to the NCHS.
As I said, the AAPC does not want to alarm the membership with premature information and so then, as you answered my question, I can go back to Salt Lake City with the information that it could be two to three years for the implementation of I-10-CM and that it would probably be premature for us to develop any educational materials at this time, but the AAPC would like to work with the NCHS in developing any of this information, disseminating the information.
And one last comment, and I'm speaking to the concerns of our coders. From what we get from our coders, the information, the calls, we are very happy to hear and heartily support the uniform coding guidelines' reporting requirements and code acceptance by all users and we agree that the current I-9 makes for imprecise coding and inaccurate documentation for reimbursement.
We thank you for the work that has been done on I-10 and we are looking forward to implementation and especially when it's going to be posted so we can start reading and reviewing. I won't read the rest of my testimony. A lot has to do with what Sue said with the implementation in Australia, and I have not changed it much from that, but I wanted to add those other comments so thank you.
MS. LEON-CHISEN: Good afternoon, Dr. Cohn, Dr. Zubeldia, subcommittee members, ladies and gentlemen. My name is Nelly Leon-Chisen. I'm the Director of Coding and Classification at the American Hospital Association, AHA. On behalf of our nearly 5,000 member hospitals, health systems, networks and other providers of care, I thank you for the opportunity to comment today on the possible future implementation of ICD-10-CM.
The AHA central office on ICD-9-CM serves as the United States clearinghouse for issues related to the use of ICD-9-CM. The office was created as a result of a memorandum of understanding between the AHA and the Department of Health and Human Services in 1963.
The office is also the publisher of coding clinic for ICD-9-CM and collectively with CMS, NCHS, the AHA and AHIMA are known as the cooperating parties.
Accurate and precise reporting of clinical codes is extremely important because clinical codes are key to benchmarking, quality assessment, research, public health reporting and strategic planning, in addition to accurate reimbursement.
For instance, ICD-9-CM codes allow hospitals to develop critical pathways for dose diagnoses or procedures that are high volume, high risk or high cost, and those in which the course of treatment is similar between patients. The hospital can perform case selections for studying best practices in disease management about using the ICD-9-CM diagnosis codes. Hospitals can use coded data to identify the most prevalent diagnoses in their communities and determine what services are most needed to provided the best care to their patients.
Why change? We've heard already this morning that ICD-9-CM has been in use for more than 20 years and an update is long over due. Over time, we have seen a shift in the way we deliver care from inpatient acute care to out patient, home care, long term care, and other settings.
So there is no doubt in our minds that ICD-9-CM diagnosis codes are outdated and insufficient to provide detail for non-acute conditions. For many years now, hospitals have been aware that ICD-9-CM was becoming outdated and would need replacement. Hospitals have been ready to accept the change to ICD-10-CM.
Dr. Sondik this morning referred to the NCVHS 1993 report and we agree with that statement by the committee, quote, because the international classification of diseases 9th revision, ICD-9, is rapidly becoming outdated and worldwide support and maintenance of the classification will terminate and because the international classification of diseases, 10th revision, ICD-10, represents significant improvements in coding of primary care encounters, external causes of injury, mental disorders and neoplasms, we have recommended that the department immediately commit resources to assess the applicability of the ICD-10 for coding morbidity, to identify problem areas and make modifications and necessary, and to develop implementation plans, end of quote.
The AHA believes that this statement made by NCVHS nearly ten years ago still holds true. Since then, ICD-10 has been reviewed and critiqued in the United States by hospitals, coders, medical special societies and researchers.
Areas for improvement have been identified and modifications made. The final and complete portion of the 1993 NCVHS recommendations still remains the development of implementation plans. The AHA continues to support the 1993 NCVHS recommendation and supports a migration to ICD-10-CM.
We have also heard this morning that ICD-10-CM is more comprehensive than ICD-9-CM and more closely parallels the vocabulary in practice with current medicine. This enables more detailed and accurate classification of diseases which leads to more efficient and effective data retrieval.
I will not list the content improvements over ICD-9-CM. These have been carefully outlined by previous speakers this morning. One never knows what other testimony is going to show up.
More detailed code assignments made possible by ICD-10-CM are needed. They would greatly reduce administrative burden for hospitals. More detailed coding would reduce the requirements for submission of additional documentation to support claims, allow the capture of accurate data on new ways of describing diseases due to advances in medical knowledge, provide data to support performance measurement, out comes analysis, cost analysis, and monitoring resource utilization and increase the sensitivity of the classification when making refinements in applications such as grouping methods.
I will now address issues related to the implementation of ICD-10-CM such as timing, cost, and testing needed.
Many of these issues are similar to the issues we have previously raised in testimony to the subcommittee related to the implementation of ICD-10-PCS. These are issues that would need to be considered in changing any clinical coding system.
With regards to timing, the time line for implementing the new system should be carefully orchestrated to minimized administrative burden to providers. Hospitals are already facing numerous regulatory changes over the next several years, including the HIPAA privacy, security and electronic transaction standards as well as implementation of new prospective payment systems, all of which will add significantly to the hospital's burden and cost.
Therefore the ability of hospitals to absorb all of these regulatory changes must be carefully taken into consideration. The vast majority of hospitals are dependent on their hospital information system vendors, or HIS vendors for programming changes. Therefore the AHA supports the HIS industry in requesting the ICD-10-CM implementation for diagnosis coding be carried out in tandem with the migration to the ICD-10-PCS procedure codes.
The AHA also supports the recommendation to implement ICD-10 three years after implementation of the HIPAA electronic transaction standards.
Lessons learned from the international implementation of ICD-10 diagnosis coding have shown that it is taking a minimum of two years to implement this system. In the United States there is a greater level of automation and a more pervasive use of clinical coding across different health care delivery systems than in the rest of the world. An additional year would be beneficial to insure that thorough testing of the information systems and the electronic transactions utilizing the codes takes place.
Accurate coding translates to accurate and more precise reimbursement for services delivered. A failure of the language used to precisely report what was wrong with the patient and what was done for the patient imposes serious financial limitations for our hospitals. The submission and payment of electronic claims is likely to be constrained due to the limitations of the current code set.
Cost. Implementation of ICD-10, both diagnosis and procedure coding will be a complex and costly process. Therefore, Medicare along with other health plans should be sensitive to these increased regulatory costs and address current and future payments accordingly. The AHA believes that the cost of implementing significant new regulations that affect provider operations should be worked into the Medicare prospective payment rate update, particularly since these regulations are not factored into the current prospective payment rates.
Further, we believe that Congress should establish grants to help hospitals with the enormous cost of complying with the HIPAA rules, including conversion to an entirely new coding system such as ICD-10-CM. For hospitals, the bulk of the cost associated with the adoption of the new classification system will be related to personnel training. Hospital support staff such as coders and billers will require training to familiarize themselves with the new coding guidelines, rules and definitions. The cost of appropriate training could be quite high for hospitals; however, the basic principles under lying ICD-10-CM are similar enough to ICD-9-CM that experienced coders should be able to quickly learn the new system.
In areas where testing occurred with a draft version of the ICD-10-CM codes, experienced coders were able to follow the draft and identify how the new codes would fit in with the existing ICD-9-CM diagnosis codes. Learning ICD-10-CM does not have to be a monumental task if sufficient time and planning efforts are undertaken.
We understand that changes to the coding system will also require extensive and costly modifications to information systems. Hospitals use a combination of purchased software and in-house developed applications. The software applications that will require modification and countless functions such as code assignments, medical records abstractions, aggregate data reporting, utilization management, clinical systems, billing, claims submission, groupers, the list is fairly long. In essence, every electronic transaction requiring an ICD-9-CM diagnosis code would need to be changed.
These changes include software interfaces, field length formats on screens, report formats and lay outs, table structures, holding codes, expansion of flat files, coding edits and significant logic changes. Hospitals will have to bear the financial burden associated with software changes as well as possible hardware upgrades while in the end the migration is beneficial to the nation's health care system, the hurdle for our members is the initial investment needed to prepare for these changes.
During the transition period, information systems will have to support both ICD-9-CM and ICD-10-CM coding systems, requiring additional data storage space. Smaller rural health care providers in particular, many of whom are facing serious financial challenges and have less sophisticated information systems are further handicapped in their ability to accommodate such changes and may require additional resources and support to help them acquire information and coding system support programs.
Testing. Although ICD-10-CM has not been thoroughly tested in the United States, the WHO version of ICD-10 as well as two different clinical modification versions, the Australian version and the Canadian version, have already been successfully implemented into real life application. Successful implementations have also taken place from western European countries with sophisticated information technology, progressive health care delivery systems similar to the US in experienced coders, to countries with minimal automation, health care delivery systems where only minimal basic services are provided and new or inexperienced coders.
The countries that have implemented ICD-10 have used it for a variety of functions, from management of utilization resources to clinical management, research, as well as reimbursement. ICD-10 has even been implemented internationally in diagnosis related groups or DRG's or DRG-like systems, some for payment and others for clinical management purposes.
America's hospitals are capable of taking on the challenge of implementing ICD-10-CM. A review of the ICD-10-CM draft back in 1998 found that the changes made to ICD-10-CM were logical, coder friendly and an improvement over ICD-9-CM. The greater detail in specifying conditions and the expansion to unique codes for conditions that were previously dedicated to catch-all categories such as all other specified will make for more specific data collection.
Many previously confusing areas have also been corrected with a greater level of detail. A review was performed from the standpoint of evaluating whether the system would be capable of answering the types of questions that a hospital coder would have with the types of questions often submitted to the central office clearinghouse by ICD-9-CM users. Additional testing is needed to insure that the alphabet is as useful and complete as possible. We would welcome the opportunity for the AHA to work with our members to conduct further testing including medical records review across different health care settings.
Additionally, the testing should also consider the comparability of the new system with current payment systems including prospective payment system. The AHA expects that over all the use of ICD-10-CM will not result in lower reimbursements as compared to levels they would receive by using ICD-9-CM.
We ask that you consider the following implementation issues. We would support migration to ICD-10-CM after testing has taken place and start-up funding is established. Medicare as well as other payers should recognize the increased cost of these regulations and should adjust their payments accordingly.
The AHA would be pleased to assist NCVHS and CMS in identifying the nature and magnitude of these additional costs. We support a well-defined implementation and maintenance classes. The process should be broad based and take into account the needs of all users. It should also be predictable and take into the account the capabilities of the users to adapt to coding changes when they occur. This includes establishment of routinely scheduled meetings to review coding changes and a date certain for using approved codes.
The AHA supports the current ICD-9-CM coordination and maintenance process and would support the same process for ICD-10-CM. The current process is well-positioned to reach out to the broadest audience possible.
We believe there should be clear, unambiguous instructions and consistent official coding and reporting guidelines. These should also be readily available and recognized as well as accepted by all payers preferably as part of the HIPAA standard code set just as the current ICD-9-CM official guidelines for coding and reporting are part of the standard code set.
AHA supports the continuation of the role of the cooperating parties in the development of guidelines and clarification on the application of ICD-10-CM and ICD-10-PCS as is done for ICD-9-CM.
The AHA, acting through a long standing memorandum of understanding with HHS stands ready to provide ICD coding advice and training and it intends to continue in this capacity under ICD-10. The AHA is uniquely positioned and ready to take a leadership role in the training of our members' facilities.
Our members include hospitals and health systems providing services across the continuum of care, rehab, skilled nursing home, health and out patient services in addition to acute, sub-acute and long term in-patient hospital care. Our members look to the AHA for guidance and support in coding training and education. Today the AHA's coding clinic for ICD-9-CM and the advisory board's purpose of nationally recognized resource for coding advice. As such, we have an established process that reduces confusion and provides clarification and consistent implication of coding and rules.
Such a process would also be beneficial for ICD-10. A detailed implementation time line along with milestones should be developed. The AHA would be happy to work with NCVHS and CMS to assist in developing a reasonable implementation time frame. Lessons learned from international implementations of ICD-10 should be applied. Backwards and forwards, electronic crosswalks between ICD-10-CM and ICD-9-CM diagnosis code should be made available free of charge or at a reasonable cost.
Again, thank you for the opportunity to provide comments to you. I will be happy to answer any questions you may have.
Agenda Item: Louise Smith, McKesson Information Systems
MS. SMITH: Good afternoon. My name is Louise Smith. I'm a Registered Nurse working as the Manager of Regulatory Assessment and Operations in Regulatory Affairs Department of McKesson Information Solutions.
McKesson Information Solutions' software applications include enterprise-wide patient care, clinical, financial and strategic management of software as well as internet based and networking technologies, electronic commerce, outsourcing and other services to health care organizations throughout the world. Our customer base include payers, hospitals, practice management, basically covering the marketplace.
On behalf of McKesson Information Solutions, I would like to thank the committee for allowing us to present our comments. My comments will be brief as I feel we have pretty much exhausted the benefits of ICD-10 but regarding the benefits, ICD-10 with ICD-10-CM, we believe the structure of 10-CM seems to be logically organized and provides greater clarity for coding diagnosis.
For example, from our review of the 1998 draft in chapter four, the endocrine, nutritional and metabolic diseases, the assignment of separate codes for type one and type two diabetes allows for greater specificity with a broader range of diagnosis codes to more accurately and precisely describe the diabetic conditions of the patient. Also, the ability to describe a condition with one code rather than two or more codes should reduce coding or errors.
The specificity of 10-CM would also facilitate DRG assignments. For example, in the fiscal year 2003 proposed in-patient prospective payment system rules, it notes the need to create new diagnosis codes for myasthenia gravis in crisis in order to better differentiate between mild and severe forms of the disease which are currently classified with the same code. The rule states that adding a new myasthenia gravis code would better insure equitable DRG assignments.
Another advantage to the expanded number of diagnosis codes is that it can facilitate a more accurate collection of outcome data by having greater precision with linking patient diagnosis to out comes. This can also lead to greater ability to measure the effectiveness of services provided and the management of disease processes as well as the safety and efficiency of care.
Additionally, migrating to ICD-10-CM allows for an expansive comparison of medical data among other countries throughout the world, countries such as we've already mentioned -- Australia, Canada and the UK -- which are already on ICD-10.
Although it is our opinion that the advantages of transitioning to ICD-10 will prove quite beneficial for clinical usage, it's important to know that as an information solution provider, our interests rest mostly with the timing of ICD-10-CM. We do not use the codes as a medical record coder or a clinician will use them. Our use of the codes is more in the area of software applications. In the second vendor session this afternoon, I'll discuss more thoroughly the application software.
Thank you.
DR. ZUBELDIA: Thank you. I have a couple of questions. First, for the entire panel. We've heard this morning that the government should support this. The government should fund it. In fact, we've heard that repeatedly. What if there is no funding? Would that change any of your views?
MS. PROPHET: Well, we believe that, and we haven't actually taken a firm position on who should fund it, but we believe that ultimately, and as some of the presenters this morning indicated, it can be tough to quantify the benefits, but we believe that there are a lot of cost reductions and other financial benefits ultimately that may not be realized so frankly, we believe that we must go forward with 10-CM even if there is no government funding to assist in doing so.
MS. LEON-CHISEN: Government funding would be helpful. It would ease the pain, but we believe that in any case we would still need to move to ICD-10-CM and ICD-9-CM is broken and it's just a matter of time. It will be costly but I think it's something that needs to be done.
DR. ZUBELDIA: Thank you. Let me follow up with a continuation of that. At least three of you prepared training materials for coders. What do you think of the government preparing training materials for ICD-10?
MS. FRAZIER: I thought about that, and I thought with the training material, if they were developed by the government, aren't they part of the public domain, that we could take the training materials and modify them for our users?
DR. ZUBELDIA: Yes.
MS. FRAZIER: So I don't, the AAPC would see no problem with the government-developed training materials as part of the public domain. Is that what you are asking? And would we see it as interference with our education process? Is that what you are asking?
DR. ZUBELDIA: Do you think it would be helpful?
MS. FRAZIER: Oh, it would be very helpful, yes.
DR. ZUBELDIA: Okay. Sue?
MS. PROPHET: I'm certain it would be helpful but AHIMA, certainly also would develop customized educational materials, both for our own members and for other users of the codes.
MS. LEON-CHISEN: I agree with Sue. I think government produced training materials would be a good starting point so that everyone starts out with the same instructions but as we have seen in the past, there is usually more detail needed and so I think there's also room for private options in terms of training.
MR. BLAIR: Chris, how many members of, is it AAPC?
MS. FRAZIER: AAPC, yes.
MR. BLAIR: Yes. How many members do you have?
MS. FRAZIER: Approximately, well, about 28,000.
MR. BLAIR: About 28,000? And when training is provided for 28,000 coders, do you feel like for the most part a large percentage of the coders are members of the AAPC? Is that a good number for us to be thinking about?
MS. FRAZIER: As far as --
MR. BLAIR: Well, sometimes there's professional associations where a minority are certified or a member of the professional association who perform those types of jobs. Are a very large percentage of coders members of AAPC? Or is it, is there many more coders than are part of your membership?
MS. FRAZIER: That's hard for me to answer. There's, I'm sure there's probably more coders than are represented by the AAPC. We do have chapters in every state and people join the AAPC for certification either on the hospital side or the provider side.
MR. BLAIR: All right, let me ask the question because I'm really getting at something which maybe you can answer without answering the first question.
Of those folks that have been coding ICD-9-CM, and you have gone through training processes and you have set up classes and I guess, Nellie, you do the same. What do you estimate is the time frame when you need to run another class because the attrition rate is such that you have a substantially new number of coders. Are you doing it every six months, are you doing it every year, are you doing it every two years or longer?
MS. FRAZIER: Oh, okay. What we do is that -- we were recommending that it takes two years of experience and education to teach someone the I-9 and procedural coding at this point, and we have a system that keeps people updated by earning continuing education units every year, 18 or 24 depending on your level of coding, and we would estimate probably, we were thinking 24 months for the transition period to transition from I-9, I-10 for the I-9 users, plus to introduce new coders into the system.
And our training seminars, we have different levels of the training seminars that are going on all the time and that is why I was wondering about the training materials because we have them developed for different levels of coders, say, beginning, advanced.
MR. BLAIR: How high is the level of turnover when you have trained somebody? Two years later, what percentage of those people are still coding?
MS. FRAZIER: It's amazing. At our national conference this year, we had people there coding, most predominantly were over five years, some 10, 15 years. It seems that people in the profession, and they were very dedicated people and they stay in the profession once they do have the certification.
MR. BLAIR: Nellie, do you have similar perception of the stability of the folks that we would be training?
MS. LEON-CHISEN: Well, there's a couple of different issues between established coders and getting new coders into the workforce. A lot of our hospitals prefer and hire credentialed coders, and Sue can probably speak a little more about how those are trained because AHIMA certifies coders and there's a constant new supply coming in through the educational program, but if your question relates to training existing coders, how long would it take to train them on ICD-10-CM, I think the training would be fairly short.
I mean, I think that it's just a matter of trying to reach all the coders across the country, and as Sue has mentioned there's many different ways of reaching them now. I mean, you don't have to have face-to-face seminars, I mean, now you have audio seminars, webcasts, video tapes and many other ways so it's just a matter of getting to them and I think ICD-10-CM is close enough to ICD-9-CM that it would be a matter of sort of going over the guidelines if there are any differences, instructing them on the code structure for ICD-10-CM and I would say that we could look at what Canada and Australia's experience has been and I would think one to two day's training sessions would be fairly sufficient.
MR. BLAIR: Sue, do you have anything to add on this? My question is related to either the turnover of the coders and also how long it might take to train them.
MS. PROPHET: (Tape failure.) Until the code set changes so we're always right now constantly having, developing new educational programs, audio seminars, we have a variety of web-based coding products, all for ongoing, experienced coders, and as Nelly pointed out, we also certify coders and as part of that requirement, both for our fully HIM credentialed individuals as well as our certified coding specialists credentials require ongoing continuing education to maintain their skill level because there's always changes in the official coding advice, there's changes in coding guidelines, there's annual updates to the codes, there's just, refresher courses, so all of that is going on.
Regarding the turnover rate, one thing I would like to remind the subcommittee of is the severe work force shortage that we have right now in the coding industry so we do have an awful lot of coders who are staying in the profession and it's a great field to go into right now because you can pretty much name your job and your career path.
MR. BLAIR: Sue, you mentioned in your testimony that Canada has gone to implementing ICD-10 codes CA, electronically?
MS. PROPHET: Yes.
MR. BLAIR: Could you each help us understand if you have begun to project forward whether the electronic implementation of ICD-10-CM would -- do you figure that that is going to reduce the requirement we would have for coders over a period of three, four, five, six years and if so, how much will it help?
MS. PROPHET: It could help to some extent that because of the ease of use, the coder friendliness that Nelly and some of the other presenters mentioned the productivity will actually increase which then could require less coder requirement and also the fact that it does use the clinical terminology that if consistent with physician's terminology so that going back and forth and trying to figure out which one of the very non-specific codes is the best non-specific code to use will be reduced to a certain extent which, of course, could help to minimize our serious work force issue at the moment.
MR. BLAIR: Thank you.
DR. COHN: Actually, Sue, this is meant as almost a follow-up of that work force issue and probably it's also a question for Nelly also. You had indicated that during the transition that there's about a four to six month period where people, I wasn't sure whether they don't code as well or as fast and I couldn't tell --
MS. PROPHET: Both.
DR. COHN: Both, okay. Well, given the work force issue, given if they are not coding as well, potential revenue implications for hospitals, is there a concern, is there a condition plan to ease that concern?
MS. PROPHET: Well, certainly any time you implement something new, there's, to a certain extent, going to be a learning curve as people learn something new. Also, of course, as I mentioned the specificity in the code set is also dependent on the medical record documentation carrying that level of specificity so also as part of that learning curve will be the clinician's documentation, hopefully improving but certainly realistically there will be somewhat of a learning period where it will drop off.
Although both Australia and Canada indicated it wasn't anything massive where all of a sudden everybody had several months' backlog in coding or anything like that but it was recognizable, a slip in the productivity and the quality until people really got proficient which is why both countries offered post-implementation refresher courses because what they found is it's one thing to teach it ahead of time before you are actually using it in real life practice but then after you actually have real records and are trying to apply the system, you have different questions and different things come up than you had in the initial training.
MS. LEON-CHISEN: I agree with Sue. I mean, we see that even now when new codes are created. You try to eliminate as much of the potential confusion as possible and like we do with coding clinic. When a new code comes up, we try to write up about it as much as possible, give the background as to why the code was created, what is it intended for and try to anticipate questions that a coder might have.
You know, what can this be confused with what would they need to know, what are some possible examples in applying this code and then what we see is that once people start using these codes, sometimes there are questions that still come up after implementation and then, this is what we were talking about having a clearinghouse and official guidelines and so forth so that there's a way to get a consensus opinion on how these codes should be applied and then disseminate that information to everyone so that it's being consistently applied.
MS. FRAZIER: And additionally, coding is not country-bound. At this time, there are third party payers who are sending their documents to be coded internationally overseas and I don't know what will happen in this situation. There's no way to predict how much we'll be going internationally and how that will affect the American coder and I suppose, with implementation of I-10 internationally, it would be much more of an exchange than it is now, I would assume.
There's nothing in HIPAA as I asked earlier about the transmission of information internationally. Nothing in privacy and the only requirement would be the X-12 as far as data transmission.
MS. BURKE-BEBEE: This is Susie Bebee. I have a question. I recently read a WEDI-Smith(?) white paper that was prepared by a subcommittee, ICD-10, and we have three of the work group members that I guess participated in writing that white paper.
What I was left with as an end result from reading the white paper is that WEDI-Smith could not endorse going forward with recommending ICD-10. The white paper seems to have been written a couple years ago and I'm wondering in the work group has done any more work on this white paper, if the opinion has evolved into something different because the testimony I hear this afternoon is different from what I came to conclude from reading this white paper.
MS. LEON-CHISEN: Yes, I was the co-chair of the group that developed the white paper and as you mentioned it's a two-year-old effort. In being the work of a subcommittee, obviously, there was not a consensus of opinion. I think at that point, AHA had already, to cite a testimony back in 1997 to the subcommittee, to NCVHS saying that we supported ICD-10-CM but because the WEDI paper was the work of many different organizations, it came short of actually endorsing ICD-10-CM because it was not just our opinion that was expressed in that paper.
MS. PROPHET: I would also like to add that our comments as well submitted for the content of that paper were all in favor of implementation of 10-CM as well.
DR. FITZMAURICE: I heard talk about a coding shortage, hard to find coders and I see people smiling and nodding their heads when that was raised. I'm wondering, is it realistic to count on computer parsing of narrative text, hospital discharges, for example, to assign codes, diagnostic codes, to the patient who was in the hospital, and does a move from 9 to 10 make that more realistic and would moving into a more common vocabulary areas also make that more realistic or would you count to make it realistic more on structuring the information as entered in by the person supplying the narratives such as the physician?
In other words, how realistic is it to move to computer parsing of this information to ease the burden on coders and is it realistic?
MS. PROPHET: Well, I think we have to first of all, differentiate on what a coder is and what a coder does. Certainly the production line coder of producing the codes from the code set and attaching them to the claim, that may be certainly reduced because of the greater specificity in ICD-10-CM but what we see is that the role of these educated coding professionals is more in the direction of clinical data management and the translation of code sets and what they mean and the logic behind them and in that arena versus the production line aspect that coding currently is.
DR. FITZMAURICE: Are you talking about something other than billing?
MS. PROPHET: Right, or even, you know, assessing the accuracy of the coding, applying the logic, the rules behind it, translating between code sets versus sitting there with a book and assigning the code and getting so many records done in a day as is often the current process of coding.
DR. FITZMAURICE: So do computers lend themselves more to the production line but they probably don't lend themselves to the more thinking part of it.
MS. PROPHET: Yes, exactly. They don't take the place of the human brain quite yet in.
MR. BLAIR: Is this synonymous with when you said that Canada had implemented ICD-10 electronically?
Are we talking about the same thing, that they wound up basically allowing you to enter textual information from the physician either entered textually or audibly and parse it into an ICD-10 code? Because that's what I thought you meant when you said that Canada had --
MS. PROPHET: I believe that much of their coding process is still done by coding professionals but they have definitely reduced the production side of it by using it in a data base, an electronic format which facilitates the whole process of coding.
MR. BLAIR: So it was not the same, okay.
MS. PROPHET: So they have not actually gone to completely automated coding from an electronic medical record.
MR. BLAIR: Is there anyplace where you know of where there is experience with ICD-10 and this is extending Michael Fitzmaurice's question because that's what I thought you meant when you said Canada had done it electronically. Is there anyplace else, whether it's Canada or Australia or elsewhere in the world, where they have started to implement it in a manner where people can enter the information verbally and textually and have it translated into codes and if so, what has that done to the requirement for manual coding?
MS. PROPHET: I'm not familiar with any. Donna, do you have?
MS. GREENBERG: Really, the main example of that, and its not in morbidity, it's in mortality and, of course, the level of clinical detail in ICD-10 is not as great as in the clinical modifications for mortality purposes, you know, you don't need all the clinical nuances and that is the US automated mortality coding system which is used by a number of countries world-wide and actually they have facilitated the implementation of ICD-10 in other countries where the physician writes the cause of death, the causes of death down on the death certificate, just the literals. And then it codes about 80 percent or so.
It doesn't do well at all on external cause of injury information but as I said, we spent well over 20 years developing this system. It is moving the world towards much more standardized, even though everyone was using the classification, it also, we have associated software systems that assign the underlying cause of death according to the rules, the WHO rules, etc. So it's moving the world towards much more standardized mortality data but it's, as I said, it's a considerably less detailed than the clinical modification but we certainly do have that experience.
The thing about the classification is it's not just, you know, you find the name and you assign the code. There are a lot of excludes and include notes and particularly in the clinical environment, it is considerably more complicated as I said. But we have a very good foundation in our automated systems for mortality, the ability to migrate that to morbidity I think is yet to be proven. I don't think there isn't any real example in morbidity.
DR. FITZMAURICE: I guess I wanted to extend my question to the other members of the panel, particularly Nelly, how are hospitals going to cope with this shortage of coders?
MS. LEON-CHISEN: I think that ICD-10-CM probably lends itself a lot better than ICD-9-CM to electronic coding. And you are right. Over time, we are seeing the application of codes with many more clinical settings. A lot more emphasis is being placed in the out patient coding, home health coding and so forth, so we are worried about whether we will one day with able to have enough coders for all the codes that need to be applied and we hope that as advances are made in artificial intelligence and that physicians one day maybe will learn to document using the terms that a computer can understand.
MS. GREENBERG: Artificial intelligence will come first.
What we have found was the automated systems for mortality is that the basic coders, yes, there's less need for them but there's a greater need for really the more advanced coders, exactly what Linda and Sue were saying and for nosologists, people who can modify the systems, who can update the system, who really understand the rules, etc., And, in fact as a result in the whole international community we are working on the training and credentialing worldwide for mortality and morbidity coders. I mean, this is a problem worldwide but it doesn't eliminate the need for professionals who understand classifications, but if you have a good automated system, it does reduce the kind of production aspect.
MS. PROPHET: This is Sue, and in answer to the work force question, I just wanted to note that the whole work force issue, not just for coders, but for the health information management profession in general is a major strategic focus of HIMA right now where we were looking at attracting people into the profession and how to address this work force issue, again, not just in the coding arena but in full health information management.
MS. LEON-CHISEN: I would like to add that for the AHA, the work force shortage is also an issue, not just in health information management but we see this as a shortage across all areas of health care delivery, starting from nurses to respiratory technicians, anybody that is needed to provide the daily care to the patient.
MR. AUGUSTINE: Nelly, this question is for you. I guess it's a given that the move to the ICD-10 will lead to more accurate reimbursement. Now did the AHA feel that that accuracy will lead to more reimbursement as opposed to the same sources with ICD-9?
MS. LEON-CHISEN: No, I mean, we have to be realistic. In a budget neutral suggestion, you are not necessarily going to get more reimbursement for a particular case but it will be more accurate reimbursement for the cases that you are actually treating so you have more detail so it will be more specific and more accurate.
MR. AUGUSTINE: For the same set of services, if it's more accurate, it may be less, it may be more, we don't really know. You are saying we don't know yet. It's a wash out.
MS. LEON-CHISEN: Right, but I think over all, I mean, we realize, especially when we are working with Medicare, it's probably going to be a budget neutral situation so it may be that for some services, you will get less reimbursement and higher reimbursement for others but I think at this point, with a lot of un-specific and vague codes, you are just almost guessing, kind of trying to group cases together.
DR. ZUBELDIA: Nelly, perhaps tongue in cheek Mary said if it's budget neutral, the cost of implementation would come out of the reimbursement for the fist two years.
I have a question for Sue Prophet. You have been talking about something that I think is very important. It's an implementation plan, and one of the recommendations that not only the code we adopted would there be an implementation plan. Could you paint to us what the implementation plan would look like?
MS. PROPHET: Well, what we're envisioning is sort of a strategy to encompass all of the different players in the system. One of the lessons I think we can learn from some of the other countries that we talked to was that some of the key players were more involved from the beginning than others, and then what happened when implementation came to play out is that meant some people were much more prepared than others so I think a strategy that encompasses the different categories of data users, who's going to be affected, what steps really need to be taken first as opposed to later, and we really feel that really a lot of the preparatory steps, as far as readying information systems and even starting to develop educational materials and that kind of thing doesn't have to wait for the publication of the final regulation, that even the original HIPAA rule on electronic transactions warned that information systems should start to build flexibility into their systems in preparation for new code sets after the year 2000.
So we feel that a lot of this implementation work can start to be done very soon.
DR. ZUBELDIA: Will the implementation plan perhaps require the payers to start accepting the goals and put the transition burden on the payers more than the providers or will it be the other way around?
MS. PROPHET: It's going to be both, obviously are going to have to play a major role. It wouldn't be one or the other regarding the payers and the rules, are you talking about the question that came up earlier about the individual payer?
DR. ZUBELDIA: No, what I'm talking about is similar to what's happening with the HIPAA trans actions. If the payers cannot accept this code, providers can't send the code because they do want do get paid so for come time the payers are going to have to start accepting the code and still getting the old ICD-9 codes coming from providers that have not converted yet. So the transition burden seems to be more and the payers have to be ready first and then the providers will become ready once all the payers are ready to start accepting the codes and then the providers can turn off the ICD-9 and only use ICD-10 but there will be some transition time.
MS. PROPHET: But one of our recommendations is that there be a standard effective date for everyone to start accepting and submitting the ICD-10 codes.
MS. LEON-CHISEN: Yes, we would agree with that. I think it would be a lot more burdensome and costly to have to maintain two sets of codes and try to figure out, well, for payer A I submit an ICD-9-CM code. For payer B, they are ready so I can start submitting ICD-10. I think when we switch other, it should be starting as of a specific date and everybody switches over with discharges or services provided as of such and such a date.
MS. PROPHET: Now, basically, they would still have to maintain two sets of codes in their systems for some time because there would be claims and services in previous encounters prior to the implementation of 10 that would still be in 9 but absolutely we would recommend that everybody have to implement on a set, given date.
DR. ZUBELDIA: Big bang approach.
MS. HUMPHREYS: I just wanted to thank Sue for reminding people that I know in this hearing or in other hearings, there are many people, I remember bill saying it, I think I have said it, too, which is it's sort of for anyone what is implementing a system between now and the future, and we were saying this in 1996 and 1997, believe it that eventually you will have to implement ICD-10-CM and, you know, let's get over the sixth digit and alphanumeric business now. I don't think anyone has been under the impression that this wouldn't eventually happen and a lot of systems have been adjusted between now and then.
DR. ZUBELDIA: Let me ask Louise. Is the alphanumeric issue an issue at all?
MS. SMITH: Well, I guess it's the next group that are going to tell you that. We will talk about that in the next session but there will be some system changes. Like there might be screen changes, formats. There are going to be system changes. That's not the major part of the system changes, but it will be a piece of it. It depends on the system.
DR. ZUBELDIA: Because you have D codes and E codes today that are alphanumeric.
MS. SMITH: The six digit --
DR. ZUBELDIA: Yes, six digit.
MR. BLAIR: Is the moral of this store that the NCVHS was really wise when it recommended NCV ten years ago because we're still recommending? Is ICD-10 -- I'm being facetious but obviously, we are still here 10 years later essentially.
DR. FITZMAURICE: I want to follow up with Louise. I'm not sure I got the impression from Louise that McKesson was for or against the move from ICD-9 to ICD-10 based upon the testimony you gave. Are you for a rapid movement to ICD-10 from ICD-9 or not? I didn't get that from your testimony.
MS. SMITH: We are not for a rapid move. We sort of, we are trying to gauge what it would take, and I'll talk about that more in the next testimony, but we are thinking it will take us three years to do it, and we see it as a natural progression and to keep up with health care that the code set needs to change so we do see it as something that needs to happen but we are real concerned about the time frame we have to do it in.
DR. ZUBELDIA: What do you think of that big bang approach?
MS. SMITH: Big bang approach?
DR. ZUBELDIA: Well, everybody switches on the same date.
MS. SMITH: Oh, a drop dead date? Yes, we believe that will have to happen. If it doesn't, there will be a lot of system issues with having to identify a code a haven't had more logic built into the system as to which codes you need to use so I think it would be more of a problem not to have a date where it's a cut-off date.
DR. COHN: Kepa, I don't know the HIPAA standard that doesn't have a drop dead date.
DR. ZUBELDIA: What I'm hearing is not a drop dead date. What I'm hearing is use the ICD-9 up to a day and on that day you switch to using ICD-10.
MS. PROPHET: For services provided after that day.
DR. ZUBELDIA: This would be for diagnosis, right?
MS. PROPHET: Right, so you should be processing claims or something for different dates of service on October 1st and so you would say that with discharges or service date encounters starting with October 1st, those encounters would all have to be coded with 10-CM but if you were submitting a claim for September 30th, you would still use ICD-9-CM.
DR. ZUBELDIA: Is the diagnosis code tied to a date the claim is submitted rather than when the service is performed?
MS. PROPHET: No.
DR. ZUBELDIA: Procedure code is tied to the date when the service is performed.
MS. LEON-CHISEN: Well, it's similar to what we currently do when diagnosis codes change October 1st, and on the hospital in-patient side it's driven by a discharge date and with the changes that are proposed now, for example, if a code becomes invalid October 1, 2002, then you would have to look at the discharge date for your patient and then figure out whether you can use this code or not and it will be kind of similar to that and it would be driven by the date, whatever date you decide on.
DR. ZUBELDIA: In the past, and as far as I know, even in the HIPAA final rule, the procedure code is tied to a date the service was rendered whereas the diagnosis code is tied to a date the claim is filed.
MS. SMITH: No. Both.
DR. ZUBELDIA: Let's take a 15-minute break and we will reconvene the next panel.
(Brief recess.)
DR. ZUBELDIA: Let's start again with the second panel. This time it's going to be the vendor perspective on the issues related to a transition between ICD-9 diagnosis and ICD-10 diagnosis. We'll start from Elizabeth Johnson and again move to the right. Please introduce yourselves when you are starting your testimony.
Agenda Item: Elizabeth Johnson, Tennant Health Systems
MS. JOHNSON: Good afternoon. I'm Elizabeth Johnson. I'm the Vice President for Clinical Information for Tennant Health System. I suppose I would offer a slightly different than vendor perspective. I would offer what the providers are thinking about this change, not from the perspective of examining the difference between ICD-9 and ICD-10 but instead looking at the operationalization of that and what it's going to take.
So there's five components I'll discuss with you and be happy to entertain your questions although we want to go through the vendors and then I'll entertain questions at that time. Either way is fine.
In essence, we really believe there's five things we need to look at -- process, technology, how it impacts other systems, training, and the perception of the HIM departments in our facilities.
First of all, in terms of process, and I think this will be good news for you, I hope it will be. This is one time when we are not here to complain that you are about to ask us to do something that is going to radically change health care and therefore throw us into another tizzy. I know that comes as a surprise to you considering some of the other testimonies I h ave heard other issues in this very room but it's not significantly different from what we do today, and that is a reality.
As you well know, ICD-9 has been around since the early 1970's. We added CM in 1979. We've had to modify it every year anyway. We are out of codes. So this isn't a hard decision for us to support kind of going forward. How you deal with CPT, we have some interest in, but that's not the topic of discussion today.
One of the things it will do is clearly give us a significant advantage in really being able to define the procedure or diagnosis that took place. We are not able to do that today as you well know because of the increase in technologies and procedures that have changed and the fact that we are out of codes, we cant adequately describe what we bill for, and obviously there is some liability in the process of doing that.
Another thing that we often get asked is with the changes that are going on with HIPAA that obviously have an impact to HIM, what is the impact going to be there?
The reality for you that don't work on the operations side is it's not the same people. Generally speaking, the people that process records and deal with release of records are a different set of people that handle coding. It's a different function within the medical records department so those people are different people and will not be impacted by the same things.
Now, I'm speaking from an organization that represents thousands of entities, but 116 provider organizations, so I'm speaking from the very small organizations to the very large. That gives you some idea of where I'm coming from. So we really believe that it will not have a significant impact to the processes within HIM.
There is a different process impact that we have some concern about and we'll talk about that in this area as well as with technology, and that is the data that is collected via the diagnosis and procedure codes issues for other functions within the hospital. We'll talk about those specifically. The question will be can the other areas handle the new length of code.
You probably heard by now the code is longer so it's more than just looking at the medical record system and the way they code, whether they do it manually or electronically. At Tennant, we do it electronically almost exclusively. However, that same information is used for credentialling, it is used in case management, it is used in disease management, it is used in cost accounting. So if those systems are not capable of handling a new data format, then we have a difficulty to overcome. I don't believe that's been examined by the industry, but I'm sure the vendors here will be able to talk about that.
The other thing is that we will need to look at those institutions that do manual coding. It will be a more laborious effort for them than those who use electronic coding or computer coding. The reason is that in the past we've often put in changes to our coding system. It's an update. It takes minimal or remedial education. It does not take a great deal. Obviously, if you are using a manual system, therefore you are using a code book, this is going to be a little different process for you.
In terms of going forward on the process, I think it's critical to remember that you develop a transition plan for old and new data. I heard the very tail end of the discussion that you had last panel. One of the things you will have to decide is, for example, when you collect the credentialing information very specifically, how many gall bladders did the general surgeon do? Your codes are going to change. You capture that information electronically now. You will have to build bridges because you will have to capture both sets of information in order to get a complete set of data. That is not new to us. It happens every year. I don't want to make that an alarmist kind of comment because ICD-9 changes every year; DRG changes every year. This is not new, but it is something you will have to take into account.
And obviously I will continue to say to you five versus six digits is an issue. So that kind of leads into technology. This is my challenge to the vendor population, not my comrade here on the provider side.
What we are expecting is that that will become part of the annual license fee. We pay a new annual license fee every year, and that is for updates in coding. That is what we're expecting. It is unreasonable to expect if they truly have to do major software changes and have to increase filled links that are not here and viable today or they have to negotiate. That is the position that we are going to take going forward and our major vendor is here to speak with you today.
So, again, you have to look at is that field link available, not only in HIM but in the supporting systems.
The next issue I would like to talk about is training. We believe this is minor. Again, we have been doing this for years and years. We generally can put a change like this in even across an entity the size of the one I represent in two to three months. It really does not take longer than that unless you are going to do something radical that we don't know about, so I went back before I came here to testify and talked with the people that do this work across the Tennant system, and they felt like that was more than adequate to get this change in place.
Then I wanted to talk about their perception. I think it's important that not only do we represent what we don't believe to be significant operational changes, but what is the attitude of the part of the organizations that's going to be impacted and I was delightfully surprised to find it as very, very positive. They are ready for this change. They have been anticipating it. We began talking at tenant about HIPAA in the very early part of 2000 and one of the things we told folks then was anticipate a change in coding systems. It is legislated that they must look at it. We anticipate that change could be coming, and so the preparatory work became evident at that point and went into place so we believe that it's positive, it will have minimal impact, it'll be very positive and other than the problems around vendors not being able to handle the digit change, it will work for us.
And then the last thing I wanted to say was, because I want to keep it short and sweet. Those who know me know I don't talk for 20 minutes, we get to the point here.
We really believe the transition planning is going to be critical. Again, the one thing we have seen the provider organizations do is go into a denial state about changes that are coming and not prepare adequately. That won't be our approach to this. We are prepared now that you are going to make this change and we have already begun to talk with our vendors about how the change is going to be handled. We've already looked across and done an inventory of our software so we know exactly what we have to make changes in and we have not limited that inventory to medical records but to all decision support systems as well.
So, as you go forward and make recommendations, I would encourage you to have others do the same and that's where Tennant Health Care stands on this situation.
Agenda Item: Chris Riopelle, Gambro, Inc.
MR. RIOPELLE: Good afternoon. I'm Chris Riopelle, Chief Privacy Officer, Gambro, Inc., parent company for Gambro Health Care. We are a large dialysis provider here in the US and internationally. Gambro provides hemodialysis, perineal dialysis and related services to more than 40,000 patients here in the US. We deliver our services in 32 states and the District of Columbia, through approximately 530 dialysis facilities and the 12,000 employee members of our patient care team. Our number one objective is delivering the best patient care possible.
I'm here today to provide some input and information relative to the impact of code set changes on a large provider who utilizes proprietary information systems.
I'll begin by giving a brief background on our information system. I'll then describe some of the impact of the transaction standards portion of the HIPAA effort, and finally I'll address some additional considerations relative to code set changes in general when you are dealing in a proprietary environment.
Unlike many of the organizations who have testified here today and at previous hearings, we are not made up of thousands of constituents. We don't have a membership that spans the universe or the US. We are simply a big provider who cares for chronically ill people. Although we don't represent other companies here today, I think you'll find that the issues we present and the challenges we present relative to a proprietary systems user may well be the same challenges that other proprietary system users face.
Let's talk about our IT environment first. We capture data on and submit claims for over 6 million patient treatments per year. The systems that capture data and submit the claims were developed internally over the last 10 years. Gambro made the decision to make instead of buy many years ago, because at the time of the decision, there really wasn't a commercially available product that suited our needs in order for us to run our business.
For the purposes of code set changes in HIPAA, we organized our systems activities around two main platforms, renal information management system or RIMS as I'll refer to it in this discussion and the billing and accounts receivable system which we call, affectionately, BAR.
The RIMS system captures clinical data, treatment level information, including the coding input, and RIMS is a key component in generating our reports of clinical activity. BAR system extracts appropriate clinical data from RIMS in order to create and submit claims to payers and, similar to RIMS, the BAR system serves as a fundamental basis for running all the necessary finance and administrative reports on our operations.
In terms of HIPAA activity and impact, like everyone in the industry, it's been a big challenge for us. Over the last two years, Gambro has been working diligently to respond to the changes that HIPAA demands. We are currently in the midst of an effort to respond both to the privacy requirements as well as the transactions components of HIPAA and, like many others in the industry we waited with great anticipation for the final, final, final security regulation to come out some time, I'm not sure when now, but it's soon I hope.
The effort for us is significant, like it is for most other companies. Major steps -- this is rocket science -- plan, assist, implement, test and go live from an IT standpoint. The planning and assessment phase is comprised several thousand hours of internal and external professional time, a significant six-figure total cost.
The implementation process for the transaction standards will involve an estimated 30,000 hours of professional time and hardware purchases totalling a mid-seven figure cost. To some gigantic organizations maybe this is a small amount of money, but for us it represents in one year more than 15 percent of our IT budget alone and that's just focussing on the IT challenges, not to say anything about the cost related to privacy and when it comes out, security.
A representative project in the implementation process for trans actions is the transition away from local code. The transition away from local codes to standard code sets, as required by HIPAA will be significant for us to under take.
It's a bit unique for us because the incidence with a CE of dialysis or renal failure in the patient population is such that we may have a payer relationship with four or five or six patients or covered lives with that payer. As such, you might imagine, we have over 13,000 different potential payer relationships in our system and over 23,000 local codes that represent how we are able to effectively bill and collect for our services.
Again, changing from local to standard is something we welcome, we support, and we think it will be great as long as those payers are in line with us but doing it, having 23,000 iterations of that task -- call to payer, change the code, test the code, send the claim -- is a tremendous undertaking for anyone whether you are a vendor who is selling systems or more importantly when you are a proprietary user, it's a tremendous undertaking for us.
The implementation process impacts many other areas within IT -- training, reports, data warehousing, planning, testing, electronic medical records, billing documentation, translators and general infrastructure -- just to name a few.
The specific areas of impact that need to be modified go beyond just IT. The transactions compliance effort requires a significant re-engineering of business processes as well. For example, we run on a regular basis over 3,000 reports out of our BAR and RIM systems. To the extent the content of those reports change; to the extent the process of running that report changed; to the extent the interpretation and the actions from those interpretations changed, those all are things we are going to have to train and educate our people on within and outside of IT.
A final consideration in the area of HIPAA impact is the timing and interconnection of each of those three subject matter areas, privacy, security and transactions as we call them within Gambro. Although the discussion today really is to focus on code set changes, privacy and security will have some system impacts for us as well. Managing the different compliance dates and their attendant systems impact is challenging, especially when the regulation has not been interpreted by those in the industry and final compliance dates and the requirements sometimes change.
It would be valuable to solicit, I know you have on many levels, input from affected organizations an how to effectively time the system modifications for all large scale changes. Some providers with their appropriate resources might say bring them all to me at once. We are going to engage a giant team of people and we are going to change our systems to comply with privacy, security and transactions. Others perhaps may identify that a phased-in approach would be more valuable.
I would urge the committee to consider all points of view in evaluating recommendations on the timing of things that are so interconnected as this.
Let's talk about some other non-HIPAA specific code set considerations. One of the efforts that is worth noting and that, a requirement that has been mentioned today, supporting old formats while adapting to new formats. Retrospective review of activity is valuable in any business setting. It's valuable to us both clinically and from a compliance standpoint. So the ability to run retrospective reports when there's a change on a large scale proportion requires us perhaps in some cases to operate two systems for a longer period of time at a minimum we have to have the data warehousing capacity to last for a number of years.
Another consideration for us is compliance because most of our patients, we all know ESRD, ultimately become the Medicare benefit beneficiary. Our most significant payer as a result is Medicare. Following necessary billing and documentation requirements for billing Medicare is a critical skill for our organization and anybody who has Medicare as their biggest customer.
To the extent coding and related documentation requirements change, retraining our work force to continue to submit accurate, compliant claims is a big undertaking with potentially severe penalties and other consequences. Medicare billing, as we all know, is complicated. Many folks have spent many years getting it down to a science to the extent code change documentation changes, narrative change, we are going to have to re-adjust how we do it, and it's important because we want to get paid accurately, but it's far more important when you consider the severe civil and criminal penalties associated with sending in mistaken claims.
Additional consideration in evaluating the impact of code set change is how different types of organizations are impacted by the changes. I would urge the subcommittee to consider the differences between a commercial software vendor and a proprietary user as it determines guidance.
As a proprietary user, there are a limited number of people adopting to new legislative changes or system enhancements. Our IT talent must necessarily spend its time responding to ongoing business issues and as a result, proprietary users like Gambro are forced to react to legislative change instead of being, as they would like to be in a world of unlimited resources, more proactively searching out potential changes.
A commercial vendor in the business of providing latest and greatest applications for a fee has a significant financial incentive to be proactive, to look for those changes in legislation coming up. They can afford to have the human capital available to study new changes and develop new products in response.
It's important to acknowledge another difference between a commercial vendor and a proprietary user and that's bearing the cost of change. A commercial vendor might be obligated, depending on the depth of a change to update through a licensing agreement for a fee in an economic price and exchange, a new version of its process. The costs of these changes are born by hundreds, perhaps thousands, of customers. Proprietary users like Gambro bears the complete cost of a change of this nature alone. We have to bear the cost in order to stay in business and we don't necessarily recognize any new revenue for making that change. It's a true cost to us, not necessarily an opportunity.
We heard some testimony earlier about the budget neutral impact of more effective reimbursement. In essence if that is the case, then it is a significant net cost to us to comply with a change of this nature.
Let me conclude. In addition to providing some examples of the challenges and difficulties of changes code sets in a proprietary environment, a key objective today was to provide some concrete recommendations to the subcommittee on how to make code set changes for effective for the proprietary user.
If there was one key driver that I could leave with you folks today is that of lead time. We all need lead time to get something as big as HIPAA or other significant code set changes done, not to say we don't adapt every year to updates in ICD-9, but what we're talking about, I think, with HIPAA in general, and some of the ICD-10 activities, is a little broader range of change.
Because of its significant budget and business resource impact on a proprietary user, we need three years at least in our environment to effectively prepare, budget, plan, and execute a change of this magnitude. The lead time needs to begin when the change is final, after it's been tested, comprehensively endorsed by the key constituents who are impacted by it, and, if at all possible, fundings identified to support the cost of such change.
We cannot afford to be in a constant state of legislative change in our health care industry. We acknowledge that we are a changing environment. We are in a dynamic business and that's part of the reason we are all in it, but at some point we need a little time to improve what we have instead of just changing it.
Final comprehensive changes need to be released with detailed action plans supporting them. New changes must be tested and endorsed by all key constituents. We can drive change in the health care business, and I think those in the business out there taking care of patients and running their companies will tell you it's much easier to pull change that have it pushed, and I think it's great that there's been lots of public comment from the industry about changes because that's going to help this get pulled as opposed to being pushed.
Additional cost, reimbursement levels will go up occasionally. They don't typically increase at a rate adequate to support a wide scale IT change. We would encourage any evaluation of a funding course to help underwrite the cost of significant changes in the future.
In order to accomplish the objectives above and thereby assure more effective code set changes, we would support the suggestion made by other organizations at previous hearings that a multi-disciplinary team be established which includes proprietary users to evaluate all of the impacts of moving from one code set to another.
I recognize some of the issues today that I have presented are not unique to you folks. You have been dealing with these policy issues much longer than most of us in the business. My remarks today haven't been technical. There's been enough people here, and I'm sure to follow who will present more technical information about the impact, specifically in the case of ICD-10 much more effectively than I could.
However, I do hope that the information has shed some additional light on the specific challenges that HIPAA and other code set changes create for those proprietary users who are trying to bear the cost of change while continuing to deliver high quality care to its patients.
In closing, let me be very, very clear. We are absolutely supportive of changes in coding that will provide for more effective capture of clinical data so we can care for our patients more effectively and more effective reimbursement. It's critical, however, that such changes are required only after the appropriate planning, testing, and lead time have been thoroughly evaluated with special emphasis on those proprietary users in the health care industry.
Thanks for having me here today.
Agenda Item: Bob Hardesty, Sarner Corporation
MR. HARDESTY: Good afternoon. My name is Bob Hardesty, and I'm a Product Manager for Patient Accounting with Sarner Corporation and have particular responsibility for our compliance effort surrounding transactions and code sets. Sarner Corporation is a leading supplier of clinical management information and knowledge systems with more than 1,500 health care organizations worldwide as clients, and as such has a great deal of interest in the subject of today's hearings in regard to the possible effects both on our own future planning from a development perspective as well as on behalf of our clients, both current and future, and we welcome the opportunity to participate.
In listening to the testimony today, it's obvious that after more than 20 years, the useful life of the ICD-9 is coming to an end. At least to me, the question becomes not if ICD-9 should be replaced with ICD-10, but when. And more importantly, both for the industry and particularly the software vendors, is the further question, I think, of how.
For many in the industry the initial reaction, I think, to converting from ICD-9 to ICD-10 is really one of incredulity. We are already as an industry struggling to achieve compliance with changing requirements for transactions and code sets, with privacy and with anticipated rules governing identifiers and security.
How, then, can the industry also absorb the sweeping change that would be required by the conversion from ICD-9 to ICD-10? The answer as of today, I think, is that we couldn't. Not within the next two to three months.
However, given the inevitability of the eventual change, we feel first that this hearing is very valuable from a standpoint of alerting the industry to the whole subject, and we feel that the migration from ICD-9 to ICD-10 is both necessary and inevitable and that the sooner that the discussion, the planning for the eventual event can begin in earnest, the better off the industry can be in developing the necessary plan for implementation.
As far as the question of when, as the committee is well aware, the initial deadline for compliance with transactions and code sets in October of this year. The privacy rule is April of 2003, and with the extension for transaction of a new code set, October of 2003. In addition the anticipated rules governing security, identifiers and additional transactions are expected soon and my own assumption would be that deadlines for those would be sometime also in 2003. Perhaps it's earlier, perhaps it's somewhat later than that.
It's difficult to believe that there won't be fallout and difficulties with these efforts and that's particularly true concerning the all-important issue of reimbursement. If we can believe the compliance with the major components of HIPAA would be achieved by the end of calendar year 2003, then it would seem to us that it would be a good idea to have a moratorium on major changes during calendar year 2004. That would allow the industry some time to assimilate the necessary changes and to mitigate any corresponding difficulties that may arise and in particular what I'm referring to there is within the electronic transactions, changes to claim submission and remittance and any types of delays in reimbursement and glitches overall between payers and providers that are caused by that type of change.
In regard to an actual implementation date for ICD-10, we feel that there are several factors that would need to be considered.
First, it would be beneficial for all of us in the industry to know that the transition to ICD-10 is, A, definitely going to happen and, B, is going to happen by a particular date regardless of when that may be.
Second, the implementation of ICD-10 should be carried out in tandem with the implementation of ICD-10-PCS and any other contemplated changes to coding classification systems and procedures.
Given these factors we feel it would be helpful for the committee to urge the Secretary to issue a guide answer on the subject that would alert the industry to the department's overall plans and directions and we would then support a three year time frame for conversion after the implementation of the major components of HIPAA.
From the standpoint of particular implementation issues, there are several general issues that depending on the software and the data architecture will confront software vendors in general in migrating their systems to the use of ICD-10-CM and I will parenthetically note here that with our, with Sarner's millennium product, many of these are not issues but they will be issues both with proprietary systems and with legacy systems throughout the industry.
First is the stored diagnosis element link. For some systems, this may present the most significant remediation issue. Diagnosis information is stored in a variety of application product areas throughout a hospital information. Some of these have been mention but they would include scheduling, registration, medical records, patient accounting, utilization reviews, decision support and others. If the data base needs to be modified in order to accommodate ICD-10 it would require a significant conversion effort in order to accommodate the necessary changes. This would require significant planning and testing as well as down time during actual implementation.
The just plain reported diagnosis element link, regardless of the data base structure, many systems that don't have the issue of data base remediation may still have an issue with the display and/or reporting diagnosis. This will require analysis by vendors of all the affected system areas, along with a corresponding software modification. Most of these changes would be relatively minor in nature. However, on particularly crowded displays or reports, a significant effort of redesign may be necessary.
Next is memory storage for diagnosis element links. All the vendors will need to analyze their internal storage of diagnosis. Even for those without issues within their own systems, they may likely be using third party software for grouping, pricing, etc., that use areas for internal storage that may need to be modify, or correspondingly tested and remediated.
And finally, it would be ICD-9, ICD-10 crosswalk. A significant area for all of us, I think, would be the probable necessity to incorporate a crosswalk of ICD-9 to ICD-10. The application and usage of the crosswalk would require significant analysis but would undoubtedly be necessary during a transition period.
The good news with the modifications that are outlined above is that there's no apparent reason why many of these sorts of changes couldn't be implemented over time using the existing ICD-9 rather than necessarily having to wait for the implementation of ICD-10. That would allow for gradual transition to compliance rather than necessitating a big bang from a conversion standpoint.
Of even greater concern from an implementation standpoint than each individual vendor's own system modification is a coordination of efforts between the vendors. Much as we may wish to be the sole vendor for a particular institution, the reality is that it's a rarity for that to be the case.
The greater challenge for any institution will be the coordination of time lines between multiple vendors, the testing of both internal and external interfaces and the further coordination of any necessary conversion outputs.
Even after all of the above has been achieved, then what about the payer? Even with an absolute date for compliance, there will undoubtedly be varying degrees of readiness and coordinating testing with payers will be a challenge and I would note that is the same challenge we are facing now with the implementation of transactions. Who's ready to accept an 837, who's not, so on and so forth. There is a significant challenge at the moment.
Given the direct correlation between diagnosis information and reimbursement there needs to be study and published guidelines that would provide a road map implementation. An additional issue that should be considered in that regard will involve managed care contracts. They are based on diagnosis that may require re-negotiation.
One of the other panelists mentioned license fees, other costs. In general, we do not charge additional software license fees for the design development of software modifications related to government regulatory change. This is covered under our ongoing agreements for maintenance support and is believed to be a fairly common approach among the vendor community.
However, given the broad impact of the change of this nature it would be expected that there would be professional service fees involved for activities undertaken involving planning, testing, training, conversion and implementation.
The amount of those costs would necessarily vary widely, depending on the size of the institution, the number of vendors involved internally and the overall number of external trading partners. Additionally, it would be expected that a cottage industry would probably arise very similar to Y-2K that was devoted to the coordination of efforts necessary for implementation.
From a testing standpoint, any change of this magnitude would require multiple layers of testing and certification efforts. Vendors would be expected to unit test each affected application product area and system test their entire product offering. In addition, testing would be required for any necessary conversion programs and for all supported interfaces affected by the change.
Vendors would also be required to test third party software that they utilize as components for their overall product offers.
On-site testing would then be required to insure vendor to vendor interface integrity and with each separate external trading partner. One thing that I should probably add to that is that given that level of testing, there may also be hardware considerations in that it would probably be necessary to create testing environment for a much broader scope of area than may be within a particular institution hardware configuration.
In conclusion, I would like to offer the following points for consideration by the committee.
First, as mentioned previously, I would like the committee to urge the Secretary to issue guidance to the industry regarding the intent to migrate from ICD-9 to ICD-10. This would be at least the "lay down the gauntlet" announcement that this is actually going to happen, and you need to start planning because honestly I don't believe that any significant planning is going to happen until there's something that looks like a date.
Second, we would urge the committee to call on organizations such as ASAK(?) To help coordinate the efforts of the vendor community. I think this would be very significant from the standpoint of trying the insure the cooperation of vendors because I truly consider that to be a larger issue than, the cooperation to be a larger issue than what happens to the individual vendor system by itself.
And last, we would urge committee to call on WEDI Smith to organize a work group to address the various issues involved from an implementation standpoint.
I would like to thank you for the opportunity of providing comments, and I would certainly be open to questions. Thank you.
Agenda Item: Louise Smith, McKesson Corporation
MR. SMITH: Good afternoon. My name is Louise Smith from McKesson Corporation. A detailed copy of our testimony is available so in the interest of time I would like to highlight some of the main points of the written testimony.
As stated earlier, from an information systems perspective our interest rests mostly in the volume of changes that will need to occur to software applications and supporting documentation. For that reason I would like to describe some of those changes in order to illustrate the importance of the timing of the implementation of ICD-10-CM.
Software applications that capture, store, maintain, receive, send or edit against diagnosis will require modifications in every location where a diagnosis code exists. Hospitals today use software applications from several vendors therefore the transition to 10-CM will necessitate modifications to the product applications into the interfaces between the various internal and foreign systems.
For example, if systems that share diagnosis codes between applications are not in synch, there will be data integrity issues with the diagnosis data. Hospital information systems often use third party grouper and encoder systems. Generally, there is a sequence in which software vendors receive file layouts, specifications and groupers, make system changes, test the changes in the software and in the interface.
Since the final version of 10-CM hasn't been released yet, it is likely that software vendors and the developers of groupers and encoders will be designing and programming software changes at the same time. While some overlap is expected, the grouper and encoder systems need to be completed and released in advance to allow smoother development cycles for providers and payers.
The state reporting files such as cost containment, discharge reporting and other state-specific files that utilize diagnosis data or the output of such codes will require modifications.
In payment systems, diagnosis codes often serve as the foundation to other applications. For this reason, system modifications in addition to field expansion and field redefinition will include revisions to the logic for fee schedule calculations, authorization screens, DRG assignments, pricing screens, data abstracts, auditing functionality and statistical analysis.
Decision support systems contain data bases for modeling purposes that will need to address the issue of current versus historical data. Once the transition to 10-CM occurs, there will be a point where part of the population is defined under 9-CM code set and the other part is defined under the 10-CM code set. In order for decision support systems to use the data for modeling or analysis, system logic will have to change, especially in defining when to invoke mapping programs.
Dual support is another important consideration. Systems will have to support dual code sets during the transition partly due to reporting purposes or the need to validate older claims. The maintenance and use of dual coding systems could impact programming and testing time. For example, systems generally maintain ICD-9-CM codes by effective date and by payer. Factoring in ICD-10-CM, potentially doubles the amount of codes that will need to be maintained.
Clearinghouses which translate non-standard claim information into standardized formats support a variety of input formats into multiple forms that are executed as the claim is processed through a clearinghouse. The formats and programs that process those formats will require modifications to the field length and field type. That field change will need to be passed onto other modules within applications.
Clearinghouses often create audit trails by means of reports to validate that all the data is accounted for. The field changes potentially impact the need to redesign those reports.
The practice management systems used for claim processing purposes that already accommodate a six character alphanumeric code, the modifications will be minimal, partly due to the fact that many practices use limited range of diagnosis codes for billing purposes.
The timing of the ICD-10-CM implementation is critical with regard to HIPAA. Currently, information system vendors are focused on meeting the compliance requirements of transactions and code sets as defined in HIPAA. There are many resources focused on the first round of HIPAA at this point in time.
It would be desirable for an implementation date of at least three years after the initial compliance date of HIPAA transactions and code sets. This would allow the industry time to design, code and test software changes and implement such system upgrades into payer and provider sites. In addition, a drop dead date and detailed guidelines for implementation for ICD-10-CM will be critical.
There are other considerations that are important to reference. As an information systems solution provider, the preference is to implement ICD-10-CM and PCS concurrently rather than at different times. This avoids the burden of maintaining version control between 9-CM procedures and 10-CM diagnosis. Additionally, the areas in the software that will need to be modified for diagnosis and procedures are the same.
Also, if ICD-10-CM is used for diagnosis and ICD-9-CM is used for procedures, DRG calculations will require grouper logic which would recognize both codes that, and define parameters to determine which code sets should be used. With grouper changes generally made once a year, mixing the two code sets adds further complexity to the system functionality, design, coding and provider usage. A mapping between the two code sets will be required, and it would be desirable to have a mapping that goes both forward and backwards.
Our experience in Canada has been that of building tables and setting parameters for users to convert to ICD-10-CA, the Canadian version. However, in Canada, the actual code conversion occurs in abstracting applications and 10-CA is not used for billing purposes. In Australia, we were required to modify tables for alphanumeric codes in expanded fields as well as storage of multiple codes for patients.
In Canada and Australia, the scope of the changes were less complicated than the US, partly due to the types of system applications we had in those countries and we didn't need to deal with groupers or DRGs.
With regard to training and resources, training in ICD-10-CM will be required for software developers and programmers to understand the new system and to facilitate applicable coding changes. Obviously the resources needed to make those changes will vary, depending on the type of system and the extent of the changes. However, we have estimated between 24 man months and 350 man months. On the last page of the testimony we have a chart that reflects the criteria for each estimated category.
On the matter of maintenance of ICD-10, the code sets, it is our opinion that public maintenance of the 10-CM and PCS code set would better serve the vendor, provider and payer industry since the existing process for ICD-9 works well. We would also like to suggest that a group be identified within the health and human services agency that could respond to questions as software developers are developing functional specifications for the required changes.
In summary, we believe the migration to ICD-10-CM will require a great deal of synchronization. The timing should be coordinated such that it is three years after the initial HIPAA implementation date. It would be better to implement both CM and PCS at the same time rather than at separate times. Public maintenance would best serve the industry and crosswalks between the two codes that would support the transition to that process.
On behalf of McKesson Information Solutions, I would like to thank the committee for this opportunity to comment. Pat Hanby, the Vice President of Regulatory Affairs has accompanied me here today and we would be happy to respond to any questions.
Agenda Item: Casey Mitchell, 3-M Health Information Systems
MR. MITCHELL: My name is Casey Mitchell. I'm with 3-M Health Information Systems and work in the Clinical and Economic Research Department.
I will be brief. Our testimony will focus on the implementation issues from a vendor perspective, potential impact to customers and their contracts and certain transition issues that we feel will be important to consider. Our written testimony with all the rest of the details is available and has been distributed.
As we've heard, the current code sets have been embedded into virtually every portion of the health care systems. It is in our vocabulary, our software and our processes. We use the code sets for a number of purposes including clinical edits, care pathways, compliance, the list is rather long as to what we are now using these code sets to accomplish. Logic data bases and software have been developed to support processes to leverage the clinical relationships and to provide efficiencies and specific solutions to the industry.
Changing to a new code set will mean that every system that we currently have in place will need to be reviewed. Changes to screens, algorithms and prompts will need to be considered. In short, all aspects of the current systems will need to be reviewed and potentially updated to reflect the impact of the new code sets.
Of particular relevance is the issue of electronic interchange of data and interfaces which is a core focus of HIPAA. In order to implement the new code sets, all interfaces between systems that transmit coded patient data must be reviewed to assess the impact of the new code sets and as needed, changed.
I don't know if anybody has commented on how many there are, but it's not three or four interfaces that make a system work. There's quite a few. Interfaces consist of two main components, the first being a communication method and the second being a data format. Code sets will have little, if any, effect on the communication methods. However, in the area of data formats, it's likely that they will change to support the new characteristics of the codes being sent.
Some examples of the data formats are, as you know, X-12, HL-7, XML, etc. Depending upon the changes needed to implement the new code sets, and how the new code sets differ physically from the current code sets will determine which data formats must change and the scope of those changes. We've heard of five versus six in testimony or four versus seven or whatever the differences might be in length, in alphanumeric versus all numeric.
For instance, current code procedure code sets contain numeric characters only, and new procedure code sets will likely contain alphanumeric characters and longer codes to increase specificity and support future expansion. These changes would require adjustment to any interface that depends upon type or length of field for accurate transmission of data and, as we heard from other speakers, data base changes would also be required to take care of the new format.
Use of industry data formats can simplify the implementation process and 3-M encourages and supports the use of standard data formats such as X-12 and HL-7.
Changing code sets will be an industry-wide challenge, but we believe a doable task. Every vendor that elects to continue to support the new code sets will need to systematically review every facet of their systems to determine where dependencies upon the current code sets exist. Any necessary changes must be designed, developed and installed within customer systems.
There seems to be a consensus on the limitation of ICD-9 and the need to move to a new code set. I'm not going to go into any detail in that.
While the effort to implement the changes will be substantial, the benefits derived from improved coding sets will outweigh the cost. The sooner decisions are made in direction and time line set, the better prepared provider and vendors will be and the sooner medical providers, consumers, and insurers will benefit.
On the impact to our customers, they expect us to solve business problems through software and consulting that address fundamental issues in the business and delivery of health care. These issues exist independent of the current code sets and significant effort has been expended to use the current code sets to support these fundamental issues. For instance, the process for assigning DRGs using diagnosis and procedure codes were first designed by Dr. Fedder. At Yale, the objective of improving the tools available for the administration of acute care hospitals in the area of analysis and process control, a need that continues to exist today.
DRGs were later used by CMS and by some of the members of the panel to address another fundamental aspect of health care payment. These two activities have been replicated multiple times around the world using different code sets and algorithms than those originally used by Dr. Fedder and CMR. These are fundamental issues our customers expect us to solve.
While the new code sets will require extensive changes to systems, the reasons customers purchase our solution have not changed. Encoding, editing for accuracy, training and education and process improvement will still be required. So unless the business problems significantly change due to the new code sets, our practice has been to update and upgrade our solutions within existing customer support contracts.
Change does bring about opportunity. So while existing software and consulting solutions will continue to be updated and enhanced to support the new code sets, we anticipate new opportunities and services will be needed by the market to fully implement any new code set.
One opportunity that may require multiple solutions for the market will be in the support of historic code sets. This issue will be important for turning of data over time, use of public data sets, and provider and hospital comparisons.
How the issue will be resolved is yet to be determine as until the new code sets are identified and documented, the implications to the market cannot be fully determined. 3-M HIS expects to survey the market and our customers and our vendor partners and to continue to be the preferred supplier for health care solutions in this area.
As stated earlier, it is 3-M HIS's opinion that the benefits derived from improved code sets outweigh the cost of implementation. We encourage the committee to set a clear direction and time line for change to allow providers and vendors the opportunity the effectively prepare for the change. The sooner the new code sets are in place, the sooner medical providers, consumers and insurers will benefit.
The linkage of HIPAA transaction formats and code sets seems to be hand in glove the format of the new code sets. There are other portions of HIPAA, however, such as privacy and security that will potentially require significant investment and infrastructure changes and process changes. Changes that tend to be time consuming and costly. While there is a definite advantage to minimizing the number of cycles of change that will be required of providers and therefore bundling all the changes up into one big event, we recommend that the committee be vigilant to the risks of combining the code set issue with other HIPAA issues that require significant changes to infrastructure and processes.
If combined, a delay in infrastructure development and/or process implementation would result in a corresponding delay in implementation of new code sets which would result in a delay to medical providers, consumers, and insurers of the benefits of the new code sets.
We would recommend that notice be provided at least one year in advance of implementation and that two years' lead time is preferred. We know this is well within the current rules. Rules have a way of changing. This is the way of saying that this is what we would prefer to have.
At the time of notice, all details related to new codes that should be publicly available and fully documented. Changes to the new code sets prior to implementation should be kept to the barest minimum. We are already going to have enough to do reacting to the code sets without having to react to version two and version three and version four.
This advance notice will allow thorough review of existing systems, collaboration with vendor partners, on embedded components and interfaces and will provide ample lead time for customer implementation, testing and certification. Less than one year is just too quick to coordinate the necessary activities.
This recommended lead time has been sufficient for other implementations of clinical modification code sets elsewhere in the world, but Canada and Australia recently implemented CM code sets and provided vendors roughly two years of advance notice. With final modifications being made in the year prior in implementation, sometimes as short as three or four or five months before final implementation to change to the final rules, were provided to the vendors and they had to react.
Even with the late changes, vendors on the whole successfully updated products and providers made the necessary changes and upgrades to the systems in order to comply with government regulations. These experiences indicate there was sufficient clarity and notice the necessary system changes can be made in a timely manner, but that full notice and a non-changing specification are a key to getting it done smoothly.
It is worth noting however, that the financial consequences of converting to a new code set in a foreign country may not be as great as in the US as a result of our prospective payment system.
In close, 3-M HIS would like to express its support of this committee and looks forward to its recommendations. 3-M HIS remains committed to makes the transition to new code sets as efficient, effective and trouble free for the industry as possible. If there is any additional information that the committee would like 3-M to provide, please to not hesitate to contact us, and I have been, I have Far Steinbeck and Phil McGrant here as well from 3-M that if there's any questions that any of us can answer an behalf of 3-M, we would be glad to answer.
Thank you very much for this opportunity to express our views.
DR. ZUBELDIA: Thank you. This is a slightly different perspective from what we heard earlier.
I would like to start with a question to the entire panel. It seems like everybody is saying we need about three years. Everybody is kind of converging on maybe October of 2005. Is that the date? Is 2005 the date? Is 2006 the date?
DR. COHN: I had the same question.
DR. ZUBELDIA: What's your recommendation?
DR. COHN: Can I set up because that's exactly the same question I was going to ask but I was going to add to this since you all mentioned the other things happening in HIPAA. I mean, let's assume that maybe security comes out before the end of this year for implementation in the end of 2004, provider ID comes out for completion, that's implementation also. So that would be done by the end of 2004. That pretty much, except for fringe attachments, is pretty much the full sweep so given that additional information as a hypothesis, when should this happen?
MS. JOHNSON: I'll talk about it first. I think from our perspective, first of all, our anticipated date those things is the same. We have gone ahead and I know this is risky but what we've done is take the security regulations. We've already written the policies and procedures and put them in place. We've already examined all applications and security functionality, including the data bases, data warehouses, our data centers so we recognize there probably will be some changes. We don't expect many so we will be ready so our anticipation, we've also looked at all of our applications for ID, ability, meeting, we've looked at all the field links to determine whether or not we can go ahead and house additional IDs going forward.
My sympathy to the vendors, and this is quite honest. We've really choked down on them, and I'm in a different situation than Chris because right now we're managing 950 vendors to HIPPA regulations. We would be managing a much smaller group for this change and they have been very, very responsive to us. They set reasonable guidelines, they have given us accurate information. It took a while to get there but we are there.
We believe that if you set a deadline, a year later we could ready for ICD-9 as long as the vendors are ready and, quite frankly, the vendors that we are dealing with I believe will be ready.
DR. COHN: So you say 2004.
MS. JOHNSON: Correct. Chris, do you want to talk about it?
MR. RIOPELLE: Do I hear 2006? No, our position would be that although there are some system impacts relative to privacy and security, the code set change impact really is one that would be married closely with the transaction code sets of HIPAA so I would advocate three years from the final compliance date, October of 2003, so three years from that date would be an appropriate time to require the subsequent code set change.
I think it's valuable to point out breathing time and breathing room is really valuable, and we need time to let ourselves get comfortable with significant change before we make any, even if it's a minor change. Adaptability is important.
For example, take a patient account rep who does processing of claims. As someone mention earlier, they get a lot of things done in the day. They are into one process, very quickly, all day long. You change part of their routine, their productivity goes down immediately. If I have to walk into the claims shop and say, "Everyone, clear your desk at the end of the day." It might take them 20, 30 minutes to clean their desks to make us compliant with privacy if that's the solution we identify.
We just lost 20 or 30 minutes of productivity. When someone is doing hundreds of transactions a day, that's an impact to us so getting used to it would be great.
MR. HARDESTY: I would think that beginning the work in 2006 would be the earliest, realistic, and quite honestly, I don't think it should extend out of 2006 so somewhere between January and October of 2006.
MS. SMITH: We were looking at the initial compliance date of HIPAA as being October 2002 so we were thinking three years after that date.
DR. ZUBELDIA: It was 2005.
MS. SMITH: Yes.
MR. MITCHELL: We may be less sensitive to the transaction set issues since we feed information into many of the systems as opposed to implementing them so ours was one to two years so we are actually on the shorter end than everyone else which --
DR. ZUBELDIA: One to three years from now?
MR. MITCHELL: From whenever you make your decision.
MR. BLAIR: I would like to make sure that we are talking about the same thing. I think that it may be that some of you from the vendor standpoint are thinking of your vendor systems that process points and transfer from the payers and thinking well, we are not going to get finished with that until October of 2003 and so you are thinking three years beyond that but, it appears to me as if many of you not only have financial, administrative transaction systems, you have physician order entry systems. You have results reported systems. You have case management systems. You have physician certification systems. On and on and on.
And you probably -- I can't see how you could wind up doing that in stages or piecemeal and be capturing ICD-10 codes in certain systems and not others so you are looking at a broad array. So while this is a big task, it also seems to me like it is the absolutely tremendous business opportunity for the vendors because it isn't just a code set replacement, it's enablement of much richer clinical applications as well, like what you said was the fourth.
Now, given what I've said, I'm wondering if your answers are going to be the same if you consider it from a much broader perspective. Could I hear your reaction?
MR. RIOPELLE: I would like ten years in that case from what you described.
MR. BLAIR: Yes, but actually, you probably don't because it means you delay the business opportunity for 10 years.
MR. RIOPELLE: Without question, there's an opportunity from a vendor standpoint, this is a tremendous business opportunity and it's a less tremendous opportunity for a proprietary user although we will reap significant benefit downstream from improved identification of clinical activities, improved description of what we're billing our payers for. There is value to that and ultimately will impact positively on quality of care.
We don't have the interim financial benefit that I think a vendor will have by whether it's rolling out new systems or continuing to service updates under their licensing agreements, those updates are priced into their licensing fee.
So I agree with you. I think it's a tremendous opportunity for vendors to do two things. One, provide more effective systems for us in the health care community and, two, recognize more revenue. I think the value to a proprietary user is that first which is have a better system of capturing data and submitting claims.
MR. BLAIR: Maybe the answer to this question requires more thought because in some cases, for many of you, you are going to have different application teams working on your clinical systems that on your patient accounting systems and so some of that work could be done in parallel so maybe you cant give an off-the-top-of-your-head answer in this context.
MR. MITCHELL: Well, from my perspective, the scenario that you have laid out is consistent with the testimony that we just provided. Change is going to bring about opportunities. It's not just system changes but new services. I believe we heard about a cottage industry coming up about how to go about doing this so I don't see anything inconsistent with the scenario you laid out with the commercial opportunities.
MR. BLAIR: Remind again of the dates you gave us, given that you feel that you have considered these.
MR. MITCHELL: One to two years after the decision of the committee to implement.
MS. JOHNSON: Yes, Jeff, I think what you are hearing is 3-M, though, and when you talk about where 3-M is, and they are in a different, they are in that niche. I mean they are the premier provider. And they are also much more a niche provider whereas if you look at McKesson, they really --
MR. BLAIR: They may not want to consider themselves a niche provider.
MS. JOHNSON: We consider them an excellent one. I think they do consider themselves that, don't you? If you talk about coding specific, would you consider that to be something you really do?
MR. BLAIR: Yes, but they also make clinical information systems.
MR. MITCHELL: Right. We have a variety of products in our portfolio, from our encoding and grouping and classification core, you are right, that we specialize in that one area. The comment is from a 3-M perspective.
MS. SMITH: Jeff, given what you said about business opportunity, I still don't think that would change our time of three years after the October 2002. I think it's more an issue that we want to make sure that the systems work correctly, that we have enough time to do what we need to do and our concern is getting it done and getting it done right.
MR. BLAIR: So your date is?
MS. SMITH: It would be October 2005.
MR. BLAIR: Okay.
MR. HARDESTY: And on the same issue, with 1,500 clients, that's really the issue, is how do you support the implementation efforts, the conversion efforts, the interface efforts and coordinate all of that within a condensed type of time frame? It's really not, at least for us it's not a software issue. It's a consulting and services issue and, you know, I mentioned January 2006 which is very similar to October 2005 and that seems like a reasonable time frame.
MR. BLAIR: That was another question I had. Some of you said you wanted early guidance, and is that separate from when an NPRM is issued? Notice of proposed rule making? Is that early guidance separate from that?
MR. HUMPHREYS: How does, excuse me, this is Betsy asking the question for those of us who've been reading the law. I basically don't understand how the Secretary provides any kind of firm guidance in the matter, plays it like this other than issuing at least a notice of intent. I mean, he's only got a few options open to him and I think a notice of some kind formally through that process is it, isn't it?
MS. TRUDEL: The whole point of the Administrative Procedure Act is that it's not done until it's done because it's dependent on public comment.
DR. ZUBELDIA: Let me read something from the act itself. Section 1174-B 2-B 2. It says if the code set is modified under this subsection, the modified code set shall include instructions on how that element of information that were encoded prior to the modification may be converted or translated so as to preserve the informational value of the elements that existed before the modification.
So it looked like the act itself required mapping between ICD-9 and ICD-10 to be part of this process.
MS. HUMPHREYS: Yes, but I guess what I'm saying is that what the community is, I'm hearing across the board here, whatever date is picked, what they want is the sign on the wall that that's the date and I believe that the only way for that to be done is an actual publication of a final rule that says, hey, here's the final rule and the final rule is telling you that it's 18 months or two years or three years or whatever it is selected.
I mean there isn't any way of giving a certain date other than publishing a final rule because in the whole NPRM process, that could be the nature of the comments which would mean even if we told you it was three years, we could get comments back that would say no, it needs to be two and then that would be in the final rule. So the only certain date comes out in a final rule as far as I can tell and even then, as we know, Congress could change it.
DR. COHN: I think we've all learned there is no certainty. Can you talk a little bit about notice of intent, though? I mean, I'm familiar with NPRMs and I'm familiar with final rules.
MS. HUMPHREYS: The notice of intent was what was used to try to get some discussion going around the very vexed question of the personal identifier, but there was a plan to issue a notice of intent and even the plan to issue the notice of intent which was intended to get comments from the field never came to pass because people were so excited about the Secretary even saying, hey, maybe we'll do this, can we get your input? It was dead in the water.
DR. COHN: Well, having had no experience with the notice of intent, notice of intent is a pre-NPRM?
MS. TRUDEL: Yes, the purpose of it is to basically open issues up to public discussion without the Secretary proposing a specific approach.
MS. HUMPHREYS: Yes, and I think that the -- you know I mean, others in the room know better than I -- but the whole use of the area would note the intent to use a notice of intent regarding the personal identifier was because really it was such a complex issue and it was so controversial that no one was willing to lay down, here's a proposal. It was just like look, we've thought about this and there are a thousand issues but, of course, it was so controversial, we weren't even able to ask people for their opinion on the set of issues.
I guess my feeling, hearing from everyone here that what we are dealing with is a date issue and not whether we should move, then I would say that this issue, you know, you can come up with a proposed rule in this area and you can pick a date for it and then we can hear the comments. I don't think it's so vexed an issue that it requires a notice of intent.
DR. ZUBELDIA: Let me ask something on procedural methods for this. I'm looking at the final rules for transactions and section 162-940 talks about exceptions from standards to permit testing of proposed modifications. Is this something that would have to go through that process to be tested before it can be adopted as modification to an existing standard or not?
MS. HUMPHREYS: That particular thing was intended in order, supposing the state of California wanted to test and, in fact, they were going to get on exception, maybe they did, they started out like they might. But supposing somebody wanted to test the use of UPNs as opposed to some HCPC for reporting on equipment and supplies or billing. They could come forward and request an exception saying we want to test the use of this and we will report back to you about the issues and problems.
There obviously has to be an environment in which a test can take place without requiring, you know, the whole industry from coast to coast to participate in this test, but if a group can come forward and say we want to test this and we see how we can do it in our environment, then there can be an approval for a test.
DR. ZUBELDIA: But it's not necessary that a modification to a code set as we are describing goes through a process like that.
MS. HUMPHREYS: No, that's just to allow if people wanted to do such a test. I don't see it as a requirement.
MR. AUGUSTINE: I was going to comment on the earlier business opportunity question that Jeff raised. From my experience in the health care industry is that the IT spending has been significantly lower in health care as opposed to other industries and now that it's finally getting up to speed and comparable to other industries, those extra resources, those increases are spent on HIPAA so there's no real opportunity or no real seizing of an opportunity by the operations side of health care to look at these business opportunities and, you know, the decision support mechanisms and framework for health care systems typically are way behind what you would find in the other industries and things like sigma are enveloping the management culture. In other areas you would barely ever hear that phrase used in the health care industry.
I would surmise that some of that has to do with health care's background from being a QA type industry as opposed to a QI or process controlled type industry and that's a culture that we are going to have to work to change.
MR. BLAIR: I'm not proposing this, I just want to explore this. I think that a compliance date for a vendor is very different than a compliance date for the industry. And I would at least in my mind think that vendors have to have systems that would be compliant with the new ICD-10 codes at least a year and a half before we could expect providers and payers to meet that compliance date and maybe two years before so when you were talking about the dates for compliance, were you factoring that into your answer?
DR. ZUBELDIA: Jeff, let me expand on that for a second. The compliance date for the receiver of the transactions is also somewhat different from the compliance date of the sender of the transactions so the receiver has to be able to accept this ICD-10 code so the sequence would be the vendor goes first, then the receiver, then the sender of the transaction.
MR. BLAIR: I'm not trying to get you to give your answer or delay it. I just really want to make sure that we are dealing with realistic information because I would much rather we go forward seriously with realistic dates and make them stick than unrealistic dates and have to back off because the industry can't make it.
MS. JOHNSON: Well, Elizabeth says it only takes two or three months after the system is in place, right?
MR. HARDESTY: I think you are absolutely correct as far as vendors needing to have the lead. I mean one of the things that I think we are commonly struggling with at the moment is the lack of the final rules for transactions with what is actually going to happen with privacy and some things of that sort that sort of left us in limbo to make final changes. Should we do the addenda, should we not do the addenda. If we do do the addenda, there are going to be changes to the addenda, etc.
And so the lead time is a very good point. In this particular area, however, I really think that the lesser problem here truly is what each individual vendor does with their product. A much greater problem is how they can coordinate that at a given site where three, four or five different vendors are concerned, all of them exchanging diagnosis information in one shape, form or fashion and synchronizing that effort and making it work and that's, you know, that's going to be the true challenge.
MS. JOHNSON: I would like to speak to that and I don't want to take away from anybody in terms of their -- I depend on these folks and if you think I don't, I would be misjudging what their value to me as a provider is. Like I said, I have 950 of them we routinely work with, but the major vendors are sitting at this table plus a couple more. We deal with these problems every single day. Everything we do at Tennant requires interface changes. Let me be just honest with you. There's nothing any more that we do because of integration that is a singular kind of, you know, we go out and do what we call a uni-directional, a one way path of data and nothing else happens. Those days are over.
So the challenges that you were talking about are absolutely real, but we deal with them every single day so that's why I believe -- I also, and maybe I have an advantage here, but I do not use the same people to work on coding that I use to work on business trans actions so TD are not being considered when I do my part of this. It is not taking us two to three months to get that portion of this in. We started on this in January of 2000.
But with coding, we make these changes every year, and we support the supporting systems every year so this is a larger change. I'm not diminishing the change and I realize we've got a field links change but again, I also don't wont to see the industry get crippled by waiting three, four or five years when we are not able to accurately reflect what we are doing for patients today and then I worry about what Chris worries about which is fright issue, quite frankly, because particularly with the potential of local codes going away, we've already had that, I'm not arguing that point. They can argue that with you tomorrow.
I'm just looking at these arrays and what we are dealing with and also there will probably be some change in CPT. So that is from our perspective what we are looking at is we need this, we are willing to take some more change. I also agree about the resting period but I'm going to make I guess what would be considered potentially a cynical comment. If we are not doing this, we would be doing something else. I've never seen this industry be still. I've been doing this 30 years and there's always something that -- no offense -- the federal government comes up with that we've got to contend with.
So, you know, we think this is a positive change so we are willing to embrace this particularly if that will then take the place of something else. That is really where I'm coming from.
DR. ZUBELDIA: Let me ask you for a second that resting period you were talking about and Keith also mentioned that once the codes are in place they cannot change. I assume that you understand that they are expected to change once per year, right?
MR. RIOPELLE: For us, the concept there is finality. We know things change. We are always in a constant state of change from updated codes, but finality here is what's going to happen on this date. We are not going to change it, we are not going to reinterpret it, we are not going to have Congress say, oh, we are going to do something else. This is what you have to deal with. Give us as a proprietary user some time to breathe, some time to adapt to other big changes and then we are probably in a position to say go ahead, bring it on but don't bring it on and two months before you bring it on say oh, you know what, Chris, we are going to change it. That is very hard for us and others to adapt.
DR. ZUBELDIA: There will be new codes changing every year but that's still finality as far as you are concerned.
MR. RIOPELLE: Yes, that is finality because you are changing very small amounts. It's not fundamentally changing to say we decided to include mental codes or exclude mental disorder codes. I mean, some large system changes like that in the last stages would be very disruptive but adding and updating the code within the last year, that is anticipated but we can't have a moving target. That makes all of our jobs much more difficult than it already is so all of our comments was based upon once you tell us what the end result is and let us get there before any major change is beyond it, let us get to that point and then we are back to business as usual which is changes every year.
DR. ZUBELDIA: I want to follow up on that with a question. How do you see a transition plan?
MS. SMITH: Do you mean as far as drop dead date?
DR. ZUBELDIA: Would you like to see a sequence of dates or vendors, payers, providers or a date where everybody switches at the same time, some test period before? How would, if you had your dream, how would you see it.
MS. SMITH: I think everyone would need, for us I think we would like to see everybody switch over at the same time, probably because we have payer systems and provider systems and it would be difficult and they pass information back and forth and I think it would be easier for them to pass information back and forth when they are on the same system so we would like to see it all occur together and all on the same time but we hadn't really talked much about testing. I think there will be testing required and we are going to have to test with the other vendors and we are going to have to test with payers. That is expected but we haven't really looked at I guess for an implementation period how long of a testing period do we need.
I'm not quite sure I could respond, given the different types of systems as to how long we think we would really need to do that testing.
MR. HARDESTY: Someone previously mentioned being able to put together a plan with given milestone dates and perhaps that's something that the WEDI Smith work group could make a recommendation on of what would be appropriate and why but something to that extent where there would be, you know, by X date this should happen and these things should have been tested because, as I mentioned, there are going to be different layers of testing, both the vendor itself layer and then the communications between layers and others and so I think it would be very helpful to have the type of study and sort of a road map of this is at least the recommended way to get from point A to point B.
DR. COHN: I apologize. It's the end of the day and I may be getting a little dense. I wanted to make sure that I was hearing everybody with the exception of Elizabeth Johnson what wants it done tomorrow, but I'm seeing everybody else relatively accurately just in the way I'm beginning to sort of think about it.
I guess what I'm hearing most of you would like is relatively soon some sort of a notice of proposed rule laying out a direction and what you would like after that is, in the not too distant future a final rule providing absolute guidance with an absolute date but you would like a three year implementation. Is that sort of what I'm hearing?
MS. SMITH: Yes, that's what we are interested in. Once we see a guideline and final rules, then that kind of tells us to start, but until we see some definitive information, then we are not going to start developing our coding on draft information until we've seen something a little for final because we don't want to make a lot of changes and then have, the final rule have even more changes or we have to reverse things that we've already done.
DR. COHN: Is there like a lesson from HIPAA that nobody does anything in place of the final rule? I just wanted to see. I don't think that's even possible, a three year delay, but it just, is that what, Robert you are sort of --
PARTICIPANT: Yes.
DR. COHN: And Keith, you are probably okay with that.
MR. BLAIR: When we are talking about three years, is the three years the January or February of 96, is that the three years? Or is it October of 96 -- excuse me, 2006.
DR. COHN: I thought that you were ending up saying it from October of 2003 when the compliance date ended for the financial administrator.
MR. BLAIR: I was rephrasing this because I was hearing two things from them. I was hearing dates related to other HIPAA rules being implemented but I was also hearing that they weren't going to start acting until they saw some certainty and the final rule is really certainty and so I think that other date is probably the more important data I think. Whenever that may happen. We know it won't be for October of 2003 probably.
MR. KOLODNER: I guess what I'm hearing is a range of from the time that you have a final rule, how long do you actually need and the three years that I was hearing was at least three years from October of 2002, maybe a little bit longer, but it isn't that it was three years from the time of the final rule.
MR. RIOPELLE: I guess I had the three years. My three years was simply let's get HIPAA digested and then we can focus on new efforts and we would think to allow appropriate breathing time to allow us to create an implementation plan with appropriate milestones. It would be a there's years from the HIPAA out of our cross hairs on a daily basis.
We are sort of on the fence here. One argument is three years from date of a final rule or three years from the HIPAA effort, and I would say if a final rule came out, frankly, in our situation, we, between now and October of 2003 we are going to be so involved in getting our transaction and code sets done, you could issue all the final rules you want and I'm probably not going to look at it.
DR. ZUBELDIA: Does it make any difference if the magic date is October 1st, January 1st or any other day of the year?
MS. JOHNSON: One of the things you could do that would help is we know the code changes come out about the same time every year. One of the things that was mentioned is we do code changes on two things. We don't want to do that that year, quite frankly, like the ICD-9 freeze, ICD-10 can come out and make the changes so we are not maintaining more than that and again, back to kind of your issues, I'm going to reiterate, not that you don't know where I stand, but as HIPAA becomes final, I believe there could be some concurrent work that could start now. Give us final rules so we know what to do, but please don't delay the industry until 2007 or 2008 before we can actually get this in place. That's too long.
So, you know, again, when I look at the NPRM process, I look at the 60 days for comments, I look at sometimes the length of time it takes to review those comments depending on the public furor related to those comments, then we finally get a final rule out and, like you said, then we see that digested and further guidelines come out. It is a very lengthy process so if you delay that process, we are way down the road and I guess that's why I'm pushing to begin the process, to begin to think through the process now to give us time to digest that and to quite honestly give the vendors time to figure out what they are going to do in response because the longer you push that, longer the whole thing gets pushed.
I can't do anything until I have vendors. As soon as I have vendors we can move. Now, Chris has got a different problem.
DR. ZUBELDIA: For the rest of the panel, does the magic date make any difference?
MR. MITCHELL: There are certain cycles within the industry. The industry, there's a rhythm to when people expect change. Vacations are planned around these rhythms; weddings, anniversaries, the whole thing, so recognizing that we do have rhythms within the industry, using them to your advantage would be a recommendations. Don't do it out of cycle.
Does that mean that January is different than October? Well, since we have an in-patient cycle and an out-patient cycle, those two magic dates are already in the rhythm but not necessarily in the middle of August so just take that into account.
As far as the HIPAA relationship, we've already started looking at and digesting what does this mean, very similar to the inventory. I mean, this is going to happen. The light is an oncoming train so to stand still and not do anything, that's why one to two years is what we are saying because that assumes that we are already looking forward to what you are doing. We've already had to do this outside the United States so we've had experiences with other countries of implementing systems so that may be to our advantage but we know that this is coming and the benefits of increased accuracy of coding should not be delayed to this 2006, 2007, 2008. The sooner that a final rule can come out and the sooner that the industry starts working, there is going to be transition issues.
The first year is not going to be fun. It's never fun, but the sooner that we get into the cycle of using the new code sets, the sooner we will get to the other side and have better data and have better accurate information to make business decisions and policy decisions.
DR. ZUBELDIA: I don't think my implementation plan question was answered. What do you guys think about what should be the implementation plan? Since you have the microphone, Louise?
MS. SMITH: I guess we would need, well, we want the lead time. We would like to have, to know how far in advance we have to develop, we would need to know like the drop dead date. I guess guidelines for all the payers, if all the payers are going to be implementing it together, that kind of information; guidelines for how to implement it. I would have to give that a little bit more thought I think.
DR. ZUBELDIA: Should the vendors be required to implement these changes before the payers and providers are required to implement the changes?
MS. SMITH: No, I think what we are looking at with our time frame is that we were talking about three years that we've it done and we are rolling it out into the customers in that three year time period so I don't think having the vendors having it ready early is going to do a lot of good for us. You would want to have it rolled out like once it's completed, do some testing and roll it out and start testing it. Does that answer your question?
MR. HARDESTY: I think our clients require us to have it ready. I don't think anyone has to externally require us to have it ready.
MR. RIOPELLE: In response, I would say that the idea of being a sender, a vendor, a sender and a receiver I think was the terminology someone said. I guess I'm a vendor and a sender and the three years I've been talking about is going to need to incorporate some internal milestones if not given to us as guidance by the government of what needs to happen and part of that is being the vendor and creating and testing as a sender with our receiver.
So I guess the guidance and direction in terms of milestones that I recommend would be valuable to us, but I would include that staged effort -- vendor, sender, receiver -- in my three years of effort.
DR. YASNOFF: I'm still not sure I'm clear on this and before our panel goes away, I would like to get further clarification. There are basically two time periods we are talking about or three time periods.
There's a time period between when we make a recommendation and there's an actual NPRM. I don't think we have control over that time period.
Then there's a time period between when the NPRM is issued. There's obviously the 60 days for comments and then there's some not clearly determined time period before a final rule is issued, and we don't have really much control over that.
So the only time period we really can make a definitive recommendation about is the time period for implementation after the final rule is adopted. So if we assume that the time period between when we make a recommendation and when a final rule is issued is certainly going to be at least a year, you'll allow me that, probably longer, then if we made a recommendation that the time period for implementation after the final rule was two years, then essentially what we would be saying today is that it would be no sooner than 2005 before this was actually implemented. Right?
So the question I have for the panel is, is that what you would like us to recommend because we can't really control the time between when we make a recommendation, the NPRM, the time between when the NPRM is issued and when the final rule is issued. That's not really under our control.
We can say once this is final, it should be implemented in a week, a month, a year, two years, ten years. So that's really the one thing we can recommend so I would like to hear what the panelists really would like us to do with that one time period we really can recommend. Given that, it will be at least a year between when we make our initial recommendation and the final rule so that means that you have to take what you are recommending, add at least one year and that's the earliest possible date when this could actually be in effect.
MS. JOHNSON: I would ask the vendors to respond first because until the vendors respond, and Chris as a vendor and then as a sender, we've already told you we can respond in that time frame. We have to have the vendors on board.
MR. MITCHELL: Two years.
MR. BLAIR: Hold it. I have got to get this pinned down. Two years from what event?
MR. MITCHELL: From the final rule.
MS. SMITH: He answered Bill's question which means two years from the time the final rule is issued.
MR. BLAIR: The question I'm asking is whenever the final rule is issued, that's the compliance date for users to be in compliance.
MR. MITCHELL: Right. We feel the absolute shortest is one year and that's really pushing it but that assumes that there's some pre-work been done. Two years is the preferred time frame from final rule. It gives you the time to develop and then as we've heard, the time to coordinate and again, back to rhythm. You have to be able to have the rhythms in place and so it gives you two years to get into the rhythm and to get everybody saying okay, it's real. All right, it's really real.
MS. SMITH: Okay, so you are saying that it would take a year, if we had the rule, you know, the semi-final rule a year in advance and then had the two years on top of it so that we ultimately had the three years --
DR. YASNOFF: You would have, whatever time period, what I'm asking for is the time period after the final rule is adopted there would be at least a year when the proposed rule that wasn't final would be floating around, but in this case since, you know, the final recommendation for what ICD-10-CM is going to be is pretty clear already. It's not going to be transformed during the time, you know, the comment period is not going to result in a total transformation of ICD-9 or excuse me, ICD-10-CM.
DR. COHN: Dr. Yasnoff, I guess I would question that. I think the comments that we keep making and yet our conversation keeps belying that is that until you have a final rule there is no certainty.
DR. YASNOFF: No, no, right.
DR. COHN: In the same way that with our HIPAA administrative and financial transactions, there is no surprise when that final rule came out. But still nobody did anything.
DR. YASNOFF: Right.
DR. COHN: So its that sort of a reality and you would never know for certain what the Secretary or when the Secretary was going to do.
DR. YASNOFF: But the time frame we are talking about is, you know, this time frame after the final rule, but that clock could not start ticking until a minimum of one year from now and that's really optimistic I think.
MS. SMITH: I think if we had, if we pretty much had the final code set and then it was a year before the final rule was published and then we had two years on top of that the, that would basically give us the three years we are asking for so I think that would probably work for us.
MR. MITCHELL: I would agree.
MR. RIOPELLE: I would not. Still, because of the challenges that HIPAA creates, I would advocate, irrespective how long it takes for final rule to come out, as a proprietary user, a three year period of implementation post-final rule.
MR. KOLODNER: Picking up on the idea that there are rhythms, if the final rule would come out in mid-August and it were a flat two years, then we might end up in the same situation that you commented on that the implementation date was the middle of August.
MS. HUMPHREYS: No, you are going to have to have a better algorithm. It's going to be the second October after the final rule is published.
MR. KOLODNER: And that's why I'm wondering about fixing it that way rather than a two year.
DR. YASNOFF: So that would provide a minimum of two years for actual implementation after the final rule was announced. Could be as much as three but it would be at least two.
MR. MITCHELL: Well, you may want to qualify it to say fiscal years. If you say fiscal years, then if you have --
MS. GREENBERG: I think we've got the gist. Now we are going to have to work out the fine points of the algorithm.
I think I asked a question like this at the April hearing, and I'm not sure that this panel is the one that would be the most perfect to answer it and some of the people are still in the room who were here earlier and some are not but given what you heard about the inadequacies of 9-CM, the international and mortality, all the different comparisons, everything you heard, and I'm certainly not going to go through it again since we heard it several times, and given the fact that it would probably be a miracle if we had a final rule out in a year, does anyone feel that this causes anyone any problem, the fact that then that makes, if we had to wait two years from the final rule, that makes October of 2005 pretty unlikely.
DR. COHN: You wonder if we think this is an emergency or just or urgency? Are you trying to ask whether you know this is an emergency or --
MS. GREENBERG: Well, I mean, we wait until 2006, 2007, 2008. I mean, well, obviously you have to wait for a final rule but there is the possibility of if the code set is completely known and the process between the proposed rule and the final rule takes a little longer than one might wish, that you could have less than a two year from the final rule if, in fact, this was going to take you into a much --
MS. HUMPHREYS: Another approach, of course, is to publish the notice of proposed rule making and say we propose this and we propose that the implementation date be 2005. We know who would faint and be upset by that. We know that. We know there might be others who would and then, of course, we would get the reaction to that so that, I mean, you could put a date certain in there and then the issue, because it does seem to me that it would be a terrible shame if this thing goes beyond October of 2006.
I mean, I think that would be an unmitigated disaster for the health care and data in my opinion. Now, it would be in my opinion much better if it were 2005 but I think we've heard from a number of people that maybe that's pushing it too hard here in terms of actually being able to do it right so that the data gets exchanged and interpreted correctly.
MR. RIOPELLE: I don't think it's valuable to put it off indefinitely. I think there are some intrinsic and some real beneficial impacts to doing this but what I don't want the subcommittee to not have the perspective on, we are not the only proprietary user out there. There are thousands, I would argue, thousands and thousands of organizations like us that are smaller that have far fewer resources than we do as a large company that are going to be sitting next to me and saying, three years, what are you talking about? We need seven.
I don't think that benefits the health care industry. There has got to be a happy medium but I think in reviewing the folks that have been testifying previously and are here today, we are talking about large associations, a large provider, and I think we represent some of the common interests in health care. I think we are probably missing some of the specific smaller rural providers, some of the smaller rural hospitals and out-patient facilities, whether they are part of the AHA or not that really don't have a voice here and that three years would be their minimum so I think the group that has been testifying has been fairly heavily weighted forwards large organizations who can assimilate change pretty effectively.
MS. TRUDEL: I really hate to add a complicating factor but enough people have said during the course of the day that CM and PCS should be adopted at the same time that I have to ask, if that were the case, does that make a difference in terms this timing issue that we've discussed, especially considering that that is adding additional complexity, and that we've also heard that that the procedure part of ICD-9 is more broken possibly than the diagnostic.
DR. ZUBELDIA: Let me add to that a little bit. The procedure part, the PCS part, is somewhat more controversial than the diagnostic part, and it could take one, two or years to resolve those controversial issues. Would you rather go with the ICD-9-CM first or wait? I'm sorry, ICD-10. No, ICD-10-CM first or wait until you can do both at the same time, knowing that it could take a long time.
MS. SMITH: Our preference was to implement the two together but we were, and we had looked at a time frame, we had gauged three years to do the two of them together. We hadn't really considered the fact that PCS may happen down the road. We were sort of looking at them happening together and we thought about it but it just seemed more likely that the two would happen together so our three year time frame considers PCS and CM being implemented at the same date.
MR. HARDESTY: We prefer they were together as well. In fact, I think Louise in her testimony brought up the difficulties with the groupers. We may want to comment on that, of using 10 for the diagnosis and the 9 for the procedures. That has got to add a degree of complexity that none of us are really thinking about.
MS. JOHNSON: I think that certainly speaks for us as well because our anticipation was when we make the change, let's make both. Let's don't do one and then drop the shoe again a year later. It's too difficult. We are already anticipating annual changes. We would prefer them together. I don't know about Chris.
MR. RIOPELLE: I'm not going to whine on this one, but I think together would be probably, not understanding the detailed complexities of it, in our shop I would say let's to it at once.
DR. ZUBELDIA: So let's answer my question. Would you rather wait and do it at once or do the diagnostic first if the procedure may be delayed?
MR. RIOPELLE: As a proprietary user, I would say let's wait and do it all at once now. I don't get any revenue for making the change part way through. I know I'll use a little bit of the benefit, I think, of a better classification environment, but I would rather see it happen all at once.
DR. COHN: Can I ask a question because I just want to push on this one a little bit because I heard Louise talking about the need to have ICD-10-PCS to do groupers and I can't think of actually a compelling reason why that's the case since right now obviously groupers work on a procedure, if there's a procedure, other wise it works on a diagnosis or works on it together but it's just this question of how you deal with a grouper. You can run a grouper on ICD-9 volume three plus ICD-10, volume one and two. There's nothing particularly space age about that. Is there something I'm misunderstanding about this, either Keith or Louise?
MR. MITCHELL: No, there's no dependency between the two. I mean, we have OPCS procedures working with I-10 diagnoses. We have I-9 diagnoses working with I-9-CM procedures and groupers that have been implemented around the world. It is just a different clinical research effort to make the groups, the DRGs that come out clinically coherent so there is relationships but there's no hard and fast dependency of one in the other. If there was, the world experiences dispelled us of that notion.
DR. COHN: Okay, so there's no dependency there.
MS. SMITH: You would have to be able to in the system decide, determine some kind of identifier to determine which code set you were using so between 9 and 10 --
MR. MITCHELL: Well, one is a diagnosis and the other is a procedure.
MS. SMITH: Yes, so you would know right there. I mean if you were using one versus the other.
DR. COHN: I'm not saying we should or shouldn't. I'm saying that's not a good reason to say no, you have got to do them together.
MS. SMITH: You build tables that have like a 9 code set and a 10 and when, you would have to have some kind of identifier to determine that you would be going to which code sets. It would identify which table you are going to go to to get the codes.
DR. ZUBELDIA: Evidently you would have to do it twice. You would have to do it with one table and then the other table so it's twice the effort. Is it worth waiting so we only have to do it once or should it be done because the ICD-10-CM is needed now?
MR. MITCHELL: I think I would encourage the committee to come to a resolution on the procedures. That seems to be where the major pain is being felt by the industry and if there is a delay in the process then focus on that part of it and come to resolution.
MS. BURKE-BEBEE: Kepa, with your scenario, if they are going to have to have a maintenance period, they are going to have to do it times two anyway, correct?
DR. ZUBELDIA: No, if the switch happens at the same time, the groupers will be doing grouping based on the new procedure codes and new diagnosis codes where as if the switch happens in two steps, there would have to be an intermediate grouper that uses the new diagnosis code and ICD-9-PCS.
MS. BURKE-BEBEE: But what I'm saying is that there is still going to be ICD-9 in the real world going on, is it not, even though there may be a day that the ICD-10 is implemented, there is still going to be a maintenance for the prior ICD-9.
DR. ZUBELDIA: But that will quickly fade away.
MS. BURKE-BEBEE: Well, maybe that's the issue. It's the evergreen issue as to when it gets turned off.
MS. JOHNSON: In other words, it's how long we have to run parallel systems, is that what you are thinking?
MS. BURKE-BEBEE: Yes.
MS. JOHNSON: Yes, that is definitely an issue. I think it's the three times. I mean, I agree they are not dependencies. It's just that we know when change comes, it's easier to do it at one time but if there is going to be a significant delay, then I think we need to rethink it. I think I would be surprised if any of us could seriously, I think kind of our lead from the industry from the you guys and others who have written on it was that they were coming together so I don't think, we haven't spent a lot of time thinking that they were to come separately, what the implications of that were.
MR. AUGUSTINE: I don't see how we can push this into 2008, 2009. I mean, this is what we are talking about now. If we wait and put the two together, I don't think we can wait that long.
MS. LEON-CHISEN: I would just like to remind the subcommittee the discussion we had at the last meeting. ICD-9-CM volume three is broken and whether we make a decision to implement 10-CM and 10-PCS at the same time or not, there still needs to be a decision made as to what needs to be done, and I think from the last hearing there was strong support to change to ICD-10-PCS to replace volume three procedures for the inpatient setting so if you are not ready to make a decision regarding a single procedure classification system, I think at the very least we could take care of the hospital in-patient side.
DR. ZUBELDIA: I think we are out of questions for this panel. I want to thank you very much for the interesting debate and your participation. Now the subcommittee is going to continue debating these issues. If you want to stay in the room, you may be able to contribute and we'll be able to ask more questions. It's up to you.
MS. HUMPHREYS: Well, I think it's very simple to come up with what you should recommend to the full committee.
DR. ZUBELDIA: I think there are some things that are well-defined like the coding guidelines for DSM-4.
MS. HUMPHREYS: No, I mean on the whole major issue. I mean, I think that the committee should recommend to the Secretary that we publish, that the department publish an NPRM and that the NPRM say that the proposed change is that we are going to move simultaneously to 10-CM and to PCS two years after the publication of the final rule on that process and then you will find out what the whole community thinks about this and you will know based on those comments whether it makes any sense or is it going to be necessary or are there sufficient issues out there in the community that you must divorce 10-CM from 10-PCS and that seems to me to be the sensible thing because you guys can hold hearings, we can all sit here for days and days and days but there are thousands of players out there and you cant get them all into this room so it seems to me that that's what you ought to recommend.
DR. YASNOFF: I agree with Betsy but I would add in the recommendation to the Secretary that the sense of the group is that the implementation date, that the final rule should be issued in time to have the implementation date be October 1, 2005. In other words --
DR. ZUBELDIA: The final rule should be at least two years before.
DR. YASNOFF: Right, not earlier. What I'm saying is Betsy is saying that the implementation should be two years after the final rule is issued but not earlier than October of 2005. I don't think we have to worry, Jeff, about it being earlier than 2005 because in order for that to happen the final rule tough to be issued before October 2002 on this plan but what I saying is we recommend --
MR. BLAIR: The final rule would have to be issued no later than October 2003. In order to get 2005 but in order to get earlier than 2005, the final rule would have to be issued before October 2002.
MS. HUMPHREYS: No, 2003. If the final rule were published in 2003 it would be two years later in 2005.
DR. YASNOFF: What Jeff is saying, we don't want the implementation date to be earlier than October 2005 and I don't think we have to worry about it being earlier than October 2005. I think we are worried about it being later.
MR. BLAIR: Well, but I think that in, with respect to the testimony we heard from the vendors that we are winding up saying two years but we are also winding up saying not earlier than October 2005 and I think that covers both of the --
DR. COHN: And I guess the other issue is remember the Secretary has the option to make a one year implementation.
DR. ZUBELDIA: Or six months.
DR. COHN: Or six month implementation so it's important to have a date there.
DR. YASNOFF: Oh, I see what you are saying but essentially saying not recommending that the rule be published with an October 2005 implementation but that it be between two and three years. In other words, the second October after the final rule, but that we are saying as an it seems to me that we would urge the Secretary to do that in such a way that the implementation date was October 2005 and that's the implementation date we hope --
DR. ZUBELDIA: With two year time frame.
DR. YASNOFF: Right. In other words, that the final rule should be published before October 1, 2003, with a two year or two-plus year compliance time.
DR. COHN: You think, I think generally I sort of agree with what you are stating --
MS. BURKE-BEBEE: I mean it's not that necessarily this is what's going to happen, but I think the way to move this process ahead and find out whether in fact what we've heard here is accurate for the country is to do this.
DR. COHN: Yes, we are not going to resolve it further.
MS. BURKE-BEBEE: How difficult is it going to be to get out the next transaction and code set, NPRM. I mean, this second NPRM that we are talking about?
MS. TRUDEL: It wouldn't necessarily be the transactions.
MS. BURKE-BEBEE: Okay, so that in mind, we already have a transaction and code set final with an addenda coming as we speak. How hard is it going to be to get out another NPRM? In the big picture, knowing that we've got security looming, and the other issues of the NPRMs and finals? So, is it going to take a year? Is it going to be less difficult? Is there going to be the resources there?
MS. TRUDEL: I think the biggest concern is that this is a major rule. It's extremely, I don't want to say burdensome, but it has a significant impact on everyone we do business with. There is going to be a need to do a certain amount of assessment in terms of impact.
MS. HUMPHREYS: And the impact assessment seems to me to be the worst part. It's actually a simple thing to write.
MS. TRUDEL: It's being able to assess and explain the impact that's really, I think, the critical factor here.
MS. BURKE-BEBEE: So the reason why I ask is that's important to know so that when we talk about two years out and the 2005, is that going to be a realistic recommendation to make?
MS. TRUDEL: I think this is something that the committee, it seems to me, needs to recommend to the Secretary and then the Secretary needs to take it from there. I mean, I think at this point the committee needs to reflect what it's heard.
DR. YASNOFF: This is why I think, you know, I was suggesting, Betsy is suggesting that we recommend that there be this minimum two year implementation. That we not say we are recommending that this be, that the implementation date for this be October 1, 2005. What we are recommending is that there be a two to three year implementation time after the final rule, and that we hope that the Secretary will make every effort to issue the final rule before October 1, 2003, so that the implementation date would be 2005 but that we, in making the recommendation that way, we recognize that it might take until October 2004 to issue the final rule. I mean, that's the reality.
MS. HUMPHREYS: I agree, and I also feel that my view in terms of dealing with obviously the concerns that -- which are real obviously about the burden of doing this and what this means to the community and the issue of PCS and so forth, and these are controversial issues -- but I think to my way of thinking, the committee in making a recommendation to the Secretary basically has to say that our way of really getting, I mean, we've done our best here, we've had N number of hearings, we've brought people until from all over.
This is the sense of what we hear from this group and there are these other issues, but the way that the secretary and the department has of really getting the feedback it needs to make this decision, the way to do it is to publish the NPRM. I mean, I don't think the Secretary has another way of figuring out exactly what the full range of issues are here. I mean, from the community except to publish an NPRM.
DR. YASNOFF: Well, the Secretary could ask us to hold more hearings.
MS. HUMPHREYS: Well, yes, but you know, how many people do you get in a room?
MR. BLAIR: And one thing, Bill, I didn't get the sense that -- I didn't hear the same thing that you heard from Betsy's statement because I think that she is saying that it's our recommendation that when the Secretary put out the NPRM, the Secretary put it out, in a sense maybe with a trial balloon, saying specifically that the compliance date will be two years from the date of the final rule. We may get some reactions to that. Let's see what happens, but that if we are accelerating this, and the NPRM does get out early, but we also put the caveat in there, but that that in no event be earlier than October of 2005 if we are fortunate and the NPRM is issued early.
MS. HUMPHREYS: Well, you know, Jeff, I was considering Bill's a friendly amendment because I think that the one thing that, you know, a friendly amendment to have a date. You know, I might, 2005 would be my preference. 2006 you might look at it based on what everybody else here was saying but the thing is that I do feel that it's important for the committee to convey to the Secretary the fact that these things are broken in the worst possible way because see, if we say, you know, if the committee says two years after the final, then somebody can say, gee, we can put this off for a long time, we'll do an incredible slow roll on the issuing of this thing and then it could be two years, we could talk 2010 but I think we really have heard from the community that this thing is, it's more serious -- I haven't heard too many people say it's a great idea to wait around 2007, 2008, 2009. Au contraire.
DR. YASNOFF: Right, and I think we also need to provide in our communication to the Secretary some sense of the history of this going back to 1993 and even in some cases further back than that but this is not exactly a precipitous, unexpected move.
DR. ZUBELDIA: Betsy, help me understand what could change as a result of the comments to the NPRM. What could change to the codes?
MS. HUMPHREYS: Well, I think it gets back to Karen's issue which Simon was also picking up on. The issue is when push comes to shove although there are, you know, people have some concerns with PCS and what have you and that to me is more controversial but on the other hand, if its possible to say so, the procedures are even worse broken, are broken in the worse way, then the diagnoses which are terrible also, so I guess the issue is what is the alternative and you had all these people saying maybe there's a real alternative.
I personally don't see it and I think Nelly's point that you could say, you know, and I think the NPRM would have to say this, but you are recommending that the PCS be a replacement for the third volume which is not required for out patient. It's only being used in in-patient now.
I could imagine that you could possibly get a huge influx of people with all kinds of other ways to solve the problem that relate to procedures but everybody really agreeing that there isn't any option and we have to go to 10-CM on the diagnosis side. I'm not saying I hope to see that happen but I could imagine that. So you could get a reaction that would split these.
DR. ZUBELDIA: That's where I'm going with what we heard today, the vendors and providers want to have two to three years to implement it from the final rule. I think that they want two to three years to implement it from the finality point, not necessarily the final rule so when they say that they are going to be mandated, they will have two to three years to that deadline because I don't think that there can be any changes here so if the proposed rule says October 2005 and the final rule doesn't come out until December of 2003, and they don't have the full two years, I don't think that has a negative impact as long as they were targeting for the October 2005 date to begin with.
MR. HUMPHREYS: And then there's the issue that we've heard from these large vendors and that there may be many people who will come back and say and maybe that will be valid, I mean something else that could happen, well, you know, too bad, 2005 cannot be done. It must be later. I mean, that could be another thing that could come out of the comments.
DR. ZUBELDIA: But I mean if the date is nailed down in the proposed rule, rather than leaving it floating up to the final rule, floating point. If it's nailed down in the proposed rule, they know that that is the target date they have nailed down and they have got three or four years to do it.
MR. HUMPHREYS: Yes, well, as I said, we have elements of the community who don't feel that's long enough but I would also tend, I mean, it's kind of like what Bill was saying, and I feel that a date should be mentioned because I think that if there are basically if there is a view even within the department that we should just wait forever, I don't think the committee would be correctly communicating what we've heard which is that waiting forever doesn't seem to be a very good option.
DR. ZUBELDIA: What is the advantage of floating the date collective to the final rule? Is there any advantage to floating that date in, say, two years after the final rule? I don't see any advantage to that.
DR. COHN: Can I comment? First of all, since we are talking about procedure codes, I do have to recuse myself from conversations on procedure codes. I thought I was going to get by today without having to recuse myself. It's only 5:10, okay. But speaking specifically about diagnosis codes, I first of all want to reflect on that we've learned a couple of things from HIPAA. I think, Kepa, regardless of what you are thinking, people don't act until there's a final rule.
Now, maybe you know people are all mobilized around security right now, that they started on the transaction rueful before the final rule same out, regardless of what was in those rules, let's take that as a lesson and not start saying, well, gee, maybe we can nail it down to one year, 11 months or one year, 10 months. This is a big deal.
The other piece I just wanted to observe here in terms of our time lines, and I'm not arguing with the nature of our recommendations is that there was a noticeable absence of any payer in any of these panels today.
MS. HUMPHREYS: They were all invited.
DR. COHN: I know, and for the record, we received a written testimony that stated we shouldn't do anything. This is from Blue Cross Blue Shield. We shouldn't do anything until all the transactions are complete and implemented.
MS. HUMPHREYS: Somebody said that today, too, right? I heard you.
DR. COHN: And that those should be other actions taken. All I'm saying is that we should not, based solely on the testimony today come away feeling very comfortable that one year 11 months, one year 10 months, is going to work.
MS. HUMPHREYS: Well, you know, I mean, the committee has to do what it does, but I really do feel that it would make a very sensible recommendation and you guys will have to figure out what it is you want to say about the dates, I think maybe something like Bill's formulation takes care of both issues which is sort of saying, you know, forever is not a great option here, but we did hear pretty strongly that at a minimum of three years and some people don't think that's enough, after the final rule is passed.
I just feel, though, that it's getting a firm proposal out there that says, okay, everybody, we are proposing move to all of these, we are proposing to move to them at the same time. We are proposing to give you a minimum of two years to implement so we can take care of all these issues that we have heard about, it probably, it seems to me, the only logical step even though I think that, you know, there's a greater than zero percentage chance that you will get comments that will leave the committee perhaps in the future to recommend a different course of action but I think in terms of moving this forward.
DR. YASNOFF: I think that while we're, certainly our charge is to move these standards efforts forward as quickly as possible, I don't see how you can conclude from this testimony, and we may not have heard from everyone so we can't say that the testimony we have heard is representative of everyone, but we didn't hear from anybody that they could do it in less that two year, after the final rule. That's what they are all saying. They are saying we need at least two years after the final rule. Except 3-M, right, with two exceptions but I think that we have a number of people saying with reasonable definitiveness that they need those two years and so, but I think we can make a recommendation that covers that and still conveys our sense, and not only our sense of urgency, but the sense of urgency that the testimony has provided as well that we also are hearing that we shouldn't wait until 2007, 2008; that these systems are both seriously broken and we do need to do something but the reality is that it does take time to do these things and that this is the compromise that we recommend and as Betsy says, if there are major problems with that recommendation, I'm sure that the comments will reveal them.
MR. AUGUSTINE: I agree with what both Bill and Betsy have said about the proposal and also the floating of a date. I think it not only alleviates the best compromise as Bill described but it also conveys urgency of action.
DR. ZUBELDIA: Okay, part of Betsy's proposal is to say both ICD-10-PCS and CM together at the same time. Personally, I think we are ready to do that except for out patient procedures.
MS. HUMPHREYS: But I think if you recommended the replacement in the context of the current rule which basically limits the use of volume three of ICD-9 to inpatient and you recommended the replacement, the movement to PCS for in that context, that even though other recommendations might come down the pike in the future, then that to me would fix what is the most broken situation and also limit what may be extremely valid concerns about issues of implementing PCS in other environments.
My personal opinions about all this aside, that seems to me to be the recommendation that would make sense from the subcommittee.
DR. ZUBELDIA: So I would like to get a sense from the subcommittee if we are ready with that decision and say measuring the current use of ICD-9 to migrate to ICD-10-PCS and CM for the same type of uses that we are using ICD-9 today. Is that the consensus? Okay, I see everybody nodding.
DR. COHN: I have to recuse myself but other than that.
DR. ZUBELDIA: There is a couple of side issues. One is the recommendation to use the DSM-4 descriptors along with ICD-10-PCS since ICD-10 already points to -- 10-CM, I'm sorry -- since --
DR. COHN: I don't think that that's really part of this letter. Do you think it's part of this letter?
DR. ZUBELDIA: Well, I see it generally points to the DSM-4 descriptors.
MS. HUMPHREYS: Yes, but I guess my feeling is that what you are recommending is in the PCS, I mean, in both of these recommendation, you are recommending that, you know, publish, that part of the business of publishing these to be used would be, there would be coding guidelines to the extent they are needed for both of them and if the coding guidelines for 10-CM include reference to DSM, doesn't that --
DR. ZUBELDIA: We don't need to express it separate? The ICD-10-CM and PCS with the corresponding coding guidelines.
MS. HUMPHREYS: That's what comes to my mind right off the top of my head.
MS. GREENBERG: I mean, you do have a separate issue where APA has made a request related to 9-CM and some kind at least, not a rule so much, but some kind of letter or ruling or the department that because 9-CM will be implemented initially with the HIPAA transactions and they are concerned about the relationship there.
MS. HUMPHREYS: I had a question about that. I really did have a question about it. I gathered that Steve was commenting that without this, there would be some great disruption but I couldn't imagine why there would be a disruption because all that matters is that when you submit your claim you have the appropriate ICD-9-CM code on it and how the individual provider is coming up with that code and sticking it on its system, as long as what gets sent with the HIPAA transaction is an ICD -- I continue fully understand it either. I just felt that it was important.
DR. COHN: The question, the note I made is whether or not we should ask the Secretary to sort of notify the APA that a crosswalk between DSM-4 and ICD-9 is compatible with the HIPAA standard. I think that's about as much. I mean, Donna, actually were producing some work on this when I -- I apologize, we are sort of short circuiting your white paper.
MR. PICKETT: From the last meeting, we were working on a white paper that initially was to make a distinction between coding guidelines and diagnostic criteria. Based on today's presentation, what kind of stuck in my mind was that, and after re-reading the HIPAA rules several times, if the Secretary is to adopt code sets, medical code sets, I think the idea of adopting diagnostic criteria is really outside the scope of HIPAA and the transactions and code sets issue so that it did concern me when I was listening to the presentation this morning and how that would affect the further work on the white paper that we were developing.
As I mentioned to Simon at lunch time, basically what had been drafted, I kind of said, we need to start over because there's a lot more information that was provided in this morning's testimony that could logically find its way into the white paper.
MS. HUMPHREYS: Yes, it did seem so me, and we went onto another topic because I was actually going to ask Steve this. I don't see in, in fact, in the current environment you do your thing with DSM-4 and what is generated on a code basically is, in fact, also an ICD-9-CM code which I believe understood him to say and I have heard you say it as well, then whatever they are doing now generates with this code. They can still do it and all we need to do is perhaps what Simon suggested which is to recommend or it may be to draft for the Secretary or somebody to send to APA saying look, this is a non-issue and you are very free to tell everyone who is a psychiatrist or any other mental health professional who is using that it is a non-issue because what they are doing right now and generating an appropriate code. They can continue to do it and they are not in non-compliance with HIPAA and they can still use DSM to produce these and there's, that is a non-issue.
DR. ZUBELDIA: My correction for April is that some payers that take those ICD-9 codes in look at them with the ICD-9 look and they are not understanding the full description of the code according to the DSM-4, they are understanding of the description of that code according to the ICD-9 and therefore that can impact the payment which is okay, it just doesn't fit anymore and that was the problem. It's not that the coding is not good. The coding is good. It's just it's being interpreted wrong.
DR. COHN: Where would if affect payment -- out patient?
DR. ZUBELDIA: They would deny some payments based on an arcane diagnosis.
DR. COHN: Well, no, the reason I'm asking is we didn't hear any testimony from the American Hospital Association that there was a problem there. What we heard was from the provider and since they typically get paid, at least currently, on the basis of their procedure diagnosis as opposed to the procedure.
DR. ZUBELDIA: But they would deny a procedure as non-covered based on the diagnosis.
DR. COHN: Well, I don't know that, do you think this is a problem we can solve right now?
DR. ZUBELDIA: Hopefully it's a problem we can solve. I think the recommendation is to say to the payers you need to use the DSM-4 to interpret these codes.
DR. COHN: Are you comfortable going that far?
DR. ZUBELDIA: Yes because the entire psychiatry universe in this country is using the DSM-4 to code the claims so when the receivers receive the claim they should use the same guidelines to interpret what the code means rather than using the ICD-9 obsolete codes.
MS. PICKETT: We actually have two issues going here for mental health disorders that are treated in a hospital, DSM-4 is not used. ICD-9-CM is the coding classification that is used in hospitals. For mental health providers on billing on a professional claim, they have historically used the DSM-4 but again, the numbers look the same, however, the code titles may be different between the two. APA did try to harmonize the code numbers but some of the descriptors are different.
Historically the mental health providers have used DSM and quite frankly, there are many in the payer world, the billing world, that don't know the difference between an ICD-9-CM code and a DSM code because they look the same but if you get to the issue so they don't mean the same.
DR. ZUBELDIA: They don't mean the same. That's the problem.
MR. PICKETT: No, some things do mean the same, but some things are slightly different and I think some of that was mentioned in Dr. Merin's presentation this morning. The descriptor may be different. The code number may be exactly the same. The descriptor may be different in that the terminology used in the DSM code is a little bit more precise or they may have taken an inclusion term from the ICD-9-CM code and actually used that as the descriptor for the code. All this was going to come out in the white paper so we'll just kind of --
MS. HUMPHREYS: Well, the issue then would, okay, well, I have --
MS. PICKETT: But the other issue that was closely tied to their presentation had to do with diagnostic criteria and in DSM you have several components of DSM and one of those is the diagnostic criteria. Now, what they are saying is that the diagnostic criteria is linked to their DSM numerical equivalent of what looks like an ICD-9-CM code and they also cited fraud and abuse as being an issue because of misinterpretation but the issue is that I have only quickly reviewed some of the federal statutes and state statutes that they had referred to, and most of those that are not focusing on a code number. They are focusing on the diagnostic criteria and clinical descriptions to meet forensic and other legal parameters.
They are not saying if you have code 309.1, they are asking for the specific clinical criteria so somehow that issue can be disaggregated but it still means that if they were to report a DSM-4 code, or billing applications, if systems have been retooled to only capture and validate ICD-9-CM codes, the code numbers would be okay but the descriptors may kick because the descriptions may be different.
DR. ZUBELDIA: The descriptors don't go in the HIPAA procedures. Only the code goes and that is a concern. They are sending a code that my be misinterpreted by the payer that gets it.
MS. HUMPHREYS: Okay, so this is a real issue. I'm sorry. It's taken me all this long to figure it out so that means that right now they send a code, they send some additional information and sometimes the payers pay appropriate attention and sometimes they don't but at least in a number of cases, the payers do pay attention and now what they are saying is they won't be able to send the accompanying information in the HIPAA transaction and therefore every time it will be kicked out and inappropriate even though it is because of what they have done?
DR. ZUBELDIA: May or may not, depending on the code.
MS. PICKETT: Depending on the code it may or may not kick because again, there are some differences between the descriptors in DSM-4 and ICD-9-CM.
DR. ZUBELDIA: They are not huge definitions. They are subtle distinctions.
MS. PROPHET: I was just going to ask Donna if it's true that, isn't that the less of a problem in ICD-10?
MS. PICKETT: We completely harmonize between DSM-4 and ICD-10-CM. It is a non-issue.
DR. ZUBELDIA: But the codes are different. ICD-10 codes are not the same as the DSM codes so there's no possibility of misinterpretation.
MS. PROPHET: Right, but the issue of the differences in the code descriptors have been harmonized. They mean the same thing now.
DR. ZUBELDIA: They mean the same but the code is different.
MS. PROPHET: In ICD-10, the problem goes away, but unfortunately we've just said that as soon as that this will, there'll at least be a two or three year hiatus when this situation will be a problem.
MS. TRUDEL: Actually that was my point that this seems to be on interim issue and maybe it is something where we want to wait for this white paper to get a whole sense of what all the issues are and treat that separately rather than putting it in these recommendations.
DR. ZUBELDIA: There has been a request from the APA that if the DSM is adopted as the HIPAA code set that they would like to be designated a DSMO to maintain the code set. And the discussion should be about 30 seconds.
MS. HUMPHREYS: They would then be a code set maintainer.
MS. TRUDEL: They would be a code set maintainer, not a DSMO.
MS. GREENBERG: We have all made that mistake at least ten times except for Karen who never makes it but I finally learned it.
DR. ZUBELDIA: Okay, I think there's nothing else for tonight. We can go home right on time and then --
MS. GREENBERG: Are you -- is there some suggestion here that the subcommittee might want to take a letter to the June meeting?
DR. ZUBELDIA: Yes, I think we need a draft a letter and maybe after we hear the testimony tomorrow we can zero in on the contents of the letter and we need to draft a letter to the Secretary to bring to the June meeting.
DR. COHN: I think the letter is probably best being done offline, I would imagine because we also have to have the letter on the NPRMs that are coming out and all that. I guess the question I'm trying to think of who could assist with drafting the letter. Betsy, is that something you will be able to --
MS. HUMPHREYS: I will discuss that with staff, with our copious subcommittee staff.
DR. COHN: Okay, we'll let our chief staff assist us with this. I think we've gotten the sense of what it is we want to do and obviously we'll have a chance to edit it.
DR. ZUBELDIA: For tomorrow, I would like to make a change in the agenda. Let's start at 8:30, and we have a lunch break from 12:00 to 1:00 and then we continue for an hour and a half, one hour and if it's okay, maybe we can work through from 1:00 to 2:00, I'm sorry, from 12:00 to 1:00 and finish at one instead of having a lunch break and reconvening after lunch.
MS. HUMPHREYS: Well, wait a minute. Who's the first testifier?
DR. ZUBELDIA: That's tomorrow afternoon.
MS. HUMPHREYS: Oh, 8:30 is not changed. Got it.
DR. ZUBELDIA: Start at 8:30 and end at 1:00. Depends on how things go.
DR. COHN: Kepa, thank you very much. It was a good day. The meeting is adjourned.
(Whereupon, the meeting was recessed at 5:31, to be reconvened the follow morning.)