Hubert H. Humphrey Bldg.
Washington, D.C.
List of Participants:
Barbara Starfield, M.D., Chair
Jeffrey S. Blair
Simon P. Cohn
Kathryn L. Coltin
Kathleen A. Frawley, J.D.
Lisa I. Iezzoni, M.D.
Vincent Mor, Ph.D.
Clement J. McDonald, M.D.
George H. Van Amburg
William Braithwaite, M.D.
Marjorie S. Greenberg
Robert Moore
TABLE OF CONTENTS
Page
Call to Order 1
Panel on Professional Health Care Providers
Dr. Kay K. Hanley 5
Dr. Larry Griffin 10
Dr. Charles Vanchiere 17
Dr. Lois Hoskins 23
Karen Pace 27
Robert Owens 32
Dr. H.C. "Moon" Mullins 36
Panel on Health Care Facilities
Nelly Leon-Chisen 87
Margaret Stewart 91
Beth Wing 128
Gary Knoble 136
Meryl Bloomrosen 141
Susan Seare 136
A F T E R N O O N S E S S I O N (1:05 p.m.)
DR. STARFIELD: Good afternoon. I'm Barbara Starfield, chair of the Subcommittee on Health Data Needs, Standards and Security.
We are halfway through April 15, and starting a day and a half of meetings on the subject of medical clinical coding and classification and issues and implementation of the administrative simplification provision, PL-104.191, the Kassebaum-Kennedy legislation.
What we are going to do is to move around the table -- these will be relatively informal hearings, although we will keep to schedule. That means eight minutes per testifier, and then limited time for discussion. But what we will do is, just so everybody knows each other, we will go around the table and around the room. I know you have all signed in, those of you sitting in the bleachers, but when you introduce yourself, could you not assume that we have all read the list of sign-ins. Say it loud, so that we know who you are. Thanks very much.
I'm Barbara Starfield.
(Whereupon, introductions were performed.)
DR. STARFIELD: Thank you very much, that's wonderful.
Just one announcement. We have many, many handouts. Everybody who came to testify this afternoon, there are a lot of you, everyone brought their own written version of their testimony, and we didn't have enough copies. It is amazing that they carried as much as they did. But I do want to remind you that everything is on the home page, the committee's home page. If you look at the reverse side of the agenda, you will find out how you can access that.
I just want to remind everybody that this National Committee on Vital and Health Statistics has a long history. We have the responsibility of advising the Secretary on everything having to do with health data needs in the country and anything to do with health services. That is more than the kinds of data we are talking about today; it also includes a lot of the survey data, all data that relates to national health needs or health care needs.
But one of our foci, especially this year when we have so many responsibilities as a result of the administrative simplification provisions, we are focusing very heavily on data that is going to be electronically transmitted. It is a monumental task, one which we take very seriously, and one that is enormously challenging to us. It is very easy to get lost in the morass of detail and forget the principles.
So what I would like to do is just to spend a minute and do a one-pager for you, just to remind you about the principles.
Challenges for administrative simplification, the goals, the principles and the mechanisms. First of all, uses of information. We are not interested in data unless we can use it for information.
What kind of information do we need? We need information for rational deployment of resources, personnel, facilities, dollars, any other kind of resources that you can think of. We cannot rationally deploy resources unless we have information on what those resources should be and where they should go.
We want to promote better quality control, better service delivery and better outcomes of care. We really can't do that without information.
Third of all, we want to advance knowledge. Evidence based medicine is the word for the day, and will be for a long, long, long, long time to come. Without information, we cannot advance knowledge.
So those are the uses of information, the things that we strive for and we think about as we do our work on this committee.
Principles for administrative simplification. First of all, in order to do the three things that I listed above, we must be able to link data across places. That is, we must have an ability to follow people across different sites and levels of care, and that of course has lots of implications for the way in which we code data. If we need to link people across places, we need to make sure that there are ways of translating information from one type of place to another type of place.
The second thing is, we need to link across time. We need to have the ability to trace peoples' problems from their onset to their resolution. That has the same kinds of implications. You cannot trace things from their onset to their resolution unless you have comparable kinds of data that goes across time.
The third principle is to link inputs to outputs, to understand why and how things happen. It is not only to describe what happens, because lots of things happen; you have to have a way of linking what you do with what is the outcome.
Now, to add to our challenge, we have to do the above while simplifying the way in which we record, use and share data, and we need to improve systems for maintaining confidentiality and security of data, because we probably still haven't -- definitely still haven't reached the ultimate in this goal of our work.
So I would ask you please all to keep these in mind as we go through the day, and maybe keep them in mind as you move through your work.
Yes?
PARTICIPANT: This is considerably larger than the scope of the administrative simplification in the legislation.
DR. STARFIELD: No, it's not larger. The legislation has to be put in this context, as far as this committee is concerned the legislation is put in this context. We happen to think that administrative simplification should help us do these things.
Does anyone have any questions about the agenda or the way we plan to proceed? If not, I think we will start. The first presenter is Dr. Kay Hanley from the AMA, a fellow pediatrician.
DR. HANLEY: Thank you. Madam Chairman, members of the committee. I am Kay Hanley. I'm a private practicing pediatrician, front lines in the trenches, in Clearwater, Florida. I am here to offer my perspective as a practicing physician on the medical and clinical coding and classification issue.
Implementing the administrative simplification provisions of the Kassebaum-Kennedy legislation will be as you surely know, no small task. Coding plays a very important part of my practice. It is my experience that when it comes to coding, first and foremost, it pays to get it right the first time.
On behalf of practicing physicians such as myself and the AMA, I strongly recommend that the physicians' current procedural terminology system be recognized as one of the initial standards for administration transactions as required by the Health Insurance Portability and Accountability Act of 1996.
I would like to mention that the AMA's written statements, which have been submitted to you, contain answers to the committee's specific questions. In addition, tomorrow the AMA will testify on the more technical aspects of coding.
I am pleased however to be before you, to offer the views of a practicing physician. In my practice as a pediatrician, I very nearly exclusively use the CPT and ICD-9 systems. I just wish I had had the opportunity to bring Debby with me. She is our office administrator. She would tell you the importance of having a coding system like CPT that works. She would tell you that CPT does a pretty good job of describing the services that I perform. She would say that having a system like CPT makes the administrative practice of medicine much simpler.
She would tell you that insurance companies that follow CPT make more accurate payments. She would also tell you that the ICD-9 becomes much more difficult when you get beyond the fifth digit.
As you know, the AMA has successfully operated and maintained the system for over 30 years. Prior to CPT's development and copyright, there were over 250 procedural coding systems. This was inefficient and detrimental to getting reliable data on medical services.
Today, acceptance for CPT is high. Estimates are that over 95 percent of all services provided by physicians are reported using the CPT coding system.
The AMA has a tested effort which continuously improves CPT through the CPT editorial panel, which is the entity responsible for its maintenance. The editorial panel is comprised of representatives of groups such as Blue Cross/Blue Shield Association, the Health Insurance Association of America, HCFA and the American Hospital Association. The panel has been expanded to include non-M.D. health providers and doctors with more specialized knowledge of managed care.
The panel itself is supported by a voluntary network of nearly one thousand experts. These committees represent medical specialty societies and organizations such as nursing, podiatry, psychology, physical therapy, chiropractic and a variety of other health care providers. CPT has met with both private and public sector critical review. For example, in 1993, CPT was chosen by NCFA as the procedural coding system for the Medicare and Medicaid program. The AMA has maintained CPT under its agreement with HCFA, providing annual improvements at no cost to the government.
CPT has shown its responsiveness to the coding needs of the changing health care system. One of the most important events in the evolution of CPT was the decision by Congress to restructure the way that physicians are paid under Medicare. In response to the transition to the RBRVS based physician payment system, the CPT editorial panel revised the codes for visits and consultations. In a recent survey, 95 percent of the physicians said they believe it is important that the coding system that they use should be developed by the medical profession.
At a recent meeting of the practicing physician advisory committee, a group of practicing physicians and providers that advise Secretary Shalala on Medicare and Medicaid issues, administrative simplification was on their agenda when the discussion turned to coding.
Incidentally, it was the chiropractic representative to PPAC who said that they couldn't imagine the government developing another coding system.
In short, PPAC was unanimous in its support for the need to continue to use the CPT system for the professional community involvement.
The cost of implementing any system other than CPT in my practice would be enormous, and for the health system at large. A recent update to a study conducted by Coopers and Ligrand estimated the cost of such a change across the system to be about one billion dollars. Such a change would also be a huge and expensive imposition on practicing physicians in terms of learning and implementing a new system as well.
On the issue of diagnosis code sets, the AMA supports the continued use of ICD-9CM. We have concerns about doing coding to the highest level of specificity versus the highest degree of certainty. It is my understanding that AMA staff is working with NCFA on this.
To date, there have not been sufficient involvement of the medical profession in issues pertaining to ICD-10. Therefore, it is not now possible for us to draw solid clinical conclusions about the ICD-10 system.
In conclusion, the cost of implementing new systems of coding would be enormous. CPT remains the clear choice and preference of practicing physicians. Thus, the AMA believes that the CPT is clearly the only coding system that is consistent with the Kassebaum-Kennedy code set requirement for the reporting of physician and professional services.
Again, thank you for the opportunity to be here today. I would be happy to respond to questions when the opportunity arises.
DR. STARFIELD: Thank you very much. I think we will hold questions until we have finished with the panel. Dr. Griffin?
DR. GRIFFIN: Thank you, Dr. Starfield, and ladies and gentlemen of the panel. I am Larry Griffin. I am an obstetrician-gynecologist and the director of program services for the American College of Obstetricians and Gynecologists, based here in Washington, D.C.
ACOG is an organization that represents more than 37,000 physicians who specialize in the delivery of health care to women. We are pleased to have an opportunity today to comment on the use of coding and classification systems for the administration of financial transactions.
The vast majority of our members are in the active clinical practice of medicine. They serve as the initial entry point into the health care system for a great number of women of reproductive age in the United States. In addition to providing a full range of reproductive health services, obstetrician-gynecologists provide a wide range
of primary and preventive health services to women.
ACOG has supported and continues to support coding and classification systems that permit accurate collection of data and appropriate reimbursement for services rendered. Our comments today attempt to address the questions put forth by your subcommittee, and reflect the impact of those proposed coding system changes on the provision of health services for women.
Clearly, physician members of the American College of Ob-Gyn predominantly use CPT for codes to report procedural services and ICD-9CM diagnostic codes to describe the justification for those procedures performed. These codes generally allow accurate reporting through the use of codes, qualifying CPT, numeric modifiers and diagnostic linkages of services rendered by physicians in any setting, that is, office, in-patient hospital, out-patient hospital, et cetera.
While CPT and ICD-9CM codes have been generally recognized by the majority of third-party payers, however, some payers still mandate the use of internally developed procedural codes which are used to control or monitor specific targeted services rendered to their beneficiary population. In addition, some payers don't recognize all the available diagnostic codes in either ICD-9CM or the procedural codes in CPT, particularly their modifiers, thus limiting the collection of accurate data.
The current strength of the CPT coding system, as Dr. Hanley just said, is the involvement of physicians throughout the process. This involvement allows physicians who actually perform the procedures listed to determine the appropriateness of the procedure or the accuracy of the condition being coded, to recommend the current and correct nomenclature when new codes are developed or old ones are revised.
The current weaknesses of this system are the length of time it takes to add new codes when regulations change, the fact that non-physician or physician-directed services cannot often be appropriately reported, and the fact that the codes are considered to be by the AMA to not be in the public domain. This latter weakness has occasionally hampered efforts to educate members of various medical communities regarding accurate and correct coding practices. The ICD-9CM diagnostic coding system has several weaknesses, at least one which we believe will be corrected with ICD-10CM. First, there is limited ability to revise or add new codes based on morbidity conditions found specifically in the U.S. Secondly, the ICD-9CM codes are reported in different ways by physicians and hospitals. For instance, a physician will link a specific diagnosis to a procedural code, whereas hospitals will list diagnoses from the patient's chart, often based on the severity and importance of the condition. If this order coincides with the primary reason the physician admitted the patient to the hospital, there is no conflict, but often there is. The strength of ICD will lie in ICD-10 because of direct input from physician users about conditions that cannot be currently coded. In our opinion, ICD-10CM will allow better information to be collected once it is implemented.
Again, however, the greatest potential weakness of any and all coding systems in our view is the inability to report information or capture complete data. Again, many third party payers do not currently capture more than one diagnostic code and often cannot accurately link procedures and diagnoses, even with electronic claims submission that refuse to recognize all the diagnostic or procedural codes currently reportable.
We strongly recommend that if a single coding or classification is eventually adopted, a uniform set of rules and definitions must also be adopted, and required by all parties involved in health care information or transaction.
At this time, we believe CPT-4 and ICD diagnostic codes should continue to be used to report physician services. Changing reporting systems before there is adequate time to prepare and accept them would be equivalent to changing to the metric system overnight.
Clearly, the most accepted system at this time is CPT-4 and ICD-9CM. Medical specialty societies and county and state medical societies, the AMA, have continued to increase the resources and effort put into physician education to insure that physicians accurately utilize these coding systems. To make a rapid change overnight would not only cause dramatic costs to an community practice, but the institutions charged with educating those physicians as well.
As we have noted above however, the acceptance of the current coding systems, that is, CPT-4 and ICD-9CM, has not precluded payers from developing their own internal coding systems for financial and administrative purposes. This not only adds to the administrative burden, it often makes it difficult to cross walk the various databases, making it difficult, as you have originally said in your prelude, track a patient across various health care systems and sites of services.
ACOG has been an active participant in the process of developing ICD-10CM diagnostic procedures that would be used in conjunction with gyn and obstetric services. This new system appears to allow for the expansion of available diagnostic codes to better serve the reporting needs for obstetric and gynecologic services within the United States.
We at ACOG have recently begun a review of the ICD-10 PCS system, and will soon be making comments to 3M Health on this system. Based on a preliminary review, however, our opinion is that it is cumbersome to use and will almost certainly lead to incorrect coding unless those responsible for assigning and recording procedural codes are exceptionally well versed in the mechanics of the procedures and the terminology used by various physicians. Using this system, procedural codes can only be formulated after careful scrutiny of the operative report. This clearly will increase the time required to record in patient data.
We have also noted that although there has been an attempt to standardize definitions, some basic procedures may in fact have more than one definition, and the definitions used often conflict with generally accepted medical terminology. Additionally and more importantly, if a decision was made to select this series of codes, that is, ICD-10 PCS for reporting all procedural information, administrative costs of doing so for physician practices could be astronomical. This could easily lead to upward pressures on the price of health care services, as well as a loss of valuable data.
Part of the administrative burden in today's health care environment involves the additional and needless cost of having to maintain extensive information about reporting the same service once again to each different third party payer. In spite of the fact that the majority of payers accept and recognize CPT and ICD-9CM codes, the rules under which these codes can be reported are varied and are often at odds with one another. Again, all of this variance translates into additional needed resources which add to practice expense and over costs.
The maintenance of medical and clinical code sets will need the input of physicians to be effective. Funding for such maintenance can be reduced when input is obtained through the voluntary participation of medical specialty societies and other interested parties.
The code sets will be more useful if kept in the public domain for contributors to the process, because it allows for unhindered access to the codes and standard guidelines for educational purposes. If the codes are kept in the public domain, even with copyright restrictions, there may also be an opportunity to offset the costs of maintaining the codes by charging a user fee to those companies or organizations that do not contribute to their development and maintenance, but who profit in the production of coding materials.
Changing to a different coding and classification system may involve several additional expenditures for physician practices. It clearly would involve additional staff and physician training, and perhaps could even lead to the need to substitute higher and higher qualified and more highly paid coding staff for current staff available.
It is the opinion of the ACOG that the current CPT/ICD-9CM system meets the needs of coding adequately today, that it currently represents the federation of medicine and practicing clinicians, that there are modifications ongoing to make the weaknesses within the system improve. We are concerned about rapid dramatic changes without implementation of the safeguards that we have tried to mention here today.
Madam Chairman, I would be glad to answer any questions. I appreciate greatly the chance to be here and address the subcommittee. Thank you.
DR. STARFIELD: Thank you very much. Dr. Charles Vanchiere.
DR. VANCHIERE: Good afternoon, Dr. Starfield and members of the subcommittee. My name is Charles Vanchiere. I am a pediatrician and represent the American Academy of Pediatrics, an organization of 53,000 pediatricians, pediatric surgical specialists and pediatric medical sub-specialists who are committed to the attainment of optimal physical, mental and social health for all infants, children, adolescents and young adults. I want to thank the subcommittee for inviting the Academy to participate in this timely discussion on the future of clinical coding and classification systems used in the administrative transactions.
In testimony today, I will stress the importance of proper clinical diagnostic and procedural coding and other classification systems, and the significant impact that their use will have on children's access to quality health care. The Academy also stresses that any coding or classification system employed for administrative transactions guarantees continuous review by the entire medical community, as well as promote the capacity for the documentation of pediatric procedures, diagnoses and diseases.
Even as we speak, the Academy and other medical societies are working actively within the CPT process and with other medical societies to examine and recommend solutions to coding disparities that may exist for pediatric care. With continued participation of medical societies and health care professionals, these disparities will eliminated and children's conditions will be given balanced representation in the classification systems.
We believe that any deviation from the current system of procedural code may damage the recent advances made for pediatric specific services. We encourage the subcommittee to maintain the current system of procedural coding while promoting its applicability to pediatric care.
Historically, pediatric services have not always been incorporated in the CPT to the degree we feel necessary. This system of procedural coding has been limited by the substantial number of advances in medicine and technology for which procedural codes have been required. Often these procedures need to be incorporated into the existing coding and nomenclature structure within a condensed time frame. The process of review and acceptance of new revised procedural codes within the CPT system also has been constrained by the limited space available for expansion.
However, it appears that those who maintain and update CPT are aware of the inefficiencies of the current system, and that these issues are being addressed. The American Academy of Pediatrics also has played an active role in reviewing various chapters and sections of the impending International Classification of Disease, Tenth Edition. We believe that these segments that we have reviewed represent a significant improvement over the current ICD-9CM coding system, and lay out in category specification for pediatric diagnosis and disease documentation.
The staff and members of the various governmental committees that maintain and operate the ICD-9 have been receptive to many of the AAP's comments.
It is essential that any diagnostic classification system require the use of ICD-E codes to document external causes of injuries and factors influencing the health status of children. We support the preliminary indications which suggest that ICD-10 will expand upon this foundation, and provide additional classifications to further clarify external causes of illness.
The benefits of using and reporting ICD-9 codes, including E codes, by physicians and hospitals are numerous. The Academy believes that more complete and accurate surveillance of injuries to children will help health professionals including pediatricians plan and implement health promotion and prevention programs, as well as evaluate program cost and quality.
The current ICD-E codes achieve this in part by assisting health care professionals in determining and examining risk factors associated with childhood injury, documenting the morbidity and mortality associated with these risk factors, and identifying demographic and geographic differences.
It is important to underscore that E codes are a source of data for health care professionals working to make health care more effective, and impart an effective means of assessing the impact of such injury prevention legislation such as bicycle helmet requirements, environmental health laws, seatbelts, car seats, et cetera.
Mental health diagnosis and services also must be appropriately documented, classified and coded within a medical clinical administrative transaction classification system. The diagnostic and statistical manual for primary care, child and adolescent version, is the most comprehensive source of mental health classification information. It was developed to facilitate the understanding and coding of child and adolescent mental health issues by primary care providers. The manual is both an operational instrument for the diagnosis and classification and a mechanism for coding of these disorders.
Coding and classification systems should also be sophisticated, yet practical enough to facilitate risk adjustment at both the individual patient and population levels. Administrative data must promote the linking of data at the patient level and over time. With such linked data, one can identify conditions treated previously that could thus be considered chronic or pre-existing. One can also look at patterns of diagnosis over previous hospitalizations or health care encounters to draw inferences about the course of illness. Longitudinal or time-linked data will facilitate patient tracking across both services and time.
The ICD-9 does not paint a complete clinical picture of all dimensions of risk. This is aggravated by the fact that CPT does not readily link to ICD-9. New coding and classification systems must be able to overcome these problems to insure the successful use of risk adjustment mechanisms in capitated environments.
It is crucial that any standardized coding in classification systems selected for administrative transactions include annual updates through a process of quarterly review that encourages physician participation. In order for physicians to incorporate a system of coding and classification to their practices effectively and efficiently, the system must be universally accepted and exclusively adopted, not adapted, by all health care providers, insurers and federal and state agencies.
Any system that is promoted for use within all phases of medical and clinical coding should be kept within the public domain without the restrictions associated with copyright laws to guarantee accessibility and availability of information for use by health care providers as well for the purpose of education research.
We believe that as the clinical practice of pediatrics has evolved, so have the current coding and classification systems. Familiarity with existing systems has created a level of comfort within the pediatric community that will only continue to grow, insuring the heightened use of these coding classification systems. However, the Academy questions the ambitious timetable for implementation of new systems.
We believe a 24-month period may not provide sufficient time for pilot testing and the incorporation of necessary modifications derived from pretesting the coding instrument. Additionally, costs for the development and implementation of a new or modified system will be astronomical for hospitals and physician practices alike, not to mention the coordination that such a dramatic and comprehensive operation of a coding practice would require.
The Academy appreciates your consideration. Thank you.
DR. STARFIELD: So far, so good. You have all been wonderful in keeping to your time, and very concise and very clear. Thank you. Okay, Dr. Lois Hoskins.
DR. HOSKINS: Thank you, Dr. Starfield and members of the committee. I appreciate being able to be here and present the opinion from the American Nurses Association. I am a member of the American Nurses Association's steering committee on databases. I am also a faculty member in the School of Nursing in Catholic University of America.
I am reporting here on nurses as users, not as developers. I have a colleague who will report tomorrow as developers. This is only as users of medical and clinical classifications.
Registered nurses as users of medical and clinical classifications for claims processing and other administrative purposes reflect only a small group. As a matter of fact, only 6.3 percent of the two and a half million registered nurses in the United States as of March 1996 were estimated to have formal preparation to practice in advanced nursing positions, and these are as clinical nurse specialists, nurse practitioners, nurse midwives and nurse anesthetists.
Legislation varies state by state under a medical framework, controlling who has prescriptive authority and who can submit billing claims. Nurses licensed with this authority are the only ones with HCFA provider identification numbers.
Now, my responses are patterned upon the questions which you submitted, and I will speak to them more or less in a bulleted response.
Question number one, related to the medical and clinical codes and classifications that nurse use and what are their strengths and weaknesses for us. Nurses primarily use the ICD diagnostic codes and CPT ICD-PCS procedural codes. These codes are inadequate for capturing nursing practice data.
Anecdotal comments that I can report from school nurse and pediatric nurse practitioners include: some things are not there, and it is a struggle to make others fit. From a nurse psychotherapist who uses DSM-4 and ICD-PCS: the DSM provides more information and makes selection more accurate. She treats eating disorders and states that if she used the ICD, she would have to code anorexia and bulimia under an appetite code.
Your question number two, given the time frame of HIPAA, what codes and classifications do we recommend as an initial standard, and specific suggestions regarding coding and classification. We want to say only that we are on record with NCVHS as supporting a single procedure classification code.
Question three, if some version of ICD is used, which should it be? We understand that some other revisions in ICD-10CM will provide more codes useful to nursing, but presently we have insufficient information to comment on this.
Your question number five with respect to the ongoing maintenance, responsibility, et cetera of medical clinical code sets, public domain versus in the private sector. We would state, the ongoing maintenance and responsibility for code sets and their continuing development should have government supported funding and should be open to entrepreneurial activities. It should not be dependent upon voluntary activity.
There are issues of dissemination, access and adoption in relation to public domain versus privately owned, with or without copyright. The American Nurses Association supports the public domain approach. We believe that having a copyrighted product may assure regular updates.
Anecdotally, nursing only began developing its vocabulary in the 1970s. Its most well-known vocabulary, that of the North American Nursing Diagnosis Association, was in the public domain until 1992. During that 20-year period, it became widely integrated into nursing education and recognized throughout this country and abroad. It is now copyrighted primarily to command income from licensing to publishers and vendors.
Question number six relating to resource implications of changing from the current coding and classification systems, costs and benefits of such changes. Initially, the costs of implementing new or different systems and user discontent will be high. But the gain in terms of adoption of a standardized nomenclature will be overriding. In the long run, standardization will simplify the processing of data and lead to more effective care and management.
In conclusion, the current systems for coding health conditions, diagnoses, services and procedures do not do a good job of representing the nursing discipline. The data captured from the medical perspective accounts for only a small portion of the variance in outcomes. Policy decisions based on current billing codes that do not account for nursing interventions lack comprehensiveness. We would like to have more of our vocabulary incorporated into the system which will become a standard.
We have a nursing vocabulary and a nursing intervention vocabulary included in SNOMED International, and we have ANA recognized vocabularies in the UMLS. UMLS is the only system that cross walks nursing terms.
Thank you.
DR. STARFIELD: Thank you very much, Dr. Hoskins. Karen Pace, please.
MS. PACE: Thank you. On behalf of the National Association for Home Care, I would like to thank the committee for giving us the opportunity to participate in this process, and also to thank you for accommodating my conflict in schedule and allowing me to participate in this first panel.
I am the vice president for research and regulatory affairs, and I should also mention that my background is nursing, and I practiced in the home care field for many years before joining the association.
We are very interested in staying informed of your work in progress and assisting in any way possible. We are the largest trade association representing home care and hospice organizations nationwide. Since home care is the fastest-growing segment of the health care industry, we think it is important that any of the coding and classification systems adequately incorporate the needs of home care providers.
We cannot respond to all of your questions. We don't have enough experience in the field of coding and classification systems as some of the other settings and disciplines, but we do offer the following comments for your consideration. We will be happy to provide any additional home care and hospice specific information that would be helpful as you respond to the charge of the committee. I will also be following your questions in terms of responses.
In terms of the codes and classification systems that are now used in administrative transactions, home health agencies and hospices use the ICD-9 for identifying patient diagnoses. In terms of procedural codes, those that have been used in home health care were developed by HCFA for Medicare reporting, but these have really not been used since 1992. I did include those in an attachment, in terms of those specific treatment codes that were required for the various disciplines.
It is important to note that these were specifically developed for Medicare coverage services, so they don't necessarily cover the whole spectrum of home care services that might be delivered.
The UV-92 is used for billing Medicare, home health and hospice services, and therefore the data required in completing that is also familiar to home care agencies. HCFA has also established revenue codes for billing home health and hospice services. These are generally based on the discipline of services being provided. They are very broad and do not distinguish the types of activities or levels of intensity of service. Again, those are included in your attachment.
Then the HCFA procedural coding system is often used when supplies are billed under Medicare, Part B.
In terms of the question about recommendations for coding and classification systems, I think it is evident, given the short time frame, that it will be necessary to use code sets already in existence. However, we feel it would be imperative to develop standards for coding and classification systems and evaluate those existing systems against the standards that you set, in terms of choosing one system over another.
We believe that nationally mandated coding systems should be in the public domain with free and easy access for those that are going to be required to use them.
In terms of changing coding systems, as mentioned previously, very few code sets are currently used by home care agencies, and many of the existing code sets do not encompass the home care setting. We feel that code sets will only assist the home care providers with simplification if all payers must use and accept the same code set. We recognize that many home care agencies will have significant difficulty implementing any new requirements.
In terms of concerns about the process, we just want to make sure that home care and hospice issues are adequately considered and incorporated. It cannot be assumed that what works for hospitals or physicians will work for home care.
An example of this that we have already encountered is with electronic remittance advice. In the initial 835 and flat file used to provide information did not really meet the business needs of home care agencies.
In regards to the additional questions, we believe it is desirable to have one coding system serve multiple purposes to the extent possible. However, it should be recognized that the definitions that go along with the coding systems are often developed for a specific disciplinary setting and are not necessarily immediately transferable.
For example, home care agencies would not want the nursing home MDS applied to home care patients, but we could strive for a core uniform data set.
To what extent is your discipline or setting represented in coding systems? Home health encompasses a variety of disciplines, including nursing, PT, TO, speech language pathology and medical social services. Nursing and others have developed classification systems and codes which need to be considered for the home setting.
For example, nursing systems include the North American Nursing Diagnosis Association system, the Iowa Nursing Intervention classification system, Omaha and Georgetown home health classification systems. Currently, there is not consensus in the home care industry on which of these to use.
The National Association for Home Care has been working on a uniform data set for home care and hospice. At this time, it would probably be most applicable for classifying types of home care and hospice organizations.
HCFA has developed the outcome assessment information set with standardized patient assessment items for adult home health patients. HCFA just recently published a proposed rule requiring its use for patients served by all Medicare certified home health agencies.
In terms of practicality in moving to a single procedure classification system, I guess we believe it would be desirable, but can only tell you that we believe that most current procedural code systems do not adequately reflect home care services. So that will certainly by a challenge for us.
In terms of, should payers be allowed to restrict reporting of certain codes, we believe that payers should not be allowed to establish different reporting requirements. We believe that that would defeat the benefits of standardization to providers.
We thank you for the opportunity to comment, and hope that you will call on us with any assistance that we can provide as you continue your work.
DR. STARFIELD: Thank you. Robert Owens.
MR. OWENS: Thank you, Dr. Starfield and members of the committee. We are certainly happy to be here and have the opportunity to present on behalf of the American Dental Association and our member dental practitioners. The American Dental Association represents 72 percent of all practicing dentists, or about 146,000.
Rather than reading through the specific answers to the questions, if I may summarize some of the pertinent points throughout this document.
What we have in use today within dentistry is primarily the current dental terminology, the CDT codes and HCVX. HCVX is based on CDT and simply offers the opportunity to make modifications in between CDT versions.
In addition, dentistry is a little bit unique, from the standpoint that we do have a segment of our industry that also uses ICD-9 and CPT. These are typically in our oral surgeons and other type of practices that relate to dental procedures.
As far as where we go from here, we would certainly recommend that the continued use of CDT and HCVX codes in the initial regulation be that recommendation from this committee, and in fact be part of the regulation for HHS. We will testify tomorrow as a developer, but here as a user, that is the only national standard right now that is recognized throughout the dental profession for procedural codes, again emphasizing the national standard.
Relative to our oral surgeons, we would like to see in that market and the medical the continuing use of ICD and CPT or the coding structures to be endorsed for this initial regulation relative to those parts of dentistry that have a medical piece.
We encourage over time that we move to a single coding system, and again, we would encourage that that single coding system be CDT for dentistry. However, in the initial regulation, depending on what we have, that may not be possible.
It is also important to note that when we talk about a single coding system, we are not talking about a single coding system for all of health care. We are talking for each segment of health care. In other words, for medicine, for dentistry, for pharmacy, those coding systems may not be identical or may not even be under the same authority, but in fact we should look at developing a single one for each segment of our industry.
We know the CDT has proven itself time and time again. We know that our development efforts and inclusion of the actual users of these codes and the testing that goes out -- and we will get more into that tomorrow, but they have proven themselves time and again. We therefore would encourage the continued use of proven coding systems when we are really trying to get at administrative simplification.
In regards to using a standard for more than one purpose, the HIPAA or K2 -- many different names -- is designed for administrative simplification. The change or the selection of a coding system should be driven by that fact. It should be driven by administrative simplification versus the need for computerized patient records.
At some point over time, those two types of coding systems may evolve, and you may be able to get certain information. In fact, as we will get into a couple of other answers, in organized dentistry we have an initiative to develop coding systems for diagnosis. However, that is to be used in our computerized patient record, not in the administrative processes at this time.
We feel that over time, the natural evolution of these coding standards will take care of themselves. They may come together, they may not, but at this point in time, in order to get the administrative simplification act implemented and to really get the industry moving forward in the use of technology to effect this administrative business, it is important to select the existing standards.
As far as some of the highlights from some of the questions, for instance, in question two, we strongly believe that there may be some economies gained in the use of a single coding structure. But when you look at what it will take to implement that, although we have not done any specific cost estimates, in talking with the many software vendors out there, it is in the thousands of dollars of cost of change to accommodate that.
As far as weaknesses, the natural evolution that we have in our development process -- again, we'll get into more of that hopefully tomorrow, or we can as you wish, but the basic premise is that we involve the industry up front. The use of the coding systems in the standard diagnostic, any weaknesses that we would have in that area we are actively addressing as we continue to evolve both the CDT code set and by the development of the diagnosis coding system. That will give us additional capabilities in outcomes research, but we look at that not as a part of the administrative process, but as a part of the clinical record process, or in our case, the dental health record.
Then just in closing, if there are any other specific questions that you would like, we will be more than happy to do that. On behalf of the American Dental Association, we certainly thank you for the opportunity.
DR. STARFIELD: We thank you, too. Dr. Mullins.
DR. MULLINS: Thank you. I am Moon Mullins. I am a family doctor who practiced in a rural community for 19 years, and have been in an academic setting for the last 20 years.
I greatly appreciate the opportunity today to represent primary care practitioners, but specifically the family practice primary care work group of AMIA(?).
My testimony today will be based on several things: the recommendations of a consensus conference on standards for clinical vocabularies, an analysis of published studies, a recent informal survey of primary care informaticians and professional organizations, and my own experience.
There is just recently an increasing realization that clinical data is essential to provide complete and accurate information needed for both clinical and administrative services in a good health care system. My remarks are in the white binder. On the second page, there is a diagram from Karl White's classic paper, the Ecology of Medical Care, which indicates that primary care covers about 95 percent of a given population. So the capture of detailed information in primary care is what will give us a lot of detailed information from which clinical data sets can be derived to support patient care, clinical audit decision support and practice based research, and management data sets can be derived to support such activities as billing, accounting and so forth.
The data flow can be accomplished in a progression of several distinct processes, terming for reporting clinical data, coding for billing, epidemiology, statistical analysis and then grouping for the costing. Each of these require different structuring language, mapped to each other: a clinical vocabulary for determining clinical information, a coding scheme for in-coding and classifying information, and case mix coding for various groups. There is a diagram there pointing out the data flow progression.
But the key building block in such an arrangement is a fully functional structured ambulatory primary care medical record from which flows all of this data needed by everybody up the chain. There is really no other way to get that detailed comprehensive data except in primary care.
In November 1995, a conference sponsored by the family practice primary care work group of AMIA, entitled Moving Toward International Standards in Primary Care, Informatics and Clinical Vocabulary, was held in New Orleans.
This was really an extraordinary meeting. It was almost magical. We had 79 people from nine countries. We achieved an unlikely result, and agreed upon shared vision consensus on clinical vocabularies for primary care. Full details of that meeting are included in your handout under Tab 3. I recommend that you read that. Excerpts are in the preamble, which says that we endorse the concept development and evaluation of a standardized vocabulary.
I might say, one of the things we realized during that meeting was, we need a glossary. We may say coding and classification in the vocabulary, and I am speaking of a good clinical vocabulary at this point.
It was a framing issue, our vision statement, and for the purposes of this hearing, the recommendations for vocabulary. The overwhelming consensus of conference participants was that the development of clinical vocabularies is a process in progress. At this time, it would be premature, misleading and destructive to recommend any one vocabulary over another. It was recommended that all of the vocabulary examined, ICPC, READCODES and SNOMEDS be aggressively used by primary care practitioners at the point of care, while researchers use standardized methods to examine and evaluate them.
Available published studies really don't answer questions about the efficacy of the various available clinical vocabularies for use in the electronic clinical information system. These studies include my own study and studies of at least one member of the panel. It will remain nameless. Obviously, there have to be additional studies, and we think that for those studies to be meaningful, it must be collected at the point of service and actual practice, functioning electronic medical records populated by the coding scheme being studied, must be used, and primary care investigators must conduct those studies.
Now to address the questions. You asked what are we using, and what are the main strengths and weaknesses. Most primary care people are using ICD-9 and CPT. The weaknesses are that they are ineffective in clinical medical records for capturing clinical care. Those codings cannot capture the true content of primary care, and they cannot capture episodes of care, which has been alluded to here earlier.
What medical clinical codes do we recommend as initial standards, what specific suggestions would you like to see implemented? ICD whatever, or ICD and CPT are recommended as initial standards for coding and classification, not as a vocabulary but as a coding classification. For clinical vocabularies, it is recommended that ICPC, READ and SNOMED be cross-mapped and used as mentioned above.
The third question, if some version of ICD is to be used, we don't really have an opinion on that.
The fourth question, how would you deal with the current coding environment to improve simplification and reduce administrative burden while obtaining meaningful information. As mentioned above, the most efficient and effective way to get information that is meaningful to clinicians is for the clinicians themselves to enter in the data at the time and point of care in a manner that works for them, using a clinical vocabulary that supports the level of detail that they wish to record and retrieve.
Implementation of this strategy, I had mentioned earlier, would allow that information to be converted into code classifications.
How should maintenance be handled? It should involve professional groups. There has to be central responsibility, long-term stability. We think copyright is essential to prevent unauthorized use and corruption. If code sets are available, stable and reasonably priced, they could reside in either the public or private sector or in a combination of both.
What are the resource implications? If the cost of the clinical system is perceived by the primary care provider as first and foremost aiding that provider in patient care, it would be enthusiastically embraced and funded by the user. The use of reported clinical data for administrative and epidemiological use would then be relatively cost free at the user level. The benefits far outweigh the costs.
Number seven, do we have any concerns about the process being undertaken by the department. Like everyone, I guess there is a concern that there will not be adequate primary care input into this process. It is interesting that there are many scattered pockets of individual and institutional excellence in primary care informatics and electronic information systems, but it is minimally organized and has no infrastructure, as you may have seen when I rolled all this stuff up here today. It is a very thin structure that we have.
This makes it very difficult for you to identify appropriate people and activities for inputting this process. Therefore, it is suggested that the department exhibit unusual aggressiveness in seeking the input and involvement of primary care on committees and in hearings. In addition, it is suggested that the extensive experience with primary care information systems in other countries be obtained by testimony, conferences and other methods.
Thank you.
DR. STARFIELD: Thank you very much. We are really on schedule. That's great. We now open the discussion, starting first around the table. My colleagues, any questions?
DR. IEZZONI: Thank you. The testimony was excellent. You really informed us in the way that we had hoped.
I have a lot of questions, but let me just start with one that was striking to me. The series of speakers were on either end. Dr. Hanley, who described herself as a primary care pediatrician practicing in a rural environment, and Dr. Moon Mullins, who described himself as a family practitioner practicing in a rural environment now, transitioning several years ago into an academic medical center. But both of you self identified as country doctors, in a sense.
You said very different things. Dr. Hanley basically was very supportive of the CPT, and had virtually no negative things to say about coding systems, and didn't speak specifically to the primary care issue that Dr. Mullins did at the end. I'm just struck by that. How do we get the primary care input in a meaningful way that spans the spectrum?
Dr. Mullins, could you perhaps -- and Dr. Hanley, can you explain to me why each of you have such different perspectives? You can't obviously read each other's minds, but what is happening here?
DR. HANLEY: Thank you. Well, I actually practice pediatrics. I am in the office six, sometimes seven days a week. Nearly every patient I see, I code and submit -- not every patient gets submitted to an insurance company, of course, but most of them are submitted to third party payers.
CPT is the system we use. CPT is the system we have used for many years. It is a system that has evolved over a period of time, and our knowledge of the system has evolved over a period of time.
Debby, who is a real person, real person, Debby has acquired a considerable amount of experience and expertise with CPT, at not an inconsequential effort on her part. That is where I speak from. I use CPT really as a service to my patients, because we submit the forms to the third party payers for the patients.
DR. MULLINS: I also am a practicing physician, seeing patients over 40 years for a long time, in many settings, solo, partnership, group, university settings, until about two months ago.
I use ICD-9 and CPT, and we have no problem with it. But I think the difference in our focus is, I only hear you talking about the billing, and what I am looking for is clinical care.
DR. IEZZONI: Right, that's what I thought.
DR. MULLINS: And actually, I have been very interested in electronic medical records. I was not interested, as long as I thought the government wanted it for resource, or somebody in academia wanted it for research. But when I saw a system, an electronic medical record, that would help me at the point of care, help me at the point of care, would give me information, not what I put in, but what it gives me at the point of care, portable, wireless, then I got excited.
To do that, it has to be structured, and it has to be structured with granularity, such that it can help you with patient care. That is a clinical vocabulary. ICD-9 doesn't have -- I don't think anybody will differ with that, you can't populate an electronic medical record with ICD-9 or 10, as far as I know.
Now, I'm not an expert on any of these things, so don't ask me bad questions.
DR. IEZZONI: Can I just follow up really quickly and clarify, Dr. Mullins? For billing, and unfortunately we are going to limit ourselves to billing, are you satisfied with the CPT-4 and ICD-9CM for billing?
DR. MULLINS: There is a little puzzle that has some dots and it says, can you put these dots together with so many lines, and to do that, you have to go outside of the dots? That is the way I felt when I was trying to talk about billing, because I don't think you can talk about billing long term without it affecting what happens clinically.
Now, getting back to your question, we are happy at our place with ICD-9 and with CPT, or with either one of those. As you know, the hospital uses ICD and we use CPT.
But the other thing that is happening is that whereas we have a small office like your office, and somebody is coding in the office, they are now moving our coders way away. There is evidence to show that the closer to the point of service the coding is done, the more complete and the more accurate. So this idea of moving something way off and hiring somebody that is a non-clinician I think is a step backwards.
DR. STARFIELD: Can I just follow up on that with Dr. Hanley? Let's suppose that we over the years move away from fee for service billing. How would you regard CPT in that scenario, where you were using coding really only to support your clinical practice?
DR. HANLEY: I'm not sure I know how to answer that, except for the fact there are still resources that you can relate to procedures that are done, whether or not you are directly reimbursed with respect to that. You still have to monitor what is actually encountered by the patient. I see that as still being something very important to know as we look at all of the resources that are used in health care.
DR. STARFIELD: And you would make no changes in CPT. Can you respond to some of the things that some of the other speakers, several of them, raised? That is, the whole issue of the proprietary nature of some of the coding systems, CPT among them? Can you see a situation where maybe that might --
DR. HANLEY: I can't respond to that directly. I think Dr. Reggie Harris is going to speak tomorrow and can probably do a better job of that. I have heard the comment that it is -- because the CPT code is proprietary to the AMA, that it is expensive. But for my office, it isn't. We can buy the CPT book for $38. It is not a significant amount of money in the day to day running of a pediatric practice.
So I can say that, and then leave the rest of the question for Dr. Harris.
DR. STARFIELD: There may be some comments on that particular issue.
DR. COHN: Can I ask a follow-up question on that? I was actually somewhat struck, as was Barbara, by the variety of views on this issue of public domain versus copyright and all of that.
Now, as I analyzed the issue, and I stay away from public domain, only because I tend to think of public domain meaning anyone can change it, which I think Moon brought out as a vital issue, but the issue of free versus cost is probably what we are really talking about.
But as I look at the issue, if a chronology is to be free to the public, it means it either has to be produced by the government at taxpayer expense, either government or else somehow subsidized, versus being paid for by the constituency using it, in which case you can have private terminology developers. I'm just wondering, those who talked about this thing being in the public domain, what is your view of that? Is it your preference to see a terminology developed by the government for you? Is that the preferred status?r what would you like to see?
DR. GRIFFIN: Thank you. I'm not quite sure that anyone would like to see a terminology developed by government by anybody, but maybe they would.
Clearly, the issue in our view --
DR. STARFIELD: ICD.
DR. GRIFFIN: ICD, right. ICD is a problem sometimes. I was particularly concerned about procedure coding. The advantage of CPT as it exists is that it really involves a compilation of information put together by volunteers from a number of different specialties. In fact, all of the specialty societies currently represented in the House of Delegates of the AMA as well as a number of other organizations, as Dr. Hanley said, most of those groups have their own coding committees that may deal with particular issues that come up, either new technologies or revisions of current procedures that may be being examined.
You can look at those procedures and services in a number of lights. Is it really contemporary practice? Is it widespread? Is it effective? Is it accepted? Is it really something that not only do we want to have included in the nomenclature of a procedure that people can accurately then code for? By putting it in the book, are we going to encourage a procedure to be used that may in fact not be valid?
Now, the various specialty societies I guess look at those things different ways. But the real nice thing about CPT is, the editorial panel then tries to filter that information that is submitted to it from the various specialty societies and says, yes, specialty society A has this standard for submitting a code and specialty society B has this standard, but we have C standard, and that is the one we are sticking to, and that has been very helpful, even though from a specialty society standpoint it can be on an individual basis sometimes frustrating. It does allow some consistency and fairness to the process. So that's the good news about it.
The bad news in relation to your question is that because it is a copyrighted particular book, for example, if a specialty society wants to publish an educational piece and educate its members about how to effectively use those codes, the potential for the cost to do that can be large, simply because of the fact you've got to go through copyright processes of a book or of the content of a material that in fact was developed in great part by the specialty societies that oftentimes are then precluded from using it without paying a royalty fee.
That is an issue that you can address tomorrow with the AMA in more detail. Now, that's not to say the AMA has not been cooperative, and it is not to say the AMA has not been willing to work with specialty societies in educating members of the medical profession, because they clearly have. Many of the medical specialty societies, certainly the ACOG, has worked with the AMA, and sells AMA CPT-4 through our own educational catalog. Our members then of course are able to get that information that otherwise they would have to go directly to the AMA to get, and many of our members quite frankly aren't members of the AMA.
So to have those resources developed by the AMA, available for sale to non-AMA members, is very, very important.
As you know, though, and I think Dr. Mullins was right and Dr. Hanley was right, many of the speakers here were right, that is, that in order to effectively code, the people providing the service, the people at the point of service have to provide the information. Although there is a Debby in every office all the way across the country, Debbys aren't enough anymore, quite frankly, because the person who signs the bill isn't Debby. The person whose name goes on the bill is Dr. X. Dr. X gets in trouble if that code is incorrect. He either doesn't get paid correctly, or he gets in real trouble in some instances.
We as a specialty society feel we have an obligation to educate our members, not coding for fun and profit, because quite frankly there is some of that that is out there. We feel we have an obligation to educate our members for accurate coding to reflect really performed clinical services.
We do that in a number of ways. We put on courses and we write monographs and we put on a number of things. But some of those things have the potential -- they haven't happened yet -- have the potential to become prohibitive from a financial standpoint for us to do if we have to worry too much about copyrights and royalties, in a system that quite frankly we helped develop. That is the issue.
DR. STARFIELD: Dr. Vanchiere wanted to say something.
DR. VANCHIERE: In my testimony, I did state that the Academy endorsed that any new coding system developed be best served to be in the public domain. But also stating that that should be developed within the medical community.
If I had to say -- and I also stated that the Academy recommends that we continue with CPT and ICD. I think that is our position. We are working within those two classification systems. It has posed some problems as it relates to pediatrics, but we have worked through those, and are effectively working through other coding issues that specifically affect pediatrics.
I think the Academy would feel best served by not a fast pace into another coding system that is an unknown, that we do want to recommend that we stay with ICD and CPT. It does serve the needs of the medical community well as it is.
MR. BLAIR: I would find it helpful to have a little bit better understanding about the process of encoding. Maybe if each of you could briefly indicate whether you all have Debbys, and if you do, whether that goes directly into a computer billing system, whether there is an encoder that helps, or whether there is some other form of automation involved, could you just help me a little bit with that?
DR. HANLEY: We mark on a thing called a master bill that is on the patient's chart, we mark the level of service, and we mark the diagnosis on the chart. Those two things go to Debby, and Debby puts the code on it. Then that is submitted electronically to the third party payer.
MR. BLAIR: Debby looks it up in the manual and puts a code on it, and then somebody else puts it into a billing system?
DR. HANLEY: She does it.
MR. BLAIR: e does it.
DR. GRIFFIN: Many of the offices of ob-gyn are set up in two ways. Because it is a combined medical surgical specialty to a great extent, about 60 percent of the services performed by obstetrician-gynecologists are office based medical services, if you will, and the remainder in the hospital, where the obstetrical services are in-patient admissions for gyn disorders or surgical services, or whatever.
Superbills are used, or master bills, or whatever you want to call them, for most out-patient encounters. Those are written -- I'm sure you're familiar with them, so please forgive me for being redundant, but they are sheets of paper that have on them the particular number which corresponds to a CPT number, which may be an intermediate office visit, under the old parlance, and then alongside that a little shorthand abbreviation of what that service is.
Usually, the service that is performed is then checked in the box by the doctor in the office doing that, because that is appended to the patient's chart when the chart comes back to the room. That is checked, the patient then takes that back to the out clerk or whoever happens to be processing patients out. That either gets paid or gets checked off, and entered in the patient's bill and then it goes to the billing office, which may be separate for them, but the code has been entered. From there, it is simply a matter of entering that code, either electronically or sometimes by visual computer scanner, or then typing it into a typewriter, if it is a claim that be submitted, or in many instances, the top sheet of that so-called superbill can be submitted directly to a third party payer and there need be no other paperwork done.
Now, the breakdown comes down when there are two or three codes that are very close to one another and there is a lot of subjective interpretation from one physician to another, sometimes even within the same practice, about the level of service that may have been provided, or that kind of thing.
Now, that has been addressed. It has been addressed by the RUC, been addressed by CPT and by HCFA, to a great extent. The definitions of what those services comprise will, if not already, soon be published and be very, very clear as to what a physician is expected to have not only performed, but have documented in the chart in order to enter that specific code. So a lot of that will be relieved.
There still is a problem from the surgical standpoint. That is, when a patient is in the hospital, a patient has a surgical procedure, is admitted to the hospital, there is not a Debby in the hospital. Debby doesn't go to the hospital, often. So when you come back from the hospital having done a surgical procedure, you go to Debby and say, Debby, I just did a hysterectomy for Mrs. Jones. There are 17 codes or so that mean hysterectomy. So Debby either has to say, Doctor, be more specific, or wait for the operative report, which we hope by our educational activities, will be clear enough and have on there enough information that Debby can accurately code, or better, and what we prefer, is that the doctor will have written out the code and the procedure precisely as he or she wants it, and given that to Debby so that there is no misunderstanding when Debby enters that into the insurance file.
That is what we hope. That is what we want. Does it always happen? You bet you not. But we are working for that.
MR. OWENS: I think from the dental perspective, there are a couple of different things that might be helpful to interject at this point. One is relative to how dental codes are developed in general. They are developed by the clinicians, both from our specialty organizations within the profession, that actually come and work with our council subcommittees. They are proven and tried out there with clinicians and approved by organized dentistry at our House of Delegates.
So it is not -- when we say we have an ADA coding structure, it is in fact an ADA standard that is improved by the profession.
Relative to specifics and how they come about coding things in the dental practice itself, as a specific service is rendered, at that point in time there is a specific procedure code that is identified with that particular service and coded on the document at that time. Although we don't have the cloudy issue of diagnosis and multiple diagnosis modifiers at this time, from dentistry's perspective, we have a much more straightforward and easy way to do that.
So we get a little bit concerned when you say proprietary standard, if you will, because when you talk about the ADA procedural codes in the current dental terminology, you are in fact talking about the industry standard for dentistry.
MR. BLAIR: Bob, the thrust of my question was a little bit more towards the process. To what degree is the coding process essentially a manual one, and to what extent may it be automated with, in your case, dental practices?
MR. OWENS: At this point in time, there is only computers in about 75 percent of dental offices, so there is a lot of manual processing that does occur. In fact, I believe the latest statistics is somewhere around five to ten percent that actually are sending electronic claims at this point. We are working to actively increase that, obviously, but at this point in time it is a more or less paper environment. But even within that paper environment, the coding is again working.
Obviously, the responsibility of the clinician, but the clinician is well trained and well provided with the tools they need to effectively manage the coding system, again because we have a more active participation process from the clinicians in the up-front development.
MS. PACE: In home care, I just want to make a comment that the coding is relatively simple in terms of getting paid. For example, for Medicare services, it is mostly just related to those revenue codes. There is one for each discipline and one for medical supplies, so that is pretty straightforward. In terms of the ICD-9 coding, although that is a required item for submitting the bill, its payment is not linked to that ICD-9 code, so there is great variability in terms of home care agencies' expertise in coding.
So I think that there is a lot of wide range in terms of how good home care agencies are in coding. It varies in terms of -- it may be the clinicians themselves doing the coding versus someone in a medical records department. Again, that depends on the size of the agency and the type of personnel that they have.
DR. HOSKINS: For nurses, it varies depending upon the practice setting they are in. If they are in a practice setting where physicians are going to use whatever coding system is there, I think largely those are still paper systems rather than electronic transmission of data.
Some nurses may be in an independent practice. In that case, it is very much likely to be by hand, whatever they submit. The codes generally do not take care of everything that nurses do, but they fit within the ICD, within the CPT and so on, struggling a little bit to get it done. You will hear tomorrow from other developers that do have coding systems that we believe that if they were incorporated, they would incorporate the things that we do.
DR. IEZZONI: I have two questions for Dr. Vanchiere. In your testimony, you stated that current code sets do not adequately allow you to risk adjust around certain clinical dimensions that are of concern to you. Could you as my first question just expand upon what you meant by that?
DR. VANCHIERE: I don't have a good answer for you. That was part of a committee report, and I wish I could answer it more specifically for you.
We have had some difficulty at times with CPT coding, that we felt was different for different age groups. Maybe that is what it was referred to, as far as the risk adjustment related to age specificity on some codes.
DR. IEZZONI: My second question was, I thought I heard you say that you preferred DSM-4 for mental health diagnoses to ICD-9CM. Was I correct in that?
DR. VANCHIERE: The information there was that the Academy has just been a part of developing for child and adolescent coding that does have more detailed information as to arriving at a diagnosis code, the criteria for establishing. But it does not always cross link with ICD.
DR. IEZZONI: Do any of the other panelists have experience with thinking about mental health diagnoses, whether you feel more comfortable with one system versus another?
DR. HOSKINS: Our nurse psychotherapists prefer the DSM codes. In support of what Dr. Vanchiere said, some of our pediatric nurse practitioners say that the specificity, the detail offered by DSM fits more of the problems they have with children.
DR. GRIFFIN: ACOG was also involved in DSM for PC development. Although ob-gyn is not what you would call in the past a mainstay of mental health professionals, on the other hand our patients suffer a number of psychiatric disorders and are missed, quite frankly, by the majority of primary care providers with whom they come into contact. The development of a manual for primary care providers that would better help to educate and pinpoint signs and symptoms of mental disorders and refer those people appropriately was seen as a key part of what we felt our health responsibilities were.
So we too would support DSM-4 and the DSM-4 PC particularly for us is a very, very helpful tool.
DR. STARFIELD: Anybody else want to comment on that particular point? If not, I want to go back to Dr. Iezzoni's first point about case mix adjustment.
One of the problems, if you are using ICD in a diagnosis, a case mix adjustment, the problem with ICD is the absence of the ability to code certainty of diagnosis, a rule-out diagnosis the same as a regular diagnosis. One of you mentioned the problem with the rules for coding certainty of diagnosis. I think it was you, Dr. Hanley. Do you want to expand on that, or does anybody else?
DR. HANLEY: Just in using the modifiers to the level of certainty, that was just that issue. I can't expand anymore on the
DR. GRIFFIN: It is a problem, because many times as you have said, you will end up coding a diagnosis as if it were a certainty, when in fact all of us know patients often leave episodes of care without a certain diagnosis. Thank God, most of them get well on their own, whether we diagnosed them or not. That is a fact, as we all know.
So sometimes we don't have a diagnosis. To not have the ability to put down rule-outs or whatever is somewhat of a problem.
DR. VANCHIERE: There are some codes in ICD that are too broad of a catchall. Like, I was reviewing the ICD-10, to document -- a patient had had myringotomy tubes as part of his ongoing past history. It was categorized with having had a cochlear implant. I said, these are two extremes of the same thing. You need separate code numbers for these.
So there are many episodes within the coding where it could be improved to help that your data reflects something that is related to risk.
DR. MULLINS: Particularly in primary care, frequently if we're honest we're not able to make a traditional diagnosis. We frequently sign out that visit in the term of symptoms, findings or problems.
Several years ago, I heard that a study had been done in Switzerland in primary care, in which they only could make a traditional diagnosis ten percent of the time. But I never could get a copy of that. It was just relayed to me.
So I think that is what we are saying with ICD. I haven't heard anything today that I wouldn't agree with, if we continue with paper records, manual coding and electronic transmission. But if we are going to have electronic medical records, then we are going to have to have some different kinds of things. ICD-9 I don't think could populate my clinical records.
The other thing I had written down here after this in support of ICPC is the whole idea about episodes of care. That didn't make any sense to me at first, and then people said, if you see somebody with appendicitis and they come to your office and then you see them in the emergency room, and then they are sent to a hospital, and then you take the stitches out and so forth, for those of you that are doing national statistics, that is six appendectomy visits. That is going to be coded as six cases or six visits of appendicitis. That is only one episode of appendicitis.
Many people say that ICD doesn't allow for the capturing of episodes. But the other thing that we --
DR. STARFIELD: Is that true or not?
DR. MULLINS: Pardon?
DR. STARFIELD: Is that true? They say that it doesn't allow for the coding of episodes?
DR. MULLINS: ICD?
DR. STARFIELD: Yes.
DR. MULLINS: No.
DR. STARFIELD: You can code episodes.
DR. MULLINS: You can?
DR. STARFIELD: I'm asking you.
DR. MULLINS: You cannot with ICD. You can with ICPC. But there is one thing we are not saying here, too. That is, it is not in any of your questions, the structure and function of an electronic medical record is going to determine a lot of these things. You can take the best coding system in the world, but if you don't have a system that operates it -- so one of the things that many people I associate with point out is that the episode of care is also a function of the medical record as well as the coding system.
DR. STARFIELD: I want to still come back to this issue of coding to the highest level of certainty. What does that do in your own practice with that instruction? What does it do?
DR. VANCHIERE: The diagnosis claim.
DR. STARFIELD: Yes, the diagnosis claim.
DR. HANLEY: One of the things that it does is, it takes out time and sometimes in takes time to make a diagnosis. It may be that you see a patient with a certain set of symptoms, but as you follow them over time, it becomes clearer what exactly their problem is. To have to code it to certainty is something you can't always do when it takes time to make the diagnosis.
That is a lot of what medicine is, I think. There are some things that are very obvious. An ear infection is an ear infection. But other things are not quite so obvious.
DR. VANCHIERE: One of the problems that it poses is, when someone out there starts looking at level of procedure coding and trying to match that with diagnosis coding. So if I see this child who has a swollen ankle today, and I have gone through a level four documentation and CPT coding of that, but I can't diagnose him as having juvenile rheumatoid arthritis or a bone tumor, something significant, and all I can code is complaint referable to ankle or swollen ankle, when that matches up it is not a good match-up in somebody's computer.
So down the road, someone is going to hold us accountable on the procedure side for what we are saying on the diagnosis side, and we need to do it better.
I do what is called cognitive coding, because if I have this child who has a streptococcal pharyngitis, and I just code that out as a level four, someone reviewing my chart might say that is a little high. But they don't know that this child a year ago had glomarial nephritis. They don't know that this same child has a parent who is incapable of receiving proper instructions from me in carrying forth -- so if I code out the codes for family situations, and past history codes, then I think it substantiates that procedural code. So that is what we need to work towards.
MS. GREENBERG: Just on that briefly, anyone can respond to this, but Dr. Hanley, do you find that what you know, and everyone agrees is often the case, is really a constellation of symptoms or number of symptoms, rather than a definitive diagnosis. If you code those symptoms rather than a definitive diagnosis it is not accepted as a valid bill.
The actual coding guidelines in the out-patient do not encourage you to code beyond your own level of certainty. Your own level of certainty is limited to the symptoms as opposed to the diagnosis. That is what the coding guidelines would tell you to do, but I know we often hear that it gets kicked back if that is what you record. So that is what I was wondering.
DR. HANLEY: Sometimes it does, and that is just the way the game is played. But I think what we want to focus on here is what is best for patients. I just keep coming back to that. That is what I do all day long.
I think as Dr. Griffin said, sometimes a diagnosis is applied to a patient that in your true belief and in your true heart you know doesn't really belong there, but it is just what you have to do.
DR. VANCHIERE: I think that to be able to code for suspected diagnoses is important, more important in some situations than others. If I am evaluating a child that has profuse vomiting, has had a five-pound weight loss, I'm wondering if this child is diabetic. So I am going to go through a few more things to determine that. The coding doesn't allow for that unless I use a generic D code for suspected condition.
Now, the door has been opened for us to do better in ICD when they gave us diagnostic codes that let us code specifically for suspected newborn sepsis, or a suspected newborn neurologic problem. So the door has been opened, and I think we can work with ICD and maybe add some codes. Hopefully we can. It makes sense to do it.
DR. STARFIELD: Can you envisage a qualifier instead of adding all those codes? Is that something that might be --
DR. VANCHIERE: Say again?
DR. STARFIELD: Using a qualifier for all diagnoses instead of adding every diagnosis.
DR. VANCHIERE: If we can -- well, the qualifier for a suspected catchall is there, but sometimes it is not accepted by third party payers.
DR. MULLINS: I was just going to say, someone down here mentioned, people don't want you to code symptoms and diagnosis. They don't accept that. I believe HCFA now is not accepting the diagnosis that we are coding. Then in our place, we are hiring all kinds of people to come in and review the clinical records to see that the criteria are there to meet the HCFA requirements.
So again, I'm speaking to have both a good coding and classification system, as well as a vocabulary. The need is already here now. HCFA has already established the need to come in and see what is recorded in your chart. If you had it electronic they said, we've got to go back and do it medically.
The other thing I wonder, I'm not sure any of us are disagreeing, on the things you were talking about over longitudinal care, if that is described as an episode and you have the different levels that are coded somehow, and then in the end there is only one episode of one thing that you saw again in primary care, that you see it early and it develops, and finally you know what it is.
So I think if we could record by episodes, if we could record with a vocabulary and then had a coding system, it probably would answer all of our questions.
DR. STARFIELD: How would you actually do that, short of computerized record? If you had a computerized record, of course, it is done automatically.
DR. MULLINS: How could you do that?
DR. STARFIELD: yes.
DR. MULLINS: I'm not sure you really could, short of a computerized record, because the clinical vocabulary would be so extensive, the way that you do it -- you had asked earlier, how do you do it; you do it from a picking list or a menu now. It is invisible to the coder. You just select whatever the finding is, and it is automatically coded. But short of an electronic medical record, I don't think that is possible.
DR. IEZZONI: Ms. Pace, I wanted to ask you a little bit more about home care. I was struck by what both Dr. Starfield talked about at the beginning, that we want to be able to track patients across different levels in different settings of care. It is obvious that in the setting of care that you represent diagnostic codes are not something that every home health agency is familiar with. It sounds from your description that the coding of diagnoses is somewhat sporadic, or the quality of it is highly variable.
In your testimony, you also talked however about the fact that HCFA has just posted a notice about the OASIS system. OASIS is obviously not a code set, but it is a list of clinical questions. You didn't say how you felt about OASIS, though, and I would be very interested in to hear, because I am hearing a lot of opinion about that right now on the street.
MS. PACE: I think basically, the industry is supportive of OASIS. It has been -- HCFA has had many years of research into it, and many home care industries have participated in the research and demonstration projects. I know the National Association for Home Care is basically supportive of that movement, in terms of having actual specified data items that are very specific in terms of patient characteristics.
The whole purpose behind OASIS is to get to quality indicator data, outcomes and certainly the industry is behind that. That data set is also being used as the core in HCFA's new study to look at case mix classification for a future prospective payment system. So from that standpoint, HCFA is looking at it as serving multiple purposes, both the quality side and as I said, the core for the case mix for prospective pay.
So in general, I think the industry is behind it. I think with anything of that nature, that kind of standardization is going to mean that every single Medicare certified home health agency is going to go through change. So you can imagine that there will be a certain amount of concern about just the volume of the change.
But I think in general, the industry supports the concept of having a standardized data set on patient characteristics, using them for quality and using them as the basis to develop a classification system.
We know in home care that it is not medical diagnosis alone that drives the amount of services that a patient needs. So that is not really even a potential at this point, in terms of having an adequate characterization of the patient in terms of some of the things we need to know.
DR. STARFIELD: Are we clear on this point that was raised? I just want to ask Mr. Owens a question. Every dentist that I have ever talked to says that the mouth signals the health of the person, of it is an ophthalmologist, it is the eye.
If you had your druthers, would you still want a separate system with industry, or would you want a system that is unified with the medical system, but that incorporates dentistry?
MR. OWENS: We would have to probably do a little more research before we answer that definitively. But I think just to maybe give you a feel for how things are evolving within the dental community and the reasoning behind some of the answers to our questions.
Given the administrative simplification or K2, it is our view that we need to get this thing implemented and get on with moving forward with administrative simplification. In order to do that, the most effective way at this point in time is to use our existing coding system, CDT, use ICD of some sort and CPT to do our oral surgery. That addresses the immediate need.
We have already had initiatives for quite some time to develop the health record. As a matter of fact, there are parts within the ADA that are working with ANC in ANC standards developing to develop the oral health record. A concept model has been developed for over a year. The logical model will be completed hopefully sometime later this fall.
Within that model, that record model, there will be need to have additional types of coding systems. To that end, the ADA has in fact participated with SNOMED in working up diagnostic codes that are appropriate for dentistry, keeping in mind that the clinician is needed to drive that. So we are serving as the catalyst or the coordinating entity for organized dentistry with all our professional groups to insure that when the diagnosing system is developed within SNOMED, that dental component will truly be reflective of dentistry.
So the answer -- over time, will you have one system like a SNOMED or something else that could be for all of health care, I really can't answer that at this point. But if you ask about how the specific components of even an overall vocabulary should be built, that should be tied to the specific segment responsible for coordinating their industry.
So no matter what coding scheme will ever be used, it is the ADA's opinion that they should be responsible both as a standards developer in the true sense of the word, both from organized dentistry and in ANSE standards developer and having that responsibility also on behalf of dentistry. As our consulting role to K2, we take that very seriously in making sure that any standard that we will put forward has the proper due diligence to serve the industry.
DR. VANCHIERE: I would like to come back to the question that I did not answer for you earlier. Having read back over the testimony, it mentioned to facilitate risk adjustment. What that was entered for is, as I am now entering a Medicaid managed care contract, and if I were to go back and look at my experience over the past X number of years of providing service to 5,000 Medicaid patients, I probably have a higher than usual case cause.
The coding that we have done in CPT does not account for those high risk children, cystic fibrosis insurers that we have taken care of with high costs. So the coding system needs to have some way of adjusting for risk.
DR. IEZZONI: You could always code the CF on ICD-9 CM.
DR. VANCHIERE: But then you would have to go back and track every one of those separate diagnoses, separate coding system, to bring it together. But it would be important. If there were a procedure coding that allowed you to code for certain conditions that was reflected in the diagnosis. There are a few of those within CPT.
DR. IEZZONI: I know. I have been a little frustrated using CPT codes that have a couple of diagnoses embedded with them, such as a procedure for either hemorrhage or infection. I haven't had a chance yet to go back and see how often there is an associated hemorrhage code or infection code coded in ICD-9CM. Do any of you know about how ell that is done for those CPT codes that do have some diagnostic information embedded within them, wither the clinicians really do also code the ICD-9CM diagnostic code to differentiate? Thank you.
DR. FRAWLEY: I have two questions for the panel. The first, it seems the majority of you in the short term, given the guidelines for the time frames in HIPAA would support ICD-9CM and CPT. I was pleased to hear that there was consensus from most of you on that.
In the long term, we heard that depending on the field testing of ICD-9CM that many of you would support the eventual migration for a diagnostic coding. Mr. Owens from the ADA talked about long-term continuing with ICD-10CM and ICD-10PCS and CPT and CDT.
I was wondering if we could get some remarks from the panel on moving towards a single procedure classification system.
DR. GRIFFIN: Excuse me, a single procedure or a single diagnostic --
DR. FRAWLEY: No, a single procedure. Many of you said you would eventually support the move towards ICD-10CM for diagnosis, depending on the field testing, in your testimony. I'm just curious on the procedure side. Short term, you all said ICD-9CM and CPT, long term, we have heard some differences he
DR. GRIFFIN: I think clearly, the college, recognizing the problems with the current coding systems, believe that currently, CPT and in at least the intermediate term CPT continues to be the best method available for procedural coding for services.
DR. STARFIELD: Does that mean across settings? Across different kind of settings?
DR. GRIFFIN: Yes, correct.
DR. STARFIELD: All settings?
DR. GRIFFIN: Yes, ma'am.
DR. DETMER: We have heard from obviously nursing and some hospice area it certainly doesn't cover --
DR. GRIFFIN: I think it is a valid criticism, but I think it has been recognized by the CPT editorial panel. I think Dr. Hanley mentioned earlier, and it is appropriate, that the panel is being expanded for non-physician professionals or other providers and to look at other sites of services to try to make sure that those deficiencies which currently exist are correct in the future editions.
So I recognize that there are some deficiencies, but none that in my view can't be easier fixed with the current system than to start de novo.
DR. HOSKINS: I think nursing, as I have stated, is on record as supporting a single procedure classification. We haven't said which one, but we would prefer a single one. We do have our own set of developers who have come up with an intervention classification, not comprehensive. We would hope that their interventions might be added to what exists, but we still prefer one.
MS. PACE: We wouldn't have any problem with the idea of having one procedural coding system, as long as it adequately incorporated what we do in the home care setting. I know right now, for example, CHAMPAS requires home health agencies to use CPT clinicals, and it is very difficult for home health agencies to use those codes for the services they are providing.
MR. OWENS: For dentistry, we in fact do have one procedure code system, that is CDT. HCFX is there and used from HCFA's standpoint. Again evolving to one, that would be a question for HCFA. Can we not evolve to the CDT code and work out any particular issues relative to why you insist on maintaining a separate coding system.
As far as one with medicine and pharmacy and all of that, again, we have a very strong feeling that those expertise lie in a specific segment of the industries, and that we do have a lot of experience in developing standards, both from an ADA perspective and from an ANSE perspective. The users, our practitioners, tell us that they want control over their industry, and they are best suited to be able to set those codes.
So from that standpoint, organized dentistry would look to maintain control of setting their own code values.
DR. HANLEY: I made a comment earlier in support of using a single procedural coding as a way to track the resources that are used. But it helps me in another way. As a practicing pediatrician, I'm the happiest in a room with the door closed, with me and a patient and a parent. But the minute I step out of that room and into the hall, I am subjected to severe pressures to comply with third party payers, government regulations, UR and QA and cost containment.
Really, the common element to help me deal with that is CPT, which I think is one of the most compelling reasons to have a single system.
DR. FRAWLEY: That brings me to my next question. You have all commented on the difficulty that you have had in the fact that many payers do not recognize official coding guidelines that you have been forced to manipulate diagnoses and procedures for reimbursement, and certainly the concerns with fraud and abuse.
I just thought, if you had any thoughts at all in terms of, should there be a central authority, and should it be in the public sector or a collaboration of public- private, in terms of trying to develop official coding guidelines enforcement. I'm just wondering if any of you had any thoughts on a central authority and if there was a role for that process.
DR. GRIFFIN: Clearly, I must admit that the process by which codes go through multiple levels of review for CPT, that is, the terminology that is used, the appropriateness of including a code in the book, and the interpretation that the panel understood the code meant when it was put in the book, can only be appropriately enforced by the same people that understood what the intent of that code was when it went in.
So it seems to me that the appropriate level of that authority should be in my view the editorial panel. Clearly, the mechanism by which that has oversight is behind me, but it makes little sense to have an editorial panel review all the information, debate it extensively, decide to include a particular service in the book, and then to have another third party payer or any other group then come in and interpret, what did they really mean.
I think that happens over and over and over again.
It doesn't matter what other secondary organization would be doing it; it would be subject to some error.
MR. OWENS: I think from the ADA's perspective, we have a similar -- we don't call it an editorial panel, but we actually have codes and vocabularies under our council in dental practice that monitors codes and continues to enhance our coding systems, and be able to move those forward.
All processes can be improved, but I think the best mechanism at this point is to let them be self policing, if you will. We have a tremendous effort to follow a due process that insures a quality coding system is maintained. Again, that continues to evolve as the needs change.
So I think rather to enforce yet another entity on top of that, it would do nothing more than just add additional administrative burden. If you look at the overall ANSE processes and standards processes within an organization such as the ADA, that should be sufficient to insure ongoing quality.
MS. PACE: I would just comment that administrative simplification does not happen for the provider if every payer is allowed to require different elements or have different definitions, or say you can use this and can't use that. Providers are working with multiple payers and the burden is put totally on the providers when that is allowed.
DR. IEZZONI: Can I just follow up on where Kathleen was going? Dr. Griffin spoke to having experience with the 3M procedure classification system that HCFA is now developing. Have any of the rest of the panel had a chance to look at that and see how it works for you? No? Dentistry hasn't? Okay, thank you.
DR. MOR: To follow up that last point that Karen made about the administrative simplification, in light of the different insurers having different desired interpretations of what a particular activity is, forget code for a moment. There is really an activity that goes on.
So really, the desire would be to have an activity stand by itself, be clinically meaningful, and then to have insurers come up with different pricing structures for that activity. Then you can negotiate pricing structures in the activity. A coding system does that better than natural language.
DR. GRIFFIN: I think it does. The coding system is really a shorthand. For example, you can say I made an incision in the abdomen and went through the fascia and tied off bleeders and did all these various parts of a particular operation, or you can say, I did a whatever, five-digit code that represents all the integral parts of that procedure, and then people really know what you're talking about.
The problem comes in, well, does that also include this, or does it also include this, or does it also include that. That has caused a number of specialty societies to have to produce at not insignificant expense and work documents that lay out what in fact is included and not included in this code that was submitted to so-and-so. That really is counterproductive.
Quite frankly, whether it will even be useful or not, nobody knows, because the insurers still have the ability to say, well, that's fine, that's your opinion. It really is a problem.
I think you could write it all out, but once again, if you're trying to simplify the exchange of information and have it meaningful, you've got to have some shorthand. That is really what the coding systems are all about, it seems to me.
DR. VANCHIERE: You mentioned having insurance companies with different reimbursement plans.
DR. MOR: Yes, it's just a different pricing table, that's all.
DR. VANCHIERE: Based on what?
DR. MOR: Activity.
DR. VANCHIERE: What is in place now gives to the insurance company a procedural code that has been giving a certain relative value of physician work and cost. So by all insurance companies using the same relative values, I think there would be less chaos than there would be if you just give them a description of a service.
DR. MULLINS: I was just going to say, the way I look at that we can't get a computer to use natural language. So we do put it in in natural language, but with something that the computer can use. That is the way I see that.
I'm like you; I'm not thinking about coding, I'm thinking, how can we take a term or a finding or something and put it in the computer so it will use it. Your question was, why isn't natural language better. I agree, it is better if we don't have to go and by hand do all that.
The second thing is, I may say nephrosis and you may say dropsy, and somebody else may say Bright's disease. If we had some kind of classification or something mapped to each other, we would all be saying the same thing. I think that is the reason coding is helpful, I think.
DR. MOR: Yes, but that basically assumes that your coders in your offices are better than the master computer that can put all that stuff together in tables, and that is highly debatable.
DR. MULLINS: I am speaking from the coding being done at the point of care. That is what I'm saying. So it is not the coders in the office.
MS. COLTIN: I heard from a couple of the speakers an acknowledgement that these data are used to support multiple purposes and not just fee for service billing, purposes such as development of different payment mechanisms for case payment and so forth, but also utilization review, quality assurance and so forth.
While I have heard a lot of support for CPT, one of the areas where CPT has been problematic when being used to support these other purposes has to do with the global codes that are used in surgery and in obstetrics in particular. The fact that there don't seem to be uniform rules for how to clinical the service dates that are associated with those codes.
What we see when we look at global codes for prenatal care that include antepartum delivery and postpartum, the service dates tend to link with the dates we get on the hospital bill for the admission and discharge. Therefore, one cannot determine when prenatal care was initiated, and you can't determine whether a postpartum visit actually ever even happened.
I would like to hear a little bit of the perspective -- and this is true for surgery, too, in terms of looking at some of the pre-op and post-op. To use this information to look at quality indicators and outcomes and so forth, what do you see as the issues? I realize it is a tradeoff in terms of coding every individual event, but what are your opinion about this?
DR. GRIFFIN: Let me go back a minute. The whole idea of global bundling for obstetrical services is different, quite honestly, than the reason for global bundling for gyn or other surgical services.
The obstetrician-gynecologist and family physicians providing obstetrics I think provide a global bundling, if you will, for obstetrical services, primarily because we believed in a total package of obstetrical services and the importance of prenatal care. The knowledge that, if you billed separately individually for each episode of a prenatal visit, you might discourage those least able to come, and therefore most important that they did, to come in for those scheduled prenatal visits for the preventive services that are so important.
As a result, the college, ACOG, has generally always supported the idea of a global package of obstetrical services which comprise all of those services that we know to be important or believe to be important for good outcomes in perinatal services.
Now, how do you best collect that information? I'm not sure, and I accept that it has been a problem. It is a problem for us, too. It is difficult for us to know when in fat a patient entered prenatal care. It is difficult for us to know exactly why she didn't, if she didn't come in early. It is difficult for us to know exactly how many visits during which that patient may have been seen and for what reasons. It is difficult for us to know if that patient came in for a routine prenatal visit or in fact really came in for another issue and was simply coded for a routine parental visit. It is difficult for us to know if she did come in for a postpartum exam or multiple postpartum exams, in many instances during that global period of time.
There clearly are ways that that could be addressed. But once again, that gets back to the issue of a uniform coding and billing policy that would be accepted by and enforced through all payers. Therefore, you could start to collect that data more appropriately. Clearly, it would be helpful to identify each episode.
We get back to episodes of care. That truly is an episode of care. Prenatal care, delivery and postpartum care is an episode of care. You say that is one pregnancy, that is one delivery. It clearly is, but it may be multiple visits, and you either code them separately in an unbundled fashion, at which point it is really difficult to figure out how many pregnancies you really had, or if it is bundled together, and then you lose the information that may be appropriate to figure out if the services delivered during that episode were adequate and appropriate.
So I recognize the deficiency. The only solution in my view is a uniform policy that would mandate the recording of that information in a global package.
The same is true for surgical services. To a great extent, we don't know how many visits or the appropriateness of those visits, either before and after in a global package. I recognize it is a problem.
DR. STARFIELD: We'll have one more question, and then we have a break. Bob Gellman gets to ask it.
DR. GELLMAN: Well, I'm not sure I have a question. It may be as much of a comment as anything. I want to turn back to the copyright versus domain issue that was discussed.
My comment really is that in some of the testimony, I think people were too quick to reach a conclusion on copyright or public domain. I think this is an issue that calls out for unbundling. That has to be broken down, and you may favor one or the other because of funding or updating or authentication. I'm not sure that under either regime, be it copyrighted or public domain or some intermediary situation that you cannot duplicate the -- you cannot find a way to address the concerns otherwise. It may be that the ultimate conclusion isn't very important, but the elements that go into this are something that need to be addressed with more precision.
DR. STARFIELD: I think that is something that we need to follow up at a later point, perhaps, rather than now.
DR. DETMER: This is a comment as well, but just on behalf of the full committee. I think this has been terrific, and I think your written materials will be very useful to the committee members that weren't here to benefit today. So just for myself at least, I think from the point of view of the full committee, I think this has been very useful.
DR. STARFIELD: Thank you, Don, you did it for me. Thank you all.
(Brief recess.)
DR. STARFIELD: This is a continuation of our panel on perspectives on medical and clinical coding and classification issues. Is there anyone who has joined us new except for the panel? We'll introduce the panel in a minute. Anybody else that has joined us that is new? Yes, please introduce yourself.
DR. MORROW: Doug Morrow, representing the American Optometric Association.
DR. STARFIELD: You missed my comment about the eye. I'll have to tell it to you later.
This is the panel on health care facilities. Could you introduce yourselves, please, then we'll go in order.
MS. LEON-CHISEN: Nelly Leon-Chisen, American Hospital Association.
MS. STEWART: Margaret Stewart, American Health Information Management Association.
MS. WING: Beth Wing, National Association of Children's Hospitals and Related Institutions.
DR. MARDER: I don't know if you want to take me with this panel. I'm a special data user.
DR. STARFIELD: No, no, introduce yourself now.
DR. MARDER: Bill Marder. I'm the vice president and general manager of the MEDSTAT group.
DR. STARFIELD: Thanks very much. Now Nelly Leon-Chisen.
MS. LEON-CHISEN: On behalf of our 5,000 member hospitals, health systems, networks and other providers of care, I would like to thank Dr. Starfield and the committee for the opportunity to provide comments. My name is Nelly Leon-Chisen. I am the director of the central office on ICD-9CM at the American Hospital Association, AHA.
The central office serves as the United States clearinghouse for issues related to the use of ICD-9CM. We are also the publishers of Coding Clinic for ICD-9CM, the official publication for coding guidelines and advice as designated by the cooperating parties, the American Hospital Association, the Health Care Financing Administration or HCFA, the National Center for Health Statistics, NCHS, and the American Health Information Management Association, AHIMA.
The need for standardization of clinical codes is extremely important to our members, particularly as they relate to transactions identified in the Health Insurance Portability and Accountability Act, HIPAA of 1996. Increasingly, diverse reporting requirements for clinical codes as imposed by payers and others have burdened our members. Wide variations in coding have not only increased our members' costs, but what is more important, they have compromised the primary purpose of clinical codes to provide an accurate record of what happened to the patient.
When providers are forced to code events differently for different payers, it becomes confusing and ultimately compromises the integrity of medical data. For this reason, AHA is very supportive of the committee's efforts.
The specifics of each question are found in the written statement. I would like now to give you a summary of what we would like to see take place.
The importance of accurately reporting clinical codes cannot be over stressed. Our members rely on clinical codes because they are key data components used for benchmarking, quality assessment, research, public health reporting and strategic planning.
To achieve these objectives requires accurate and comparable data. These goals were always important to providers, but are even more critical now, as many individual providers are moving to integrate services across settings. Standardizing the language to be used to communicate will greatly enhance their ability to delivery high quality, well coordinated care. We cannot achieve this unless we have uniformity and standardization of clinical data reporting requirements.
By building on the strengths of the existing infrastructure, namely, the ICD-9CM and the CPT-4, and by making some simple changes in the use of these coding classifications, our nation can make substantial progress toward the administrative simplification goals outlined in HIPAA.
Many questions outlined talk about the use and development of ICD-9CM and CPT-4 and how it has affected users. Other questions asked whether the process of code development belongs in the public or private domain. Each has advantages and disadvantages. However, for our members, the important issue is whether the code development process incorporates the following principles: sensitivity to all user needs, recognizing that organized delivery systems shift from in-patient to ambulatory procedures and from physician offices to ambulatory treatment facilities, structured so that code usage can be easily accessible and widely distributed, maintained through a clearly defined structure in order to achieve a high level of coding integrity.
Complementing these principles is a well-defined maintenance and implementation process. Input into the process should be broad-based, and changes to the coding system should take into consideration the needs of all users. Overall, the process should also be predictable and take into account the capabilities of the users to adapt to coding changes when they occur.
Providers should be able to count on routinely scheduled meetings to review coding changes and a certain date for when approved coding changes take effect.
In the transition to national standards, it is imperative that the following requirements be adopted. For the near term, establish a requirement that all payers and providers use and accept nationally established guidelines for coding classification usage, including V codes and E codes, whether they are in an in-patient or ambulatory setting. We recommend retaining ICD-9CM for in-patient transactions in a combination of ICD-9CM diagnosis and CPT-4 procedural code for ambulatory transactions, and also eliminating the use of all local HCPX level three codes.
For the long term, if a single classification system is developed, it must not be implemented without a complete cost benefit analysis, and should be implemented in a predictable time frame for all payers and providers. The process for maintaining the system should be broad based, well defined, and address the needs of all users. Any new system must undergo a thorough testing for compatibility to the DRG system, ASC rates and other ambulatory payment systems like APGs or simple fee schedules to prove their value and importance. The new system must also demonstrate its ability to provide an accurate record of the patient's condition and what happened to the patient before and after the rendering of care.
The AHA and its members appreciate the opportunity to provide comment. I will be happy to answer any questions you may have.
DR. STARFIELD: Thank you very much. Margaret Stewart.
MS. STEWART: Good afternoon, Dr. Starfield and members of the subcommittee. My name is Margaret Morgan Stewart, and I am president-elect of the American Health Information Management Association, AHIMA.
On behalf of AHIMA's 37,000 members, thank you for the opportunity to address the subcommittee on issues regarding clinical coding and classification. AHIMA is very much aware of the importance of the task which the National Committee on Vital and Health Statistics has been charged with under the administrative simplification subtitle of the Health Insurance Portability and Accountability Act of 1996. Our complete written statement answering the questions posed in your letter of invitation has been submitted for the record. My brief statement is intended to present the critical issues of concern to AHIMA and its members.
Currently, ICD-9CM and CPT are the most common coding systems used for administrative transactions. One criticism of ICD-9CM is that it does not provide sufficient clinical detail to describe the severity or complexity of diagnoses or procedures, nor does it provide sufficient codes for health care encounters for reasons other than treatment of disease, such as preventive medicine.
Another problem is that there is no modifier or descriptor at the present time to indicate that a condition is suspected and not confirmed at the time of coding. Since suspected conditions are coded as though they are present for in-patient encounters, the incidence of many conditions is overstated in statistical data.
One problem with CPT is that many of the codes include a variety of procedural combinations, making it difficult to know what was done. It is also difficult for facilities to use CPT codes because some of the code descriptions and coding rules are more applicable to physician services. Both ICD-9CM and CPT are updated annually. Their differing structures and purposes prevent development of a useful cross walk between systems. Differing coding rules between the two systems further compound the lack of data comparability. It is not possible to compare procedures reported by hospitals with those reported by physicians, because the procedures are coded in two different systems. It is also not possible to conduct longitudinal studies across care sites when multiple systems are used.
In order to compare complete data on a given procedure, it is often necessary to code the procedure using both systems, resulting in additional costs and administrative burden. Health care data is inaccurate and unreliable today because of the way coding systems are being used, not because of the coding systems themselves. The current way in which both ICD-9CM and CPT are being used for reimbursement and clinical data analysis is having a profoundly adverse impact on the accuracy of coded data. Clinical data are being compromised, and excessive cost is incurred due to lack of recognition for official coding guidelines and rules. Multiple payment policies among various payers are dictating code assignments with little regard to consistency or national coding rules and guidelines.
ICD-9CM becomes effective October 1 of each year, and CPT codes become effective January 1. However, many payers implement new codes arbitrarily. Some Medicaid prospective payment systems are utilizing code sets from several years ago. This requires providers to map current codes back to codes which have since been revised or deleted.
Many payers have arbitrarily chosen not to accept certain codes such as V codes, which describe factors influencing health status in contact with health services. Some payers will not accept certain valid digits with codes such as .9, which often describes a non-specific condition. However, there are many instances when the documentation supports the assignment of these codes.
Guidelines, rules, conventions and valid codes are not recognized by all payers consistently, causing providers to alter their claims according to the individual payer's instructions for coding. In addition to the tremendous administrative burden and waste of resources in modifying claims to fit a particular payer's requirements, consistency and comparability of clinical information is destroyed. With an increased focus on fraud and abuse, providers are faced with the dilemma of defining correct coding.
If a payer's policy conflicts with official coding guidelines or coding system rules, which way is correct? Given the time frames in HIPAA, we believe that in the short term, ICD-9CM as it is used for diagnostic coding and in-patient hospital procedural coding, and CPT for physicians' services should continue to be utilized as the initial standards for administrative transactions.
The administrative burden of duplicative systems is immense. In 1993, we issued a position statement entitled National Data Quality calling for a single procedural classification system. This position statement has been included with our written testimony for your review. AHIMA remains strongly supportive of the universal adoption of a single procedural code classification system. Although we have not yet had the opportunity to review ICD-10CM, we are hopeful that it will address some of the limitations and areas of confusion within ICD-9CM.
The parent system, ICD-10, is already being used in several other countries. It is important for the U.S. to maintain data comparability with other nations in order to conduct global research studies. ICD-10 will be implemented for mortality reporting in the U.S. in 1998. We recommend that the U.S. continue with its plans to move from ICD-9CM to ICD-10CM.
ICD-10PCS is still under development and testing. AHIMA participated on the technical advisory panel and in the system testing. We believe that ICD-10PCS has significant potential as a procedural classification system.
Volume Three of ICD-9CM, the procedural portion, has become outdated. It is impossible to capture precise procedural data, because distinct procedures may be assigned to some other code. It is conceivable for one system to serve most if not all purposes, so long as there is a mechanism for representing the needs of the various users.
We have poor data comparability and consistency today due to the use of multiple coding systems, disregard for the system rules, and adaptation to accommodate specific purposes. Rather than compounding this issue and causing further deterioration of the quality of our data, we need to limit the systems in use and develop better controls over how they are being used.
Numerous classification systems, nomenclatures and vocabularies are being developed in the hope they will be adopted for use in the CPR. All of these systems must be carefully and thoroughly evaluated, compared and field tested by an entirely objective entity before being selected for implementation. If more than one system is chosen, these systems must be evaluated for compatibility with regard to uniform definitions.
Our ultimate goal is standard data representation of health care services, not multiple systems providing conflicting, overlapping or duplicative information. All processes for developing and maintaining a data dictionary and classification system should include collaboration of the public and private sectors and be completely open, allowing input from individuals and organizations with a broad range of interests and expertise.
Any adopted system should incorporate an established organized process of annually updating the coding system, and this process should permit broad input from the public. There should also be an established process for developing rules and guidelines for the correct consistent use of the system and for addressing users' concerns.
The entity responsible for guideline development should include broad representation from users of the coding system. Any adopted coding system should be in the public domain without a copyright. This promotes open access to the process for updating the system and usage rules. It also lessens the cost to the health care system, since there are no licensing fees to be incorporated into the cost of products and educational services.
AHIMA recognizes there are significant resource implications in changing coding and classification systems. They include cost of new software for providers, payers and data users, interfaces between computer systems, massive educational initiatives for those currently working in the coding field, and complete restructuring of existing educational programs, cross walk development among coding systems in order to maintain comparability with historical data and complete restructuring of all systems based on classifications such as EPS, payment policies, performance measurement systems and provider profiling systems. However, with the development of a long-term transition and implementation strategy, there would be tremendous savings and improved process efficiencies and greatly improved clinical data.
We commend the Department of Health and Human Services for the establishment of a coding and classification implementation team. We are concerned however about the need for ongoing input from the health care industry. While this hearing will provide valuable input, we would recommend that there be established mechanisms such as an advisory panel to provide assistance to the department in the process.
In summary, AHIMA has identified the following key issues. Uniformity and consistency must be required of all health care organizations, payers and other data users. A central authority comprised of representatives from the public and private sector should be created to insure compliance with and enforcement of coding system rules and official guidelines. A realistic implementation plan for the introduction of any new coding system, including migration to a single procedural classification system, must be established.
Data quality is based on the data source, clinical vocabulary and classification system. It is important not to focus on only one component. A standard data set must be established. It is important to maintain data comparability with other nations in order to conduct global research studies.
AHIMA has had a longstanding relationship working with the department on coding and classification issues, and is committed to providing assistance to carry out the requirements of the law.
Thank you for the opportunity to testify. I'll be happy to answer any questions.
DR. STARFIELD: Thank you very much. Beth?
MS. WING: Thank you for the opportunity to speak to the subcommittee today.
NACHRI, the National Association of Children's Hospitals and Related Institutions, currently has 151 member and supporter hospitals. I would like to discuss how we use coding and classification systems at NACHRI, particularly in our classification research program area, and offer comments and recommendations for consideration to this subcommittee.
NACHRI uses the international classification of diseases, 9th Revision, clinical modification, ICD-9CM extensively for evaluation of in-patient, out-patient and population-based health care databases. NACHRI provides comparative analyses to the member hospitals using ICD-9CM. NACHRI also extensively uses and analyzes ICD-9CM through the development and maintenance of two classification systems. These two systems are the out-patient refined diagnosis related groups, APR DRGs, which NACHRI co-developed with 3M Health Information Systems, and is used for in-patient care analysis, and the NACHRI classification of congenital and chronic health conditions, CCCHC, used to evaluate population-based databases for disease prevalence, utilization profiling, capitation risk adjustment and linkage to measures of patient satisfaction.
Through the development of these two classification systems, NACHRI has carefully reviewed the content, structure and clinical meaningfulness of ICD-9CM, specifically as it is used to describe and classify neo-natal, perinatal and pediatric conditions. Review of the ICD-9CM codes has entailed checking for all index entries for a given code, researching conditions, diseases assigned to a code, and seeking medical input from our medical advisory committee.
Additionally, interviews have been conducted with all medical and surgical division chiefs at two of our member children's hospitals. Throughout this process, we have identified and received affirmation of the strengths and weaknesses of ICD-9CM.
The strengths of ICD-9CM, which represents anevolution of the ICD classification system, is that it currently contains the ability to classify thousands of known causes of death, established diseases and other conditions for which patients encounter health care. It is a widely known and accepted disease classification system, used for statistical and reporting purposes. Hospitals, physicians and other health care providers have long used these codes to classify diseases, index health records, conduct research and most recently, to bill for patient care.
The codes are used as a base of HCFA's prospective payment systems, diagnosis related groups, DRGs, and many other payment systems, and therefore ICD-9CM has assumed a central role for payment systems.
One of the greater weaknesses of ICD-9CM in terms of pediatric classification would be the lack of specificity in many codes in the congenital anomaly and perinatal sections. Some codes have been found to contain many different diseases, with a few codes containing as many as 100 conditions. Although many of these diseases may be rare or lower in prevalence, they can account for extensive in-patient hospital stays, multiple surgical encounters and out-patient health care consumption. There is a need to have many of these conditions further broken out of these existing ICD-9CM codes to more closely monitor the care and outcomes.
NACHRI has been active in the identification of necessary changes to ICD-9CM by regularly submitting proposals to the ICD-9CM coordination and maintenance committee, which meets twice per year. Some of these proposed changes have been adopted into ICD-9CM.
Although NACHRI is grateful to the purpose of this committee, the process is slow in meeting the requests of NACHRI for pediatric classification changes. Sometimes only three to four of NACHRI's proposals are able to be accepted for consideration per meeting agenda.
Many categories in ICD-9CM which need to be more finely divided have no more available codes, such as available fifth digits. The need for further refinement is great, such that the adoption of an ICD-10CM would achieve the greatest improvement in the shortest time frame.
Many of NACHRI's proposals for changes to ICD-9CM are based on reviewing the classification of diseases in the existing ICD-10 classification system. For pediatric conditions, ICD-10 has provided more specific codes, particularly for congenital anomalies, perinatal and other hereditary conditions.
Attached as an example are our most recent proposals submitted for the upcoming June meeting. Many of the proposals are recommending the ICD-10 classification system breakout of these conditions.
ICD-10 has been adopted into use in many other countries. It has been considered for implementation in the United States with a clinical modification as ICD-10CM around the year 2000. NACHRI strongly recommends the adoption of ICD-10CM as a standard of coding and classification of diseases and health care encounters no later than the year 2000.
Changing from the ICD-9CM to the ICD-10CM classification system will require resources for training and changing computer databases. This is to be expected with any major system change. However, the need for greater specificity far outweighs this temporary expenditure of resources. The level of specificity of pediatric diseases is needed to further refine the earlier described classification systems, as well as to improve upon ongoing analyses of pediatric health care.
In addition to ICD-9CM disease classification review, NACHRI has been actively involved with the development of ICD-10PCS for procedural classification through HCFA's ICD-20PCS technical advisory panel. We are familiar with the latest refinements towards the development of this NACHRI and its differences to the existing ICD-9CM procedure classification system.
As with its disease classification counterpart, ICD-9CM for procedures needs to have many categories subdivided in more detail. Again, the needs over extend the available space. It has been strongly recommended that there be one system adopted for procedural classification. It is confusing to remember which system, ICD-9CM or CPT-4, is used, depending on which care setting the patient encounters and which payer is reimbursing the treatment.
In addition, because both systems are not consistently applied to all treatments, it is difficult to report and analyze the care. Cross walks are currently used between these two systems, but at times are confusing.
While NACHRI does not at this time endorse one classification system over the other, we would like to recommend that there be only one adopted for procedural classification, whether it be ICD-10PCS or CPT-4 or a combination of both. This would simplify collection and evaluation of the data.
Another group of disorders that can be difficult to evaluate under the current practice are neoplastic diseases. The second edition of the International Classification of Diseases for Oncology, ICDO, provides the ability to classify the morphology of the neoplasm. Having this additional information available can help greatly with profiling, prognosis and evaluation of utilization of services for these conditions.
Most of the neoplasms and especially the solid tumors can be better classified using both of these classification systems together. Many facilities and health care providers have internal tumor registries and already collect the ICDO code in that database. These codes are usually not externally reported for other purposes, and therefore only the ICD-9CM code, which usually only describes the site of the neoplasm, exists in a health care facility's central database or outside agency databases.
Many neoplasms are unique only to their age group, in that they do not occur in adults and vice versa. Studying these diseases strictly by site and age is not as clinically meaningful as knowing its tissue of origin or histological behavior. Since many facilities already internally assign and collect these codes in a separate database, NACHRI recommends that a standard be adopted to require the collection of both the ICD-9CM and ICDO codes when reporting neoplasms externally, as well as in a given health care provider's central database.
Currently, the guidelines for coding and reporting define different rules, depending on the type of health care setting that the patient encountered. There are differences in these guidelines, depending if the patient was treated in an out-patient or an in-patient setting.
An example of this difference would be the reporting of a suspected condition. If the patient was admitted to an in-patient setting for a certain condition as ruled out or suspected, the in-patient coding guidelines instruct that the condition is reported as a confirmed diagnosis. For example, if the patient is admitted for the diagnosis of seizures rule-out, the ICDO for epilepsy would most likely be reported. If the patient was seen in the emergency room with the same diagnosis, it would be coded and reported as seizures.
NACHRI recommends that these guidelines be revised to adopt one set for all health care encounters. ICD-9CM currently has a section of V codes which are supplementary classification of factors influencing health status. In ICD-10, these have been retained as part of the classification system. Many of these codes are necessary to know the current status of a disease such as transplant status, personal history of a neoplasm, presence of a device and the dependence on a device. Many times, these codes are not reported because either they are not reimbursed or a health care facility adopts an internal policy that the information is not needed. NACHRI recommends that V codes or their equivalent ICD-10 codes be required where applicable to report the status of disease processes.
As previously mentioned, NACHRI actively participates in the semiannual meetings of the ICD-9CM coordination and maintenance committee. The role and function of this committee provides health care providers and others interested in the classification of diseases and procedures with the opportunity to regularly propose updates to this classification system. This means as new diseases are discovered, or disease process change or are differently understood, the classification system can be changed to accommodate those needs.
NACHRI recommends that this process be continued beyond the adoption of standards addressed by this committee, whether it be continuing to update ICD-9CM or hopefully moving into ICD-10CM, but maintaining that system as needed, and that this process be fully staffed.
In conclusion, NACHRI would like to emphasize the importance of clinically meaningful classification systems to continue to study the morbidity and mortality of disease processes, and consistency across all patient settings is most important. We believe the above recommendations will help accomplish this goal.
NACHRI appreciates the opportunity to have provided input to this subcommittee for this important decision process.
DR. STARFIELD: Thank you. That was a powerful panel. We are open for discussion. Questions?
DR. VAN AMBURG: Miss Stewart, you talked about potential inaccuracies for statistics using ICD-9CM and CPT-4, where the panel member on your left talked about this being an accurate representation of patients' conditions. Would you like to discuss that further?
MS. STEWART: I'll be happy to start if you would like for me to. I feel like our view was that the data, because there were two different systems used to describe the same procedures often, like in the in-patient setting, it is coded one way for the DRGs and prospective payment system via ICD-9 and by the physician for his practice on the HCFA 1500, where the CPT codes did it.
There really and truly is no -- in our view, accurate cross walk that can be used. So therefore, the data is often inconsistent, because the two systems really and truly are not in sync. That was our --
DR. VAN AMBURG: I thought you said the counts of diseases were inaccurate.
MS. STEWART: The what?
DR. VAN AMBURG: I think in your testimony you said the counts of number of people with diseases was inaccurate.
MS. STEWART: Oh, what we were talking about there is, often as she mentioned, in the in-patient setting you code a rule-out or a suspected diagnosis as if it were established. So it looks like, even though it may not be concluded, that is inconsistent with how the physician would code the same diagnosis for an out-patient encounter. Someone in the ER would be coded differently than in the in-patient setting. So it may skew the data in that way.
MS. LEON-CHISEN: I think one of the other things too was that there are codes that are perfectly good codes as far as the classification system is concerned, where that is the only way you can code it. One of the examples was the V codes. For example, someone comes in for care of their gastrostomy tube, and that just happens to be a V code, and some insurance companies will not accept it. Sometimes it is forced into another code so that it will fit that computer system.
DR. BRAITHWAITE: I'd like to bring up the question of the timing of the switch from ICD-9 to ICD-10. Miss Stewart, you recommended that we move to ICD-10 in 1998 as programmed. Yet, you recommended that we adopt the initial standard, the ICD-9CM standard under the HIPAA or K2 regulation. But that doesn't go into effect, in terms of people actually adopting and using it, until the year 2000. So there seemed to be a bit of inconsistency there.
Then Miss Wing, you might want to make some comment about that as well, because you recommended that we move well before the year 2000 to ICD-10, if that is possible. So I would like to get some comment going between the two of you about that.
MS. STEWART: When we were talking in our testimony about ICD-10 for mortality reporting, not the ICD-10CM, clinically modified, would be in use or should be in use by 1998. That is what we were referring to there.
For a procedural coding system, we said that CPT as it is now for out-patient procedures, it would be our recommendation to continue with that until such time as the ICD-10PCS, the procedural coding portion of that, could be evaluated. We needed further evaluation for that.
So if that helps you to clear that up, the ICD-10 for mortality, not the CM.
MS. GREENBERG: I should maybe just clarify that although the original date for moving to ICD-10 for mortality was 1998; it is now 1999.
MS. STEWART: We hadn't communicated that as well as we should.
MS. WING: I think NACHRI was just concerned about the implementation of ICD-10CM going beyond 2000. With the proposals that we submit each year, many times they just reflect what is in ICD-10 right now. We think it would be a quicker process to implement ICD-10 than to continue to recommend changes to ICD-9.
DR. STARFIELD: And write ICD-10CM?
MS. WING: Yes. I guess we're thinking of an ICD-10CM system, yes.
DR. BRAITHWAITE: I have heard from other sources that converting to ICD-10CM however should take longer than the year 2000, because of the effort that needs to go into training and getting people ready for that. Do you have any experience with that?
MS. WING: I have experience with training, but I don't think I have any further comment on the time frame.
DR. BRAITHWAITE: Do you think it can be done before the year 2000?
MS. WING: I think if we put our minds to it.
MS. LEON-CHISEN: One of the concerns that we had regarding ICD-10PCS is, we would like it to be tested first, and then allow us time to train our members before we would go ahead and implement it. I think maybe that is what you were speaking to when you were talking about training the coders, the physicians, and changing the information systems that should be able to accept this. With ICD-10PCS, you are talking about a different code structure altogether, and more digits to include in the information system.
DR. BRAITHWAITE: Well, the year 2000 will be a great year to make that change.
MS. LEON-CHISEN: Yes, I know, make all the changes.
DR. BRAITHWAITE: That's right.
MS. GREENBERG: You were talking about ICD-10PCS. I guess I am a little unclear on your position about ICD-10CM.
MS. LEON-CHISEN: For us, I see 10CM and PCS were going together. So I wasn't making a distinction between the two.
DR. STARFIELD: Do you have any questions of each other?
DR. HUMPHREY: I just want to follow up on the last comment. You said that you felt that ICD-9PCS required field testing --
DR. STARFIELD: ICD-10.
DR. HUMPHREY: ICD-10, required field testing and what have you, and that certainly it has been announced in the past that ICD-10CM and ICD-10PCS would be implemented at the same time.
But I would just like the panel to clarify whether they feel differently in terms of the expense required to implement or the difficulty of training or whatever on the CM side of 10 than they do on the PCS. I think it would be useful to have that distinction made by the panel. Thank you.
MS. STEWART: Do you want me to start?
DR. STARFIELD: Go ahead.
MS. STEWART: Going back to ICD-8 when it crossed over to ICD-9, that was a diagnostic system and a procedural system that both went to -- there was no distinction like there will be with 10 with PCS. That was very resource intensive. There is training, education, educational materials, et cetera.
Our feeling that to go from an ICD-9 as we have it now system to ICD-10 for diagnosis coding would be -- our history would say that that would be something that we could accomplish. Where we see it being more of a challenge is for all settings to go to one procedural coding system. That is what we feel like will be more resource intensive. But we, as I mentioned in my testimony, are not necessarily in a position to recommend ICD-10PCS or CPT either one, but we feel like there are obviously costs involved if you've got CPT on the resource -- fee schedule side, but you also have ICD-9 procedural coding, which makes up the DRGs.
So either way, if we go to one, it is going to be more significant than our change from eight to nine was. So I think that is a lot of what we had to consider in looking at the two systems and the switch.
MS. LEON-CHISEN: I think also, going from ICD-9 to ICD-10, it is not like you are going to a totally new system, where they don't have any overlap. Sure, there is going to be training --
MS. STEWART: You're talking about ICD or the CM?
MS. LEON-CHISEN: The CM, ICD-9CM to ICD-10CM on the diagnostic portion. There are still going to be some familiar things that coders can grasp and move from there, and it is not going to be totally new.
DR. MOR: You each alluded to the difficulty associated with keeping codes pure in the face of insurers wanting and providers wanting to up-code or change codes, in accordance with some set of reimbursement schemes.
Could you comment, each of you, on the viability of totally separating payment from coded activities and clinical activities?
DR. STARFIELD: I think you need to amplify your question.
DR. MOR: Okay. I suppose if all episodes of care, if everyone was capitated and you didn't have to make any notation about how to charge, these are just clinical events, would that affect your recommendations as to which way to go with one of these coding systems or not?
DR. FRAWLEY: I guess I would just like to comment. We used to code for many years before we had DRGs. So our members have a history of using many different classification systems. I hate to admit, when I was in school, the different classification systems that are no longer in existence that I was trained in.
But we have a long tradition of coding for data quality without issues related to reimbursement. It has become very difficult since the advent of prospective payment, because of the fact that we've got all these different rules and guidelines.
DR. STARFIELD: I think that is the importance of the question. I think we need to think about this.
MS. STEWART: Also, I would like to say that even if -- take away the reimbursement and our recommendations in talking about how it is difficult to deal with the different payers, clinical events or the procedures that are performed on patients, there still needs to be a standard and a central authority on the way those codes are assigned. Otherwise, it will just lead to mass confusion, which it is now. You won't be able to compare and base clinical outcomes that will have no validity, because they won't be following a system that is endorsed and mandated.
DR. STARFIELD: Would it change your recommendation about the coding system if you weren't thinking about payment?
MS. STEWART: No, it wouldn't. I don't think so.
MS. WING: In our recommendations, we were trying to exclude the coding as it is used for payment. It is a reality that it is used for payment now, but take away that and use it for disease prevalence and study of disease, there is still a need for greater specificity. ICD-9CM has outgrown that capability. I think we do need to move therefore to ICD-10.
MS. LEON-CHISEN: For our members, we stress that coding should not be just for reimbursement, because then you don't have good data that you can use for all the other reasons that you collect data. So it should not be the only thing that you look at.
So I don't think that it would make any difference whether the coding system was tied to payment systems or not, because what you're trying to do is collect accurate information on what happened to that patient, the condition of the patient, the services that were provided, whether you are getting paid for them or not, and whether this will provide you a higher payment or not shouldn't matter.
DR. STARFIELD: Does that answer your question?
DR. MOR: Yes.
DR. IEZZONI: I had a question about adding a flag to in-patient diagnoses that indicates whether the condition was pre-existing or during current hospitalization. It is an effort that a number of places like New York and California are trying to implement to improve the utility of diagnostic coding for quality assessment and performance measurement.
There are conflicting reports about how well that is working. I just wonder from your experience as coding experts what you feel about the accuracy of such flags and whether it is something that should be more widely implemented.
MS. LEON-CHISEN: I think it all depends on the accuracy of the documentation in the record. Oftentimes we forget that the coders, at least institutional coders, are working with a paper record, and there are different times that pieces of that record will not be complete. Lab results may not be there, and so forth. So oftentimes they will have to go through the documentation in the record and make a decision.
So if the documentation is there where the coder can actually make a distinction whether something occurred before or after the hospitalization, I think they would be able to collect the data.
DR. IEZZONI: But do you know what the current situation is? How is it working in California and New York, do any of you have any sense of that?
MS. LEON-CHISEN: I don't.
DR. IEZZONI: No? Thank you.
DR. STARFIELD: As an aside, I think we are hoping to bring in some of that state experience at our June National Committee meeting.
DR. SCANLON: All of you have a lot of experience in classification issues in settings beyond hospital care, acute hospital care and ambulatory settings, long-term care settings, for example. What other classifications do you typically come across, in terms of research uses or reimbursement uses or clinical uses? You spoke of CPT and ICD largely. Are there other classifications that you use, or have come across in other settings? There is an international classification for handicaps and others, functional status and disability. Are those used very often outside research settings?
MS. STEWART: There are several systems. In fact, as part of our testimony we started to bring that out, the number of systems, and we determined that it was really and truly too many to list, if you consider all the different entities and specialties that have the DSM and different things. But for our use, I would say the majority of our members that are coding health care data, it is predominantly ICD-9CM and CPT.
DR. COHN: I actually wanted to follow up some of our discussions around the ICD-10, in this case, procedure, not diagnosis. As we all know, it is in the final stages of being developed. But I'm curious from your view, if you were to mandate its use, what sort of field testing would you feel would be necessary to make you comfortable with having it out and being used in the year 2000 or 2001 or at some point in that frame? What would be convincing and comforting to you?
MS. LEON-CHISEN: We have started some of that testing. HIMA and AHA jointly sponsor a training session, where coders would go back to their hospitals and ambulatory care settings to actually code with real records. I think that is a good test of the system; is the information in the record matching what is on the classification system.
But in addition to that, we would like to see that the testing is carried forth in terms of the payment systems that are in existence now, how does it fit in with the DRGs, with APGs or ARPs or any other payment methodology that would be used.
MS. STEWART: Also, I think it would be important to apply it to all settings, not just hospital and ambulatory care. If we are going to talk about a procedural classification system that is applied to each setting, then see if in fact it can be used in the physician office for physician services, see if long-term care -- see if in the procedural classification system, is there a way that you can code for office visits and the consults like the ENM codes are now, is there something that can cross all those boundaries.
So it would have to be not just tested in a particular setting, but if it is going to meet all the needs, it needs to see if it can apply to all those settings.
MS. COLTIN: I think all of you made a recommendation to move toward consistency in the area of how diagnoses are coded in the in-patient and the ambulatory setting with regard to suspected or confirmed diagnoses. I know that the argument right now that I am hearing is in the in-patient setting, we see over diagnosis, that we may be overstating conditions.
I think we heard somewhat from the previous panel that that also is true in the out-patient setting even with coding guidelines, because of payment constraints that lead them to code things when they are not certain of the diagnosis.
I didn't see any specific recommendations as to how you would like to achieve consistency. One could try to enforce adherence to the current ambulatory coding guidelines, and then you would have to bring payers into line. You could move toward what is currently occurring in the patient setting, but then you've got the overstatement problem.
One of the suggestions we heard was a modifier code. Do you have an opinion on what might be the best way to reach consistency?
MS. STEWART: I think it would have to be some sort of descriptor or modifier that would say that this condition is suspected. Regardless of the setting, you can look at the data and tell that that was the case; it was a rule-out diagnosis. That would be what we would recommend.
MS. LEON-CHISEN: I think one of the concerns too is that there are guidelines, but not everybody is following them. Some payers accept them, some payers create their own guidelines and they don't even share them with the providers.
There is also the issue of education. The guidelines exist, but does everybody know that they exist and apply them consistently.
MS. WING: I would be in agreement with the knowledge and education of the guidelines. Obviously you have to study the guidelines and decide if there is a way to compromise the in-patient/out-patient world.
What we encountered when we went out and performed these physician interviews at our member hospitals is that a lot of times, people are just unaware of the guidelines, or as Nelly said, the payers have modified them to meet the reimbursement needs.
DR. STARFIELD: I am hearing some discrepancy. Margaret, you said you really think you couldn't get around it without a modifier. Consistency of the guidelines is another issue, because you could have different guidelines for in-patient, different guidelines --
MS. WING: I'm not in disagreement with the modifier. That is one area where the guidelines differ. I think the modifier is a good solution. But the guidelines in general, I think a lot of people just aren't aware of them.
DR. STARFIELD: But you think there should be consistency in the guidelines in the in-patient and out-patient setting?
MS. WING: Yes, I think there should be one -- try to come to one consensus on one setting of guidelines. Now, if it means introducing the modifier, I guess that is one solution. But I think it is confusing to have two sets of guidelines for two different patient types.
DR. STARFIELD: Does anybody else want to follow up on that one?
MS. GREENBERG: HIMA and AHA, are you also in agreement that the guidelines should be consistent across settings?
MS. LEON-CHISEN: Yes, but I also want to clarify that the guidelines that I think both of us are talking about are only guidelines for ICD-9CM.
MS. GREENBERG: Right.
MS. LEON-CHISEN: They don't exist for CPT.
MS. GREENBERG: Right.
MS. LEON-CHISEN: Or diagnosis.
MS. GREENBERG: It is a different kind of problem with procedure.
MS. LEON-CHISEN: Right.
DR. STARFIELD: Is there a question in back? No.
DR. RUSSLER: If I could ask a question of the coding experts.
DR. MOR: Would you identify yourself, please?
DR. RUSSLER: Oh, excuse me. I'm Dan Russler, a physician with HBOC and Company, HBO and Company, excuse me, a practicing physician out of Wisconsin, way back when.
At any rate, my interest in coding is going back to the original charge to this committee, the need to track resources and attach those resources to the treatment of diseases or problems or conditions. The question is, as I try to look at the coding systems and try to do that, I see for example, there is a procedure regarding taking of vitals. Taking of blood pressure is a big cost of treating hypertension.
But the question is, in your institutions, are you able to answer the question of how often are blood pressures taken and how often are those blood pressures connected to the disease of hypertension. I think that comes to the essence of coding. When we talk about the completeness of a procedure coding system that doesn't include the procedure of taking blood pressure, or a procedure coding system that doesn't include the procedure of prescribing erythromycin, those aren't coded procedures as far as I know in CPT or in any of the other procedure intervention codes. But how often do physicians prescribe?
Now, the reason that we probably don't have codes for them is, those individual activities haven't been paid for in the past. But the fact is, if we are going to try and measure quality and resource allocation connected to diseases, we need the procedural codes.
What I haven't heard is anyone addressing the simple daily practice of medicine, and being able to address the completeness of codes to describe the daily practice and attach it to the common everyday diseases that we treat every day, hypertension, diabetes, et cetera.
Maybe comments on that. Thank you.
DR. COHN: I would like also to defer to the coding specialists here, but I guess in my own paradigm, I will try to answer that one, at least try to.
I think Moon Mullins earlier talked about the issues of encoding as well as terminology. I think that were are obviously recognizing the need. I think this whole committee recognizes the importance of having very granular information about the process of health care. We however recognize that that is an issue that probably needs to wait for increased computerization of clinical data and probably computer-based patient records. So it becomes abstract if captured during the process of care.
I think we recognize that most of the clarifications we are talking about are some abstraction of that. Certainly as I am listening to Nelly, for example, talking about coding guidelines, I recognize that that is not something that as a clinician I would generally want to have to deal with. I would want to have that handled through a computer-based patient record.
But recognizing we aren't there yet, I think we need to look towards reasonable classifications. We need to look at, as I think this committee has commented on, things that don't get in the way of further and eventually increased granularity of data. But I don't think we can do that for the year 2000. Are there comments by the committee, or by our coding specialists?
DR. STARFIELD: We certainly have to do it by the time we make recommendations on the computerized patient record.
DR. COHN: Right, exactly.
DR. STARFIELD: Which isn't so far off.
DR. COHN: Agreed.
MS. GREENBERG: And unfortunately, if you think about the longer term issue, as we talk about classification systems, we have to keep thinking about clinical vocabulary and a standardized data set. These are all interlocking pieces as we move towards the CPR.
DR. STARFIELD: Fortunately, the committee has done some work on this. I'm sure you are all aware of the core data elements project in the report this summer. So I think we're not starting from zero.
MS. GREENBERG: Do any of you have any comments on the last question?
MS. LEON-CHISEN: Well, I do. In one sense, I understand your point about wanting to collect information on the daily routine activities that go into caring for a patient. But in this day and age, when we are still talking about people manually going to a record, looking up the code either in a book or in the computer, we're talking about trying to balance the fact that you do want to collect data, and at the same time, how much is it going to cost you to collect that data, and what are you getting out of it.
So most hospitals that I know of -- matter of fact, I probably would say none of the hospitals that I know of are collecting that level of data.
DR. STARFIELD: Unless they have electronic medical records.
MS. LEON-CHISEN: Right.
DR. STARFIELD: Which some do. I think we need to move to the next panel. Bill, you're on. We're going to try to be very efficient in the next hour, because we would like to try to adjourn between 5:30 and quarter of six.
We would like everyone to introduce themselves, please. Meryl?
MS. BLOOMROSEN: I'm Meryl Bloomrosen. I'm the vice president for health information services at Aspen Systems Corporation.
DR. STARFIELD: At Aspen Systems Corporation?
MS. BLOOMROSEN: Yes.
MS. SEARE: Hello, I'm Susan Seare, vice president and publisher for Medicode, Incorporated, in Salt Lake City, Utah.
DR. MARDER: I'm Bill Marder, vice president and general manager for the MEDSTAT Group, from Cambridge, Massachusetts.
MR. JIMENO: I'm Julius Jimeno, the chair of the ANSI Z16.2 Committee.
MR. KNOBLE: Gary Knoble of The Hartford, here representing IDMA, the Insurance Data Management Association.
MR. BROWN: Todd Brown with the Texas Workers Comp Commission. I am representing the IIBC.
DR. STARFIELD: So you are all special data users, is that right? Why don't we start with you, Bill? We'll go in the order that you are listed on the agenda.
DR. MARDER: Thank you. Good afternoon. My name is William Marder. I am vice president and general manager with the MEDSTAT Group. I have overall responsibility for our business with federal government research organizations and our commercial research clients. It is a pleasure to be here to discuss medical clinical coding and classification issues in connection with the recent enactment of the Health Insurance Portability and Accountability Act.
Prior to becoming a manager of research activities, I was an active health economists with interests in a variety of health services research topics, especially topics around physician supply. I mention this background because my training as an economist is similar to the experience that MEDSTAT brings to this panel. Economists are trained to analyze data sets that typically are designed and implemented by others. Training in economics contains large doses of complaints about data quality and hard work associated with doing the best you can with the data that are at hand.
Early in my career, I became convinced that timely investments of efforts in the design and implementation of data collection strategies would pay large dividends, especially when combined with sophisticated approaches to analyzing the resulting data sets. I joined MEDSTAT because the company operates in a way that is broadly consistent with this mixed strategy: try to collect the right data, but be sure to make the most of the data at hand.
MEDSTAT specializes in strategic application of health information and offers knowledge-based systems, consulting and research for improving the quality and total value of health care. The company services in excess of a thousand health care purchasers, managed care organizations, policy researchers and providers in the public and private sectors. MEDSTAT has been providing health care information products and services to our clients since 1981. With headquarters in Ann Arbor, Michigan, we have over 725 employees in offices across the United States.
MEDSTAT is in the health care information business, focusing on what could be termed post-encounter decision support. Our customers are trying to take action on important issues, but they are not trying to affect patient care as it is being provided. We are well known for developing claims and encounter databases that help purchasers or managed care insurance companies monitor and evaluate their impacts on covered populations.
I want to focus on four different MEDSTAT projects that form the basis for my comments on coding issues. The four projects are: standardizing health insurance claims and encounter data for self insuring employers, standardizing Medicaid claims data for the Health Care Financing Administration, standardizing hospital discharge abstracts for the Agency for Health Care Policy and Research, and reviewing and reporting HEDIS data for managed care plans.
In each of these projects, MEDSTAT staff encountered significant issues which are relevant to this panel. These issues are concerned with the comparability of data from different sources.
On the first project, MEDSTAT was founded in 1981 to help large self insuring employers understand and better manage the cost and quality of the health care utilized by their employees and their employees' dependents. National employers contract with different insurance companies around the country; are the local managed care and insurance companies providing equal access in quality at comparable cost. MEDSTAT builds analytic databases and decision support software and combines these products with consulting services to answer this question.
To build our decision support tools, we needed to deal with the large local variation in coding practices. Comparing access quality and cost after all depends on an understanding of how the health care needs of local populations might differ. Our goal was to provide appropriate adjustments to the data before advising our clients about alternative actions. We invested considerable time and energy to understand the coding practices that were acceptable to private insurers around the country. We did this to distinguish between real differences in disease and treatment and variations in coding practices.
One of the first lessons learned from this exercise was that when the national coding system does not closely meet the needs of those paying the bills, there will be a proliferation of local solutions. MEDSTAT training is like the training I described for economists. We learn how to cope with local coding variations and to derive as much value as we could from the imperfectly comparable data that are available.
The second project is to standardize Medicaid claims data. The research and policy division of MEDSTAT that I head was built from Systemetrics, a California based company that provided research database building projects to the federal government.
One of the longstanding contracts that was started by Systemetrics and that was only recently completed was known as the tape to tape project. Tape to tape was funded by HCFA to provide comparable data from a limited number of state Medicaid programs. We encountered in the tape to tape project the same issues described above for private insurance. Even in the context of federal-state partnership to administer a single program, we incurred significant costs imposing the standardization that was needed by NCFA funded researchers.
As part of tape to tape, we learned all about the difficulties of mapping ICD-9CM procedural codes into comparable CPT codes. We grappled with cross walks among HCPX, UB-82 and UB-92 revenue codes, the SM-4 and other difficult problems. In general, we learned that no one system will meet user needs, and that users will adapt any system in order to meet the perceived requirements of their local situation.
The costs associated with the tape to tape project are a matter of public record. This historical record can provide a beginning estimate of the costs associated with the standardized claims data system. Take this current experience with the successor system to tape to tape, the Medicaid statistical information system can provide more recent data about the challenges of standardized administrative data systems in the Medicaid program.
Within the past two years, we have continued to provide technical assistance to HCFA and its regional offices as they grapple with the encounter data systems that Medicaid managed care plans are asked to develop. Our extensive work in this area depends on our ability to work with any classification system in use within the health care industry. As the move into managed care continues, there will be a growing number of information systems that use medical clinical coding systems. Introducing change into this pluralistic system will be challenging and inevitably costly.
The third project is the standardized hospital discharge abstracts. We worked under contract to AHCPR and its predecessors to construct the health care cost and utilization project, HCCUP database. HCCUP is a research ready file of UB-82 and UB-92 hospital discharge from a representative sample of acute care hospitals around the country. ICD-9CM coding seems to fit this in-patient discharge data very well. The level of abstraction provides reasonable diagnostic and procedural detail, without swarming the analytic files with minor differences that could either represent differences in care or differences in coding practice. Even in the hospital environment, with significant coding expertise in a well-adapted system, we incurred substantial costs standardizing data from different hospitals.
When HHCUP expanded to provide ambulatory surgery data, we encountered in this project the same problem that could be noted above for claims data. ICD-9CM and CPT contain different levels of detail for surgical procedure. There is no one to one correspondence between the two systems. If the institutions doing the coding need the detail for reimbursement purposes, then CPT will be used and used correctly. If no reimbursement depends on it, the detail of CPT is not likely to be consistently implemented.
In the context of the HHCUP project, we developed a disease staging approach to grouping diseases and examining complications. Disease staging abstracted further from the ICD-9CM view of diagnostic information provide a more manageable number of categories in a hierarchy of disease severity. We continued to use disease staging as a core method, but we applied the data from many sources to assist in the process of fairly comparing providers and plans.
The fourth project is on HEDIS data audits. MEDSTAT has been auditing self reported HEDIS data since the National Committee for Quality Assurance, NCQA, implemented the program. Even with the relatively straightforward indicators included in the HEDIS data sets, we have found that there is enough ambiguity in administrative systems that self reported data do not provide a reliable basis to compare managed care plans.
This year for the third year in a row, we are providing technical support to the California health care reporting initiative, CCHRI. CCHRI is a coalition of health care purchasers that provides a comparison of the major managed care plans in the state. As a result, we developed samples of medical records to investigate the status of each HEDIS indicator.
We continue to use medical records abstraction as the gold standard for HEDIS reporting. The diversity of information systems in place and the difficulties encountered when plans use their own administrative systems for quality measurement lead CCHRI and others to continue this relatively costly procedure.
In a variety of settings, we have found that the information systems of managed care plans and providers can provide useful information. As a provider of software, we have developed techniques that compute the HEDIS indicators from the administrative data systems. The software, coupled with samples of medical records to validate accurate reporting, has considerable promise for the future.
My written statement includes responses to specific questions. But in conclusion, I would like to thank you for the opportunity to address these important issues. Quantification is an important step in understanding and managing complex processes. Quantifying health care needs and services in a pluralistic system such as ours depends on standardizing information. The coding systems that you are examining will provide the information that companies like mine will use in the next century to help purchasers, managed care insurance companies, providers and researchers do their jobs more effectively.
We hope these comments have been helpful.
DR. STARFIELD: Thank you very much. They have been. Thank you for sticking to the time. Julius Jimeno.
MR. JIMENO: Gary Knoble will give you our remarks.
DR. STARFIELD: You are working as a team?
MR. KNOBLE: Actually, there are three of us here. The agreement was that I would deliver the presentation, and Julius and Todd would answer all the questions.
A full background of our effort is included in the material that we passed out to you. But the short story is that about nine months ago, a group of individuals drawn from the insurance industry, from the stage workers comp areas and the safety folks at the federal level and several other state initiatives, we all identified a common problem. The problem was that we did not have a good way of identifying occupational injury and illnesses in the workplace.
So we set about to see if we couldn't answer this question. I think that the conclusion that we have reached definitely has some impact on the issues that you are studying, and that is why we are here today, to present to you our recommendations that I think have an impact on what you're doing.
It is a slightly different look at the picture than you have heard to date, but I'll go right to the bottom line. We are recommending that we standardize the description and codification of occupational injuries and illnesses in the workplace. We think that if we do that, we can then use those standards as guiding principles in developing our systems as we move forward. When I say our, I am really talking about the systems of the constituents that participated in the process.
The proposal has three aspects. First, we are proposing that we use the ICD-10CM. We did this because the time frame we are looking at is beyond the implementation of ICD-10CM, that we use ICD-10CM for nature of injury and illness and part of body, that we use the ANSI Z16.2 standard, formerly called the OIICS standard, to identify event, source and secondary source. Those of us in the insurance industry refer to this as cause of loss. Then the third aspect of the proposal is that we standardize the reporting of occupational injuries and illnesses using the standards that are being developed by the IAIAVC, which is the National Association of Workers Comp Commissioners.
Why did we not come up with one method? Obviously, that would have been the preferred way. The reason is that we could not find one coding method that we felt met the needs of the users. Those needs were the need for high quality data, the need for compatibility with legacy systems to address the implementation problems, and finally, a system that had proven and recognized credibility.
The strengths of our proposal. We think it does several things. We think it improves the safety and health prevention programs. That is clearly one of its foundations. We think it improves data quality, because everyone will be using the same standards. We think it increases efficiency of the workers comp claims process. We think it reduces administrative costs, enhances analytical capabilities, and enables comparisons across users, so the different users have a way of comparing the results in different systems.
Why did we choose ICD-10CM for part of body, nature of injury, and illness? First of all, because as you have heard today, it has broad acceptance in the medical community. Second, it has international recognition. It is used by several other countries. Third, it has insurance acceptance. The insurance industry is seeing this as the method that the doctors use to describe nature of injury and illness, and we believe the doctors are the ones that ought to be describing that nature, not someone else.
There is certainly a receptivity on the part of the folks that developed the standard, to making changes to the standard. We identified early on some problems, and they came back -- the particular problem was identifying the dominant side, right or left. They came back with a modification to the system very quickly to recognize that problem. So we feel that they are receptive to change, receptive to working with us.
There is a process for timely updates. It is in the public domain. We have heard a lot of discussion about that today. It has statistical robustness. By statistical robustness, we mean that it is demonstrated to be usable. It is demonstrated that this system produces data that can be analyzed and can be used. Finally, there is the availability of a training program, a very good training program to train users in this system.
Why did we choose the ANSI Z16.2 standard for cause of loss or for event source and secondary source? Some similar reasons. Specialized system design; this refers to the fact that it was specifically designed for this purpose. Again, it has widespread acceptance and usage. It has insurance acceptance. It has a process for timely updates. It is in the public domain, and it also has been demonstrated to have statistical robustness.
In summary, we feel that using this combination to identify occupational injuries in the workplace has the right people collecting the right data at the right time. What do we mean by that? The medical diagnosis information will come from the physicians or the health care workers, where it should come from. The description of the incident will come from the employers or the employees. So the folks that are in the position to know the most about what is going on will be the folks that will be providing the data.
We feel it reduces duplication of coding. Today, we have multiple systems for identifying these types of injuries, and employers are often asked to fill out as many as three or four different surveys, using different coding systems to describe the same injury. There is a single data flow and coding scheme, rather than the multiple schemes that exist today. We feel that it improves communication across the different users of the data, and people are using a standard language. We feel that is very important, because that definitely leads to quality data.
So in short, we do believe that the employers, the employees, the community at large will all benefit from going to this kind of a coding structure. As I say, the details of the proposal are in your pass-out, along with the list of the individuals who participated in the development of this proposal. We obviously will be happy to answer any questions that you have at the end of the presentations.
DR. STARFIELD: Thank you very much.
MS. BLOOMROSEN: Subcommittee members, invited guests and speakers, it is a pleasure to be here today to discuss standards for coding and classification systems. My name is Meryl Bloomrosen, and I am the vice president of Health Information Services for Aspen Systems Corporation, which is an employee owned information management and technical consulting firm here in Rockville, Maryland.
For over 30 years, we have provided services to clients in both the public and private sectors. We have had the opportunity and the privilege to participate in numerous projects dealing specifically with coding and classification systems for health care providers, health data organizations and state and federal government agencies.
My remarks today are based on our collective knowledge and experience, as well as our enthusiasm and concerns that we share with you about the standardization of coding and classification systems.
By way of background, I also offer you my personal viewpoint as a professional who has worked in the industry as a health care administrator, a health policy analyst, and a data quality consultant.
We have worked with hundreds of health care providers throughout the country, reviewing hundreds upon thousands of clinical records for a variety of research and clinical financial and research oriented projects. On behalf of our many federal and state government clients, we have ourselves had the opportunity to collect, merge, analyze extensive data sets, and then publicly disseminate information on various health care and health related topics. We are very familiar with the overlapping needs of health data.
While the administrative simplification we are focused on during these hearings revolves around those data sets needed for administrative transactions, surely we all accept the fact that these data are routinely used for ever-increasing numbers of other purposes.
We have reviewed the questions that face the committee, and our remarks are organized to follow the committee's specific questions and conclude with some general comments that we offer for your consideration. We are pleased to share our views with you during your hearings, and would be happy to provide additional assistance as you proceed to finalize your recommendations.
I won't repeat the questions, I'll just walk you through the responses that I think are most important for this afternoon, and given the lateness of the day.
As I mentioned earlier, we work directly with providers in their efforts to report data to state and federal health programs, submit financial data for payment, and implement clinical information systems throughout their delivery systems. Our involvement with coding and classification includes ICD-9CM, CPT-4, HCPX, DRGs and APGs. Some strengths of the current coding and classification systems include their widespread use and acceptance and the structured nature already in place for the maintenance of these systems. The systems have been around a long time and generally speaking, the user community knows when to use which code system. This familiarity should not be taken for granted. The complexity of updating any coding system from the user's perspective is vital to their understanding of and ability to use the system itself.
Data systems have been set up to recognize and accept these codes. In general, users know when to report which codes and where to go for technical assistance regarding coding decisions. Systems, processes and procedures are already in place for the certification and continued checking of credentials of coders using these systems. Mechanisms have been established and refined over the years to address code interpretation, implementation and updates. Nonetheless, there are perceived and real weaknesses and challenges inherent in all these systems and in the established mechanisms for their updates and revisions. These include their limitations in keeping up to date with new and changing medical technology, devices and treatments.
While the codes themselves can be blamed for their failure to reflect the most current technology or treatment modality, one can also look to the fact that there has to be some control over the proliferation of codes to respond to every special interest group's needs or perceived needs. Thus, the time lines and the timing of adding or revising codes to reflect current medical practices is crucial for all code systems.
Other challenges include the need for each system to reflect the ever-increasing diversity of care provided by many different types of practitioners across an ever-expanding variety of provider settings and their inability to do so. Other issues relate to structural and taxonomic limitations of some of the systems due to compression problems, making the addition of new and revised codes difficult.
There is also well known incongruence between existing procedural coding systems. Additionally, there is a lack of uniformity in reporting requirements and code acceptance across data collecting entities. Data reporting instructions are often shadowed by whether a treatment or diagnosis is covered by a health plan or a program. This in turn influences whether codes that are established for a particular treatment or diagnosis and when codes do exist, whether they are accepted where they are reported.
The committee should explore what additional efforts are necessary to establish definitive limitations on variations of codes, code rules, interpretations of guidelines and reporting mandates. The health care industry cannot simply do business as usual when it comes to coding.
Finally, the burgeoning health care industry and its ever-increasing reliance on coded data as drivers of decisions makes the timely distribution of new code information, code application, interpretation and co-changes the keystone of the entire infrastructure.
DR. STARFIELD: Forgive me for interrupting.
MS. BLOOMROSEN: I'm not going to read the whole thing.
DR. STARFIELD: Thank you.
MS. BLOOMROSEN: I'm just mentioning things that I think are similar to some of what you've heard this morning and this afternoon.
With question two, I'll summarize for you, Barbara. Given the time frame, we do not believe it is feasible, practically and logistically speaking, to do anything other than accept existing code systems, ICD-9CM and CPT-4, in the immediate future. However, we believe that no matter what systems are ultimately endorsed for the future, that it is an opportune time for a strong public-private partnership to successfully comply with the regulations that we need to implement as it relates to administrative simplification. We offer as examples specific issues relating to the number of federal and state programs, information systems, commercial products and the like.
We talk a lot in answer to this question about the need for a seamless transition to any new coding system, so that we all don't have the industry come to what we call a grinding halt. It would be nice to believe that the transition can be smooth and seamless, and therefore we are suggesting strongly that the committee look at short-term and long-term recommendations.
We suggest that the most prudent approach at this time would go with the existing systems, but the fact that there are systems under development and in various processes of being tested should be looked at strongly.
One point I would like to bring your attention to is our response to question number three, which asks us whether we think the movement to ICD-10CM or ICD-10PCS should move forward, assuming that quote, field evaluations are generally positive. We don't think generally positive is a strong enough field test answer. We suggest that -- I don't know enough specifically myself, other than, I understand there are field tests under way. We suggest that -- why can't the testing include an assessment of the comparable performance of both new and any proposed revised systems against a set of measurable and perhaps weighted criteria, in much the same way researchers have scrutinized the various case mix measurement systems and severity adjustment systems over the years. The evaluation should help us understand how any new system can and will perform in in-patient and out-patient settings, in hospitals, in ambulatory care clinics, in the physician office setting, under fee for service or capitated and managed care scenarios, for all types of practitioners, including physicians, nurses and others.
We raise the question relating to, can the systems be tested and evaluated to look at how they will affect morbidity and mortality statistics so our trend data are not greatly skewed, and what precautions we will be able to take to mitigate inappropriate interpretation of the systems, and what performance tradeoffs is the community at large willing to accept, knowing that no system is ideal.
We are suggesting that we don't think it will happen before the year 2000 if the community at large does it right, in terms of migrating to a new system. There are many of us sitting in the audience and at this table who have lived through and can talk about the migration from I8 to I9. We know it can be done, but not without extensive public and private sector collaboration and involvement that needs to start extensively now if we know that we are migrating to a particular system.
We feel strongly that the education of the numerous users and coders across the country is so much more mammoth than it was 20 years ago when we sat through the I8 to I9 migration, that we should not take it just as a matter of happenstance that training will occur. We need to see how it will occur.
I think in the interest of time and so that you can ask some questions, because we have gone into quite a lot of detail for you with each of our responses, I'll just thank you for the opportunity to present to you, and offer our corporate support if you would like any additional information.
DR. STARFIELD: Thank you. Susan?
MS. SEARE: Hello. I'm Susan Seare. I'm the vice president and publisher of Medicode. I have so much to say and so little time; it is a great topic.
A very brief background about Medicode. We are in the business of developing and maintaining information databases. We are a health care information company. We disseminate that information in many, many ways.
I would like to just go briefly through some of the questions. I have just highlighted some of the main classification and coding systems that we use in our products, CPT-4, ICD-9CM diagnostic, procedural, CDT, the ASA guide, which hasn't been discussed here as yet, NCD. We do more workers compensation fee schedules for state jurisdictions than any company in the country, and we deal with HCPX as well.
We strongly feel that the claim and classification systems, all of them, should be evaluated as we are struggling with diminishing the costs and time spent on administration of health care.
We think the ultimate goal should be the fewest number of systems, with the greatest degree of usability. The granularity, type and use of information needed by this industry should be determined before any decisions on how that information is reported is made.
It is a little bit dismaying that I don't see a lot of that happening. We don't have a clear picture of what it is we are trying to collect and why we are collecting it before we talk about how we are collecting it. I think we should be drawing that map for ourselves.
CPT has several strengths. It is obviously used by a majority of our industry, and everybody knows it, it is very common. It is updated yearly. The AMA maintains this, they have an interest to see that it is disseminated widely. It includes some definition guidelines that aid in implementation.
Some of the problematic areas are that industry wide needs are not currently represented, and there is very little potential for review and public comment as the system is updated. It is nota pure system. The procedural codes have diagnostic terms, which limits some of their uses, and that should be evaluated as well.
Index terms don't provide ample granularity for insuring the correct code choices. I think granularity is an issue that needs to be heavily addressed by this committee.
The expandability of this system is limited, and escalating royalties and complex contracts do have a significant cost impact, which is contrary to the goals of administrative simplicity and cost containment.
ICD-9 has several strengths. It is open to public input, which assures that the classification meets the industry needs. The public domain allows that its costs can be market driven as its distribution is done by the private industry at large. And index terms in ICD-9 have a better developed granularity, although not where it needs to be.
I would think this committee should be a little bit concerned that several speakers here today are talking about developing alternative or adjunct diagnostic coding systems, because the current system isn't meeting their needs. We've got the ADA, who is developing their own diagnostic granularity, that is over and above that in ICD-9. I think these are the types of things that should be saying to you all that I-10 is something that maybe we should go to, that it is something we have to go to, and it is a matter of when, and not if.
The clarification is running out of space in some sections, which makes it a little more difficult for our group reporting. The conventions are not consistent throughout ICD-9, and there are too many opportunities for non-specific code choices. Even though in some instances, that is the correct code choice, there are many other instances that are provided that are not necessary.
The current processes allow protracted time frames for the time frames of new codes and terms. Vital statistical information is often lost, because the E codes and V codes are appendixes and not in the body of the classification. That has been addressed by others as well.
We could go through some of these others. I think just as a means of example, there are a great deal of overlaps in some of these systems. For example, HCPX has the entire CPT and then some within it, and CDT is part of HCPX. There are differences in the first digit. Coders often don't know which one to use and how to report it, so they are both reported. I suspect there is a significant loss of potential data because of that.
Moving on to CDT, it is only updated every five years. Anybody knows that a system updated every five years in the current medical industry is something that needs to be evaluated. The ASA guy is indeed more intended to be a relative value guide. It is updated yearly. It incudes definitive guidelines. While it also is contained in CPT, it also has codes that are not contained in CPT. Again, you are developing information systems that make it difficult to cross all of this data, and you are losing a lot of potential information that allows committees such as yourself to make decisions.
I have also incorporated in here the state workers compensation fee schedules as a coding system, which should raise your eyebrows a little bit. But in working with these various states, they are quite insistent on including their own state-specific codes. Some states have included as many as 700 codes that are specific to that state in spite of our recommendations and abilities to convince them to the contrary.
You will find that providers will in turn incorporate those codes into their systems, which will include a greater level of granularity than that available to them in the existing systems, and they will in turn bill that information to a payer, a commercial payer, who will in turn add that information to their systems to allow for payment. You have just altered their coding schema in a geographic area.
Again, these are the kinds of things that are saying there isn't enough granularity. The market is asking for more. I know that there is concern about ICD-10 implementation being very difficult. To that, I would have to say, DSM-4 is based on ICD-10, and the market has embraced DSM-4. I have not heard people saying, boy, this is really hard to use. It is easier to use. It gives people the granularity they need.
I have listed some things in here that you can go through, steps that I think you can take in the near future that would allow this committee to make the process a little easier, should the decision be made to go to a single classification system. I would leave that to your perusal.
I would suggest that you take a look at the NDC codes. In trying to map some of those codes, our company has determined that the various publishers of NDC databases have different ways of looking at it, different ways of publishing it, and the standards probably be chosen for this country.
Another concern I have is that people are worried about the complexity of ICD-10PCS. I think it is an intriguing system. It is infinitely expandable. It easily lends its results to statistical analysis. The current system is no such animal. Again, being compassionate to the pain level that change inflicts, the transition does need to happen. Even though the system is different, it still operates in a manner consistent with how business is done within the organizations required to use these systems.
Once the necessary coding systems diminish and the data focus within each system is the same, transitioning to a file national standard, if that is indeed your goal, is a little more manageable task.
I think paying attention to the indexes is very important. I know we have had discussion on a standard coding nomenclature or a standard medical nomenclature. I think that is important, especially with regard to levels of severity of diseases. But I think to limit to a simple standard vocabulary could probably be problematic, because you will lose granularity and the ability for the system to be as robust as it might be, given a greater number of choices and information allowing you to arrive at the right codes.
Public domain I know has been a concern for you. I have listed opinions of Medicode with regard to public domain and private sector pros and cons. Medicode doesn't have a great stand on a choice of one or the other, although we think there should be one procedural coding system, so that you're not losing data.
One of our concerns as a data company is that the data that we receive is very dirty. We have spent a great deal of money developing algorithms, some of which we have patented, others which are just standard logic-based algorithms to scrub the data.
One of our goals is to try to get the data to be as clean as possible when it goes into the system. To that end, we have tried to develop editing systems for providers prior to claim submission, and even at the point of service, when the physician is in the room.
I think that it would be unwise to think that we can have a data set that is clinical and a data set that is administrative. The ultimately way that this industry is moving is for this to happen all at once, and that needs to be kept in mind. The granularity should be there, regardless. All of this is going to be electronic. In the interim, the systems that we have don't function all that well, not if you're looking at data collection and being able to use this information to make data based decisions.
We have been mapping databases for some time. I went out on a limb and gave you what I thought a potential cost estimate would be, in the answer to question six. Since we have been doing this for several years in our company, I think the costs are going to fall with the vendors. I estimate for my company, the cost of updating product to these new coding systems -- and we have done some preliminary analysis in this area -- would be somewhere from $500,000 to $750,000. Now that cost would be spread across all products.
Whenever you start talking about change, you talk about an increase in number of sales, so the cost per unit would probably diminish. But I don't think individual providers would bear the brunt of these costs; it would be the people who developed the systems for them.
We have many contracts that are multiple year contracts that only allow for a CPI increase. So regardless of how all of this is implemented, it is the vendor who will make the changes and pass that information along to the end users.
So I think to that end, the health care vendors should have a significant say. They might help you in some of these decisions. I think that is one way to look at this.
There is obviously going to be a significant impact on training, and that is where I think a lot of the dollars will be, is in training, and then in the physician world of using physicians and clinicians and nurses to make the medical decisions that go along with these changes.
I appreciate the opportunity to speak to you today.
DR. STARFIELD: Thanks. We'll move directly into questions.
DR. COHN: I guess I'll start out. There is so much to ask, it is hard to know where to start. I actually thought I would have a question for those involved with ANSI Z16.2 standard. I would ask them if it is possible to get a copy of that standard, just because I'm not familiar with it.
MR. JIMENO: Yes, it is, and I will deliver it to you within seven days. You have to realize, that standard is based on -- if you look at the footnote on the bottom of the second page, it will tell you that that standard is based on the Bureau of Labor Statistics occupational and illness classification system, which is on the World Wide Web. The notation for that Web site is there.
So I'll give you the standard, and if you want the classification system, I have it on my desk and can deliver it to you within a short period, maybe not 25 pages. Being from EPA, we don't like to spread paper around.
DR. COHN: I was hoping maybe you would give it to Marjorie, and she would pass it around. The real question I had as I am looking at issues around events source -- maybe I'm mistaken, but it sounds like injuries for an E code? It seems to me, in the interest of administrative simplification, being a practicing clinician at a time seeing occupational injuries that I don't discover are occupational injuries until they are six months old. It seems to me that it might make sense that there be a standardized way to deal with these sorts of events or injuries, as opposed to one that is specific for occupational as opposed to others.
MR. BROWN: Let me say, I think that is exactly what our point is. A lot of coding systems try to encompass things that they really can't provide good information on. For example, when an employer records an injury, they can't tell you what that injury is. They know the person got hurt. So the information you get from the employer, or even the employee at the time of the injury may not be that -- they can tell you what happened.
So we think that information ought to come from the employer, as far as what happened, tell us the events. When it gets to the extent of the injury, we feel it needs to come from the health care provider that does the diagnosis and provide that information.
We collect that information now in Texas on every medical bill. We receive about three million medical bills a year on all injuries and accidents in the state, and we are able to analyze the types of injuries occurring off the ICD-9CM code. It is not as good as the ICD-10CM, but we are able to do that now. We receive those electronically from the carriers in the state. We have had that system in place for about three years now.
So that is precisely what we are saying. Instead of each coding system trying to answer all the needs of everybody, you need to break it up and allow those people that can provide the information at the right time and the right place to provide that information, and leave the other information to be provided by those people that can best provide that.
DR. STARFIELD: I just want to make sure, you're talking about the E codes when you talk about the ICD, is that right?
MR. KNOBLE: Yes. It may not have been clear from the presentation, but the participants have agreed that if we go to this, the ICD structure would drop its description of the incident and would adopt the ANSI standards to describe the incident.
So we are saying, we would take half of the ICD- 10CM and you would take the ANSI secondary source, and combine them to describe the part of body, nature of injury and the source, or the thing that caused the injury.
DR. STARFIELD: What part of the ICD are you not going to take then?
MR. KNOBLE: The E codes.
DR. COHN: And the NCHS has agreed to that?
MR. KNOBLE: In principle. Through these discussions, they have been participating in the discussion.
DR. STARFIELD: I think there is some confusion here. Your program is to not use the E code system, but to use the ANSI system.
MR. KNOBLE: Right.
DR. STARFIELD: I had a question as to what comparability there is between the two of them, both nationally and internationally. But I think we need to be on the record that this would not be a substitute -- this wouldn't happen in the use of ICD by anybody else. There is no intention as I understand it to substitute this --
MR. JIMENO: This would be for the occupational environment, for workers comp claims. The importance of that is that you have to recognize that there are standards from the Occupational Safety and Health Act. There are obligations by the Bureau of Labor Statistics to gather data, to monitor the occupational injuries and illnesses that occur in this country. We can't lose that knowledge. We can't lose that ability to make progress in our occupational environment.
If we are able to recognize and reconcile that difference, then you may be able to eliminate a lot of duplication, overlap and duplication for the physicians that don't have to worry about the source and injury or the E codes for the ICD-10.
DR. STARFIELD: But the medical community always does have to worry about that. If we're going to do anything about understanding why things happen, we can't separate that off.
See, we talked earlier about the need to link information from source to source and across patients. What I think you are recommending is to have two different systems, one for medical practitioners and one for occupational --
MR. JIMENO: I guess that is my point. When you talk about what is the extent of the injury, that is the medical. The doctor is there, they are looking at the person, they are saying here is the extent of the injury. As far as what actually happened to cause that injury, the doctor wasn't privy to that. They are only taking what the person told them or they are making certain surmises.
We think the information that should be relied upon to determine nature and cause of injury ought to be at the source that it happens, not necessarily in the doctor's office, because they are not at the site when the injury occurs. They are not on site, watching the injury. They are either taking it from what an employer told them or an employee told them.
DR. STARFIELD: But we're not talking about who gets it, we're talking about whether you code it the same way.
MR. KNOBLE: I think you're missing the point. It is only applicable to the occupational part, because that is the only arena we have been dealing with. But what we are saying is that the physician simply doesn't have access to knowledge about what happened. It is either the employer or the employee that has knowledge of what happened, and the structure that we have found, that we think is best suited to describe that, the coding structure we have found, is the ANSI Z16.2. That is just an observation we are making to you, that in the analysis that we have done, we feel that is superior.
MR. BROWN: But the one we are talking about is the BLS codes that have been in existence for 40 or 50 years, ostensibly to look at causes of injury in the workplace, to do the type of research that will result in good preventive actions occurring. So it is not that this is a group that has dealt with this for quite some time.
DR. STARFIELD: The comparability then comes from everybody's comparable use of the ICD.
MR. JIMENO: Yes, from the medical side. For the first time in this country, we may be able to link occupational and non-occupational injuries and illnesses that occur, and take a look at rehabilitation issues, prevention issues, whether they are occupational or non-occupational. This brings a partnership between the two groups of people that are trying to deal with minimizing the suffering that occurs for injuries and illnesses in the country.
DR. STARFIELD: Would you do away with the E codes?
MR. BROWN: Part of it is, right now when you talk about the amount of problems the health care provider has with all the paperwork, you've got the same problem on the employer's side. They basically have to file a different report with different mechanisms for OSHA, different report with different mechanisms with the state workers comp, and different mechanisms for BLS. Then they are also talking now about trying to use the ICD-10CM in some aspects.
So maybe it is just in that realm that you would use it for the occupational injury and illnesses, but by doing this, we feel that we can move towards eliminating those three type of report forms for employers. They would fill out one report form, and the users would go to one source.
We are already going to be testing that this next year with BLS. We have already proven it with OSHA in this thing. But it is literally the elimination of those three paperworks for employers that they are having to keep up with.
DR. MOR: I have a question. Is this kind of health care transaction covered under the law?
MR. BROWN: Let me just say, the IBC has spent the last ten years trying to bring together a single source report of injury for the country out of the 50 states. We have worked with all 50 states, at least 300 carriers and numerous employers. We are already successfully transmitting first reports of injury electronically in a number of states.
DR. COHN: I think we need to get some supporting documentation on this, because it is hard to talk about without having looked at the coding system.
MR. JIMENO: I understand exactly what you need, and I'll deliver it to you. We will deliver you our answers to the questions within a couple of weeks to 30 days.
MR. BROWN: Would you like to see the information regarding first report of injury with the standards that have already been developed and used in 50 states right now?
DR. COHN: Yes.
MR. BROWN: We'll get that to you as well.
DR. MOORE: There is also first report of injury out of X-12 that was submitted.
MR. JIMENO: That is the same group.
MS. GREENBERG: I assume that your system is much more detailed for occupational information about the causes of occupational injury than is in the E code system, either in ICD-9CM or is intended for ICD-10CM. But that is clearly the point that you're making here.
But has there been a mapping? Does this greater detail map back to the less granular E codes for similar types of injuries? Or is there -- that is the question I would have. Is there comparability? If you wanted to look at injuries that were suffered on the site at occupation -- at the job site, and similar injuries that somebody suffered from recreational activity, would your codes collapse back into some of these broader codes or in the E codes, or is there just no relationship?
MR. BROWN: At the present time, there is very little relationship, primarily because all the work historically has been trying to go between what the insurance had, called the DCI, detailed claim codes, information, back to the BLS codes. No one fathomed that possibly because of the Kassebaum bill, this was going to force going back in the ICD and CMs.
Can it be mapped back? Obviously, I think it can be. I don't think it is an impossibility. I think you could map it back. Will it take some work and energy and effort? It certainly will. But I think it would not necessarily be a reason to hold up implementation of some of these things. I think as long as you allow people to, they will wait until next year.
MR. JIMENO: We will address that in our submittal.
DR. COHN: This is toward Susan and Meryl. I am impressed with the two of you sitting together, because you represent seemingly major extremes, in terms of your views around implementation of the coding systems and all of this.
I wanted to hear a little bit more from each of you. I heard from Meryl a very strong statement of, this needs to be tested, it needs to be better than just okay, it needs to be -- maybe the testing needs to be part of the implementation. Susan, you said we've done it before and we'll do it again.
Can I hear a little more? Is there some consensus between the two of you? I need to understand a little better, because it is so black and white between the two of you, on this one. Meryl, do you want to start out?
MS. BLOOMROSEN: I think rather than just say that any one of or one or two of the systems that are being discussed, ICD-10CM as an example, I have been looking forward to ICD-10CM in some sort for the last 20 years. We have been told it is coming now for 10, 20 years. We grew up thinking ICD was revised every 10 years. The 10-year mark came and went, and now we're waiting for the next 10-year mark. So I am patient.
I'm just saying that as part of the assessment, why not give it a little more rigor, and make sure that the deficiencies that we are saying exist, the biggest ones, ICD-9CM doesn't address the needs of multiple -- and any other system, for that matter, CPT or ICD -- doesn't address the needs of managed care, it doesn't address the needs of nurses. We have heard that today. Let's be sure that whatever we are replacing it with does better.
In our remarks, we say throw the baby out with the bathwater for the hell of it, give that a code and move on to the next system. I think there is probably not that much difference. I think we are saying that a new system probably is valid, but not just to jump without looking at it a little more carefully.
I'm not comfortable in answering anything more definitive, because I'm not really sure if I am privy to know exactly how the currently proposed systems are being tested. I am familiar with field testing. It is my understanding, for instance, that there will be some records reviewed using this new proposed system. I believe the records are coming from hospitals.
I don't know the answer to how the field testing is currently envisioned or proposed. I know that there has been a lot of effort in the past to more rigorously look at other systems. The example in my remarks was severity systems and case mix systems, because that has been scrutinized quite a bit and it continues to be. Why just accept at face value that the proposed new system will solve all the deficiencies.
I'm not saying we have to do a -- it potentially would be extremely costly, but perhaps there is at least some pilot testing that could be done. We are also suggesting a public-private sector partnership. I think that the amount of dollars that will be spent by both the public and private sector, the providers, the vendors, the systems people and the like, is going to be large. That doesn't mean we shouldn't go forward with a change, but we should be looking to pooling our resources potentially to collaborate to see how it can happen best.
I would suggest that a pilot kind of a test would allow us to see if there are major deficiencies in any of the proposed systems, which could be addressed before their implementation.
So I think we are in probable agreement in the sense of migrating towards a transition, but I don't think -- if all of that were to occur in a systematic, strategically planned way, we are going to do it in the next year or two.
DR. FRAWLEY: Just to make a comment, Meryl was part of the team that was doing the training workshops all over the United States when we were converting to ICD-9CM. That is an enormous educational effort. I think her testimony highlights that point. It has got to be really well thought out. It will be a massive educational initiative in terms of the time frames.
DR. STARFIELD: I heard you say not only in hospitals.
MS. BLOOMROSEN: Our clients are hospitals, HMOs, single providers, multiple clinics. The range of what they are doing today includes having no code books at all in front of them. They are using what you and I would call cheat sheets. It is the 10 or 25 most commonly used codes. Those are the ones they are used to. So the talk of technology and computer systems would be shocking to these people. There's resource limitations.
People we have worked with, clients that are using information systems for a lot of things, but not necessarily for assigning codes. The codes that are in the systems are not necessarily the ones that are on the bills. I would highly support that the committee and the department could contribute significantly to mandating no deviation from established guidelines. I think you have heard that from almost all of us as a uniform kind of statement.
What providers and clinicians and vendors are dealing with across the country by interpretation and payer specific data collector specific nuances about how the system uses the codes is what is really driving people crazy. And administrative simplification can in large part occur by not allowing all this diversity of what I have to report from one payer or another, this kind of a claim and so on.
I think we are suggesting because of testing of this sort, it might be expensive, that there might be some collaboration that is available to look at how people would like to see it happen, and do it right.
MS. SEARE: I mainly agree with Meryl, but I will say that coders and people involved in that process are a whole lot more sophisticated today than they were when we moved to ICD-9, a lot more sophisticated. We receive at our help line desk about 3,000 calls a month. We code over 5,000 operative reports a month. The kinds of questions I get today in no way, shape or form match what I got five years ago.
I get asked statistical questions, I get asked questions that I can't even answer, and I have been in this business for a long time. So I'm not so sure that it is going to be that big a leap in the minds of coders to make some of these changes. I think many of them will embrace them, because the systems will answer the deficiencies they have been trying to deal with for years.
In answer to your original question, which I have just forgot was, -- what was your original question?
DR. COHN: You are actually doing a fine job. The question was, why are you both in such disagreement?
MS. SEARE: Right, but coders are more sophisticated. That is one aspect. I think that the other aspect is that systems are more sophisticated. I can tell you that our sales of low end look-up coding software has grown in multiples of ten. I don't think it is going to be that big a stretch for people.
I also don't think that we are talking about systems that are new per se. ICD-10CM has been around for awhile. It is being used in other countries. We are not talking about something that -- I think HCFA has looked at in depth. I know that studies are ongoing on the clinical modification and the recommendations for that. That has been happening for at least a year, if not more. ICD-10PCS has had incredible input from the industry.
So I don't think we are talking about systems that are so new that they haven't had testing and haven't had some significant input, because they have.
MS. BLOOMROSEN: I guess that is probably our main difference of opinion. I think that there has not been a rigorous testing, and there should be.
MS. SEARE: Yes, more testing, but --
MS. BLOOMROSEN: There has been a lot of input. I think that -- and my remarks say that when we look at the year 2000, the literature out there that suggests how all of us across not just health care but industry in general are scared to death about making a two digit change in our computer systems related to the year 2000. We are talking about making significantly more digit changes with any of these proposed systems.
I think we are talking about mammoth systems implications, not that we can't overcome technologically, but from a people perspective. There are still doctors documenting in medical records. There are still coders trying to interpret what they are saying, either through an automated end coder or a book.
I think that we should look at it as an opportunity to do something right, and make some enormous progress, so that the health data that we are collecting is sound and reliable health data for all of us, not just the hospital based people, but throughout the health care environment in the country.
DR. STARFIELD: I'm going to excuse myself, leave not so quietly, and turn over the chair to Kathy Coltin. Thank you very much.
MS. COLTIN: Are there any other questions in the committee? From the audience?
MS. HUMPHREY: I would like to ask Dr. Marder, I was interested in your comments. You said that you have done all these projects and you have done all of these studies and you know that every payer is changing the system because it doesn't meet their needs. So I was wondering if there is any pattern to the way they change it.
It seems to me that on the one hand, I think I have heard that it is not specific enough, but on the other hand, I have heard many people say that when they try to use the specificity that is in the system, the payer says I don't accept that credit.
So is it mostly a trimming of codes which is the changes they are making? Can you characterize them?
DR. MARDER: Yes. They are going for the extra granularity when they want to get paid more money. Frequently they are successful. Our task ex poste is to figure out what these local codes mean.
MS. HUMPHREY: (Comments off mike.)
DR. MARDER: I think the best examples will probably be HCFX level 3 codes that show up in the Medicare program to describe something that isn't readily available. One can collect from the public use files lists of same. But that would be the best class of examples.
Our traditional goal has been to contact the insurer, ask what does this mean, develop relationships with the coding specialists at the insurance companies. So the investment on our side is in the information transfer that happens over the phone. We got your last claims, we can't map these five things to anything. What do they mean? You can get answers to that question.
MS. COLTIN: Other questions? I think we can call it a day. Thank you all.
(Whereupon, the meeting was concluded at 5:45 p.m.)