of the
to the
Subcommittee on Health Data Needs, Standards and Security
National Committee on Vital Health Statistics
Department of Health and Human Services

Re: Physicians' Current Procedural Terminology (CPT)

Presented by: T. Reginald Harris, MD

April 16, 1997


Mr. Chairman and Members of the Subcommittee: My name is T. Reginald Harris, MD. I am a practicing internist and Chairman of the American Medical Association's (AMA) Current Physicians' Procedural Terminology (CPT) Editorial Board. I want to thank you for this opportunity to present the AMA's views before your Subcommittee to discuss the CPT and its future. As I will describe in more detail, the AMA has a long and successful history with respect to the research and development of CPT. The AMA continuously updates the CPT with the advice and counsel of physicians, payers, policy-makers and other interested parties.

We welcome the opportunity to talk with you about this important matter and, in doing so, hope we can help the Subcommittee and the full Committee in its extremely important task of making recommendations to the Secretary of Health and Human Services (HHS) on the appropriate code sets for usage in standard financial and administrative transactions, as required by the Health Insurance Portability and Accountability Act (HIPAA) of 1996 (PL 104-191). There is no question that the recommendations made by the Committee, and eventually by the Secretary, will have profound implications for not only the world of coding systems and data collection but for the future of the entire health care system.


As you know, CPT is an organized listing of descriptive terms and identifying codes for reporting the services of health professionals. The purpose of the terminology is to provide a uniform language that will accurately describe medical, surgical, and diagnostic services. Accordingly, CPT provides an effective and unmatched means of communication among physicians, patients, and third-parties. November of last year marked the 30th anniversary of CPT. The first edition of CPT, or CPT-1, was created and copyrighted by the AMA in 1966. When the AMA began the development of CPT in 1966, it was one of many procedural coding systems that was either in existence, or that would be developed, within the next 10 years. In fact, in the late-1970s it was estimated that there were over 250 different procedural coding systems in use in the United States.

In 1970, the AMA published a second edition of CPT which, in addition to adding many newly described medical procedures, converted CPT to a five-digit system. A third edition of CPT was introduced in 1973; the distinguishing factor in this third edition was the use of two-digit modifier codes that allowed physicians greater precision and flexibility in reporting. The fourth edition of CPT, which is commonly known now as CPT-4, was first published in 1977. This edition reflected numerous revisions prompted by rapid technological development. It also included a totally revised approach for reporting of anesthesiology services. Perhaps the most important substantive revision to CPT since the introduction of the fourth edition occurred in 1992 with the introduction of an entirely new series of codes and descriptors for the reporting of physicians' evaluation and management (E/M) services.

Beginning in 1984, the AMA began publishing the CPT book annually and those various annual revisions are now referred to as CPT 1996, CPT 1997, etc. All of these later revisions are still within the framework of the fourth edition (i.e., CPT-4). Obviously, with so many different coding systems in use in the late 1970s, it was very difficult to collect any reliable data on medical services being provided to patients, and made implementation of national payment policies, such as the type often needed in the federally supported health care programs (e.g., Medicare and Medicaid) nearly impossible. Because of this difficulty, in 1981, the United States government undertook a study to evaluate the existing procedural coding systems and to determine if any of them were suitable for adoption by Medicare as a national uniform standard. Based on the government's independent evaluation, it was determined that CPT was the best available system, and was adopted by HCFA for the Medicare program for the following reasons:

! CPT could be implemented nationally with a minimum of disruption to existing data processing activities;

! CPT could be implemented without fear of increasing costs to the health care system;

! CPT was acceptable to the medical profession; and

! CPT had professional commitment to maintain it.

In 1983, the United States government and the AMA entered into a formal agreement where CPT was adopted by HCFA for the reporting of physician services under the Medicare program. After this agreement was finalized, many private insurers and most other government programs also began to convert their systems exclusively to CPT. In this same year, CPT was adopted as part of the Health Care Financing Administration's (HCFA) Common Procedure Coding System (HCPCS) which is actually comprised of three volumes. Thus, by the late-1980s, CPT had in fact become the single uniform coding system for reporting of physician services.

Today it is estimated that over 95 percent of services provided by physicians are reported using the CPT coding system. Accordingly, CPT serves a wide variety of important functions. In addition to payment purposes, CPT is used widely for administrative management and the development of guidelines for medical care review. CPT is likewise applicable to medical education and clinical research by providing a useful basis for local, regional, and national utilization comparisons. Since 1983, the AMA has maintained the CPT system under its agreement with the federal government and has provided annual updates to HCFA and its agents at zero cost.


The AMA has developed a comprehensive structure and made a substantial financial and organizational investment and commitment to maintain and update CPT over the past thirty years. This has been a process that has been designed and operated to meet the needs of the diverse parties that rely upon CPT--physicians and their organizations, other health care providers, public and private payers, and others who rely upon accurate data on the services that physicians provide.

There is a continuous process centered around the maintenance of CPT by the CPT Editorial Panel, which is responsible for maintaining CPT. This Panel has recently been expanded and, effective in May of this year, will include sixteen physicians (including one non-MD provider), twelve nominated by the AMA and one each nominated by the Blue Cross and Blue Shield Association (BC/BS), the Health Insurance Association of America (HIAA), the Health Care Financing Administration (HCFA), and the American Hospital Association (AHA). The Panel's Executive Committee includes the Chairman, the Vice Chairman and three other members of the Panel, as elected by the entire Panel. Of the three Executive Committee members-at-large, one must be a payer representative. The most recent expansion of the Editorial Panel reflects the provision of full-voting privileges to a non-MD health professional and the addition of a physician who has specialized knowledge of the coding needs of the managed care community.

Supporting the Editorial Panel in its work are the CPT Advisory Committees. These Committees are comprised of physicians and other health professionals, representing those national medical specialty societies in the AMA House of Delegates and national organizations representing chiropractic, nursing occupational therapy, optometry, podiatry, physical therapy, physician assistants, psychology, social work and speech pathology/audiology. The Advisory Committees, which now number over 100 individuals, meet annually to discuss items of mutual concern and to keep abreast of current issues in coding and payment. Typically, members of the Advisory Committee chair their own specialty society coding committees, and it is thus estimated that CPT is supported by a network of nearly 1000 practicing physicians and other health professionals.

Specific formal procedures exist for addressing suggestions to revise CPT. These procedures, which have been developed in concert with the all the participants in the editorial and maintenance process, are clearly articulated in The CPT Process brochure which accompanies this statement. The CPT Editorial meets four times per year to consider a variety of coding issues. In a typical year, the Panel addresses 200 major topics, which normally involve more than 2,500 votes on individual items.


A system of modifiers is also an integral part of the CPT coding system because it allows for greater flexibility in the coding process. These two-digit numeric modifiers provide a method to:

The modifier system has proven extremely effective in providing needed flexibility and "granularity" to the coding system without having to add a significant number of new codes and descriptors. While the third-party representatives to the CPT process indicate that their members have the ability to handle the two-digit modifiers, we continue to hear anecdotal evidence that such modifiers are not routinely considered in the claims payment and review processes. This is an issue that is of great concern to us, and one that we hope the Committee will address as its recommendations are developed. The only real alternative to uniform acceptance of modifiers, is an explosion in the number of CPT codes, a phenomenon that the Editorial Panel has been working very hard to avoid.


One of the most important events surrounding the more recent development of CPT was the decision by Congress to restructure the process by which physicians are paid for their services under the Medicare program. The transition to the Medicare Resource-Based Relative Value Scale (RBRVS)-based physician payment system began on January 1, 1992, culminating nearly a decade of effort by the medical profession and the government. An RVS is simply a list of physician services ranked according to value, which when multiplied by a monetary conversion factor becomes a payment schedule.

The enactment of the RBRVS legislation brought tremendous attention and critical review to CPT. Use of an RVS as a basis for determining payments was a familiar concept for physicians and insurers. Once the Secretary of HHS selected CPT as the coding system that would underlie the RBRVS when implemented, many specialty societies realized that the existence of codes and the manner in which procedures were segmented might have a profound influence on the eventual flow of Medicare dollars to their members. CPT had naturally and appropriately developed to complement charge-based type systems (like customary, prevailing and reasonable charges, known as CPR). In many instances it was alleged that the structure of the codes was not sufficient to adequately, fairly, and uniformly reflect the "resource" costs of providing services.

This "gap" in CPT was believed to be particularly acute in what were then referred to as the CPT codes for visits and consultations. Data indicated that these codes were used in significantly different ways by different physician specialties and across geographic areas. This disparity would not be acceptable in an RBRVS system which, by and large, prohibited Medicare from making differential payments for the "same service" across specialty or geographic area. Thus in 1989, the AMA's CPT Editorial Panel began revising the CPT coding system for visits and consultations.

The CPT Editorial Panel developed new codes for office visits, hospital visits, and consultations, taking into account recommendations of the panel's own Ad Hoc Committee on Visits and Levels of Service, a special AMA/Physician Payment Review Commission Consensus Panel, and research from the Harvard study from which the development of RBRVS is based. Issues that the panel considered included the appropriateness of using time in visit coding, the number of levels of service, the need for different codes for different sites of service, and the need for different codes for new and established patients.

The 1992 CPT codes for evaluation and management services differ fundamentally from the previous version in the way they define and categorize codes. The familiar levels of service such as brief, minimal, and intermediate were replaced by a more precise method of assigning codes based primarily on the extent of history, examination, and on the complexity of medical decision-making. Other factors that may affect the level of service were identified, including: counseling, coordination of care, severity of presenting, and face-to-face time taken to perform the service.

The development of these new codes is an excellent example of the capacity and willingness of the CPT process to make fundamental changes in the coding system, responding to new health care information system needs, in a careful and tested manner. It also provides useful insights on the magnitude of the educational and other required implementation steps that would be required by any major departure from CPT for reporting health care professionals services.

In addition, operating somewhat in tandem fashion to the CPT Editorial Panel, the AMA, working with the national medical specialty societies, established in 1991, the AMA/Specialty Society RVS Update Committee (RUC) to provide annual recommendations to HCFA on assignment of RBRVS values for new and revised CPT codes. This process has been a great success. The RUC has submitted nearly 2,000 relative value recommendations for new and revised codes since 1993 and HCFA's acceptance rate for these recommendations has increased to more than 90% annually. During the recently completed five-year Review of the RBRVS the RUC examined 1,118 codes and achieved an acceptance rate of 93%. It is important to note that the AMA, along with participating specialty societies and organizations representing non-MD/DO health care professions, supports this highly resource intensive process at no cost to the federal government, other payors, providers, or vendors who rely on its operations.

In addition to the important changes in E/M services, the CPT Editorial Panel has made in excess of 1000 modifications to CPT since 1989 (including those necessitated by the Congressional mandated five-year review of the RBRVS) so that the coding system may truly reflect the resource costs of the particular service being described. Over the past few years, CPT has been increasingly oriented to support the resource cost distinctions of the Medicare fee schedule.


With this as background, we would also like to take the opportunity to respond to some specific questions or concerns that have been raised during recent deliberations. First, on the critical question of professional acceptance of code sets, the AMA investigated this as part of its annual survey of physician opinions. Preliminary results, based on telephone interviews of 779 randomly selected physicians involved in direct patient care, conducted by Gordon S. Black, Inc. in January/February of this year, 95 percent of physicians indicated that they think it important that the codes used to describe the services physicians provide to patients are developed and maintained by the medical profession -- 78 percent believing it is very important. Fewer than one-third of the physicians surveyed thought that private organizations that develop health care information standards (32 percent) or the federal government (22 percent) should develop or maintain the codes.

These results very clearly indicate the tremendous importance that physicians place on being able to describe the services they provide in terms that they believe are clinically meaningful and underscore our belief that compliance with coding rules and conventions is very highly correlated with the source of the code sets. The results of this survey reinforce our long-held view that the medical profession must maintain its code sets and that the AMA, given its structure and resources, in the body best suited to handle that responsibility. We can simply not envision any other process or body that could maintain a professional services coding system with such a high degree of responsiveness to diverse needs and timely and low-cost distribution of annual revisions.

Next, on the question of a single, uniform coding system for health professional services, the AMA's position remains as we have previously testified. Ideally, there should be one system that is used uniformly by health professionals, in all sites of service. Such a system would allow for true administrative efficiencies, would reduce burdens on those who are currently mandated to use multiple systems, and would help facilitate the creation of data bases to effectively analyze differences in treatment patterns and outcomes.

At the same time, however, we believe that the costs of creating a new coding system to accomplish this would be extraordinarily prohibitive and disruptive. We have separately provided to Committee staff independent studies done by Coopers and Lybrand which demonstrate the enormous costs to the health care system of moving to a new coding structure. That September 1989 study suggested that the costs of a newly developed uniform coding system could approximate $700 million. Today's costs would, obviously be substantially greater, for a variety of reasons including: the implementation of the RBRVS, growth of computerized billing and record-keeping systems, usage of CPT by non-physicians and institutions, etc. Recent estimates from the study's authors conservatively indicate those amounts, in 1997 dollars, as between $866 million and $993 million (Argus/Arista 1997).

AMA believes that with appropriate modification, CPT can easily and readily (i.e., within the time frames specified by the legislation) become the single, uniform coding system that so many have advocated. On numerous occasions, AMA has indicated its willingness to work with all interested parties to make that happen and in light of the requirements of the HIPAA legislation, and particularly because of its emphasis on administrative transactions, it is very clear that the time to move forward on that is now. When one goes beyond the question of code sets for administrative and financial transactions, as specified in the legislation, the AMA is less certain that a "single uniform coding system," which meets everyone's purposes, for every application, is truly attainable.

Third, on the issues of efficient and low-cost distribution, AMA places a great deal of emphasis on ensuring that CPT is made available in a variety of formats for an ever-growing variety of users. In print form CPT is now available in soft-bound, spiral-bound, and a new three-ring bound professional edition. Electronically, the data are available on magnetic tape, diskette in a variety of common software formats. A CD-ROM version has been made available for 1997. In addition, CPT is currently licensed to many software vendors and publishing companies.

CPT is available at low cost through the AMA or through the AMA's licensing activities. CPT can be purchased from the AMA in print format for as low as $38 and can be purchased from the AMA in electronic formats for as low as $149. Licenses for CPT in print products are granted for as low as $2 per product. Licenses for CPT in electronic products are granted for $50 per products license with a minimal additional user fee for multi-user versions. Of the 615 licenses that are currently in effect, over one-third of those are completely royalty-free.

The AMA has taken additional steps to make CPT available over the Internet and is expected to complete an agreement with the HCFA in the very near future. Under the agreement, complete public access to HCFA data files containing CPT will be available, free of charge, both domestically and internationally. We are confident that this is the first step toward making CPT generally available over the Internet, although as this may be tied to commercial applications in some cases, it can be expected that some nominal charges may be associated with it. Our recent agreement with the National Library of Medicine (NLM) to make CPT available as part of the UMLS, again free-of-charge to researchers, is another important step that has been taken in this regard.

Fourth, on the issue of proprietary versus public-domain systems, the AMA feels strongly that it is in the public's interest for the Secretary to advocate the "best coding systems to do the job" regardless of whether such systems are public-domain or not. There is nothing inherent in the proprietary status of a code set that makes it more or less useful for the purposes specified by the legislation. The legislation is very straightforward in providing the Secretary the flexibility to choose both proprietary and public domain systems. It is clear that for maintenance purposes, and because of concerns over continuity, accuracy, professional acceptance, and so on, that it is entirely appropriate for some systems to be copyrighted while others may not be. The recent decision by the World Health Organization (WHO) to copyright ICD-10, even though it is to become an international standard for reporting of morbidity and mortality statistics, is an excellent example of this need. The Committee should be careful to clearly separate any issues pertaining to public domain, from the issues of openness of input, and distribution. The costs of obtaining CPT publications is not materially different from obtaining publications of public domain code sets, such as ICD-9-CM made available from a variety of commercial publishers.

Fifth, on the issue of diagnosis code sets, the AMA supports the continued usage of ICD-9-CM. To date, there has been insufficient involvement of the medical profession in issues pertaining to ICD-10, so that it is not now possible to draw solid clinical conclusions about its practical superiority to ICD-9-CM. Accordingly, discussions of the need for the development of an ICD-10-CM are also premature. Having said that, the AMA is more than willing to participate in meaningful educational programs about the philosophy and structure underpinning ICD-10 and, pending the outcomes of those reviews, we see no reason why that could not be done on a timetable to meet the implementation dates specified by the legislation. The network of physician coding experts that we have assembled through the CPT Advisory Committee referred to earlier would be an excellent way to pursue that review in a timely and serious manner and the AMA would be more than willing to help facilitate that.

Sixth, with regard to the development of the ICD-10-PCS, AMA has followed its development closely and has appreciated the opportunity to participate in the Technical Advisory Panel established by HCFA for this project. We think that the ICD-10-PCS project has made valuable contributions to many issues relating to coding and terminology and that those contributions may, and ought to, find numerous applications. At the same time, our preliminary review of the system, along with that of many medical specialty societies, has identified a number of problems with the system that, in our view, would pose great limits on its acceptability by practicing physicians--not the least of which is the introduction of an overwhelming number of new codes (perhaps up to 150,000) being introduced into the health care system.

With respect to its implementation for administrative transactions, our views here are closely tied to our position stated above on the single-uniform coding systems. If HCFA's views that ICD-9-CM Volume 3, is hopelessly flawed and out of date, then it seems we face an historic opportunity to replace that volume with a system that will provide true uniformity across care setting. We see no particular merit in spending considerable sums of money to implement a new system that, even if it is successful, will only perpetuate a dual reporting system that medical professionals find awkward and wasteful.

If a decision is reached to moved forward on the ICD-10-PCS we believe that rigorous field testing is essential. Any field test should not simply evaluate whether the ICD-10-PCS "works" but the degree to which this system presents a real, quantifiable improvement over the existing systems like ICD-9-CM Volume 3, and CPT. Evaluation should include carefully designed experiments to test the degree to which this new system would improve inter-or intra code reliability relative to other systems that could be put in place at significantly lower system costs. AMA would oppose any decision by the Secretary to move forward with ICD-10-PCS in the absence of such extensive field testing and without a comprehensive evaluation of the other options that are available.

Lastly, with respect to additional administrative efficiencies that might be achieved within the context of the HIPAA, there are two important ones that we would like to address at this time, within the framework of CPT. One, as previously mentioned, is the usage of modifiers. For CPT to work correctly, and for the all available efficiencies to be manifest, it is essential that all health insurance companies accept modifiers as part of their claims processing systems. A second is uniform adherence to the CPT rules.

We believe that all insurance companies should be encouraged to follow the notes, guidelines and instructions that accompany the codes and descriptors. They should also be required to use the most recent issue of CPT. Without such adherence physicians are faced with multiple sets of coding rules, that often distort the intended meaning of the codes and, in effect, may require physicians to submit codes that are less than fully descriptive of the services actually provided. The AMA is willing to work with all interested parties, to add to or revise the guidelines as appropriate, to ensure that they are clear, unambiguous and can be uniformly administered.


While the fourth edition of CPT has had a very successful life, the AMA has already begun to explore the longer-term needs of procedural coding systems and is now addressing the question of whether, or how long, CPT in its current form/structure can continue to meet the growing demands being placed on it. In this sense the impetus of the HIPAA legislation and the questions that it has generated pertaining to the standards for code sets in the arena of administrative transactions provides an extremely useful template for those discussions and planning efforts.

Accordingly, AMA has appointed an Exploratory Committee to begin investigation of issues such as whether there is need to:

! make additional changes in CPT to address payer concerns about potential abuse of the coding system;

! make additional changes in CPT to accommodate the needs of non-physician users of the system (professional and institutional);

! whether, and how, managed care organizations make use of CPT and to consider changes in CPT to accommodate their special needs;

! make changes in CPT as a result of movement toward a computerized medical record;

! make changes in CPT as a result of the increased interest in developing and using practice guidelines, "report cards," uniform data sets, and other quality measurement devices; and

! modify the basic structure of CPT by, for example, gravitating to a coding structure that is more obviously hierarchical in nature.

By virtue of its work thus far, the Exploratory Committee has been made more keenly aware of some of the issues that are of concern to some in the managed care industry regarding the need to improve CPT, and has begun to consider these factors. Although most managed care plans have indicated that they use CPT quite well, there have been a number of concerns that will require further study and evaluation. For example, plans have found it useful to create their own codes to describe certain services such as those provided by telephone consultations, various types of preventive and counseling services, and case management.

In addition, certain plans have suggested that some CPT codes do not provide for a sufficient amount of clinical detail. For example, some have stated that the "with or without" language for some codes makes it more difficult for them to determine exactly what services have been provided to an individual patient. Again, most of these issues are not new and it is too early to tell at this stage how pervasive these points are since most evidence is anecdotal. We do, however, take the issues very seriously and are committed to making whatever changes in CPT necessary to meet the ever-changing needs of the medical community. In particular, we believe that CPT can provide that level of "granularity" that meets the diverse needs of its various users, including clinicians, payers, managed care plans, and researchers.


In conclusion, the AMA believes that the strengths of CPT are considerable and have been proven effective over many years. The CPT system has clearly and repeatedly demonstrated its flexibility and ability to adopt to changes in the health care environment and, as always, we are open to additional modifications that are deemed necessary and appropriate. The AMA has made great strides in assuring that the code sets are available to all interested parties, in an efficient, low cost, and fair manner. The costs of implementing any system other than CPT would be enormous and we AMA believe that any attempt to move away from CPT would be unnecessary, confusing, fiscally irresponsible, and would pose great harm to data collection activities for both federal and private insurance entities.

Importantly, CPT remains the clear choice and preference of practicing physicians. Thus, the AMA believes, that CPT is clearly the only coding system that is consistent with the HIPAA code set requirements for specification as the coding system for reporting physician and similar professional services.

Again, thank you for this opportunity to discuss CPT. I would be happy to respond to any questions you may have.