Physician Fee Schedule

September 15, 2000

Administrator Nancy Ann Min DeParle
Health Care Financing Administration
Department of Health and Human Services
Hubert H. Humphry Building, Room 443-G
Attn: HCFA-1120-P
200 Independence Avenue, S.W.
Washington, D.C. 20201

Re: HCFA-1120-P

Dear Administrator Min DeParle:

The Society of Vascular Technology ("SVT"), the Society of Diagnostic Medical Sonographers ("SDMS"), and the American Society of Neuroimaging ("ASN") appreciate this opportunity to comment on the Notice of Proposed Rulemaking ("NPRM") regarding "Revisions to Payment Policies Under the Physician Fee Schedule for Calendar Year 2001" published by the Health Care Financing Administration ("HCFA") on July 17, 2000. 65 Fed. Reg. 44176.

SVT is a professional society comprising over 3,500 members who consist of registered vascular technologists, nurses, and some 1,000 physicians. The members of SVT provide high-quality vascular ultrasound services.

SDMS is a not-for-profit professional organization which represents multispeciality, cardiac sonographers and vascular technologists. Our 11,000 members are professionals involved in the delivery of high-quality echocardiography, ophthalmology, neurosonography, obstetrical and abdominal ultrasound, and vascular ultrasound services.

ASN is an international, professional organization representing 820 neurologists,
neurosurgeons, neuroradiologists, and other neuroscientists who are dedicated to the advancement of techniques used to image the nervous system. Its purpose is to promote the highest standards of neuroimaging in clinical practice, thereby furthering ongoing improvement in the delivery of medical care.

Ultrasound imaging provides a means of looking into the body of a patient and examining structures, such as organs, vessels and the fetus. By using sound waves to obtain images of the interior of the body, vascular technologists and sonographers provide a critical, low-cost diagnostic service to physicians and other clinicians that can, in many circumstances, eliminate the need for much more costly and invasive diagnostic studies.

I. Technical Component Only Services and the Resource-Based PERVU System

SDMS, SVT, and ASN understand the difficulty entailed with implementing the resource-based practice expense relative value unit system ("PERVU") given the complexity of the Medicare program and the lack of comprehensive data upon which to base the system. Nevertheless, the reductions in PERVUs for technical component-only ("TC") services proposed in the NPRM for the 1999 physician fee schedule illustrated fundamental problems with the PERVU methodology as it was applied to TC services and providers. 63 Fed. Reg. 30818 (June 5, 1998). In 1998, when the resource-based PERVU implementation was proposed initially, both SDMS and SVT commented to HCFA that two of the most critical defects in the methodology design as it relates to TC services include (1) the paucity of TC services data captured by the American Medical Association Socioeconomic Monitoring Survey ("SMS") which was used to establish practice expense and (2) that the methodology necessitates a service have a physician work component in order to assign a PERVU. / We further explained, that the CPEP data does not reflect accurately equipment costs nor the salary and benefits costs for sonographers and vascular technologists. In fact, the salary and benefits information is so dated as to be obsolete.

We asserted then, and still hold today, that the combination of these deficiencies lead to the drastic and unsupported cuts in reimbursement originally proposed for TC services procedures, / and therefore, the PERVU methodology cannot be applied to TC services.

I. The "Zero-Work" Pool: An Interim Fix

Recognizing the substantial and inappropriate reductions in PERVUs for zero-work or technical component-only services, and the possibility they might be related to flaws in the SMS and CPEP data as well as a biased methodology for allocating practice expense indirect costs, HCFA created, as an interim measure, a separate practice expense pool for services without work RVUs regardless of the specialty that provides them (the "zero-work" pool). We are most grateful for HCFA's careful response to our concerns in this regard. Its willingness to respond to our comments is yet another example of the discipline and professional process that HCFA follows. For purposes of the zero-work pool, HCFA used the "all physicians" category for practice expense per hour and allocated the pool using existing 1998 charge-based RVUs instead of the CPEP data.

It our understanding that a permanent solution has not been decided upon, and that HCFA has directed The Lewin Group to provide the Agency with recommendations on how to set the practice expense of services without physician work. It is because this issue is to be considered in the upcoming months, that we submit these comments and reassert our position that the unique concerns related to TC services dictate that (1) practice expense data must be collected from TC-only providers to capture and quantify accurately the resources used to deliver services without a professional component, and (2) a separate practice expense methodology must be used to allocate payments equitably.

SVT, SDMS, and ASN believe that the zero-work pool concept should continue to be used until alternative methodology based on sufficient hard data is developed. Significantly, HCFA received few complaints regarding the zero-work pool, suggesting most technical component groups supported this solution. A minority of speciality groups asked that an alternative to the zero-work pool mechanism be used for selected services on the grounds that a small number of technical component services specific to their medical specialties were disadvantaged by the zero-work pool, 64 Fed. Reg. 59380, 59406 (Nov. 2, 1999), but HCFA has addressed these isolated issues since other specialty groups opted to remain in the pool.

SDMS, SVT, and ASN would disagree emphatically with any decision to default to the standard PERVU methodology used for services with physician work because, without doubt, TC services would once again be faced with drastic and unjustified underpayments. Moreover, TC-only providers simply will be unable to sustain the tremendous financial losses they will incur when providing ultrasound services to Medicare beneficiaries which will have grave implications for access and the quality of ultrasound services.

II. Background of the Problem

A. Determining Practice Expense Related to Technical Component Services

To reinforce the foundation of our position we offer the following brief background. Like the vast majority of providers, SDMS, SVT, and ASN are very concerned about the general accuracy and reliability of the ("SMS") data and has submitted comments to HCFA to this effect. We understand that through work with the Lewin Group and the willingness of some specialty provider groups to bear the financial burden of documenting their practice expenses through independent surveys, some of the statistical concerns with the SMS data are being addressed. These efforts will not change, however, the fact that the data does not, in any meaningful way, capture the practice expense for TC services.

The SMS data itself demonstrates its lack of information regarding technical component ultrasound services, and the need to collect practice expense data specific to this issue. For example, cardiac-thoracic-vascular physicians represent a significant source of ultrasound services. Therefore, it would be expected that if, the practice expenses related to ultrasound were captured through questions about physicians' office expenses, the data from these specialities would reflect the technical component costs. Yet, this was not the case.

According to the CPEP data a single ultrasound machine costs something on the order of $250,000. / Over 74.8 percent of the small number of cardiac-thoracic-vascular surgeons that responded to the SMS survey, however, reported equipment expense costs of less $5,000. Indeed, almost 90 percent of the respondents reported equipment costs of less than $25,000 emphatically, suggesting that the SMS sample was skewed toward professional-component only providers.

The same appears to be true in connection with a number of other specialties that provide significant percentages of the technical component of ultrasound services. For example, almost eighty-eight (88) percent of OB-GYN SMS respondents reported equipment costs of less than $25,000, with almost fifty (50) percent listing equipment costs of less than $5,000. The results were essentially the same in radiology, where eighty-six (86) percent of all respondents reported equipment costs of less than $25,000 and over seventy (70) percent of all respondents listed equipment costs of less than $5,000, yet equipment costs likely exceed $250,000.

Even if a reasonable percentage of technical component providers were surveyed in the SMS data, the effect of the SMS methodology would be to dilute higher technical component costs with lower professional component costs. Accordingly, we believe that the only way to accurately determine practice expense costs for technical component services is to perform a practice expense survey of TC-only providers and use that survey, with appropriate adjustments, to fix the practice expense for technical component services.

B. Technologist Time Versus Physician Time

Furthermore, rather than expressing the practice expenses in terms of those expenses per physician hour, we believe that the integrity of the practice expense calculation can only be maintained if it is expressed in terms of sonographer/technologist hours. Physicians typically do not play any significant role in the actual provision of the technical component of these services.

To illustrate our concerns regarding payment under the original method for determining the PERVUs for technical component ultrasound services, we provided HCFA in the past with data from two published studies undertaken by well-known and well-regarded vascular practitioners. J. Dennis Baker, M.D. of the University of California at Los Angeles completed the first study ("the Baker study") in 1992. Mark F. Fillinger, M.D., Robert M. Zwolak, M.D., Ph.D., Anne M. Musson, B.S., R.V.T., / and Jack L. Cronenwett, M.D. of the Dartmouth-Hitchcock Medical Center completed the second study ("the Zwolak study") in 1993. /

The Baker study involved the use of data from a simple and well-designed expense survey that was completed by 140 different providers of vascular TC services located across the country and in various settings, including teaching hospitals, non-teaching hospitals, clinics, physician offices, and independent physiological laboratories. Significantly, the 142 respondents is more than three (3) times the number of cardiac-thoracic-vascular surgeon respondents in the SMS data and more than fourteen (14) times the number of vascular surgeon respondents. The survey was mailed to potential respondents, a mechanism that should have provided more careful and better-researched answers than those provided in the SMS "telephone only" survey. In order to ensure the accuracy of the data obtained, respondents were contacted by telephone to confirm their written answers. The mean costs per vascular study reported were $181.

To obtain a figure for expenses per technologist/sonographer hour from this data, we took the CPEP estimates for each of the vascular ultrasound imaging codes to secure an average study length. This led us to calculate an expense per technologist/sonographer hour for vascular ultrasound services of $120.50.

We note that the Zwolak study confirmed these costs. That study, which was based on a review of the Dartmouth-Hitchcock laboratory's costs and augmented by extrapolations to national numbers, found national technical component costs in the range of $143 to $173. Based on this data, we urged HCFA, prior to the establishment of the zero-work pool, to recalculate, on an interim basis, the vascular technical component PERVUs using a figure of at least $120 an hour in expenses, updated to reflect inflation for the years since the study was undertaken. / We believed then, and still do today, that this was an extremely conservative figure to choose, as the Zwolak study actually demonstrated higher costs. If the zero-work pool concept is abandoned, the findings of these studies should be kept in mind. Indeed, the practice expenses associated with providing the various ultrasound speciality services should be validated.

C. Biased Allocation Methodology

In addition to the difficulties with the SMS data, SDMS also had concerns with the negative bias against technical component reimbursement reflected in the mechanism used to allocate indirect costs. As originally proposed, indirect costs were allocated based on direct costs and physician RVUs. Because there are no physician RVUs in technical component services, this approach necessarily led to a decrease in the recognition of the indirect costs associated with technical component services. This effect is particularly inappropriate because space and utility costs, two of the prime indirect costs, should be higher for technical component services than for professional component services.

D. Independent Physiological Laboratory Costs

Another obvious problem in the calculation of technical component services in the original PERVU proposal was the treatment of independent physiological laboratory for practice expenses. If we understand the data correctly, the CPEP equipment expenses for IPLs was scaled from $197,000,000 to just $6,000,000, or a mere 3.04 percent. Supply costs of $105,000,000 were scaled to just 13,000,000, or a mere 12.4 percent. Clinical costs were scaled from $74,000,000 to $21,000,000, or 28.0 percent. The weighted average scaling for IPLs was 10.6 percent, far below the scaling used for the average specialty.

The use of scaling also appears to be fundamentally inappropriate in connection with IPLs now operated under the IDTF classification. As we understand it, the scaling factors were designed to bring the CPEP data into accord with the SMS data. In the case of the SMS data, however, we are unaware of any data that was collected for IPLs. Our understanding is that the SMS survey was developed, designed, and used solely and exclusively for physician practices.

In the absence of any IPL data from the SMS survey, we urged HCFA to use the Baker and Zwolak studies to provide expenses per hour for the ultrasound clusters generally. The Baker study demonstrates that IPL costs per study were $256. Using the same average length of a vascular ultrasound examination figure discussion above, we converted the Baker study figure to an hourly expense of $176.40.

Furthermore, the ratio of IPL costs to other technical component provider costs established by the Baker study was 1.46 percent ($265 (IPL cost)/$181 (mean cost)). As such, we advocated that HCFA set the IPL costs in its calculations by taking the practice expense for the physician specialty providing the most of a particular type of ultrasound service and multiplying that expense figure by the conversion factor of 1.46 to secure the IPL expense figure for that service. In echocardiography, for instance, where cardiology tends to be the dominate provider, HCFA should take its cardiology expense per physician hour of $84 and multiply it by 1.46 to secure an IPL expense per hour of $122.

III. Deletions from CPEP Data

SVT, SDMS, and ASN also urge HCFA to remove Doppler from the list of items it intends to delete from the CPEP data equipment file. Doppler and color Doppler are essential pieces of equipment for ultrasound procedures, and not a standby item. The introduction to the non-invasive vascular diagnostic studies section of the CPT Manual itself defines "duplex scan" which encompasses most ultrasound procedures as "an ultrasonic scanning procedure with display of both two-dimensional structure and motion with time and Doppler ultra-sonic signal documentation with spectral analysis and/or color flow velocity mapping or imaging."

Thus, the expectation is that the Medicare beneficiary undergoing a "duplex" ultrasound procedure will undergo a study involving three components: (1) an ultrasound picture of the structure of the relevant arteries or veins (i.e., the "two-dimensional imaging"), (2) data regarding the rate of the blood flow (i.e., the "Doppler signal documentation"), and (3) "color flow velocity mapping," which consists of a pictorial representation, in color, of the rate and the direction of the blood flow. If HCFA removes Doppler from the CPEP, the result will be to require TC providers to finance Doppler equipment for the benefit of Medicare beneficiaries without any reimbursement. This is not fair and inconsistent with this services (1) satisfaction of statutory coverage criteria and (2) the congressional mandate to pay for services based on their practice expense.

IV. Crosswalk

Lastly, HCFA requested comment as to the appropriateness of the crosswalk of a transvaginal ultrasound to a pelvic ultrasound. We agree that this crosswalk is reasonable.

* * *

Thank you, once again, for the opportunity to provide these comments regarding the problems with the resource-based PERVU system as it pertains to technical component services. As we have offered in the past, SDMS, SVT, and ASN are willing to provide any type of additional information or support to HCFA or the Lewin Group as you review this issue. Thank you once again, for your careful consideration of these issues.

Sincerely,

Patricia Marques, RN, RVT, FSVU
President SVT

Stephen McLaughlin, BS, RT,RDMS
President SDMS

Charles Tegeler, M.D.
President ASN