Policy Implications of the ComputedTomography (CT) Scanner: An Update

January 1981

NTIS order #PB81-163917

‘Library of Congress Catalog Card Number 81-600006

For sale by the Superintendent of Documents,U.S. Government Printing Office Washington, D.C. 20402

Foreword

In August 1978, OTA published a report called Policy Implications of the Com-puted Tomography (CT) Scanner. The report has generated much interest in both theCT scanner itself and Federal policies aimed at rationalizing its diffusion and use. Dur-ing the summer of 1979, OTA staff reviewed the status of policies concerning the scan-ner and found that they were very much in the process of change. A number of staff ofcongressional committees (especially the Senate Committee on Finance, which had re-quested the original study) expressed interest in an update of material in the originalreport. In addition, OTA continued to be consulted by outside organizations andgroups, including Federal and State government agencies, on implications of CT scan-ners, For these reasons, OTA decided to update the 1978 report.

This paper does not repeat material in the original report. For the most part, it isassumed that the reader is familiar with the 1978 report. Thus, basic descriptions ofthe scanner and of certain Federal laws and policies are found in the earlier report.

As part of this paper, OTA has updated its list of operational CT scanners. TheOTA list apparently continues to be the most complete and reliable inventory of scan-ners. An analysis of data from the list is presented in chapter 2.

This paper considers the CT scanner in the context of the entire field of what hascome to be called “diagnostic imaging” (making pictures of the inside of the humanbody for the purposes of diagnosis). A number of new applications of existing tech-nologies as well as several new technologies have developed. The field of diagnosticimaging is developing rapidly, and presents a significant challenge to policy makers.Can the relative advantages of the different technologies be demonstrated? CanFederal policies rationalize the use of the many technologies? Or will the newtechnologies merely be added on to the existing methods, driving up costs and con-tributing only a small marginal benefit to people’s health?

Since this is a background paper, no policy options are presented. The purpose ofthis paper is to summarize the most important development concerning CT scannersthat have occurred over the past 21/2 years. However, since the policy options of the1978 report seem generally valid, they are reprinted in appendix A.

Drafts of the final paper were reviewed by the Health Program Advisory Com-mittee chaired by Dr. Sidney Lee and by a number of other individuals and groups rep-resenting manufacturers, radiologists, and Federal agencies (see app. I). We aregrateful for their assistance.

JOHN H. GIBBONSDirector

111

Health Program Advisory Committee

Sidney S. Lee, ChairmanAssociate Dean, Community Medicine, McGill University

Stuart H. AltmanDeanFlorence Heller SchoolBrandeis University

Robert M. BallSenior Scholarinstitute of MedicineNational Academy of Sciences

Lewis H. ButlerHealth Policy ProgramSchool of MedicineUniversity of California, San Francisco

Kurt DeuschleProfessor of Community MedicineMount Sinai School of Medicine

Zita FearonConsumer Commission on the Accreditation

of Health Services, Inc.

Rashi FeinProfessor of the Economics of MedicineCenter for Community Health and

Medical CareHarvard Medical School

Melvin A. GlasserDirectorSocial Security DepartmentUnited Auto Workers

Mark LepperVice President for Inter-Institutional AffairsRush-Presbyterian Medical SchoolSt. Luke’s Medical Center

C. Frederick MostellerProfessor and ChairmanDepartment of BiostatisticsHarvard UniversitySchool of Public Health

Beverlee MyersDirectorDepartment of Health ServicesState of California

Mitchell RabkinGeneral DirectorBeth Israel Hospital

Frederick C. RobbinsPresidentinstitute of Medicine

Kerr L. WhiteRockefeller Foundation

Computed Tomography Scanner Project Staff

Joyce C. Lashof, Assistant Director-, OTAHealth and Life Sciences Division

H. David Banta, Health Program Manager

Karen Kohlhof, Co-Project DirectorH. David Banta, Co-Project Director

Chester Strobel, Research AssociateJohn Bell, Computer Prograrner

Kerry Britten Kemp, * Editor- and WriterVirginia Cwalina, Administrative Assistant

Shirley Ann Gayheart, SecretaryNancy L. Kenney, Secretary

Publishing Staff

John C. Holmes, Publishing Officer

John Bergling* Kathie S. BossDebra M. Datcher Patricia A. Dyson*

● Contract personnel

Contents

Chapter Page

1.

2.

3.

4.

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

NUMBER AND DISTRIBUTION OF CT SCANNERS . . . . . . . . . . . . . . . . . 9

Diffusion of Medical Technology—Some General Considerations. . . . . . . . . . . 10Distribution of Operational Scanners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Geographic Distribution. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12Institutional Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 15

Trends in the Type and Manufacture of Scanners . . . . . . . . . . . . . . . . . . . . . . . . 20

CHANGES IN FEDERAL POLICIES TOWARD CT SCANNERS . . . . . . . . . 27

Research and Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27Evaluation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .......27Regulation of Efficacy and Safety . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31Diffusion and Utilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..... 32

Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32Utilization . . . . . . . . . . . . . . . . .. .. ... ... ... .. 36

SUMMARY AND CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

APPENDIXES

A.

B.c.

D.E.F.

G.H.I.

Policy Alternatives (Reprinted From Policy Implications of the ComputedTomography Scanner, Office of Technology Assessment, August 1978). . . . . . 47

Research and Development of CT and Other Diagnostic Imaging Technologies 59CT-Related Policies of the Department of Defense and theVeterans Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72Indications for CT Scans. . . . . . . . . . . . . . . . . . . . . . ..........74National Guidelines for Health Planning: Standards for CT Scanners. . . . . . . . 92Amendments to Regulations Governing Reviews of CT Scanners Under 1122and CON Programs(April 25, 1979). . . . . . . . . . . . . . . . . . ... 94Methods of Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..98Glossary of Acronyms and Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. .. .. ... ... 105

LIST OF TABLESTable No. Page

1.

2.3.

4.

Medical Technology Development and Use: Formal Programs of theDepartment of Health and Human Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Distribution of CT Scanners by Region and State. . . . . . . . . . . . . . . . . . . . . . . . 12Ranking of States by CT Scanners per Million Population as of February 1979and May 1980 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 14Distribution of Installed CT Scanners in l0 Countries . . . . . . . . . . . . . . . . . . . . 15

Contents–-continued

Table No. Page

5. Number of CT Head and Body Scanners by Health Service Areas . . . . . . . . . . . 166. Distribution of CT Scanners by Type of Facility . . . . . . . . . . . . . . . . . . . . . . . . 187. Distribution of CT Scanners in Community Hospitals by Hospital Size . . . . . . . 188. Manufacturers of CT Head and Body Scanners in Use . . . . . . . . . . . . . . . . . . . . 20B-1. Overview of Diagnostic Imaging, United States . . . . . . . . . . . . . . . . . . . . . . . 59B-2. Sales of Diagnostic Imaging Equipment in the United States, by Year . . . . . . . 60B-3. Types and Models of CT Scanners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61B-4. Federal Funding for R&D in Ultrasonic Imaging Diagnostic

Instrumentation , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

LIST OF FIGURESFigure No. PageI. Cumulative Number of CT Scanners Installed . . . . . . . . . . . . . . . . . . . . . . . . . . 92. A Scheme for Development and Diffusion of Medical Technologies. . . . . . . . . . 103. Diffusion of Chemotherapy for Leukemia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114. Cumulative Number of CT Scanners Installed, by Manufacturer. . . . . . . . . . . . 21B-1. Principal Components of the DSR Imaging System . . . . . . . . . . . . . . . . . . . . 62B-2. Principal Components of Real-Time B-Mode Ultrasound Imaging System . . . 68

1.

lntroduction

1. ■

Introduction

This paper is an update of the OTA reportPolicy Implications of the Computed Tomog-raphy (CT) Scanner, published in August of1978 (129). The CT scanner remains an instruc-tive case study of Government involvement inthe policy areas of evaluation, regulation of dif-fusion and use, and financing of medical tech-nologies.

After 7 years’ use in the United States, the CTscanner has established itself as a revolutionarydiagnostic device (69,81,137). It has given phy-sicians a diagnostic capability that they pre-viously lacked. The development of this andother diagnostic technologies has made possiblethe definitive and conclusive diagnosis of someconditions. These technologies can sometimesguide physicians to appropriate treatments, pre-venting deaths and disability and relieving painand suffering. These basic activities are unques-tionably valuable. There seems to be little doubtthat CT scanning has been a remarkably usefuladdition to the array of medical technology.During the past few years, however, both theavailability of a wide variety of new diagnostictests and the strong incentives to use them haveenormously increased the use of these tests. Infact, there appears to be virtually no upper limiton the number and kind of diagnostic tests thata cautious and caring physician can order.Likewise, hospitals desire to acquire newtechnologies such as the CT scanner for a varie-ty of reasons, not least of which is to make theirprogram more effective in relieving human suf-fering and sometimes saving lives. In the case ofthe CT scanner, radiologists felt (and continueto feel) that the improvement in imaging, andthus in diagnosis, was so evident as to allowreasonable clinicians to accept the new instru-ment readily. For the radiologist, CT scanningwas easier, safer, and in many cases more re-liable than the X-ray procedures in use,

However, the CT scanner appeared in theUnited States at a time when the benefits, risks,

and costs of medical technology were of increas-ing concern. Because of this concern, CT scan-ning has been evaluated more than is usual.Thus, the CT scanner itself is not the problem.The problem is much broader and concernsappropriate use of medical technology in socie-ty. Perhaps a detailed examination of someaspects of policy toward CT scanning can in-dicate how far we are from having effectivepolicies to promote the efficient expenditure ofour health care dollar (142). In particular, OTAis concerned about regulatory approaches beingconsidered to control CT scanners and othertechnologies in the absence of definitive scien-tific information that will allow wise decision-making by Federal officials or insurance com-panies.

The purpose of the original OTA study on theCT scanner (129) was twofold. First, it was toexamine the usefulness and costs of CT scan-ning, the effect of CT scanners on medical caredelivery patterns, and ways to improve plan-ning affecting such devices. In the backgroundwas a concern about the implications of costlynew technologies such as CT scanners. The sec-ond purpose was to examine policies toward CTscanners. The 1978 study examined emergingand existing policies concerning the develop-ment, evaluation, diffusion, use, and financingof the CT scanner. It attempted to determine theeffects, both real and potential, of those poli-cies, and to identify problems experienced in im-plementing them.

Like the original report, this update (coveringthe period since August of 1978) documents thechanges in the number, distribution, and diffu-sion of CT scanners. It also summarizes changesin Federal policies, agencies, and programs dur-ing that time that affect research, development,and diffusion of CT scanners, and the evalua-tion and financing of CT scanning. Wherever

3

4 ● Policy Implictions of of the Computed Tomography (CT) Scanner: An Update

possible, it focuses on the relationship betweenchanges in policy and in the numbers and distri-bution. Although the 1980 CT scanner is quitedifferent from the 1973 edition, most of the dis-cussion treats all scanners as if they are the same(see app. B). Nevertheless, the continued de-velopment of technological improvements inCT scanners and the concomitant documenta-tion of new uses for scanners, has posed aserious problem for policymaking.

The dramatically rapid rate of diffusion ofscanners during 1975 and 1976 set the stage forOTA’s original study. An equally dramaticdrop in this rate during 1978, 1979, and 1980provides the backdrop for this update. During1977, the rate of installation of scanners wasabout 40 per month. (ch. 2 compares this rate ofdiffusion with that of other technologies. ) In1978, the rate fell by half to about 20 per month.

Whereas about 480 scanners were installed in1977, only 270 units were installed in “1978.These turnabout trends in the installation ofscanners are also reflected in the manufacture ofCT scanners. The consolidation of productionevident in 1978 is in sharp contrast to the ex-pansion that had occurred steadily since themid-1970’s (see ch. 2).

This diffusion pattern has occurred during aperiod of change in Federal policies towardmedical technology. With recent changes in Fed-eral law, the Federal Government is involved inevery stage of research, development, diffusion,and use of CT scanning (see table 1). TheGovernment has invested in R&D on CT scan-ning. But the Government also regulates CTscanners through the Food and Drug Adminis-tration, which approves medical devices formarketing. Since 1974, a nationwide network of

Table 1 .—Medical Technology Development and Use: Formal Programs ofthe Department of Health and Human Services

Policy area Stage of development Function Agency or programR&D . . . . . . . 1. Research and development

a. Basic researchb. Applied research

Evaluation. . 2. Demonstration of safety, efficacy,and cost effectiveness

a. Clinical trials

b. Assure efficacy and safety ofdrugs and devices

c. Provide economic analyses

d. Evaluate social, ethical, andpolitical impacts

Regulation . 3. Diffusion

Financing . . 4. Widespread use

● Support and planning of research Support and planning of research

● Test safety● Test efficacy● Protect human subjects● Control of testing procedures. Postmarketing surveillance● Cost-benefit analysis● Cost-effectiveness analysis● Technology assessment

● Premarket approval of drugs anddevices

● Encourage distribution byinformation dissemination

Control distribution through CON,review of purchase

● Assure appropriate use● Monitor practice. Reimbursement● Define benefits package● Set reimbursement levels

NIH, other smallNIH, other agencies andprograms

NIH, other small

FDA

NIH (limited)NCHCT, NCHSR

FDA

NIH (limited)

HRA

PSRO certification programsPSROs (limited)Medicare (elderly)Medicaid (poor)

CON = certificate of need. NCHSR = National Center for Health Services Research.FDA = Food and Drug Administration NIH = National Institutes of HealthHRA = Health Resources Administration. PSRO = Professional Standards Review Organization.NCHCT = National Center for Health Care Technology.

SOURCE: H. D Banta, 4’Publlc Policy and Medical Technology: Critical Issues Reconsidered,” presented at the conference ‘( POIICY Imnovatlon and the Service Sector,”Berlln, West Germany, June 13-16, 1978.

health planning agencies has had approval zation program has had the authority since 1972power over capital investments such as that re- to review medical services provided under thequired to purchase a CT scanner. The medicare medicare and medicaid program for medical ap-and medicaid programs pay for CT scanning. propriateness. The impact of these policies andAnd the Profession] Standards Review Organ i- programs is explored in succeeding chapters.

2 .

Number and Distributionof CT Scanners

2 ●

Number and Distribution of CT Scanners

By May 1, 1980, there were 1,471 operational end of 1976, the number of operational scannersscanners in the United States. This number of had increased to 475. Diffusion was even moreoperational scanners has been rapidly attained rapid in 1977, when about 40 scanners were in-(see figure 1). ’ At the end of 1974, only 45 scan- stalled per month. During 1978, however, theners were in operation. Two years later, at the rate of installation of scanners fell by nearly

half, In 1979 and the first 4 months of 1980, therate fell a little more, to about 17 scanners per

‘Based on data collected through April 1980. month.

Figure 1.— Cumulative Number of CT Scanners Installed (1973-80)

-

1,400 .

1,300 —

1,200 —

1,100 —

1,000 —

900 “

800 —

700 —

600 —

500 —

400 —

300 ‘

200 -

100

o1973 74 75 76 77 78 79 1980

Year

SOURCE Off Ice of Technology Assessment

DIFFUSION OF’ MEDICAL TECHNOLOGY–SOME GENERAL CONSIDERATIONS

The process by which a technology enters andbecomes part of the health care system is knownas diffusion. The diffusion of a technologyfollows the stage of R&D and may or may notoccur following careful clinical trials to dem-onstrate efficacy and safety. Descriptive re-search has shown that the diffusion process forany technology usually follows an S-shaped orsigmoid curve, relating the percentage of poten-tial adopters to actual adopters (see figure 2).Generally, there is an early phase of diffusionthat is somewhat slower. This has been inter-preted as indicating caution on the part of users(145), although it could also indicate problemsof communication of information about the in-novation (126). As experience indicates that the

technology does indeed have some benefit, ac-ceptance increases. Finally, when most potentialadopters have accepted the innovation, diffu-sion slows and the curve flattens. Althoughmost of the work demonstrating the S-shapeddiffusion curve is outside the health care area,this curve has been documented for such medi-cal technologies as intensive care units (1.46),cardiac pacemakers (126), respiratory therapy(146), diagnostic radioisotope facilities (1.46),and electroencephalographs (146).

The diffusion of medical technologies doesnot always follow the sigmoid curve. One majordeparture from this model occurs when diffu-sion reaches a high rate almost immediately

SOURCE Off Ice of Techology Assessment, U S Congress, Deve/oprnerr/ o/ Med/ca/ Techrro/ogY Opporfurr/f/es for Assessment (Washington, D C GovernmentPrlntlng Off Ice, August 1976)

after the technology becomes available (seefigure 3). This pattern has been referred to as the“desperation-reaction model” (182). A firstphase of rapid diffusion seems to occur becauseof the provider’s sense of responsibility to helpthe patient and their mutual desperation. Later,results of clinical tests and experience begin toinfluence the physician’s behavior. If results oftests are positive, diffusion may continue rapid-ly. If the evidence is not clear cut, there may becaution and slow diffusion. If the evidenceseems negative, use of the technology graduallydeclines.

Figure 3.— Diffusion of Chemotherapy for Leukemia

I I

SOURCE K Warner, “A ‘Desperation-React!on’ Model of Medical Dlffuslon,”Healfh Serv/ce Research 10369, 1975. Redrawn by the Off Ice ofTechnology Assessment

Whatever its initial pattern of diffusion, atechnology may eventually be partially or com-pletely abandoned. The rate of tonsillectomy(surgical removal of the tonsils), for example, is

presently declining (119). Such a decrease in usecan result from additional knowledge or the in-troduction of a more effective technology. Theintroduction of polio vaccine, for example,almost overnight entirely supplanted the costlyhalfway technology of rehabilitation centers(176).

Little work has been done on the diffusion ofspecific medical technologies, but some compar-isons can be made. Intensive care is an expen-sive technology that had its most rapid spread inU.S. hospitals from 1960 to 1968. The most rap-id diffusion rate was slightly over 200 per year,or less than 20 per month (146). Another tech-nology, nuclear medicine, spread at the rate ofalmost 200 facilities per year during the period1969 to 1972 (141), As noted above, the diffu-sion of CT scanners was considerably morerapid than the diffusion of either of these twotechnologies. The more rapid diffusion of CTscanners could be due in part to the change inreimbursement policies since the 1960’s.

Technologies have been observed to diffusemost rapidly among large hospitals (146). Earlydiffusion to hospitals affiliated with medicalschools was observed for intensive care (146)and nuclear medicine (141). Cromwell, et al.(39), however, found that when size and long-term debt were held constant, medical school af-filiation had little effect on equipment expendi-tures. These investigators also showed thattechnologies diffuse more rapidly as the percent-age of hospital resources from third parties in-creases. As seen below, except for its rapidity,the diffusion of CT scanners generally followsthe pattern predicted by previous research.

The sections that follow give detailed infor-mation on the diffusion and present distributionof CT scanners.

DISTRIBUTION OF OPERATIONAL SCANNERS2

Geographic Distribution

Table 2 summarizes informationtion of scanners by State. All Statesone scanner. There are no scannersSamoa, Guam, the Trust Territory

tion. States with high scanner-to-populationratios include Nevada (12.8), Florida (10.9),

on the loca- California (10.5), Missouri (9.4), North Dakotahave at least (9.1), Arizona (9.0), Nebraska (8.3), and Newin American Mexico (8.0). States with the lowest scanner-to-of the Mari- population ratios include South Carolina (2.4),

ana Islands, or the Virgin Islands. The national Rhode Island (3.3), Idaho (3.3), Delaware (3.4),average is now about 6.7 scanners per million Michigan (3.6), New Jersey (3,7), Kentuckypopulation. Washington, D. C., has the highest (3.7), and Montana (3.8). Puerto Rico has onlyratio, with 16.7 scanners per million popula- about 1.6 scanners per million population.

Table 3 shows that a ranking of States accordingto scanner-per-population ratios changed little

‘Numbers In this paper may dltfcr tr-om those ]n the 1978 O T Areport on CT w~nners becdu~ e they include ackiltl<~nal scanners with the addition of new scanners between Feb-[den t I t ICLi, replclcemen t+, d nci \LI on. ruary 1979 and May 1980.

Table 2.— Distribution of CT Scanners by Region and State (May 1980)

Region and State

New England. . . . . . .Maine. . . . . . . .New Hampshire ... ... ...Vermont. . . . . . . . . . . . . . . .Massachusetts. . . . . . . ... ., . . . .Rhode Island. ... ... . . ... .Connecticut . . . . . . . . . . . . . .

Middle Atlantic . . . . . . . . ... .,New York. ... . . . ... ., . . . . .,New Jersey . . . . . . . . . . . . . . . . . . . . .Pennsylvania, ... ... ... . . ...

East North Central . .Ohio . . . . . . . . . . . ... ... ...Indiana. . . . . . ., ., ... . . . . .Illinois . . . . . . . . ... ... . . . . . . .Michigan . . . . . . . . . . . . . . . . ., . . .Wisconsin . . . . ... ... ... . . .

West North Central . . . . . . . . . . .Minnesota. ... . . . . . .,Iowa ..... . . . . . . . . . . .Missouri . . . . . . . . . . . . . . . . . . . .North Dakota. . . . . . . . .South Dakota . . . . . . . . . . . . . .Nebraska . . . . . . . . .Kansas . . . . . . . . . . . . . . . . . . . . . .

South Atlantic ......., ... . . . . . . . . .Delaware . ., ...... . . . . . . . . . . .Maryland . . . . . . . . . . . . . . . . . . . . .District of Columbia . . . . . . .Virginia . . . . . . . . . . . . . . . . . . . .West Virginia . . . . . . . . . . . .North Carolina . . . . . . . . . . . .South Carolina . . . . . . . . . . . . . .Georgia . ... . . . . . ... ... . .Florida . . . . . . ... ... . ... .

Table 2.—Distribution of CT Scanners by Region and State (May 1980)–continued

Region and State

East South Central ...,Kentucky, . . . . ..., ..., ..., ...,Tennessee. ..., ..., . . . . . . .Alabama ., ..., ..., ...,Mississippi ., ., ..., ..., ..., . . .

West South Central . . . . . . . . . . . . . . . . . . . . .Arkansas . . . . . . . . . . . . . . . .Louisiana . . . . . . . . . . . . . . . . . . . . .Oklahoma . . . . .Texas . . . . . . . . . . . . . . . . . .

Mountain . . . . . . . . . . . . . . . . . . . . .Montana . . . . . .Idaho . . . . . . . . . . .Wyoming. . ., . . . . . . . . . . ...,Colorado . . . . . . . . . . . .New Mexico . . . . . . . . . . . . . . . . . . . . . . . . . .Arizona . . . . . . . . . . . . . . . . . . . . . . . . .Utah. . . . . . . . . . . . . . . . . . . . . . . . . . .Nevada. . . . . . . . . . . . . . . . . . . . .

Pacific. . . . . . . . ..., . . . . ...,Washington. . . . . . . . . ...,Oregon. . . . . . . . . . . . .California. . . ., . . . . . . . .Alaska . . . . . . . . ..., ..., .Hawaii ., . . . . . ..., . . ..., ..., . . . . .

Puerto Rico. . . . . . . . . . . . . . . . . . . . . . .

Subtotal . . . . . . . . . . . . . . . . . . . . . .

Total scanners in the United States ..., . . . .

aPopulatlon datawereobtalned from the U S Bureau of the Censusb[nclud[ng I mOblle scanner

Clncludlng 2 mobde scanners

Number of CT scanners Number of CT scanners per million populationa

Hospital Hospital Office Total

67 611 2 3.1 0.6 3.724 3 5.5 0.7 6.222 0 5.8 — 5.810 1 4.1 — 4.1

126 19 1 4.1 0.5 4.6

2 5b

3 6.2 0.7 6.915 0 5.2 . 5.277 25 5.8 1.9 7,762 14

3 0 3.8 — 3.83 0 3.3 — 3.32 0 4,4 — 4,4

17 3 6.1 1.1 7,25 5 4.0 4.0 8.0

I 8b

4 7.3 1.6 9.06 1 4.4 0.7 5.18 1 11.4 1,4 12.8

239 5419 11 4.8 2.8 7.616 0 6.3 — 6.3

196e 43 8.6 1,9 10.53 0 7.4 — 7.45 0 5.5 — 5.52 3 0.6 1.0 1.6—

1’, 193 2 7 8 5.4 1.3 6 . 7—1,471 6.7 -

The ratio of scanners per million populationis often used as a standard by which to comparescanner availability in the United States to scan-ner availability in other countries. Table 4 givesthe number of CT scanners in the United Statesand in a number of other industrialized coun-tries early in 1979. It seems apparent from thesedata, and from other sources, that the UnitedStates at present has the greatest number of CTscanners of any country in the world. This in-formation is not easy to interpret, however,because the appropriate number of scanners isnot known. One also needs to consider that theUnited States has, in addition to scanners, thegreatest amount of other diagnostic technol-ogies such as conventional X-ray (120) and alarge number of surgeons per capita in compari-son to such countries as Canada and the UnitedKingdom (28,178).

dlncludlng 3 scanners at [he National Institutes of Health If they are removed

from the totals, Maryland has 58 scanners per mll I on populationelncludlng 8 mobile scanners

Within the United States, the ratio of scannersper million population is often used as an in-dicator of relative geographic maldistributionfrom State to State, as the discussion above il-lustrates. The ratio is inadequate as an indicatorof relative access, however, because it does notincorporate the geographic dimension of access.The ranking of States by number of square milesis shown in the last column of table 3. It is strik-ing that the 10 States with the highest scanner-to-population ratios are all relatively largeStates characterized by relatively low popula-tion densities. Several of these States are furthercharacterized as mostly rural, so their popula-tion may be expected to be dispersed over theState.

The point to be made is that both populationand geography are essential factors to consider

Table 3.—Ranking of States by CT Scanners per Million Population as of February 1979 and May 1980

States (ranked by May May 1980 ratio of February 1979 ratio of State ranking based on1980 ratio of scanners/ scanners/miIIion scanners/million 1979 ratio of scanners/ State ranking based onmillion population) population population million population size in square miles

1. Nevada. . . . . . . . . . . .2. Florida . . . . . . . . . . . .3. California. . . . . . . . . .4. Missouri. . . . . . . . . . .5. North Dakota. . . . . . .6. Arizona. . . . . . . . . . . .7. Nebraska. . . . . . . . . .8. New Mexico. . . . . . . .9. Texas . . . . . . . . . . . . .

10. Georgia . . . . . . . . . . .

11. South Carolina . . . . .12. Washington. . . . . . . .13. Alaska . . . . . . . . . . . .14. Colorado . . . . . . . . . .15. Illinois . . . . . . . . . . . .16. Louisiana. . . . . . . . . .17. Maryland . . . . . . . . . .18. Oregon. . . . . . . . . . . .19. Pennsylvania. . . . . . .20. Tennessee. . . . . . . . .

21. Virginia . . . . . . . . . . .22. Iowa . . . . . . . . . . . . . .23. Kansas. . . . . . . . . . . .24. New York. . . . . . . . . .25. Alabama. . . . . . . . . . .26. Ohio . . . . . . . . . . . . . .27. Minnesota . . . . . . . . .28. North Carolina . . . . .29. Hawaii . . . . . . . . . . . .30. Indiana. . . . . . . . . . . .

31. Wisconsin . . . . . . . . .32. Massachusetts. . . . .33. West Virginia. . . . . . .34. Oklahoma . . . . . . . . .35. Utah . . . . . . . . . . . . . .36. New Hampshire . . . .37. Arkansas . . . . . . . . . .38. Maine. . . . . . . . . . . . .39. Vermont. . . . . . . . . . .40. South Dakota . . . . . .

41. Wyoming . . . . . . . . . .42. Connecticut . . . . . . .43. Mississippi . . . . . . . .44. Kentucky . . . . . . . . . .45. New Jersey . . . . . . . .46. Michigan . . . . . . . . . .47. Delaware . . . . . . . . . .48. Idaho . . . . . . . . . . . . .49. Rhode Island. . . . . . .50. Montana. . . . . . . . . . .

12.810.910.59.49.19.08.38.07.77.67.67.67.47.27.16.96.56.36.26.26.25.95.95.95.85.85.75.75.55.55.55.45.35.25.14.74.64.64.64.44.44.24.13.73.73.63.43.33.33.0

12.810.39.88.37.77.88.34.26.37.5

2.16.37.47.26.05.66.35.24.56.0

4.54.24.74.84.94.46.04.94.55.3

5.23.64.34.65.14.74.23.64.24.4

2.53.63.33.22.63.43.45.82.23.0

1234765

3712

8

5013

910141811203016313626252332152429192141342822273540383348394445474342174946

722

319176

1552

21

4020

18

2431411033343625143029351228473836454218114427394316

94832374623491350

4

SOURCE. Office of Technology Assessment.

in determining access, This is of particular im- availability of CT scanners in the largest State,portance when making comparisons of access Alaska, and the smallest, Rhode Island. 130thbetween States or countries. An intuitive appre- States have three scanners, but Alaska has 7.4ciation of the relationship between population scanners per million population while Rhode Is-and geography is illustrated by comparing the land has only 2.2. Few would infer from this in-

Table 4.— Distribution of Installed CT Scanners in 10 Countries (1978 and 1979)

March 1978-.

1979

ScannersNumber of scanners per million

Country a Head Body Total population

United States. . . . 3 3 7 668 1,005 4.6Japan ., . . . . . . . 180 112 292 2.6W e s t G e r m a n y . 51 42 93 1.5Australia. . . . . . . . . . u u u uCanada. . . . . . . . . . . . u u u uSweden. . . . . . . . . . 8 5 13 1.6N e t h e r l a n d s b . u u u uUnited Kingdom ., 36 16 52 0.9France c. . . . . . . . . . 10 2 12 0.2Iceland . . . . . . . . . . . 0 0 0 0.0

Number of scanners

Head Body Total

400 854 1,254304 212 516

u u 1607 21 289 29 388 6 14

u u 2039 18 5720 10 30

0 0 0

Scannersper millionpopulation

5.7 (Feb.)4.6 (Apr.)2.6 (July)1.9 (Jan.)1.7 (May)1.7 (Feb.)1.4 (Jan.)1.0 (Jan.)0.6 (Jan.)0.0 (Jan.)

Key to symbols U = UnknownaRanked by scanners per m(ll(Ofl popu[atlofl In 1979bThe Netherlands has planned to Install 30 head scanners and 8 body scannersCln France an additional 21 scanners were authorized {n JulY 1979

SOURCE Reprinted from reference 128 Data sources can be found there

formation, however, that Alaskans have betteraccess to scanners (according to geographicavailability) than Rhode Islanders. On the otherhand, the population of Rhode Island is almosttwice that of Alaska. Consequently, its greaterpopulation density implies that access may begreater in the geographic sense, but less in termsof the greater population served.

OTA suggests consideration, therefore, of analternative index of scanner availability thatwould incorporate the geographic dimension ofaccess. The index is based on the index used tocompare physician availability for any des-ignated unit of analysis (124). The unit ofanalysis that would be most appropriate in thiscase would be individual health services areas. 3

First, a ratio of the number of scanners in ahealth service area to the number of scanners inthe entire United States would be computed.Next, a ratio of the population density (personsper square mile) of the health services area tothe population densit y of the United Stateswould be determined. The availability index ofthe geographic unit of analysis would be theweighted average of the ratio between the firstratio and the second. This index would have theadvantage of incorporating the relative impactsof geographic effects and population effects onaccess. The calculation of such an index would

‘Health xr~lce areas are the ge<)ptllltlcal areas served by corre-spondtn~ health systems a~encles ( HSAS ).

not give a clear indication of what appropriateaccess should be, since the only point of ref-erence would be the national average. Althoughthis index would not be the ideal indicator,given the data currently available, it would bean improved indicator to use in discussions ofcomparative accessibility.

In terms of health service areas, the distribu-tion of scanners has improved. In 1979, therewere 16 health service areas with no CT scan-ners, but now there are only 3. Table 5 showsthe number and type of CT scanner (head orbody) by health service area of May 1980. Al-though health service areas are smaller unitsthan States, and therefore give a better sense ofgeographic distribution, some encompass theentire State. However, the ratio of scanners topopulation still varies greatly from one healthservice area to another.

Institutional Distribution

In May 1980, 18.9 percent of the 1,471 opera-tional scanners were in private offices andclinics: This is very close to the figure of 19 per-cent observed in the May 1977 data presented inthe original OTA CT report (129). Table 2shows the number of scanners in noninstitu-tional settings by State. More importantly, table2 also shows the ratio of private office scannersto population. States with high ratios includeNew Mexico, with 4.0 scanners per million per-

Table 5.— Number of CT Head and Body Scanners by Health Service Areas (May 1980)

Health service area (by number)

State

-

1 2 3 4 5 6 7 8 9 10 11 12 13- 14 Total2B 1B 6B 2B 4B 2B ●

Alabama . ... . . . . . 2H 1H 2H —

I 2B 1B I I I IAlaska. ., ... . . . . . . . . .

A r i z o n a

A r k a n s a s . . . . , . . . ,

C a l i f o r n i a . , . . . , .

Colorado . . . . . . .

Connecticut, ., ., ..., . . .

Delaware. . ...,

D i s t r i c t o f C o l u m b i a

Florida ..., ..., ...,

Georgia . . . . . . ..., ..., ...,

Hawaii. ..., . . . . . ...,

Idaho, . . .

Illinois. . . . . . .

Indiana. . ..., ...,

Iowa . . . . . . . . . . . . . . . . .

Kansas . . . . . . . . . . . .., ....,

Kentucky . . . . . . . . . . . . . . . .

Louisiana ..,...... . .

Maine ..., ..., . . . . . . . . . . .

M a r y l a n d . . . . . . .

Massachusetts . . . . . ...,

Michigan. . .

Minnesota . . . . . . . . . . . . . . . . .

Mississippi. . . . . . . . ..., .,..

M i s s o u r i . . . . . . .

Montana. . ... . . . . .

Nebraska . . . . . . . . . . . . . . . ,.

Nevada, . . . . . . . . . . . . .

New Hampshire. ...., ..., . .

New Jersey. . . . . . . . . . . . . . .

Table 5.— Number of CT Head and Body Scanners by Health Service Areas (May 1980)–continued

SOURCE Office of Technology Assessment

sons, Florida (3.0), Washington (2.8), NewYork (2.4), and Vermont (2.3). A number ofStates have no private office scanners.

The proportion of scanners located in privateoffices versus hospitals raises concern over theissue of access. Data on the hospitals by type

and size do little to assuage this concern. Tables6 and 7 present data on the distribution of CTscanners by type of facility and for short-term,general community hospitals, by size of hos-pital. Of a total of 5,881 short-term general hos-pitals (12), 1,01.5 or 17.3 percent have CT scan-ners. As shown in table 7, 361 hospitals, or 35.6

Table 6.—Distribution of CT Scanners by Type of Facility (May 1980)

Facilities with CT scanners

PercentageType of facility Number of totalAll hospitals. . . . . . . . . . . . . . . . 1,041 78.7%

Community hospitals . . . . . (1,015) (76.7)Other short-term hospitalsb. (26) (1.9)

Mobile scanners . . . . . . . . . . . . 18 1.4Office and clinics . . . . . . . . . . . 264 19.9

Total . . . . . . . . . . . . . . . . . . . 1,323 1OO.OO/O

1,175 79.9%(1,147) (77.9)

(28) (1.9)18 1.2

278 18.9

1,471 100.0%

alncl udes pror, rletary, public, and voluntary community hospitalsblncludes 17 VA hospttals and 8 armed forces hoswtalsSOURCE Of fl(:e of Technology Assessment and Amertcan Hospital Association

Table 7.—Distribution of CT Scanners in Community Hospitals”by Hospital Size (May 1980)

All hospitals Hospitals with CT scannersPercentage Percentage Number of

Size of hospital Number of total Number of total CT scannersO- 99 beds. . 2,833 48.20/0 14 0 . 5 % 14

100-199 beds. . 1,401 23.8 129 9.2 133200-299 beds. . 713 12.1 218 30.6 225300-399 beds. . 380 6.5 220 57.9 228400-499 beds . . 243 4.1 170 70.0 187500 and over . . 311 5.3 264 84.9 360

Total. . . . . . 5,881 100.0% 1,015 1 7.3=X. ‘1 ,147

alnc ludes proprietary, public, and VOIU ntary hospitals Does not Include federally surworted hospitalsSOURCE Office of Technology Assessment and American Hospital Assoclatlon

percent of the total community hospitals havingCT scanners, are less than 300 beds in size.

Of the total short-term general hospitals,1,832 are supported by State and local govern-ments, and only 161, or 8.8 percent, have CTscanners. When size is taken into consideration,this point becomes even more striking. A short-term general hospital with 500 beds or more isalmost certain to have an active emergencyroom, a neurosurgery service, and other special-ized and acute care services that virtually re-quire a CT scanner for the provision of appro-priate care. But of the 47 local-government-sup-ported community hospitals of at least 500 beds,only 32 have CT scanners. New York City alonehas six such hospitals with no CT scanner.These include Bellevue Hospital (1,258 beds),4

Harlem Hospital Center (884 beds), Metropoli-tan Hospital (754 beds), and the City Hospital

—4Money is currently budgeted for a scanner for Bellevue Hos-

pital.

of Elmhurst (816 beds). Other important publichospitals in the United States without CT scan-ners include Cook County Hospital in Chicago(1,326 beds), D.C. General Hospital in Wash-ington, D. C. (600 beds), and San Juan Munici-pal Hospital in San Juan, Puerto Rico (687beds). Not only are the patients of these hos-pitals poor, but they are often members ofminority groups.

The problems related to the distribution ofCT scanners in hospitals are not confined tothose in urban ghettos. The Department of De-fense and the Veterans Administration (VA)operate large hospital systems. Although thesehospitals do not run the busy emergency roomsof the urban public hospitals, they do servelarge populations. Only 17 of 171 VA hospitalsand 8 of 135 armed forces hospitals currentlyhave CT scanners. There are 44 VA hospitalsacross the country with 500 beds or more thathave no CT scanner. (The average bed size ofthis group of VA hospitals is over 800 beds. )

There is 1 armed forces hospital of over 500beds with no scanner. (See app. C for moredetails on the VA and armed forces’ policiestoward CT scanners. )

As shown in table 7, 84.9 percent of short-term general hospitals of over 500 beds nowhave at least 1 CT scanner; 264 such hospitalshave 360 scanners. Thus, there is a fair propor-tion of community hospitals with more than onescanner. This category is comprised of severaltypes of hospitals including voluntary, public,and proprietary. Most of the community hos-pitals with CT scanners are voluntary. Thecategory of public hospitals includes hospitalssupported at the level of hospital district, city,county, and State governments, but excludesfederally supported hospitals. Of the State-sup-ported hospitals with over 500 beds that haveCT scanners, all but one are affiliated with theState university. This reflects the concentrationof diffusion of CT scanners in hospitals affili-ated with virtually all medical schools in thecountry. However, not all university teachinghospitals are large, and some major ones lack aCT scanner (e.g., Beth Israel in Boston).

The plight of local-government-operatedhospitals has already been discussed above. Thecase of proprietary hospitals also illustrates in-equity in distribution of CT scanners. There area total of 81 such institutions with CT scanners.In general, proprietary hospitals tend to besmaller, in terms of bed size, than other com-munity hospitals. Of the 80 proprietary hospi-tals with scanners, 40 have less than 200 beds.

In total, there are 97 hospitals of all typeswith 500 beds or more which are still without aCT scanner. The 44 VA hospitals constitutealmost half of these, or 45.4 percent. The 15large, publicly supported urban hospitals and 1military hospital discussed above compriseanother 16.5 percent, and the remaining 36hospitals, or 40.3 percent, are “private” com-munity hospitals, including voluntary and pro-prietary hospitals. 5 There are, then, 51 com-munity hospitals of 500 beds or more without

‘The discrepancy of two hospitals from the data in table 6 is dueto the fact that two large hospital systems were counted as onehos.pltal each In the American Hospital Association data, In OTA’sanalysis, the scanners were Ilsted under individual hospitals.

scanners, which account for the 16.5 percent ofall hospitals of that type and size that are stillwithout CT scanners. b

Based on these data, a seemingly clear case ofmaldistribution of scanners within the categoryof hospital settings emerges. It is not clear,however, which type and size of hospital mayderive the greatest benefit from having a CTscanner. Modest evidence from a new study sug-gests that scanners may have a greater diagnos-tic and therapeutic impact in a public universi-ty-affiliated hospital than in a private medicalcenter with a similar affiliation (14). Yet it isthese hospitals for whom the economic andtechnical support a scanner requires may be lessfeasible.

The capital expenditures and technical sup-port required may prohibit the hospital of lessthan 200 or 300 beds from installing a scanner.Table 7, showing the proportion of communityhospitals by bed size and the proportion of eachof these groups that has a CT scanner, wouldlend support to this hypothesis.

This analysis of type of setting, and type andsize of hospital, suggests another issue besidesthat of institutional distribution of scanners:The commercial market for CT scanners, atleast in voluntary community hospitals of ap-preciable size (500 beds), may be approachingsaturation. Such a conclusion, however, is sub-ject to the qualification of type and generationof scanner being considered. Thus far, in dis-cussing the diffusion and distribution of scan-ners, the technical capabilities (beyond thoseindicated by dedicated head v. total body CTunits) of scanners have not been explicitly con-sidered. Clearly, any statements regarding satu-ration of the market are a function of the factthat these facilities merely have a CT scanner—not that they have the CT scanner of a type thatthey might need or desire. One outcome of the“rush” for scanners in 1975 was that a greatmany hospitals purchased scanners representingthe state of the art of CT technology at thattime—typically an early head scanner. Sincethat time, improvements in scanning speed and

“There is also one Public Health Service hospital not included inthese figures that has .500 beds but lacks a CT scanner.

20 ● Policy Implications of the Computed Tomography (CT) Scanner: An Update

image resolution, as well as the potential toreduce radiation exposure to patients, have oc-curred in successive models of scanners, creat-ing a concomitant demand for these new state-of-the-art CT scanners,

In considering the question of whether healthplanning policies have influenced diffusion,either in terms of the aggregate number of scan-ners, the rate of purchase, and/or market satu-ration (as qualified above), the concomitant ef-fects of the distribution and technical capabilityof existing operational scanners have been ig-nored. The focus on whether these policies havebeen effective in either limiting diffusion or pro-

moting market saturation reflects concern foronly one of the two objectives of the healthplanning laws— the containment of costs. But if,in the attempt to control diffusion, the law andrelated regulation can be shown to have effectedan inequitable distribution of medical technol-ogy that is inadequate for the needs of varioushealth care providers, then they have failed toassure the second objective—namely, ensuring

access to and quality of care. Perhaps more im-portant is the issue of whether existing healthplanning policies will be able to redress distribu-tional inequities and resolve problems related toappropriate technology in the future (see ch. 3).

TRENDS IN THE TYPE AND MANUFACTURE OF SCANNERS

The CT scanner market has undergone dra-matic changes since EMI, Ltd., developed thefirst commercial head scanner in the early1970’s. By May 1980, there was a strikingchange in type of scanner being sold (see table8). Only slightly more than half of the EMIscanners are now head scanners, compared to92 percent of the EMI scanners installed as ofMay 1977, Body scanners have increased theirdomination of the market, and by May 1980, al-most 69 percent of operational scanners were

body scanners. During the 24 months from Jan-uary 1978 through December 1979, however,Ohio-Nuclear installed 83 head scanners.7

Since the sale of the first scanner in this coun-try by EMI in 1972, the CT market has under-gone “see-saw” changes in both the number ofcompanies manufacturing CT equipment and in

7Comparison with FDA data on scanners reported installed indi-cates that a good portion of these wt>re the new lower priced scan-ners.

Table 8.—Manufacturers of CT Head and Body Scanners in Use (May 1980)

Head scanners Body scanners Total scanners

Percentage Percentage PercentageManufacturer Number of total Number of total Number of total

EMI, Ltd.. . . . . . . . . . . . . . . . . . . 284 61 .60/0 267 26,40/. 551 37.5%Ohio-Nuclear . . . . . . . . . . . . . . . 109 23.6 309 30.6 418 28.4General Electric. . . . . . . . . . . . . 16 3.5 221 21.8 237 16.1Pfizer. . . . . . . . . . . . . . . . . . . . . . — — 107 10.6 107Artronix . . . . . . . . . . . . . . . . . . .

728

.36.1 4 0.4 32

Syntex. . . . . . . . . . . . . . . . . . . . .2.2

11 2.4 17 1.7 28 1.9Picker . . . . . . . . . . . . . . . . . . . . . — — 22 2.2 22 1.5Elscint . . . . . . . . . . . . . . . . . . . . — — 13 1.3 13 0.9Varian . . . . . . . . . . . . . . . . . . . . — — 16 1.6 16 1.1AS&E. . . . . . . . . . . . . . . . . . . . . – — 11 1.1 11 0.7Philips. . . . . . . . . . . . . . . . . . . . — — 10 1.0 10 0.7Omni. . . . . . . . . . . . . . . . . . . . . 9 1.9 — 9 0.6Other a. . . . . . . . . . . . . . . . . . . .

—3 0.6 14 1.4 17 1.2

Total . . . . . . . . . . . . . . . . . . . . 460 100.0% 1,011 100.0% 1,471 100.0%

aTflree Companies— Neuroscan, Sle Tens, and cGRSOURCE Off Ice of Technology Assessment

Ch. 2—Number and Distribution of CT Scanners ● 21

their respective shares of the CT market. EMIdominated the American (and world) marketthrough mid-1975 (see figure 4). Although sixother companies were marketing CT scanners inthe United States by May of 1977, EMI still hadsold almost 60 percent of all operational scan-ners at that time.

However, the rapidly increasing number ofnew companies entering the market, as well asthe new generations of scanners they introducedto the commercial marketplace, brought aboutsome abrupt changes in the share of the marketcontrolled by early manufacturers. By March1978, there were 15 companies worldwide thathad CT scanners in operation: Only 4 of these(EMI, Pfizer, General Electric (GE), and Sie-mens) had ever manufactured a rotate andtranslate, dual-detection scanner; 7 more of

these (Philips, Elscint, Picker, Ohio-Nuclear,Syntex, Hitachi, and CGR) had entered the mar-ket with a rotate and translate, multiple detec-tion scanner; and 4 more (Varian, Artronix,Searle, and American Science and Engineering(AS&E)) had entered with a rotate-only scanner(see table B-3 in app. B) (65,120).

The most dramatic change in the U.S. marketshare occurred in 1977 with the sharp increase inthe number of scanners installed by GE. This isprimarily attributable, not to expansionarymarket trends, but to GE’s introduction of itsnew rotate-only scanner (which had been pio-neered by that company) to the commercialmarket. Both GE and Ohio-Nuclear expandedtheir share of the market during 1978, so that by1979, EMI’s share had fallen to 40 percent ofoperational scanners. By 1980, EMI’s share had

Figure 4.—Cumulative Number of CT Scanners Installed, by Manufacturer

1,500

1,400

1,300

1,200

1,100

1,000

900

800

700

600

500

400

300

200

100

0 L

Other

Ohio-Nuclear

1973 74 75 76 77 78 79 1980Year

SOURCE Off Ice of Technology Assessment

been further reduced to about 37 percent. Figure4 shows the changing relative market shares ofmanufacturers, as measured by the proportionof operational scanners in the United States,over the past several years.

Another indicator of changing market sharesis worldwide sales of scanners. According toone source, estimated 1979 sales yield a rankingof manufacturers as follows: GE ($100 million),Siemens ($50 million), Johnson & Johnson(Ohio-Nuclear) ($45 million), both Picker andPfizer ($30 million), EMI ($15 million), Elscint($l4 million), Philips ($10 million), and bothToshiba and Hitachi ($5 million)8 (48). Natural-ly, different companies tend to be more suc-cessful in certain countries, usually their own orcontiguous countries. For example, Siemens, aWest German company, tends to dominate themarket in West Germany. Despite an early leadby EMI, the Japanese companies are gainingdominance in Japan. Siemens dominates inBelgium, with CGR, a French company, havingmajor success as well. Other countries have alarger spread of manufacturers, although EMItends to have larger percentages because of itsearly domination of the market. GE is apparent-ly the only U.S. company that has had signifi-cant success outside of the United States (33).

The precipitous decline in both the number ofscanners sold and in scanner sales in 1977 and1978 marked the beginning of the end of pre-vious expansionary market trends. As the first,and for many years dominant, manufacturer ofCT scanners, EMI aptly illustrates the variousaspects of the troubled CT market over the pastfew years, Following a $29.1 million profit forthe fiscal year of 1977 (115), the medical elec-tronics division of EMI, including the CTscanner business, incurred major losses in bothfiscal years 1978 ( – $28.7 million) and in 1979( – $27.8 million) (172). In early 1979, it wasreported that EMI had begun to seek a merger ofits medical division with a U.S. company inorder to cushion these losses (31). In December1979, EMI was acquired by Thorn Electrical In-dustries, Ltd., of Great Britain, but in April1980, Thorn attempted to sell its newly acquired

—“Data collected by U.S. ma lufacturers indicate that the estimate

for the Japanese companies is much too low.

EMI scanner interests with a sale to GE (26,165).GE sought an advisory opinion from the U.S.Department of Justice, however, and was in-formed that such a takeover would probably befound to violate antitrust law. The upshot wasthat GE only acquired EMI’s non-U. S. opera-tions, leaving EMI’s U.S. operation in limbo. In1977 and 1978, EMI initiated litigation againstTechnicare of Ohio-Nuclear (45), Pfizer (113),and GE (@I). The suits filed by EMI against thesecompanies sought damages for alleged infringe-ments of its many patent rights on the CT scan-ner.9 Part of GE’s agreement in purchasing EMIwas the settlement of this patent litigation (26).

Further signs of the troubled CT scanner mar-ket are evident in the trend toward market con-solidation as measured by the number of othercompanies that have merged, are seeking to sell,or have already sold their scanner interests. Thedepressed state of the CT scanner market in1978 is reflected by the fact that by the end ofthat year, at least two companies (Searle andSyntex) went out of the CT scanner business(26,43); another (AS&E) sold its rights tomarket and produce the scanner it had pio-neered (to Pfizer) (42); and one of the leadingmanufacturers of body scanners at that time(Ohio-Nuclear) was acquired by the singlenewcomer to the CT scanner market (Johnson &Johnson) in 1978, (44). In 1979, Varian also putits CT scanner division on the market, with theintent of eliminating the divison if it could notfind a buyer (114). By October 1979, Neuroscanwas no longer making scanners, and Artronixhad notified the Food and Drug Administration(FDA) that it would cease to market scanners(90).

Thus since the beginning of 1978, eight com-panies, EMI, Searle, Syntex, AS&E, Ohio-Nu-clear of Technicare, Varian, Neuroscan, andArtronix, have left the CT market (in somefashion), and only one, Johnson & Johnson, hasentered it. As of September 1979, there were 10companies which still had CT scanners certifiedas meeting FDA performance standards mar-keted in the United States (90): These included

‘ln a separate agreement with EMI, Johnson & Johnson agreedto pay EMI $15 million to drop suit against Technicare of Ohio-Nuclear (44).

GE (United States), Siemens (West Germany),Johnson & Johnson (formerly Ohio-Nuclear,United States), Picker (United States), Pfizer(United States), EMI (Great Britain), Elscint(Israel), Philips (Netherlands), CGR (France),and Omni Medical (United States).

With the apparent exit of EMI in late 1979,nine companies remain. Counting the threeJapanese companies (Shimadzu, Toshiba, andHitachi), there are now 12 companies world-wide still manufacturing CT scanners. It isbelieved that the remaining market for scannerswill not support all of these companies,however, and further consolidation is predictedfor the future (26).

Manufacturers have cited Federal interven-tions as the culprit behind the millions of dollarslost on the CT scanner market over the last fewyears. Specifically, this calamitous turn ofevents has been blamed on the implementationof the health planning laws enacted in 1974 andon the consequent certificate-of-need (CON)regulations imposed through local health sys-tems agencies (HSAs) since 1976 (26,65). How-ever, it is also true that the expansionary trendsexhibited in the mid-1970's could not continueforever: The number of scanners that couldultimately be sold was not limitless, and thatnumber could have been reached by far fewermanufacturers than the number of manufac-turers that rushed to share in profits such asthose EMI was realizing in 197510, In addition,companies like EMI in 1976 which in 1976 faceda backlog of 250 unfilled orders, had geared upproduction capacities to meet the wildly esca-lating demand for scanners. Thus, it appearsthat there may have been substantial overes-timation of the potential market for scanners onthe part of manufacturers.

In the wake of this controversy, there havealso been modifications in marketing strategies,

‘“A. U.S. market estimate In 1975 prepared by Kidder, Peabody,and Co., predicted that a total of 1,425 CT units would be in placeby the end of 1980. In fact, th]s number was probably attained bythe end ot 1979 (according to manufacturers’ sales data). In report-ing the above projection, however, it IS interesting to note that theauthor, cri tlcal of the lmpos,ltion of health planning measures in1976, noted: “The growth curve was well on the way to reachingthat level ( 1425} until It encountered the Federal and State CONlaws that were Imposed” (65 ).

some of which appear to be in response to theCON review process and specific regulations.The advent of the new low-priced scanners, inparticular, has drawn the attention of policy-makers. At least three companies have modelsof a head scanner having a list price of less than$200,2000 (49). Four of these models sell for lessthan $l50,000, the threshold figure at whichCON approval is required for purchase. Onecompany has a body scanner whose purchaseprice is less than $100,000, plus maintenancecosts (49). The interim regulations issued inApril 1979 by the Bureau of Health Planning(BHP), however, have countered this particularstrategy as a means of skirting the purview ofCON review (see ch. 3).

According to FDA data on scanners reportedas installed between June 1978 and June 1979,there were 39 scanners listed that sold for lessthan $200,000 (95). Ohio-Nuclear has been par-ticularly active in selling these scanners, havingsold 16 of the model 150 Delta-Scan head scan-ner that costs approximately $145,000 and 6 ofthe model 110 Delta-Scan that is priced at$96,500 (49). Omni Medical has also been activein the promotion and sale of these scanners andhas reportedly concentrated its production in alow-cost (sub-$l50,000) highly reliable cranialCT scanner (40). The technical capabilities ofthese scanners are more limited than those of themore expensive and technically sophisticatedmodels, and this reduces their appeal to manypotential buyers. Still, these new lower pricedscanners avail themselves to a new market ofsmall hospitals and private offices (49).

Another strategy of some manufacturers hasbeen to upgrade and refurbish older scanners;this includes modifying head scanners to bodyunits. Several of the “new” cheaper scanners areactually older scanners that have been boughtback, or traded in on more advanced newerequipment, and then refurbished by the manu-facturers (90). EMI and Pfizer have both beenengaged in programs of updating older modelsto the latest specifications. Generally, refur-bishing can be done for less than $100,000. Thechange in definition from CT scanning equip-ment to CT scanning services (again by virtue ofthe April 1979 interim regulations issued by

BHP), however, means that changes such asupgrading a head to a body machine are subjectto CON review (see ch. 3).

Another strategy manufacturers have used todiversify their CT markets has been to installscanners in a mobile environment. At least twocompanies are now selling various models oftheir scanners installed in special vehicles. Ac-cording to OTA’s data, the number of mobilescanners as of February 1979 had doubled, go-ing from 7 to 14 in less than a year; by May of1980, the number had increased further to 18.For a while, the market potential for mobileunits appeared substantial, since these scannerswere not subject to CON review. It was also ex-pected that medicare would soon begin to payfor scans done on mobile scanners. In an-ticipation of that announcement by medicare,however, mobile scanners were placed withinthe purview of CON review (once again, underthe interim regulations issued by BHP in April1979) (see ch. 3). Furthermore, when the an-nouncement did come from medicare, the reg-ulations for reimbursement stipulated that reim-bursement would be made only for scans doneon CON-approved mobile scanners (85). None-theless, the number of mobile scanners seemscertain to increase. One company is now serv-icing the needs of 44 hospitals in southernCalifornia for CT scanners (166), and that com-pany reports to OTA that it expects to expandits present stock of 21 operational scanners (asof October 1980) by 1 to 11/2 per month. This

development is clearly to the advantage of thesmaller hospitals that cannot support a scanneron their own, and it may well be an efficientway to provide access to CT scanning services(70). So far, however, Federal policy with re-spect to mobile scanners has been conservative,and sharing has not been encouraged.

It would seem fair to conclude that manufac-turers have attempted to place the blame for thechanges that have occurred in the CT scannermarket over the past 2 years on cost-consciousFederal policies. Although it is unlikely thatthese policies are solely to blame for the ratherabrupt turn of market events, it is clear that thetrend in Federal policies toward the CT scannermarket over the past 2 years has been one of in-creased restraint in a kind of “cat and mouse”game with the manufacturers in the name ofcost containment. It would seem that Federalpolicymakers and manufacturers alike couldbenefit from taking a broader, more compare-hensive view of the forces shaping these events,and developing a more balanced appreciation ofthe two objectives of ensuring access and quali-ty care and containing costs. One of the forces,research and development of existing andemerging diagnostic imaging modalities withwhich CT is competing (or will eventually com-pete) for a place in medical practice, is discussedin appendix B. Federal policies toward CT scan-ners and changes in those policies since 1978 aresummarized in the next chapter,

3.

Changes in Federal PoliciesToward CT Scanners

3 ●

Changes in Federal PoliciesToward CT Scanners

This update of information presented inOTA’s 1978 report on the computed tomogra-phy (CT) scanner (129) focuses on four areas ofGovernment policy: 1) research and develop-ment, 2) evaluation, 3) regulation, and 4) fi-nancing. Generally, the policies in these fourareas address issues characterizing the stages ofresearch, development, and demonstration ofefficacy and safety, diffusion, and widespreaduse of medical technology, respectively. Thesefunctions are performed largely by agencies andprograms within the Department of Health andHuman Services (DHHS), as shown in table 1 in

RESEARCH AND DEVELOPMENT

As previously stated, biomedical R&D hasbeen generously supported by Federal funds.Nevertheless, Federal investment in the develop-ment of CT scanning was small. Because CT “iswell beyond the initial stage of development,Federal support has now largely ended. The lastmajor CT project funded by the National Insti-tutes of Health (NIH) ended in April 1978 (22).It supported the development of a more techni-cally advanced, ultrafast prototype scanner.Basic R&D agendas at present reflect the pursuitof alternative imaging techniques that promiseeven greater technical capabilities than CT.Such new technologies may eventually become

EVALUATION

The available efficacy information on CTscanning in 1978, OTA found, was inadequatefor the purposes of planning agencies and otherorganizations in need of such information (129).Planning agencies, Professional Standards Re-

chapter 1. 12 The changes that have occurred inthese functions, programs, and agencies since1978 are discussed below.

I Following the crea tlon of a separate Department oi Education,the name of what was formerly the Department of Health, Educa-tlc~n, and ~~’elfare ( DHEF7’ ~ wa~ c h a n g e d Ettectlve Nla} 4, 1 Q80,the new name IS the Department t>t Health and Human ScrvlcesI DHHS)

‘OTA’s ~rl~lnal report ( 129) describes the involvement of nine(~t her agenc]e+ (Jr departmcn t> CIU t +]de DH E\\r that were I n v<ll~’edI n w~me wav WI th CT \ca nn I n~. hl ~l~t ~verc c<>ncerned WI t h wicty~nd radla tt(}n expo\LI re. Two ot these, t h e Lreterans Admlnlstra -tl{~n and the Department (~t Detenw, purchaw <canners and pr(~-vldc wrv Ice>. An update (~1 act I VI tws and po] ]c)e~ Of t hew I w()a~enc]e~ 1> presented I n app. C-. tlt t h IS LI pda t c.

alternatives to, and may even surpass the capa-bilities of, CT scanning (122).

At this time, there are a number of new imag-ing techniques evolving that have present andgreater future clinical applications. These in-clude ultrasound tomography, scintigraphy, dy-namic and spatial reconstructive CT, and elec-tronic recording. In addition, there are two newtechniques that will require considerable capitalinvestment: positron emission transaxial tomog-raphy scanning and nuclear magnetic resonancescanning. The latter two applications are de-scribed in some detail in appendix B.

view Organizations (PSROs), and third-partypayers did not have the information they re-quired to determine the need for scanners, ap-propriate standards of use, and appropriate in-dications for reimbursement, respectively. The

2 7

1978 report on the CT scanner (129) stated:“The development and diffusion of CT scannersoccurred without formal and detailed proof oftheir safety and efficacy . . . Nonetheless, by1977, efficacy and safety have been more thor-oughly assessed for CT scanning than for manyother medical technologies at a similar stage ofdevelopment and use. The evidence has notcome from well-designed, prospective clinicaltrials, but as typical for medical technologies, ithas been obtained from analyses of clinical ex-perience. ” This summary statement is still gen-erally valid. The Federal Government must takea large share of the blame for this situation. Fed-eral investment in clinical trials of efficacy andsafety has been small (125).

As defined in the original CT report, efficacyis more than a simple consideration of benefits.No technology is beneficial in the absolute. It isbeneficial only when used in an appropriatemanner—for a defined population, for a givenmedical problem, under certain conditions ofuse, and for a specified outcome(s) (125). Deter-mining efficacy thus becomes a matter of deter-mining indications for use, stated in terms of allfour of these criteria. The task is a formidableone.

For a diagnostic technology such as CT, eval-uating these benefits can become very compli-cated, depending on the type of impact specified(125). For example, should a diagnostic devicebe evaluated for its impact on diagnosis, ontreatment, or on patient outcome (13,61)? Mostof the available evaluations of CT scanning arelimited to evaluations at the levels of technicalcapability and diagnostic accuracy. Few addressdiagnostic and therapeutic impacts. Fewer stillare available that attempt to determine efficacyin terms of patient outcome (102,181,185,186).The focus of research is often on the methods forgoing about the task of evaluation (27,173,185).There has been little Federal initiative to under-take or support such evaluation of efficacy ofCT scanning (60,181). Little information that in-dicates in a definitive way either appropriateuses or the benefits of use beyond diagnostic ac-curacy has been forthcoming (32).

Available information regarding the efficacyof CT scanning is much more conclusive for

head than for body scanning (92,93), and it isgenerally acknowledged that efficacy is muchbetter established for the former (3). As yet,there are insufficient numbers of adequate stud-ies or patterns of use available to ascertain infull the proper indications for CT body scanning(167). However, it should be noted that bodyscanners are able to perform head scans whosequality is at least as good as that of head scan-ners. To the extent that a body scanner is usedto scan heads, its usefulness may be said to bemore firmly established.

There have been numerous evaluations enun-ciating the comparative benefits of CT andother diagnostic modalities for specific clinicalindications in the case of head scanning (19) andfor certain anatomical regions in the case ofbody scanning. CT of the head has been foundto perform favorably in comparison with sev-eral neurodiagnostic procedures and to havepartially supplanted the use of some of these (4).Body scanning capabilities have most often beencompared to ultrasonography, but have not al-ways been found to be decisively superior ( 1,2).To a large extent, the inconclusive efficacy

status of body scanning is a function of themany more possible clinical indications andorgans to which computed body tomographymay be applied, as well as the large number ofalternative imaging and other diagnostic tech-nologies with which it must be compared (167).In addition, any physician or institution has alimited number of patients with a specific condi-tion, and outcome data are generally lacking inall medical care (125).

To complicate matters further, a more impor-ant question now emerging is when to use CTscanning vis-a-vis other modalities rather thanwhether or not to use it at all. Evaluations of ef-ficacy that compare the benefits of applying onetechnology to those of another for a given prob-lem can provide information that will enable ef-ficient, as well as efficacious, application forthese technologies (184). To this end, the objec-tive of comparative evaluation should be notonly to determine whether one modality cansupplant or replace another, but also to deter-mine whether, when, and how the modalities

might be used in a complementary way toachieve even greater benefit and efficiency.

Even if there were available efficacy informa-tion for CT scanning that was complete ac-cording to the criteria of application, benefit,relativity, and complementarily, the questionwould remain as to whether planners, PSROs,and third-party payers would then have the in-formation they need to make decisions requiredof them. The decisions made by planning agen-cies and PSROs should be in keeping with theirtriple mandate to contain the cost of medicalcare while simultaneously assuring quality andaccess. For third-party payers, the availabilityof even the best efficacy information may ad-dress only the problem of reimbursement for in-efficacious procedures and technologies. Theimportant policy question is whether it is possi-ble to encourage a choice between competing al-ternatives or develop methods to assure com-plementary uses of them based on diagnosticsuperiority.

An idea gaining prominence is that the needsof decisionmakers in these agencies and pro-grams can be met by information from eco-nomic evaluations, perhaps in the form of cost-benefit analyses (CBAs) or cost-effectivenessanalyses (CEAs). The momentum in the re-search community toward these formal analytictechniques is based on the premise that the tech-niques of CEA/CBA can contribute to achievingcost-containment objectives—an assumptionthat may be untenable. Countering this op-timism is a growing body of skepticism re-garding the potential use and usefulness of theseeconomic analyses (127). Nevertheless, the newNational Center for Health Care Technology(NCHCT) (see below) has a specific mandate todevelop such information.

Review of the cost impact literature of CTscanning reveals a myriad of approaches to eco-nomic evaluation (20,50,51,54,55,89,187), fewof which offer real assistance to planners, uti-lization review groups, or third-party payersfaced with resource allocation decisions (180).To date, the bulk of most economic evaluationsof CT have been analyses of costs of CT scan-ning only or of the impact of CT on diagnosticcosts (180). Still, CT scanning is probably the

medical technology which has been most oftensubjected to economic evaluation, and specifi-cally, to so- called CEA/CBA.

The difficulties in applying CEA/CBA tomedical technologies in general are well doc-umented (127). But applications to diagnostictechnologies present even more difficult prob-lems (183). In the application of CEA/CBA toany technology, there are tremendous problemsin estimating both costs and benefits (or effects).The ability to conduct the CEA/CBA is depend-ent on (among other things) the availability notonly of good cost estimates, but also of validefficacy studies which are the basis for quan-titative estimates of benefits (183). This method-ological role underscores the need to developscientifically based efficacy information. Lackof such information greatly exacerbates themethodological difficulties of any analyses at-tempting to rationally compare costs and out-comes. The inclusion of CEA in the mandate ofthe new NCHCT is appropriate, but high expec-tations regarding its contribution to policy ob-jectives may not be, owing both to a continuedlack of adequate efficacy and effectiveness in-formation and to the other unanswered ques-tions concerning the methodological validity ofthe analyses themselves and their usefulness indecisionmaking.

Despite these complications, well-designedstudies are possible. The relative paucity ofscientifically derived efficacy information per-sists, and repercussions continue to be felt bythe affected agencies and organizations. Somechanges along the lines of proposals containedin OTA’s 1978 report regarding efficacy havebeen made. The most promising of these is thelegislation authorizing NCHCT. Newly man-dated by the Health Services Research, HealthStatistics, and Health Care Technology Act of1978, 3 this fledgling organization has now beenin operation a little less than 2 years. The effectsof this organization lie in the future, however,because staff and resources so far have beenlimited.

The mandate of NCHCT is a broad one, re-lating in some fashion to most technology-—

3Health Serz~ ices Research, Health Statistics, a}~d Health CareTec-hnology Act of 1978 (Public Law 95-623).

related issues and activities within DHHS. As ofOctober 1979, CT scanners were 1 of 16 tech-nologies on the NCHCT list of priorities forassessment. One of NCHCT’s most importantauthorities, however, is its responsibility to rec-ommend to the Health Care Financing Admin-istration (HCFA) what technologies should orshould not be reimbursed by the Federal Gov-ernment. This determination is to be made pri-marily on the basis of available information re-garding the safety and efficacy of the technol-ogy. This formal link between safety and ef-ficacy information and reimbursement decisionsrealizes one of the proposed options presented inOTA’s original CT report (129) (see app. A).In this advisory capacity, NCHCT formallyassumes the function previously served by thenow defunct Office of Health Practice Assess-ment (OHPA) in the Office of the AssistantSecretary for Health (OASH). To date, thelimited staff and resources of the Center havebeen primarily devoted to answering inquiriesfrom HCFA regarding these reimbursement de-cisions (177). However, NCHCT has a specificmandate to develop information on efficacy,safety, and cost effectiveness.

A second development relating to evaluationefforts at a Federal level is the series of consen-sus development conferences being sponsoredby the Office for Medical Applications of Re-search of NIH. These meetings convene over aparticular technology or disease category andattempt to reach a consensus judgment regard-ing efficacy and appropriate conditions of use.These conferences provide a forum for bringingtogether representatives of the academic andpracticing medical communities. The outcomerepresents a consensus falling on a middleground between analyses of clinical practice andscientifically derived evidence from clinicaltrials of efficacy and safety. The first of theseconferences was held in September 1977. Alto-gether, 12 were held in 1977 and 1978, and 28had been held by October 1980 (58). Both CThead and body scanning are on the conferenceagenda. The first consensus conference on CTwill focus on CT scanning of the central nervoussystem and is scheduled to be held in 1981.

The consensus development conferenceplanned for CT scanning will be jointly spon-sored by NCHCT and NIH. The Center’s re-sponsibility will be to provide cost-effectivenessinformation, while the responsibility of NIHwill be to provide the medical and technicalevidence. The conference on CT scanning willbe one of the first that will include cost-effectiveness information (131). What sort ofcost information NCHCT will supply, however,remains to be seen.

Because scientific evidence, clinical experi-ence, and expert opinion regarding the use ofCT are fragmented, and because practitionersand policy makers have had an immediate needfor efficacy information, various scientific orga-nizations, professional medical societies, andpeer review groups have reviewed and weighedthe available evidence (see app. D). Severalhave reached a consensus and issued formalpolicy statements on appropriate applications ofCT in medical practice. For example, in 1977,the Institute of Medicine (IOM) published a listof indications for appropriate use of CT scan-ning of the head and body as part of a policystatement (116). In July 1979, part of this listwas updated and augmented by the Society forComputed Body Tomography (SCBT), whichpublished a list of indications for extracranial(other than brain) applications of CT (164). InSeptember 1979, the American College of Radi-ology (ACR), the professional organization ofradiologists, issued a formal policy statement onCT scanning which concluded that the diagnos-tic efficacy of CT is no longer in question andcited six roles for CT scanning in medical prac-tice (10,11). These roles were offered as ageneral guide for use of CT, while the specificclinical areas and indications for CT scanningwere left to be determined locally by hospitalmedical staffs or other recognized peer reviewgroups. Also the Radiological Society of NorthAmerica held a convention in November 1979,at which papers documenting new uses of CT inclinical practice or reviewing evidence of es-tablished clinical uses were presented and dis-cussed (134). These mechanisms are critically

important to the practicing medical community

in establishing the role and application of CTscanning in medical practice.

It is critical to develop some form of informa-tion that addresses the issue of resource alloca-tion underlying all policymaking. From theplanner’s perspective of allocating resources, theimportant question is not simply whether theprocedure or the technology producing it is jus-tified on the basis of its having some efficacy, oreven whether its introduction and use mightraise or lower total health care costs: It is howthe diagnostic capability should be used in thepractice of medicine (180), Only by being ableto identify which patients should receive a pro-cedure during their treatment is one able toknow whether there is too little or too much CT

capacity to meet the needs of any community.This requires balancing benefits and risks (125).Similarly, evaluations of CT, whether economicor some other type, should be able to addressthe incentives toward excessive use that char-acterize current reimbursement policy. Presentmethods of reimbursement decisionmaking pro-mote the use of additional technologies andprocedures—not tradeoffs between them (184).Evaluations that could identify when, if at all,CT should be used in the diagnostic evaluation,treatment, or monitoring of a given patientcould provide the necessary information toenable reimbursement policy to encourage themost efficient—as well as the most efficacious—use of technologies in patient care.

REGULATION OF EFFICACY AND SAFETY

Somewhere between the policy areas of eval-uation and regulation lie the medical devicesprogram and the radiation safety program ad-ministered by the Food and Drug Adminis-tration (FDA). At the time of OTA’s original CTreport (129), FDA regulated CT scanners to en-sure minimum radiation exposure with anequipment standard and it was beginning to im-plement the enabling Medical Devices Amend-ments of 1976.4 Under the Medical DevicesAmendments, CT will be categorized as a class11 device, which means that CT scanners will berequired to meet specified technical perform-ance standards. These standards have beendeveloped by the Bureau of Radiological Health(BRH) within FDA. By virtue of an interbureauagreement with the Bureau of Medical Devices,BRH assumed the lead role in FDA for allradiological devices as of April 1979 (18,189).Safety of radiological devices is also regulatedby FDA through BRH, as described in OTA’s1978 report (129). CT scanners became subjectto the 1974 performance standard that appliedto diagnostic X-ray equipment. Since 1976, BRHhas been in the process of developing amend-ments to the general X-ray performance stand-ards to include criteria specific to CT scanners.

In March 1978 and October 1978, draft amend-ments were sent out for comment. The finalanalysis of comments has been completed, andfinal rules are expected to be published in 1981.

These amendments to the X-ray standard willbe the first performance standards written spe-cifically for CT scanners. The amendments pri-marily address radiation safety of CT systemsand require information on the imaging per-formance and radiation dose to be provided topurchasers (86). Image information from agiven CT system is proportional to the radiationdose. With a particular CT system, slower scantime results in both a higher radiation exposure,as well as a better image. In addition, it is gen-erally true that increased radiation providesmore image information.

The amount of radiation to which the patientis exposed is partially dependent on how theclinician using the scanner specifies certainvariables such as scan time. Usually this is deter-mined by the clinician’s preference. There issometimes a tendency to opt for more imageinformation at the expense of a higher dose ofradiation (129). Another problem found wasthat because of the complexity of CT equip-ment, there is potential (through suboptimal

performance) for obtaining a poor quality im-age even at a higher close of radiation (86).

The proposed amendments require informa-tion concerning the absorbed dose delivered byCT systems to a standard phantom (a test ob-ject) and the imaging performance correspond-ing to this dose, within the normal range of sys-tem operating conditions. This information willhelp to estimate the relationship between doseand imaging performance. However, the clini-cian will continue to be expected to exercise pro-fessional judgment in selecting conditions of op-eration of CT scanners.

It is difficult to summarize available informa-tion on radiation dose from CT scanning. Max-imum doses from a number of systems andunder a number of conditions of operation wererecently examined and were found to rangefrom less than 0.5 rad to almost 10 rad for asingle scan (158). As noted by the investigators,however, “All of the systems are capable ofalternate conditions of operation which willresult in different doses than those reportedhere, many of them significantly larger. ”

DIFFUSION AND UTILIZATION

Diffusion

This area of policy has been the site of thegreatest controversy over CT scanning duringthe last several years. Contention has sur-rounded the health planning law, 5 certificate-of-need (CON) review mandated in that law, andthe National Guidelines for Health Planning. b

Although it is difficult to say which, if any, ofthese have had an impact on the rate of diffusionof scanners, and/or the current aggregate sup-ply of CT scanners, the most heated debate hasfocused on the standards pertaining to CT thatare set in the national guidelines.7

— —5Health Pla?lning and Ri’sources Development Act of 1 9 7 4

(Public Law 93-641 ).‘National Guidelines for Health Planning, Federal Register,

Mar. 28, 1978, p. 13040.‘As noted in OTA’s 1978 rt~port (129), section 1122 of the Social

Security Act gave many States the authority to review capital ex-penditures over $100,000. This mechanism was available begin-ning in 1972 and was used by a number of States early in CT scan-ner diffusion. However, 1122 reviews are being replaced by theCON process. Now 36 States have CON laws.

The language of the Medical Devices Amend-ments of 1976 specifies assurance of effective-ness, but this is apparently used as a synonymfor efficacy. FDA approaches efficacy from arather technical standpoint. It interprets itscharge as one of assuring that the products soldin the marketplace are safe and technicallycapable of their professed abilities. It does notinterpret or perceive its purpose as being one ofdetermining how, and under what conditions,those products are to be applied by practi-tioners. FDA has supported research related toefficacy and safety of CT scanners. A survey ofthe system performance of CT scanning inselected U.S. hospitals has produced data fordeveloping dosimetry standards and technicalspecifications of scanner performance (157,1.59).FDA has also awarded a contract that will eval-uate utilization of CT head and body scanners.Survey items on its impact on diagnosis andtherapy relating to management and patientoutcome have been included to examine clini-cian perspective and motivation (30).

The National Guidelines for Health Planninghave been controversial since their inception,following the enactment of the Health Planningand Resources Development Act of 1974 (117).The guidelines established standards for 11 tech-nologies which were to be used by local healthsystems agencies (HSAs) and State health plan-ning agencies in reviewing and approving ap-plications for capital expenditures by hospitalsof greater than $l50,000. Published in Septem-ber of 1977,9 the first public request for com-ment elicited more than 50,000 responses, mostof which protested the proposed standards.Several months of deliberations ensued. A re-vised set of guidelines was issued in January o f

“The use of th]s term is not intended to be negative. The Impact(d the discussions and documents described below to promotelearning and mutual understanding ot the Issues on the part of allparties must be recognized and valued.

“National Guidelines for Health Plannlng, Advance Notice ofProposed Rulemaklng, Federal Regi.;ter, Sept. 23, 1977, VC1. 42, p.48502.

1978, and a set of standards for nine technol-ogies became official in March 1978. 11

The three standards set forth in the section ofthe National Guidelines for Health Planningpertaining to CT scanners (see app. E) are asfollows:

1.

2.

3.

A computed tomographic scanner (head andbody) should operate at a minimum of 2,500medically necessary patient procedures peryear, for the second year of its operationand thereafter.There should be no additional scanners ap-proved unless each existing scanner in thehealth service area is performing at a rategreater than 2,500 medically necessary pa-tient procedures per year.There should be no additional scanners ap-proved unless the operators of the proposedequipment will set in place data collectionand utilization review systems.

The current round of debate was instigated bya request for comment and recommendationsconcerning the existing guidelines for CT scan-ners. In issuing the final rules, DHHS had madeit clear that the newly established standardswould continue to be open to discussion andsuggestions for change. In keeping with its com-mitment, on March 23, 1979, almost 1 year afterthe guidelines had become effective, the Depart-ment issued a public call for comment on thestandards for CT scanners. 12 The notice wasprompted by recognition of the fact that becauseCT scanning is a rapidly changing field, new de-velopments, experience, and data may haveemerged since publication of the original stand-ards just a year earlier that might provide abasis for altering that standard.

Comments and suggestions received in re-sponse to the notice have been considerable andreflect the divergent opinions of various interestgroups. Among the suggestions have been in-creasing or lowering the number of patient pro-cedures required; developing a population-based standard for determining need; use of a

‘“h’at]onal Guidelines for Health Planning, Proposed Rule-maklng, Federal Register, Jan. 20, 1978, VO]. 43, p, 3056.

] ‘See footnote 6, p. 32.I l~otlce 0[ Request for Comment on the CT Scanner GuidelmeS

of the National Health Planning Gu]dellnes, Federal Register Mar.23, 1979, VOt. 44, p. 17760.

weighting formula; further specification of cir-cumstances for adjustments; no change in theexisting standard; elimination of the quan-titative target; and elimination of the standardfrom the national guidelines (72,98,99,100,108,109,135,168).

The request for comment renewed a vigorousand intense debate over the CT standards(52,112). Responses focused on the question ofwhether there is sufficient evidence to suggestthe need for changing the single quantitativestandard of 2,500 patient procedures per scan-ner per year. In general, the response from pro-viders and private associations (including man-ufacturers) has been that the target levels areunrealistically high. In support of this position,new evidence from a national survey of CTcapacity sponsored by the National ElectricalManufacturers’ Association (NEMA) was pre-sented which found that 61 percent of the 441 in-stallations surveyed could not meet the existingstandards (82). The response from planningagencies, on the other hand, has been the op-posite, i.e., that the target levels may be toolow, but in any case are not unreasonable (97).The experiences of HSAs that had establishedstandards higher than the 2,500 scans per yearwere brought to bear on their case.

Several months of deliberations ensued over awide range of options for change suggested inpublic comments (36). On September 13, 1979,DHHS proposed changes that would provide in-creased flexibility in the standard to take intoaccount the proportion of head and body scansand double studies13 performed, and would en-sure that access to necessary CT services ismaintained (139). Commitment to further studyof alternative weighting approaches was alsorecommended. However, no change in thequantitative target of 2,500 procedures was rec-ommended at that time.

Considerable support for the incorporation ofa weighting scheme in the guidelines had been

I JThe term “double study” refers to a series of two CT examinat-ions consisting of an unenhanced study, followed by an enhancedstudy. An enhanced study is one in which [one of several) contrastagents is administered to the patient prior to the examination, theobjective being to obtain a clearer image of an abnormality. Theimprovement in diagnostic information resulting from this pro-cedure has been debated (129).

expressed in the comments; no fewer than 33different weighting formulas and sliding scaleshad been suggested (8). The HECT formula(Head Equivalent Computed TomographyUnit), based on results of the capacity surveyand proposed by NEMA, received the mostvigorous and consistent promotion (35,171).While the concept of a weighting approachfound widespread support, the lack of consen-sus on the specifics argued that mandating theuse of any one approach would be premature.As a next step, DHHS initiated efforts to eval-uate alternative weighting approaches inselected areas (111).

With respect to the addition of provisions forincreased flexibility, the reaction to the pro-posed changes was generally supportive bothfrom planners, and from providers and privateassociations. However, while planners con-curred with the need for further study ofweighting approaches, NEMA and ACR pro-tested this recommendation, arguing that therewas already sufficient evidence on which to basea weighting system. The recommendations forincreased flexibility prepared by the Health Re-sources Administration (HRA) and approved bythe Subcommittee on National Guidelines,Goals, Standards, and Priorities of the NationalCouncil for Health Planning were sent to healthplanning agencies by DHHS, In November1980, HRA was in the process of preparingrevised standards that would incorporate aweighting formula. Such a standard must bepublished for comment in the Federal Registerfirst, but it could be functioning by some time in1981.

Other developments regarding the Federalregulation of diffusion of CT scanners tookplace on April 25, 1979. ’4 On that date, BHP ofHRA issued interim regulations regarding re-views of proposed capital expenditures for CTservices under the capital expenditure reviewprogram of section 1122 of the Social Security

terim Regulations, Federal ii egister, Apr. 25, 1979, vol. 44, p.24428.

I ~ ~oc-ltl~ S[,(-lJ rl tv A ~~t~~~ldl~l~>~l ts of 1972 ( I]ubllc Law 92-603), sec.

221, Llmltatl{]n on Federal Partlclpatlon for Capital Expenditures,1972.

view program of the Health Planning and Re-sources Development Act (see app. I). Expan-sion of the 1122 review authorities was broughtabout by concern on behalf of DHHS over theappearance of head scanners in 1978 that werebeing sold at prices well below the thresholdfigure for review of $100,000 (74). The reg-ulations were changed to cover any CT scannerunder a “change in service” review trigger. Theother target of amendments to both CON and1122 regulations was the growing market formobile scanners: According to OTA data, thenumber of mobile scanners doubled (from 7 to14) during 1978. The potential market for mo-bile scanners appeared to be great because CONregulations did not yet cover mobile units. Fur-ther, the anticipated change in medicare reim-bursement policy to cover scans performed onmobile units was expected to make the purchaseof the units even more attractive and to increasesales significantly. Consequently, it was felt thatmobile scanners should be subject to the reviewprocess (74).

There was limited public reaction to the April25th issuance, but as was the case during thecontroversy over the guidelines discussedabove, there was protest from individual pro-viders, provider associations, and manufac-turers of CT equipment, and support frommany local and State health planning agencies(74). Objections to the additional restrictions onthe purchase of CT equipment focused on thebelief that the regulations were yet another ex-ample of overregulation of the CT scanner asthe “scapegoat’ -unfairly singled out whenother hospital equipment more costly than CTscanners was not subject to review.

The changes in the regulation of capital ex-penditures under 1122 also specified the reviewof proposed changes in CT “services” (in con-trast to CT equipment), the implication beingthat replacing a dedicated head scanner with abody scanner and/or upgrading existing equip-ment is now subject to review. Under CON,such changes have always been regarded as newservices and have therefore been subject toreview. One outcome of the “rush” for scannersin 1975 was that scanners purchased at that timehave since been outmoded by total body scan-

ners with markedly improved scan time and im-age resolution. The updating and replacementof CT units within these health facilities hasbecome an issue of great concern for providersas new generations of scanners have becomeavailable.

Taken together, the new rules promulgatedmight be interpreted as concrete evidence of theincreasing emphasis placed by DHHS on the ob-jective of cost containment—possibly at the ex-pense of access and quality of care concerns.

The final change occurring in 1979 with thepotential for affecting the diffusion and, moreimportantly, the distribution of CT scannerswas the enactment of new amendments to theHealth Planning and Resources DevelopmentAct of 1974. Those amendments were signedinto law in September 1979.16 Under the provi-sions of the 1974 law, all major capita I expendi-tures by physicians for out-of-hospital settingswere exempt from CON review (37). OTA, inits CT report of 1978 (129), cited this exemptionas one of the greatest weaknesses in the originalplanning legislation. At that time, OTA pro-posed expanding regulations to cover all pur-chases of major medical equipment regardless ofsetting or ownership (see app. A).

The 1979 amendments only partially addressthis weakness. The health planning law now re-quires State review and approval of equipmentoutside hospitals, regardless of ownership orphysical location, if the equipment is to be usedto provide services for hospital inpatients. Thisamendment represents a compromise resolutionfrom the 1978 Senate bill, which had requiredcertification for all major medical equipmentpurchases irrespective of ownership or setting.It is aimed at the loophole in the previous lawwhereby physicians could make a private pur-chase of a scanner for a hospital that mighteither have applied and failed to win approvalor wished to avoid the CON process entirely,and then could locate that scanner in thehospital setting.

While the above amendment extended CONreview beyond purchases by health facilities

(i. e., hospitals), another amendment to the lawresulted in the exemption of certain health main-tenance organizations from the CON reviewand approval process.

Finally, the new law includes a preemptionprovision barring States from passing CONlaws that are more stringent than the Federalstatute after September 30, 1982. At last report,seven States had broader certification require-ments than Federal law stipulates (130).

Overall, the new health planning law addedlimited new regulatory authority, and it post-poned the date (January 1980, set in the 1974law) for a pending cutoff of certain Federalfunds to States that had not yet enacted man-dated CON programs by that date (71). Cur-rently, only 36 States have enacted CON laws(73,161).

The major gap in the health planning law re-mains, and what was intended to partially closean existing loophole in the law affecting the dif-fusion of scanners may have a perverse effect onthe distribution of scanners. Under the old law,CT scanners owned by physicians but operatedin hospitals may have skirted CON review, butthey were at least more accessible to the com-munity in these settings than in private offices.Hospitals have always had a more difficult timepurchasing scanners than private physicianshave. The new amendments, leaving them withone less option, however, place hospitals at aneven greater disadvantage.

Thus, the price exacted by curtailing the dif-fusion of scanners (i.e., the aggregate number ofscanners) may possibly be increased institu-tional maldistribution of scanners: The law nowfavors not only private purchase, but private lo-cation of scanners as well. This is another exam-ple of the preeminence of the cost-containmentobjective—possibly at the expense of access andquality of care concerns—found in Federal poli-cies toward the diffusion of CT scanners. It islittle wonder that the debate over the guidelinesis long and loud and hotly argued by those par-ties that are subject to them.

To summarize, the emphasis on cost-contain-ment objectives may be to omit other importantconsiderations such as access, medical effec-

tiveness, equity of distribution, and safety, aswell as other (besides capital) costs. This is notto suggest that the emphasis on cost contain-ment in the case of CT is unwarranted or that itnecessarily should be lessened, but is to suggestthat these other considerations should not besacrificed (either unwittingly or intentionally) inrestricting the deployment of scanners in thename of cost containment. Such policies maystrike particularly at the poor and underpriv-ileged (14). The tradeoffs between containingcosts on the one hand and assuring access andquality on the other should be made explicit,and a better balance struck between them. Inkeeping with the statutory mandate of thehealth planning program, it is critical to rec-ognize a much broader set of indicators thancost moderation in estimating the impact ofplanning and regulatory activities on the de-ployment of CT scanners.

Utilization

The history of coverage of CT scanning bypublicly funded third-party payers has been oneof increasing expansion, but it also is the first in-stance of a policy decision by HCFA to withholdreimbursement payments for a particular newprocedure pending evidence of efficacy (184).17

Eligibility for reimbursement of CT scansthrough the medicare program administered byHCFA has always been restricted by the typeand manufacturers of scanners used, and to a setof conditions deemed appropriate for use. Scansof the head, when performed on an EMI, Ltd.,head scanner, have been reimbursed since Sep-tember 1976 (103, 129).

——-‘“Pr}vate third-party payers have exhibited similar kinds of pol-

icy declslons with respect to CT that have also been precedents.Blue Cross and Blue Shield, (or example, has kept a list of pro-cedures and services (such as gastric freezing) that are widelyagreed to have no medical benefit, and for which they do not reim-burse. Blue Cross and Blue Shield also withheld payments for CTscans for some time, and it was that insurance company who re-quested the study on CT scal\ning that resulted in the first pub-lished concensus on indications for CT scanning (116). The in-fluence of the private sector on the acceptance of CT scanning inmedical practice is, therefore, -ecognized as being significant but isnot the focus of this discussion. The potential for leverage on dif-fusion and practice patterns tk rough private sector health insurerswarrants further investigation

Scans of the body, however, have been reim-bursed only since August 1978 (103). Under the“reasonable and necessary” clause of the SocialSecurity Act authorizing medicare payments(118), HCFA alread y had a mechanism fordenying payment for clearly antiquated pro-cedures. Based on a broad consensus that theprocedures were not useful, rather long lists ofsuch procedures were sent to medicare in-termediaries. Using this same clause, medicaredenied payment for body scans for almost 2years, pending study and recommendation bythe now defunct OHPA in OASH for reimburse-ment of certain indicated body scans (184). InJanuary 1978, OHPA made its determination(107). Eight months later, medicare began reim-bursing for certain body scans in addition tohead scans, based on detailed medical indica-tions for scanning.

Until April 1979, reimbursement for bothhead and body scans was limited to scanners in-stalled in a fixed location. But again, based onthe findings of a 15-month study carried out byOHPA, that Office recommended in June 1978that scans done on mobile scanners also be reim-bursed (105). Fourteen months later, coveragewas extended to scans done on mobile units(104).

Increasingly, the areas of reimbursement pol-icy and planning are being tied together. For ex-ample, medicare instructed its intermediaries in1979 to pay for scans from mobile scanners onlyif they have been approved by CON review(104). The regulations discussed above regard-ing reviews of proposed capital expendituresunder section 1122 of the Social Security Actalso state that denial of reimbursement underthe medicare, medicaid, and maternal and childhealth programs may be the penalty for capitalexpenditures that fail to conform with thereview plans, standards, and criteria.

Other, more subtle disincentives concern lev-els of payment. In August 1978, HCFA in-structed its carriers by letter (intermediary letterNo. 78-38) that services on CON-approvedscanners would be reimbursed at cost, whileservices on scanners without approval (e. g.,those in private physicians’ offices or thoselocated in hospitals but owned by physicians)

would be subject to ceilings. Mobile scannersnot owned by hospitals would also be subject toceilings. These changes were in part an attemptto counter reimbursement incentives toward thepurchase and use of scanners outside the plan-ning review and approval process (106). How-ever, in March 1980, in a case in South Carolina(Starns v. Harris), a U.S. District judge enjoinedHCFA from continuing the policy based on itshaving been promulgated without due process.Rather than appeal, HCFA announced its inten-tion to reissue the policy and make it applicableto other expensive technologies as well.18 Theproposed rule should be published in the FederalRegister within a year.

Besides being linked to planning policies, re-imbursement policy is also being increasinglytied to evaluation policies. HCFA now has ac-cess to an institutionalized resource in the newlymandated NCHCT and its functions to which itmay direct reimbursement inquiries regardingefficacy of medical technologies and their ap-plications (133). OTA’s 1978 report (129) previ-ously proposed that rates of reimbursement bebased on efficient use of technologies and thatthe payment system be fundamentally restruc-tured to encourage providers to perform and useservices efficiently (see app. A). To the extentthat NCHCT can develop cost-effectiveness in-formation, HCFA will be better able to translateit into a structure that might promote cost-effective physician behavior. Whether the infor-mation will be developed and, if so, whether itcan be translated into effect through reimburse-ment policies remains to be seen. A recent OTAassessment (127) examined some of the dif-ficulties of applying cost-effectiveness tech-niques in reimbursement. In addition, there is apossible ethical question involved in with-holding a service or procedure on the basis ofthe question, is it worth its cost? rather than onthe question, does it confer a health benefit? Theuse of such a criterion in providing services foronly that part of the population receiving pub-licly financed health care has obvious ethicalramifications that might cast doubt on the de-sirability of reimbursement policy based on it.—

“Medicare Program; Reasonable Charge Llmitatlons, Fedcru/Register, May 29, 1980, VO]. 45, p. 36100.

Finally, one of the major expressions of Fed-eral policy toward the use of medical services,including CT scanning, is the PSRO program es-tablished by law in 1972.19 PSROs are separateand independent organizations covering almost200 areas of the country. Each PSRO must besubstantially representative of all practicingphysicians in an area. The program operates bysetting standards and criteria for the desiredlevel and quality of medical services and byevaluating against these standards the servicesactually provided. This process is designed toensure that payment will be made only whenservices are medically necessary.

OTA’s 1978 report on CT scanners (129) de-scribed the PSRO program in detail, and thatmaterial is not repeated here. The only majorchange that has occurred since 1978 that couldaffect CT scanning is that the national PSROprogram distributed draft screening criteria forbody and head CT scans on February 22, 1979(160). These criteria, which were developedby the American Association of ProfessionalStandards Review Organizations, reflect thelack of well-validated information on efficacyand appropriate use of CT scans (see app. D).The body criteria are taken virtually word forword from the IOM report of April 1977 (116).In July 1979, SCBT published a list of indica-tions intended to “clarify, update, and augmentthe indications published in the April 1977 pol-icy statement of the Institute of Medicine” (164).Thus, by the time the PSRO draft guidelineswere beginning to be applied, the body criteriawere out of date, according to the most expertgroup dealing with the subject. (The NationalProfessional Standards Review Council, recog-nizing this problem, suggested to potential usersthat the criteria should be reevaluated at leastevery 6 months and updated if necessary. ) Thisis not to judge the validity of the recommenda-tions themselves, since they were based largelyon clinical experience, and not on well-designedstudies.

By October 4, 1979, eight PSROs had com-pleted medical care evaluation studies on CTscanning (188). Four others were carrying out or

‘* SC>cIal .+cur~~y An~c)~dmLIHts of 1’472 (Publ]c Law 92-b03 ), sec.301.

38 . Pollicy lmplications of the Computed Tomography (CT) Scanner: An Update

planning CT scan review at that time. On May28, 1980, OTA staff visited one of the PSROsthat was studying CT head scanning. The draftreview criteria had been used by that PSRO toproduce a list of 21 criteria justifying CT headscans, arrayed in order of importance (see app.D, exhibit 4). The first eight criteria related toevaluation of suspected or previously knowndiagnoses, the next eight related to abnormalphysical findings, and the next five related tosymptoms noted on a medical record when nosuspected diagnosis was listed. Only 8 of 427scans the PSRO reviewed did not meet thesecriteria. Of these 427, however, 58.3 percentyielded negative results. The PSRO concludedthat CT head scans were used judiciously in thatregion.

There are numerous reasons that this PSRO’sconclusion cannot be supported. One is that theindications written in medical records as indica-tions for procedures are known to lack validity.Secondly, the indications are broad and generalenough so that almost any patient would qual-ify (one of the criteria is simply “headache”).But perhaps most important is that the criteriahave not been firmly connected to evidence ofefficacy. The truth is that it is not known in thatPSRO area, or in any other, whether the CThead scans are done judiciously. What can beobserved is that PSROs deal primarily with ex-treme cases, and thus cannot be expected tohave a great impact on the utilization of any

procedure that is accepted by the medical com-munity. In the absence of scientific efficacy in-formation, existing practice may become thestandard of practice-whether or not it is

“appropriate. ” Established patterns have thehabit of lingering in medical practice even aftersuch time as efficacy information becomesavailable (60).

An interesting pilot project is attempting touse evidence of efficacy of an X-ray procedure,pelvimetry, to significantly reduce the use of X-rays. FDA’s BRH developed a consensus policystatement concerning the lack of efficacy of X-ray pelvimetry. The statement was endorsed byACR and the American College of Obstetricsand Gynecology. In the study project, PSROsintend to change the practice norm by movingto eliminate these X-ray procedures for purposeswhere they are proved not to be efficacious.This project demonstrates the promise ofPSROs, and with the development of better in-formation on efficacy, can perhaps become thenorm rather than the exception.

In summary, utilization policies toward CTscanning are still very much in the process ofchange. HCFA perceives that it has a role incontrolling technologies such as CT scannersand will undoubtedly make further changes inits payment and review policies. Further regula-tion through these mechanisms seems inevit-able. In October 1980, HCFA had drafted pro-posed regulations (not yet available) that willdefine “reasonable and necessary, ” the criteriaspecified for payment for services in the medi-care law, 20 According to HCFA staff, the def-inition will include costs and broader social im-plications in addition to efficacy and safety.

4Summary and Conclusion;

4.Summary and Conclusions

The computed tomography (CT) scanner re-mains an instructive case study which illumi-nates both the process of innovation and Federalpolicies toward medical technologies. Althoughthe CT scanner is not in itself a major healthpolicy issue, it can be used to understand prob-lems in Federal policies.

The trends in diffusion of CT scanners havebeen the cause of much controversy. Federalpolicies have been cautiously developed to cur-tail the rapid diffusion of medical technologiessuch as the CT scanner. Critics have leveled ageneral charge that Government interference isinhibiting the process of innovation— one of thecritical signs of a robust, dynamic economy. Butit is difficult to ascertain whether the existenceof the Health Planning and Resources Develop-ment Act itself, the process of review by healthsystems agencies (HSAs) and approval by Statehealth planning and development agencies(SHPDAs) of applications for large capital ex-penditures by hospitals under certificate-of-need(CON) provisions of that Act or section 1122 ofthe Social Security Act, or the standards for CTscanning set in the National Guidelines forHealth Planning to assist HSAs and SHPDAs inthese functions have influenced the CT scannerdiffusion rate. Although one would expect thediffusion and distribution of scanners to berelated to Federal policies addressing them,there is really no good evidence available to in-dicate whether and to what extent any or all ofthese factors have influenced the diffusion of CTscanners.

Opposition to Federal policies concerning CTscanners has focused on the National Guidelinesfor Health Planning. For over a year now, thedebate between manufacturers, providers, plan-ners, and Federal authorities has honed in on thespecific standards set in those guidelines. Al-though the guidelines have the potential to re-strict diffusion and affect the distribution ofscanners, diffusion slowed before developmentof the guidelines (see figure 1 in ch. 2). The

standards set in the guidelines became effectiveonly in March 1978. It should also be noted thatthe standards are advisory rather than man-datory. Scanners installed during 1979 wereprobably ordered months earlier. The rigor withwhich the guidelines are applied by HSAs inreviewing CON applications and the extent towhich SHPDAs adhere to them in deliberatingapproval of CON applications are unknown.All these factors make the impact of theguidelines uncertain, It may be that the opposi-tion to the guidelines themselves and the debateover the specific standards for CT scanners inthem are based largely on the potential impactthat the guidelines may have in the future,rather than on any effect that has been wit-nessed in the recent past. In addition, the role ofthe Federal Government in regulating diffusionof medical technology is being questioned.

The impact of CON regulations on the diffu-sion and distribution of scanners is, in general,currently unknown. (One study that is investi-gating the correlation between the implementa-tion of State CON laws and the diffusion rate ofCT scanners on a State-by-State basis is under-w ay (21). ) The manufacturers of CT scannersbelieve that the guidelines and CON regulationshave had an impact. While requests for CTscanners are approved far more often than theyare disapproved (74), the effect of the healthplanning process of discouraging applicationsmust be considered as well. It may well be thatthe extremely high number of scanners sold in1975 was induced by the anticipation of the im-peding sanctions embodied in the upcominghealth planning regulations (15). It also seemslikely that a number of scanners were purchasedfor physicians’ offices because of delays in ob-taining permission for hospital scanners (15).Thus, the health planning program has partialresponsibility for the existing maldistribution ofCT scanners (see ch. 2). With the large numberof older scanners in place, the health planningprocess may impede appropriate replacementand upgrading as well.

41

This discussion suggests a partial alternativeexplanation regarding the slowdown of the dif-fusion rate that would be the logical outcome ofthe intense market activity of 1975. That is thatthe market for CT scanners may be beginning toreach its limits. It is supported by close inspec-tion of the data on distribution and diffusion ofCT scanners to May 1980, when there were1,471 operating scanners. Analysis of the insti-tutional placement of scanners indicates thatmore than 80 percent of all large hospitals, orthose over 500 beds, now have operationalscanners. Even in early 1979, 62 percent of hos-pitals over 300 beds and 46 percent of those over200 beds had scanners, A July 1979 presentationby the Technology Marketing Group, Ltd. (66),showed that of 2,250 hospitals with over 100beds that did not have a scanner, only 23 per-cent were considering purchase of a scanner,while only 22 percent of the 624 hospitals withscanners were considering purchase of an addi-tional one.

That the slowdown is more a natural phe-nomenon than a result of Government policymay be supported from a theoretical point ofview as well as an empirical one. The cumula-tive diffusion curve of the installation of CTscanners in the United States is a textbook il-lustration of a theoretical innovation diffusionprocess (see ch. 2). The logistic curve is a modelof that process that holds true of innovationssuch as automobiles, televisions, and automaticwashing machines. Following the introductionof an innovation to the market, a certain per-centage of interested parties will purchase it.Through time, this percentage will accumulateuntil virtually all of those who are expected topurchase will have done so. As more parties ofthis target group make their purchase, a smallerpercentage are left to make theirs, so that overtime the rate of diffusion must slow down andeventually level off. This phenomenon may justbe manifesting itself in the 1978, 1979, and early1980 data on CT scanner installations.

In addition to the question of whether or notFederal policies embodied in health planningprograms have had an impact on the number ofscanners, the question of whether they have hadan impact on the distribution of scanners also

remains at issue. Questions persist about the ef-fectiveness of health planning laws with respectto the optimal location of scanners. Provisionsof the health planning laws do not require CONapproval of large expenditures made by privatephysicians. Currently, about 19 percent of thetotal number of scanners in this country arelocated in nonhospital settings. In particular,large urban hospitals typically serving a pre-dominantly indigent clientele and large VeteransAdministration hospitals lack scanners. Geo-graphic maldistribution is also evident. Someurban areas have exceptionally high scanner-to-population ratios (the District of Columbia, forexample, has 15.9 scanners per million popula-tion, and the Los Angeles area has 14.0 scannersper million); but some rural areas have no ac-cessible CT scanner. Health planners have lim-ited tools to assure placement in appropriatesites—their powers are largely negative. Thismay be the greatest problem with the healthplanning program.

Although Federal health planning programsmay aim to curtail the diffusion of CT scanners,the stance assumed by the Federal Governmentin its policies toward other stages of develop-ment and use of scanners has tended to fosterdiffusion and widespread use. As noted earlier,Federal policies address all of the four stages inthe development and use of medical technol-ogies: R&D, demonstration of efficacy and safe-ty, diffusion, and widespread utilization. In par-ticular, the Federal Government has tradition-ally been a generous supporter of biomedicalR&D. The real boon to diffusion and use of CT,however, has been in the Federal policy area offinancing. Through its reimbursement policies,the Federal Government continues to assume analmost open-ended commitment to pay for CTscans. This posture has doubtless played an im-portant role in the rapid acceptance of CT scan-ners and scanning in medical practice, therebyinfluencing the rate of diffusion and the aggre-ate supply of scanners.

Thus, a number of factors have affected CTscanner diffusion, not least of which is the revo-lutionary nature of the technology itself and itspotential for improving diagnosis. The relative

impact of each of these factors will probably the problems identified in OTA’s 1978 reportnever be fully understood. (129) remain largely unaddressed. For this

reason, the policy alternatives of the 1978 reportAlthough some changes have been made in are reprinted in appendix A. They still seem to

policies regarding CT scanners since 1978, the have relevance to those interested in improvingunderlying programs remain little changed, and Federal policies toward medical technologies.

ell. 4- S1llnnlllr.valld('(1I1dIlSi'11L~ • 43

Appendixes

Appendix A.— Policy Alternatives(Reprinted From Policy Implications of the Computed Tomography (CT)

Scanner, Office of Technology Assessment, August 1978)

The computed tomography (CT) scanner is a newdiagnostic device that represents an important ad-vance in medical detection. Studies show that CTscanners perform reliably and provide accurate diag-noses of abnormalities in the head and abdomen. Asa relatively safe and painless procedure, CT scanningcan replace several less safe and more painful tech-nologies, such as pneumoencephalography. CT scan-ning has been readily accepted by the medical pro-fession, and its use is expanding rapidly. To the ex-tent that a fundamental problem with CT scanningexists, it lies not in the existence of the technology,but in its appropriate use.

Although this study focuses on CT scanners, itsfindings are applicable to the general problem ofappropriate use of diagnostic medical technologies.Appropriate use includes considerations of safety, ef-ficacy, and cost. Overuse of a technology may leadto both excessive expenditures and unwarranted riskto patients; underuse may result in delayed detectionor prolongation of medical problems. In either case,the study demonstrates basic policy problems relatedto the appropriate use of medical technologies.

Use of a diagnostic medical technology such as aCT scanner depends on many factors: Some increaseand others restrict use. A principal and obvious fac-tor is the desire of physicians to provide good carefor their patients. Attempts to identify medical prob-lems and to refine diagnoses lead physicians to usethe technologies available to them. Medical educa-tion also predisposes physicians to liberal use of diag-nostic technologies by emphasizing thoroughnessrather than discrimination and concern for costs. Thecurrent medical malpractice situation further en-courages the use of diagnostic tests to avoid error. Insome instances, patients themselves request thatphysicians perform diagnostic tests. Although theseare important issues, this report has not addressedmedical education, malpractice, and patient demand.Rather it concentrated on available information,governmental regulation, and financing.

After their formal training, physicians continue toreceive information about medical technologies fromscientific meetings, professional publications, col-leagues, manufacturers’ representatives, and theirown clinical experience. Two Federal agencies, theFood and Drug Administration (FDA) and the Na-tional Institutes of HeaIth (NIH), develop and dis-seminate such information. By law, manufacturers of

drugs and medical devices must submit to FDA datathat supports claims made in labeling. NIH conductsevaluations of certain medical technologies andmakes the results available to the public. However,as illustrated by this study, no single Federal or pri-vate policy establishes a formal, systematic processto develop needed information about medical tech-nologies. Nor is there a clearly defined mechanismfor disseminating what is known to all appropriateparties.

Without such information, physicians appear totest new technologies using a variety of methods todevelop a sense of their worth empirically. Un-fortunately, these methods are often not designed toyield statistically reliable information. This informalexperimentation can both retard the early applicationof valuable technologies and advance the use of ques-tionable ones. Without valid information obtainedfrom well-designed studies, physicians face a verydifficult task in deciding on the appropriate use ofnew technologies.

Prevailing methods of financing medical care pro-vide incentives for additional use of technologies,regardless of their marginal value. Health insuranceprograms have continued previously existing fee-for-service payment of physicians; performance of addi-tional tests thereby generates additional revenue forthe physicians. Hospitals are reimbursed on the basisof their costs or charges. These methods at the leastfacilitate and at the most stimulate providers toprescribe additional use. Under such a system, pro-viders have little incentive to weigh the benefits andcosts of additional tests.

The regulatory framework created by FDA, theProfessional Standards Review Organizations(PSROs), and capital expenditure laws also affectsthe use of medical technologies, in a restrictive sense.FDA requires proof of safety and effectiveness beforedrugs and devices may be marketed. The PSRO pro-gram was designed to establish norms and standardsfor hospital utilization and medical care providedunder medicare and medicaid. And review of pro-posed capital expenditures is aimed at avoiding un-necessary duplication of facilities and promotingtheir efficient use. Unlike many of the other factorsaffecting technologies, these programs may restricttheir use. The PSRO program and capital expend-iture review were created in part to counter incen-tives for greater use, especially from financingmethods.

47

48 ● Policy Implications of the Computed Tomography (CT) Scanner: An Update

The following sections present alternatives thatmight improve the use of medical technologies suchas CT scanners. The alternatives are presented inthree sections, each addressing a specific category ofgovernmental policy: Section 1 focuses on develop-ing and disseminating information on efficacy andsafety; section 2 on regulatory policies; and section 3on financing. The alternatives in each of these sec-tions illustrate, but do not exhaust, possible options.Nor are they necessarily mutually exclusive. Eachalternative should be measured against the con-tinuance of current policies and their consequences aswell as against the consequences of the alternativeitself. These alternatives represent broad guidelinesfor policy. As such, they do not consider in depth themore technical aspects of implementation, such asthe mechanisms for evaluating efficacy, specificcriteria for utilization review, methods of cost ac-counting, or details of ratesetting.

1. Information on Efficacy and Safety

Many decisions concerning the use of a medicaltechnology depend—directly or indirectly—on anassessment of its efficacy and safety. Much of theavailable information on efficacy and safety is notderived from well-designed controlled clinical trials,epidemiological studies, or analyses of clinical ex-perience. Instead, informal judgments evolve, judg-ments based primarily on the experience and per-ceptions of individual physicians. Judgments of thistype, when they do not accurately reflect the efficacyand safety of a technology, may contribute sub-stantially to inappropriate use.

The development of information on efficacy andsafety involves identifying the technologies to bestudied, conducting the appropriate evaluations, andsynthesizing the results of those evaluations and rele-vant clinical experience. The synthesized informationmay then be disseminated to the individuals and or-ganizations most in need of guidance. Although sim-ple to delineate on paper, this process of synthesisand dissemination can be complex and difficult toimplement.

This section presents two policy options designedto address the needs of medical care decisionmakersfor efficacy and safety information. The first con-cerns the development and dissemination of the in-formation. The second requires the type of synthesisthat analyzes information to produce formal policy .judgments about a technology’s efficacy and safety.This section and the alternatives presented in it areconcerned only with developing and disseminatinginformation.

Together, the two alternatives, if adopted, wouldincrease the amount of information available tophysicians in their use of medical technologies Theinformation would also be helpful to planners,regulators, and public policy decisionmakers. As ex-plained in alternative 3 of the following section, FDAalready requires the development of information andmakes certain policy judgments about the safety andefficacy of medical technologies. The alternatives inthis section would substantially enlarge these existingprocesses,

As discussed in chapter 3, information about ef-ficacy is used or could be used by many Federal pro-grams, as well as by providers of medical care. Deci-sions and policies based on efficacy may now beinconsistent as each user defines efficacy in its ownway. As described in chapter 3, only FDA has a for-mal definition of efficacy at present, and that defini-tion merely ensures that the evidence substantiatesthe claims of the manufacturers. But FDA’s decisionson efficacy and safety are of limited value to healthplanning agencies, PSROs, and reimbursementprograms.

A general definition of efficacy could be developedfor all types of medical technologies—preventive,therapeutic, and diagnostic. No medical technologyis beneficial in all circumstances, and some tech-nologies can be extremely beneficial only if used invery limited situations. Therefore, the efficacy of aparticular technology must be related to a definedpopulation, a given medical problem, and particularconditions of use. A complete specification of ef-ficacy encompasses all three of these factors. ’

Alternative 1: Establish a formal process toidentify medical technologies that should beassessed for efficacy and safety; conduct thenecessary evaluations; synthesize the resultsfrom the evaluations and from relevant clinicalexperience; and disseminate the resulting in-formation to appropriate parties.Except for new drugs and, potentially, new med-

ical devices, the Federal Government’s identificationof technologies warranting study occurs in an ad hocmanner. Often, decisions to evaluate a technologydepend on the curiosity of investigators or Federalprogram administrators. Few efforts have been madeto coordinate the selection of technologies to bestudied with the informational needs of relevant gov-ernmental agencies and private groups.

IEfflcacy IS def}ned as the potent]al ber eflt to Individuals In a definedpopu]atlon from a med]cal technology applied for a given medical p~oblemunder Ideal conditions of use. These Ideal condltlcms may be approa( hed Inresearch se t t ings , but are unllkely m al erage practice. Efficacy, then,represents an outer IImlt to benefit.

Appendix A. —Policy Alternatives ● 49

No existing Federal procedure systematically iden-tifies those technologies that are most in need of in-vestigation. Indeed, no formal set of criteria has beendeveloped for establishing such priorities. The pri-vate sector identifies medical technologies to beassessed for efficacy and safety through an even moreinformal process. As described in chapter 6, how-ever, some efforts have been initiated by organi-zations such as the Federal Health Care FinancingAdministration and private Blue Cross-Blue Shield toidentify and develop information on possibly inef-ficacious or unsafe technologies.

Various Federal agencies currently have respon-sibility for conducting or funding studies on efficacyand safety, although in each case their mandate islimited and often ambiguous. The NIH effort is byfar the largest; that agency spent approximately $100million on more than 750 studies during fiscal year1975. The emphasis at NIH is on new technologies,rather than on those already diffused; z thus, existingtechnologies receive relatively little scrutiny. Similar-ly, drugs and biologics receive more attention thandevices or medical and surgical procedures.

No Federal policy focuses responsibility for the dis-semination of efficacy and safety information.Although NIH and FDA both disseminate substantialamounts of information, their efforts are hamperedby various factors. For example, NIH historicallylacks working relationships with many of the partiesin need of the information. Although FDA obtainsinformation on efficacy of drugs and devices frommanufacturers, most of that information is con-sidered to be proprietary and is not released in thatform by FDA to the public or to providers. In addi-tion, the information disseminated is often not in aform readily usable by parties in need.

This study of the CT scanner illustrates some of theconsequences of using the present informal assess-ment process, Although the CT scanner has been thesubject of much publicity since its introduction, fewwell-designed evaluations of its efficacy and safetyhave been conducted, Despite this dearth of informa-tion, CT scanning has been more fully evaluated thanmany other diagnostic technologies.

Instead of continuing the present informal assess-ment system, the process could be made explicit andformal as indicated by this alternative. The processcould be applied to both existing and new medicaltechnologies. With the implementation of an explicit,formal system, criteria could be developed forscreening the thousands of existing and future med-ical technologies to establish priorities for investiga-

‘Dttuslon ot a techn[)logy reters to the p r o c e s s of adopt]on from d e -velopment un t)] general acceptance

tion. These criteria could take into account factorsnow excluded or only minimally included in theprocess of assigning research priorities. Such factorsas needs of health planning agencies and third-partypayers and the level of expenditures for thetechnology could be included in the criteria to beestablished.

Also, under this alternative, an agency or agencieswould be given explicit responsibility for conductingstudies of efficacy and safety or ensuring that theyare conducted, for synthesizing information to ap-propriate parties, (Two bills before Congress, H.R.12584 and S. 2466, would create an office withinthe Department of Health, Education, and Welfare(DHEW) to evaluate medical technologies, ) Thedirect anticipated result of this alternative is the pro-duction of science-based information for use bymedical professionals, policy makers, Governmentagencies, and the public.

This alternative is not designed to change the cur-rent processes of introducing and using medical tech-nologies except to increase the amount of validatedinformation available. The present process allows abroad and varied experimentation process to occurwith new medical technologies. Through its proc-esses of careful human experimentation, the presentsystem also permits technologies to be used early intheir development. Controlled clinical trials, epi-demiological studies, and other forms of technologyevaluation are often lengthy activities. Thus, thedevelopment of information on efficacy and safetycan be a time-consuming process. Under this alter-native, diffusion and use of a medical technologywould not necessarily be postponed until the conclu-sion of the evaluation process.

Implementation of this alternative could be costly.Controlled clinical trials are expensive: An averagetrial funded by NIH costs more than $100,000 peryear, and those for surgical procedures or expensivetechnologies may be several times higher. Formaliz-ing activities under this alternative is likely to in-crease substantially the number of trials because thescreening and synthesizing processes would identifyproblems with technologies and gaps in efficacy andsafety information. A large number of medical tech-nologies might warrant careful examination, requir-ing complete reviews of available information and at-tention to clinical experience. The process outlinedwould make cooperative trials (such as many ofthose of the National Cancer Institute) more feasible,a development that could reduce the magnitude ofthe increase in the trials.

A distinction can be made between changing thetotal use of medical technologies and reducing inap-propriate use (e.g., of technologies that are under-

used or overused). This alternative makes the latterpossible, though it does not ensure it. Reduction inthe use of certain technologies, following evaluation,might be offset by increased use of other technol-ogies, some of which may themselves be unevalu-ated. The relative magnitude of these three factors—reducing use of overutilized technologies, increasinguse of underutilized ones, and the unpredictableshifting of utilization patterns from one technologyto another—will in part determine the effect of thisalternative on total use of medical technologies andon expenditures for medical care.

Alternative 2: As part of alternative 1,establish a formal process for making officialjudgments about the efficacy and safety ofmedical technologies.Under current law, FDA must determine the ef-

ficacy and safety of a drug or device before it can bemarketed. No Federal organization is responsible forofficially determining the efficacy and safety of med-ical and surgical procedures. At least two com-ponents of the Public Health Service (NIH and theOffice of Health Practice Assessment) are attemptingto develop formal systems to synthesize informationand arrive at decisions on particular medicaltechnologies.

The synthesis process of alternative 1 could takemany forms. It could collect and analyze existing in-formation, or it could attempt to identify gaps in ex-isting knowledge as a guide for further research.Under this second alternative, synthesis would in-volve collecting and analyzing available informationin order to produce official policy judgments aboutthe efficacy and safety of the technologies underexamination.

This alternative would establish a process wherebyrelevant information on a medical technology iscritically evaluated. The evaluation would result in ajudgment, or policy decision, as to a technology’s ef-ficacy and safety. This alternative would be in-tegrated with alternative 1. The judgments couldcontain detailed information on a wide range of in-dications for appropriate use of the technology.Thus, they could be broader than FDA’s currentdeterminations for marketing approval,

Providing official judgments to relevant in-dividuals and organizations would add to the in-formation available to them for making decisions.However, those individuals and groups would stillmake the final decisions. The judgments about ef-ficacy and safety might be issued as guidelines or asrecommendations. They would not be binding. Thissecond alternative would only produce information;it would not be a regulatory process.

Such official information might reduce the errorsin judgment that such individuals and organizationsmake. However, mistakes made by the group devel-oping the judgments, while perhaps fewer in number,would have broader ramifications because of theirofficial nature. Since mistakes are inevitable andjudgments of efficacy and safety can change as addi-tional information becomes available, this alter-native would require a substantial degree of flexibili-ty in operation. The process outlined in this alter-native and alternative 1 could be used initially for asmall number of technologies to test its feasibility.An evaluation of CT body scanning, for example,could produce judgments about the types of benefitslikely to result for certain kinds of patients andspecific medical conditions.

This second alternative would almost certainlyhave an effect on the current medical malpracticesituation. The existence of official, though volun-tary, statements as to the efficacy and safety of atechnology might become the standard for judgingwhether a provider properly used that technology.

The major controversy surrounding this alterna-tive would be determining the process that would beused to make such scientific judgments. Because suchjudgments could be used to decide whether a tech-nology is to be reimbursed and where it can beuseful, this alternative could become the focus ofconsiderable political and economic pressure. Carewould have to be taken to see that the process is bothtimely and scientifically appropriate.

2. Governmental Regulatory Policies

In an attempt to offset powerful incentives en-couraging the use of medical technologies, Congresshas established three regulatory programs: the FDA,the PSRO program, and capital expenditures review.FDA regulates the marketing of drugs and devices.Marketing requires prior FDA approval that the tech-nology is safe and effective, and advertising is limitedto the approved conditions. FDA does not haveauthority to restrict subsequent use by physicians orpatients. PSROs evaluate appropriateness of caregiven to medicare and medicaid patients. PSROsmay establish standards for the use of specific med-ical technologies, such as CT scanners, although fewsuch standards have yet been developed.

State certificate-of-need laws require prior ap-proval for capital expenditures greater than a certainamount, usually $100,000 to $150,000. Federal andmost State laws cover hospitals, but exclude privatephysicians’ offices. In general, capital expenditurelaws do not regulate use of facilities or equipment

Appendix A. —Policy Alternatives ● 51

once they are in place. The Social Security Act alsorestricts payment under medicare to services that arereasonable and necessary for diagnosis, treatment, orimproved functioning.

Inadequate information about efficacy and safetyhandicaps the effectiveness of these three programs.FDA obtains information about efficacy and safetyfrom manufacturers, but that information is limitedto certain uses of the drug or device. PSROs, reim-bursement agencies, and State and local planningagencies need information about the appropriate useof a technology—the population benefiting, the med-ical problems affected, and the conditions of useunder which the technology is safe and effective. Fur-ther information is required concerning the substitu-tion of a new technology for existing ones. Both thePSRO and the health planning programs are new andnot yet fully implemented. In addition, lack ofuniversal coverage facilitates circumvention of theseprograms.

This section includes alternatives concerning theuse of medical technologies, capital expenditure re-view, and medicare reimbursement. Alternatives 1and 2 from section 1 would facilitate alternative 3and would be necessary for alternative 4. Alternative3 would restrict the use of medical technologies tothose indications approved by FDA for marketingpurposes. Alternative 4 would link medicare reim-bursement to the information and judgments of alter-natives 1 and 2. And alternative 5 would expand theregulation of capital expenditures to include all pur-chases of medical equipment regardless of setting orownership.

Alternative 3: Authorize a Federal regulatoryagency, such as FDA, to restrict the use of med-ical technologies to the conditions of usespecified in the FDA-approved labeling.When FDA approves a drug or device for mar-

keting, it also approves the specific wording of theproduct’s labeling, i.e., the written information usedby the manufacturer to describe the product. Label-ing (which includes package inserts) lists medicalconditions (and possibly populations) for which thedrug or device is deemed to be safe and effective andwarns about possible side effects.

These “indications for use” are usually not ex-haustive. A manufacturer that has conducted pre-marketing clinical tests to evaluate safety and effec-tiveness for defined medical conditions and popula-tion groups could then seek marketing approval onlyfor those conditions. Thus, the FDA marketing ap-proval process might consider only a portion of thepossible indications or contraindications for a newdrug or device.

Use of drugs and devices by physicians and pa-tients, however, is not restricted to the approvedconditions. Although the manufacturer providesonly the approved information to physicians andother providers, this information is in effect merelyadvice. Nothing in the law prevents the use of drugsor devices for conditions other than those specified.(A bill before Congress, S. 2755, would restrictdistribution of drugs to particular providers. )

Uses of a technology for conditions other thanthose approved by FDA are not necessarily inef-ficacious. Conceivably, some potentially efficacioususes are not evaluated prior to initial marketing ap-proval by FDA. However, the absence of a particularuse from the list of approved uses implies that suffi-cient information is not available to determine thetechnology’s efficacy for that use.

Examples can be cited of beneficial uses that wereneither anticipated nor evaluated by the manufac-turer but were later adopted by practitioners. Use ofthe drug propranolol for treating hypertension (highblood pressure) is such an example. Other uneval-uated uses, however, have been shown to be medical-ly unjustified when investigated after the drug ordevice was marketed. For example, chloramphenicolhas often been used for upper respiratory infectionswhen equally effective and less toxic drugs wereavailable. The balance between positive and negativeeffects of unapproved uses of drugs and devices is dif-ficult to determine. One factor is clear—unapproveduses usually have not been verified by the rigorousclinical research that is necessary to gain FDA ap-proval.

Allowing physicians to use technologies for unap-proved uses has resulted in a de facto research or ex-perimentation process. Formal clinical investigationsof a new use must proceed under an FDA-monitoredInvestigational New Drug (IND) process for drugsand under a similar process for devices. Unapproveduse by physicians and patients could be consideredan unofficial clinical investigation. This result can beeither beneficial if a new efficacious use is found orharmful if the use is unsafe or ineffective. Also, asidefrom the technical questions of efficacy and safety,moral or human rights questions may be raised bythis unapproved application.

This third alternative would make FDA decisionsbinding on physicians. Drugs and devices could beused legally only in accordance with the indicationsfor use specified by FDA’s marketing approval.Other uses would be allowed only as part of an ap-proved IND or an investigational process for devices.The investigational process for unapproved uses, themechanics of which could be similar to the current

process, could replace the present practice of unap-proved use. A scientific process evaluted by FDA oranother agency charged with the task could addvalidated indications or contraindications to the ap-proved labeling for a drug or device. This alternativeis based on marketing approval, which is now limitedto drugs and devices; it would not cover medical andsurgical procedures.

The indications for use comprise one aspect of ef-ficacy and safety, as noted above. Therefore, thisthird alternative would be most effective if generallyaccepted and comprehensive definitions of efficacyand safety were developed. In addition, a publicationlisting the FDA-approved indications for use of allcovered technologies might be necessary to informphysicians who rely on these technologies.

The principal intention of this alternative is to im-prove the quality of medical care by ensuring moreappropriate use of medical technologies. Fewer pa-tients would then be subjected to unapproved andunscientific uses of technologies. Instead, medicaltechnologies would be more likely to be used in ac-cordance with valid scientific information.

A probable consequence of implementing thisalternative would be an increase in premarketingclinical investigation to determine appropriate in-dications for use, The number of such investigationswould depend on the proportion of potential usesthat had already been investigated.

This alternative could affect the timing of using atechnology for a new indication. Use of the technol-ogy for the new indication would not be permitteduntil the experimentation process had been com-pleted (although some use would obviously occur aspart of the experimentation process itself). However,once a use had been demonstrated to be efficaciousand safe, the manufacturer would be allowed toadvertise that use. This advertising promotion mightresult in diffusion of the new use to a larger numberof individuals in a shorter period of time than occursunder the present system. However, if no firm orother organization decided to conduct investigationsand seek approval for a particular condition of use,that potential use might go undetected.

The financial costs of this third alternative are notpredictable. Additional clinical trials would increasethe costs of bringing a technology to market. The netcost to manufacturers is not clear. They would bearthe costs of extra clinical trials, but might receiverevenue from addition sales if a new use gained ap-proval. A system of financing additional evaluationsof efficacy and safety could be developed, possiblythrough a combination of manufacturers, patients,and third-party payers. Expenditures for the use ofmany technologies might fall if third-party payers

and patients did not have to pay for unapproveduses, But expenditures on new uses might rise.

Adoption of this alternative would require a sys-tem for ensuring compliance. One can imagine veryelaborate enforcement measures requiring additionalpaperwork and specialized personnel that are notreadily available. A more simple approach wouldrely on the good faith of providers. A provider foundto be noncompliant would be penalized, but com-pliance would otherwise be assumed.

The practicality of this third alternative is ques-tionable. Although laws and regulations can man-date this alternative, their enforcement could becumbersome and expensive. Monitoring, let alonealtering, physicians’ use of medical devices and drugsis difficult. In addition, the cost of enforcement mightexceed the benefits. At a minimum, however, enact-ment of this alternative might increase providers’awareness of their legal liability in using technologiesfor unapproved uses and might lead them to operatewithin the approved investigational process. In fact,approved uses might serve as a basis for liability.

Alternative 4: Link medicare reimbursementto the information and judgments about a tech-nology’s efficacy and safety that would resultfrom alternatives 1 and 2.Medicare administrators have interpreted the pro-

vision of the Social Security Act limiting payment toreasonable and necessary services as allowing med-icare to withhold payment for experimental pro-cedures whose efficacy has not been determined. Itwas under this provision that medicare withheld pay-ment first for CT head scanning and then for CTbody scanning pending evaluation of efficacy. His-torically, medicare has denied reimbursement foroutmoded procedures rejected by the medical com-munity. But medicare’s action on CT scanning usedefficacy and safety criteria to make a more contro-versial decision. And overall medicare policy sup-ports strengthening the dependence of reimburse-ment on efficacy and safety. It is medicare’s policy torestrict reimbursement for drugs to conditions of useapproved by FDA. FDA’s evaluation of devicesunder the Medical Device Amendments of 1976 doesnot yet provide a sufficient basis for medicare action.For advice on procedures and devices, medicare con-tinues to rely mainly on the Office of Health PracticeAssessment of the Public Health Service.

Although medicare policy links reimbursement toefficacy and safety, major problems remain. As dis-cussed in section 1, information on the efficacy andsafety of devices and procedures is insufficient forreimbursement purposes. These deficiencies rangefrom inadequate clinical data through incompletesyntheses of existing information to the processes

Appendix A. —Policy Alternatives ● 53

used in making judgments. The task of evaluation ismuch beyond the present capability of the Office ofHealth Practice Assessment. Besides an inadequateinformation base, the Office has a small staff and noformal process for evaluating technologies. FDAlabeling provides more available and useful informa-tion on drugs.

This fourth alternative suggests linking medicare’sreimbursement for use of a technology to the infor-mation provided by alternative 1 and to thejudgments about efficacy and safety reached underalternative 2. Medicare would not only refuse pay-ment for a technology considered inefficacious or un-safe, but would also limit payment to conditions forwhich the technology was deemed efficacious andsafe. The Office of Health Practice Assessment couldcontinue to advise medicare. It could secure the rele-vant evaluations, digest them for medicare purposes,and point out areas needing further information. Al-ternatively, medicare could deal directly with anynew office established.

Theoretically, the same procedure could apply toreimbursement under medicaid, but such a stepmight require amending the Social Security Act. Al-though medicare officials have already decided thatthe program has administrative authority to denyreimbursement for new technologies, medicaid ad-ministrators are less certain of medicaid’s legal au-thority at the Federal level. States have the authorityto deny medicaid reimbursement and have exercisedthat authority.

As a probable consequence of this fourth alter-native, judgments about efficacy and safety wouldaffect the use of medical technologies. To the extentthat payment by medicare is important to hospitals,physicians, and patients, all three groups would havean incentive to follow the judgments made. As a re-sult, this alternative could help prevent inappropriateand harmful technologies from being introduced, dif-fused, and used, and could reduce expenditures onthem for medicare patients. At the same time,however, this alternative is less intrusive than direct-ly prohibiting the use of a technology. Providersmight use unapproved technologies, but would thensimply forego medicare reimbursement.

Substantial changes in the medical care systemcould flow from this alternative. The traditionalprocess of third-party payment by Governmentwould change. Government has traditionally leftdecisions of appropriate technologies and conditionsof use to practicing physicians. To the extent thatGovernment reimbursement exerts leverage on pro-viders, this alternative would restrict the use oftechnologies.

Implementing decisions at the local level to denyreimbursement would pose difficult technical prob-lems. Medicare already transmits to its carriers andintermediaries instructions on particular technologiesand conditions of use for which reimbursementshould be denied. These medicare agents in turn havethe responsibility of informing providers and en-forcing the restrictions. Because of the magnitude ofservices involved, implementation depends primarilyon the good faith of providers and secondarily onselected audits.

Billing practices, for example, make monitoringthe use of specific technologies difficult. CT scansmay be reported under the general category of radio-logical procedures. The present level of detail rarelyindicates specific drugs or their conditions of use. Intheory, Government agents adjust cost reimburse-ment for institutions to exclude costs of disallowedtechnologies, such as CT body scans. If imple-mentation of this alternative made these adjustmentstoo intricate and lengthy, the Government mightchoose to drop cost reimbursement and switch topayment by service, even in institutional settings.

This alternative could substantially lengthen thetime required to introduce an innovation into medi-cal practice. As discussed in section 1, the mere ex-istence of information and judgments might influencethe use of technologies. By denying Governmentreimbursement for unapproved uses of technologies,this alternative would give substance to thosejudgments. Providers would be reluctant to adoptprocedures for which they and their patients couldnot receive payment. And the longer time required tointroduce an innovation would apply to both ef-ficacious and inefficacious technologies.

Linking medicare reimbursement to more system-atic evaluations of efficacy and safety could occuronly as a gradual process and over a long period oftime. Clinical studies, syntheses, and judgments areall lengthy undertakings. A practical approachwould be an incremental process of making reim-bursement contingent on comprehensive evaluationsas they become available. Or in the case of new tech-nologies, the Government could mitigate the prob-lem of delay by screening and permitting reimburse-ment for those with the potential to save patients forwhom no efficacious technology exists. A newsurgical procedure, for example, might be reim-bursed for patients suffering from an otherwise fatalcondition.

While a new technology is undergoing evaluation,medicare could pay for it only in designated loca-tions. The choice of centers would have to take intoaccount access for patients throughout the country.

These centers could provide data for evaluating thetechnology; their participation in controlled clinicaltrials could be a condition of their designation. Thesetrials could generate data for analyzing efficacy andsafety without widespread dissemination of the tech-nology. This alternative might reduce innovation be-cause it would make the process of innovation riskierfor developers of new technologies. If other third-party payers followed medicare’s lead and if thispolicy affected use and sales of a technology, innova-tion could become more risky.

Another consequence of this fourth alternative isthat reimbursement would be withheld for patientscovered by governmental programs, but not forother patients. Medicare and medicaid cover certainsubgroups of the population because they havegreater medical need or less ability to pay. Restrictingreimbursement for these patients would probablyresult in their receiving different services from otherpatients because marry medicare and medicaid pa-tients would be unable to pay for their own medicalservices. Such a consequence could protect these pa-tients from harmful and inefficacious services, as wellas prevent their receipt of efficacious and safe serv-ices. Other third parties such as Blue Shield are start-ing to make payment contingent on efficacy. To theextent that other insurers followed the same course,medicare and medicaid patients might not be re-stricted more than other patients with insurance.

The Department of Health, Education, and Wel-fare is already linking reimbursement and efficacythrough administrative action, as discussed inchapter 6. DHEW’s decisions, then, may make con-gressional action superfluous,

Alternative 5: Expand regulation of capital ex-penditures to cover purchases of medical equip-ment regardless of setting or ownership.Under the provisions of the National Health Plan-

ning and Resources Development Act (Public Law93-641), capital expenditures over $150,000 are sub-ject to certificate-of-need review only if made byspecific medical care facilities. These facilities includehospitals and certain categories of ambulatory carefacilities, but exclude private physicians’ offices.Similarly, section 1122 of the Social Security Act ap-plies to capital expenditures over $100,000 only ifmade by the same types of facilities. Therefore,unless State certificate-of-need laws authorize suchregulation, purchases of equipment by physicians inprivate offices are not subject to review by planningagencies. At the end of 1977, the laws of only sevenStates covered physicians’ offices.

These State laws encourage circumvention of theregulatory process by treating the same kinds ofequipment differently, depending on ownership or

setting. Physicians and other individuals may leaseor purchase capital equipment, such as a CT scanner,place it near a facility that is regulated, and be ex-empt from review. To the extent that the nationalguidelines issued under Public Law 93-641 increasethe stringency of criteria for regulated providers, theguidelines will further induce placement of equip-ment in unregulated settings.

Incomplete coverage of capital expenditures mayfoil the plans developed by local agencies. A plan-ning agency may decide that a certain number of CTscanners is appropriate for its area and approve thatnumber of applications from regulated providers.Purchase of scanners by other unregulated providerswould counteract the local plan, but would lie out-side the planning agency’s jurisdiction.

This fifth alternative suggests amending currentlaws to cover capital expenditures over a certainamount, regardless of the ownership or setting wherethe equipment is operated. A planning agency wouldthen have more complete control over the numberand distribution of such equipment in its area. By ex-panding the regulation of capital expenditures tocover providers such as physicians’ offices that arenow exempt, the alternative would remove the pres-ent incentive for providers to place equipment in un-regulated settings. This alternative would not givepreference to one setting or form of ownership overanother. Planning agencies could still set prioritiesamong applications and exercise discretion over theplacement of equipment. (Two bills, S. 2410 and S.2551, that would so amend Public Law 93-641 arenow before Congress. ) The Social Security Act andthe National Health Planning and Resources De-velopment Act differ in the amount of the expendi-ture that triggers coverage. Legislation could makethese amounts uniform, but that is an issue separatefrom this alternative,

The broadening of the planning provisions underthis fifth alternative would necessitate arrangementsfor physicians to have access to available equipment.Since laws now generally apply to hospitals, any newproblems of access would be limited to ambulatorypatients; these patients could be transported betweenfacilities. Many planners already include sharing ofservices in their criteria (see ch. 4). Ensuring access toequipment for physicians might require changes inthe legal liability that a medical practice bears. Apractice, which is now responsible for its own staffphysicians, might otherwise become responsible forthe actions of other physicians who are using thefacility’s equipment,

Implementation of this fifth alternative would in-crease the workload of the regulatory process, Thetotal number of purchasers of equipment covered by

Appendix A. —Policy Alternatives ● 5 5

the law would increase substantially, with a probablerise in the number of certificate-of-need applications.Administrative costs of capital expenditure regula-tion would increase accordingly. To the extent thatnewly regulated purchasers of medical equipment re-quired additional personnel time to apply for cer-tificates of need, their costs would also rise. Oneshould note that regulated providers already bear thecost of applications.

An increase in the level of regulatory activity couldalso slow the diffusion of new medical equipment.The implications for quality of care are unclear, sinced e l ay would affect efficacious and inefficacioustechnologies alike. Likewise, the effect on expendi-tures for a given technology is difficult to determine.The certificate-of-need process may deter some po-tential purchasers. Later purchasers of new productsmay benefit from lower prices as a result of com-petition or decreased manufacturing costs. Or theymay face higher prices due to inflation, increased de-mand, or product development.

A related issue is the effect of this fifth alternativeor any such regulation on total capital expenditures.Practical limitations of time and money require aminimum expenditure threshold for certificate-of-need review. But it has already been observed thatregulated providers such as hospitals shift theircapital expenditures to less regulated technologies.Such substitution is sometimes possible within thesame category of equipment; some models of CTscanners sell for less than $100,000. This situation ispart of the larger context wherein a new technologyis not necessarily substituted for another. Rather thenew are typically added to the store of existing tech-nologies. This alternative, then, will not in itself limiteither total capital expenditures on medical equip-ment or expenditures on the use of that equipment.

3. Financing Methods

The financing of medical care influences use of andexpenditures for technologies through incentives toproviders and patients and through restrictions oncoverage and payment. The Federal financing pro-grams, medicare and medicaid, have largely con-tinued the reimbursement methods that prevailed inthe private insurance field (see ch. 6). Payment bythese programs to hospitals on the basis of costs in-curred, and to physicians on the basis of charges, hasresulted in an open-ended commitment by these Fed-eral programs to finance the use of covered services.

In the course of financing medical care, public andprivate third-party payers have restricted the extentof coverage and payment. They have, in effect, de-fined the product for which they will pay. Medicare

and certain private third parties in some cases havelimited coverage to efficacious technologies. On thatground, medicare refused payment for CT bodyscans. (Setting maximum rates of payment for certainservices are more widespread. Medicaid, for exam-ple, has placed ceilings on its reimbursement fordrugs, and most third parties place some limits ontheir payment of physicians’ charges. ) Ironically,Federal financing—like health insurance in general—has encouraged the use of services such as CT scans,but not efficient methods in their performance ortheir substitution for other services. No restrictivemechanism such as a finite budget induces providersto make tradeoffs between increased information orbenefit and increased costs from using technologies.On the contrary, financing methods reward withhigher revenue those providers who perform addi-tional services, regardless of their marginal value orefficient performance. As a result, providers have lit-tle incentive to choose among alternative proceduresor to perform services efficiently. Prevailing third-party payment thus insulates providers as well as pa-tients from the financial consequences of usingtechnologies.

Contained in this section are two alternatives toaddress problems with current financing methods.Under the first, medicare and medicaid would con-tinue to use costs or charges as the basis for reim-bursement, but would base their rates on efficientmethods of performing services. The second alter-native would fundamentally change the paymentmethod in order to create incentives for providers tobecome cost conscious in using and producingmedical services. Although the alternatives in thissection are mutually exclusive, either could be com-bined with alternatives from the previous sections oninformation and regulation.

Alternative 6: For services paid by medicareand medicaid, establish rates of payment thatare based on efficiency.The Department of Health, Education, and Wel-

fare has set limits on routine hospital operating costsand charges of drugs payable under medicare andmedicaid, respectively. However, reimbursementlimits on routine hospital costs are only very general-ly related to efficiency of operation. And withroutine costs of a hospital day limited, hospitals havea strong incentive to allocate costs as much as possi-ble to ancillary services, which are often not limited.

These policies give providers who receive costreimbursement little incentive to be cost conscious intheir services and production methods. As a result,governmental payments probably exceed those thatwould result from limits based on a tighter definitionof efficiency.

Similarly, reimbursement to physicians is basednot on standards of efficient operation, but oncharges prevailing in a given area. Nor does govern-mental policy coordinate payments to hospitals andphysicians’ offices to ensure comparable payment forcomparable services. Medicare, for example, couldpay different amounts for the technical component ofan ambulatory CT scan depending on the settingwhere it occurred, And the charge for that service ina physician’s office is typically higher than its cost ina hospital.

Under this sixth alternative, rates of paymentwould be based on the basic costs necessary to op-erate a facility or piece of equipment at an efficientlevel. Soliciting bids from manufacturers might be re-quired to lower purchase prices of equipment. Tomake payments consistent for comparable servicesthat are based on charges in one setting and on costsin another, fee schedules would be developed forservices paid by charges. Fees paid to physicianswould also be based on costs using efficient methodsof operation. To that basic amount would be addeda predetermined profit margin to arrive at th e

allowable fee, This alternative could apply to allpayers or all third-party payers, not just medicareand medicaid. In that case, the alternative would en-tail the establishment of national ratesetting formedical services.

Under this alternative, medicare and medicaidwould not pay for inefficient methods of operation orfor high profits. Rates could be reviewed to enablemedicare and medicaid to take advantage of changesthat had resulted in lowered costs, such as reductionsin prices of equipment or improvements in methodsof operation. Of course, changes in these factorscould lead to increases in rates as well as decreases.

Under the assumption that medicare and medicaidpayments exert a degree of leverage over providers,these federally set rates could encourage the per-formance of services in ways considered desirable bythe Government. The relative rate structure for dif-ferent settings, different tests, and different types ofphysician specialists could provide incentives favor-ing one over another. For example, the Governmentcould establish rates for CT examinations and alter-native diagnostic procedures, such as arteriograms,that would encourage the relative level of use of eachtest that was considered desirable. If all physicianswere considered equally capable of reading CT scans,all could be reimbursed at the same rate. If some wereconsidered capable and others not, reimbursementcould be limited to those considered capable.

Considerable technical expertise would be neededto set, monitor, and review rates under this sixth al-ternative. For both hospitals’ and physicians’ rates,

the Government would require experts with detailedknowledge of such factors as budgets, methods ofperforming services, and types of equipment. Also,to set fees and monitor costs, hospitals and physi-cians would have to adopt uniform methods ofrecording and reporting their costs. (Public Law93-641 mandated the development of uniform ac-counting and reporting, and Public Law 95-142 re-quired uniform reporting for institutions. ) If pay-ment under medicare and medicaid were based on theefficiency of services provided, hospitals would haveto apportion costs to specific services, not to depart-

ments or functions as is currently done.Whether the ratesetting described here would re-

sult in lower net expenditures on medical services isnot clear. Rates would probably be lower for medi-care payments, but total expenditures would not nec-essarily rise more slowly or decline absolutely. Othergovernments, such as those of the Canadian Prov-inces, have found that rates of use and therefore totalexpenditures have risen when rates of payment wereheld fixed. The costs of hiring the new technicalexperts required would also add to government ex-penditures. Despite the time and expense involved,this alternative would not necessarily lower pay-ments under medicaid. Since 1972 when the law wasamended, medicare’s definition of reasonable costsfor hospitals has been a maximum limit for medicaidpayment; many States pay less. Medicaid’s limits forphysicians’ services are also typically below those ofmedicare.

Certain adverse consequences might result if medi-care rates paid to physicians were reduced belowtheir current levels. For example, fewer physiciansmight be inclined to accept assignment for medicarepatients (acceptance of medicare rates as full pay-ment); the rate of assignment is already falling. Insuch circumstances, medicare patients with some fi-nancial means could pay the difference betweenphysicians’ charges and medicare’s allowable fee, Butpatients with less ability to pay might have to rely onphysicians with lower charges.

Overall, ratesetting entails detailed considerationof each service, the method of performing that serv-ice, and the profit margin. This course of actionwould be time-consuming and expensive for pro-viders and governmental agencies alike. Implement-ing this sixth alternative might result in the Govern-ment’s questioning in detail how medical services areprovided. Furthermore, ratesetting would not affectthe incentives of present reimbursement methods thatencourage additional medical services, such as diag-nostic tests, regardless of their marginal value.

Alternative 7: Fundamentally restructure thepayment system to encourage providers to per-form and use medical services efficiently.

Appendix A. —Policy Alternatives ● 57

Present retrospective payment of costs and chargesand fee-for-service payment contain perverse incen-tives, as discussed in alternative 6. These paymentmethods, used by public and private third parties andby self-payers, reward physicians and hospitals withhigher revenue when they provide additional serv-ices. This result occurs regardless of whether theservices substantially improve patient care orwhether they are produced efficiently. Medicare, forexample, pays for a CT head examination regardlessof any other neurodiagnostic tests that have beenperformed and the information that may have beengained from them.

This study has identified the incentives of the pres-ent reimbursement system, but has not systematical-ly analyzed possible changes in that system. This al-ternative, then, suggests a general restructuring ofpayment methods, but does not propose a definitesubstitute. The altered payment system would con-tain incentives for physicians and hospitals to pro-vide appropriate care and to do so efficiently, insteadof present incentives that conflict with these goals.Rather than control rates of payment for each serviceas in alternative 6, this alternative would indirectlyor directly fix the total revenue of a provider in ad-vance of the delivery of medical care. Payment bycavitation (per person) would do so indirectly, whilereview of providers’ budgets would fix that revenuedirectly.

The consequences of a restructured payment sys-tem would depend on the specific plan put into effect.Nevertheless, certain generalizations are possible.Limiting total revenue would both enable and forceproviders to make choices among alternative servicesand among alternative methods of performing thoseservices. Within the predetermined revenue, a pro-vider could choose which services to perform andhow to perform them. With total revenue limited, forexample, a hospital’s administrator and physicianswould decide whether to operate a CT scanner, howmany scans to perform annually, which patients toscan, and how to combine CT scans with otherdiagnostic procedures.

Furthermore, physicians and hospital adminis-trators rather than Government would make thedecisions. The Government would set the cavitationpayment or budget limit, but would not become in-volved with production methods, use, or paymentfor particular services. Providers could consider thecost implications of their actions, choose services toprovide, and determine how to perform those serv-ices. The factors that physicians and hospitals weighwhen making decisions would undoubtedly undergogreat change. Additional services would no longer

automatically increase their revenues and might evendecrease their incomes by increasing their costs.

This seventh alternative could pertain either toFederal financing programs alone or to all payers ofmedical care. However, if only medicare and medic-aid limited their payments, a provider could increasecosts and charges and generate additional revenuefrom other third parties and self-payers. The alter-native could also cover either hospitals or physicians.But some services that are performed in bothhospitals and physicians’ offices, such as ambulatoryCT scans, are often substitutes for each other. Ifrevenue were limited only for hospitals, one wouldexpect payments to rise for nonhospital providerswhose revenues were not limited. Although thisalternative would clearly be most effective if ap-plicable to all payers and providers, such an ap-proach would represent a major policy decision.Private payers could, of courser follow any Federallead. This alternative would also be compatible withnational health insurance, for the Federal Govern-ment would then be the major payer of health care.

Calculating cavitation levels or revenue limitswould require the responsible Government office tohave much technical expertise. Experts would have toidentify variables that cause costs to differ amongproviders or consumers and adjust payment levelsaccordingly. (Such efforts have not proved very suc-cessful in the past. ) Governmental experts would alsohave to review rates periodically. The ways in whichrates changed would greatly influence total medicalexpenditures. For example, a system of basing therate of change on an indicator within the medicalcare system could simply accept and transmit in-creases with a lag of 1 year. Rate changes could bebased on broader economic indicators, such as theGNP deflator, which would not necessarily be self-generating, But broader indicators might be insen-sitive to changes specific to the medical care sector.

Although the changed payment system wouldcreate an environment with different incentives, thisseventh alternative would not necessitate substantialchanges in the way providers are organized. Pro-viders could continue to deliver medical care undercurrent practice arrangements. Compared to the cur-rent situation, the new environment would enhancethe competitive position and perhaps stimulate thegrowth of health maintenance organizations (HMOs)and other providers currently paid by cavitation,Such groups now compete for physicians, supplies,and enrollees with providers who gain more revenuefrom the provision of additional services. If cavita-tion payment or budget limits applied to all pro-viders, all would have similar incentives and be sub-

ject to similar restrictions under the payment meth-od. But the relative position of providers now paidby cavitation would be improved if others facedsome limit on their total revenue.

The presence of different incentives would affectthe kind of medical care delivered and expenditureson that care only over a long period of time. Similar-ly, any effect on the nature of medical care deliveryand the strength of HMOs would occur over severalyears.

Changing payment to providers as described inthis seventh alternative would be compatible withregulatory programs of certificate-of-need and uti-lization review, and might make these programs evenmore valuable than at present. Under this alterna-tive, providers would have an incentive to under-

serve patients in order to stay within their budgets.Minimum standards of appropriate use might haveincreased importance in this new context. Utilizationreview under the PSRO program currently appliesonly to medicare and medicaid patients, as describedin chapter 5. To prevent providers from economizingon service to nonmedicare and nonmedicaid patients,PSRO review could be broadened to cover all pa-tients. Such an expansion of the PSRO programwould represent a major policy decision and wouldsubstantially increase PSRO regulatory activities andadministrative costs. Utilization review might alsoguard against the tendency of providers to considercosts exclusive of benefits in order to meet theirbudgets. Standards of appropriate use would therebyfunction as a counterweight to the possibility of in-creased cost consciousness by providers.

Appendix B.— Research and Development of CT andOther Diagnostic Imaging Technologies

The computed tomography (CT) scanner was de-veloped with little involvement of U.S. Governmentresearch agencies. Nonetheless, Federal support forR&D of the CT scanner has been substantial in thepast. It is clear that this support has decreased signifi-cantly and steadily in the past few years (23,110).Meanwhile, private industry has assumed an increas-ing share of further basic R&D of CT scanners.

The National Institutes of Health (NIH) has beenthe major source of Federal funding for R&D of CT.Of the Institutes, the National Cancer Institute (NCI)has been the most active, supporting an estimatedtotal of over $4 million in CT-related research proj-ects over the past several years. The last major proj-ect funded by NIH (NCI) concerned with developingtechnological improvements in CT scanners, how-ever, terminated in April 1978: This extramurallysupported research yielded the fixed-detector geome-try type scanner developed by the American Scienceand Engineering Co. (AS&E) (1 10).

Currently, most CT-related research funded byNIH is concerned with new and improved uses of CTscanners and/or applications of CT scanning. Thefunding levels of current projects, however, are muchmore modest than those of earlier projects concernedwith basic R&D of CT itself. More importantly, NIHresources currently being allocated to CT pale incomparison to NIH moneys being allocated to theR&D of other imaging technologies.

For example, NIH is currently supporting basicR&D of the dynamic spatial reconstructor (DSR) im-

aging system; positron emission transaxial tomogra-phy (PETT); and zeugmatography, or the applicationof principles of nuclear magnetic resonance (NMR) toimaging techniques, In addition, ultrasound (whichmuch preceded CT historically) continues to be re-searched at NIH for improvements in the technologyitself, as well as for new and improved applications.

Theoretically, the imaging capabilities of someof these new technologies exceed those of CT. Someof these technologies may also be safer than CT, be-cause they do not use ionizing radiation. Given theseadvantages, the development of these technologiesand their eventual emergence into clinical use couldplay a decisive role in the future of CT scanning.One trait that the new technologies (in particular)have in common with CT that might dampen this po-tential effect, however, is their costliness. In somecases, their estimated cost not only rivals, but ex-ceeds, that of the most advanced CT equipment cur-rently available.

Consequently, these emerging technologies willsoon face many of the Federal policies established inthe wake of the introduction, diffusion, and wide-spread use of CT scanners. Just how these expensive—but nonetheless, miraculous—technologies willfare when they encounter Federal policies toward theevaluation, diffusion, and reimbursement of newhigh-cost technologies will be interesting indeed. Thefield of diagnostic imaging is already a large and ex-pensive one, as shown in tables B-1 and B-2.

Table B-1.–Overview of Diagnostic Imaging in the United States (1977 and 1980)—

Number of Number o f p rocedures ‘- ‘ -

(millions) Costs (millions)hospitals with --– –.. ---–- —–— —--—-–—-— --– .—–-—— -–- -—- –-capability 1977 1980 1977 1980

Diagnostic X-ray. 7,000a 158b 171b $5,300 b $ 7 , 6 0 0 c

CT scanning 1,OOOC 1-1.4C 3.4C

$300d $875 cd

Nuclear medicine. 3,300e 8.2f 11.1b $800 f $1 ,250b

U l t r a s o u n d All Approximately 4g na $360g

aThls is the apprommate number of hospitals In the United States It IS assumed that all have such equipment—

bEst[mates of Bureau of Radlolog]cal Health 186) The diagnostic X.ray figures Include dental XraYCOTA est!matesdTh, s f,gure ,$ Partlal[y offset by reductions In other dlagnost[c procedures Estimates are presented In OTA’S 1978 CT rePort

11291‘American Hospital Association Hosp(tal S(a(/st/cs 7978 Ed/t/on (Chicago, Ill 1978)fL Russell Technology ,n Hospl(a/s (1461 Russell s 1975 estimates are extrapolated to 1977gNo rellable figures are ,aVallable The State of New York survey of Rochester N Y If projected nationally, would Indicate 76

mfltron procedures In hospttals alone (29) Sources In the Bureau of Radtologlcal Health cite Informal estimates of 1 4 mill (onprocedures In 1979 excludlng obstetrical use (a large Componf?ntl, but also !nd!cate that that est!mate seems too (OW Ultrasound use IS growing rapidly The Stanford Research Institute estimates 12 m I [lion to 14 ml Illon procedures In 1979 growingto over 125 mill on In 199011 63)

NOTE Estimates are approximate for dlustrat]on only The Bureau of Radlologlcal Health IS presently beglnnlng a survey of125 hospitals to determine the rates of use of dlagnosllc X r~y, ultrasound, and nuclear medlclne This study WI I givemuch more conclusive figures than those shown above

Table B-2.-Sales of C* Diagnostic Imaging Equipment in the United States, by Year (1977-83)

Sales (millions of dollars).1 9 7 4 1975 1976 1977 1981 1982 1983

Diagnostic X-ray. . . . . . . . . . . . . . . . . . $265’ $300 b $230 b $280 b $375 b

CT scanning . . . . . . . . . . . . . . . . . . . . —

—.100b 12ob 160b 2 0 0b

Nuclear medicine. . . . . . . . . . . . . . . . 40 C — — 100d —

Ultrasound . . . . . . . . . . . . . . . . . . . . . 65a — — 160e — 269f $490e

NOTE The validity of the sales fig~les IS not known They are undoubtedly rough They are Included here as general indicators only The Stanford Research Illstltuteestimates expenditures for diagnostic lma91n9 equlprnent, sumlles, accessories, and rn’lnten’flCe Of $’2 bllllon In 1978, rlsln9 to almost $65 bllllon In 1990 (163)

SOURCESaE/ectr/ca/ News Mar 29, 1976, p 59 dE/ectrtc Business, May 1979, P 66bE/ecfron/c, Jan 5, 1978, P 148 eJ E/cc Eng, November 1979, P 14

f,nst Tech April 1978, P 18cpredi 88, May 1531973, P 24

Basic and Applied Research on CT

Current CT Scanners

Since the publication of the 1978 OTA report onCT scanners (129), the technical capabilities of CTscanners have increased as new models have been de-veloped. This increase has expanded the potentialusefulness of these scanners. The new scanners offertechnically improved image resolution, largely byvirtue of reduced scanning times and the consequentminimization of problems associated with patientmotion. The scan times of the most recent CT scan-ners are less than 5 seconds for a single cross-sectionimage. The most recent scanners are capable ofachieving image resolution of as little as 0.61 mm (seetable B-3).

The scanners listed in table B-3 were developed pri-vately with the exception of the AS&E scanner.AS&E received considerable Federal support fromNCI of NIH ending in April 1978 (22). AS&E, how-ever, only sold a few of the new scanners. In January1978, Pfizer, Inc., made an agreement with AS&E topurchase the rights to market and produce the scan-ner. Using the AS&E gantry, Pfizer made certaintechnical modifications (primarily in the electroniccomputer of the scanner) and now markets a hybridof the AS&E scanner known as the 0450 Pfizer/AS&E scanner. The scanner has a price tag of ap-proximately $650,000 to $700,000. According to the

‘The term ‘generation” IS often ,ipplled to describe the type O( scannerThe first scanners to be developed all used a stmllar approach, and haveoften been labeled ‘ first ~enerattc)r ,‘ The prtmary mean]n~ of the phrasewas to ]ndlcate that that type of scanner was the t]rst to be developed, How-ever, there is an Inevitable Impress ion conveyed that the ‘second gcmera -tlc~n” IS super]or to the ‘ first genera t [on. ‘ For this reason, OTA, In consulta-tion with the manufacturers, has a(’cepted labels that are more descriptiveand not as m]sleadln~, In particular, this was done because what has beencalled the ‘t(mrth ~eneratlon ” scanner IS not superior to the “third ~enera -tt(m” scanner acc<)rdln~ to both the National Electrical ManufacturersAsw)clatton and the Bureau of Radlolo~lcal Health ot the Food and DrugAdmlnlstratlon. The columns ]n table B-3 are in order of development, withthose scanners w)metlrnes called “third generation’ and “fourth genera tlonf’t(>gether In the column labeled “rota e-only.

Food and Drug Administration (FDA), 13 of thesescanners were reported to be sold in the United Statesbetween June 1978 and 1979 (95).

The Dynamic Spatial Reconstructor

Development of the DSR imaging system at theMayo Clinic is currently receiving substantial NIHsupport (23). The DSR system adds the critical di-mension to computerized tomography that is neces-sary for accurate imaging of moving organ systems(such as the heart and lung) and for studies of three-dimensional anatomy and circulatory dynamics in allregions of the body (88). These capabilities are de-pendent on the development of high-speed electronicdata processing and digital computing techniqueswhich is an integral part of the R&D of the DSRsystem (23).

Developers of the DSR system do not believe thatit represents an extension of previous CT scanningprinciples and logic. They are reluctant to call it anadvanced CT scanner (67). The DSR system does useX-ray (as does CT): But whereas CT is capable ofproducing only a 2-mm thick cross-section at a scantime of just a few seconds, the DSR when completedwill be able to scan up to 240 l-mm thick cross-sections in 11 msec, repeat the complete scan pro-cedure at intervals of l/60th of a second, andreconstruct the entire three-dimensional volume of awhole organ, as well as dynamic changes in shapeand dimension of moving structures (88). The princi-ple components of the system are shown in the illus-tration below (see figure B-l). The DSR is describedas follows (88):

. . . A set of 28 rotating-anode X-ray sources, inde-pendently controlled, is arranged around a semicirclewhose radius is 143 cm. Abutting this arrangement isanother semicircle that contains 28 independentlycontrolled image intensifiers and image isocon cam-

Table B-3.—Typesand Models of CT Scanners(1980)

Motion of gantry:

Scanners no Ionger availablecommercially in the UnitedStates as new equipment

Current models

Scanners-announced but notyet available commercially

Rotate and translate,dual detector

4-6 min scan timeSingle pencil beam X-ray

source

2 detectors

Source and detectors trav-erse gantry in parallel, gan-try rotates through smallangle, process repeats,

E M I M a r k I General Electric Neuroscan

CT/NPfIzer0100Siemens Siretom

SOURCE National Electrical Manufacturers Association 1980

eras. These two semicircles make up one circle ofequipment. Inside the circle is a 30 cm-wide floures-cent screen, bent to form a semicircle with a 58-cmradius. The 28 Image intensifiers and image camerasproduce 28 images on the flourescent screen .

The entire assembly—X-ray sources, floures-, . .cent screen, and video camera chains—is mounted ona cylindrical gantry, which is rotated at 90 degrees persecond about a horizontal axis. Each X-ray source ispulsed on sequentially to irradiate the patient for 350seconds. Simultaneously, the image intensifier andvideo camera for each X-ray source are activated torecord the image on the flourescent screen. The heavyreliance on image intensification lowers, in effect, theX-ray dosage to levels no greater than those now em-ployed for X-ray for X-ray procedures in general med-ical practice .

Rotate and translate,multiple detector

20 see-2 min scan time2 or more pencil beams or

Single fan beam X-raysource

3-60 detectors

Sources and detectorstraverse gantry in parallel,taking more readings androtating through largerangle than dual detector.

Elscint 850EM I CT 1010EM I CT 5005EM I CT 7020Ohio-Nuclear 25Ohio-Nuclear 50Ohio-Nuclear 50FSPicker TR-120Philips Tomoscan 200Syntex System 60Syntex System 90Toshiba TCT-10A

CGR ND 8000Elscint 905Hitachi CT-WOhio-Nuclear 100Omni Medical 4001Pfizer 0200FSToshiba TCT-30

Rotate only

Under 5 sec scan timeSingle fan beam X-ray

sourceHundreds of contiguous

detectorsRotation motion only. In somemodels, source and detec-tors move together; in othermodels only source moves.

AS&E 500Artronix 1100Artronix 1120EM I 6000EM I 7070Searle Pho/Trax 4000Siemens Somatom IVarian V-360

General Electric CT/T 7800General Electric CT/T 8000General Electric CT/T 8800Ohio-Nuclear 2005Ohio-Nuclear 2010

O h i o - N u c l e a r 2 0 2 0Omni MedicalPfizer 0450Philips Tomoscan 300Picker Synerview 300Picker Synerview 600Siemens Somatom IIToshiba TCT-60A

CGR CR 10000Philips Tomoscan 310

The DSR system, somewhat reduced for fundingreasons, is currently being tested on animals on a lim-ited basis. Researchers estimate it will be at least 2 to3 years before it will be used to scan the first patient.At this time, it is being developed for medical re-search purposes, and not wit h an eye towards massclinical application (67). The system will cost about$5 million, and might cost $3 million in mass produc-tion (34). The ultimate use of the multimillion dollarDSR system in the practice of medicine is viewed as atertiary, or even a Quarternary tool, with perhaps 5to 10 serving the entire country (67,83).

NIH has been the primary supporter of R&D of theDSR imaging system. The National Heart, Lung, andBlood Institute (NHLBI) has been the major source of

62 . Policy implications of the Computed Tomography (CT) Scanner An Update

28

Image

Figure B-1 .—Principal Components of the DSR Imaging System

IHigh-speedreconstruction

1

sourcesVideo taperecorder

SOURCE :IEEE Spectrum, January 1979, p 77

funding for this project since one of the major objec-tives is to permit accurate measurement of the struc-ture and function of the diseased and normal heart(144). NHLBI support totals about $2 million overthe past few years for development of the imagingdevice itself. Development of the high-speed com-puter system necessary to the device has been sup-ported by the Division of Research and Resources(DRR) of NIH, which has spent about $1 million dur-ing fiscal years 1978 and 1979 (23).

Research on Applications of CT

It is difficult to compile an inclusive listing of proj-ects related to the applications and uses of CT scan-ning at NIH. First, since such projects are organizedby disease and organs as the Institutes are, identifica-tion of CT-related research is difficult. Secondly,even when such projects can be identified, it is dif-ficult to determine the proportion of moneys thatshould be apportioned to research on CT. Without aformal survey of the Institutes, therefore, preciseestimates of such projects and their funding levels areunavailable. Consequently, the projects discussedbelow are meant only to indicate the kinds of ongo-ing research being supported by NIH, Similarly, theaccompanying dollar figures are provided as a rough

i ’.

computerII interface .I

estimate of current Federal investments in this type ofCT-related research project.

Formerly the major NIH backer of research (on CTscanning, NCI spent only approximately $75,000 infiscal year 1979 on research for scanner development(11), In addition, however, NCI spent approximately$400,000 in that year for CT-related studies withsuch objectives as developing better contrast agentsand new algorithms for diagnostic use to reduce radi-ation exposure (110). Also, in that year, NHLBI sup-ported some extramural research grants involving theuse of CT scanners in diagnostic methods for par-ticular cardiac diseases (110), DRR, a major fundingsource of the DSR imaging system discussed above,also supported about 15 projects involving the use ofCT scanners through its biomedical research supportprogram: These few projects, however, are verymodest totaling approximately $65,000 in fiscal year1979. In addition to this research, about 3,000 pa-tients per year are scanned in the NIH ClinicalCenter. Most of these represent patients who are onprotocols requiring a CT scan (110).

The major project at NIH related to the applicationof CT is funded by the National Institute of Neuro-logical and Communicative Disorders and Stroke(NINCDS). In fiscal year 1979, the Institute funded a$500,000 project investigating the use of CT scanning

in the diagnosis of head trauma (110). The Institutealso had supported about $50,000 in intramural re-search projects related to the use of computed headtomography for diagnosis of diseases particularlyrelevant to it, such as brain tumor (110).

In conclusion, although there are still many initia-tives at NIH related to applied, as opposed to basic,research on CT scanners and scanning, the cumula-tive resources devoted to these activities do not beginto approach the levels of funding for the ongoing de-velopment of the DSR system, for example. At thistime, the Federal Government is not a significantcontributor to R&D of CT scanners and scanning: Itstime has come and gone. Instead, Federal support ofbiomedical R&D is concentrated on new imagingtechnologies.

Emerging Imaging Technologies

There are a number of new technologies and tech-nological applications in the imaging field that holdgreat promise for medical research and eventual clin-ical application. These will not be covered in detail.However, there are two new technologies that areparticularly exciting and at the same time raise manyof the same policy issues characteristics of CT scan-ners. These are PETT and zeugmatography, or theapplication of principles of NMR to imaging tech-niques. Various Institutes at NIH are supportingR&D of both of these imaging techniques, and thereis considerable private (worldwide) R&D investmentbeing made in them as well.

There are now only a few PETT and NMR scan-ners throughout the world, and these so far havebeen limited to experimental clinical use (with humanpatients). However, the unique capabilities and at-tributes of these two imaging techniques have gen-erated a great deal of excitement in the medical re-search community, and the possibilities for clinicalapplication have sparked even greater enthusiasm forthese technologies. Speculation regarding their rolein clinical practice, associated operational costs, andcommercial viability has already captured the atten-tion of the media (83,96). One reason for excitementis that these two technologies may provide the meansto image tissue function, whereas present CT andultrasound techniques provide the means to imagetissue structure.

The excitement, enthusiasm, and speculation sur-rounding these technologies has also drawn the atten-tion of the Office of Health Regulation of the HealthCare Financing Administration (HCFA), and the Na-tional Center for Health Care Technology (NCHCT)of DHHS (143). For example, NCHCT is preparingan overview paper on NMR that reviews the efficacy,

diffusion, and utilization questions surrounding theintroduction of new medical devices (73). It appearsthat PETT and NMR have already been flagged bythese two Federal agencies, and that if and when theyare ready to be introduced into medical practice,these technologies will undoubtedly be subjected toFederal policies toward the evaluation, diffusion,use, and reimbursement of high-cost medical technol-ogies—many of which were formulated around theCT scanner. ’

Positron Emission Transaxial Tomography

PETT is the latest of several radionuclide imagingsystems belonging to the family of nuclear medicinetechniques. Although ionizing radiation is used inPETT, the technology differs significantly from CTin principle and in capability. A PETT scanner maybe briefly described as “. . . a large, computer-con-trolled tomography unit that maps the distribution ofpositron-emitting pharmaceuticals in order to con-struct detailed images of organ metabolism, physiol-ogy, and function” (96).

In the PETT scanning procedure, radioactive iso-topes of elements such as oxygen, carbon, fluorine,and nitrogen are administered to the patient, usuallyby injection, but also sometimes by inhalation. Thisis in contrast to the manner in which CT scanners(and conventional X-ray techniques) expose the pa-tient to ionizing radiation by means of an external X-ray tube. The radionuclides are administered asmetabolically active compounds, such as glucose, oras naturally occurring compounds, such as carbonmonoxide, which may be used as tracers. The imagesproduced by PETT scanners are based largely on thedetection of the distribution of the radioactivitythrough body tissue. Reconstructed images producedby PETT scanners, therefore, may reflect compart-mentalized localization, flow, or biochemical andmetabolic activity, whereas CT scanners basicallydetect and display anatomical structure, although theuse of iodinated contrast media may give significantfunctional information. The difference in informa-tion presented in a comparable cross-section of thebrain produced by these two technologies is il-lustrated by the fact that while CT scans of a cadaverand a live human would show a similar image, aPETT scan of a cadaver would show a relativelyblank screen image in comparison to the scan of alive human, since due to lack of flow, the radioactivematerial would not have been transported (143).

‘The memorandum from the Otf]ce (~t Health Regulation In HCFA su~-~ests that local health systems a~encles and State health plannln~ anddevelopment a~encles be alerted to the ]mpendln~ ]ntroductlon of these twotechnologies as well t 143)

Ter-Pogossian and his coworkers (175) have re-cently described PETT as follows:

In this technique a chemical compound with the de-sired biological activity is labeled with a radioactiveisotope that decays by emitting a positron, or positiveelectron. The emitted positron almost immediatelycombines with an electron, and the two are mutuallyannihilated with the emission of two gamma rays. Thetwo gamma rays fly off in very nearly opposite direc-tions, penetrate the surrounding tissue and are re-corded outside the subject by a circular array of detec-tors. A mathematical algorithm applied by computerrapidly reconstructs the spatial distribution of theradioactivity within the subject for a selected planeand displays the resulting image on a cathode-rayscreen . . . . With suitable interpretation PETT im-ages can provide a noninvasive, regional assessmentof many biochemical processes that are essential to thefunctioning of the organ that is being visualized.NIH investment in the basic R&D of the use of pos-

itrons for imaging which led to PETT has been con-siderable, amounting to almost $9 million in grantsto one research center alone over an 18-year period(143). Significant support of PETT continues and isprojected for the next few years. For example,NINCDS initiated a series of new PETT projects in1979. In the first year of support, the Institute spent$5.9 million in grants to establish five university-based neurology centers of research around the coun-try and the construction of a positron emission to-mography scanning instrument in each (151). Con-tinued grant support for these centers is projected forthe next 3 years. The Institute’s interest in PETTstems from the expectation that PETT will enablephysiological research of cerebral metabolism just asCT enabled research of cerebral vascular anatomyand flow (151). Thus, the purpose of research onPETT by the Institute has been to understand normalbrain biochemistry and metabolic disorders and tostudy the effects of lack of oxygen, various pharma-ceuticals, trauma, and varieties of stress on neuraltissue (151).

There are probably more than 20 experimentalpositron emission scanning devices in the world atthis time, half of which are located in the UnitedStates at 10 different locations: At least three morePETT devices are scheduled for installation (all atU.S. locations); these are also to be used for experi-mental purposes (143). 3 In addition to the investiga-tion of brain functions PETT scanning is also beingused for a variety of other research purposes, most of

IAS ot August 1980, I’ETT scar ners are located at Washington Umverslty(2 ) , Massachuse t t s General Hospital (2), UCLA, Unlvers]ty of Mlaml’sBrookhaven Laboratories, Un]vt rs]ty of Pennsylvania, University of Ch]-cago, Oak Ridge National Labori]tory, Sloan Kettering Institute for CancerResearch , and the Nat]{lnal Institutes ot Health (7), The Un]verslty ofMich]gan, Johns Hopkins University, and Houston are reported to have or-dered PETT devices also ( 143).

which are related to heart and lung functions (96).The strategy in research is to administer differentpositron-emitting chemicals which respond to dif-ferent metabolic pathways in the target organ. Bymeasuring the behavior of these chemicals at varioustimes, information concerning the function of theorgan can be obtained. For example, red cells “la-beled” with the positron emitting carbon-n monox-ide will show the blood distribution in the heart.Clearly, the number of positron-emitting radiophar-maceuticals and biological pathways that can bepaired for study presents an almost infinite numberof permutations. This potential suggests that PETTwill play an important role in research in both organphysiciology, and in basic physiological research(96).

The estimated cost of a PETT scanner and its asso-ciated equipment (i. e., a cyclotron or linear ac-celerator for the preparation of positron-emittingisotopes, and the computer software and hardwaresystems necessary for imaging) is from $1.35 million(143) to $1.94 million (7).4 Such a high cost suggeststhat the use of PETT scanners might be restricted toresearch purposes, since the cost would be prohibi-tive to all but the most major institutions. Never-theless, the potential of PETT technology for clinicalapplication and use has been recognized by manufac-turers of medical equipment, and it is reported that atleast a few have undertaken feasibility studies formarketing PETT scanners (143).

Nuclear Magnetic Resonance Tomography

Although the principles of NMR were discoveredby atomic physicists at least 30 years ago, and havebeen incorporated in the techniques of NMR spec-troscopy developed and used by chemists in analyti-cal chemistry almost since that time, NMR tomog-raphy, or zeugmatography, has only been under de-velopment for the past several years (96). Since 1973,when Paul C. Lauterbur of State University of NewYork (SUNY) at Stony Brook first demonstrated ameans for reconstructing an image in two (and eventhree) dimensions based on NMR signals, zeugma-tography has been the most rapidly expanding ap-plication of NMR in medicine (76). 5 A variety of

‘Almost half of the total estimate ot $1,35 mllll<~n is represented by theestlma ted cost ot cyclotron at $600,000. The computer systems necessarytor Imaging represent another $250,000, and the PETT scanner Itself, ap-proximately S500,000 (143)

‘Examples ot ongoing research on NhlR at various Institutes z nd dlv]-sions at NIH were presented at a science writers’ seminar on NMR held cmApr 23, 1980. These included. studies of he structure and moblllty of DNAand proteins by NMR techniques, NMR studies O( sickle cell in t Ie intactred blood cell, and NMR studies of the molecular structure of c ollagen.These Intramural projects were In addition to the presentation of the prcqectInvolwng production of two- and three-dimensional Images by NIM R tomo-graphic methods and a discussion of the] r potential diagnostic applicationsby Dr. Hoult (122),

techniques have been developed by numerous re-searchers in Europe and the United States since thefirst experiment by Lauterbur. While it remains to beseen which method(s) will gain acceptance, the tech-nology is ready to be clinically evaluated.

Since 1973, zeugmatography has made significantadvances in the clarity of computer-generated imagesof the body (143). It is estimated that approximately200 individuals have been subjects of NMR scans(79). Theoretically, the resolution potential of zeug-matography is much greater than X-ray, nuclear, orultrasound imaging techniques (for reasons whichwill be shown below) (77). However, this potential isnot the sole—or even the major—reason for the ex-citement surrounding NMR tomography. Rather, theexcitement stems from the fact that NMR does notuse ionizing radiation (either X-ray or gamma-ray), isnot “stopped” by bone, and most importantly, canyield metabolic information with appropriate ad-justments (77). The relatively greater potentialcapabilities of NMR tomography in comparison toother imaging technologies (including PETT) impliestremendous potential for application to a wide rangeof diagnostic and treatment monitoring functions.However, the effects of the magnetic fields used inNMR are unknown. Although NMR may be saferthan X-ray, it is much too soon to know for certain.

Hounsfield, who was awarded the Nobel prize forhis work in CT scanning, described the principles ofNMR as follows (80):

When hydrogen protons are placed in a magneticfield they will precess (or “wobble”) around the fielddirection just as a spinning top precesses around itsvertical gravitational field. This precession occurs at adefinite frequency, known as the Larmor frequency,and is proportional to the magnetic field intensity.

The usual NMR procedure for imaging is to apply astrong magnetic field along the body to be studied.After a short period of time, the nuclei will align withtheir magnetic moments along the field. A radio fre-quency tuned to the precession frequency of the hy-drogen nucleus is then applied at right angles to themain field by means of a set of coils at the side of thebody, This causes some of the hydrogen nuclei to pre-cess—all keeping in step. After the radio receiver fieldhas been switched off, the nuclei will continue to pre-cess in phase, generating a similar radio frequencywhich can be picked up in receiver coils placed at theside of the body, these signals detect the water contentof the body. It will take some time for the precessionto die away, as the nuclei again realign themselveswith the magnetic field. The measurement of this timeis important as it gives us some information about thenature of the tissue under investigation.This knowledge immediately suggests the compari-

son of recovery times of hydrogen atoms in healthyversus diseased tissue. In early 1971, Raymond

Damadian at SUNY Downstate Medical Center inBrooklyn published research suggesting that theNMR signal from the water in tumor cells differedfrom that in normal cells, the signal from cancerouscells being much longer than that from normal cells(96). The possibilities of a noninvasive, highly sen-sitive diagnostic tool based on chemical informationat a cellular level were obvious.

In 1973, these subtle differences in chemical infor-mation in human tissue were displayed in the firstNMR tomographic pictures, published by Paul C.Lauterbur. Lauterbur realized that by changing thedirection of the magnetic field (in which the patient,or portion of the patient is placed) gradient, and re-peating the experiment at a variety of orientations(i.e., taking projections at many different angles, andthen reconstructing them by computer), it was possi-ble to picture the subject in two (and potentially)three dimensions (77).6 Lauterbur named this tech-nique “zeugmatography” (from the Greek “joiningtogether”) based on the underlying physics wherebythe magnetic field gradient joins together frequencyand spatial information (77). Although a variety ofNMR tomographic techniques are currently beingpursued, all are based on the phenomenon of reso-nance of hydrogen atoms in body tissue. The out-come is a reconstructed image of an organ or wholebody cross-section which appears on a screen (143).Differences in body tissue are thus detected by theirintrinsic chemical differences, rather than by theirdensity or absorpability of X-rays as in CT scanning,or by tracing administered positron-emitting isotopesas in PETT scanning.

Perhaps the most exciting potential of NMR, how-ever, is the potential for metabolic studies that will berealized over the longer term. The dimension ofmetabolic information is already represented in NMRimages. Eventually, it may be possible to “zoom in”on part of an organ, such as the ventricle of the heartor hemisphere of the brain, to obtain metabolic infor-mation in that specific region (79). In these extendedcapabilities, one may envision the imaging ofmetabolic information that would be comparable tothat currently obtainable only by biopsy. It is thecombination of metabolic information (not intrin-sically available in any other imaging technique) withthe image that makes the NMR technique potentiallyso powerful.

Hoult (79) has described one potential application,a scan of a baby’s head for hydrocephalus or intra-cranial bleeding. NMR could locate a particularartery, measure the blood flow in that artery, and

then check that the oxygen uptake of a hemisphere isadequate. All of this would be done totally nonin-vasively, and perhaps without risk to the infant (77).In the Biomedical Engineering and InstrumentationBranch of the Division of Research Services at NIH,such an experiment is under way. The imaging sys-tem developed at NIH has been almost entirely builtwithin the Branch (apart from the computer re-quired). The crux of this particular NMR tomograph-ic system is a novel magnet design that has two mov-able hemispherical windings which can generatepowerful transverse magnetic field gradients (78).Construction of the magnet was near completion inMay 1980. Initially the equipment will be used withphantoms and animals to obtain experience andverify safety. Imaging of human subjects is to beginin spring 1981 (approximately). Eventually, the NIHsystem will be used to scan premature neonates in aseries of experiments. It is hoped that the NMR in-strument will provide a major imaging facility at NIHfor diagnosis and repeated observation of diseases towhich NMR is particularly suited (77).

Besides these capabilities, the final and immediateadvantage of NMR over other imaging techniques isthat it may be safer because it does not use ionizingradiation (79). However, there are real and potentialhazards from strong magnetic fields, especially withpulsed or alternating polarity fields, and resulting in-duction currents (101). Although some have pro-posed that NMR is particularly well suited for usewith infants and fetuses of pregnant women, FDAspokesmen urge caution in applying it to infants orpregnant women (86).

Meanwhile, the expectations based on the capabil-ities and attributes of NMR have attracted the intenseinterest of researchers throughout the world. Uni-versity and research centers developing NMR scan-ning techniques include Nottingham, London, andOxford Universities in England, and SUNY at StonyBrook, SUNY at Downstate New York Medical Cen-ter, the University of California at Berkeley, theUniversity of Illinois, and Johns Hopkins Universityin the United States (143).7 Damadian, who is affili-ated with Downstate has formed his own company,FONAR Corp. and plans to place an instrument in adiagnostic center in Cleveland, Ohio, for clinicalevaluation (101 ).

In addition, there is substantial private investmentcurrently being made in R&D that will translate theprinciples of NMR tomography into devices that maybe commercially marketed. In the United States, Pfi-

‘A150 in this c(mntry the Nld%dchusetts Irrstltute o f Technology, Har-vard UnLverwtv, and Bell Labora tor]es are devel(~p] ng NNIR scanners Theyhave not yet produced Images that are ctlmparable t~~ those of the lnstltu-tlons Ilsted in the text, hcwvever ( 143).

zer has a scanner at the University of California inSan Francisco (143). Johnson & Johnson (Techni-care), General Electric, and Intermagnetics are alsoreportedly involved in the commercial developmentof NMR scanners (79,101). In Europe, there are fourcompanies known to be developing NMR scanners.These are: EMI, Ltd. (United Kingdom), BroukerWest Germany), Siemens (West Germany) , andPhilips (Holland) (143). The intense involvement ofthese companies attests to their expectations regard-ing the potential marketability of NMR techniques.The estimated cost of an NMR for whole-body scan-ning is about $500,000, but with the addition of com-puter equipment necessary to provide the imagingcapability of the scanner, total costs would approach$750,000 (101,143) .

Interestingly, EMI, Ltd. (now Thorn EMI), whichpioneered the R&D of the CT scanner, is actively in-volved in NMR (80,101). Both EMI and NottinghamUniversity have recently produced images of bodysections using NMR. These models have been thebasis for research by several of the U.S. manufac-turers (143). In May 1980, EMI installed its prototypeNMR scanner in Hammersmith General Hospital inLondon with the purpose of evaluating the deviceunder conditions of hospital use (79). It is conjec-tured that EMI will begin manufacturing and market-ing this device in the near future (143).

Ultrasonography

Ultrasonography is not a new technology: Its de-velopment preceded that of CT scanners by at least20 years and it has been used in the clinical practiceof obstetrics since 1956 (59), Ultrasound has ex-perienced a much slower developmental history thanCT scanning, and it has been slower than CT to gainwider acceptance by practitioners and broader ap-plication in medical practices (153). However, thereare now several indications that suggest that ultra-sound is rapidly coming of age 8

Recent improvements in ultrasound instrumenta-tion have resulted in enhanced image quality and reli-ability, convenience of use, and quicker study times,all of which have heightened the appeal of ultrasoundto practitioners. These emerging improvements in thetechnical performance of ultrasound imaging sys-tems, however, are not the sole reason for its rela-tively newfound appeal. There is also increasing im-portance being placed on cost and safety —two at-tributes that have always made ultrasound appealingrelative to CT scanners for some uses. Ultrasound

“Nlarket trends ]nd]cate that ultrasound IS currently the m<]st rapidl y

growung market of ]maglng products (see tahle B-2)

equipment is much less expensive than any CT scan-ner (in terms of capital cost): The most technological-ly advanced, fully automated ultrasound imagingsystems now commercially available still sell forabout $150,000, a price that is about one-fifth that ofsome present generation CT scanners. 9 A real-timeultrasound scanner costs around $50,000 (174). Inaddition, ultrasound units are smaller than CT, re-quire no elaborate installation, and are portable,allowing them to be used in areas such as the new-born nursery and the intensive care unit.

The second attribute of ultrasound that has alwaysmade it preferable to CT scanning (and other X-raymodalities) is that it is based on the physics of soundrather than radiation, and therefore does not imposethe risks associated with ionizing radiation (136). Thehigher growth rates for ultrasound sales observed in1979 data and projections through the coming decadesuggest a market trend toward imaging devices thatdo not use X-ray (91). In part, this may reflect theheightened public awareness of the harmful effects ofradiation (91). However, relative to CT, ultrasoundhas the limitation that it cannot penetrate bone andthus cannot be used to image the adult brain, and thelimitation that it cannot penetrate gas, and hencecannot be used to image structures surrounded bygas- filled loops of bowel (136). It likewise has no rolediagnosing disease in the lungs.

Although it is true that ultrasound is safer than CTbecause it does not use X-ray, more cautious observ-ers point to the possibility that ultrasonics may in-volve other risks yet unknown (86). Until recently,ultrasound has been presumed to be harmless, and itssupporters have insistently promoted it on this basis.Indeed, this assumption has been one of the primaryreasons for its near-routine application in the practiceof obstetrics. 10 Now, however, the possibility of riskentailed in using ultrasound is becoming the focus ofconsiderable concern particularly because of its prev-alent application in obstetrics (17). It is more realis-tic, and safer, to say that the risk associated with ul-trasonic energy is unknown rather than nonexistent.

The principles of ultrasonics can be clarified by adiscussion of one of the two types of ultrasound,pulse echo imaging (59). The key element of theultrasound system is the transducer that changes

—‘This IS th(, prlct, tlt t h e [latas~~n manufactured by C,enera I Electrlc and

Intr{duc ed ]n i\’c~\,em twr I Q7Q, wh]ch Intcgra tes the B-mtxie and r e a l t]mescanning capabllltles ( 123 ~

“The technology was t [rst found to be very usetul ]n obstetrics because ofthe fact that sound transmits ver} well through water, [Jr tluld such as thattc~und In the amn]t~t]c sac Its safety was assumed and later accepted basedt~n dccumulatl ng c1 I n]ca I t,xperlence Th}s appl]ca tl[>n IS based (~n t h e st)narprlnclpie It dl~pldy< !mages t~} sectl(lns [,t the human body S(J rapidly thatIn ternd 1 m{lvemen t ma} ht, cietec ted Its main uw IS rec(~rdl ng the tetal heartr~ te dU ring ldb(~r

voltage into high-frequency sound by means of apiezo-electric crystal: This crystal also has the capa-bility of picking up reflected sound and changing itback into electricity. This electronic input is thenconverted into visual data. There are several formatsfor display, not all of which provide a two-dimen-sional image. The format most closely approximatingthe X-ray view supplied by CT scanner is that of theB-mode compound scanning method which providesa two-dimensional, cross-sectional view of a bodytissue or structure (59). Its principal components in-clude the transducer, transmitter and receiver, digitalor analog processor, and display monitor. There mayalso be a television camera, video tape recorder, andrecord monitor so that image sequences of particularinterest may be recorded for later analysis (88). Anultrasound imaging system is shown schematically infigure B-2.

One problem with ultrasound is that the qualityand reliability of the images depends directly on theskill of the person operating the equipment. Recentrefinement and automation have not yet solved thatproblem (136). In addition, some observers have at-tributed the rather long developmental path of ultra-sound, as well as its consequent slow application andacceptance by practitioners in diagnostic capacities,to the fact that ultrasound has no “natural” constitu-ency among the medical specialties (153). There is nomedical specialty to which ultrasound is particularlygermane (outside obstetrics), although it is now beingapplied in many: Cardiology is one specialty withrapidly expanding applications of ultrasound (150).Many other specialities including ophthalmology,pediatrics, and neurology, are now acquiring theirown units, which helps explain the recent explosivegrowth (174).

Historically, ultrasound has received fairly largefunding support from NIH. Table B-4 shows Federalinvestment in 1975—present levels are probablycomparable. NCI has done some recent work: Twoprojects are currently in progress, each funded atabout $500,000 in fiscal year 1979 (153). NHLBI isalso investigating the use of ultrasound diagnostictechniques in cardiology. The most active researchunit for ultrasound applications at NIH is the Divi-sion of Radiology in the Clinical Center of NIH.Most of the work done in that division relates toclinical applications, but research has involved ad-vances in instrumentation as well. New equipment isbeing developed in the field of real time scanning(153).

In spite of these past trends in development andadoption into wider use in medical practice, how-ever, ultrasound is now being applied to a variety ofmedical problems outside its longstanding and now

68 . Policy Implications of the Computed Tomography (CT) Scanner: An Update

Figure B-2. -Principal Components of Real-Time B-Mode Ultrasound Imaging System

Transmitterandreceiver

b 1

SOURCE:IEEE Spectrum, January 1977, p 80

Table B-4.—Federal Funding for R&Din Ultrasonic ImagingDiagnostic Instrumentation (1975)

National Bureau of Standards . . . . . . . . . . . . . . ... . . . . . . . . . . . . . ... ... ... ... ..$ 100,000Department of Defense:

Army. . . . . . . . . . . . . . . . . . . , . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120,000Navy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ., . . . . . . . . . 285,535

Energy Research and Development Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80,000Department of Health, Education, and Welfare:

Food and Drug Administration . . . . . . . . . . . . . . . . . . . . . . . . . ... . . . . . . . 841,459Health Resources Administration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .National Institutes of Health:

25,000

National Cancer Institute. . . . . . . . . . . . . . . . . . . . . ., . . . . . . . . . . . . . . . . . . . . . . 418,514National Heart and Lung Institute. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2,851,165National Institute of General Medical Sciences. . . . . . . . . . . . . . . . . . . . . . . . . . . 1,530,166National Institute of Arthritis, Metabolic, and Digestive Diseases . . . . . . . . . . . . 118,964National Eye Institute. . . . . . . . . . . . . . . . ... , . . . . . . . . . . . . . . . . . . . . . . . . . . . . 439,297National Institute of Neurological and Communicative Disorders and Stroke. . . 379,905Division of Research Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20,000Division of Research Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50,000

Social and Rehabilitation Service ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24,851National Aeronautics and Space Administration. ., . . . . . . . . . . . . . . . . . . . . ., . . . . . . 360,000National Science Foundation ., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 818,850Veterans Administration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20,500

Total ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... ... .$8,484,206

aDoes not I nclu{je al I Intramural programs, whlc h are considerable

SOURCE: All lance for Englneerlng In Medicine and Biology, D/rectory of Federa/ Programs In Med/ca/ D/agrrost, c U/trasourrd(Chev{ Chase, Md , 1976).

near-routine use in the practice of obstetrics. Ap- glands), as well as fluid collections in the abdomenplacations of ultrasound have expanded to include (147). With the emergence of these new applications,studies of the brain, eyes, and various organs and ultrasound has become the diagnostic imaging mo-structures of the abdomen (including the liver, gall dality that is currently most often compared to CTbladder, spleen, pancreas, kidney, and adrenal scanning for studies of the abdomen.

Thus, ultrasound is being applied in diagnosticroles that compete with and/or complement thosetypically performed by CT scanners and/or other ra-diological diagnostic imaging modalities. Ultrasoundhas proved particularly popular in applicationswhere the risks associated with ionizing radiation areespecially high (as has always been the case with theuse of ultrasound is obstetrics). The recently pub-lished results of a clinical trial testing the efficacy ofusing ultrasound for breast cancer screening fortumor, 11 for example, showed ultrasound to be ableto accurately and reliably diagnose tumors of thebreast when they are fairly large (25).

The successful application of ultrasound for breastcancer screening for tumor would be significant inthat it would offer an alternative to X-ray mammog-raphy, a procedure for breast cancer screening thathas been the focus of much controversy. Breast can-cer screening was the topic of the first consensusdevelopment conference sponsored by NIH (24).Consideration of X-ray mammography, as used forscreening rather than diagnosis, was an importantpart of that conference. The risks and potentialbenefits of X-ray mammography screening were suchthat the panel recommended routine screening forwomen age 50 and over, but that women between 40and 49 years be routinely screened only if they haveeither a personal or family history of breast cancer,and that women under 40 years of age not be routine-ly screened unless they have a personal history ofbreast cancer (24).

Other potential applications of ultrasound may be-come increasingly important. One is carotid arteryscanning to diagnose occlusion (blocking) of theartery (156). Another is the use of ultrasound to char-acterize tissue such as liver (155) and pancreas (154)to diagnose such diseases as pancretitis. In terms ofits capability to diagnose some diseases, ultrasound isnot superior to or even equal to CT: For other dis-eases, the two may be about equal. In those caseswhere the images produced by each modality canenable accurate and reliable diagnosis, and onemodality involves irradiating the patient while theother does not, it stands to reason that the obviouschoice would be to avoid imposing the risk asso-ciated with radiation.

It is important to stress, however, that ultrasoundcannot be assumed to be harmless because no ioniza-tion occurs with the interaction of ultrasonic energyand human tissue. Rather, the associated risk is un-known and is cause for growing concern by morecautious observers (148). Proponents of ultrasound

‘ ‘These are the tlr>t cl]n~cal tr]als {~f th]s appllcat]~,n ot ultrast,und H(N+ever this equ]pment w,a~ marketed commerciall y a tew ~rears ag(>, and thereare al read} a t least three cc}m merclal m{xiels presently ava] I able ~ I S3 )

maintain that the risk is negligible, noting that noadverse effects attributed to ultrasound have beenreported by either obstetricians or pediatricians (59).Proponents further substantiate this claim by point-ing to the fact that the developing embryo or fetus istremendously susceptible to traumatic influences andthat such a fragile organism would be the first tomanifest any ill effects. Critics argue that the absenceof reported hazards does not constitute proof of safe-ty (59). Although there have been no adequate hu-man studies of the risk entailed in the use of obstetricpractice to date, experimental laboratory studieswith mice and primates have indicated a variety ofproblems as a direct result of using ultrasound at highlevels (148).

The possibility of risk associated with intrauterineexposure to ultrasonic energy is particularly poignantgiven its prevalence of use in the United States (59).Virtually every large labor room in the country isequipped with ultrasound for the purposes of moni-toring the fetus during labor; an application now re-garded as routine practice. More recently, there hasbeen a trend toward the routine use of ultrasound formonitoring the embryo and fetus in early stages ofgestation as well (59). The current high use levelsobserved and expanding routine application of ultra-sound may not be justified in terms of the benefits at-tained by the monitoring procedure (17).

Thus, at the same time that considerable concern isbeing expressed over the safety of ultrasound as ap-plied in obstetrics, it is being more liberally applied:The controversy over risks and benefits has placedthe technology at the center of a heated debate re-garding its appropriate use in obstetrics. Certainlythere is the potential for abuse in applying the tech-nology. At the least, unnecessary use of ultrasoundcould result in unnecessary costs. But at the worst, itcould result in unknown damage in a generation ofchildren. The controversy points out, in the mostdramatic way, a great need for basic informationregarding the safety and efficacy of ultrasound.

The Future of CT Scanning

As recognized by the honor of the Nobel Prize be-stowed on its originators in 1979, the CT scanner un-doubtedly remains a remarkable advance in diagnos-tic medicine. With CT technology now well beyondthe phase of basic R&D to which Federal funding

sources are primarily oriented, Federal funding hasrecentl y supported the R&D of new imaging technol-ogies such as NMR, PETT, and ultrasound, the capa -

“There are n i n e h u m a n stud]e~ but all are method{~l(,~]cal 1}, tlawed( 153).

bilities of which may exceed those of current CTscanners.

This is not to suggest that the limits of the CT scan-ner, based on the principles of radiology and CT,have been fully realized. But continued refinementsand improvements in CT technology are now morethe concern of those private companies that currentlyhave considerable vested interest in the future of CTscanning. Among the performance improvementsthat current CT technology may now be capable ofsupporting are subsecond, high-resolution, and/orthree-dimensional reconstructions (64). Proponentsof CT are confident regarding the continuing tech-nical evolution of the technology through the re-mainder of this century (64).

Meanwhile, however, private, as well as Federal,investments in the R&D of new principles of imaginghave resulted in the emerging technologies of NMRand PETT scanning. While the contribution of CTimaging to biomedical research and medicine was toprovide studies of anatomical structure in a nonin-vasive, automated mcde, the techniques of PETTand NMR provide studies of physiology and func-tion, and metabolism (respectively) in that samemode (151). In the case of NMR, these capabilitiesare particularly enticing since they are achieved via atechnique that does not involve radiation exposure,either from X-ray or from the administration of ra-dionuclides.

It is not only the new and emerging imaging tech-niques that are poised to present a challenge to CTscanners, but also other older techniques such asultrasound. Continued research on ultrasound diag-nostic imaging techniques has resulted in improve-ments in equipment and procedure that have broughtabout comparable diagnostic capabilities for certainconditions, as well as convenience of use for practi-tioners. At the same time, the appeal of diagnosticultrasound has been enhanced as increasing emphasishas been placed on cost and safety of equipment andprocedure. Diagnostic ultrasound, which is assumedto be not only safer than X-ray imaging techniquesbut also far less expensive even in its newest forms,can be expected to continue its competitive positionin medical practice, and consequently in the commer-cial marketplace.

Several indicators already reflect the increasingpreference and demand for less expensive, noninva-sive, non-radiation-emitting modalities such as ultra-sound. For example, an estimate of 1979 (worldwide)dollar volume sales in ultrasound imaging equipmentindicates an increase of 40 percent over sales in 1978,while an increase of 10 percent over 1978 was esti-mated for CT scanner sales (48). Projections of an-nual growth rates from 1979 through 1982 show an

extension of these trends: It is expected that sales ofCT scanners will continue to increase at an annualrate of approximately 10 percent, but ultrasound isexpected to show a 31-percent rate of increase peryear (48). Increased utilization of ultrasound in med-ical practice may also be expected in the comingyears based on such indicators as the papers pre-sented at the annual meeting of the Radiological Soci-ety of North America in November 1979 (134). Whilethe number of papers on applications of ultrasoundincreased from the previous year by about 15 per-cent, the number of papers on applications of CT de-creased by about 10 percent (91). It is already pre-dicted that the increased use of ultrasound in diag-nostic capacities enabled by recent refinements inultrasound technology will affect the future sales ofCT scanners (91).

Eventually, the diagnostic imaging instrumentmarket will be further altered by the introduction ofnew technologies. However, it also seems to be rec-ognized that these new imaging techniques will besubjected to an increasingly critical and extendedperiod of evaluation to establish efficacy, and alsocost effectiveness (64). Further, primarily because oftheir costliness, they will come under particularscrutiny as their diffusion and widespread utilizationin medical practice become imminent (143). If well-designed studies are not done of their clinical utility,two equally undesirable outcomes are possible:Rapid spread without demonstration of usefulness orconcerted attempts by Government to restrict diffu-sion without a good basis on scientific studies onwhich to rely.

Expectations surrounding the introduction and useof new technologies have given rise to a certain skep-ticism regarding their becoming generally available(64). Some believe that the cautious environment in-to which expensive new diagnostic modalities such asNMR will be introduced will have a net effect offavoring continued evolutionary changes in diffused,accepted technologies and procedures— in this caseCT scanners and scanning. The recent trends ob-served with respect to the development and use ofultrasound equipment and procedure are a pertinentexample of this prediction. 13

Ultimately, the way in which these new and im-proved diagnostic imaging modalities will competewith CT scanners in the marketplace will be deter-mined by the way in which their capabilities are used

“Although ultrasound M not nearly as capital ]ntenslve as either CT 1s, orNMR IS expected to be, expenditures associated with Its use now ri~al thoseassociated with CT because of the prev,ilent and frequent application ofultrasound in its current capacities. And as the diagnostic appllca tions ofultrasound expand, these expenditures can also be expected to rise, TheIssue of capital costs v. the cost associated with actual utlllzatlon (that ultra-sound so aptly Illustrates) is an Important one.

to complement, supplement, or replace CT scanningin medical practice. Essentially, this is tantamount tosaying that the clinical efficacy of CT scanning—aswell as that of emerging and improved diagnostic im-aging techniques—must be evaluated. The future ofCT scanning lies in determining what the potentialimpact of CT scanning can be and under what condi-tions these benefits can be attained. For example,comparisons of CT and ultrasound for abdominaldiagnoses show that CT is generally the better imag-ing technique for corpulent patients, ultrasound thebetter for thinner ones (153). Since ultrasound doesnot require the patient to be motionless, it is also bet-ter than CT for imaging very young, elderly, and agi-tated patients (153). To be sure, the principles under-lying each technique, as well as the attributes of theassociated procedure, will aid in determining the ulti-mate place that each will occupy in the practice ofmedicine.

Until very recently, CT has been used primarily asa supercapable X-ray machine. But new applicationsof CT need to be explored and refined if i t is toestablish a legitimate position relative to otherdiagnostic technologies that are being, or soon willbe, used in medical practice. Investigation of the useof CT in capabilities that lie beyond its traditionaldiagnostic role is especially important to extendingand establishing the boundaries of its domain. Exam-

ples of such applications are the use of CT in theplanning and delivery of radiation and chemother-apy treatment, and the monitoring of cancer patientsunder treatment (5,68, 75, 169). Another is its applica-tion in emergency medicine for head trauma (9 I ). Afinal example is the use of CT as a guide in biopsyingtumors, aspirating cysts, and draining abscesses ofthe brain.

Applications of CT outside conventional diagnos-tic roles will be important to establishing its clinicalefficacy. The benefits accruing from the use of CT intherapeutic capacities (e. g., in conjunction with ra-diotherapy for cancer) are more readily discernedthan those accruing from the use of CT in diagnosticcapacities, partially because they have a more directpotential influence on health outcome ( 102,183,185,186). These new applications could provide a broad-ened base for arguing the need for additional CTscanners. However, they have not yet been raised asa major issue in the current heated public debate sur-rounding the National Health Planning Guidelines.Presently, the evaluation of efficacy and cost effec-tiveness, the regulation of diffusion, and the financ-ing of CT scanning are primarily based on the ap-plication of CT in its diagnostic capacities. New ap-plications being made outside the diagnostic role inwhich CT was born, however, will be a Critical factorin determining its future.

Appendix C.— CT-Related Policies of the Department ofDefense and the Veterans Administration

The Department of Defense

The Department of Defense (DOD) Health Coun-cil, including the Surgeons General of the Army,Navy, and Air Force, and the Assistant Secretary ofDefense for Health Affairs, reviews all requests forcapital expenditures exceeding $100,000. Acquisitionof computed tomography (CT) scanners proceeds inthe same way for each of the three armed services.The Commanding Officer of a hospital will make theinitial request for a scanner to the Surgeon General’sOffice. Justification of the purchase request will statethe need for the scanner in terms of expected patientload, the beneficiary population, and geographic ac-cess. Local health systems agencies (HSAs) are con-sulted, since military requests are to be coordinatedwith those of local civilian facilities. The SurgeonGeneral will then bring the request and justificationbefore the Health Council for authorization. Fundingof authorized expenditures is a separate process, andmoney may not be appropriated for some time fol-lowing authorization of purchase. The time from ini-tial request to installation of a new scanner may be aslong as 2 years.

At last report, a decision had been made to pur-chase six scanners, two for each of the armed serv-ices. As of May 1980, the Army has two operationalscanners, and the Navy has four; the Air Force hastwo, and a third one is being installed. Since July 1,1977, when only one scanner was operational, sevenmore have been added for a total of eight scannersfor the entire DOD. All are total body scanners andare located in large medical and teaching facilities.Others have been approved and funded, but are notyet installed. Military hospitals and clinics withoutscanners continue to use CT scanning facilities ofcivilian institutions and pay for such scans out ofoperating budgets.

DOD is currently not supporting any major re-search related to CT scanning, although it does fundhealth research. The Air Force, however, is a partialsupporter of the dynamic spatial reconstructor imag-ing system being developed at the Mayo Clinic. Itsinterest in that project does not lie with the diagnosticcapabilities of the technology, but with the genericproblem of high-speed processing of information forimaging that is inherent in it.

The Veterans Administration

There has been a major shift in Veterans Admin-istration (VA) policy toward the purchase of CT

scanners for the VA system of hospitals. The pre-dominant means of obtaining scans for VA patientshas been for VA hospitals without scanners to con-tract with local civilian institutions and to pay forsuch scans out of operating budgets, As of August1978, a total of 14 scanners were either operational,being installed, or were ordered for the 171 VA hos-pitals. Currently, there are 17 operational scanners inthe VA system (5 head and 12 body), another 6 totalbody scanners have been purchased, and there are 2mobile scanners—for a total of 25 scanners in the VAsystem. As noted in chapter 2, a number of large VAhospitals do not have scanners.

The shift towards preference for purchasing CTscanners rather than contracting for services isjustified in a 1978 VA report by the Special CentralOffice Advisory Group for Computerized Tomogra-phy Units (179). The study compared the cost of ob-taining scans for its patients by these two methodsand found that the cost of performing a CT examina-tion on VA-owned and operated scanners was onlyabout 60 percent of the cost of the same exam ob-tained under contract from a civilian institution. Theconvincing cost differential suggests that the VA can-not afford not to buy its own scanners. This evi-dence, in combination with VA’s special character-istics of a fixed operating budget, and the legislativemandate it has to serve veterans, has led the VA toadopt a policy of purchasing CT equipment when-ever possible,

The report suggests optimal guidelines for the pur-chase of CT scanners under this new policy. Theseguidelines state that at least 500 exams per yearwould be required to cost justify purchase of adedicated head scanner costing $l50,000 or less;1,500 exams per year would be necessary to justify atotal body scanner costing $450,000 or less; and morethan 1,500 exams per year would be required to jus-tify one costing more than $450,000. At least 2,500exams per year would be necessary to cost justifypurchase of a second CT unit in a facility. Under nocircumstances would purchase of a CT scanner beconsidered if a VA or other federally owned scanneris within 30 minutes from the facility in question.

It should be emphasized that these guidelines areonly optimal. While more generous than the pro-posed National Health Planning Guidelines of 2,500exams per year for both a first and a second scannerin a given facility, and the VA is legally exempt fromthat standard, they are attempting to conform asclosely as possible. In practice, the VA is using a

72

guideline of about 2,000 exams per year as a thresh-old for considering a request for purchase of CTequipment. It is also continuing its policy of coordi-nating purchase decisions with existing local CTcapacities by seeking local HSA certification of need,although this review is not required by law.

Decisionmaking on purchase and placement of CTscanners is done centrally, although it is responsiveto local level requests. The VA Advisory Group onCT Units, including the directors of Medicine, Sur-gery, Neurosurgery, and Neurology, have developeda list of VA hospitals that are priority candidates forplacement of additional CT scanners in the VA sys-tem. At this time, purchase of the total body scannershave been given precedence so that only the biggest,most well-equipped and staffed, and busiest hospitalsare being considered for placement of these scanners.Being at the top of this list, however, does not ensurethat the hospital will receive a scanner. Local avail-able CT capacity may mediate this apparent neces-sity. Acting within the constraints of a fixed annualbudget, the decisions of purchase and placement ofCT units for the system must be based on the cri-terion of maximizing total VA scanning capacity, bywhichever means it may be obtained.

Purchase of scanners by the VA is itself a uniqueprocess. Once a decision has been made to buy a unitfor a particular facility, or facilities, the VA requestsbids from manufacturers quoting prices of modelsthat meet the particular specifications of the scannersrequired at each site. The VA accepts the lowest bidmade by manufacturer for a machine that meets thenecessary specifications. The large purchasing powerof the VA allows them to purchase scanners in thisway, typically resulting in a purchase price between$100,000 and $200,000 below the usual market price.Even so, the manufacturers bids seem very high inlight of the fixed annual budget within which the VA

must operate. Price tags quoted for total body scan-ners being considered for purchase by the VA at thistime were between $750,000 and $900,000. This mayexplain the fact that the Office of Management andBudget disapproved VA’s plan to buy nine CT scan-ners with year-end money in September 1980. How-ever, information recently collected by the VA froma number of its large hospitals indicates that the costsof buying outside CT scans has increased since the1978 report. It may cost as much as three times asmuch to obtain scans in this way.

Several alternatives are open to the VA as possiblesolutions to their apparent dilemma. One, the pur-chase of the new cheaper scanners, is not now beingconsidered by the VA. Such scanners are not believedto be adequate to the task required in the largehospitals now being given priority for placement ofCT units. Another avenue of approach recently usedwas the purchase of a refurbished EM I head scanner,originally costing $450,000, for $160,000. This scan-ner was purchased for the Palo Alto VA hospitalwhich had been ranked number one on the prioritieslist for placement of a total body scanner. Because ofthe nearby scanning capacities of Stanford-Univer-sity-owned body scanners, however, it had not beenapproved, A compromise solution was made where-by at least the head scans may be done in-house.

A final alternative that holds promise for the VA isthe purchase and operation of mobile scanners. Fi-nancially, such units are advantageous for the VAbecause they can use their own physicians, techni-cians, attendants, and vehicles. A research projectfunded by the VA is just getting underway to investi-gate this possibility. The study will evaluate the shar-ing of VA-owned equipment with four satellite hospi-tals in the Boston area. A radiologist at the centralVA hospital will read all scans performed by thisunit.

Appendlix D.— Indications for CT Scans (Exhibits 1 -4)

Exhibit 1:Draft Screenin Criteria for Body and Head Computerized Axial Tomography (CT)Scans. Memorandum from Director, Office of Professional Standards ReviewOrganizations, Health Standards and Quality Bureau, Health Care FinancingAdministration, DHEW, to Planning and Conditional PSROs, Statewide Councils andRegional PSRO) Officers, Feb. 22, 1979.

MEMORANDUM DEPARTMENT OF HEALTH, EDUCATION, AND WELFAREHEALTH CARE FINANCING ADMINISTRATION

TO

HEALTH STANDARDS AND QUALITY BUREAUOFFICE OF PROFESSIONAL STANDARDS REVIEW ORGANIZATIONS

: Planning and Conditional PSROs, DATE: FEB 22 1979Statewide Councils, RegionalPSRO Project Officers GENERAL MEMORANDUM NO. 3.79

FROM :

SUBJECT:

Director

Draft Screening Criteria for Body and Head Computerized Axial Tomography(CT) Scans

Attached are sample screening cr i ter ia for body and head CT scans col-lected from several sources by the Ad Hoc Computerized Axial Tomog-raphy Cri ter ia Commit tee of the American Associat ion of ProfessionalStandards Review Organizations (AAPSRO) . These critertia were accept-ed without change by the National Professional Standards Review Council(NPSRC) . PSROs may wish to adopt and adapt the criteria for local use.The criteria should be helpful to PSROs that now review CT scan proceduresor plan to do so.

Due to rapid developments in the f ie ld, the A A P S R O C o m m i t t e e r e c o m m e n d e d

t h a t t he c r i t e r i a be eva lua t ed in s ix m o n t h s f o r n e c e s s a r y r e v i s i o n s .This recommendation was approved by the NPSRC. To assist AAPSRO in thiseffort, please address your comments to Lloyd Cloud, DDS, Chief, AlliedHealth Branch. The mail ing address is Health Standards and Quali tyBureau, Dogwood East Building, 1849 Gwynn Oak Avenue, Baltimore,M a r y l a n d 2 1 2 0 7 .

Dennis F. Siebert

Attachment

NOTE: The PSRO standards (exhibit 1) are virtually identical to the Institute of Medicinestandards, so the latter are not reproduced here.

Criteria for CAT Head Scans

CAT scans of the head should be covered for the following signs,symptoms, and/or disease processes:

A. Symptoms - persistent symptoms after physical examination includingn e u r o l o g i c a l e v a l u a t i o n .

1 . H e a d a c h e o f s i g n i f i c a n t m a g n i t u d e2. P e r s i s t e n t v e r t i g o3 . Persistent seizures, adult onset; in the absence of d r u g /

alcohol withdrawal or recent head trauma4. Acute or progressive focal necrologic findings, when systemetic

or metabolic origin has been excluded, such as: a) apasiab) ataxiac) paresis ord) sensory deficit

5. Unexplained dementia; progressive organic mental deteriorationunexplained by systemic disease (e.g. , memory loss)

B. Physical Findings

1. Papilledema, or other signs of increased intracranial pressure2. Apraxia or aphasia3. Visual field defects4. Cerebella dysfunction signs5. Hemiparesis6. Exophthalmos after thyroid disease has been ruled out7. Other focal neurological signs

c. Unresolved Medical Problems

1. VascularSuspected Intracranial hemorrhage, such as:

a) subarachnoid hemorrhage,b) subdural hematoma,c) bleeding arteriovenous malformatd) bleeding aneurysm,e) complications of anticoagulation

in patient on Coumadin, Heparinf) intracerebellar or intracerebral

2. Traumatic

(e.g., Progressive headache

hematoma

Suspected lesion secondary to trauma (significant head injury)with progressive neurological findings.

3. Neoplastic

Suspected neoplastic lesion, such as: a) primary brain or meningealtumor or cranial nervetumor or

b) intracranial metastasisc) paranasal sinuses and

nasopharynx

4. Congenital Lesions

Congenital lesions, such as: a)b)

5. Calverial lesions (skull); lesionsx-rays.

6. Detection of cerebral metastasisthoracic surgery

)

hydrocephalusencephalocelesanomaly of brain

not fully defined by skull

in proven lung cancer prior to

7. Evaluation of effectiveness of treatment of documented cerebrallesion including:

subdural hematomaneoplasm, after surgery, radiation, and/or chemotherapyhematoma, arteriovenous malformation or aneurysmhydrocephalus, after shuntmanagement of brain abscessw h e n s i g n s a n d s y m p t o m s s u g g e s t p r o g r e s s i o n , r e c u r r e n c e , o r l a c ko f r e s p o n s e t o t h e r a p y

Appendix D-Indications for CT Scans ● 77

The following problems are generally not considered to be appropriatesituations for C.A.T. use and will be reviewed:

1. Vertigo as an isolated symptom.

2. Syncope as an iso

3. Migraine headache

4. Febrile seizures

5. Alcohol withdrawal,

evaluation.

ated symptom.

uncomplicated

n children under six years of age.

repeated, with seizures, upon initial

6. T.I.A. on hospitalized patients unless cerebral arterio-

graphy and surgical re-vascularization.

7. U n c o m p l i c a t e d m e n i n g i t i s

8 . A h e a d i n j u r y f o l l o w e d b y a t r a n s i e n t l o s s o f c o n s c i o u s -

ness (concussion) admitted for hospital observation which in

twenty-four hours resolves without persistent neurological

signs.

9. Completed Stable Cerebral Infarction (Stroke).

At the present time, indications for contrast studies varyaccording to the diagnostic problem and the judgment of the radio-

l o g i s t s a n d c l i n i c i a n s . Therefore, decisions concerning the use ofcontrast are not addressed in these screeninq guidelines.

Patients having more than three scans should be subject topeer review.

Criteria for Body CAT Scan

Neck -- CT scanning is not indicated at this time.

Chest --

● Pleura

--Detection of chest wall lesions.

.Lung

--Detection of more have been

pleural metastasis and other

multiple tumor modules where one crfound by conventional x-ray techniques.

--Search for a primary tumor when a positive sputumfor malignant cells has been obtained, but no evidencehas been found through conventional x-ray techniques.

--Determination of extent of spread tO adjacent lobesin patients with impaired pulmonary function.

--Differentiation of solid, cystic, fatty, inflammatoryand vascular masses.

--CT is not indicated for detection of pulmonary emboliat this time.

(If there is clearcut evidence of bilateral tumor in-volvement, CT is not appropriate. )

● M e d i a s t i n u m

- - D e t e c t i o n a n d e v a l u a t i o n o f m a s s e s .

- - D i f f e r e n t i a t i o n o f s o l i d , c y s t i c , f a t t y , i n f l a m m a t o r y ,a n d v a s c u l a r m a s s e s .

--Determination of extent of primary or secondary tumor.

H e a r t

--Studies of the heart are not indicated at this time

Great Vessels (including abdominal aorta)

-- CT scanning is not indicated in the aorta and great vesselsexcept in the few post-operative patients in whom aortic graftabscesses are suspected.

S p i n e a n d C o n t e n t s -

● Spinal Cord

--CT is not indicated for disease of the spinal corda t t h i s t i m e .

● Spinal Column

--Determination of content and extent of meningocelesand meningomyeloceles.

--Biopsies under CT guidance.

--Otherwise, CT scanning of the spinal column isindicated only where other procedures, includingconventional tomography, radionuclide scanning,and myelography have failed to detect primarytumors, metastasis, and inflammatory diseases int h e p r e s e n c e o f p e r s i s t e n t s y m p t o m s o r s i g n s .

Abdomen -

● Retroperitoneal Area

--Diagnosis and staging of nodal and extranodalextension of Iymphomas, determination of extent ofretroperitoneal involvement with lymphomas, and extentof other types of retroperitoneal metastasis fromvarious primary sites.

--Detection of primary malignancies such as those ofmesenchymal, neural, lymphatic, embryonic rest origin,melanomas, and benign conditions such as cysts whichmay mimic malignancies. Trauma with suspected retro-peritoneal hemorrhage.

Peritoneum -

--Detection and aspiration of abscesses and cysts.

Liver -

--Search for primary and secondary tumors and somelife-threatening benign lesions such as liver celladenomas and cavernous hemangiomas and abscesses.

--Determination of extent of tumor and differentiation

o f s o l i d , c y s t i c , i n f l a m m a t o r y , v a s c u l a r , a n d f a t t y

lesions.

-- Biopsies under CT guidance.

Spleen -

--CT ‘is not indicated at this time.

Pancreas

--Search for primary and secondary tumor. W h e nprincipal diagnostic consideration is pancreatictumor, CT should precede and when positive supplant suchless sensitive studies as upper GI, barium enema,liver and spleen scans.

--Determination of extent of tumor.

--Differentiation of solid cystic, inflammatory,vascular, and fatty lesions.

--Biopsies under CT guidance.

Kidney -

CT scanning of the kidney is indicated only when preceded bya conventional IVP study, and then for:

--Search for primary and secondary tumor.

--Determination of extent of tumor.

- - D i f f e r e n t i a t i o n o f s o l i d , c y s t i c , i n f l a m m a t o r y ,v a s c u l a r , o r f a t t y l e s i o n s .

--Biopsies or aspiration under CT guidance.

Gall Bladder -

--CT is not indicated at this time.

Biliary Tree -

--Differentiation of obstructive from non-obstructivejaundice in t h o s e c a s e s w h e r e c h o l e c y s t o g r a m a n d / o rultrasound fails to define cause.

Gastrointestinal Tract -(Stomach, Small and Large Bowel)

--CT is not indicated at present. Except for determinationof extent of tumor spread to other organs (see other indications).

A d r e n a l G l a n d s -

- - S e a r c h f o r p r i m a r y a n d s e c o n d a r y t u m o r .

--Determination of extent of tumor.

--Differentiation of solid, c y s t i c , i n f l a m m a t o r y ,

v a s c u l a r} o r f a t t y l e s i o n s .

--Biopsies under CT guidance.

Pelvis -

● U t e r u s a n d O v a r i e s

--CT scan is appropriate for the staging and evaluationof extent of tumors.

Indication for CT is limited and cases should besubject to individual review. Pelvic exam andul t rasonography should define most masses.

● Bladder, Ureter, Prostate, Testicles

--CT scan is appropriate for the staging and evaluationof the extent of tumors.

CT adds little information and cases should be subjectto individual review.

--Differentiation of solid, cystic, inflammatory vascular,or fatty tumors.

( F o r r e t r o p e r i t o n e a l p r i m a r y a n d s e c o n d a r y , s e e r e t r o -p e r i t o n e a l . )

Bones

--Evaluation of bone lesions.

--Biopsies under CT guidance.

Extremities -

--CT is indicated for determining the local extent of atumor and presence of regional metastasis.

Therapy Planning & Follow-up -

--CT may be indicated for collection of information on cross-sectional anatomy and attenuation coefficients of bone

and soft tissue in tumor-bearing areas for planningsurgery and radiation therapy.

--CT may be indicated in follow-up evaluation ofeffectiveness of radiotherapy, surgery, or chemo-t h e r a p y in cancer pat ients a t pr imary or metasta t ict u m o r s i t e s w h e n p a r t o f a n e s t a b l i s h e d a n d a c c e p t a b l efollow-up protocol or when signs and symptoms suggestprogression, recurrence or failure or therapy.

Foreign Body -

Foreign body localization anywhere in the body when otherconventional techniques have failed to resolve the problem(e.g., F.B. :orbit, globe of eye, intracranial or extremity).

C o n d i t i o n s fo r wh ich CT scanning is more hazardous than or d iagnost ica l lyi n f e r i o r t o o t h e r p r o c e d u r e s w e r e n o t i n c l u d e d i n t h e l i s t o f i n d i c a t i o n s .For some indications listed, other tests may be more appropriate inparticular patients. If other diagnostic tests have permitted a definitivediagnosis to be made, CT scanning IS justified only for planning treatment.

Conversely, if a CT scan establishes a definitive diagnosis, additionaldiagnostic tests are unjustified. Sometimes, tests may complement eachother either by providing difterent information or when one test succeedsafter the first has failed to yield useful information. Recent studiescomparing CT scanning with ultrasonic imaging of the abdomen suggeststhe two methods are complementary. (20)

Based on current evidence, CT is not superior in all applications. Fordynamic studies of the circulatory and digestive systems and for high-resolution radiography in which structural details below a millimeter mustbe discerned, CT cannot compete with conventional radiographic techniques.In mammography, for example, x e r o r a d i o g r a p h y p r o v i d e s d e f i n i t i v e d i a g n o s t i ci n f o r m a t i o n a t a l o w e r c o s t , a l t h o u g h a t a h i g h e r r a d i a t i o n l e v e l .U l t r a s o n i c i m a g i n g i s s a f e r , a n d , t h e r e f o r e , d i a g n o s t i c a l l y s u p e r i o r t oCT scanning in obstetrics and gynecology. In cardiology, TM mode andreal-time ultrasonic imaging provide more valuable data than do currentlya v a i l a b l e C T s c a n n e r s . C T s c a n n i n g c a n n o t r e p l a c e t h o s e n u c l e a r m e d i c a ltechniques that provide unique information about body functions andbody chemistry, as in the case of thyroid scans.

Because CT scanning of the body is an efficacious diagnostic tool forthe conditions listed above on the basis of current standards of evidence,the committee recommends that CT scanning of the body when used forappropriate indications be recognized as a covered service under third-party reimbursement plans until and unless a decision is made to requirem o r e d e m a n d i n g s t a n d a r d s o f e v i d e n c e f o r t h e s e d e c i s i o n s . However ,

e x p e r i e n c e w i t h b o d y s c a n n i n g i s e v o l v i n g r a p i d l y a n d t h e l i s t o f i n d i c a t o r sfor which coverage is warranted should be reviewed at least every sixmonths. Therefore, the committee recommends that:

● CT scanning of both the head and body, when

a p p r o p r i a t e l y u s e d f o r s p e c i f i e d i n d i c a t i o n ss h o u l d b e a c o v e r e d d i a g n o s t i c s e r v i c e u n d e rt h i r d - p a r t y r e i m b u r s e m e n t p l a n s , a c c e p t i n ga s c r i t e r i a o f e f f i c a c y t h e u s u a l s t a n d a r d so f c l i n i c a l p r a c t i c e .

As with any radiologic procedure, the clinician must exercise cautionin ordering number and extent of studies and repeat studies, sinceradiation dose varies widely with number of slices and area examined,and with the equipment used.

Exhibit 2:New Indications for Computed Body Tomography. Update to indications publishedin the April 1977 policy statement of the Institute of Medicine by the Society forComputed Body Tomography, July 1979. Published in the American Journal ofRoentgenology, July 1979.

SpecialReport

Society forComputed Body

Tomography

1979 American Roentgen Ray Society, Reprinted with permission.

New Indications forComputed BodyTomography

The Society for Computed Body Tomography has prepared the following listof indications for computed tomography in extracranial applications. These newguidelines are intended to clarify, update, and augment the indications published

INDICATIONS FOR CT AJR 133 , July 1979

Indications for Body CT

Neck

● Determinationsmatton of the extent of primary and secondaryneoplasms of the neck.

. Evaluation of bony abnormalities of the cervical spree

● Determination of extent of intrathoracic spread in selectedpatients with bronchogenic carcinoma including medias-tinal or pleural Invasion.

including neoplasms, fractures, dislocations, and congen-ital anomalies. Chest Wall

● Localization of foreign bodies in the soft tissues, hypo- ● Determination of extent of neoplastic disease.pharynx, or larynx and assessment of airway integrity — Assess bone, muscle, and subcutaneous tissuesafter trauma. — Detection of intrusion into thoracic cavity or spinal

● Evaluation of retropharyngeal abscesses canal

Mediastinum

● Evaluation of problems presented by chest radiograph— M a s s .

— Differentiation among cystic, fatty, or solid na-ture

— Localization relative to other mediastinal struc-tures.

— Mediastinal widening.— Assessment of whether cause IS pathologic or

anatomic variation— Distinction of solid mass, vascular anomaly, or

aneurysm, and physiologic fat deposition.— H i l u m

— Differentiation of enlarged pulmonary arteryfrom solid mass when conventional tomographyfails or IS not capable of making this distinction

— Paraspinal line widening.— Distinction among lymph node enlargement,

vascular cause, or anatomic variant

● Search for occult thymic lesion— Detection of thymoma or hyperplasia in selected

patients with myasthenia gravis when plain chestradiography IS negative or SUSPICIOUS

Lung

● Search for pulmonary lesions— Detection of occult pulmonary metastasis when

— Extensive surgery IS planned for a known pri-mary neoplasm with a high propensity for lungmetastasis or for apparent solitary lung metas-tasis.

— Detection of primary tumor in patient with positivesputum cytology and negative chest radiographyand fiberoptic bronchoscopy.

—— Assessment of lung and mediastinum for underly-ing pleural effusion and the postpneumonectomyfibrothorax for recurrent disease

● Search for diffuse or central calcifocation in a pulmonarynodule when conventional tomography IS Indeterminate

Percutaneous Needle Biopsy

● Assist biopsy of Iesions when fluoroscopic guidance in-inadequate.— Certain mediastinal masses.— Mass low in costovertebral angle or obscured by

overlying bone.

Heart

● Examinations of Intracardiac anatomy are not indicatedat this time. Future advances in CT equipment may allowmore clinically useful demonstration of cardiac anatomy

and physiology.● Distinction of cardiac (e. g., ventricular aneurysm) from

pericardiac (e.g., mediastinal or pulmonary Iesion) mass.● Detection of aortacoronary vein graft occlusion IS possible

with Intravenous contrast medium bolus with third- andfourth-generation scanners.

Major Blood Vessels

● Evaluation and detection of thoracic aortic aneurysms● Screening and measurement of abdominal aortic aneu-

rysms when ultrasound fails or IS unavailable● Detection of Intraluminal clots, chrome leakage, and rup-

ture of thoracic and abdominal aneurysms.● Evaluation of aortoprosthetic disruption● Evaluation of suspected infection of synthetic grafts of

the major vessels● Detineation of relation of major vessels to retroperitoneal

tumors, infections, or other abnormalities.● Demonstration of invasion of vena cava by tumor

Spine

● Type I examination. No contrast medium Type II exami-natlon: Dilute metrizamide. Type Ill examination Concen-trated metrizamide Installed originally for conventionalmyelography with subsequent CT, performed within 4hours after metrizamide instillation.

A.JR 133, JUIV 19 ?9 INDICATIONS FOR CT

●

●

●

●

●

●

●

Evaluation (type 1) of spinal stenosm to determine extentand specific causes of bony and soft tissue encroach-ment.— Diffuse Wmal stenosis, congenital or acquired.— Localized spinal stenows, associated with degen-

erative disease or malahgnment.— Posttraumatic stenosw: detection of fracture frag-

ments or hematoma.— Post:;pmal fwon stenosm: f u s i o n b o n e o v e r -

growth.— Detection of midline or foramenal spurs not seen on

plain films.— Combined causes mcludmg degenerate, latro-

gemc, traumatic, Infection/tumor, as well as her-mat!cms of the nucleus pulposus.

Evaluation (types I and 11) of congenital dysraphlc abnor-mal i t ies (splna blfida, menmgomyelocele, menmgocele,dlastematornyeha).Evaluation !type I or 11) of spinal cord and/or nerve rootmasses, usually as secondary procedure to further deter-mme nature and extent of Ieslon.Locahzatlor procedure (type O for CTqulded btopsy oraspwation.Evaluation (type 1) of nature and extent of boney orparaspinal tumors and Inflammatory masaea.Following nmdiagnostlc conventional myelography (typeI or II procedure) using myelogram and/or climcal findingsto spectfy CT level(s).Altemamre procedure (type 1) m Sltuatlons precludingstandard myelography as primary examination (allerglchwtory, mechanical dlfficultles, emOtlOnal faCtOrS).

Rotroperitonelum

●

●

●

●

●

DetectIon of pnma~ malignancies such as those of mes-enchymal, neural, Iymphatlc, and embryomc rest ongm,melanomas and bemgn conditions, such as cysts thatmay mimic nalignancles.Staging of nodal and extranodal extension of Iymphomasand other t@es of retropentoneal metastasis from var-IOUS primary sites (e g., mltlal staging or detection ofrecurrent metastatlc testicular tumor).Detectton of retropentoneal abscess or hemorrhage (he-matoma); localization for needle aspwatlon.Further evaluation when other radlologlc studies unex-pectedly suggest abnormality, such as dewated ureter bynormal relropentoneal fat.

Gutdance fcjr retropentoneal b!opsy

Peritoneum

● Detectton and differential dlagnosls of free or IoculatedIntrapentoneal flutd collections and inflammatory proc-esses.

● DetectIon clf Primary or secondary peritoneal masses(neoplasms and abscesses, etc.)

. Guidance for the asptratlon of Intrapentoneal fluld collec-hons and peritoneal masses.

Liver

. Evaluation of space+xcupymg lemons.— Primary and secondary malignant neoplasm and

clinically significant bentgn lesions, such as ade-nomas, cavernous hemanaiomas, and abeceasea.—

—

—

—

——

Initial detection; whether liver is primary organof interest or examined as part of CT evaluationof other suspected abdominal disease, such aspancreatic carcinoma, m which knowledge ofassociated hepatic Ieslons IS of clinical impor-tance.Confirmation of the presence or clarification ofthe nature of hepat!c Ieslon(s) suspected orfound on other !magmg procedure. such as anmconcluswe or nonspeaftc radlonuchde scan.Dlfferentlatlon of sohd, cystic, inflammatory,and vascular lesions.Assessment of location, extent, and number ofIestons, when such mformatlon IS of clinicalImportance.Guidance for hepatic biopsy and aspwatton.Assessment of response to ncmoperatwe ther-apy.

. Evaluation of trauma.— Detection of hepatlc Iacerat!on and mtrahepatic and

subcapsular hematoma, and determination of extentof Injury m cases of blunt or penetrating trauma

. Evaluation of diffuse liver dmeaae.— CT currently of limited value, but may be useful m

specific cwcumstances, such as cletection of fattyInflltratlon of the liver and conditions of exceasweIron depmwtion (hemochromatosw) a n d glycogenstorage disease m chddren.

Spleen

● Detection and estimation of age of subcapsular hema-toma.

● DetectIon of Intrasplemc mass and differentiation of solld,cystic, and inflammatory Ieslons.

Pancrees

●

●

●

Evaluation for possible mass Ieslon.— DetectIon of primary tumor and its extent.— Search for primary lesion m patient with distant

metastasis.— Evaluation of jaundiced patient.— Evaluation of suspected pancreatitis.— Evaluation of patient with posstble upper abdommal

masses.— Sertal assessment of regression or persistence of

tumor during and after therapy.Dlfferentlatlon of pancreatic from paraptmcreatlc mass.— Dwtmct!on among sohd, cystic, vascular, inflam-

matory, calclfied, and fatty lesions.DetectIon of comphcatlons of acute or subacute pancrea-tltls.

INDICATIONS FOR CT AJR 133, July 1079

— Detection of pseudocysts, thew number, size, andextent

— Serial assessment of paeudocyst followlng medicalor surgical management.

— Oetection of abscess: determination of stze andextent.

. Guidance of percutaneous pancreatic biopsy and aspi-ration procedures.

Kidneys

. Evaluation of kidneys when excretow urography or art-glography IS contraindicated by risk of serious reaction tocontrast medium.

. Evaluation of renal mass or suspected mass detected onanother Imagmg procedure.— Dtfferentlatfon of an anatomic variant from a patho-

logic process,— Dlfferentlatlon of a benign fluld-fdled cyst from a

cyst and/or solld renal mass.— Oetermmat!on of the extent of renal neoplasm be-

fore and after treatment.. Evaluation of selected patients, suspected clmlcally of

renal neoplasm, when excretory urogram IS negatwe.. Evaluation of juxtarenal (para- or perirenal) Iestons seen

or suspected on excretory urography— OMerentlatlon of anatom!c vanant from pathologic

process.— Oetermmahon of the cause, Iocatlon, and extent of

a Ieslon.● Evaluation of urograph!c nonfunctlonmg ktdney(s)

— Assessment of stze, outline, and parenchymal thick-ness.

— DetectIon of obstruction, determination of site,cause, and extent of disease process

— Documentation of congemtal absence.— DetectIon of mlnlmally calclfled r e n a l calculi n o t

demonstrated by conventional techmques.. Oetermlnatlon of cause of renal and perlrenal calcifica-

tion● Assessment of extent of renal trauma.. Guidance for antegrade nephrostomy, renal biopsy, or

mass as~lratlon.

Gallbladder

s CT IS not Indicated at th!s time unless oral and IntravenouscholecystograPhY and ultrasonography are mdetermmateor unobtainable

Biliary Tree

. Dlfferentlatlon of obstructive from nonobstructwe jaun-dice.

● Determlnatlon of site and etiology of obstruction. Determlnatlon of etiology of obstruction

Gaatrointeatinal Tract

. CT IS useful m the assessment of extent or recurrence oftumor or tumorhke condition mto the mesente~ or adja-cent Ofgans. CT is not currently indicated for the detectionof mucosal Ieslons.

Adrenai Gland

. Evaluation of patients with biochemical ewdence of ad-renal hyperfunction

. Evaluation of patients with susplcton of adrenal massfound on conventional radiographic exammatlon.

● Guidance for adrenal biopsy.

Uterua and Ovariee

● Evaluation of mass detected by cltnlcal examination, afterposmve biopsy, after fatlure of ultrasound examtnatton, orwhen strong chnlcal suspicion exists for a mass lesion.

● Evaluation of primary tumor and Its extent of spread: andevaluation of secondaw tumor.

. Olfferentlatlon of solid, cystic, inflammatory, vascular, orfany masses.

● Guidance for uterine and ovanan btopsy.

Bladder, Ureters, Proatate, and Seminal Veaiciea

● Evaluation of primary and secondary tumor, Includingextent of tumor.

● Dlfferentlatton of solid, cystic, inflammatory, vascular, orfatty tumors.

. DetectIon of obstructing, mlrvmaily calctfled ureteral cal-CUII not detected by conventional studtes

● Guidance for biopsy

PISIVIC Bones

● Evaluation of bone Ieslons and accompanying soft tissueextent

. Guidance for biopsy.

Muacuioakeietal System

●

●

●

●

●

Evaluation of selected patients with known or suspectedpnma~ bone tumors.Evacuation of patients with suspected recurrence of bonetumors.Evaiuat!on of patients with suspected but Indefinite signsof skeletal metastasis when Conventional studies fall toclarify.Evaluation of joint abnormalities dlfflcult to detect byconventional methods.Evaluation of patients with soft tissue tumors, either

AJR 133. Jutv 1979 INDICATIONS FOR CT

known or suspected to confwm presence and deterrmneextent.

● Guidance for biopsy.

Therapy Planning and Followup

Defmmon of cross-sectional anatomy and attenuationcoefftclen of bone and soft tmsue m tumor-bearmg areasfor the purpose of planning radiation therapy.Prowslon of baseline prior to radlatton therapy andchemotherapy from whtch effectweness of these treat-ment modallt!es can be judgedConformance as part of an estabhshed and acceptable

●

follow-up protocol.Evaluation of s!gns and symptoms suggesting progres-sion, recurrence, or fadure of therapy.

Foreign Body Localization

● In chest and abdomen when other traditional lmagmgtechmques prowde mrsufficlent Information.

ACKNOWLEDGMENT

The Soclefy IS greatly mdebled to Dr Peter L.wmgston, HollywoodMemorial Hospital, Florida for hm input and Joan Collie for her

assistance m manuscript preparation

Members and Officers of the Society for Computed Body TomographyRalph J, Alfldl, M.D.Case Western Reserve Umverslty School of Medcme

Ronald G. Erens, M.D.Mallmckrodt Institute of Rad!ology

Wilham Gler n, M D.Long Beach Memorial Hospital

John Haaga M DCleveland C Inlc

George S Harell, M.D.Stanford Un verstty Medical Center

Robert R Hattev, M.D.Mayo Clmlc

Ehas Kazam, M D.New York Hmspltal-Cornell Medcal Center

Stuart S. Sagel, MDMalhnckrodt Institute of Radiology

W Frederick Sample, M.D.UnwersNy of Cahfomla Center for Health Sctences

Dieter Schelhnger, M.D.Georgetown Unwerslty School of Medicine

Patrick F Sheedy, 11, M.D.Mayo Cltnlc

Robert J. Stanley, M.D.Mallmckrodt Institute of Radtology

Dawd Stephens, M.D.Mayo Clime

Jack Wittenberg, M.D.Massachusetts General HospNal

Melvyn Korcbkln, M DDuke Umverslty School of Medlcme

Exhibit 3:New Policy CT Approved. Policy statement by the American College of Radiology,October 1980. Published in the ACR Bulletin, October 1980.

American College of Radiology, reprinted with permission.

New Policy on CT ApprovedComputed tomography is a proven radiologic modality which provides valua-

ble clinical information in the early detection, differentiation and demarcation-of

disease.Abundant documentation of its safety and diagnostic efficacy has been pre-

sented in the scientific literature, It has totally changed the practice of radiologyand has become the primary diagnostic modality for a variety of presentingproblems. it is even more widely accepted as a supplement to other imagingtechniques. It is particularity helpful in solving problems where there is conflict-ing information from other radiologic or laboratory studies. It frequently re-places other examinations, many of which carry greater discomfort and expense.

APPROPRIATE SUPERVISION OF CT FACILITIES1) Computed tomography IS a form of medical imaging which, like other x-ray

and radionuclide procedures, involves the exposure of patients to ionizingradiation. Its use should be limited to physicians with the necessary trainingin radiation protection to optimize examination safety. Radiation physicssupport and a trained technical staff must be provided.a) Necessary training in radiation protection to optimize examination safety

should include formal structured didactic and practical courses in radia-tion physics, monitoring and safety including actual experience in the useof radiation monitoring equipment and the design and use of equipmentfor radiation protection. Certification by the American Board ofRadiology would be acceptable as verification of this level of competencefor a physician,

b) Radiation physics support should include regular, periodic inspection,and performance and quality testing of both the scanning equipment andthe protection devices. Access to the consultative services of a qualifiedradiation physicist should be readily available at all times

c) At least one registered radiologic technologist should be present at ailtimes during the actual operation of the scanning equipment for patientuse. Trained technologists with prior experience in operation of theequipment must be available in sufficient numbers to allow patient accessto the equipment over a broad range of time.

2) The different imaging procedures now available in diagnostic radiology, i.e.,angiography, ultrasound, radionuclide imaging and computed tomographymake selection and interpretation of appropriate studies for a specific patientmore complex. A referring physician should consult with a radiologist withexperience in imaging with ail available modalities concerning the pro-cedures and sequence best suited to answer a specific clinical question.

Such experience in the selection, execution and interpretation of appropri-ate imaging procedures is usually obtained during radiological residencytraining or fellowships and supervised clinical experience following suchradiological training. Any training program or experience should emphasizegross and cross-sectional anatomy, radiation absorption characteristics of theinvolved tissues and the pharmaco-physiology of contrast media in the organsystems of interest,

3) Each computed tomographic procedure must be individually designed for theproblem presented by the patient. The radiologist should be directly involvedin the performance of the procedure, determining its extent, administeringand/or supervising the use of contrast media and modifying the study afterimmediate interpretation of the initial images obtained.

Direct involvement in the performance of scanning procedures requires:a) Selection of the appropriate scanning sites, levels and sequences for each

patient.b) Determination of the need for contrast media (if any), the type and quan-

tity of contrast, and the route and method of administration.C) Administration or supervision of the administration of the contrast.d) Recognition and proper treatment of any and all contrast reactions

Council OKs New CT Concepte) Modification of the procedure after viewing the initial scans in order to

optimally demonstrate the appropriate findings in each case.ALL of the elements listed are equally necessary regardless of the anatomicalarea being examined.

APPROPRIATE UTILIZATION OF CT SCANNERS

The diagnostic efficacy of the CT is no longer in question. In general,guidelines for its utilization are based upon:1) Determining the site, type and extent of disease.2) Immediate diagnosis of trauma and other medical emergencies.3) Problem solving in patients when conflicting information exists.4) Radiation therapy planning and monitoring.5) Follow-up of treatment results.

6) Guidance for biopsy control.

There are many specific clinical areas in which CT is recognized as a necessaryand definitive diagnostic modality. However, specific indications for CT scan-ning should be determined locally by hospital medical staffs or other recognizedpeer review groups.

APPROPRIATE DISTRIBUTION OF COMPUTEDTOMOGRAPHY SCANNERS

Adequate distribution of safe and reliable CT scanning service is necessary toassure accessibility to appropriate and equitable medical care for all patients. CTscanners should be located in facilities which permit their availability to patientsof all physicians. No one set of criteria meets the dual requirements of medicalneed and economic justification. Determination of need for a CT scanner shouldbe made at the local level.

In addition to demographic and geographic factors, special considerationsshould be given to the capabilities and demands of the medical community.Teaching and research centers, regional medical facilities, cancer treatment pro-grams, neurological facilities and trauma centers all have a demonstrated needfor CT scanning capabilities. The existence of any of these may produce numeri-cal relationships between CT scanners and need indicators at a variance fromcommunity criteria based upon population or utilization projections.

The location of a CT unit in a physician’s office should meet the same criteriaas for institutions. Instances when a CT unit is located other than in an acutecare hospital should be infrequent due to the interdependency of CT scanningand specialized medical services.

The economic justification of a CT scanner depends upon sufficient patientdemand to allow reasonably full utilization of the unit. Capital costs are high anddepreciation over five years is prevalent. Operating costs are substantial, particu-larly when CT scanners are operated and available outside of normal workinghours. Provision must be made for updating and replacement of obsolescentequipment. Utilization goals should be directed toward optimal performance,allowing for maintenance, research and patient handling. They should not be setso high as to generate marginal patient referrals or to impose unrealistic workingconditions for staff and supporting institutions.

SUMMARY

When appropriately located, properly utilized and correctly supervised, CT scan-ners can have a positive impact upon both the cost and the quality of medicalcare. Through reduction or elimination of hospital stays, replacement of otherexpensive and more hazardous diagnostic studies and avoidance of some opera-tive procedures, CT scanning can contribute positively to cost containment. Ear-lier and more precise diagnoses will provide opportunities to modify therapeuticapproaches which may be expected to improve the outcome and/or the qualityof life. Access to this technology must therefore be assured to all patients whomay be expected to benefit from CT scanning.

Exhibit 4:Criteria for Clinically Indicated Head Scans. Used by the Colonial VirginiaFoundation for Medical Care, Virginia Beach, Va., in a concurrent review of CT ofthe head in PSRO Area V, Va., 1980.

Criteria for Clinically Indicated Head Scans (16)

Criteria1—Suspected Intracranial Hemorrhage2—Suspected Lesion Sec. Head Trauma3—Suspected Neoplastic Lesion4—Congenital Lesions5—Skull Lesions Undefined by X-Ray6—Detection Cerebral Met. Before Tharacic Surgery7— Eval. Treatment of Documented Cerebral Lesion8—Delineation of Residual Structural Abnormality After Neur. Disease, Injury9—Papilledema, Obtundation or Coma

10—Apraxia or Aphasia11 —Visual Field Defect12—Ataxia, Nystagmus, Dsmetria, Tremor or Incoordination13— Hemiparesis or Hemiplegia14—Exophtalmos After Thyroid Disease Ruled Out15—inequality of Pupils, Ocular Palsies, Ocular Ptosis16—Proptosis (suspected orbital tumor)17— Headache18—Persistent Vertigo Unresponsive to Outpatient Management19—Seizures (grand mal or complete, focal or partial, with altered level of

consciousness)20—Acute Focal Neurological Symptoms21 —Unexplained Dementia, Progressive Mental Deterioration22— Does Not Meet Criteria

Appendix E.— National Guidelines for Health Planning:Standards for CT Scanners

(Federal Register, March 28, 1978)

Section 5121.210Computed Tomographic Scanners

(a) Standard. A computed tomographic (CT) scan-ner (head and body) should operate at a minimum of2,500 medically necessary patient procedures peryear, for the second year of its operation andthereafter.

(2) There should be no additional scanners ap-proved unless each existing scanner in the healthservice area is peforming at a rate greater than 2,500medically necessary patient procedures per year.

(3) There should be no additional scanners ap-proved unless the operators of the proposed equip-ment will set in place data collection and utilizationreview systems.

(b) Discussion. Because CT scanners are expensiveto purchase, maintain, and staff, every effort must bemade to contain costs while providing an acceptablelevel of service. Intensive utilization of existing units,regardless of location, will prevent needless duplica-tion and limit unnecessary health care costs.

Estimates and surveys for efficient utilization ofCT scanners range from 1,800 to over 4,000 patientprocedures a year. (One patient procedure includes,during a single visit, the initial scan plus any neces-sary additional scans of the same anatomic area ofdiagnostic interest. )

The Institute of Medicine, the Office of Technol-ogy Assessment, and others have carefully reviewedthese data and the capabilities of various availableunits. The Department has reviewed these analysesas well as the extensive literature that has beendeveloped on CT scanners.

In arriving at a standard for the use of thesemachines, the Department has considered a varietyof factors, including the difference in time requiredfor head scans and body scans, the need for multiplescans in some patient examinations, variations in pa-tient mix, the special needs of children, time requiredfor maintenance, and staffing requirements. More-over, the Department considered the actual operatingexperience of hospitals and institutions reflected inreports on the use of CT scanners.

The standard set in the Department’s guidelines isintended to assure effective utilization and reason-able cost for CT scanning. These machines areexpensive, and therefore must be used at levels ofhigh efficiency if excessive costs are to be limited.

The Department recognizes that the cost of somemachines is declining, particularly those that performonly head scans which require less time. For ma-chines that do predominantly head scans, the stand-ard represents an efficient but more easily attainablelevel of utilization.

For scanners capable of performing both head andbody scans, it is imperative that they be effectively

used in order to spread the high capital expendituresover as much operating time as possible. As the In-stitute of Medicine report stated, “The high fixed costof operating a scanner argue for as high a volume ofuse as the equipment allows without jeopardizing thequality of care. ”

The Department believes that a .50- to 55-houroperating week is both consistent with the actualoperating experience of many hospitals and a reason-able target. Based on reported experience for the timerequired for both head scans and body scans, theDepartment estimated that a patient mix of about 60-percent head scans and about 40-percent body scans,making allowance for the other factors identifiedabove, would allow a CT scanner to perform about2,500 patient procedures per year if it is efficientlyused about 50 to 55 hours per week. This estimateassumes a higher percent of body scans than is cur-rently being performed. If fewer than 40-percentbody scans are performed, then 2,500 patient proce-dures would involve less than 50 to 55 hours perweek. Basing the standard on a higher percentage ofbody scans also takes account of current trendstoward increased proportions of such scans.

The Department believes that sharing arrange-ments in the use of CT scanners is desirable, in linewith the national health priorities of section 1502.Individual institutions or providers should notacquire new machines until existing capacity is beingwell utilized.

In planning for CT scanners, the health systemsagency should take into consideration special cir-cumstances such as: 1) an institution with more thanone scanner where the combined average annualnumber of procedures is greater than 2,500 per scan-ner although the unit doing primarily body scans isoperating at less than 2,500 patient procedures peryear; 2) units which are, or will be, devoting a signifi-cant portion of time to fixed protocol institutionally

approved research projects; and 3) units which are,

or will be, servicing predominantly seriously sick and adequately plan the distribution and use of CTpediatric patients. scanners in the area. The data to be collected should

A summary of the data collected on CT scanners include information on utilization and a descriptionshould be submitted by the operators to the appro- of the operations of a utilization review program.priate health systems agency to enable it to

Appendix F.—Amendments to Regulations GoverningReviews of CT Scanners Under 1122 and CON Programs

(April 25, 1979)

Review of Proposed CapitalExpenditures for CT Scanners Underthe Capital Expenditure Program ofSection 1122 of the Social Security Act42 CFR Part 100

Inclusion of Computed Tomographic ScanningServices.

Agency: Public Health Service, HEW.Action: Interim regulations.Summary: This notice sets forth interim rules re-

garding reviews of proposed capital expenditures forcomputed tomographic (CT) scanner “services” un-der the capital expenditure review program of section1122 regulations with minor revisions, a polic y

notice on this matter which has already been issuedby the Department. Interested persons are invited tosubmit written comments and suggestions concerning

these interim rules.Dates: These regulations are effective on April 25,

1979. Comments must be received on or before June25, 1979.

Address: Interested persons may submit writtencomments on these interim regulations to the ActingDirector, Bureau of Health Planning, Health Re-sources Administration, Center Building, Room6-22, 3700 East-West Highway, Hyattsville, Md.20782. The comments will be available for public in-spection at the above address between the hours of8:30 a.m. and 5:00 p.m., Monday through Friday.

For Further Information Contact: Colin C. Riorrie,Jr., Ph. D., Acting Director, Bureau of Health Plan-ning, 3700 East-West Highway, Center Building,Room 6-22, Hyattsville, Md. 20782, 301 /436-6850.

Supplementary Information: Section 1122 of theSocial Security Act (42 U.S. C. 1320a-1) provides fora program for reviews of certain proposed capital ex-penditures by designated planning agencies (DPAs)in participating States to determine their conformitywith applicable health plans, standards, and criteria.Subject to certain procedural requirements, the De-partment will not provide reimbursements, under themedicare, medicaid, and maternal and child healthprograms for expenses related to capital expendituresfound by DPAs to use out of conformity with theseplans, standards, and criteria. Section 1122(g) of theSocial Security Act defines a capital expenditure sub-ject to review as one which under generally accepted

accounting principles is not properly chargeable asan expense of operation and maintenance, andwhich: 1 ) exceeds $100,000 or 2) changes the bedcapacity of the facility with respect to which the ex-penditure is made, or 3) substantially changes theservices of the facility with respect to which the ex-penditure is made. The third of these categories isfurther defined in the regulations under section 1122(42 CFR part 100) as including an expenditure “whichresults in the addition of a clinically related (i. e.,diagnostic, curative, or rehabilitative) service notpreviously provided in the facility . . . (42 C F R100. 103(a)(2)(iv) ).

On February 3, 1978, the Department issued sec-tion 1122 Notice 78-05 to clarify the requirements ofsection 1122 with respect to CT scanner services. Thepurpose of these interim regulations is to incorporatethat policy notice with minor revisions, into the sec-tion 1122 regulations, to the extent it is not already apart of these regulations.

The Department recognizes that the existing reg-ulations do not explicitly include all aspects of theFebruary 3 notice and accordingly amends the reg-ulations so that they will, subject to the followingrevisions. First, because the Health MaintenanceOrganization (HMO) Amendments of 1978 (PublicLaw 95-559) deleted from section 1122 all referencesto HMOs, expenditures by or on behalf of an HMOare no longer subject to review, unless they are alsoon behalf of a health care facility which is subject toreview. Thus, if an HMO proposes to purchase a CTscanner on behalf of a hospital, the proposed expend-iture is subject to review. Second, the regulationsspecify that the proposed expenditure for a CT scan-ner by or on behalf of a health care facility is subjectto review, whether it is for a fixed or a mobile CTscanner. Third, the purchase of an additional CThead scanner by or on behalf of facility whichalready has such a scanner is not subject to review ifit costs less than $100,000 because this is a servicewhich was “previously provided in the facility. ” (See

Accordingly, the regulations are amended as setforth below, so that the acquisition of a CT scannercosting $100,000 or less will be governed by the fol-lowing principles:

A. The purchase of a CT scanner by or on behalfof a health care facility involving a capital expend-

iture less than $100,000 is subject to review if it re-sults in the addition of a new diagnostic service.

B. Such a purchase involving a capital expenditureof less than $100,000 is considered to be the additionof a new diagnostic service unless the CT scanner isin addition to or replaces an existing scanner.

C. If a health care facility proposes to purchase aCT full-body scanner to replace an existing CT headscanner or to purchase equipment to upgrade an ex-isting head scanner to a full-body scanner, this pur-chase is considered to result in a new diagnostic serv-ice for the facility and, therefore, is subject to section1122 review.

D. If a health care facility proposes to offer theservices of a mobile CT scanner, this is consideredthe addition of a new diagnostic service unless it is inaddition to or replaces an existing fixed or mobile CTscanner of the same type (head or full-body scanner).

E. The lease (acquisition through a comparable ar-rangement) or the donation of a CT scanner by or onbehalf of a health care facility is also subject to sec-tion 1122 review if its purchase, under the principlesnoted above, would have required review. (See 42CFR 100.103 (b).)

F. Any capital costs associated with installing a CTscanner, as well as the costs of any renovations to ac-commodate its installation or use, are to be includedin the estimated cost of the proposed capital expend-iture under the section 1122 review program.

In relation to these regulations, attention is calledto another interim regulation, also being issued inthis edition of the Federal Register, which amends 42CFR parts 122 and 123 to require review of fixed andmobile computed tomographic scanners in satisfac-tory certificate of need programs under title XV ofthe Public Health Service Act.

For the reasons set forth below, the Secretary hasdetermined that public participation in rulemakingbefore issuance of these regulations and a delay intheir effective date would be impracticable, unneces-sary, and contrary to the public interest. First, this isin large part simply a clarification of the Depart-men t’s interpretation of the existing regulations. Sec-ond, given the recent proliferation of CT scanners, adelay in implementing these revisions and clarifica-tions would likely result in the purchase or otheracquisition of scanners which are not needed. Third,because these amended regulations complementamendments to regulations governing certificate-of-need reviews under title XV of the PHS Act andbecause those regulations are being issued on an in-terim basis to give those States needing revised legis-lative authority the maximum time possible to obtainit. Proper coordination of reviews requires that theseregulations also be effective upon publication. As

noted above, however, the public is invited to submitcomments on these amended regulations during thenext 60 days, and the Secretary will revise the regula-tions further as warranted by his evaluation of thecomments received.

The Assistant Secretary for Health, with the ap-proval of the Secretary of Health, Education, andWelfare, amends 42 CFR Part 100 as set forth below.

Dated: January 30, 1979. ●

Julius H. Richmond, Assistant Secretary forHealth.

Approved: April 16, 1979.Joseph A. Califano, Jr., SecretaryS e c t i o n 1 0 0 . 1 0 i s a m e n d e d b y a d d i n g a t

its end the following sentences:

(a) . . .(2) . . .(iv) . . . The addition of CT scanner services not

previously provided in or through the facility is asubstantial change of services within the meaning ofthis subparagraph, whether these services are pro-vided through a fixed or mobile CT scanner. The ad-dition of CT full-body scanner services is included inthe previous sentence if it is added to or replaces ex-isting CT head scanner services.

(Sec. 1122, Social Security Act, 86 Stat. 1386 (42U.S. C. 1320a-1): sec. 1102, Social Security Act, 49Stat. 647, as amended (42 U.S. C. 1302). )

(FR Doc. 79-12637 Filed 4-24-78, 8:45 a.m. ) BillingCode 411 O-83-M.

Reviews of Proposed New InstitutionalHealth Services by SHPDAs and HSAsUnder State CON Programs

42 CFR Parts 122 and 123

Inclusion of Computed Tomographic ScanningServices.

Agency: Public Health Service, HEW.Action: Interim regulations.Summary: The Assistant Secretary for Health,

with the approval of the Secretary of Health, Educa-tion, and Welfare, proposes to amend the regulationsgoverning reviews of proposed new institutionalhealth services by State health planning and develop-ment agencies (SHPDAs) and health systems agencies(HSAs). These regulations set forth requirements forsatisfactory State certificate-of-need (CON) pro-grams. The amendments would require review ofradiological diagnostic health services which are pro-posed to be offered in, at, through, by, or on behalfof a health care facility or health maintenanceorganization, which are to be provided by fixed or

mobile CT scanning equipment whether or not an ad-dition to or replacement of these services offered byfixed or mobile CT equipment. Interested parties areinvited to submit written comments and suggestionsconcerning these proposed amendments.

Dates: These regulations are effective on April 25,1979, subject to the discussion set forth under “Sup-plementary Information” below. Comments receivedon or before June 25, 1979, will be considered.

Address: Interested persons may submit writtencomments on the interim regulations to the ActingDirector, Bureau of Health Planning, Health Re-sources Administration, Center Building, Room6-22, 3700 East-West Highway, Hyattsville, Md.20782. The comments will be available for public in-spection at the above address between the hours of8:30 a.m. and 5:00 p.m., Monday through Friday.

For Further Information Contact: Colin C. Rorrie,Jr., Ph. D., Acting Director, Bureau of Health Plan-ning, 3700 East-West Highway, Center Building,Room 6-22, Hyattsville, Md. 20782, 301/436-6850.

Supplementary Information: Section 1523(a)(B) ofthe Public Health Service Act (“the Act”) requireseach fully designated SHPDA to “administer a StateCON program which applies to new institutionalhealth services propose to be offered or developedwithin the State and which is satisfactory to theSecretary” of Health, Education, and Welfare, Sec-tion 1523(a)(5) of the Act requires each SHPDA tomake findings as to the need for proposed new insti-tutional health services, after consideration of recoin-”mendations submitted by HSAs. Section 1513(f) re-quires HSAs to assist SHPDAs by reviewing and

making recommendations concerning the need forproposed new institutional health services. Institu-tional health services are defined in section 1531(5) of

Code of Federal Regulations. Sections 122.304 and123.404 establish thresholds for review of institution-al health services. The amendments add an addition-al threshold: radiological diagnostic health servicesprovided by fixed or mobile CT scanning equipment.

The Secretary, wishing to contain the rising costsof health care, is concerned with the high cost of CTscanning, both as to the initial cost of the equipmentand its subsequent operating expenses, These amend-ments require review of CT scanning services pro-posed to be performed in space leased or made avail-able to any person by a health care facility or HMO.As to mobile scanners, although it is realized thatthere may be cost saving, since a number of institu-tions may share the same equipment, the offering by

a health care facility or HMO of services of mobilescanners is also subjected to review by health plan-ning agencies.

The Secretary notes that the amendments do notnecessarily require changes in any State’s statutes orregulations or in any lists of services which may beembodied therein. The amendments require simplythat, for a State CON program to be satisfactory}’, theproposed services be required to be reviewed, andthose services not be offered or developed without aprior determination of need and issuance of CON.

In relation to these amendments, the Secretarycalls attention to another interim regulation, also inthis edition of the Federal Register, which amends theregulations for review of proposed capital expendi-tures under section 1122 of the Social Security Act toclarify the coverage of CT scanners under that pro-gram.

Effective date provisions. -–For the reasons setforth below, the Secretary has determined that publicparticipation in rulemaking before issuance of theseregulations and a delay in their effective date wouldbe impractical and contrary to the public interest.First, given the recent proliferation of CT scanners, adelay in implementing these regulations would likelyresult in the purchase or other acquisition of scannerswhich are not needed. This is especially true withregard to mobile CT scanners. Hospitals can now re-ceive reimbursement through medicare for fixedscanners, but not for scans from mobile scanners.Medicare, however, will soon begin reimbursing hos-pitals for scans from mobile scanners as well. As aresult, it is expected that sales of mobile scanners willincrease significantly. Second, in order to give thoseStates which need revised legislative authority to im-plement these amendments the maximum time possi-ble to obtain it, these regulations should be effectiveimmediately.

As noted above, however, the public is invited tosubmit comments on these amended regulations dur-ing the next 60 days, and the Secretary will revise theregulations further as warranted by his evaluation ofthe comments received.

As noted above, these regulations are effectiveupon publication in the Federal Register. However,because the question of when the Secretary willdetermine whether a State’s CON program is satis-factory is not addressed in the regulations them-selves, the Secretary has decided as follows.

Initially, the Secretary notes the relevant statutoryprovisions under section 1521(b)(2)(B) of the Act theterm of a conditional designation agreement ofSHPDA may not exceed 36 months. A fully desig-nated SHPDA must, under section 1521(b)(3) becapable of performing all of the functions specified insection 1523, including CON reviews, during its firstyear of full designation, If, on September 30, 19&0, adesignation agreement under section 1521 is not in ef-

feet in a State, the Secretary is prohibited by section1521(d) from paying certain Federal funds for thedevelopment, expansion, or support of health re-sources in that State.

Accordingly, in determining whether a SHPDA iscapable of administering a satisfactory CON pro-gram (which is a necessary element in establishingeligibility for full designation), the Secretary will re-quire compliance with these revised regulations asfollows:

(1) For States in which SHPDAs do not require ad-ditional legislative authority to implement the revi-sions to these regulations, the Secretary will requiretheir implementation within 6 months after publica-tion of this document in the Federal Register, and inaccord with other SHPDA designation requirements.

(2) For those States in which the SHPDAs do re-quire additional legislative authority to implementthe revisions to these regulations, the Secretary willrequire their implementation within 6 months afterthe end of the earliest legislative session in whichlegislation to permit this implementation may be in-troduced and acted upon, and in accord with otherSHPDA designation requirements.

After consulting with their legal counsel, SHPDAsshould contact the appropriate DHEW RegionalOffice to determine into which of these categoriesthey fall.

Accordingly, 42 CFR Part 122, subpart D, and 42CFR part 12, subpart E, are amended in the mannerset forth below.

Dated: April 13, 1979.Julius B. Richmond, Assistant Secretary for

Health.

Approved: April 16, 1979.Joseph A. Califano, Jr., Secretary.1. Section 1 2 2 , 3 0 4 of part 122 of title 42 is

amended by adding to its a new paragraph (a)(5), toread as FOllOWS:

$122.304 of part 122 of title 42 is amended by add-ing to it a new paragraph (a)(5), to read as follows:

(a) . . .(5) Radiological diagnostic health services which

are offered in, at, through, by or on behalf of ahealth care facility or HMO (including services of-fered in space leased or made available to any personby the health care facility or HMO), which are pro-vided by fixed or mobile CT scanning equipment ex-cept where these services are an addition to or re-placement of the same service offered in, at, through,by, or on behalf of the health care facility or HMO.For purposes of this subparagraph, a CT head scan-ner and a CT body scanner, do not provide the sameservice and a CT fixed scanner and a CT mobile scan-ner do not provide the same service.

2. Sect ion 1 2 3 . 4 0 4 of part 123 of title 42 isamended by adding to it a new paragraph (a)(5), toread as follows:

to review.(a) . . .(5) Radiological diagnostic health services which

are offered in, at, through, by, or on behalf of ahealth care facility or HMO (including services of-fered in space leased or made available to any personby the health care facility or HMO), which are pro-vided by fixed or mobile CT scanning equipment, ex-cept where these services are an addition to or re-placement of the same service offered in, at, through,by, or on behalf of the health care facility or HMO.For purposes of this subparagraph, a CT head scan-ner and a CT body scanner do not provide the sameservice, and a CT fixed scanner and a CT mobilescanner do not provide the same service.

(Sec. 215 of the Public Health Service Act. 58 Stat.690 (42 U.S. C. 216), sees. 1501-1536 of the PublicHealth Service Act, 85 Stat. 2225-57 (42 U.S. C. 300k-l-300n-5). )

(FR DOC. 79-12636 Filed 4-24-79, 8:45 a.m. ) BillingCode 4110-83-M.

Appendix G.— Methods of Data Collection

The OTA list of computed tomography (CT) scan-ners from 1976 included 321 scanners, listed by loca-tion, type, and manufacturer of scanner, and date ofinstallation. That list, as well as the 1977, 1978, and1979 updated lists, were developed using multiplesources. First, the previous list was updated bychecking against a listing from the Food and DrugAdministration (FDA). In 1978, the updated list wassent to all State planning agencies and selected urbanhealth systems agencies (HSAs). Those sources con-tacted were asked to make corrections on the up-dated list. Because of the necessity of many telephonecalls to HSAs in 1978, inquiry was made to all Statehealth planning agencies and HSAs in 1979. The firstletter was sent in February. Two followup letterswere sent to nonresponders. When State informationwas incomplete and HSAs had not responded, theywere contacted by telephone. The same procedurewas followed in 1980, with the first letter sent inMay.

All States had information on the institutional lo-cation of CT scanners, but the information was oftenincomplete. Some States knew only the names of in-stitutions that had had an application for a certifi-cate-of-need approved. Others knew which institu-tions had operational scanners, but did not knowhow long they had been in operation or the manufac-turer. When incomplete information on an opera-tional scanner was received, and the HSA informa-tion was either incomplete or in conflict, the institu-tion or physician was called. A special attempt wasmade to identify out-of-office scanners. Frequently,staff of HSAs were more aware of the locations ofsuch scanners than were staff in State agencies.

Only scanners delivering clinical services to pa-tients were included. Thus, scanners registered by a

manufacturer, by a leasing company, or by a com-pany providing mobile scanners were included onlyif the site of clinical services could be identified.

The main effort in this study has been to ascertainthe geographic and institutional location of scanners.No attempt was made to ascertain the owner of an in-dividual scanner. Some States collect informationdifferentiating between scanners owned by radiolo-gists but located in a hospital, and scanners ownedby a hospital. It was assumed that the location—hospital or private office—was the important factorin terms of access to the entire community. Scannersowned by radiologists but located in hospitals weretreated exactly as were scanners owned by hospitals.Even when a scanner was registered to an individualphysician, if its address indicated location within ahospital, it was considered to be a hospital scanner. Ifthe scanner was located in a private clinic or physi-cian’s office building close to the hospital, however,it was considered to be a private scanner.

Other data for this report were collected from pub-lished literature and extensive interviews. Many ofthe interviews are cited, as the information is notavailable in any other source. Many individuals andgroups also furnished helpful written materials (seeapp. I).

Based on these materials, a draft of the policyinformation was developed and reviewed by theOTA Health Program Advisory Committee at itsmeeting of April 28, 1980. A draft of the entire reportwas then developed and was sent to the committeeand to more than 100 individuals and groups forreview on August 1, 1980. Most comments were re-ceived by October 1, 1980. The final report was thenwritten based on the many excellent comments andsuggest ions.

Appendix H. —Glossary of Acronyms and Terms

Glossary of Acronyms

AAPSRO — American Association of Professional

ACRAHPAAS&EBHPBMDBRHCBACEACFRCONCT

DHEW

DHHS

DCDEPDODDPADQADRDRR

DSRFDAFRGEHCFA

HECTHMOHRA

HSAHSPIOMMCENAS

Standards Review Organizations—American College of Radiology—American Health Planning Association— American Science & Engineering Co.– Bureau of Health Planning (HRA)— Bureau of Medical Devices (FDA)– Bureau of Radiological Health (FDA)— cost-benefit analysis— cost-effectiveness analysis— Code of Federal Regulations— certificate of need— computed tomography, computerized

tomograph y— Department of Health, Education, and

Welfare— Department of Health and Human

Services (formerly DHEW)– Division of Compliance (BRH)– Division of Electronic Products (BRH)— Department of Defense— Designated Planning Agency– Division of Quality Assurance (BRH)— Division of Radiology (NIH)— Division of Research and Resources

(NIH)— dynamic spatial reconstructor— Food and Drug Administration (PHS)— Federal Register— General Electric— Health Care Financing Administration

(DHHS)— head-equivalent computed tomography— health maintenance organization— Health Resources Administration

(PHS)--- health systems agency— Health Systems Plan– Institute of Medicine (NAS)— Medical Care Evaluation (studies)— National Academy of Sciences

NCHCT — National Center for Health CareTechnology (OASH)

NCHS — National Center for Health Statistics(OASH)

NCHSR — National Center for Health ServicesResearch (OASH)

NCI — National Cancer Institute (NIH)NEMA — National Electrical Manufacturers

AssociationNHLBI – National Heart, Lung, and Blood

Institute (NIH)

NIH —National Institutes of HealthNINCDS — National Institute of Neurological and

Communicative Disorders and StrokeNMR — nuclear magnetic resonanceOASH — Office of the Assistant Secretary for

Health (DHHS)OHPA — Office of Health Practice Assessment

(defunct)OMAR — Office for Medical Applications of

Research (NIH)PETT —positron emission transaxial

tomographyPHS – Public Health Service (DHHS)PSRO — Professional Standards Review

OrganizationRSNA — Radiological Society of North AmericaSCBT — Society for Computed Body

TomographySHPDA — State health planning and development

agencySUNY — State University of New YorkUSC —United States CodeVA — Veterans Administration

Glossary of Terms

Cost-benefit analysis (CBA): An analytical] techniquethat compares the costs of a project or technologi-cal application to the resultant benefits, with bothcosts and benefits expressed by the same measure.This measure is nearly always monetary.

Cost-effectiveness analysis (CEA): An analyticaltechnique that compares the costs of a project orof alternative projects to the resultant benefits,with cost and benefits effectiveness not expressedby the same measure, Costs are usually expressedin dollars but benefits effectiveness are (ordinarily

expressed in terms such as “ii\’es saved, ” “disabil-ity avoided, ” “quality adjusted life years saved(QALY), ” or any other relevant objectives. Also,when benefits/effectiveness are difficult to ex-press in a common metric, they may be presentedas an “array. ”

Device (medical): Any physical items, excludingdrugs, used in medical care (including instru-ments, apparatus, machines, implants, and re-agents).

Diffusion: The process by which a technology entersand becomes part of the health care system. It hastwo phases: adoption and use of the technology.Most studies of diffusion have only examined theadoption phase.

Effectiveness: Same as Efficacy (see below) exceptthat it refers to”. . . average conditions of use.”

Efficacy: The probability of benefit to individuals ina defined population from a medical technologyapplied for a given medical problem under idealconditions of use.

Medical technology: The drugs, devices, and medicaland surgical procedures used in medical care, andthe organizational and support systems withinwhich such care is provided.

Procedure: A medical technology involving anycombination of drugs, devices, and provider

skills and abilities. Appendectomy, for example,may involve at least drugs (for anesthesia), mon-itoring devices, surgical devices, and physicians’,nurses’, and support staffs’ skilled actions.

Risk: A measure of the probability of an adverse oruntoward outcome’s occurring and the severity ofthe resultant harm to health of individuals in a de-fined population associated with use of a medicaltechnology applied for a given medical problemunder specified conditions of use.

Safety: A judgment of the acceptability of risk (seeabove) in a specified situation.

Appendix 1. —Acknowledgments

In addition to reviews by the members of the Health Program Advisory Committee, OTA received reviewsof drafts or parts of drafts of this report from the following people and groups. These reviewers do not necessari-ly approve, “disapprove, or endorse the report.

Leonard J. Bisaccia, M.D.DirectorRadiology ServiceVeterans Administration

Bruce G. BreamerGovernment Affairs ManagerTechnicareCleveland, Ohio

Russell A. Brown, M.D.General Electric Co.Milwaukee, Wis.

David O. Davis, M.D.Professor and ChairmanDepartment of RadiologyThe George Washington University Medical Centerwashington, DC.

Michael Eliastam, hf. D., M.P.P.Special Assistant to the DirectorNational Center for Health Care TechnologyDepartment of Health and Human Services

Henry A. Foley, Ph.D.AdministratorHealth Resources AdministrationDepartment of Health and Human Services

Jean R. Herdt, M.D.Deputy ChiefDiagnostic Radiology DepartmentThe Clinical Center National Institutes of Health

Gordon C. Johnson, M.D., and staffBureau of Radiological HealthFood and Drug Administration

Otha W. LintonDirector of Governmental RelationsThe American College of RadiologyWashington, D.C.

Joseph A. Marasco, Jr., M.D.Director, Diagnostic RadiologySt. Francis General HospitalPittsburgh, Pa.

Barbara P. McCool, Ph.D.Acting DirectorNational Center for Health Services ResearchDepartment of Health and Human Services

Robert G. McCuneDivision Staff ManagerDiagnostic Imaging and Therapy Systems DivisionNational Electrical Manufacturers AssociationWashington, D.C.

David A. Pistenma, M.D.Chief, Radiotherapy Development BranchNational Cancer InstituteNational Institutes of Health

Dennis F. SiebertDirect orOffice of Professional Standards Review

OrganizationsHealth Standards and Quality BureauHealth Care Financing AdministrationDepartment of Health and Human Services

Thomas H. Shawker, M.D.Chief, Ultrasound SectionDiagnostic Radiology DepartmentThe Clinical CenterNational Institutes of Health

Michael Walker, M.D.DirectorStroke and Trauma ProgramNational Institute of Neurological and

Communicative Disorders and StrokeNational Institutes of Health

Daniel ZwickAmerican Health Planning AssociationWashington, D.C.

101

References

References

1.

2.

3.

4.

5.

6.

7.

8,

9.

10.

11.

12.

13.

Abrams, H. L., “Computed Body Tomogra-phy: A Background Paper, ” draft, NationalCenter for Health Care Technology, DHHS,Rockville, Md., July 1980.

, panelist, American College of Radiolo-gy Press Seminar on CT Scanning, press kitmaterials on “Computed Tomography in Ma-lignant Disease, ” Washington, D. C., Oct. 22,1979.

, and McNeil, B., “Medical Implicationsof Computed Tomography (’CAT Scanning’ ),”N. Eng. }, Med. 298(5):255 and 310, Feb. 2 and9, 1978.ACR Faculty Panelists, American College ofRadiology, statements at press seminar on CTscanning (from transcript of proceedings),Washington, D. C., Oct. 22, 1979.Alfidi, R. J., et al., “Computed Body Tomogra-phy and Cancer, ” A)~I, J. Roerltge?lol, 1 2 7 :1061, 1976.Alliance for Engineering in Medicine and Biolo-gy, Directory of Federal Progra)~ls itl Medica/Diag~zostic LJItrasound, C h e v y C h a s e , Md.,1976.Allsopp, P., A Prelii}rirlary lr~vestigatio~~ of t}z~Dezlelopmellt a~zd Costs of Positrotl EmissionTramaxial Tonlograp/ly Scan~zers (PETTs),a~lci t~le llegulatcry arid Rein] bu rseme~z t IssuesRaised by Their l)~lpe?lding Diffusio?z (Wash-ington, D. C.: Office of Health Regulations,Health Care Financing Administration, DHHS,August 1980).Alternative Weighting Systems Proposed forthe CT Scanner Standards in the NationalGuidelines for Health Planning, Office of Plan-ning, Evaluation and Legislation, Health Re-sources Administration, DHHS, Hyattsville,Md., Nov. 8, 1979.Ambrose, J., et al., “EMI Scan in the Manage-ment of Head In juries,” La}~cet 1 :847, Apr. 17,1976.American College of Radiology, “ACR OffersGuidelines for Supervision, Use of CT, ” ACRBull. 35:4, October 1979.

“New Policy on CT Approved, ” ACRBull., October 1980.American Hospital Association, Hospital Sta-tistics, 1978 Edition (Chicago, 111., 1978).Baker, C., and Way, L. W., “Clinical Utility ofCAT Body Scans, ” Am. ]. Surg, 136:37, J u l y1978.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

Banta, H. D., “Computed Tomography: CostContainment Misdirected” (editorial), A)n. /.Public Health 70:215, 1980.

“Diffusion of the Computed Tomogra-phy (CT) Scanner in the United States, ” IHt, ].Health Sem. 10(2):251, 1980.

r “Public Policy and Medical Technol-ogy: Critical Issues Reconsidered, ” presented atthe conference “Policy Innovation and theService Sector, ” Berlin, West Germany, June13-16, 1978.

, and Thacker, S., “Costs and Benefitsof Electronic Fetal Monitoring, ” DHEW pub-lication No. PHS 79-3245 (Hyattsville, Md.:National Center for Health Services Research,DHHS, April 1979).Barone, G., Chief X-Ray Product Branch, Divi-sion of Compliance, Bureau of RadiologicalHealth, Food and Drug Administration,DHHS, presentation at Seminar for Specifica-tion, Evaluation and Quality Assurance of CTScanners, Hyatt Regency Hotel, Washington,D.C. , July 9-10, 1979.Bartlett, J. R., and Neil-Dwyer, G., “A ClinicalStudy of the EMI Scanner: Implications forProvisions of Neuroradiological Services, ” I?ri.Med. ~, 2(6140):813, Sept. 16, 1978.

, et al., “Evaluating Cost-Effectivenessof Diagnostic Equipment: The Brain ScannerCase, ” Bri. M e d . J. 16(6140):815, Sept. 1 6 ,1978.Bice, T. W., “Regulation of the Adoption andDiffusion of CT Technology, ” contract No. HS03750, National Center for Health Services Re-s e a r c h , DHHS, Hyattsville, Md., p r o j e c t e d1981.Blackwell, Q., National Cancer Institute, Na-tional Institutes of Health, DHHS, Bethesda,Md.; persona] communication, July 24, 1979.Blumstack, M. B., Special Assistant for Man-agement Planning and Evaluation, Division ofResearch and Resources, National Institutes ofHealth, DHHS, “Development Use and ~orEvaluation of Computerized Axial Tomogra-phy (CAT), ” memorandum to Chief, Legisla-tive Liaison and Analysis Branch, Division ofLegislative Analysis, National Institutes ofHealth, DHHS, Bethesda, Md., July 5, 1979.“Breast Cancer Screening, ” Consensus Devel-opment Conference sponsored by the NationalCancer Institute, National Institutes of Health,

25<

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

DHHS, National Institutes of Health Corrsen-sus Del~elopment Conference Summaries 1:5,1977-78.“Breast Scanning Makes Waves as Alternativeto Mammography },” M. W,N. 20(18):16, Sept.3, 1979.Bronson, G . , “Sober Diagnosis: As Once-Bright Market for CAT Scanners Dims, Small-er Makers of the X-Ray Devices Fade Out, ”Wall Street ]our~lal, p. 48, May 6, 1980.Brown, R. A., et al., “Stereotaxic Frame andComputer Software for CT-Directed Neurosur-gical Localization,” lnl~est. Radio, 15(4):308,July-August 1980.Bunker, J., “Surgical Manpower, ” N. E/~g, ],Med. 282:135, 1970.Bureau of Health Management Standards, “Di-agnostic Ultrasound—The Second Report” (Al-bany, N. Y.: New York State Department ofHealth, November 1979).Bureau of Radiological Health, Food and DrugAdminis trat ion, DHHS, Rockville, Md., re-quest for proposal No. 223-79-6024, “The Col-lection of Data on the Utilization of ComputedTomography (CT), ” June 8, 1979.Business Week-, “Why EMI Seeks a Merger inthe U.S., ” p. 52, Apr. 9, 1979.Carrera, G. F., et al., “Computed Tomographyof the Brain in Pati ~nts With Headache of Tem-poral Lobe Epilepsy: Findings and Cost-Effec-tiveness, ” ]. Co~np~{t. Assist. To~~zog, I(2),1977.CGR Co., information furnished to OTA, Au-gust 1980.Cohn, V., “New X. Ray Machine Depicts Bodyat Work, ” Was)zi~zgto~~ Postr Oct. 7, 1980.“ C o m m e n t s o n N a t i o n a l H e a l t h P l a n n i n gGuidelines for CT Scanners, ” National Elec-trical Manufacturers Association Responsesubmitted to Office of Planning, Evaluationand Legislation, H~alth Resources Administra-tion, DHHS, Hyattsville, Md., May 4, 1979.“Consideration of the CT Scanner Standards inthe National Guidelines for Health Planning, ”draft, Office of Plar\ning, Evaluation and Legis-lation, Health Resources Administration,DHHS, Hyattsville, Md., Aug. 30, 1979.“Coverage of Physicians’ Offices Under StateCertificate of Need Programs, ” memorandumfrom Gregory Raab, Public Health Analyst,Health Resources Administration, DHHS, toDirector, Division of Regulatory Activities,Health Resources Administration, DHHS, Hy -attsvi]]e, Mcl., Mar. 20, 1978.

38.

39.

40.

41.42.43.44.45.46.47.48.

49.

50.

51.

52.

53.

54.

55.

56.

Creditor, M. C., and Garrett, J. B., “The In-formation Base for Dilfusion of Technology:Computed Tomographic Scanning, ” N. Eng. ].Med. 297(1):49, July 7, 1977,Cromwell, J., et al., ]ncentives and DecisionsUnderlying Hospitals’ Adoption and Utiliza-tion of Major Capital Equipment, report to theNational Center for Health Services Researchand Development, DHEW (Cambridge, Nlass.:ABT Associatesr 1975).Dietrich, W. C., Vice President of Marketing,Omni-Medical, “Future of CT Scanning, ” ;rrter-nal memorandum, May 8, 1980.E/ectrica/ News, p. 59, Nlar. 29, 1976.

p. 24, Jan. 23, 1978.~ p. 57, Jan. 1, 1979.

p. 60, Feb. 26, 1979.Elec. Week, p. 24, Mar. 1, 1978.Electronic, p. 148, Jan. 5, 1978.Elec. Bus,, p. 66, May 1979.Emmitt, R. B., “Market Scan: Dynamic 40 Per-cent Increase in Ultrasound Sales, ” Diag.Zmag., November 1979.Enlow, R. A., and Ehlert, K., “Recent Develop-ments in United States Commercial ComputerTomography Systems, ” working paper, HealthServices Research Center and Health CareTechnology Center, University of Missouri-Columbia, grant No. HS 02569, NationalCenter for Health Services Research, DHHS,Hyattsviller Md., Jan. 20, 1978.Evens, R. G., “An Update on the Economicsand Utilization of CT—Its Impact on HealthCare Costs in the U.S. “ presentation at theAmerican College of Radiology press confer-ence, Washington, D. C., Oct. 22, 1979.

“The Economics of Computed Tomogr-aphy: ‘ Comparison With Other Health CareCosts, ” Radio. 136:509, August 1980.

“National Guidelines and Standardsfor Health Planning: Their Relation to Radiolo-gy, ” Am. ], Roeiztgenol 131:1 101, December1978.

“New Frontier for Radiology: Com-puted ‘Tomography, ” Am. ], Roentget?ol,126(6): 1117, June 1976.

and Jest, R. G., “Economic Analysis ofCompu’ted Tomography Units, ” Am. ]. Roent-genol, 127:191, 1976.

and Jest, R. G., ‘ Economic Analysls ofBody Computed Tomography Units IncludingData on Utilization, ” R a d i o . 127:151, April1978.

, and Jest, R. G., “Utilization of Body

57

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

Computed Tomography Units, ” Radio. 131:695, June 1979.

and Jest, R. G., “utilization of HeadCompu’ted Tomography Units, ” Radio. 131:691, June 1979.“Executive Summary: NCHCT Policy and Pro-gram, ” draft, National Center for Health CareTechnology, Rockville, Md., June 28, 1979.Fact File’ Medical X-Ray Forum No. 36, “Ultra-sonography: Seeing With Sound, ” presentationat Eastman Kodak, Inc., New York, N. Y., pre-pared under the supervision of forum panelistsand the American College of Radiology, 1977.Fineberg, H. V., and Hiatt, H. H., “Evaluationof Medical Practices: The Case for TechnologyAssessmen t,” N, Eng. ], M e d . 301(20):1086,Nov. 15, 1979.

, et al., “Computerized Cranial Tomog-raphy: Effect on Diagnostic and TherapeuticPlans, ” ]. A. M.,4. 238(3):224, July 18, 1977.

, et al., “CT Scanners: Distribution andPlanning Status in the United States, ” N. E)Ig.). Med. 297(4):216, July 28, 1977.Fox, J. L., “Meth(d\ Map Bra in Functi[~nsChemically, ” C&E Nrus 56(481:20, Nov. 27,1978.Frank, R. B., Vice-President and Corporate Di-rector of Planning and Development, PickerCorp., “Diagnostic Imaging in the 1990’ s,” inl?e/Jclrt of t}lt? Procecdi}zgs of t)le Natiotlu/ S<Y??l-posiuttl 0)2 Diag}lostic I?llagillg, Ottawa, Cana-da, Oct. 23-25, 1978, published by authority ofthe Minister of National Health and Welfare,1978.Freedman, G. S., “CT: A Financial and Legisla-tive Reappraisal, ” App. Radio., November-December 1978.Gallagher, J. J., Principal, Technology Market-ing Group, Ltd., “Market Surveys of CT Inter-est, ” presentation at Seminar for Specification,Evaluation and Quality Assurance of CT Scan-ners, Hyatt Regency Hotel, Washington, D. C.,July 9-10, 1979.Gilbert, B., Biodynamics Research Unit, MayoClinic, Rochester, Minn., personal communi-cation, June 20, 1979.Coitcin, M., “C~>mputed Tomography in Plan-ning Radiation Therapy, ” l)zt. ], Rad. 0. 5(3):445, hfarch 1~79.Gordon, R., “Image Reconstruction From Pro-jections, ” Sri. A))~, 233(4):56, October 1975.Greenwald, H. P., et al., “Transportation CJfCT Scanners: A Theory and Method of HealthRest) urces Allc~catic~ n,” Hea/t/z se)~’. Res. 14(31:207, fall 1979.

71.

72.

73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.85.

86

“Health Planning Amendments, ” coF~g-, Qtrly.,1979.Health Resources Administration, DHHS,memorandum, Jan. 21, 1980.Henderson, H., “A Sideways Look at Scan-ners, ” hlcu~ Sci,, Dec. 6, 1979.Hobbins, C . , Office of the Administrator,Health Resources Administration, DHHS, Hy -attsville, Md., personal communications,December 1979, February 1980.Hobday, P. E, et al., “Computed TomographyApplied to Radiotherapy Treatment Planning:Techniques and Results, ” Tl~erap. Rad., 133:477, November 1979.Hoult, D . , “An Overview of NMR in Medi-cine, ” a report prepared for the NationalCenter for Health Care Technology, Biomedi-cal Engineering and Instrumentation Branch,Division of Research Services, National In-stitutes of Health, DHHS, Bethesda, Md.,August 1979.

, “The Productic}n of 2-D and 3-D I m -ages by NMR, and Potential Diagn{wtic Appli-cation, ” paper given at Science Writers Semi-nar on Nuclear Magnetic Res[~nance, NationalInstitutes of Health, Bethesda, Md., Apr. 23,1980.

Biomedical Engineering and Instru-mental ion Branch, Divisi(>n (~t Research Serv-ices, National Institutes t~f Health, DHHS,Bethesda, Md., personal communication, Sep-tember 1979.

Biomedical Engineering and Instru-mentation Branch, Division of Research Serv-ices, National Institutes oi Health, remarksfollowing presentation at science writers sem-inar on Nuclear Magnetic Resonance, NationalInstitutes of Health, Bethesda, Md., Apr. 23,1980.Hounsfield, G. N.r “Computed Medical Imag-ing, ‘r Scie~~re 210:22, 1980.

“The EMI Scanner, ” Proc. R. SOL-.La}ld. B. 195:281, 1977.Hughes, G. M. K., “National Survey of CT Ca-pacity, ” Radio. 135(3):699, 1980.Ingrassia, L., and Stevens, C. W., “MayoClinic X-Ray Scanner Promises Tt) Add toMedical Knowledge—and Costs, ” Wall StreetJour)lal, Mar. 6,1980.ZIrst. Tec/~., p. 18, April 1978.Intermediary letter No, 78-38, pts. A and B, re-garding “Maximum Reasonable Charges forComputer Tomography (CT) Diagnostic Serv-ices, ” refer to IHI-3 12, September 1978.Jt~hns[~n, G . , Divisit~n ~)f Qualit)r Assurance ,

87.88.

89.

90.

91.

92.

93.

94.

95.

96.

97.

98.

Bureau of Radiological Health, Food and DrugAdministration, DHHS, Rockville, Md., per-sonal communications, September 1979, Oc-tober 1980.]. Elect. Eng., p. 14, November 1979.Jurgen, R. K., “Electron@ in Medicine: TheMicroprocessor Takes Over and the UltimateCAT Scanner Nears Completion, But CostsStill Soar, ” ZEEE Spectrum, p. 76, J a n u a r y1979.Knauss, W. A., and Davis, D. O., “Utilizationand Cost-Effectiveness of Cranial ComputedTomography al a University Hospital, ” J.Comput. Assist. Tornog. 2(2):2, 1978.Knox, H., Food and Drug Administration,DHHS, Rockvillc, Md., personal communica-tion, September 1979.Kuntz, E., “Medicine Turns to Non-X-RayModes, ” Mod. Heulthcm-e 10:58, January 1980.Larson, E. B., et al., “’Computed Tomographyin Patients With Cerebrovascular Disease: Im-pact of a New Technology on Patients Care, ”Am. ]. Roentgenol. 131:35, July 1978.

, et al., “Impact of Computed Tomogra-phy on the Care of Patients With SuspectedHydrocephalus, ’ A m . ], Roentgenol. 131:41,July 1978.Latchaw, R. E., et al., “A Protocol for the Useof Contrast Enhancement in Cranial ComputedTomography, ” Radio, 126:681, March 1978.Laze]], A., Assistant Director, Bureau ofRadiological He:lth, Food and Drug Adminis-tration, DHHS, Rockville, Md., FDA data oninstallations of CT scanners for the periods ofJuly 1977 throul;h June 1978, and July 1 9 7 8through June 1s79, furnished by Dr. StuartCarlow, FDA, Rockvil]e, Md., data releasedOct. 10, 1979.Lerch, 1. B., “Beyond the CT Scanner: InSearch of a Point of Light, ” The Sciences 20(2):

6, February 1980.Letter and enclo~ures from Anthony T. Mott,Executive Director, Finger Lakes Health Sys-tems Agency to (Iffice of Planning, Evaluationand Legislation, Health Resources Administra-tion, DHHS regarding the proposed recom-mendations on the CT scanner standards in theNational Guidelines for Health Planning,Rochester, N. Y., Feb. 12, 1980.Letter from the American College of Radiologyto office of Planning, Evaluation and Leg-islation, Health Resources Administration,DHHS, May 2, 1979 in response to the Notice

99.

100.

101.

102.

103.

104.

105.

106.

107.

108.

of Request for Comment in Federal Register,Mar. 23, 1979.Letter from Dean Montgomery, ExecutiveDirector, Northern Virginia HSA pursuant toJune 5, 1979 meeting with Office of Planning,Evaluation and Legislation, National ElectricalManufacturers Association, American Collegeof Radiology, and others, to James Stockdill,Director, Office of Planning, Evaluation andLegislation, Health Resources Administration,DHHS, Falls Church, Va., July 2, 1979.Letter from OTA, U.S. Congress, to Office ofPlanning, Evaluation and Legislation, HealthResources Administration, May 2, 1979 in re-sponse to the Notice of Request for Commentin Federal Register, N4ar. 23, 1979, Washing-ton, D.C.Marx, J. L., “NMR Opens a New Windclw Intothe Body, ” Science 210:302, 1980.McNeil, B. J., “Pitfalls in and Requirements forEvaluations of Diagnostic Technologies, ” inMedical Technology, DHEW publication No.(PHS) 79-3254, NCHSR Research ProceedingsSeries, Urban Institute Conference, West PalmBeach, Fla,, Dec. 10-12, 1978 (Washington,D. C.: DHEW, September 1979).Medicare Carriers Malzual, sec. IO-12 of appen-dix on Coverage Issues.Medicare Part A intermediary Manual, Part 3,transmittal No. 765, May 1979, regardingChange in Policy: Effective for Mobile CTScans Performed on or After Apr. 25, 1979.Memorandum, American Hospital Associa-tion, Chicago, Ill., to the record, “Mobile CTScanners: State of the Art,” Oct. 13, 1978.Memorandum from Director, Medicare Bu-

reau, to all Regional Medicare Directc)rs, re-garding “Identical Memorandum No. 77-90—Maximum Reasonable Charges for CT Diag-nostic Services, ” refer to IHI-312, Bait imore,Md., December 1977.Memorandum from Director, Office of HealthPractice Assessment, OASH, to Director, Divi-sion of Provider and Vledical Services Policy,Program Policy, Medicare Bureau, regarding“Computed Tomograph y (CT) —Body Scans(with exception of crarial CT), ” Jan. 13, 1978.Memorandum from James Stockdill, Ast;ociateAdministrator, Office of Planning, Evaluation,and Legislation, Health Resources Administra-tion, DHHS, to Henry Foley, Administrator,Health Resources Administration, [)HHS,regarding Proposed Revision of the CT Scanner

109.

110.

111.

112.113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

Standards in the National Guidelines forHealth Planning, Hyattsville, Md., Sept. 1 0 ,1979.Memorandum from Laurel Shannon, SpecialAssistant to Associate Administrator, Office ofPlanning, Evaluation and Legislation to JamesStockdill, Associate Administrator, Office ofPlanning, Evaluation and Legislation, regard-ing Standards for CT Scanning in the NationalGuidelines for Health Planning, Hyattsville,Md., Aug. 29, 1979.Miers, M. L., Division of Legislative Analysis,National Institutes of Health, “Information Re-quest on CT Scanners, ” memorandum to theRecord, July 10, 1979.Minutes from meeting of Subcommittee on Na-tional Guidelines, Goals, Standards and Prior-ities of National Council on Health Planningand Development, Los Angeles, Calif., Nov. 8,1979.

Hyattsville, Md., January 1980.Mod. ~ealthcare, “AS&E Loses CAT License, ”7(9):14, September 1977.

“Varian P l a n s A x i n g CTS, ” 9:71,March ~979.

“Hospitals Expand Microsurgery Abili-ty as M’ethods Become Routine, ” 9:36, Novem-ber 1979.National Academy of Sciences, Institute ofMedicine, Computed Tomographic Scanning,A Policy Statement, IOM publication No.77-02 (Washington, D. C.: IOM, April 1977).

Health Planning in the United States:Zssues in Guideline Development, IOM pub-lication No. IOM 80-01 (Washington, D. C.:IOM, March 1980).National Center for Health Care Technology,“Coverage Recommendations by the PublicHealth Service, ” NCHCT, draft, part of packetat first meeting of NCHCT Council on July 12,1979.National Center for Health Statistics, NationalCenter for Health Services Research, Health,United States (table 107), DHEW publicationNo. (PHS) 78-1232 (Hyattsville, Md.: DHEW,1978) .National Electrical Manufacturers’ Associa-tion, Washington, D. C., personal communica-tion, September 1980.

response to Request for Comment onthe CT’Scanner Guidelines published in FederalRegister, Mar. 23, 1979.National Institutes of Health, Science Writers’Seminar on Nuclear Magnetic Resonance, Be-thesda, Md., Apr. 23, 1980.

123.

124.

125.

126.

127.

128.

129.

130.

131.

132.

133.

134.

“New Products: GE Introduces New Ultra-sound Scanner, ” M o d . Healthcare 10(1):101,January 1980.Office of Graduate Medical Education, HealthResources Administration, “Interim Report ofthe Graduate Medical Education National Ad-visory Committee to the Secretary, ” DHEWpublication No. (HRA) 79-633 (Hyattsville,Md.: DHEW, April 1979).Office of Technology Assessment, U.S. Con-gress, Assessing the Efficacy and Safety ofMedical Technologies, GPO stock No. 052-003-00593-O (Washington, D. C.: U.S. Govern-ment Printing Office, September 1978).

Development of Medical Technolo-gies: O~portunitie.s for Assessment, GPO stockNo. 052-003 -00217-5 (Washington, D. C.: U.S.Government Printing office, August 1976).

The Implications of Cost-EffectivenessAnalys’k of Medical Technology, GPO stockNo. 052-003-00765-7 (Washington, D. C.: U.S.Government Printing Office, August 1980).

The Zmp/ications of Cost-EffectivenessAnalys’is of Medical Technology/BackgroundPaper #4: The Management of Health CareTechnology in Ten Countries, GPO stock No.052-003 -00783-5 (Washington, D. C.: U.S.Government Printing Office, October 1980).

Policy Implications of the ComputedTomography (CT) Scanner, GPO stock No.052-003 -00565-4 (Washington, D. C.: U.S.Government Printing Office, August 1978).Padila, E., and Strasberg, M. A., “Policy En-vironment of CT Scanners and Other Capital-Based Technologies, ” paper, Graduate Schoolof Public Administration, New York Universi-ty, summer 1980.Perry, S. M., Director, National Center forHealth Care Technology, Rockville, Md., per-sonal communication, September 1979.

Director, Office of Medical Applica-tions ~f Research , National Institutes ofHealth, DHHS, “Memorandum-NIH Outlinefor Technology Study,” Aug. 26, 1977.

Director, Office of Medical Applica-tions ~f Research , National Institutes ofHealth, DHHS, panelist at presentation ofASPA Conference, Medical Technologies Sec-tion, Baltimore, Md., Apr. 4, 1979.Perspective RSNA 1979: Scientific Program,65th Scientific Assembly and AnnuaI Meeting(including sessions sponsored jointly with theAmerican Association of Physicists in Medi-cine), sponsored by the Radiological Society ofNorth America, Atlanta, Ga., N O V. 24-29,1980.

135.

136.

137.

138.139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

“Plethorao fUnder-Utilized Body CT Scannersin States, Metro Areas Creates ‘Moratorium’on New Machines, Wash. U. Radiologist TellsACR, ” Blue Sheet, Oct. 31, 1979.Popp, R. L., and Macovski, A., “Ultrasonic Di-agnostic Instruments, ” Science 210:268, Oct.17, 1980.Potsaid, M, S,, “Diagnostic Imaging in Per-spective,” ]. A.M.A. 243:2412, June 20, 1980.Predi 88, p. 24, hIay 15, 1973.Program Information Letter No. 80-20 to StateHealth Planning and Development Agencies,Statewide Health Coordinating Councils,Health Systems Agencies, and Centers forHealth Planning from the Bureau of HealthPlanning, Health Resources Administration,DHHS, regarding the Proposed Recommenda-tions on the CT Scanner Standards in the Na-tional Guidelines for Health Planning, Hyatts-ville, Md., Jan. 9, 1980.“Prospective Reimbursement Needed to Har-ness Health Technology . . .“ Blue Sheet 2:3,Oct. 31, 1979.Rapoport, J., “13iffusion of Technological In-novation Among Non-Profit Firms: A CaseStudy of Radioisotopes in U.S. Hospitals, ” ],Econ. Bus. 30:18CI, 1978.Relman, A. S . , “CAT Scanners—Conferring‘The Greatest Berlefit on Mankind’, ” N. Eng. ].Med. 301:1062, November 1979.Reiss, J., Director, Office of Health Regulation,Health Care Financing Administration, “NewHigh Technology Medical Diagnostic andTreatment Monitoring Devices—PositronEmission Transaxial Tomography and NuclearMagnetic Resonance (NMR) Scanners, ” memo-randum to the Administrator, Health Re-sources Administration, DHHS, Washington,D. C., Apr. 17, 1980.Ritman, E. L., et al., “Three-Dimensional Im-aging of Heart, ‘Lungs and Circulation, ” Sci-ence 210:273, 19813.Rogers, E. M., and Shoemaker, F. F., Cornrnu-~lication of Innovations, A Cross-Cultural Ap-proach (New York, N. Y.: The Free Press,1971).Russell, L., Techfiology in Hospitals (Washing-ton, D. C.: The Brookings Institution, 1979).Sample, W. F., “(~omparison of Diagnostic Ul-trasound and CT Scanning in the Torso, ” paperpresented to the Bureau of Radiological Health,Food and Drug Administration, DHHS, Rock-ville, Md., Jan. 12, 1977.Scheidt, P. C., and Lundin, F. E., Division ofBiological Effects, Bureau of Radiological

149.

150.

151.

152.

153.

154.

155.

156.157.

158.

159.

160.

Health, Food and Drug Administration,DHHS, “A Review of the 9 Human Studies andTheir Methodological Limitations, ” FDAbackground paper, Rockville, Md., 1979.Schwartz, H., “The Government Puts a Dam-per on the Scanner Bonanza, ” New YorkTimes, Dec. 18, 1977.Shah, P., Wadsworth Veterans AdministrationMedical Center, “Ultrasound Diagnostic Tech-niques in Cardiology, ” summary of paper inpreparation for the National Center for HealthCare Technology, DHHS, Washington, D. C.,August 1979.Shakhashiri, Z., National Institutes 0[ Neu-rological and Communicative Disorders andStroke, National Institutes of Health, DHHS,Bethesda, Md., telephone conversation duringwhich Dr. Shakhashiri related the text of theNINCDS official statement on the PETT proj-ects of that Institute, October 1980.Shapiro, D. B., and Allen, N., “Computer-Assisted Axial Tomography–CAT or E M IScans, ” Med. Tr, T. Q. 23(2):116, fall 1976.Shawker, T. H., Abdclminal Ultrasound Unit,Department of Radiology, NIH Clinica,l Cen-ter, National Institutes of Health, Bethesda,Md., personal communication, October 1979,October 1980.

“Amplitude Analysis of Pnacreatic B-Scans: ‘A Clinical Evaluation of Cystic Fibro-sis, ” Ultrasonic Zmag. 2:55, 1980.

“Amplitude Analysis of the B-Scan inPatient; With Diffuse Infiltrative Liver Dis-ease, “ in Proceedings of the American in,; titu teof /J/trasound and Medicine, in press, 1980.

letter to OTA, Aug. 27, 1980.Shope, ’ T., Division of Electronic Products,Bureau of Radiological Health, Food and DrugAdministration, DHHS, Rockville, Md., per-sonal communication, October 1979.Shope, T. B., et al., “Radiation Dose FromComputed Tomography: A Survey of 20 Sys-tems Using Thermoluminescent and Film Dosi-metry in Cylindrical and AnthropomorphicPhantoms, ” manuscript, Bureau of Radio-logical Health, Food and Drug Administration,DHHS, 1980.Showalter, C., Food and Drug Administration,DHHS, Rockville, Md., “Computed Torriogra-phy: Benefit/Risk and Elenefit/Cost Considera-tions of Computed Tomography (CT) -–His-tory of FDA Actions, ” manuscript, July 23,1979.Siebert, D. F., Director, Office of ProfessionalStandards Review Organizations, Health

161,

162,

163,

164,

165

166.

167.

168.

169.

170.

171.

172.

173.

174.

Standards and Quality Bureau, Health CareFinancing Administration, “Draft ScreeningCriteria for Body and Head Computerized Ax-ial Tomography (CT) Scans, ” General Memo-randum No. 3-79 to Planning and ConditionalPSROS, Statewide Councils, Regional PSROproject officers, Feb. 22, 1979.Simmler, S., “States Face CON Revision Dead-line, ” Mod. Hea/t/~care 9(12):16, D e c e m b e r1979.Skalnik, B., “CT Body Scanning: A Questionof Efficacy, ” App. Radio., July-August 1977.Smithson, L., “Diagnostic Imaging in theEighties, ” manuscript, Stanford Research In-stitute, Menlo Park, Calif., September 1979.Society for Computed Body Tomography,“New Indications for Computed Body Tomog-raphy, ” special report, AI~l. ]. Roerltge~rol.133:115, July 1979.Spivak, J., “EMI Ltd. Leaves X-Ray ScannerBusiness, Plans Sale of the Assets to GE in theU.S., ” Wall Street ]our~ml, Apr. 30, 1980.State of California, Office of Statewide HealthPlanning and Development, material sent toOTA, Apr. 30, 1980.State of Massachusetts, Department of PublicHealth, Commissioner J. E. Fielding, “Determi-nation of Need of Guidelines for CT Scanners, ”Boston, Mass., 1978.“Statement of Robert M. Moliter on behalf (~fthe National Electrical Manufacturers Associa-tion, ” at the meeting with Office of Planning,Evaluation and Legislation, Health ResourcesA~m i n ist ra t ion, DHHS, J a m e s St(}ckdill, Di-rectt~r OPEL, Chairman, Hyattsville, McI.,1une 5, 1979.Stewart, J. R., et al., “Computed Tomographyin Radiation Therapy, ‘ r 1t~t. ]. Rad. O. 4:313,March-April 1978.Stocking, B., “X-Rays Highlight the Doctor’sDilemma, ” New Sci., Jan. 11, 1979.Summary of comments received regarding pro-posed recommendations on the CT ScannerStandard issued Jan. 9, 1980, Office of Plan-ning, Evaluation and Legislation, Health Re-sources Administration, DHHS, Hyattsville,Md., April 1980,“Suppliers Briefs: EMI Profits Down, ” Mod,Hea/thcare, November 1979,Swets, J. A., et al., “Assessment of DiagnosticTechnologies, ” Science 205(24):753, Aug. 24,1979.Technology Marketing Group Ltd., “Ultra-sound Forum 1979, ” Rad, Nuc, Med. M., p. 4,August 1979.

175.

176.

177.

178.

179.

180.

181.

182.

183.

184.

185.

186.

187.

Ter-Pogossian, M. M., et al., “Positron-Emis-sion Tomography, ” Sri, AI)I, 243: 170, 1980.Thomas, L., T}IC l-i~~es o) a Cell (New Y o r k :The Viking Press, 1974).Towery, O. P. , National Center for HealthCare Technology, Office of Health Research r

Statistics and Technology, DHHS, Washing-ton, D. C., personal communication, October1979.Vayda, E., “A Comparison of Surgical Rates inCanada and in England and Wales, ” N. E?~g, ],Med. 289:1224, 1973.Veterans Administration, A Reporf o~l tileE~?a/llatiotl of Co}?~put~rized Tot~lograp/l<y {CT)L/)zits Witl~it? tl~e Vetcra/~s Ad}~/i~~istratio~~ Hm-pitals, from the Special Central Office Advi-sory Group, Washington, D. C., 1978.Wagner, J. L., “The Feasibility of EconomicEvaluation of Diagnostic Procedures: The Caseof CT Scanning, ” i n T17Q lt~zplicatiotzs of ccxt-E~~ectiz?e~~ess A~zalysis o f Mcdica/ Tec\ztzol-ogy ‘l?ack~rou?zd Paper # 2 : CasLT St2~dies o fMedical Tecl~}zologies, prepared by OTA, U.S.Congress (Washington, D. C.: in press, 1980).

, “Toward a Research Agenda on Med-ical Technology, ” in Medical Tccllilology,DHEW publ icat ion No, (PHS) 7 9 - 3 2 5 4 ,NCHSR Research Proceedings Series, UrbanInstitute Conference, West Palm Beach, Fla.,Dec. IO-12, 1978 (Washington, D. C.: DHEW,September 1979).Warner, K., “Treatment Decision Nlaking inCatastrophic Illness, “ Med. Care 25:I9, 1977.Weinstein, M. C., “Economic Evaluation ofMedical Procedures and Technologies: Prog-ress, Problems, and Prospects, ” in MedicalTec/~~~o/ogy, DHEW publication No. (PHS)79-3254, National Center for Health ServicesResearch, Research Proceedings Series, UrbanInstitute Conference, West Palm Beach, Fla.,Dec. 1o-12, 1978 (Washington, D. C.: DHEW,September 1979).Willems, J., presentation as panelist on “Effi-cacy and Safety Information and Cost Contain-merit, ” Medical Technologies Section, ASPAConference, Baltimore, Md., Apr. 4, 1979.Wittenberg, J., et al., “Clinical Efficacy ofComputed Body Tomography, ” A/)1. ], Roe/zt-#VIO/. 131:5, Ju)y 1978.

, et al., “Clinical Efficacy of ComputedBody Tomography, II, ” Am. ]. Roentge~zo).134:1111, June 1980.Wortzman, G . , and Holgate, R. C., “ R e a p -praisal of the Cost-Effectiveness of ComputedT o m o g r a p h y in a Government-Sponsored

Health Care System,” Radio, 130:257, January istration, DHHS, Washington, D.C., personal1979. communication, Oct. 4, 1979.

188. Yaeger, D., Office of Professional Standards 189. Yin, L., Bureau of Medical Devices, Food andReview Organizations, Health Standards and Drug Administration, DHHS, Rockville, Md.,Quality Bureau, Health Care Financing Admin- personal communication, October 1979.

0

130:257,

Admin-

DHHS,

DHHS, Rockville, 1979,