8420

8/14/2019 8420

http://slidepdf.com/reader/full/8420 1/151

Nuclear Magnetic Resonance ImagingTechnology: A Clinical, Industrial, and

Policy Analysis

September 1984

NTIS order #PB85-146207

8/14/2019 8420


—

HEALTH TECHN OLOGY CASE STUDY 27:

Nuclear Magnetic Resonance

Imaging TechnologyA Clinical, Industrial, and

Policy Analysis

SEPTEMBER 1984

This case study was performed as a part of OTA’s Assessment of

Federal Policies and the Medical Devices IndustryPrepared for OTA by:

Earl P. Steinberg, M. D., M.P.P.Assistant Professor, Johns Hop kins School of Medicine an d John Hop kins School of Hygiene and Public Health;

Senior Associate, Office of Medical Practice Evaluation;also w/ Hen ry J. Kaiser Family Found ation Faculty Scholar in General Internal Medicine

an d

Alan B. Cohen, Sc.D.Associate Director, Center for Hosp ital Finance and Man agement ;

also w/ Assistant Professor of Health Policy and Management,Johns H opkins School of Hyg iene and Pu blic Health

f I

OTA Case Studies are documents containing information on a specific medicaltechnology or area of app lication that su pp lements formal OTA assessments. Themater ial is not norm ally of as immed iate policy interest as that in an OTA Repor t,nor does it present options for Congress to consider.

CONGRESS OF THE UNITED STATES

Office of Technology AssessmentWashington, D C 20510

8/14/2019 8420


Recommended Ci ta t ion : Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: AClinical, Industrial, an d Policy Analysis (Washington, DC: U.S. Congress, Office of Tech-nology Assessment, OTA-HCS-27, Septem ber 1984). This case study was perform ed a spart of OTA’s assessment of Federal Policies and the Medical Devices Industry.

Library of Congress Catalog Card Number 84-601123

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

8/14/2019 8420


Preface Nuc lear Magnet ic Resonance Imaging T e c h -

nology: A Clinical, Industrial, and Policy Anal- ysis is Case Study 27 in OTA’s Health TechnologyCase Stud y Series. This case study has been prep ared

in connection with OTA’s project on FederalPol ic ies and the Medical Dev ices Indus t ry , r e -quested by the Senate Committee on Labor and

Human Resources and endorsed by the SenateCommittee on Veterans’ Affairs. A listing of othercase studies in the series is included at the end of this preface.

OTA case studies are designed to fulfill twofunctions. The primary purpose is to provideOTA with specific information that can be usedin formin g general conclusions regardin g broaderpolicy issues. The first 19 cases in the H ealth Tech-nology Case Stud y Series, for example, were con-ducted in conjunction with OTA’s overall projecton The Implications of Cost-Effectiveness Anal-

ysis of Medical Technology. By examining the 19cases as a group and looking for common prob-lems or strengths in the techniques of cost-effec-tiveness or cost-benefit analysis, OTA was ableto better analyze the potential contribution thatthose techniques might make to the managementof medical technology and health care costs andquality.

The second function of the case studies is toprovide useful information on the specific tech-nologies covered. The design and the funding lev-els of most of the case studies are such that theyshould be read primarily in the context of the as-sociated overall OTA projects. Nevertheless, inmany instances, the case studies do represent ex-tensive reviews of the literature on the efficacy,safety, and costs of the specific technologies andas such can stand on their own as a useful contri-bution to the field.

Case studies are prepared in some instances be-cause they have been specifically requested bycongressional committees and in others becausethey have been selected through an extensive re-view process involving OTA staff and consulta-tions with the congressional staffs, advisory panel

to the associated overall project, the Health Pro-gram Advisory Committee, and other experts invarious fields. Selection criteria were developedto ensure that case studies provide the following:

Ž examples of types of technologies by func-

tion (preventive, diagnostic, therapeutic, andrehabilitative);examples of types of technologies by physicalnature (drugs, devices, and procedures);examples of technologies in different stagesof development and diffusion (new, emerg-ing, and established);

examples from different areas of medicine(e.g., general medical practice, pediatrics,radiology, and surgery);examples addressing medical problems thatare important because of their high frequen-cy or significant impacts (e. g., cost);examples of technologies with associated highcosts either because of high volume (for low-cost technologies) or high individual costs;examp les that could provide information ma-terial relating to the broader policy and meth-odological issues being examined in theparticular overall project; and

examples with sufficient scientific literature.

Case studies are either prepared by OTA staff,commissioned by OTA and performed under con-tract by experts (generally in academia), or writ-ten by OTA staff on the basis of contractors’papers.

OTA subjects each case study to an extensivereview process. Initial drafts of cases are reviewedby OTA staff and by members of the advisorypanel to the associated project. For commissionedcases, comments are provided to authors, alongwith OTA’s suggestions for revisions. Subsequent

drafts are sent by OTA to numerous experts forreview and comment. Each case is seen by at least30 reviewers, and sometimes by 80 or more out-side reviewers. These individuals may be fromrelevant Government agencies, professional so-cieties, consumer and public interest groups, med-ical practice, and academic medicine. Academi-cians such as economists, sociologists, decisionanalysts, biologists, and so forth, as appropriate,also review the cases.

Although cases are not statements of officialOTA position, the review process is designed to

satisfy OTA’s concern with each case study’sscientific quality and objectivity. During the vari-ous stages of the review and revision process,therefore, OTA encourages, and to the extentpossible requires, authors to present balanced in-formation and recognize divergent points of view.

,..111

8/14/2019 8420


Health Technology Case Study Seriesa

— . . . —— . . —. . . .-. —.. ————...— ——— —— .Case Study Case Study

Series Case stud y title; aut hor(s); Series Case stud y title; auth or(s);number OTA publication nu mber b

number OTA publication nu mber b

1

2

3

4

5

6

7

8

9

10

11

12

13

Formal Analysis, Policy Formulation, and End-StageRenal Disease;

Richard A. Rettig (OTA-BP-H-9(1))C

The Feasibility of Economic Evalua tion of Diagnostic Procedu res: The Case of CT Scanning;

Judith L. Wagner (OTA-BP-H-9(2))Screening for Colon Can cer: A Technology

Assessment;David M. Eddy (OTA-BP-H-9(3))

Cost Effectiveness of Automated MultichannelChemistry Analyzers;

Milton C. Weinstein and Laurie A. Pearlman(OTA-BP-H-9(4))

Periodontal Disease: Assessing the Effectiveness andCosts of the Keyes Techniqu e;

Richard M. Scheffler and Sheldon Rovin(OTA-BP-H-9(5))

The Cost Effectiveness of Bone Mar row Transplan tTherapy a nd Its Policy Imp lications;

Stuart O. Schweitzer and C. C. Scalzi(OTA-BP-H-9(6))

Allocating Costs and Benefits in Disease Prevention

Program s: An Ap plication to Cervical CancerScreening;Bryan R. Luce (Office of Technology Assessment)(OTA-BP-H-9(7))

The Cost Effectiveness of Upper Gastrointestinal

Endoscopy;Jonathan A. Showstack and Steven A. Schroeder(OTA-BP-H-9(8))

The Art ificial Heart : Cost, Risks, and Benefits;Deborah P. Lubeck and John P. Bunker(OTA-BP-H-9(9))

The Costs and Effectiveness of Neonatal Intensive

Care;Peter Budetti, Peggy McManus, Nancy Barrand,and Lu Ann Heinen (OTA-BP-H-9(1O))

Benefit and Cost Analysis of Medical Interventions:

The Case of Cimetidine and Peptic Ulcer Disease;Harvey V. Fineberg and Laurie A. Pearlman

(OTA-BP-H-9(11))Assessing Selected Respiratory Therapy Modalities:Trends and Relative Costs in the Washington, D.C.Area;

Richard M. Scheffler and Morgan Delaney(OTA-BP-H-9(12))

Cardiac Radionuclide Imaging and CostEffectiveness;

William B. Stason and Eric Fortess(OTA-BP-H-9(13))

14

15

16

17

18

19

20

21

22

23

24

25

26

27

Cost Benefit/ Cost Effectiveness of MedicalTechnologies: A Case Study of Orthop edic JointImplants;

Judith D. Bentkover and Philip G. Drew(OTA-BP-H-9(14))

Elective Hysterectomy: Costs, Risks, and Benefits;

Carol Korenbrot, Ann B. Flood, Michael Higgin s,

Noralou Roos, and John P. Bunker(OTA-BP-H-9(15))

The Costs and Effectiveness of Nurse Practitioners;

Lauren LeRoy and Sharon Solkowitz(OTA-BP-H-9(16))

Surgery for Breast Cancer;

Karen Schachter Weingrod and Duncan Neuhauser(OTA-BP-H-9(17))

The Efficacy and Cost Effectiveness of Psychotherapy;

Leonard Saxe (Office of Technology Assessment)(OTA-BP-H-9(18)) d

Assessment of Four Common X-Ray Procedures;Judith L. Wagner (OTA-BP-H-9(19))e

Mandatory Passive Restraint Systems in

Automobiles: Issues and Evidence;Kenneth E. Warner (OTA-BP-H-15(20)) f

Selected Telecommunications Devices for Hearing-

Impaired Persons;

Virginia W. Stern and Martha Ross Redden(OTA-BP-H-16(21))g

The Effectiveness and Costs of AlcoholismTreatment;

Leonard Saxe, Denise Dougherty, Katharine Esty,

and Michelle Fine (OTA-HCS-22)The Safety, Efficacy, and Cost Effectiveness of Therapeutic Apheresis;

John C. Langenbru nner (Office of TechnologyAssessment) (OTA-HCS-23)

Variation in Length of Hospital Stay: TheirRelationship to Health Outcomes;

Mark R. Chassin (OTA-HCS-24)Technology and Learning Disabilities;

Candis Cousins and Leonard Duhl (OTA-HCS-25)Assistive Devices for Severe Speech Impairments;

Judith Randal (Office of Technology Assessment)(OTA-HCS-26)

Nuclear Magnetic Resonance Imaging Technology:A Clinical, Indu strial, and Policy Ana lysis;

Earl P. Steinberg and Alan Cohen (OTA-HCS-27)

available for sale by the Superintendent of Documents, U.S. Government dBackground paper #3 to The Implications of Cost-Effectiveness Analysis of

Printing Office, Washington, D. C., 20402, and by the National Technical Medical Technology.Information Service, 5285 Port Royal Road, Spring field, Va., 22161, Cal] eBackground paper #5 to The lmplicat;ons of Cost-Effectiveness Analysis of

OTA’S Publishing Office (224-8996) for availability and ordering infor- Medical Technology.fBackground paper #l 10 OTA’S May 1982 report Technology an d ~=fi-at ion .bor igina] publication numbers appear in parentheses. capped People.cThe first 17 cases in the series were 17 separately issued cases in Background gBackground Paper 42 to Technolog y and Handicapped People.Paper #2: Case Studies of Medical Technologies, prepared in conjunction

with OTA’S August 1980 report The Implications of Cost-Effectiveness Anal-

ysis of Medical Technology.

iv

8/14/2019 8420


OTA Project Staff for Case Study #27

Jane E. Sisk, Project Director

Katherine E. Locke, Research Assistant

Christopher G. Bailey, Research Assistant’

H. Christy Bergemann, Editor

Virginia Cwalina, Administrative Assistant

Rebecca I. Erickson, Secretary/Word Processor Specialist

Brenda Miller, Word Processor/P. C. Specialist

Clyde J. Behney, Health Program Manager

R o g e r H e r d m a n 2 and H . Dav id Ban ta , 3 Assistant Director, OTA

Health and Life Sciences Division

ISummer 1984.IFrom December 1983‘Until August 1983.

u

8/14/2019 8420


ADVISORY PANEL FOR FEDERAL POLICIES ANDTHE MEDICAL DEVICES INDUSTRY

Richard R. Nelson, C h a i r Institute for Social and Policy Studies, Yale University

New Haven, CT

William F. Ballhaus

International Numatics, Inc.Beverly Hills, CA

Ruth FarriseyMassachusetts General HospitalBoston, MA

Peter Barton HuttCovington & BurlingWashington, DC

Alan R. KahnConsultantCincinnati, OH

Grace KraftKidney Foundation of the Upper MidwestCannon Falls, MN

Joyce Lashof School of Public HealthUniversity of CaliforniaBerkeley, CA

Penn LupovichGroup Health AssociationWashington, DC

Victor McCoyParalyzed Veterans of AmericaWashington, DC

Robert M. Moliter

Medical Systems DivisionGeneral ElectricWashington, DC

Louise B. RussellThe Brookings InstitutionWashington, DC

Earl J. SaltzgiverForemost Contact Lens Service, Inc.Salt Lake City, UT

Rosemary StevensDepartment of History and Sociology of ScienceUniversity of PennsylvaniaPhiladelphia, PA

Allan R. ThiemeAmigo Sales, Inc.Albuquerque, NM

Eric von HippelSloan SchoolMassachusetts Institute of TechnologyCambridge, MA

Edwin C. WhiteheadTechnicon Corp.Tarrytown, NY

8/14/2019 8420


Contents

. —

CHAPTER 1: INTRODUCTION AND SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . .Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

NMR—Historical and Technical Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Clinical Applications of NOR. . . . . . . . .The NMR Imaging Device Industry . . . .

Hospital Costs and Strategies . . . . . . . . .History of Funding for NMR Research .FDA Regulation . . . . . . . . . . . . . . . . . . . . .Third-Party Payment Policies . . . . . . . . .State Certificate-of-Need Programs . . . .Regulatory Overview . . . . . . . . . . . . . . . .

CHAPTER 2: NMR–HISTORICAL ANDHistorical Background . . . . . . . . . . . . . . . . .Basic Technical Background . . . . . . . . . . . .Magnets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Magnet Type . . . . . . . . . . . . . . . . . . . . . . .Field Strength . . . . . . . . . . . . . . . . . . . . . . .Bore Size . . . . . . . . . . . . . . . . . . . . . . . . . . .Homogeneity of Field . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

TECHNICAL BACKGROUND. . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CHAPTER 3: CLINICAL APPLICATIONS OF NOR . . . . . . . . . . . . . . . . . . . . . . . . . . .Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Potential Biological Hazards of NMR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Proton Imaging: A Clinical Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Brain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Mediastinum, Hilum, and Lung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Heart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Breast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Abdomen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Kidneys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Pelvis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Musculoskeletal System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Clinical Oncology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Blood Vessels and Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Putting Proton NMR Imaging Into Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Relaxation Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Putting T1 and T2 Relaxation Times Into Perspective. . . . . . . . . . . . . . . . . . . . . . . . . . . .In Vivo Spectroscopy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Putting In Vivo Spectroscopy Into Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CHAPTER 4:THE NMRIntroduction . . . . . . . . . . .Industry Structure . . . . . .

Seller Concentration . .

Buyer Concentration. .Barriers to Entry . . . . .Diversification of FirmsAcquisition and Merger

Industry Conduct . . . . . . . .Product Pricing PoliciesNonprice Competition .

IMAGING. . . . . . . . .. . . . . . . . .,... . . . . .

,... . . . . .. . . . . . . . .. . . . . . . . .Activity .. . . . . . . . .. . . . . . . . .. . . . . . . . .

DEVICE INDUSTRY. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page

3

3

4

4

4

5

6 7

8

9

1010

15

1516

1920222222

2525252727272728282929

2929293032333435

39393939

42445050545455

vii

8/14/2019 8420


8/14/2019 8420


Contents—continuedPage

APPENDIX B.–SURVEY OF MANUFACTURERS: METHODS. . . . . . . . . . . . . . . . . . 122

APPENDIX C.–MANUFACTURERS OF NMR IMAGING DEVICES. . . . . . . . . . . . . 123

APPENDIX D.–GLOSSARY OF TERMS AND ACRONYMS . . . . . . . . . . . . . . . . . . . 140

APPENDIX E.–ACKNOWLEDGMENTS AND HEALTH PROGRAMADVISORY COMMITTEE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

TablesTable No . Page

1.

2.3.4.5.6.7.

8.9.

10.

11.12.

13.14.15.

16.

17.

Suggested Guidelines for Safe Operation of NMR Imagers . . . . . . . . . . . . . . . . . . . . 26Demonstrated and Potential Advantages of NMR Imaging . . . . . . . . . . . . . . . . . . . . 30Current and Potential Disadvantages of NMR Imaging . . . . . . . . . . . . . . . . . . . . . . . 30Likely Areas for Future NMR Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32The NMR Imaging Device Industry, October 1983 . . . . . . . . . . . . . . . . . . . . . . . . . . . 40The NMR Imaging Device Industry: Company Profile. . . . . . . . . . . . . . . . . . . . . . . . 41The NMR Imaging Device Industry: Market Share as Reflected inClinical Placements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42Status of NMR Imaging Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43Research and Development Expenditures Among Firms in the

NMR Imaging Device Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47Manufacturers’ Collaborative Arrangements With Universities/ Medical Centers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48Diversification Among Firms in the NMR Imaging Device Industry . . . . . . . . . . . . 51Acquisitions, Mergers, and Key Trade Agreements in theNMR Imaging Device Industry Since 1971 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52Diagnostic Imaging Industry Sales Growth Projections . . . . . . . . . . . . . . . . . . . . . . . 58Estimated Worldwide NMR Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58Range of Estimated Costs for NMR Imaging Systems byType of Magnet, 1983 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64Estimated Annual Costs for NMR Imaging Systems byType of Magnet, 1983 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651983 Estimated Costs per NMR Imaging Study by Type of Magnet . . . . . . . . . . . . 65

FiguresFigure No. Page

i .2.3.4.5.

6.7.

8.

9.10.11.12.

First NMR Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16First In Vivo Human Whole-Body NMR Scan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17First NMR Image of a Human Head . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181983 NMR Image . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18An NMR Image of a Normal Head From an Axial ViewWith Changes in Pulse Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19Schematic Diagram of NMR Scanner Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . 20Modern Day NMR Imager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21Mean and Standard Deviations for T 1 and T2 of Various Tissues ina Live Rat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

NMR Phosphorus Spectrogram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34Chronological Development of the NMR Imaging Device Industry . . . . . . . . . . . . . 45Cumulative Number of Worldwide Clinical NMR Placements Over Time . . . . . . . 59How To Get a New Medical Device to Market.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

A-1. Energy Levels of Hydrogen Atoms in Applied Magnetic Fields . . . . . . . . . . . . . . 119A-2. Schematic Diagram of NMR System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120A-3. Experimental Sample in Magnetic Field Gradient . . . . . . . . . . . . . . . . . . . . . . . . . . 121

il

8/14/2019 8420


8/14/2019 8420


8/14/2019 8420


Introduction and Summary

INTRODUCTION

Nuclear magnetic resonance (NMR) imaging’ isan exciting new diagnostic imaging modality thathas captured the interest of the medical profes-sion for a number of reasons. First, it employsradiowaves and magnetic fields rather than ioniz-ing radiation, thus eliminating the risk of X-irradiation that is associated with use of devicessuch as X-ray computed tomography (CT) scan-ners. Second, in addition to providing excellentdistinction between adjacent structures (spatialresolution), the technique uses differences in thedensity and the molecular environment of dif-ferent substances to provide excellent tissue con-trast without the need for injection of potentiallytoxic contrast agents. Third, because bone doesnot interfere with NMR signals (the absence of signal artifact from bone), physicians can visualizeareas such as the posterior fossa, brain stem, andspinal cord with NMR that previously were notwell seen with other noninvasive imaging tech-niques. Finally, and potentially of greatest impor-tance, because NMR imagers are sensitive to fun-damental physical and chemical characteristics of cells, the technique offers the possibility of detect-ing diseases at earlier stages than is currently pos-

sible and of permitting accurate diagnoses to bemade noninvasively.

Along with these attractive attributes, however,NMR has its disadvantages. At present, NMR im-agers are expensive, and installation of them iscostly and logistically difficult. Furthermore, until(and possibly after) more experience with themodality is obtained, NMR imaging may requiremore physician time in performance of patient ex-aminations than is the case with X-ray CT or otherimaging techniques. Moreover, despite the rapidimprovement in the quality of NMR images thathas occurred over the past several years and the

increasingly large number of clinical situations inwhich NMR imaging might prove to be of value,the exact role of NMR imaging in clinical medi-

IThe term “NMR imaging, ” used in this case study, is increas-ingly being replaced by the term “magnetic resonance imaging. ”

cine, particularly its efficacy compared to otherimaging modalities, has yet to be defined.

Despite these concerns, NMR imagers are dif-fusing very rapidly. In January 1983, 14 units werein place in the United States outside manufac-turers’ facilities. By October 1983, 34 units hadbeen installed in the United States, and by Augu st1984, at least 145 units were installed worldwide,of which 93 were in the United States.

Given the rapid rate of change in both the clin-ical and scientific status of NMR imaging, as wellas in the number of units being installed world-wide, it is impossible to publish a review that ac-curately describes the “current status” of NMR im-aging in almost any dimension. Such a reviewquickly becomes outdated as the field continuesto evolve. This case study was written, therefore,with the following limited goals in mind:

To provide a vehicle for gaining insight intothe impact that Federal policies have had onthe development of NMR imaging as a mo-dality, on the industry that manufactures theimagers, on the hospitals and medical centersthat might consider acquiring NMR imagers,

and on a public interested not only in thetimely introduction of valuable innovations,but also in protection from unsafe devicesand rapid increases in health care costs. Byidentifying and analyzing a number of pol-icy issues, the case study is intended to helpthe Federal Government and other interestedparties assess the process through which newdevices are made available.

To make available a large amount of tech-nical, clinical, industrial, and policy infor-mation under a single cover, and in the proc-ess to provide a “snap shot” view of the status

of NMR imaging in several dimensions. z

‘The material was first compiled in fall 1983. App. C and policiesof the Food and Drug Administration and third-part y payers wereupdated in August 1984.

3

8/14/2019 8420


4 • Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology : A Clinical, indust rial, and Policy Analy sis

SUMMARY

The body of the case study is organized intonine chapters. Each of the chapters is briefly sum-marized below.

NMR—Historical andTechnical Background

The existence of the NMR phenomenon wasfirst demonstrated in 1946 by two American scien-tists, Felix Bloch and Edward Purcell, who jointlyreceived the Nobel Prize for Physics in 1952 fortheir discovery. The first NMR image (of twotubes of water) was published by Paul Lauterburof the State University of New York (SUNY) atStony Brook in 1973, the same year X-ray CTscanning was introduced into the United States.Remarkable progress in the quality and capabil-

ities of NMR imaging has been made in the yearssince Lauterbur imaged his two tubes of water,with no plateau in the rate of improvement insight.

The nucleus of the hydrogen atom (proton)3 hasbeen most successfully exploited to produce high-quality NMR images because of its desirable mag-netic properties and the high concentration withwhich it is present in the body. NMR images arefundamentally different from X-ray CT images.The latter rely on partial absorption and partialtransmission of X-rays (linear attenuation) to pro-duce images that reflect differences in the electron

density and specific gravity of adjacent tissues.Proton NMR images are formed without the useof ionizing radiation and reflect the proton den-sity of the tissues being imaged, as well as the ve-locity with which fluid is flowing through thestructures being imaged and the rate at whichtissue hydrogen atoms return to their equilibriumstates after being excited by radiofrequency energy(proton relaxation time). The excitement aboutand investment in NMR have arisen from thebelief that enormous clinical benefits might de-rive from the ability to obtain information aboutboth the tissues of the body and certain kinds of chemical activity.

3Since the hydrogen atom has one unpaired proton, the termshydrogen atom and proton are used interchangeably.

Clinical Applications of NMR

Concerns regarding the safety of NMR imag-ing have focused on magnetic fields and radiofre-quency energy. To date, since adequate precau-

tions have been taken, no significant biologicalrisks associated with use of NMR have been iden-tified. Other potential sources of concern relateto damage that could be caused by the possibil-ity that metallic objects in the vicinity of NMRmagnets could become projectiles, or that thestrong magnetic fields used in NMR imaging coulddamage computer tape or other objects in the sur-rounding environment.

The National Radiological Protection Board inthe United Kingdom is maintaining a record of patients and volunteers who have undergone

NMR imaging studies in order to evaluate prob-lems that arise in the future in individualsundergoing NMR scanning. The American Col-lege of Radiology is attempting to collect similarinformation in the United States. It would seemadv isable to establish u niform g uidelines for w orld-

wide collection of this type of data, at least forthe near future. Issues of who should be respon-sible for collecting and maintaining such data, andat whose expense, as well as issues pertaining topatient confidentiality, remain and need to be re-solved.

The clinical application of NMR imaging inwhich the most experience has been gained andwhich so far has proven most efficacious is im-aging of the brain and central nervous system.Results of studies of NMR imaging of the heartand pelvis are also particularly promising.

The scope of the role of NMR imaging in medi-cine is yet to be determined. Although there issome plausibility to the hundreds of applicationsthat have been cited for NMR imaging, the majority of such applications must, for now, be con-sidered potential rather than demonstrated.

Although future NMR production models arelikely to simplify image acquisition for physicians,and although the time required to produce high-quality NMR images will likely decrease, it shouldnot be assumed that images of the same quality

8/14/2019 8420


. — .Ch. 1—Introduction and Summ ary 5

. — . .-

as those being published by research institutionswill necessarily be produced immediately in hos-pitals that are not able to spend equivalent timeand effort on their production.

In the early stages of evaluation of NMR im-aging, many, if not most, of the patients that have

been studied have appropriately been patientswith known pathologies. It is not necessarily thecase that NMR will be shown to have the samesensitivity and specificity when used to image pa-tients with unknown pathology. Given the rapidimprovements taking place in NMR imaging,however, current assessments may underestimatethe ultimate sensitivity and specificity of NMR im-aging in many applications.

It is likely that algorithms with pulse sequences(patterns of radiofrequency energy used to exciteprotons) specific to different pathologies will be

built into NMR software in the future. While thismeans that NMR images of individual types of pathology are likely to become even better thanthey are today, it also means that if patients withunknown types of pathology are referred for a“screening” NMR scan, multiple scans, usingmultiple pulse sequences, may have to be per-formed in order to exclude with a reasonable de-gree of certainty the existence of an abnormality.Such use of multiple pulse sequences may increasethe time and expense required to perform NMRstudies.

To the extent that use of multiple pulse se-

quences does increase patient examination times,a tension may develop between the economicpressure to maintain reasonable patient flow andthe clinical requirement that pathologic abnor-malities be excluded with a high degree of cer-tainty. To the extent that the latter predominates,the number of patients seen may decrease, pro-ducing a rise in average cost per NMR study, Tothe extent that the former predominates, the sen-sitivity and specificity of NMR may decrease.

Because the risks of NMR imaging appear tobe low, NMR scans maybe performed on patients

repeatedly over time to monitor therapeutic prog-ress or the natural history of disease. Such usagecould result in increased demands being placedon NMR machines and in increased health carecosts .

Within certain numerical ranges, relaxationtimes may provide sufficient pathologic specificityto be clinically u seful. Because of overlap betweenthe relaxation values of normal and abnormaltissues, however, relaxation times alone are unlikely

to permit reliable pathologic diagnoses, despitethe theoretical attractiveness of using such meas-urements. The possibility exists that nontoxic con-trast agents can be devised that will enhance thepathologic specificity associated with relaxationtime values. Considerable research remains to bedone in the exploration of what physical, chemi-cal, and biological factors give rise to and influ-ence NMR relaxation times. Only through answersto these questions will it be possible to exploitrelaxation times’ full medical and scientific po-tential.

NMR is also used to perform in vivo phospho-rus NMR spectroscopy, in which the “chemical

shift” phenomenon is used to provide an indica-tion of the relative concentrations in which com-pounds such as phosphocreatine, adenosine tri-phosphate, and inorganic phosphate are presentin intact human tissues or organs. Much addi-tional research is required before an assessmentcan be made of the extent to which in vivo NMRspectroscopy can be used to provide diagnosticallyuseful information regarding the metabolic andfunctional status of normal and abnormal tissues.

It should also be recognized that the technol-ogy required for in vivo human NMR spectros-

copy is considerably more sophisticated than thatrequired to perform proton NMR imaging. Thus,most of the NMR imagers that are generally be-ing installed in hospitals today cannot currentlybe used to perform NMR spectroscopy. Hospi-tals desirous of performing spectroscopy and im-aging may need either to obtain more than oneNMR machine or to tolerate potentially costlyamoun ts of time w hile field strengths are changedand the NMR machine is not operational. For thepresent, in vivo NMR spectroscopy should beconsidered an exciting and promising area of re-search that is of questionable feasibility for most

hospitals.

The NMR Imaging Device IndustryThe NMR imaging device industry, as it now

exists, is both dynamic and intensely competitive.

8/14/2019 8420


6 Ž Health Technology Case St udy 27: Nuclear Magnetic Resonan ce Ima ging Technology : A Clinical, indust rial, and Policy A naly sis

Since 1976, at least 23 companies worldwide havedecided to enter the NMR imaging marketplace.Eight firms have reached an advanced stage of development, whereas at least three others areengaged in intermediate-level activities. The in-dustry has a multinational character, with firmsbased in the United States, Japan, West Germany,

Great Britain, France, Israel, and The Nether-lands. All but three of the firms have multipleproduct lines. The industry appears concentratedamong four firms, which accounted for 79 per-cent of the 145 known worldwide placements ex-isting in August 1984.

At present, small firms can enter the market,but entry depends on several key factors, includ-ing their ability to attract capital and scientific ortechnical talent for research and development(R&D), to develop strong university or medicalcenter ties for collaborative research, and to mar-

ket products once they have been developed. Thepathways to market entry are varied, but involveessentially four different routes: government-supported R&D, university-based R&D, acquiredtechnology, and internally based R&D. Initialcapitalization for market entry is estimated to bebetween $4 million and $15 million. Universityor medical center research ties are consideredessential in the industry, and every firm that hasattained either intermediate or advanced stageR&D has a close collaborative relationship withone or more universities or major medical centers.

The existence of at least 19 NMR imaging de-vice manufacturers suggests that patents have notcreated a significant barrier to the entry of com-petitors into the marketplace. Whether patentablediscoveries will emerge, prohibitively expensivecross-licensing agreements will be devised, or pen-ding lawsuits will be settled in such a way as tochange this situation is difficult to predict. It isalso difficult to assess how beneficial the protec-tion afforded by patents has been to the commer-cial development of NMR imaging in this coun-try. It is possible, if not likely, that many manu-factur ers have opted to retain discoveries as “trad e-

secrets, ” rather than to reveal confidential infor-mation in patent applications.

There is considerable diversity in the productlines and operations of firms in the NMR imag-ing industry. Sixty-three percent of the companiesmanufacture non-health-care related products

either directly or through a parent firm. Since the1970s, the NMR imaging device industry haswitnessed a large number of acquisitions, mergers,and trade agreements. At least three mergers inthe industry have involved vertical integrationeither to acquire magnet manufacturing capabil-ities or to expand sales or distributorship networks

to specific geographic areas. Vertical integrationis expected to increase in the industry over thenext 2 to 5 years.

Most firms in the industry believe that non-price factors will prove more important thanproduct price in determining future NMR salesand market share. Product differentiation is ex-pected to figure prominently in the non-price com-petition strategies of NMR imaging device firms.Manufacturers believe that the most importantfactors are likely to be image quality, productfeatures or capabilities, product reliability, and

product service. Various manufacturers are plac-ing different emphasis on these factors as part of their marketing strategies. Buyers’ perceptions of a corporation’s chances for long-term survival willprobably also be important.

It is expected that NMR imaging sales willbecome an important source of compan y revenuesfor many manufacturers over the next few years.Firms are expected to maintain heavy investmentin R&D activities even after receiving Food andDrug Administration (FDA) premarket approvaland introducing commercial NMR imaging pro-

totypes. NMR sales could increase from $100 mil-lion per year in 1983 to $2.5 billion per year in1988, amounting to an annual rate of growth insales of 90 percent. The percentage of diagnosticimaging industry sales attributable to sales of NMR imaging systems could increase from 2.5percent in 1983 to 30 percent by 1988.

Hospital Costs and Strategies

One of the major concerns that has emergedregarding NMR imaging relates to the impact thisnew technology will have on health care costs.

These concerns derive in part from the high an-ticipated costs associated with the purchase andinstallation of an NMR imaging system and fromuncertainties regarding the extent to which NMRimagers will be used in conjunction with otherdiagnostic modalities.

8/14/2019 8420


—— . ———

Ch. 1—Introduction and Summ ary 7

Capital and operating expenses for NMR im-aging are primarily determined by the type of magnet (resistive, perman ent, or superconducting)used to produce the static magnetic field. Purchaseprices range from approximately $800,000 for aresistive system to $1.5 million for a permanent

magnet system or a 0.5 tesla superconducting sys-tem and to over $2 million for a 1.5 tesla super-conducting system. Installation costs range from$25,000 to $75,000 for a permanent magnet sys-tem to up to $1 million for a 1.5 tesla supercon-ducting system. Estimates of the average cost of an imaging study, exclusive of professional fees,are difficult to make at this time, but range fromas low as $180 for a resistive system to as highas $700 for a superconductive system. These esti-mates are quite sensitive to a number of keyassumptions, such as the time needed to processpatients.

The likely effect of NMR imaging on h ealth carecosts will depend on how it is employed by phy-sicians in actual practice situations. Several fac-tors need to be considered in this regard. First isthe extent to which NMR imaging is performedinstead of other diagnostic modalities in the man-agement of specific patient complaints or diseaseentities. Second is the extent to which NMR isused in situations in which no diagnostic modalityis currently used. Such situations are likely to in-clude the use of sequential NMR scanning to mon-itor the natural history of diseases and the prog-ress of chemo- and other therapies. Finally, muchwill depend on such factors as how much surgeryis avoided, whether hospital lengths of stay areshortened, and whether diagnostic workups thatwere performed in the hospital are shifted to theoutpatient setting.

Most of the early NMR units acquired by hos-pitals have been installed in university teachinghospitals. This situation is not surprising, giventhe interest such hospitals have in performing re-search and being at the “cutting edge” of medicaldevelopments, and given the research needs of manufacturers in order to obtain FDA premarket

approval. In addition, university hospitals havebeen able to use their special strengths to obtainNMR imaging systems at decreased or nominalcost. Price and operating costs of experimentalsystems have frequently been further subsidized

by research grants from manufacturers and haveoften been shared between hospitals and univer-sities. These observations suggest that many of the university hospitals that have obtained NMRimaging systems to d ate may have done so in p artbecause they did not have to be so concerned with

acquisition costs and early operating costs as otherhospitals have to be.

In 1983 the Veterans Administration (VA) de-cided to initiate a staged program of acquisitionof NMR devices with a single NMR demonstra-tion and evaluation project. The decision to ac-quire an NMR device for the VA system derivedfrom an interest in “helping the VA march intothe future” (171). No estimates of the impact of NMR on the cost of patient care were made. Thedecision to restrict the initial purchase to a singleunit emanated from a concern about the rapid rateat which NMR technology was changing and the

resultant desire to avoid installing a large num-ber of systems that might soon become obsolete.NMR manufacturers have suggested, however,that due to the ability to upgrade their systems,early obsolescence may be less of a problem withNMR imagers than it was with X-ray CT.

Investor-owned hospitals have also followed acautious approach to acquisition of NMR imag-ing equipment. The Hospital Corp. of Americaand Humana, for example, have each decided toacquire a small number of systems in the nearfuture in order to conduct in-house evaluations

of the cost, utility, and ideal configuration of NMR imaging systems in the community hospi-tal setting. Others, such as American Medical In-ternational, National Medical Enterprises, andLifemark, have postponed acquisition of NMRequipment until additional information regardingthe cost, utility, and reimbursement rates forNMR imaging is available. Finally, investor-owned companies that operate hospital chainsplan to use their ability to buy in volume to ob-tain special price consideration from manufac-turers.

History of Funding for NMR Research

In the United States, both the National Insti-tutes of Health (NIH) and the National Science

25-341 0 - 84 - 2 : QL 3

8/14/2019 8420


8 • Health TechnologyCase Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis

Foundation have provided considerable supportto basic NMR research over the past decade. NIHis currently funding approximately $2 million inresearch in at least 10 different institutions relatingto NMR imaging or in vivo spectroscopy. An-nouncement of awards from the Diagnostic Im-aging Research Branch of the National Cancer In-stitute to assess the comparative efficacy of NMRimaging and other diagnostic modalities weremade in mid-1984.

At least three different noncommercial entitiesprovided support for NMR research in Englandand Scotland over the past decade. These includethe Wolfson Foundation, and two governmententities, the Medical Research Council and theDepartment of Health and Social Security inEngland.

Certain contrasts between the history of the de-velopment of NMR imaging in the United Statesand Great Britain can be identified. Unlike the sit-uation in the United States, in Britain the govern-ment undertook a concerted effort to developtechnology that might be of use specifically in hos-pitals. This effort was focused through a programfunded by the Department of Health and SocialSecurity which lent considerable financial supportto the development of NMR imaging techniques.It is interesting to note, however, that once itbecame apparent that the development of NMRimaging systems was not only commercially vi-able, but also potentially extremely profitable,

U.S. manufacturers rapidly and intensively beganinvesting in NMR imager development programs.

In Britain there also seem to have been severalinterdisciplinary groups that collaborated on thedevelopment of NMR imaging techniques. In theUnited States, in contrast, most of the early workon NMR imaging was done by Lauterbur andDamadian with apparently little, if any, interac-tion between the two, despite the fact that bothwere at campuses of the State University of NewYork. There also seem to have been fewer centersin the United States in which scientists with var-ied backgrounds collaborated on the type of in-

terdisciplinary research that resulted in the ad-vances in NMR imaging that took place in Britain.

FDA Regulation

FDA authority over NMR imaging devicesderives from two Federal acts: the Radiation Con-trol for Health and Safety Act (RCHSA) of 1968and the Food, Drug, and Cosmetic Act (FDCA),as amended in 1976. FDA has not established ra-

diation emission performance standards for NMRdevices under its RCHSA authority, and it is notlikely that the RCHSA will have a significant im-pact on the development of NMR imaging as amedical diagnostic modality. The FDCA, in con-trast, has had and continues to have a significantimpact on the development of NMR imagingdevices.

The 1976 Medical Device Amendments requirethat all medical devices be classified into one of three regulatory categories based on the extent of control necessary to provide reasonable assurance

of safety and effectiveness. NMR imaging devicesare the first imaging devices to be classified intoClass III for which premarket approval (PMA) hasbeen required. The premarket approval applica-tions (PMAAs) submitted by three companieswere d eemed “app rovable” by the FDA RadiologicDevices Advisory Panel in July 1983, and weregranted formal premarket approval by FDA inspring 1984.

Some general insights into the PMA process canbe gained from examining how NMR has faredin its interactions with it to date. It should be

realized, however, that the experiences that an ex-tremely promising, high R&D-cost device such asNMR has had with the FDA may not be repre-sentative of those that other devices may have inthe future.

In the case of NMR, it appears that the FDAPMA process is primarily playing a quality-assurance role—a role that Congress intended itto play. PMA does not ap pear to h ave constrainedNMR technological development. However, in itsattempt to assist manufacturers and institutionalreview boards to define when experimental useof NMR does not pose a significant risk, FDA mayhave influenced the technological development of NMR devices.

8/14/2019 8420


FDA clearly has not constrained the number of NMR imagers that could be installed on an ex-perimental basis in the United States. Of the ap-proximately 34 NMR systems installed in theUnited States by October 1983, 15 were by a singlemanufacturer. It appears, therefore, that the FDA

PMA process will not act as a major constrainton the rate at which NMR devices are adoptedand used throughout the United States. This sit-uation may, in large part, be a result of the longgestation period required for development of aproduction model of a high R&D-cost device, suchas an NMR imager.

If PMA is not granted to other manufacturersin a timely fashion, however, manufacturers maybegin to suffer from delays in receiving revenuesto cover their development costs. Because theHealth Care Financing Administration (HCFA) re-

quires FDA approval of a device before it ap-proves coverage for it, undue delay in PMA couldinjure manufacturers because of the constraininginfluence that the absence of Medicare reimburse-ment would have on hospital acquisition decisions.

Two final impacts of the FDA PMA processshould be identified. First, in their quest for PMA,manufacturers have subsidized a considerableamount of research in order to establish the safetyand effectiveness of NMR imaging devices. Howmu ch of this research w ould h ave been subsidizedor performed by manufacturers in the absence of the PMA process is impossible to estimate. Fi-

nally, it appears that the PMA process may provecapable of conferring a competitive advantageupon those manufacturers who are first to receivePMA, particularly if third-party payers decide toapprove coverage only for those manufacturers’devices that have received PMA. How much of a financial benefit, in both the short run and thelong run, accrues to those “early bird” manufac-turers who obtain PMA while others still awaitit may help determine not only the future of theNMR manufacturing industry, but also the speedwith which manufacturers pursue development of other new technologies in the future.

Third-Party Payment Policies

In determining coverage policy for new medi-cal technologies, third-party payers look first toFDA for some indication of a device’s status.

Ch. 1 —In t roduct ion and Summ ary 9 — — . .

Third-party payers generally will not reimbursefor clinical services performed with “investiga-tional” devices. HCFA will provide coverageunder the Medicare program only for those de-vices, services, or procedures that are determinedto be both “reasonable and necessary. ” HCFA

generally does not approve coverage of a new de-vice unless FDA has already found it to be “safeand effective. ” FDA determination of safety andeffectiveness, however, does not ensure that thedevice will satisfy HCFA’s criteria of reasonable-ness and necessity.

Other third-party payers, such as State Med-icaid programs, Blue Cross and Blue Shield plans,and private insurance companies consider simi-lar factors in making coverage decisions, but varyin their general procedures, methods of assess-ment, and decision criteria.

HCFA conducts the most in-depth assessmentof a new technology, with the aid of the PublicHealth Service’s Office of Health TechnologyAssessment (OHTA).4 In performing a technol-ogy assessment, OHTA gathers and analyzes rele-vant data on clinical safety and efficacy fromvarious public and private sources. The assess-ment process often takes between 8 and 18 monthsto perform.

The national Blue Cross and Blue Shield Asso-ciation also conducts technology assessments atthe request of member plans. Association staff re-

view available literature and elicit expert opinionfrom medical specialty societies in determining thesafety and effectiveness of a new device. Staff assessments of new technologies frequently resultin Uniform Medical Policy statements, which areintended only to guide coverage policy decisionsof member plans. Each plan, however, may makeits own independent coverage decision.

Commercial insurance companies follow a lessformal procedure in conducting technology assess-ments. The Health Insurance Association of Amer-ica (HIAA), a private organization serving thecommercial insurance industry, furnishes infor-

mation on new technologies to its members. Atthe request of a member company, HIAA will so-licit an expert opinion regarding a new device

‘This executive office differs from the Health Program in the Con-gressional Office of Technology Assessment.

8/14/2019 8420


10 Ž Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis

from the Council on Medical Specialty Societies.The information will be synthesized and forwarded

to member companies, who independently inter-pret it and make coverage policy decisions.

The major third-party payers also differ in thecriteria they employ in setting payment levels for

covered services. Important factors in these deci-sions include where the technology will be used(e.g., hospital, physician’s office), in what circum-stances it will be used (e.g., certain clinical situa-tions or diseases), and by whom it will be used(e.g., physicians with general versus specialtytraining). Payment levels are generally based oncriteria of “prevailing, customary, and reason-able” charges, allowing for differences in geo-graphic area, past experience of individual prac-titioners, and prevailing market prices or fees.

Third-party payers are evaluating their cover-

age of NMR imaging. Some third-party payershave already begun to pay for NMR scans. Tech-nology assessments of NMR imaging are now be-ing performed by OHTA (for the Medicare pro-gram) and by the Blue Cross and Blue ShieldAssociation.

State Certificate-of= Need Programs

Although State certificate-of-need (CON) pro-grams were never specifically intended to con-strain the diffusion of medical technology, theyconstitute one of the major policy mechanisms

available to health planners for control over tech-nology adoption. CON review of “need” may bebased on numerous factors, including clinical useof technology, institutional characteristics, eco-nomic and financial effects, and population-basedconsiderations. In the past, CON programs haveemployed at least four different policy orienta-tions or strategies regarding technology introduc-tion and distribution: pro forma denial, formal-ized strategy of delay, predetermined limits ondiffusion, and uncontested approval.

The CON experience with X-ray CT scannerspoints out two problems that could arise in thefuture with NMR imaging: the fragility of shared-service arrangements am ong hospitals and the cre-ation of incentives that encourage “anticipatoryacquisition” of new technology. The latter situa-tion can produce a “franchising” effect wherebyhospita ls that adopt technology early—often

while the technology is still considered “investiga-tional’’—become well-positioned to keep the tech-nology once its status changes and diffusion ac-celerates. Those hospitals that wait to submitCON applications risk being “disenfranchised”from obtaining the technology.

Various State and local planning agencies re-port increasing CON activity related to NMR im-aging. As of September 12, 1983, at least 33 CONapplications for NMR had been reviewed nation-wide. Of these, 19 had received approval: 16 byState Health Planning and Development Agen-cies and 3 by local Health Systems Agencies.Twenty-five health planning agencies across theNation also reported that they either had NMR-specific review criteria in force or were planningto develop them in the near future. Pending orrecently enacted State legislation or regulationsrelated to NMR were reported in at least six

States.Several distinct CON strategies regarding NMR

appear to have emerged among th e States. For ex-ample, New York, Illinois, Ohio, New Jersey, andKentucky have each ad opted p redetermined limitson NMR imager diffusion. The Southeast KansasHealth Systems Agency has invoked a moratori-um on NMR until community hospital planninghas been completed. The District of ColumbiaCON program also has statutory power to employa formalized strategy of delay. Nebraska, by con-trast, is encouraging group applications involv-

ing shared-service arrangem ents among hospitals.Utah and California, through recent amendmentsto their respective State CON laws, appear to befollowing a strategy of uncontested approval forNMR imagers. No CON program, on the otherhand, has adopted a policy of pro forma denial.It is anticipated that CON agencies will witnessa rapid increase in the number of NMR applica-tions filed by hospitals once HCFA policies re-garding NMR are finalized.

Regulatory Overview

Since FDA has granted premarket approval tothe first NMR imaging manufacturers, third-partypayers have a position of major influence overthe rate at which NMR imagers are acquired byhospitals. This influence will derive from theirdecisions regarding: 1) whether to cover use of

8/14/2019 8420


Ch. l—introduction and S u m m a r y 1 1

NMR imaging at all; 2) whether to cover NMRdevices only of those manufacturers that have re-ceived premarket approval or those of any man-ufacturer; 3) which types of NMR scans to cover(e.g., head stu dies only or head and body stu dies);4) the monetary level at which use of NMR willbe reimbursed; and 5) the level at which profes-

sional fees for NMR imaging are set. If initial cov-erage of NMR is limited to a small number of clin-ical circumstances or reimbursement rates do notreflect the increased professional time that will ini-tially be required for NMR scanning, hospitalsmay be restrained in the speed with which theyacquire NMR devices.

The introduction of prospective payment basedon diagnosis related groups (DRGs) under Medi-care is also expected to affect the rate of diffu-sion of NMR devices into hospital settings. Hos-pitals now have to weigh financial considerationsagainst patient care benefits more carefully whendeciding whether to acquire an NMR imager andin deciding how an acquired NMR scanner is tobe used. For some hospitals, such as municipalfacilities serving large Medicare and Medicaidpopulations, the DRG payment system may ex-acerbate an already financially distressed situa-tion and further impede those institutions’ effortsregarding capital formation. The net effect maybe to weaken the hospitals that serve as primarysources of care for disadvantaged populations.The ultimate impact of the prospective paymentsystem on acquisition of NMR scanners is likelyto depend on future HCFA decisions regardingrecalibration of DRG payment rates to take ac-count of introduction of new technology over timeand regarding inclusion of capital expenditures inthe DRG rate,

The final major regulatory influence on the rateat which new technology, such as NMR imagers,diffuses throughout the medical system is Statecertificate-of-need (CON) policies. There isalready evidence that CON agencies are delay-ing the acquisition of NMR devices by some hos-pitals. Whether State agencies are adequately in-formed to be able to make appropriate decisions

regarding whether and when NMR scanners shouldbe introduced into hospitals is questionable.

A number of problems with CON policies thathave appeared over the past decade in the experi-

ence with X-ray CT are likely to affect the courseof NMR as well. Evidence for “franchising” and“anticipatory acquisition” of NMR is alreadyavailable and will need to be addressed by CONprograms. In addition, there is evidence of con-siderable interest on the part of private radiologygroups, as well as hospitals, in establishing out-

patient diagnostic centers which will include, butnot be limited to, NMR devices. In most States,such ambulatory placements do not require CONapproval. If State agencies are interested in con-trolling the introduction of new technologies, suchas NMR, they will have to address themselves tothis limitation in their purview. Alternatively,CON agencies could leave control over the “in-troduction” of technology to the FDA and third-party payers and concentrate on playing a com-plimentary role by assuring equitable distributionof new technologies within their jurisdictions.

In addition to these influences on the rate atwhich new technologies such as NMR imagers dif-fuse, two final policy issues should be addressed.First, there appears to be a large amount of du-plicated effort on the part of FDA, third-partypayers, CON agencies, and hospitals with regardto the assessment of new technology. Althoughit is unclear whether it would be beneficial to in-crease the coordination among these separatetechnology assessment efforts, the issue should beaddressed. If HCFA is to continue relying on thePublic Health Service’s OHTA as an impartialsource of advice, attention should be given to

whether the resources available to OHTA areadequate.

Finally, as more technologies become available,it becomes increasingly important that the com-parative efficacy of each be adequately evaluatedand defined. How such comparative efficacy datawill be acquired and who w ill fund th e studies nec-essary to generate them are increasingl y impor-tant issues that the Federal Government andothers need to address if appropriate reimburse-ment policy decisions are to be made. In the case

of a rapidly evolving technology, such as NMRimaging, the question of when to perform suchcomparative assessments also needs to be ad-dressed. This “moving target” issue has ham peredcomparative efficacy assessments in the past.

8/14/2019 8420


2.NMR—Historical andTechnical Background

8/14/2019 8420


2.

NMR— Historical andTechnical Background

It is not possible to provide either a comprehen-sive historical profile of nuclear magnetic reso-nance (NMR) or a detailed technical explanationof the NMR phenomenon within the scope of thisdocument. In order to fully appreciate the excite-ment about the implications of being able to pro-duce hydrogenNMR, however,some historical

or other atomic images usingit is essential that the reader haveand technical background. The

HISTORICAL BACKGROUNDThe existence of the phenomenon of nuclear

magnetic resonance was predicted by a Dutchphysicist, Gorter, in 1936. Gorter sought, unsuc-cessfully, to demonstrate the NMR phenomenonin lithium fluoride. A decade later, in 1946, twoAmerican scientists, Felix Bloch and EdwardPurcell, working independently, simultaneouslydiscovered and demonstrated the existence of NMR. Bloch’s observations w ere mad e with stu diesof water at Stanford; Purcell’s with studies of par-affin wax at H arvard. The two were jointly award ed

the Nobel Prize for Physics in 1952. Since thenchemists and physicists worldw ide have r outinelyemployed uniform magnetic fields in what cannow be considered “conventional NMR spec-troscopy” to study the molecular structure anddynamics of small homogeneous specimens (8).The NMR imaging techniques that have evolvedover the past decade derive in large part from the25 years of experience that had been accumulatedprior to 1973 in the application of NMR spec-troscopic techniques to the study of solids andliquids.

The establishment of a magnetic field gradient(a magnetic field that increases or decreases instrength in a given direction along a sample)across a sample was the key to going from spec-troscopy to spatially encoding the informationthat forms the basis of NMR tomographic imag-

following sections attempt to provide that back-ground. The first section discusses the historicaldevelopment of NMR. The second section pro-vides basic technical background about NMR andNMR imaging, including a description of the tech-nical components used in NMR imaging systems.The final section introduces the types of magnetsused in NMR imaging. Appendix A contains ad-ditional technical information.

ing. Although magnetic field gradients had beenemployed by scientists since the 1950s in studiesof molecular diffusion in liquids (78), phase sepa-ration (separation of homogeneous but physicallydistinct portions of matter) in helium solutions(199), and methods of information storage (7), itwas not until 1971 that Paul Lauterbur workingat the State University of New York (SUNY) atStony Brook conceived of the idea of manipulat-ing magnetic field gradients to obtain a two-dimensional NMR image (116).1 In his now classic

experiment in w hich the first NMR image was p ro-duced, Lauterbur rotated magnetic field gradients(changed magnetic field gradients) in a techniquehe called zeugmatography to reconstruct a two-dimensional image of two tubes of water (115) (seefig. 1). In d iscussing the implications of his resu lts,Lauterbur recognized the potential applicabilityof his technique to the imaging of soft tissue struc-tures and malignant growths in vivo (115).2

ILauterbur was aware of the studies performed by Damadian (42)and Hollis (93), which demonstrated that excised tumors manifestedprolonged NMR relaxation times. Recognizing that it might betremendously beneficial to be able to make such measurements in

vivo, Lauterbur worked to develop a technique in which NMR couldbe used to produce images (116).

‘In 1973, Mansfield and Grannell published a letter in which theydescribed a method, involving magnetic field gradients, throughwhich NMR could be used to determine spatial structure in solids(121). No mention is made in the letter, however, of a proposal touse NMR to produce images.

15

8/14/2019 8420


16 • Health Technology Case Stud y 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, industrial, and Policy Analysi s.- -. .— — —

Figure 1 .—First NMR Image

t

:.

.

NMR image of two tubes of waterSOURCE P C Lauterbur, “Image Formation by Induced Local Interactions. Ex-

amples Employing Nuclear Magnetic Resonance, ” Nature 242 (5394)190.191, Mar 16, 1973,

Remarkable progress in the quality and capa-bilities of NMR imaging has been made in the 10years since Lauterbur imaged his two tu bes of wa-

ter. In 1974, Peter Mansfield and his colleagues

BASIC TECHNICAL BACKGROUND

An understanding of nuclear magnetic reso-nance (NMR) requires a familiarity with certainnatural phenomena. The first phenomenon is thatall atoms, of which everything in nature is made,contain nuclei which, in turn, are made up of par-ticles called protons and neutrons. It is theseatomic nuclei to which the “N” in NMR refers.

The second natural phenomenon pertinent toan understanding of NMR is that certain nuclei,namely those that contain an odd number of pro-tons, or an odd number of neutrons, or both,

at Nottingham University pu blished the first crudeNMR medical image (of a human finger) (122).Only the gross outline of the finger without anyinternal detail was revealed. Improved images of human fingers were produced by the same group2 years later using a different imaging techniquethat relied on selective radiation 3 of the specimen

in sw itched m agnetic field grad ients (123). In 1976,Damadian and colleagues, working at SUNY atBrooklyn, employed a Field Focussing NuclearMagnetic Resonance technique (FONAR) to pro-duce the first NMR image of a tumor in a live ani-mal (44). A year later a human wrist was imaged(91) and the first in vivo human whole-body N MRtomographic scan (image of an individual slice)was produced (43). In the latter scan, crude bycurrent standards, the heart, lungs, mediastinum,and descending aorta could be detected (43) (seefig. 2). In 1978a team led by Hugh Clew and IanYoung, working at English Music Industries’(EMI’s) laboratories in London, produced whatis believed to be the first NMR image of a humanhead (96) (see fig. 3). Since then considerable im-provements have been made in NMR imaging of both the head and body, with no plateau in therate of improvement in sight (see fig. 4).

3The radiant energy used for NMR is low-frequency, non-ionizing

radiofrequency waves, not the high-frequency waves used in X-rays.

possess an intrinsic angular momentum, called“spin.” Since nuclei are electrically charged, thosenuclei that spin generate tiny magnetic fields. Thatis, they are magnetic. Only those nuclei that aremagnetic, such as 1Hydrogen, 13 Carbon, 19Flourine,23Sodium, and 31P h o sp h o ru s , can be exploited in

NMR experiments. It is this phenomenon of nu-

clear magnetism to which the “M” in NMR refers.Supplying radiofrequency energy of the appro-

priate rotational frequency will excite hydrogennuclei from a lower energy level, E1, to a higher

8/14/2019 8420


— —..—.

Ch. 2–NMR Historical and Technical Background 1 7

Figure 2.–First In Vivo Human Whole-Body NMR Scan

SOURCE Provided by Raymond Damadien, President, FONAR Corp

level, E2. If the radiofrequency energy is turned

off after the n uclei have been raised into th e higherenergy level, the excited hydrogen nuclei dropback down to level E1, i.e., they relax. In the proc-ess of relaxing, the hydrogen nuclei re-emit theenergy they had initially absorbed, If this radiofre-quency energy is repeatedly applied, hydrogennuclei will oscillate, or resonate, back and forthbetween El and E2, alternately absorbing and emit-ting energy. It is this type of radiofrequency-induced resonance to which the “R” in NMRrefers.

Since the NMR signals that are emitted by mag-

netic nuclei are extremely weak, atoms must bepresent in sufficient concentration in order to pro-duce an NMR signal that is strong enough to beconverted into an image exhibiting clinically use-ful spatial discrimination. To date, the nucleus of

the hydrogen atom, which is the most prevalent

in the body and has a single unpaired proton, hasbeen most commonly exploited to produce high-quality NMR images.4

As Raymond Andrew has explained in his re-view of NMR imaging (8), NMR images are spatialrepresentations of NMR signals. Although thesignal detected in proton imaging is proportionalto proton density, the image is not just a two-dimensional representation of that proton d ensity.Rather, the signal also depends on three otherparameters.

The first parameter is the velocity with whichfluid is flowing through the structure being im-aged, since the movement of protons in that fluid

4Recently, sodium and phosphorus have also been imaged in someresearch centers.

8/14/2019 8420


18 • Health Technology Case Stu dy 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, an d Policy Analysis

rameter, called the “spin-lattice” relaxation time,or Tl, is a time constant that reflects the rate atwhich excited protons exchange energy with thesurroun ding environm ent. The other, called “sp in-spin” relaxation time, or T2, is a time constantthat reflects the rate of loss of coherence (the rateat which protons stop rotating in phase with each

other) due to the local magnetic fields of adjacentnuclei. Naturally occurring variations in relaxa-tion times may have biomedical significance.

The extent to which any single NMR imagereflects each of these four parameters (proton den-sity, flow, T1, and T2) depends on the particularradiofrequency pulse sequence employed to ex-cite the protons in a region being imaged (see app.A). Thus, there is no such thing as a unique NMR“picture” of any region of the body. Rather, asis illustrated in figure 5, NMR images of a singleregion vary depend ing on the pu lse sequence used

to produce them.NMR images thus are fundamentally different

from computed tomographic (CT) X-ray images.Whereas the latter rely on the linear attenuationof ionizing X-radiation to produce images that re-flect differences in the electron density and spe-cific gravity of adjacent tissues, NMR images areformed without use of ionizing radiation and re-flect fundamental physiochemical differences be-tween adjacent tissues. It is from the belief thatenormous clinical benefits might be derived fromobtaining information at a nuclear level through

NMR, that the excitement about and investmentin NMR have arisen.

Except for the addition of a computer and a sys-tem for producing a magnetic field gradient, thebasic components used in modern day NMR im-aging devices (see figs. 6 and 7) are qualitativelysimilar to those employed in the first NMR ex-periments performed by Bloch and Purcell in1946. These components include: 1) a m a g n e t wh ose aperture or bore (diameter) is large enoughto enclose the structure being imaged (the magnetis used to pr oduce a h ighly uniform m agnetic fieldaround the structure being imaged); 2) a set of

gradient cods to impose the magnetic field gra-

8/14/2019 8420


.— . . . ———

Ch. 2–NMR Historical and Technical Background 1 9————— — -- ——

Figure 5. —An NMR Image (3 mm slice) of a Normal Head From an Axial View With Changesin Pulse Sequence

SOURCE General Electric Co , 1984

client required to provide the system with spatial during the process of relaxation; 5) a computer discrimination; 3) a radiofrequency transmitter to to con trol inst rument operat ion and to reconst ructproduce radiowaves that excite the nuclei being and store the image produced from the NMR fre-imaged ; 4) a radiofrequency receiver to detect the quency signals being detected; and 6) a displayradiofrequencies being emitted by excited nuclei system.

MAGNETSAlthough small-bore magnets had been em- bores large enough to accommodate a human be-

ployed in conventional NMR spectroscopy for ing were designed and built. Much of the recentmany years, it was not until interest in NMR im- research and development on magnets for NMRaging emerged in the 1970s that magnets with imaging and in vivo spectroscopy h as been funded

8/14/2019 8420


20 ŽHealth Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A C/inical, Industrial, and Policy Analysis

Figure 6.—Schematic Diagram of NMR Scanner Instrumentation

MAGNET

I

CONTROLINTERFACE

ANALOGTO DIGITAL

CONVERTER/ \

IMAGEDISPLAYSYSTEM

4

COMPUTER

SOURCE C L Partain, R, R. Pricel J, A, Patton, et al, “Nuclear Magnetic Resonance Imaging,” in Radiological Society of North America, Inc.l 1984, figure 11, D 13.(Courtesy of C. L. Partain).

in part by the NMR imaging industry and car-ried out by magnet manufacturers. The design of magnets manufactured specifically for NMR im-aging, however, is still in an early stage of evolu-tion, with improvements likely to be made as in-terest intensifies.

There are four main characteristics of magnetsused in NMR scanners with which one shouldhave some familiarity: magnet type, field strength,bore size, and homogeneity of field.5

Magnet Type

Two different classes of magnets can be usedto produce the static magnetic field employed inan NMR scanner: electromagnets (either resistiveor superconductive) or permanent magnets. Re-sistive magnets use electric current carried by cop-per or aluminum wire to create a magnetic field.Because copper and aluminum offer resistance tothe flow of electric current, power must be sup-plied to force current through the wire. Energysupplied to power the system is lost as heat, re-quiring employment of a cooling system (usually

5Individuals interested in more than the following brief descrip-tion of these features can consult standard physics texts or reviewswritten about magnets used in NMR imaging systems (57,126).

cold water). Resistive magnets are comparativelylight and inexpensive, and, because they can beshipped in parts, the costs of installing them arecomparatively less than that for superconductivemagnets. Resistive magnets do have high powerrequirements (50 to 100 kilowatts) resulting inoperating costs of about $20,000 per year (6),however, and field strength is limited by coolingconsiderations to about 0.15 tesla 6 (126).

Superconductive magnets also utilize electriccurrent to create a magnetic field, but instead of employing resistive materials such as aluminumor copper wire to carry the current, they use wiremade from superconductive materials such asniobium-titanium alloys. Such alloys offer no re-sistance to current flow when operated at tem-peratures near absolute zero (126). In contrast toa resistive magnet, a su perconductive magnet doesnot require an external source of electrical energyto sustain current flow once it has been started,as long as temperatur es are maintained near abso-lute zero (126). Cooling of superconductive mag-

nets is accomplished through use of liquid heliumand liquid nitrogen. Helium needs to be replen-

bMagnetic field strengths are measured in units of tesla (T).1 T = 10,000 gauss (10 kilogauss). For perspective, the magneticfield strength of the Earth is approximately half a gauss.

8/14/2019 8420


Ch. 2—NMR—Historical and Technical Background 21 ——. ——— . — —.——

Figure 7.—Modern Day NMR Imager

SOURCE Picker International

ished approximately once per month, whereas ni-trogen needs to be replenished every one to twoweeks.

Manufacturers are developing systems to reducethe loss, and therefore expense, of these coolants.Considerable research, for example, is being di-rected at development of techniques for recycl-ing and liquefying the helium that now boils off into the atmosphere. In the event of excessivehelium boil off, a superconductive magnet can

quench (i.e., lose its superconductive properties).At least a day can be required to recool the mag-net, during which time the instrument is unusable(126). The primary advantage of superconductive

magnets is that they can carry high current den-sities, enabling generation of very high magneticfield strengths (see “Field Strength” below). Super-conducting magnets can also provide magneticfields that are both highly uniform and stable,once equilibrium is established.

Probably the most serious problem associatedwith use of resistive or superconductive magnetsfor NMR imaging derives from the external mag-netic fields produced by the magnets and disrup-

tions in the magnet’s magnetic field produced byferromagnetic objects (e.g., passing vehicles orelevators) in the vicinity of the magnet. The fringefield produced by electromagnets can erase mag-

8/14/2019 8420


22 Health Technology Case Study 27: Nuclear Magnetic Resonance imaging Technology: A Clinical, industrial, and Policy Analysis

netic tape and disrupt pacemakers. Magnetic objects in the environment, in turn, can cause unac-ceptable distortions in the primary magnetic fieldthat degrade image quality. Because of these po-tential distortions, expensive preventive site prep-aration or renovation is necessary.

Problems related to stray magnetic fields do notoccur with permanent magnets (126), resulting infewer siting problems with the installation of per-manent magnets. Because permanent magnetseliminate the need for either electrical power orliquid helium, they also have the advantage of lower operating costs. Permanent magnets tendto be extremely heavy (as much as 100 tons), how-ever, often creating a need for reinforced floors.The field strength of most currently available per-manent magnets does not exceed 0.3T.

Field Strength

The optimum field strength for proton NMRimaging is the subject of intensive research anddebate. It is likely that no one field strength willbe optimum for all NMR applications. Higherfield strengths might be preferable to lower onesbecause a higher field strength increases the NMRsignal/ noise ratio, and increased signal to noisetranslates directly into improved image quality,finer spatial resolution, or reduced scan times, allother parameters being equal (21). Still at issue,however, is whether improvem ents in image quality

achieved through increases in field strength above0.3T will result in clinical benefits.

Bore Size

The size of specimens that can be imaged withNMR is limited by the diameter of the magnet

bore. Whole body NMR scanners therefore re-quire m agnets with an effective bore diameter suf-ficient to accommodate a human body. Approx-imately 1 to 3 percent of patients that have beenimaged to date have complained of feeling claus-trophobic in the magnet.

Homogeneity of Field

Inhomogeneities in the magnetic field (lack of uniformity in magnetic field strength) can resultin clinically important distortions in NMR images.As mentioned previously, both stationary and

moving ferromagnetic objects can produce suchinhomogeneities in the fields of resistive andsuperconductive magnets. Many of these prob-lems can be minimized (albeit at significant ex-pense) by magnetic field shimming (adjustmentssuch as addition of special coils made to elimi-nate inhomogeneities in the magnetic field) andappropriate site renovations. Higher degrees of field homogeneity are required to perform 3

1Pspectroscopy than to perform proton imaging.

8/14/2019 8420


3.Clinical Applications of NMR

8/14/2019 8420


3 .

Clinical Applications of NMR

INTRODUCTION

When a new imaging modality is evolving asrapidly as NMR imaging, it is impossible to pub-lish a review that accurately describes the “cur-rent status” of its clinical use, By the time sucha review is published, it is out of date. Therefore,the goals of this chapter are: to provide a “snap-shot” view of the state of the clinical applicationsof NMR in mid-1983 and to attempt to put thecurrent status of various clinical applications intoperspective, highlighting current limitations aswell as advantages and d istinguishing between po-tential applications and those that have been

demonstrated with scientific rigor. More detailedclinical reviews are available (see, for example,27,86,125,137,142,148,166) .

The first section in the chapter provides a brief overview of potential biological hazards of NMR.The rest of the chapter is organized around thethree clinical applications of NMR in which themost research had been performed as of thiswriting: 1) imaging of protons to assess normaland abnormal structure, metabolic processes, andfluid flow; 2) use of proton relaxation times1 inthe diagnosis of disease; and 3) use of the chemi-cal shift phenomenon to assess biochemical proc-esses in vivo (in vivo spectroscopy ).2

‘These relaxation times are estimated from NMR images.‘Imaging of nuclei other than hydrogen, such as sodium or

phosphorus, has recently been accomplished, but is not included

The second section provides a clinical overviewof proton (hydrogen) imaging, briefly summariz-ing results of research on proton imaging of thebrain, mediastinum, hilum, lung, heart, breast,abdomen, kidneys, pelvis, musculoskeletal sys-tem, clinical oncology, and blood vessels. Readerswithout clinical background may wish to skip thesecond section and proceed directly to the thirdsection (Putting Proton NMR Imaging Into Per-spective), which discusses the advantages and dis-advantages of NMR imaging and addresses anumber of clinical and practical issues which help

put the status of proton imaging into perspective,Section 4 (Relaxation Parameters) provides a brief overview of research relating to the clinical useof measurements of proton relaxation times. Again,readers without clinical or scientific backgroundmay w ant to skip th is section and proceed directlyto section 5 (Putting T

1and T

2 Relaxation Times

Into Perspective), which discusses the conclusionsto be drawn from the research described in sec-tion 4. Finally, sections 6 and 7 briefly addressin vivo spectroscopy. Non-scientists may wish toread only section 7, which attempts to place thecurrent status of in vivo spectroscopy into per-

spective.

in this list. A discussion of the accumulated experience with suchimaging is beyond the scope of this case study.

POTENTIAL BIOLOGICAL HAZARDS OF NMR

One of the primary advantages of NMR imag-ing is that it does not employ X-rays. Althoughthere is thus no safety concern regarding X-irradia-tion with NMR imaging, concern has arisen re-garding other potential hazards associated withuse of NMR.

NMR safety concerns have focused on threeNMR parameters: 1) time-varying magnetic fields,

2) radiofrequency electromagnetic fields, and 3)static magnetic fields.

Although the possibility exists that time-varying

magnetic fields could induce electric currents thatcould cause ventricular fibrillation, cardiac arrest,inhibition of respiration, or tetanization (sustainedmuscular contraction), none of these potentialadverse effects has been observed in over 2,000

8/14/2019 8420



NMR studies performed in England (179). Norhave significant problems relating to local heatinginduced by radiofrequency pulses, or pathologi-cal, developmental, genetic, or behavioral hazardscaused by prolonged exposure to static NMRfields been observed (161). In fact, the only com-plications reported at a July 1983 FDA Radiologic

Devices Panel m eeting by Professor Robert Steiner,of Hammersmith Hospital, London, in his discus-sion of the British NMR studies, were that onebaby had vomited and aspirated during position-ing in the NMR scanner and that 3 percent of thesubjects had opted out of having an NMR scansecondary to development of a feeling of claus-trophobia when positioned in the magnet. Othercenters have reported claustrophobic feelings in1 percent of subjects (147).

Potential nonbiological hazards include projec-tiles of metallic objects in the vicinity of NMR

magnets and damage to computer tape or otherobjects in the surrounding environment.

The unblemished N MR safety record is perhapsrelated to the extreme caution that has been ex-ercised by clinical researchers working with NMR.Researchers have, for example, tended to excludefrom their studies patients in whom cardiac pace-makers, large pieces of metal (such as an artifi-cial hip), metallic cranial aneurysm clips, or metal-lic intrauterine devices have been implanted.Pregnant wom en have also tended to be exclud ed.It is possible that these restrictions will be relaxed

in the future. For now, however, no harmful ef-fects have been associated with NMR imagingunder existing conditions of use.

In 1980, Great Britain’s National RadiologicalProtection Board (NRPB) issued interim safetyguidelines pertaining to NMR imaging of humans(table 1); revised guidelines are expected to bepublished soon. In 1982, the U.S. Food and DrugAdministration (FDA) also issued guidelines, butFDA’s guidelines were intended only to assist In-stitutional Review Boards in making decisionsregarding what constituted “significant risk. ”

Table 1 .—Suggested Guidelines for Safe Operationof NMR Imagers

National Bureau ofRadiological RadiologicalProtection Health

Board (FDA)(U. K.) 1980a (USA) 1982bc

Static magnet field(tesla) (whole orpartial bodyexposure) . . . . . . . . . . <2.5 T <2 T

Time-varying magneticfields (whole orpartial bodyexposure) . . . . . . . . . . <20 T/sec <3 T/see

for pulses for> 10 msec

Radiofrequency electro-magnetic fields . . . . . <15 MHz < 0.4 W/kgd

(<70 w averagedaverage power over whole

absorbed) body; <2W/kg over anyone gram of

tissue

aR. D. Saunders and J S. Orr, “Biologic Effects of NM R,” in Nuclear MafvreflcResonance (NMR) /rnagirrg, C L Partain, A. E James, F. D. Rollo, et al. (eds )(Philadelphia: W, B Saunders, 1983).

bW. E Gundaker, Guidelines for Evaluating Hecfromagrretic Risk for Tr/a/s of

C/lnica/ NMR Systems, FDA Bureau of Radiological Health, Dwision of Com-pliance, Feb. 25, 1982

‘ T he NRPB guidelines specify Upper limits that I17USt nOt be exceeded. ‘he ‘DA

guidelines, in contrast, are intended to serve as an aid for Institutional ReviewBoards making determinations of what constitutes “significant risk” under anInvestigational Device Exemption. Thus, the FDA does not prohibit use of NMR

dlmagers that operate outside these recommendationsAverage specific absorption rate rather than fixed frequency.

Some attempts are underway to maintain rec-ords of individu als who have und ergone NMR im-aging studies so that they can be contacted in the

future. The NRPB is collating this type of ex-posure record, and the NMR Commission of theAmerican College of Radiology has distributedpatient data forms that are to be completed eachtime a patient undergoes an NMR imaging study.It would be advisable to establish uniform guide-lines for collection of this type of data worldwide,at least for the near future. Who should be respon-sible for collecting and maintaining such data andwho should bear the expense need to be deter-mined. In addition, patient confidentiality issuesneed to be resolved.

8/14/2019 8420


Ch. 3–Clinical Applications of NMR 2 7

PROTON IMAGING: A CLINICAL OVERVIEW

Brain

Thus far, the widest and most efficacious useof clinical proton NMR imaging has been for thebrain and central nervous system. NMR’s prow-

ess in demonstrating the anatomy of the centralnervous system derives in large part from twofacts. First, although the proton density in thegray and white matter of the brain is approx-imately equal, gray matter contains approx-imately 15 percent more water than white mat-ter. This difference in the chemical environmentin which gray and white matter protons exist,which is reflected in differences in the T1 relaxa-tion time of gray and white matter, results in asignificant level of gray versus white matter con-trast in NMR images (30,52). This differentiationis much more striking with NMR imaging than

with X-ray computed tomography (CT) becauseX-ray CT’s linear attenuation coefficient for graymatter is only slightly different from that for whitematter (24). Second, because cortical bone has avery low NMR signal strength, it is possible toimage with NMR areas such as the posterior fossa,brain stem, and spinal cord that are not well seenby CT because the skull and vertebrae are imper-vious to the X-rays used in CT (136). In workdone in Aberdeen, Scotland (166) and at Ham-mersmith Hospital in England (30), for example,NMR has produced clear images of the cerebellumand brain stem, and has demonstrated brain stem

hemorrhages and cerebella tumors not seen withCT.

NMR also appears to be more sensitive thanX-ray CT to the demyelinating lesions that occurin multiple sclerosis (207).

In a report on cranial NMR scans performedon 14 0 patients with a broad spectrum of neuro-logic diseases and on 13 healthy volun teers, Bydd er,

et al., demonstrated NMR’s ability to detect tumors,

hemorrhages, infarctions, and ventricular size(30). It remains to be seen whether NMR will besuperior to X-ray CT in demonstrating these types

of pathologies in areas of the brain other than theposterior fossa and brain stem.

One final interesting central nervous system ap-plication of NMR is the study of brain develop-

ment. Researchers at Aberdeen and Hammer-smith, for example, have begun to explore thetime course of myelination in newborns (166,178).Given the presumed safety of NMR imaging, fur-ther studies can be anticipated.

Mediastinum, Hilum, and Lung

Although recent evaluations of NMR imagingof the thorax suggest that th e technique m ay proveuseful in assessing mediastinal, hilar (referring tothe anatomic area where the bronchus bloodvessels, nerves, and lymphatic enter or leave thelung), and lung abnormalities (71,166), the dataare still insufficient for comparing the efficacy of NMR w ith other imaging techniques in evaluatingthese structures. The possibility exists that varia-tions in the T1 relaxation time between blood andsoft tissues might permit clinically useful discrim-inations between blood, fat, and hilar tumors tobe made without use of intravenous contrastmaterials. Clinical trials are necessary, however,to assess the sensitivity and specificity of NMRimaging in these applications. Early work at Aber-deen suggests that some peripheral lung tumorsmay not be so well seen with NMR as with X-rayCT (166). If this proves to be the case, cliniciansmight use X-ray CT to evaluate lung parenchymaand NMR to evaluate the hilum. Finally, the pos-sibility of exploiting d ifferences in relaxation tim esbetween flowing, static, and clotted blood sug-gests that NMR might be used to diagnose pul-monary emboli. Respiratory gating (a techniquein which image acquisition is coordinated with thebreathing cycle) may imp rove the diagnostic quality

of all types of NMR studies of the lung.

Heart

Researchers have begun investigating the po-tential for using NMR to produce tomographicimages of the heart muscle, chambers, and val-vular structures (27,43,56,83,142). Although the

spatial resolution in images of the heart obtainedto date is crude compared to that of images of thebrain, advances in cardiac imaging are being mad eas techniques for gating of images to the electro-cardiogram are developed (114) and refinements

8/14/2019 8420


28 Ž Health Technology Case Study 27: Nuclear Magnetic R esonance imaging Technology: A Clinical, Industrial , and Policy Ana lys is

are made in the application of high-speed imag-ing techniques (such as the echo-planar tech-nique—a particular type of NMR imaging inwhich an image of a plane is obtained from anexcitation pulse) (124) to visualization of thebeating heart (27,142). Although NMR tomo-graphic imaging of the anatomical structures of

the heart must for now be considered experimen-tal, evidence is accumulating that NMR mayemerge as the modality of choice in the imagingof certain structures. (See, for example (176),relating to imaging the pericardium with NMRversus X-ray CT. )

Another potential cardiac application of pro-ton NMR imaging is discrimination between in-farcted, ischemic, and normal myocard ium(27,142). In early work performed at Massachu-setts General Hospital, researchers were able todistinguish ischemic from non-ischemic caninemyocardium with, but not without, use of aparam agnetic (a substan ce with a small but positive

magnetic susceptibility (magnetizability) that mayincrease the contrast between tissues and NMRimages) (4) contrast agent (76). Although man-ganese, the contrast agent used, may be too toxicto be used in humans, other less toxic paramag-netic agents might be developed (23,76). Morerecently, investigators at the Massachusetts Gen-eral Hospital have used anti-myosin (antibodiesdirected against myosin, a component of muscle)monoclinal antibodies labeled with manganeseand injected into the coronary arteries to visualize

acutely infarcted myocardium in dogs (110).Finally, researchers at the University of Califor-nia at San Francisco have been able to distinguishin vivo between norm al tissue and experimentallyinduced infarcts in the lower extremities of ratswith in 24 hours of infarction (89). Further researchneeds to be performed before the implications of these findings to the diagnosis of myocardialischemia and infarction in humans can be de-termined.

Breast

Ross, et al., reported on a series of NMR scansof 12 8 breasts examined in 65 patients (155). Scanswere obtained using a magnetic field strength of 0.045 T. The investigators found that althoughnormal breast tissue could often be distinguished

from abnormal breast tissue, there was some dif-ficulty in localizing abnormalities, as well as somedegree of overlap of T1 values obtained frombenign and malignant breast tissues. The lattermay limit NMR’s diagnostic efficacy in this ap-plication.

Ross, et al., also cited the prolonged time re-quired to do an NMR breast scan as a possiblelimitation of the technique (at a relatively lowmagnetic field strength). In Ross’ study, the aver-age total examination time per p atient was 1 hour.Pairs of transverse sections of the breasts wereproduced in 10 minutes; T1 measurements weremade in 6 minutes. The average total examina-tion time per patient exceeds the time required toproduce individual transverse sections because im-ages of several sections are used in the process of completing a patient examination.

Yousef, et al., employing a 0.3 T superconduct-

ing magnet, recently reported experience withNMR imaging of two patients with breast abnor-malities (58). Three-dimensional NMR images inone case and single-slice planar images (imagestaken through a single plane) in the other cor-related well with mammograms performed in thesame patients. The authors suggest that NMR im-aging may prove useful as a noninvasive probeof breast lesions, but acknowledge that it ispremature to predict the efficacy of NMR in theassessment of human breast disease (58).

Abdomen

Although the T1 values of the major abdominalorgans are sufficiently specific to make recogni-tion of those organs relatively easy with protonNMR imaging, there may not be sufficient differ-ence between the values in normal and diseasedstates to permit pathologic diagnoses to be madewith a high degree of accuracy (166). Results of small clinical series of NMR imaging of the liver(53,132,167) and pancreas (169,175) have beenreported, but further comparative studies need tobe performed before the clinical utility of NMRimaging of space-occupying lesions in the ab-

domen can be defined.

There is some indication that metabolic liverdisease may also be amenable to evaluation byhydrogen NMR imaging techniques. NMR can be

8/14/2019 8420


—-

Ch. 3–Clinical Applications of NMR • 29

used to measure liver iron overload in patientswith hemochromatosis (173), for example, and todemonstrate focal areas of inflammation in pa-tients with chronic active hepatitis (174).

Kidneys

Although several studies of proton NMR im-aging of the kidney have been reported (98,118,168,170), NMR investigations of the anatomy andfunction of the kidneys must be considered ex-perimental at this time. The possibility of distin-guishing between renal cortex and medulla withNMR (98) has been demonstrated, but furtherstud ies are needed before the clinical utility of suchNMR data can be defined. Similarly, NMR’s rolein the clinical evaluation of renal diseases suchas glomerulonephritis remains to be demonstrated(98). One potential disadvantage of using NMRto evaluate the kidney is its inability to detect

calcifications. In another area, Ackerman, et al.,have begun exploring the possibility of employ-ing NMR to evaluate transplanted kidneys (2).

Pelvis

Researchers at Aberdeen were able to correctlydifferentiate between benign prostatic hyperplasiaand prostatic carcinoma with proton NMR in 27of 30 men with symptoms of urethral obstruction.Two cases, however, were misdiagnosed as can-cer when, in fact, prostatitis was present (166),again illustrating that pathologic diagnoses made

with proton NMR may not be sufficiently specificto obviate the need for biopsy. Bladder tumorshave also been imaged (178). It is too early toassess the utility of using NMR in the staging of cervical, uterine, and ovarian cancer. The fact thatNMR imaging d oes not employ ionizing radiation,however, makes it attractive as a potential meansof imaging the pelvis, particularly in children andpregnant women. More recent developments inNMR imaging of the pelvis have been reviewedelsewhere (97).

Musculoskeletal SystemAlthough NMR does not visualize bone cortex,

it can be used to demonstrate bone marrow andalterations in bone architecture. ConsequentlyNMR has been used to demonstrate osteomyelitis

and tumor metastasis in vertebral bodies andpelvic bones (166). Because of its ability to pro-vide sagittal and coronal images, NMR may proveuseful in examination of the spinal canal, butagain studies comparing N MR to other techniquesneed to be performed before NMR’s role in thisregard can be defined. Early work has suggested

that both the spinal cord and nucleus pulposuscan be discerned in NMR images. Images of mus-cles, tendons, and ligaments demonstrating greatdetail have also been produced with NMR (131),suggesting that NMR imaging is likely to provevaluable in the evaluation of musculoskeletaldisorders.

Clinical Oncology

As has been discussed, further research needsto be performed before the role of proton NMRimaging in the diagnosis of cancer can be defined

(14). Additional techniques need to be developedbefore NMR can be used to distinguish with clin-ically useful accuracy between benign and malig-nant lesions. Because NMR does not employionizing radiation, it might be used frequently toclosely monitor the progress of pediatric and adultcancer patients being treated with radiation orchemotherapy. Finally, the exciting possibility ex-ists, but remains to be demonstrated, that NMRcould prove capable of detecting malignant ab-normalities at a stage earlier than is currently pos-sible. If that potential is realized, the additionalpossibility exists that clinicians will become moresuccessful in treating the malignancies that aredetected.

Blood Vessels and Flow

NMR imaging offers two different approachesto the evaluation of vascular disease, Althoughit is still too early to assess the sensitivity andspecificity of NMR techniques in the detection of vascular obstructions, a number of investigationshave suggested that proton NMR imaging can beused not only to demonstrate atheroscleroticvascular disease, but also to qualitatively assessflow in major vessels (3,41,81,108,109). Thesefindings are particularly interesting since NMRimaging does not require the potentially toxicradiopaque dyes and catheterization of traditionalangiography.

8/14/2019 8420


30 Ž Health Technology Case Study 27: Nuclear Magn etic Resonance Imaging Technology: A Clinical, Indu strial, and Policy Ana lys is

PUTTING PROTON NMR IMAGING INTO PERSPECTIVE

NMR imaging has captured the interest of themedical profession for a number of reasons. First,it eliminates the risk of X-radiation exposure thatis associated with use of devices such as X-ray CTscanners. Second, in addition to providing ex-

cellent spatial resolution (on the order of 1.0 milli-meter), the technique provides excellent contrastresolution without the need for injection of po-tentially toxic contrast agents. Third, because of the absence of NMR signal artifact from bone,physicians can visualize areas such as the posteriorfossa, brain stem, and spinal cord with NMR thatwere not well seen with other noninvasive imag-ing modalities. Finally, because NMR images aresensitive to fundamental physical and chemicalcharacteristics of cells, the technique offers thepossibility of detecting diseases at earlier stagesthan is currently possible and of permitting ac-curate pathologic tissue diagnoses to be madenoninvasively. These and other potential clinicalor practical benefits associated with NMR imag-ing are summarized in table 2.

NMR imaging is not without its disadvantages,however (see table 3). NMR imagers, for exam-ple, are expensive and their installation is costlyand difficult. In addition, the magnetic fields usedto generate NMR images with resistive and super-conducting systems are sensitive to radiofrequencyinterference and may damage computer tape orother objects in the surrounding environment.

Furthermore, it is currently thought not to be safeto perform NMR imaging on certain patients (e.g.,those with pacemakers or metal implants, suchas aneurysm clips). Finally, until more experiencewith the modality is obtained, NMR imaging mayrequire more p hysician time in p erformance of pa-tient examinations than X-ray CT or other imag-ing modalities.

Because the number of clinical applications of NMR imaging and the utility of each is constantlyevolving (table 2), NMR’s role in medicine has yetto be determined. As Kaufman, Tosteson, et al.,

have pointed out, many of the claims that NMRwill provide detailed chemical information spe-cific enough to be diagnostically useful are pre-mature (109). Although there is some plausibilityto the hundreds of applications that have been

Table 2.—Demonstrated and Potential Advantagesof NMR Imaging

A. Demonstrated benefits1.

2.

3.

4.

5.

Uses no ionizing radiation; currently free of known bio-logical hazards

Sensitive to differences in proton density and relaxationparameters, providing a basis for a high degree of softtissue contrastDoes not require injection of potentially toxic contrastagents (In the future paramagnetic contrast agents maybe developed for use with NMR. The risk associated withuse of such agents will have to be evaluated.)Low-intensity signal from bone, resulting in improvedvisualization of the posterior fossa, brain stem, and spinalcordCan produce three-dimensional as well as tomographic im-ages in virtually any imaging plane without reformatting(computer manipulation of the image)

B. Potential clinical or practical benefits 1. Increased sensitivity to physiochemical characteristics

of tissues may permit enhanced detection of disease (if,for example, alterations in tissue physiochemical char-

acter can be detected prior to alterations in organ func-tion or morphology)2. Increased sensitivity to physiochemical characteristics

of tissues may permit accurate pathologic tissue diag-noses to be made noninvasively (The development of non-toxic paramagnetic contrast agents would be particularlyuseful in this regard.)

3. Potential for analysis of flow of blood and other fluids, suchas cerebrospinal fluid

4. Absence of moving parts may decrease maintenance costsand increase the useful lifetime of NM R imagers

SOURCE E P Steinberg, Johns Hopkins Medical Instltutlons, Baltlmore, MD,1983

Table 3.—Current and Potential Disadvantagesof NMR Imaging

A. Current disadvantages 1. NMR imagers are expensive2. Difficult and expensive site preparation3. Exclusion of patients with pacemakers; metal implants,

such as aneurysm clips; or claustrophobia. (Patients whohave undergone brain surgery in whom there is uncertaintyregarding whether metal clips were implanted may needto be X-rayed before being imaged with NMR. Metal detec-tors are also being used.)

4. Insensitivity to dense bone and calcifications5. Relatively prolonged acquisition times and sensitivity to

motion result in degradation of images of moving anatomicstructures

B. Potential disadvantages 1. May require more physician time in performance of patient

examinations compared to other imaging modalities2. Examination times could prove to be longer than with X-

ray CT or ultrasound, resulting in fewer patients examinedand increased average cost. (Newer developments in im-aging techniques may change this)

SOURCE: E P Steinberg, Johns Hopkins Medical Institutions, Baltimore, MD,1983

8/14/2019 8420


Ch. 3–Clinical Applications of NMR 31

cited for NMR, the majority of such applicationsmust be considered potential rather than demon-strated. Although T1 and T2 relaxation times havebeen found to vary in different pathologic proc-esses, for example, researchers have so far beenunable to characterize tissue patholog y as specif-ically as had originally been expected using T 1 an d

T 2 values, The possibility exists, however, thatthrough the development of nontoxic paramag-netic contrast agents, much of the noninvasivediagnostic potential of NMR that had been hopedfor can be realized.

Whether the results that have been observed inresearch institutions to date will be easily dupli-cated by other institutions remains to be seen. TheNMR imaging that has been performed so far hastaken place in research centers that possessed oracquired considerable background and expertisein NMR prior to initiation of an NMR imagingprogram. Most, if not all, such centers have uti-

lized interdisciplinary teams consisting of physi-cians, physicists, engineers, and others to producethe images that have been p ublished so far. In ad -dition, the exquisite images that are helping to fuelexcitement about NMR imaging have often re-quired prolonged amounts of time that it may notbe practical to expect most hospitals to expend,Although future NMR production models arelikely to simplify image acquisition for the phy-sician, and although the time required to producehigh-quality images is likely to decrease, it shouldnot be assumed that images of the same qualityas those published or displayed to date will nec-essarily be produced immediately in hospitals thatare not able to spend equivalent time and efforton their production.

Furthermore, although in most research studiesperformed to date several physicians have inter-preted NMR scans independently and without ac-cess to comparative X-ray CT scans to guide them,in many, if not most, instances patients withknown pathologies were being imaged. This is,of course, not only reasonable, but also advisablein the early stages of evaluation of a new tech-nology. Nonetheless, NMR will not necessarily

be shown to have the same sensitivity and speci-

ficity when used to image patients with unknowntypes of pathology.3 (See (149), for example. )

Although researchers have reported the abilityto produce head and body images with NMR intimes comparable to those required for X-ray CTscanning (41), it is possible that the time requiredto perform NMR studies on patients with un-

known types of pathology will be longer than th atrequired by experts imaging patients with knownpathology (table 3). Given the variability in NMRimages depending on the pulse sequence em-ployed, researchers are finding that certain pulsesequences are better for demonstrating one typeof pathology, while different pulse sequences arebetter able to demonstrate other types of pathol-ogy. Optimal pulse sequences for different pathol-ogies are therefore being sought. It is likely thatin the future pulse-sequence algorithms that arebest suited to imaging specific medical problemssuch as abscesses, hemorrhages, and tumors willbe built into NMR software. Although this de-velopment implies improvement in NMR imagesof individual types of pathology, it also impliesthat patients with unknown types of pathologywho are referred for a “screening” NMR scan inthe workup of symptoms such as fever, weightloss, or pain will have to receive multiple scansusing multiple pulse sequences to exclude an ab-normality. Such use of multiple pulse sequencesmay increase the time required to perform NMRstudies.

The relationship between pulse sequence and

pathology suggests two additional issues. The firstis the issue of how many different pulse sequenceswill have to be employed before an examiningphysician is comfortable with concluding that apatient is normal. The second is that unless exam-ining physicians are thorough, existing pathologymay be overlooked. It is thus possible that a ten-sion will develop between the economic pressureto maintain scheduling of a reasonable numberof patients and the clinical requirement that pa-

30n the other hand, given the rapid improvements taking placein NMR imaging, current assessments may underestimate the ulti-

mate sensitivity and specificity of NMR in many applications.

8/14/2019 8420


32 • Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis

thologic abnormalities be excluded with a high de-gree of certainty. To the extent that the latterpredominates, fewer patients may be seen, pro-ducing a rise in average cost per NMR study. Tothe extent that the former predominates, the sen-sitivity and specificity of NMR may decrease.

Finally, since NMR imaging employs no ioniz-

ing radiation, it is possible that patients can safelyhave repeated NMR scans over time to monitortherapeutic progress or the natural history of dis-ease. To the extent that such monitoring is demon-strated to provide clinically useful information,the temptation to scan a single patient more fre-quently may increase. Such an effect would in-crease demands on NMR machines, which wouldresult in hospitals’ having to either ration time onNMR machines or purchase more NMR scanners.A second potential effect of “sequential scanning”is that in the absence of prospective payment (asis now the case for patients in many States), health

care bills might increase (see ch. 8). The occur-rence and magnitude of such increases depend onthe extent to which sequential NMR scanning de-creases or increases total patient managementcosts and improves patient care.

In conclusion, NMR imaging is a new modalitythought to be capable of providing not onlyanatomic information comparable or superior tothat obtainable with other im aging techniques, butalso chemical and metabolic information thatcould yield noninvasive insights into tissue his-tology and function. Were NMR to be considered

just another means of visually demonstrating

RELAXATION PARAMETERS

As has been mentioned previously, protonNMR images reflect not only the proton densityof the tissues being imaged, but also the protonrelaxation time characteristics of those tissues. Inaddition to exploiting these relaxation parametersto enhance the contrast of NMR images, research-ers over the past decade have been investigating

the extent to which quantitative measurements of T1 and T2 can be employed to make precise, yetnoninvasive pathologic assessments.

anatomy, therefore, its true potential would bedrastically underestimated. The lack of ionizingradiation or demonstrated hazard associated withNMR imaging, as well as the apparent lack of need to employ toxic contrast agents to obtainhigh-contrast images, would suggest an importantclinical role for NMR imaging in medicine evenif it proved to be no more efficacious than X-rayCT scanning. Considerably more research, suchas that described in table 4, will be required beforeNMR’s potential is determined and its most app ro-priate roles in clinical medicine are defined.

Table 4.—Likely Areas for Future NMR Research

1.

2.

3.

4.

5.

6.7.

8.

9.

10.

11.

Determination of clinical utility of NMR imaging comparedto other imaging modalitiesImprovements in magnet design (e.g., increased fielduniformity; increased field strength; techniques for con-serving liquid helium and nitrogen)

Determination of magnetic field strength that yields op-timum image quality for hydrogen, for other nuclei, or forcombinations of hydrogen and other nuclei)Identification of pulse sequences optimal for demon-strating different types of pathologiesImprovements in imaging software and techniques affect-ing image quality, resolution, and scan timeOptimization of radiofrequency coilsDevelopment of new and improved surface coilsImprovements in siting and shielding techniquesDevelopment of paramagnetic agents for assessing tissuepathophysiologyFurther development of whole-body spectroscopic tech-niques and applicationsImaging of nuclei other than hydrogen (e.g., sodium orphosphorus)

SOURCE E P Steinberg, Johns Hopkins Medical Institutions, Balt!more, MD,

1983

The first indication that relaxation times mighthave diagnostic significance emerged from theearly work of Raymond Damadian at the StateUniversity of New York, Brooklyn. In 1 9 7 1 ,Damadian demonstrated that the proton spin-lattice relaxation time (T1) in excised tumors thathad been experimentally induced in rat livers was

prolonged compared to the T1 values of normalrat liver tissue (42). The T1 relaxation times of twobenign fibroadenomas were likewise found to be

8/14/2019 8420


—— — .—— .—————Ch. 3–Clinical Applications of NMR 3 3

distinguishable from those of the malignant tissue(42). Similar results were observed by othersshortly thereafter (32,84,93,160). In 1974, Dama-dian extended this work to humans, finding theT1 values of 106 surgically excised human tumorsto be longer than the T1 values of the correspond-ing normal tissues (45,46). About the same time,

Weisman and Bennett discovered that they coulduse NMR to differentiate in vivo between the nor-mal tissue of a mouse’s tail and a malignantmelanoma transplanted into the tail (202).

Out of concern that experimental malignanttumors with slower growth rates might manifestT1 values within the range observed for normaltissues of nontumor bearing animals, Hollis, etal., studied intermediate and slow-growing m alig-nant tumors in rats (92). The authors found thatthe T1 values of these more slowly growing tu morswere again longer than those of normal rat liver,

but overlapped T1 values of other types of nor-mal tissues (92). In view of this finding, Hollis andhis co-workers voiced the concern that “. . . theoverlap of T1 values for malignant and normalspecimens raises the possibility of confusion be-tween primary tumors adjacent to sites of simi-lar or higher T1 value, or between normal tissueand metastatic tumors having a similar or lowerT 1 value” (92).

Similar observations were subsequently madeby others. Herfkens and co-workers at the Univer-sity of California at San Francisco, for example,

found that the T1 values of in vivo tumors in ratsestimated by NMR imaging techniques variedwidely, and, a lthough mean T 1 values of thetum ors were high, individual T1 values overlappedthe full range of normal and other types of ab-normal tissues (87). The same group also exploredthe possibility of using both T 1 and T2 values esti-

mated by NMR imaging techniques to distinguishbetween different types of pathologic abnormal-ities and found that contrast differentiation be-tween normal and abnorm al tissues was improved,albeit still imperfect (87) (see fig. 8) .

Figure 8.— Mean and Standard Deviations forT1 and T2 of Various Tissues in a Live Rat

?11 0 I 1 1 1 1 I 1 b 1 1

00

4 0

PUTTING T1 AND T2 RELAXATION TIMES INTO PERSPECTIVE

A number of conclusions can be drawn fromthe results of studies on the potential use of quan-titative T1 and T2 values for pathologic diagno-sis. First, within certain numerical ranges, relax-ation times may provide sufficient pathologicspecificity (to distinguish normal from abnormal,and one disease from another) to be clinicallyuseful in understanding disease processes. Second,in instances in which overlap exists between theT1 and T2 values of normal and abnormal tissues,it will not be possible to use relaxation valuesalone to make reliable pathologic diagnoses, de-spite the theoretical attractiveness of doing so. “ .

Third, even if “natural” T1 and T2 values do notprove to provide the pathologic specificity thathad been hoped, the y may be useful in under-standing disease processes. In addition, there isthe possibility that nontoxic paramagnetic con-trast agents (21,23,156) may modify relaxationproperties and enhance diagnostic specificity.Fourth, the possibility exists, yet remains to be

demonstrated, that detectable changes in T1

an dT2 values may precede the development of notice-able anatomical changes in ways that will be clin-ically useful. Studies of early changes in T1 an dT2 during oncogenesis, radiotherapy, and chemo-

8/14/2019 8420



therapy, for example, may provide immediateclinical dividends, as well as insights into theetiology of cancer and new modes of cancer ther-apy. Finally, considerable research remains to bedone in the exploration of which physical, chem-

IN VIVO SPECTROSCOPY

Although small concentrations of phosphorusare present in the body, researchers have recentlydetected NMR signals from phosphorus of suffi-cient intensity to produce images. Researchershave also been able to perform in vivo phospho-rous NMR spectroscopy in which the “chemicalshift” phenomenon is used to indicate the relativeconcentrations of compounds such as phospho-creatine, adenosine triphosphatase (ATP), and in-organic phosph ate in hum an tissues or organs (see

fig. 9). Such a noninvasive metabolic probe hastremendous potential applications in medicine.

It has been demonstrated, for example, thatlevels of inorganic phosphate rise, while the con-centration of phosphocreatine falls, in cells deprived

ical, and biological factors give rise to and influ-ence NMR relaxation times. Only through answersto these questions will it be possible to exploitrelaxation times’ full medical and scientific po-tential.

of oxygen (165). The possibility thus exists thatNMR spectroscopy could be employed not onlyto assess the utilization of ATP and the viabilityof oxygen-deprived brain or heart tissue (22), butalso to gain insight into the functional and meta-bolic status of brain, heart, or peripheral muscletissues in various physiologic and pathologicstates. It has been demonstrated, for example, thatthe pH of oxygen-deprived tissue can be calcu-lated from the positions of inorganic phosphorus

peaks observed with NMR spectroscopy (22,26)and that changes in muscular metabolism duringexercise can be detected by NMR (70,145), Rarediseases due to inborn errors of metabolism, suchas McArdle’s syndrome, can also be diagnosedusing in vivo NMR spectroscopy (146).

Figure 9.— NMR Phosphorus Spectrograma

IPhosphocreatine

Inorganic phosphate

Signalstrength

Sugarphosphates

ATP

1 1 1 1 I T - – 5 o 5 10 15 20 Chemical shift p.p.m

%raphic depiction of the Indwidual components of NMR signals from phoshorus-containing compounds arranged according to the frequency of electromagnetic radiation

SOURCE D I Hoult, “An Overview of NMR in Medicine, ” U S. Department of Health and Human Services, National Center for Health Care Technology, February 1981

8/14/2019 8420


.

Ch. 3–Clinical Applications of NMR 3 5—— — . . — —

PUTTING

Much addit

IN VIVO SPECTROSCOPY

onal research is required before an

INTO PERSPECTIVE

neither proton imaging nor in vivo specassessment can be made of the extent to whichin vivo NMR spectroscopy can be used to pro-vide diagnostically useful clinical informationregarding the metabolic and functional status of

normal and abnormal tissues. It should also berecognized that the technology required for invivo human NMR spectroscopy of organs suchas the heart or brain is considerably more sophis-ticated than that required to perform proton NMRimaging. Superconducting magnets operating athigh field strengths (on the order of 1.5 T orgreater) with extremely high levels of field uni-formity, for example, seem to be required to pro-vide adequate resolution of the peaks emanatingfrom various phosphorus-containing compounds.Thus most of the NMR imagers that are gener-ally being installed in hospitals today (i. e., ones

with field strengths less than 1.5 T) may not becapable of performing NMR spectroscopy, In ad-dition, even though it has been proven feasibleto perform spectroscopy on the 1.5 T imaging sys-tems currently being installed, it is possible that

roscopycan be performed optimally at a field strength-of 1.5 T.4 Hospitals desirous of performing spectros-copy and imaging, therefore, may need either toobtain more than one NMR machine or be con-

tent to tolerate potentially costly periods duringwhich field strengths are changed and the NMRmachine is therefore not operational.5 Consid-erably more research is also required before spe-cialized equipment, such as auxiliary surface coils,are designed that permit NMR spectroscopy of phosphorus and other nuclei to be performed op-timally. For the present, therefore, in vivo NMRspectroscopy should be considered an exciting andpromising area of research that is of questionablefeasibility for most hospitals.

4

High-quality images of both the head and body have recentlybeen produced at 1.5 T. It is too early to know whether images pro-duced at Is T will be better or faster than those produced at lowerfield strengths,

5In addition, alterations in software may need to be made inassociation with adjustments in the field strength, and such soft-ware changes may prove to be too time-consuming to be practical.

8/14/2019 8420


4.The NMR Imaging Device lndustry

8/14/2019 8420


4.

The NMR Imaging Device lndustry

INTRODUCTION

NMR imaging is fast emerging as the “growth”technology in the diagnostic imaging field. Thepossibility of unrivaled image quality, coupledwith the perceived demand for a less risky alter-native to X-ray computed tomography (CT) andradionuclide scanning, has led numerous firms toinvest heavily in NMR imaging research and de-velopment. Since 1976, at least 23 companiesworldwide have entered the NMR imaging mar-ketplace. The industry is both dynamic and in-tensely competitive, Sales of NMR imaging de-vices over the next 5 years have been estimatedby one source at $6.4 billion (60), an estimate that

some manufacturers contend is conservative. Thefuture of the industry, however, will depend notonly on the internal composition and behavior of the industry, but also on external economic andregulatory forces, including Federal and State pol-icies toward technology development and diffu-sion, and third-party d ecisions regarding paym ent.This chapter will focus on the NMR imaging indus-try, while succeeding chapters will address exter-

INDUSTRY STRUCTURE

The NMR imaging device industry may be ex-amined through several important structural fea-tures, including seller and buyer concentrations,barriers to entry, diversification of firms, and ac-quisition and merger activity. The findings pre-sented here reflect our own interpretations of datafrom multiple sources and do not represent theofficial views of the firms concerned. (A descrip-tion of the methods employed in our survey of manu facturers ap pears as app. B; detailed descrip-tions of manufacturers, products, and clinicalplacements of NMR imaging units appear in app.

c.)

Seller ConcentrationThe NMR imaging device industry may be

divided into th ree groups of manu facturers, based

nal economic and regulatory forces affecting thatindustry.

To understand the forces driving the NMR im-aging device industry, it is necessary first to focuson three elements that have become standard inthe analysis of American industry: structure, con-duct, and performance (31). Structure refers tothe composition and boundaries of an industry,i.e., the number and size distribution of its firmsand their ability to enter the marketplace. Con-duct pertains to the behavior of such firms oncethey gain entry to the marketplace, e.g., their pol-icies toward setting prices, differentiating theirproducts, and engaging in competition with oneanother. Performance relates to the results of firms’ behavior, i.e., how well the industry is ableto achieve recognized economic goals of efficiencyand profitability. Industry structure often influ-ences the nature of market conduct, which, in turn ,may affect the quality of industry performance.

on their stage of research and development (R&D)as of October 1983 (see table 5). Of the 20 firmsfor which we have information, seven had reachedan advanced stage in which they had conductedextensive preproduction technical testing and hadplaced numerous units in clinical sites outside theirfactories. Each of these advanced stage firms hadalso developed a commercial prototype system 1

that was available for placement. Two companies,Diasonics and Technicare, obtained FDA pre-market approval for the sale of their devices inMarch 1984, and a third, Picker International, inMay 1984.

‘See subsequent discussion of industry development for defini-tion of a commercial prototype system. By August 1984 Elscint ha dalso reached the advanced stage. See app. C.

39

8/14/2019 8420



Table 5.—The NMR Imaging Device Industry,October 1983

Companies in advanced stage of development Engineering model(s) complete; mu/tip/e clinicalplacements outside factory; ongoing clinical studiessince 1982 or earlier; commercial prototype system(s)available for placement:

Bruker Instruments

Diasonics Inc.a

Fonar Corp.Philips Medical SystemsPicker Internationala

Siemens Medical SystemsTechnicare Corp.a

Elscint Ltd.b

Companies in intermediate stage of development Engineering model(s) complete; limited clinicalplacements outside factory; generally limited clinicalstudy thus far; commercial prototype system(s) generallynot yet available for placement:

General Electric Co.M&D Technology Ltd.c

Toshiba Corp.

Companies in early stage of development: d

Engineering model(s) under development; no clinicalplacements outside factory; commercial prototypesystem(s) still to be defined:

ADAC LaboratoriesCGR Medical Corp.e

Fischer Imaging Corp.f

Hitachi Ltd.JEOL USANalorac Cryogenicsg

OMR Technologyh

Sanyo ElectricShimadzu Corp.

aFDA granted premarket approval in SPrin9 19~

bln the advanced stage by August 1964 (see aPP. C)cHad extensive clinical experience prior to formation of company.dT wo other firms that have announced plans to develop NMR ima9in9 sYstems

are Ansoldo SPA and Instrumentarium Oyengineering model complete, but clinical placement iS not expected Until 1964.fHad been developing its own NMR imaging systems until acquired by DiasonicsInc in May 1983.gAs Indicated in ‘pp. c, Nalorac Cryogenics had two cl!n!cal placements by

August 1964 However, these are small bore, high field strength systems thatare currently being used for research purposes only.

h Ac q u l r e d by )(onics Inc. In late 1963

SOURCES Interviews with manufacturers; American Hospital Assoclatlon, 1963(6); Boteler, 1963 (20); and “Imaging Equipment Sales Close In On $4Billion Mark, ” f2/ag Irrrag 5(11)”55431, November 1963

By October 1983 four firms had progressed toan intermediate stage of R&D in which engineer-ing and experimental mod els had been completed,but commercial prototype systems had either notyet been developed or not yet been installed inclinical sites. The extent of clinical study also var-

ied widely among these manufacturers. M&DTechnology Ltd. had the benefit of clinical experi-ence acquired by its founders at the Universityof Aberdeen (Scotland) prior to its incorporation,

but it had yet to reach the advanced productionprototype stage.

Nine other manufacturers could be character-ized as engaged in early R&D work in October1983. Of these, only one (CGR Medical Corp. )had completed an experimental prototype on

which clinical testing was expected to begin in1984, One firm, Fischer Imaging Corp., had beenacquired (May 1983) by a manufacturer in the ad-vanced R&D group, Diasonics Inc. Completedetails on the future organizational structure of the two companies and their respective NMR im-aging programs are not available at this time.

Industry Profile. The multinational characterof the NMR imaging device industry is reflectedin the 19 firms listed in table 6.2 In October 1983,U.S. companies accounted for 37 percent of thetotal, while Japanese corporations comprisedanother 26 percent. Five other nations had entriesin the world market: West Germany and GreatBritain with two companies each; and France,Israel, and the Netherlands with one apiece.3

Thirteen of the manufacturers (68 percent) arepublic corporations, some of which are giants inother fields (General Electric, Hitachi, Philips,Sanyo Electric, Siemens, Toshiba). Of the six pri-vately held firms, two are owned by small groupsof investors (Nalorac Cryogenics, OMR Technol-ogy 4), while three others are subsidiaries of ma-

jor corporations (Picker International, BrukerInstruments, CGR Medical Corp.). M&D Tech-

nology Ltd. is a unique entity financed by a com-bination of private individuals and public trustsin Aberdeen, Scotland.

In terms of organizational structure, 11 NMRimaging device manufacturers (58 percent) are in-dependent firms, seven (37 percent) are whollyowned subsidiaries, and one is the Medical Sys-tems Division of a major public corporation (Gen-eral Electric). Fifteen NMR manufacturers (79 per-cent) have multiple product lines. Some of these

‘Because Fischer has been acquired by Diasonics, we have excludedit from table 6.

‘Two other firms that have announced plans to develop NMRimaging systems are Ansoldo SPA of Genoa, Italy, and Instrumen-tarium Oy of Helsinki, Finland.

40MR Technology has recently been acquired by Xonics Inc. , apublicly owned multiproduct firm in the United States.

8/14/2019 8420


8/14/2019 8420



Table 7.—The NMR Imaging Device Industry: Market Share as Reflectedin Clinical Placementsa

Current placements(as of August 1984) Current market sharec

Company b Worldwide U.S. only Worldwide U.S. only

Technicare Corp. . . . . . . . . . . . . . . . . . . 44 36 30 ”/ 0 39 ”/ 0

Picker International . . . . . . . . . . . . . . . . 28 12 19 13Diasonics Inc. . . . . . . . . . . . . . . . . . . . . . 23 19 16 20

Siemens Medical Systems . . . . . . . . . . 19 10 13 11

Fonar Corp. . . . . . . . . . . . . . . . . . . . . . . . 9 6 6 6Philips Medical Systems. . . . . . . . . . . . 5 1 3 1

Bruker Instruments . . . . . . . . . . . . . . . . 4 2 3 2Elscint Ltd. . . . . . . . . . . . . . . . . . . . . . . . 4 2 3 2General Electric Co. . . . . . . . . . . . . . . . 4 3 3 3M&D Technology Ltd.d . . . . . . . . . . . . . 2 0 1 0Nalorac Cryogenics . . . . . . . . . . . . . . . . 2 2 1 2Toshiba Corp.d . . . . . . . . . . . . . . . . . . . . 1 0 1 0

Totals . . . . . . . . . . . . . . . . . . . . . . . . . . 145 93 100 % 100%aNMR imaging systems Flaced in clinical sites outside facto~, human Systems onlY (whole body and head).bin descending order of worldwide market share, aS Of August 19~cExpressed as ~rcentage of total current placements Detail may not sum to 100 Percent because of roundingdAs of October 1983.

SOURCES Interviews with manufacturers, Boteler, 1983 (20), American Hospital Association, 1983 (6); and “Imaging Equip-ment Sales Close In On $4 Billion Mark, ” LVag . /rnag. 5(1 1):55-61, November 1983

percent of those in the United States (36 of 93).Picker International was second, with 28 unitsworldwide (19 percent) and 12 in the United States(13 percent). Diasonics Inc. had placed 23 unitsworldwide (16 percent), with 19 of those in theUnited States (20 percent of the U.S. market).Siemens Medical Systems had slightly fewer place-ments, with 19 worldwide (13 percent) and 10 inthe United States (11 percent).

NMR Imaging Systems. An important deter-minant of industry growth and seller concentra-tion will be the product features offered in the

NMR imaging systems. Manufacturers are in-vesting great energy in product differentiationstrategies designed to segment the market forNMR imaging devices (see discussion of nonpricecompetition policies of firms in the industry con-duct section of this chapter). Considerable con-troversy exists over the optimal design and con-figuration of NMR imaging units (20). Muchof the debate centers on magnet design (6),with various manufacturers pursuing differentstrategies.

At present, M&D Technology Ltd. is the only

company that appears committed to resistivemagnet design operating at relatively low fieldstrengths (see table 8). At least five firms (Dia-sonics, Philips, Siemens, General Electric, and

Nalorac Cryogenics) are strongly committed tosuperconducting magnet technology only. Philipsand Siemens now offer both 5 and 15 kilogausssystems, whereas General Electric plans to mar-ket a 15 kilogauss model in 1984. Nalorac Cryo-genics is developing three superconducting sys-tems intended largely for research applications,with magnet strengths ranging from 10 to 40kilogauss (see app. C).

Superconducting magnet systems are now of-fered by at least four other manufacturers, butthree of them also offer resistive systems (Picker

International, Technicare Corp., and Bruker In-struments) and one is experimenting actively withpermanent magnets (Elscint Ltd.). Fonar Corp. isthe only manufacturer that now bases its systemdesign on permanent magnet technology, includ-ing a 3 kilogauss mobile, whole body unit. ADACLaboratories is also developing a permanentmagnet NMR imager and expects to have a pro-duction model ready in late 1985.

Buyer Concentration

Unlike the high seller concentration in the NMRimaging device industry, the number and diver-sity of potential buyers in the market is extraor-dinarily large, covering research laboratories and

8/14/2019 8420


Ch 4–The NM R Imaging Dev ice Indus try 43—— — . . .— —

Table 8.—Status of NMR Imaging Systemsa

—

NMR imaging system —

ClinicalMagnet Field strength Bore Year first patients studied

Companyb type (kilogauss) size available to datec

Bruker Instruments R 1 .3d B 1979 100s 47 A 1979s 19d A or H 1982R 2.4 B 1984

CGR Medical Corp. R 1.5 B 1982 0s 3.5 B 1983s 5 B 1983

Diasonics Inc. s 5 d’e ‘ B 1981 NA

Elscint Ltd. s 5 B 1982 N A

Fonar Corp. P 0.4 B 1980 2,200P 3d

B 1983P 3d BM 1983

General Electric Co. R 1 .2f

B 1982 600R 1.5f

B 1983s 15d B 1984

M&D Technology Ltd. R 0.4 B 197? 1,200R 0.8 B 1982

Philips Medical Systems R 1.5 B 1982 300

s 30 A 1982s 15d B 1983s 5 d B 1983

Picker International R 1 .5d

B 1978s 3 B 1981 NAs 5d B 1983

Siemens Medical Systems R 1,2 B 1980R 2 B 1981 800s 5 d B 1983s 15d

B 1983

Technicare Corp. s 15d A 1980 4,750R 1 .5d

H 1981s 3 f B 1982s 5 d B 1983s 6 B 1983s 15 B 1983

Toshiba Corp. R 1.5 B NA NANA = Not availableKEY Magnet type P = Permanent

R = ResistiveS =Superconductlng

Bore size A = AnimalB = Whole bodyBM = Whole body (mobile)H = Head

aAs of August 1984bln alphabetical orderCAS of October 1983dprobable commercial prototype System(S)esystem operating at 35

‘No longer available

SOURCES Interviews w!th manufacturers, Boteler, 1983 (20), and American Hospital Association, 1983 (6)

various types of clinical facilities. Likely buyers in the United States (5) will purchase NMR im-

in the clinical segment of the market include hos- agers by 1990. The prime buyers will be thepitals, private radiology groups, and health main- leading teaching hospitals and medical centers,tenance organizations (HMOs). Many manufac- followed by large urban and moderate-sized com-turers are optimistic that more than half of the munity hospitals with bed capacities of at least5,900 non-Federal, short-term general hospitals 200. In the United States alone there are over

8/14/2019 8420


44 Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis

1,700 hospitals meeting this description (5) anda large number in Canada and Western Europe.NMR imaging manufacturers also expect to makein-roads into other segments of the U.S. hospitalindustry, including the smaller independent com-munity hospitals (100 to 199 beds); Federal Gov-ernment hospitals in the Veterans Adm inistration,

Department of Defense, and Public Health Serv-ice systems (numbering around 350 facilities); andlong-term and specialty hospitals (roughly 1,000).Hospital chains, particularly investor-owned cor-porations, are expected to be prime purchasers of NMR imagers (see the discussion of hospital strat-egies in ch. .5).

Several hundred NMR imaging units are ex-pected to be sold worldwide to private radiologygroups and to physicians’ offices outside hospi-tals. The approximately 23 6 HMOs and prepaidhealth plans in the United States (193) are another

potential source of buyers, with some likely topurchase multiple units for outpatient as well asinpatient settings.

Finally, the medical research community isviewed as an important market segment. At leasttwo manufacturers (Nalorac Cryogenics and JEOL)are firmly committed to developing NMR imag-ing systems specifically intended for researchapplications. Both firms are investing in super-conducting magnet systems that will operate atrelatively high field strengths and be capable of performing phosphorus spectroscopy as well asproton NMR imaging.

Barriers to Entry

The ability of relatively small firms to enter theNMR imaging device industry d epends on severalkey factors: their ability to attract adequatecapitalization and technical/ scientific talent forR&D, the development of strong university tiesfor collaborative research, and the ability to mar-ket products once they have been developed. Atpresent, three small, single-product firms comprise16 percent of the total number of firms in the in-

dustry (3 of 19 firms). Among them, one (FonarCorp. ) has attained advanced R&D status, anda second (M&D Technology Ltd. ) stands on thethreshold of commercial production. In order tound erstand th e importance of these achievements,

it is necessary first to examine the chronolog-ical development of the NMR imaging device in-dustry.

Industry Development. The birth of the NMRimaging device industry can be traced to 1976when EMI began work on building an NMR im-aging machine. In 1977, two other companies(Bruker Instruments and Philips Medical Systems)embarked on parallel courses of NMR imagingR&D (see fig. 10). Between 1978 and 1980, fiveadditional firms entered the industry. Fonar,draw ing on several years of research by RaymondDamadian, was the first American company (andthe only small single-product firm) to make a firmfinancial commitment to NMR imaging R&D dur-ing this period. Since 1980, the industry has ex-perienced rapid growth, with four new entries in1981 and another four in 1982. Data for the Jap-anese NMR imaging device industry are incom-plete, but it is believed that several Japanese com-panies also entered the market during this time.

The pattern of NMR imaging development, i.e.,the sequence of steps through which a manufac-turer must pass in order to reach full productioncapability, generally consists of four major steps(see

1.

2.

3.

fig. 10):

Corporate decision— the manufacturer makes

a corporate decision to invest in R&D activ-ities and marshals its resources (capital, staff,facilities, materials) to assemble a programdevelopment effort whose first objective is

to produce an experimental prototype orengineering model. Experimental prototype—upon attaining thisgoal, the manufacturer can begin in-housetesting that proceeds through several stagesusing “phantom” (inanimate) objects at first,and then laboratory animals and humans asimaging subjects. The knowledge and experi-ence gained in the process facilitate the re-finement of both system hardware and im-aging techniques.Clinical placement outside the companyplant—manufacturers differ somewhat in

their approach to clinical testing. Some pre-fer initial testing with humans on experi-mental in-house systems before seeking out-side clinical placements of investigationalunits. Others choose to perform all clinical

8/14/2019 8420


8/14/2019 8420


46 Ž Health Technology Case Study 27: Nuclear Magnetic Resonanc e Imaging Technology: A Clinical, industrial, and Policy Analysis —

In the case of NMR imaging R&D, the timespan required for completion of all four steps ap-pears to have decreased over the years (see fig.10). The early firms entering the market (e.g.,EMI, Bruker Instruments, Philips Medical Sys-tems, Siemens Medical Systems, and Fonar Corp. )each took 2 to 4 years to produce their first ex-

perimental systems. By contrast, firms enteringthe market in the last 2 to 3 years either have at-tained, or plan to attain, this goal in significantlyless time. Overall, the actual or projected timeframes of these late participants in NMR imag-ing are very much compressed, owing to thestrong pressures of competition and to the knowl-edge about NMR imaging design conferred uponthem by the pioneering efforts of their prede-cessors.

The pathways for entry 6 into the market havevaried by manufacturer (see fig. 10). Essentially,

four different routes have been followed, somesimultaneously:

Government-supported R&D—EMI enteredthe NMR market in 1976 with grant supportfrom the Department of Health and SocialSecurity (DHSS) in Great Britain. BritishGovernment support of university-basedR&D at Nottingham and Aberdeen duringth e 1970s also later benefited Picker Interna-tional and M&D Technology Ltd., respec-tively, when they decided to enter the NMRimaging market. Three firms (Bruker, Philips,and Siemens) have received grants from theWest German Government, but only aftereach had initiated NMR program develop-ment with company resources.University-based R&D—all four small, sin-gle-product firms emerged as a direct resultof university-based R&D at the following in-stitutions: State University of New York(Fonar Corp.); University of Aberdeen (M&DTechnology Ltd.); University of Nottingham( N a l o r a c Cryogenics); a n d U n i v e r s i t y o f California, Los Angeles (OMR Technology).In the case of Nalorac Cryogenics, the com-

pany’s founder, James Carolan, had worked

‘The paths for entry do not necessarily reflect subsequent R&Dstrategy and are not mutually exclusive.

at Nottingham and later at Bruker beforeestablishing his own firm.

Acquired technology—two firms have suc-cessfully employed this strategy to acceler-ate their market entry and their progresstoward advanced R&D. Picker Internationalin 1981 purchased all NMR imaging technol-

ogy that had been developed by EMI of England since 1976. That sam e year, DiasonicsInc. purchased the rights to all patentableNMR technology developed by Pfizer underan agreement with the University of Califor-nia, San Francisco. A third firm, Fischer Im-aging Corp., also sought to purchase NMRimaging technolog y from other manufac-turers, but eventually was acquired by Dia-sonics Inc. in May 1983 .

Internally based R&D—the remaining firmsin the industry have generally relied on in-ternal R&D operations to develop their own

NMR imaging technology. General Electric,Philips, Siemens and Technicare are examplesof large companies marshaling their consid-erable R&D resources for directed NMR pro-gram development. Elscint and ADAC Lab-oratories have also committed themselves tointernal R&D without benefit of governmentfunding or of “off-the-shelf” technology. Atleast three firms (Bruker, Philips, and JEOL)have been able to draw directly on their cor-porate experience in manufacturing researchlaboratory NMR spectrometers.

The major elements affecting a company’s abil-ity to enter the marketplace generally have in-cluded the availability of capital, staff expertise,corporate experience, and collaborative links withmajor university research groups.

Capital Requirements for Market Entry. Inter-views with manufacturers suggest that capital re-quirements for R&D have not been unduly ex-cessive for those who have entered the field.Industry sources estimate that a new firm, or afirm lacking prior experience in NMR spectros-copy, requires between $4 million and $15 mil-

lion for initial capitalization of R&D. This esti-mate does not include the capital required forexpanding production capacity or for vertical in-tegration of NMR imaging-related products, e.g.,the capacity to build one’s own magnets.

8/14/2019 8420


Ch. 4– The NMR Imaging Device industry 4 7

Table 9 shows the R&D expenditures incurredup to October 1983 by 12 companies involved inNMR program development. Expenditure levelsare reflections, in part, of a company’s stage of R&D effort. For instance, the four firms report-ing expenditures of $1 million to $5 million haveonly recently entered the market and are engaged

in early R&D. By contrast, the six companiesreporting expenditures in excess of $10 million areinvolved in inter-mediate (two firms) or advanced

(four) stages of R&D. There are exceptions to thispattern: the two firms with expenditures in the$6 million to $10 million range have both attainedadvanced R&D status.

Staff Requirements for Market Entry. The im-portance of staff expertise to R&D activities inNMR imaging cannot be overstated. As men-tioned earlier, Fonar Corp. and M&D Technol-ogy Ltd. were formed around innovative scien-

tists an d their specific techniques or m ethodologies.The technical complexity of NMR imaging dic-tates that manufacturers assemble R&D teamswith expertise in such fields as physics, chemistry,engineering, and compu ter science. Specific knowl-edge of NMR spectroscopy is valuable. Severalfirms, including Technicare, Picker International,and General Electric, have aggressively recruitedindividuals who conducted some of the earliestNMR research in England in order to augmenttheir in-house R&D staff talent.

For small firms, staff recruitment and develop-

ment may be a constraint. A “critical mass” of at least five to six scientists appears to be neces-sary before a company can actively initiate R&D.Continued staff growth, as R&D activities mature,is vital to company survival. In the face of tightresource constraints, some firms have had to aug-

Table 9.—Research and Development ExpendituresAmong Firms in the NMR Imaging Device Industry

R&D expenditures Number ofto datea f irms report ing Percentage

<$5 million . . . . . . . . . . . 4 33%$6-10 million . . . . . . . . . . . 2 17

$11-20 million . . . . . . . . . .1

8>$20 million . . . . . . . . . . 5 42

Total . . . . . . . . . . . . . . 12 100%0aAs of October 1963

SOURCE Interviews with manufacturers

ment in-house expertise with outside consultants.Equally important for sustained program devel-opment is the need to establish a top-notch mar-keting and sales force. The larger, well-establishedfirms with existing sales networks hold a criticalcompetitive edge over smaller companies that lacksuch organization and expertise. Marketing and

sales acumen may prove to be a decisive factorin the competitive marketplace that is rapidly de-veloping.

Collaborative Research With Universities and Major Medical Centers. University or major med-ical center research ties are considered essentialin the industry, Every manufacturer engaged ineither intermediate or advanced stage R&D in Oc-tober 1983 had a close collaborative relationshipwith one or more universities or major medicalcenters (see table 10). The lack of university re-search links early in R&D is not necessarily a bar-

rier to market entry for small firms, but futurecompany survival—particularly in the clinicalphases of product testing—may depend on thenature and quality of such agreements. Largefirms also recognize the importance of collab-orative research. Technicare, for example, has astated policy of not placing units in clinical sitesunless close working arrangements with the in-stitutions can be established. Acquisition of clin-ical data is extremely important to the manufac-turer in preparing for FDA premarket approvaland for coverage decisions by third-party payersin the health care system.

Patents. Among the Federal policies that havebeen developed to promote innovative researchand product development are those related to theprotection of discoveries by patents. Although athorough discussion of patent law and its com-mercial and societal ramifications is beyond thescope of this report (and the expertise of itsauthors), a few comments can be made regardingthe NMR-related patents and their impacts.

A number of components of NMR imaging andspectroscopy systems would seem patentable.Among these are designs of: 1) the magnet used

to produce the static magnetic field; 2) the radio-frequency coils used to emit and receive radiofre-quency waves; 3) the gradient coils used to per-mit spatial encoding; and 4) the software techniques

8/14/2019 8420


48 Ž Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, industrial, and Policy Analy sis

Table 10.—Manufacturers’ Collaborative Arrangements With Universities/Medical Centersa

Company:b Company: b

university/medical center university/medical center

ADAC Laboratories: Negotiated, yet to be announced.

Bruker Instruments: Baylor College of Medicine, Houston, TXYale University, New Haven, CT

CGR Medical Corp.: None in USA (number in Europe not available)Diasonics Inc.:

University of California, San Francisco (UCSF)University of Texas, DallasUniversity of Michigan, Ann Arbor

Elscint Ltd.: Hebrew University, Jerusalem, IsraelWeitzman Institute, Rehovoth, Israel

Fonar Corp.: University of California, Los Angeles (UCLA)

General Electric Co.: Medical College of Wisconsin, MilwaukeeUniversity of Pennsylvania, PhiladelphiaYale University, New Haven, CTDuke University, Durham, NC

JEOL USA:

None at presentM&D Technology Ltd.: University of Aberdeen, Scotland, U.K.

Nalorac Cryogenics: None at present

Philips Medical Systems: Neurological Institute, Columbia-Presbyterian Hospital,

New YorkUniversity of Leiden, The Netherlands

Picker International: University of Nottingham, Nottingham, U.K.Royal Postgraduate Medical School and Hammersmith

Hospital, London, U.K.Mount Sinai Hospital, Cleveland, OH

Mayo Clinic, Rochester, MNBowman Gray Medical School, Winston-Salem, NCUniversity of British Columbia, Vancouver, CanadaCity of Faith Medical and Research Center, Tulsa, OKNational Institutes of Health, Bethesda MD

University of lowa, Iowa CityQueens Square Hospital, London, U.K.Siemens Medical Systems:

Washington University, St. Louis, MOUniversity of Hanover Medical Center, Hanover, West

GermanyRadiological Institute, Frankfurt, West GermanyRadiological Institute, Munich, West GermanyMount Sinai Medical Center, Miami, FLAllegheny General Hospital, Pittsburgh, PA

Technicare Corp.: Massachusetts General Hospital, Boston, MAUniversity Hospital, Cleveland, OHCleveland Clinic Foundation, Cleveland, OHUniversity of Kentucky, LexingtonIndiana University, IndianapolisHershey Medical Center, Hershey, PA

Millard Fillmore Hospital, Buffalo, NYSt. Joseph’s Hospital, London, Ontario, CanadaJohns Hopkins University, Baltimore, MDOntario Cancer Institute, Toronto, CanadaCharlotte Memorial Hospital, Charlotte, NCNew York Hospital, New YorkVanderbilt University, Nashville, TNDefalque Clinic, Charleroi, BelgiumUniversity of Florida, GainesvilleBaylor University Medical Center, Dallas, TXRush Presbyterian-St. Luke’s Medical Center, Chicago

Toshiba Corp.: Toshiba General Hospital, University of Tokyo,

Japan—

a~of AuguSt 1 g84, except M& D Technology Ltd, and Toshiba Corp. are as of October 1983 I nformatlon not avallabie for H Itachi Ltd , OM R Technology (fIOW xOfIICS

Inc.), Sanyo Electric, and Shimadzu Corpbln alphabetical order

SOURCES Interwews with manufacturers and American Hospital Association, 1983 (6)

used for spatial encoding, data gathering and im-age reconstruction.

Neither Lauterbur nor SUNY-Stony Brookpatented Lauterbur’s original NMR imaging tech-nique or the apparatus (115).7 On March 17, 1972,however, Raymond Damadian filed a patent ap-plication for an “Apparatus and Method for De-tecting Cancer in Tissue” and received a patentin February 1974. Damadian has apparently filedan additional patent application in the United

States, as well as patent applications in foreign

‘Apparently, the SUNY-Stony Brook was advised by an outsideconsultant not to proceed with a patent application.

countries (65). A number of NMR-related patentsare apparently also held in a patent portfolio bythe British Technology Group, formerly the Na-tional Research Development Corp. (120). Infor-mation regarding the types of license agreements,if any, related to such patents is not available.No attempt has been made to gather comprehen-sive information for this report regarding thenumber and types of patents held by NMR man-ufacturers.

A primary concern regarding patents is thatthey might create undesirable barriers to the en-try of potential NMR manufacturing competitorsinto the marketplace. However, the existence of

8/14/2019 8420


—

— —

at least 19 manufacturers of NMR imaging sys-tems suggests that patents have not created sucha barrier; the manufacturers we interviewed con-curred with this view. Whether patentable dis-coveries will emerge, prohibitively expensivecross-licensing agreements will be devised, orpending lawsuits’ will be settled in such a way

as to change this situation is difficult to predict.A second policy concern regarding patents is

that manufacturers might: 1) stifle the promptdissemination of scientific discoveries made bythose university-based researchers whom theysupport in order to provide time for filing patentsor 2) redirect the focus of university-based re-search away from “basic science” and toward thedevelopment of p atentable d evices and techniques.The existence of a large number of industry-university collaborative NMR research relation-ships (see table 10) suggests that universities have

not found such research agreements prohibitivelyrestrictive. 9

This empirical inference was confirmed bydiscussions with a number of investigators whoseNMR research is being supported in part by in-dustry (1,184). Others, however, voiced concernthat the scope of their research was more con-strained when sponsored by industry than byNIH. Such a concern would seem to be more anargument for Federal research funds than an in-dictment of patents.

Finally, it is difficult to determine how bene-

ficial the protection afforded by patents has beento the commercial development of NMR in this

“On Sept. 20, 1982, Fonar Corp. and Dr. Raymond Damadianfiled suit in the U.S. District Court in Massachusetts against Johnson& Johnson and its subsidiary, Technicare Corp. (65). The suit chargesJohnson & Johnson and Technicare with willfully infringing onDamadian’s patent for using Nuclear Magnetic Resonance in detect-in g and diagnosing human disease and with unfair competition andinterference in Fonar’s ability to successfully market the apparatuscovered by the patent. The defendants have denied the allegationsand requested a judgment declaring the patent invalid. The matteris in the discovery stage (65).

‘Nlany research agreemen t~ between industry and academia enableuniversities to benefit financially from the discoveries made by theirfaculty. Diasonics Inc., for example, holds the exclusive right to ob-

tain an exclusive license to all patentable NMR technology discoveredpursuant to the research project it supports at the University of

California Under the terms of the license, the university is entitledto a royalty of 0.56 percent of the selling price of any NMR systemsold by Diasonics that includes technology patented by the univer-sity (51 ).

Ch. 4–The NM R Imaging Device Indust ry 4 9

country. It is possible, for example, that manymanufacturers have relied more on maintainingdiscoveries as “trade-secrets,” rather than reveal-

ing confidential information in patent applica-tions. Of interest in this regard, however, is thebelief voiced by Lauterbur that acquisition of apatent by either SUNY-Stony Brook or himself

would have accelerated commercial developmentof NMR imaging devices in the United States byvirtue of providing a means of protecting a man-ufacturer’s competitive advantage (116).

Regulatory Policies. We surveyed NMR man-ufacturers about their perceptions of the impactof various regulatory policies on the placementof their products in clinical sites.10

Of the various Federal and State policies affect-ing technological development and diffusion, nonewas perceived by manufacturers to be a seriousconstraint on NMR development. The FDA pre-

market approval (PMA) process is generally re-garded as a time-consuming “hurdle” that is notoverly obstructive. None of the firms interviewedfelt that the PMA process had influenced eitherthe pace of R&D activities or the placement of investigational units at clinical sites. (For a morecomplete discussion of issues pertaining to theFDA and its PMA process, see ch. 7.)

By contrast, third-party payment policies, andto a lesser degree, State certificate-of-need pro-grams, appear to cause major concern amongmanufacturers as potential barriers to NMR dif-

fusion. Coverage policy decisions of the FederalMedicare program, State Medicaid agencies, localBlue Cross/ Blue Shield plans, and commercial in-surance companies are considered critical to thefuture marketing of NMR imaging devices. Un-favorable coverage decisions—or even moderatedelays in decision making—by the Health CareFinancing Adm inistration (HCFA) and other third-party payers could pose serious financial prob-lems for those manufacturers in advanced stagesof R&D. Coverage denials for NMR imagingcould conceivably destroy the hospital segmentof the market and militate strongly against entry

of new firms into the industry. State prospective

—.

IL’ The views expressed by the manufacturers should not be con-sidered to represent the views of either the authors or OTA.

8/14/2019 8420


50 Ž Health Technology Case Study 27 : Nuclear Magnet ic Resonance Imaging Technology: A Clinical, Industrial, a n d P o l i c y A n a l y s i s— .—

payment programs are viewed by manufacturerswith considerably less trepidation since, undermany such programs (e.g., Maryland, New York,Massachusetts), hospitals have retained widediscretion in their capital-equipment purchases.

State certificate-of-need (CON) programs, on

the other h and , are perceived as potentially trouble-some constraints that might delay—or even limit—the placement of NMR imaging devices in speci-fied geographic areas. Some manufacturers feelthat CON policies could prove unusually restric-tive in some areas of the country despite favorablecoverage decisions by third-party payers. Shouldthis occur, NMR diffusion could slow noticeablyin the United States, sending discouraging signalsto firms contemplating market entry.

Diversification of Firms

The firms that constitute the NMR imaging de-vice industry display considerable diversity intheir product lines and operations. Twelve com-panies (63 percent) manufacture nonhealth carerelated products either directly or through aparent firm (see table 11). These products rangefrom assorted electrical equipment and householdappliances to electron microscopes and instru-ments for testing. In many instances, sales of theseproducts far exceed those of health care relatedequipment and products.

Of the 15 firms identified in table 11 as multi-product entities, all but two (Bruker Instruments

and JEOL) prod uce diagnostic imaging equ ipmentother than NMR imaging. Of these, six (CGR,Elscint, General Electric, Philips, ” Picker, andSiemens) offer full diagnostic imaging product-lines, including CT, ultrasound, nuclear medicine,digital radiography, and conventional X-ray andfluoroscope. An additional four firms (ADACLaboratories, Hitachi, Technicare, and Toshiba)manufacture products in three or more diagnos-tic imaging modalities.

The diverse nature of industry firms serves tobenefit their R&D efforts in NMR imaging by:

offering technical expertise gained in the de-

I Iphi]ips manufactures nuc]ear cameras, sold in the United States

through ADAC Laboratories.

—

velopment and marketing of other diagnos-tic imaging modalities, such as CT and nu-clear medicine;accelerating product development based oncorporate experience with related technol-ogies in nonhealth care fields, such as NMRspectroscopy; and

increasing the R&D resource base availableto NMR imaging development through thesales of various other products in both healthcare and nonhealth care fields.

Diversification in the industry is likely to in-crease in the future, as some small firms expandoperations into new product lines and as somelarge companies augment their already diverseportfolios.

Acquisition and Merger Activity

Since its inception in the mid-1970s, the NMRimaging device industry has witnessed a consid-erable number of acquisitions, mergers, and im-portant trad e agreements among its member firms.Acquisition and merger activity may be classifiedinto four major groups (9,31,139):

Product extension—in which a companygains entry into a related market by acquir-ing a firm that sells products not presentlyproduced by the parent.Market extension—in which a company con-solidates or increases its market share by ac-quiring a firm in the same product line.

Conglomerate merger—in which a parentcompany acquires another company that isunrelated in either product or market.Vertical integration-in which a company ac-quires another firm whose activity is impor-tant to the processing, manufacturing, sale,or distribution of the parent company’sproduct.

Most acquisitions and mergers in the industryhave been oriented toward product extension in-volving various diagnostic imaging modalities (seetable 12). Among these have been two cases in-

volving NMR imaging. In one instance, DiasonicsInc., acquired the rights to NMR imaging tech-nology developed under an agreement betweenPfizer, Inc. and the University of California, SanFrancisco (UCSF) Radiological Imaging Labora-

8/14/2019 8420


8/14/2019 8420


52 Ž Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, industrial, and Policy Analysis

Table 12.—Acquisitions, Mergers, and Key Trade Agreements in the NMR Imaging Device Industry Since 1971

Year and nature of acquisition/ merger/ trade agreement activity Comment

ADAC Laboratories: 1981: Agreement with Picker International for manufacturing digital

angiography systems1982: Agreement with Fischer Imaging Corp. for manufacturing digital

angiography systemsBruker Instruments:

1982: 25% of company ownership acquired by IBM

1983: Acquired Oxford Research Systems from Oxford Instruments

CGR Medical Corp: 1971 : Created by the acquisition of Westinghouse Medical X-Ray Division

by CGR of France1979: CGR of France merged with Thompson-CSF to form Thompson-

BrandtDiasonics Inc.:

1981: Acquired rights to NMR technology developed under agreementbetween Pfizer and the UCSF Radiological Imaging Laboratory

1981: Acquired rights to cardiology ultrasound technology developed byVarian Associates

1982: Acquired Sonotron Holding A.G.

1983: Acquired Sonics Imaging, Inc.

1983: Acquired Fischer Imaging Corp.

Fischer imaging Corp.: 1980: Acquired the Medical Ultrasound Division of EMI1982: Agreement with ADAC Laboratories (see above)1983: Agreement with M&D Technology Ltd. to become exclusive

marketing agent for M&D NMR imager

1983: Acquired by Diasonics Inc.General Electric:

1980: Acquired Thorn (CT Scanning Division of EMI)JEOL USA:

1973: Parent firm, JEOL of Japan, acquired by Mitsubishi1982: Agreement with Smith, Kline & French Laboratories for joint

research into NMR spectroscopyM&D Technology Ltd.:

1982: Formed and financed by a combination of private and publicinvestors

1983: Agreement with Fischer Imaging Corp. (see above)Nalorac Cryogenics:

1977: Acquired by Nicolet Instruments

1982: Divested by Nicolet Instruments and reestablished as independentfirmPicker international:

1981 : Created by the acquisition of the Picker Corp. by GEC of England,and its subsequent merger with GEC Medical and CambridgeMedical Instruments

1981 : Acquired rights to NMR technology developed by EMI1981: Agreement with ADAC Laboratories (see above)

Technicare Corp.: 1979 Acquired by Johnson & Johnson, Inc., from Ohio Nuclear1982: Acquired Magnet Corp. of America

—

—

Agreement terminated same year

IBM provides grant support to MIT and Harvard for

NMR imaging research at Brigham & Women’sHospital (Boston) using Bruker equipment

—

Market extension: X-ray equipment

Conglomerate merger

Product extension: NMR imaging

Product/market extension: phased array ultrasound

Vertical integration: Western Europeandistributorship

Vertical integration: Southeastern U.S.distributorship

Product extension: X-ray equipment

Product extension: ultrasound(See ADAC Laboratories above)Product extension: NMR imagingb; agreement

terminated same year following acquisition byDiasonics (NMR) Ltd.

(See Diasonics Inc. above)

Market extension: CT scanning

Conglomerate merger: NMR spectroscopi c

—

—

(See Fischer Imaging Corp. above)

Product extension: superconducting magnets

Product extension: X-ray equipment, CT scanning

Market extension: NMR imaging(See ADAC Laboratories above)

Product extension: CT scanningVertical integration: superconducting magnets

aDia~~ni~~, prime ~urpo~e inacquiring Fischer Imaging corp. Was to Obtalrl radiographic and fluoroscopic equipment to which it could add i!S comPuter software.

Fischer Imaging Corp had, by May 1983, already made a commitment to NMR imaging, but had not yet begun extensive R&D.bFischer,s May 1963 marketing agreement with f$f&D Technology Ltd. was an attempt to extend its product line into NMR imaging without having to conduct extensive

R&D efforts, Two weeks after signing the agreement, Fischer was acquired by Dlasonics.CJEOL of Japan had been manufactur~ng NMR spectrometers since Iw o Followlng its acquisition by fditsublsht, R&D efforts continued but It was not Untl[ 1982 that

the company formally entered the NMR imaging field.dJohnson & Johnson had developed interest in NMR imaging as early as 1977, but it was not until after the acquisition of Technicare (and Its CT technology) that

serious R&D efforts into NMR imaging were undertaken

SOURCES Interviews with manufacturers, Arthur Young & Co., 1981 (9), and Emmitt & Lasersohn, 1983 (60)

8/14/2019 8420


——

Ch. 4–- The NMR Imaging Device Industry 53 — ——

Diasonics’ acquisition of the Pfizer-UCSF technol-ogy, Picker used its new technology to acceler-ate its market entry and to catapult to the indus-try forefront.

At least three mergers in the industry have in-volved vertical integration. In one case, integra-tion has been “backward”: Technicare’s purchaseof Magnet Corp. of America for the purpose of building its own superconducting magnet systems(see subsequent discussion in this chapter of themagnet manufacturing industry). The other twomergers represent “forward” integration wherebyDiasonics acquired companies to expand its salesand distributorship network to specific geographicareas (see table 12 and later discussion in thischapter of vertical integration under “industryconduct”).

Trade agreements involving marketing and dis-tribution rights are fairly common in the indus-

try, and those listed in table 12 are probably buta subset of all transactions that have taken place.Joint research ventures among firms, on the otherhand, are rare, if not nonexistent. Manufacturerstend to be secretive about their NMR imaging de-signs and place units in clinical settings only if thehospitals agree not to accept a companion unitfrom a competitor for comparative study pu rposes.

As the NMR imaging device industry matures,one may expect further market extension andproduct extension mergers as some smaller firmsare acquired by larger competitors or by firms

seeking to enter the industry .12 A high degree of vertical integration is also likely, as many firmswill seek to expand internal capacity for market-ing and distribution of products and for produc-tion of NMR component parts (e.g., magnets,cryogenic systems, and computer consoles andsoftware). Magnet production capabilities are par-ticularly important to manufacturers who wishto minimize both production costs and delays inreceiving supplies in order to stay competitivewith other companies. In addition to Technicare,which owns a magnet company, at least five otherfirms (Bruker, Diasonics, Elscint, Fonar, and

Nalorac Cyogenics) possess in-house magnet man-ufacturing capabilities, while another seven plan

“The acqul>] t io n of ON!I< Technology>’ b} Xonlcs in late 1 Q83 IS

a turtht’r example of produ[ t twtens](>n,

to vertically integrate this function over the next2 to 5 years.

The Magnet Manufacturing Indus t r y . T h emagnet manufacturing industry is considerablymore concentrated than the NMR imaging sys-tem manufacturing industry, Only a small num-ber of firms make superconducting magnets, andlittle is known about manufacturers of resistivemagnets.

According to a report from Hambrecht & Quist,as of September 1982 the majority of supercon-ducting magnets used in NMR imaging systemsworldwide had been supplied by a single manu-facturer, Oxford Instruments, Ltd., a companybased in the United Kingdom (80). For the yearended March 1983, Oxford Instruments had salesof 30 million English pounds, with profits of over2.5 million pounds, an increase from 17.7 millionp o u n d s i n s a l e s a n d a p p r o x i m a t e l y 2 m i l l i o n

p o u n d s in profits in 1981-82 (1 19). In 1983, Ox-ford Instruments produced about s ix magnets permonth and had secured long- term contracts tosupply magnets to several NMR imaging manu-facturers, including General Electric and Siemens(119). These orders would require an increase i nOxford’s production capacity to about 12 magnetsper month (119). To fulfill this increased demand,Oxford planned to hold a public stock offeringin 1983 to secure funds to expand its productioncapability (119), and opened a manufacturing fa-cility in the United States in a joint venture withAirco, Inc. (51). (Airco, a subsidiary of the Brit-

ish Oxygen Co. International, Ltd., has the ca-pability of producing superconducting materialsrequired in the manufacture of superconductingmagnets. )

According to a 1982 prospectus issued by anAmerican magnet manufacturer , In termagnet icsGeneral Corp. (IGC), there are at least six Amer-ican manufacturers selling superconducting mag-nets to NMR imaging manufacturers . To date ,however, compared to Oxford Instruments, Ltd.,these magnet manufacturers have not supplied sig-nificant numbers of superconducting magnets to

N M R i m a g i n g m a n u f a c t u r e r s .13

11 l~;~- Ic)r ~XamPle h as been i nvo] veci tor over 10 years i n themanufacture ot the materials from which super-conducting magnetsa re const ructecl. Rec-entl}’, it h as begun appl}’ing it~ expertise in \ u p e r-

L onduct in g techn[)log}” to t h e dci’elopment of super c(~nduc t in~

8/14/2019 8420


54 • Health Technology Case Study 27: Nuclear Magnetic Resonance imaging Technology: A Clinical, Industrial, and Policy Analysis

Two principal superconducting materials arecommercially available for the construction of superconducting magnets: niobium - titanium(Nb-Ti) wire and niobium - tin (Nb 3Sn) tape.According to the 1982 IGC prospectus, there areseveral foreign manu facturers of Nb-Ti and Nb 3Sn.and IGC is the leading domestic producer of bothmaterials (150). Airco, Inc.; Magnet Corp. of America (now a subsidiary of Technicare); andSupercon, Inc., are the other domestic suppliersof Nb-Ti and Nb3Sn .

magnets for use in NMR imaging systems. IGC increased its R&Dexpenditures from $264,000 in fiscal year 1981 to $1.5 million infiscal year 1982 to help develop its magnet manufacturing capacity(104). It is currently manufacturing 0.5 tesla (T) and 1.5 T magnets.IGC supplied its first 1.5 T magnet (to Columbia-PresbyterianMedical Center) in March 1983 and planned to produce one to threemagnets per month for the remainder of 1983. IGC has also begunconstruction of a new factory, which should be operational by 1984an d which will double its magnet production capacity. As of theend of August 1982, IGC had a backlog of $2.2 million in ordersfor superconducting magnets from NMR imaging manufacturers(105).

INDUSTRY CONDUCT

The structural characteristics of aging device industry (i.e., its high

the NMR im-seller concen-

tration, relatively easy market entry, considerablediversification, high degree of acquisition andmerger activity, and low buyer concentration)have conflicting implications for competition

among manufacturers. The behavior, or conduct,of the market is likely to be influenced not onlyby the policies and actions of individual firms,but also by their reactions to the policies of theirrivals, Two important aspects of industry andmarket conduct are product pricing policies andnonprice competition strategies.

Product Pricing Policies

Based on interviews with manufacturers in1983, the estimated sales price of a resistive mag-net system is likely to range from $800,000 to $1.2

million. Superconducting magnet systems, de-pending on size and field strength, are expectedto comman d prices between $1 million and $3 mil-lion, with the median expectation closer to $2 mil-

Oxford Instrum ents is the major sup plier world-wide of resistive magnets for NMR imaging sys-tems (80). Technicare and Bruker manufacturetheir own resistive magnets, and Fonar and OMRmake their own permanent magnets.

With magnets accounting for an estimated 30to 50 percent of the cost of NMR imaging sys-

tems,14

it is not surp rising that N MR imaging m an-ufacturers are seeking to develop their own ca-pacity to produce magnets. As stated previously,our survey of manufacturers found that six firmsnow produce at least some of their own magnetsand seven others plan to develop their own ca-pacity to manufacture magnets. According toIGC, however, it is unlikely that NMR imagingmanufacturers will be able to meet their magnetsupply needs themselves, and they are likely towant to have more than one source of magnets(154).

.—IdAccording to Ivl. J, Ross of IGC, 0.5 T magnets cost $300,000

to $350,000 and 1.5 T magnets cost over $500,000 (154).

lion. Since the FDA has only recently grantedpremarket approval for NMR imaging devices,there has been little experience with product pric-ing and sales.

Most of the manufacturers queried about salesprice felt that it would not be a significant factorin determining future company market share.They instead stressed the importance of nonpricefactors in d ifferentiating their pr odu cts from thoseof competitors (see discussion of product differen-tiation in this chapter). Only four firms viewedsales price as key to the coming competition formarket share. Two companies expressly plan tosegment the market on the basis of price, withlower magnet strength, less expensive NMR sys-tems being offered to community hospitals andprivate radiology groups that may lack the req-uisite financial resources for p urchasing th e higher

magnet strength, more costly models. One firmintends to develop medium-sized superconductingmagnet systems that would sell for as low as$500,000 to $700,000. All four believe, though,

8/14/2019 8420


that industrywide prices will decrease in the long-term (3 to 7 years from now) if for no other rea-son than that: 1) new magnet designs may leadto some efficiencies, 2) increased vertical integra-tion in many companies should reduce produc-tion costs and create economies of scale, and 3)further experience with NMR imaging in clinicalapplications may point to an optimal system con-figuration that is less expensive to produce. Itshould be noted, however, that increased verticalintegration could actually result in higher pricesif such activity serves to diminish competition inthe industry.

At least two manufacturers also believe thatprice cutting will not evolve simply as a conse-quence of the factors cited above, but ratherbecome a conscious policy of the larger firms in-tent upon weakening and acquiring, or possiblydriving out, smaller competitors. Such “predatory

pricing” policies are employed in other industries(31) . Their appl icat ion here would , in the lon g

run, make the NMR imaging device industry mo reconcentrated than the current trend suggests (seeearlier discussion of seller concentration and mar-ket share in this chapter). On the other hand, if one draws inferences from the experience with X-ray CT scanning, price competition may play littleor no role in the coming industry “shakeout. ”Rather, as the next section suggests, nonprice fac-tors may prove more important to companystrategies.

Nonprice CompetitionProduct differentiation and vertical integration

are both expected to figure prominently in thenonprice competition strategies of NMR imagingdevice firms, Given the diversity of potentialbuyers, the ability to differentiate one’s productfavorably from that of a rival may prove impor-tant to future company sales and market share.Vertical integration, in addition to its obvious eco-nomic benefits, may offer further advantage byraising barriers to entry for potential rivals (31),

Product Differentiation. Interviews with man-ufacturers have led to the identification and rela-tive ranking, by tier, of nine nonprice factors con-sidered imp ortant to NMR pr oduct d ifferentiation

Ch. 4— The NMR Imaging Device Industry 55—— —

in future sales efforts. In descending order of rela-tive importance, 15 these elements are:

First Tier (4 factors):

1.

2.

3.

4.

Image quality—high-resolution images of various soft tissues in the head and body areconsidered essential to product sales. Almostwithout exception, this factor ranked first oramong the top tier of elements.Product features and capabilities—productfeatures refer to the magnet type and fieldstrength, bore size, radiofrequency coil de-sign, computer system console and software,cryogenic systems for superconducting mag-nets, magnetic shields, etc. Product capabil-ities refer to measurement of T1 and T2 re-laxation times, imaging capabilities, andspectral analysis capabilities in addition toproton NMR imaging. The relative impor-

tance of each feature or capability to a pro-spective buyer will depend on the buyer’sfundamental imaging needs (e.g., clinical v.research) and level of sophistication. Inno-vative product capabilities, such as multisliceimaging, are imp ortant means of pr oduct d if-ferentiation.Product reliability—reliability is essential tothe continuous operation of an imager and,therefore, is valued highly by prospectivebuyers. Lack of product reliability, such asthe tendency for a magnet to “quench” (i. e.,lose its magnetic properties), can have seri-

ous adverse effect on imager sales.Product service—timely and responsivemaintenance and repair service is importantboth for ensuring client satisfaction and forpreserving company image. Distributor andservice networks covering broad geographicareas are an important asset to marketing theproduct.

Second Tier (3 factors):

5. Delivery time—at present, delivery time canbe very important to some buyers. Overtime, however, as the industry matures and

“The reader should bear in mind that these views are those ex-pressed by NMR manufacturers, which may or may not coinc]de

with the percept ions of potential buyers and users of the techno]og}r.

8/14/2019 8420



the production of NMR imaging units isstreamlined in many f i rms , del ivery t imeshould become less impor tant to buyers . Long-term viability of the manufacturer—the larger, more well-established firms be-lieve that size and tradition are importantassets, and that buyers respond positively to

companies whom they perceive to be viablefor years to come. The smaller, newer firmsconcede this point, but argue that product

characteristics (e.g., featur es an d capabilities,image quality, reliability) will take prece-dence over company characteristics in deter-mining future NMR imaging sales.

7. Guarantee against technological obsoles-cence—when purchasing expensive, newtechnologies, buyers frequently want assur-ance that the model they p urchase today willnot become obsolete in a short period of time. With a technology that is evolving and

changing as rapidly as NMR imaging, suchguarantees are difficult to make. Severalmanufacturers, therefore, have either: 1) de-layed introduction of a commercial proto-type until optimal system design can besatisfactorily determined, or 2) designedNMR systems that can be “upgraded” to ac-commodate new imaging needs or new ad-vances in technology as they arise. However,when compared with other factors listed inthe first tier, guarantees against product ob-solescence were secondary in importance.

Third Tier (2 factors):8.

9.

Collaborative research—at present, in thepremarket stage of NMR imaging R&D, col-laborative research w ith clinical centers holdsgreat importance for manufacturers and hos-pitals alike. In the future, however, manyfirms expect that collaborative research willhold no more than tertiary importance (rela-tive to the preceding factors) in influencingbuyers’ purchase decisions.Training and education—a few firms believethat providing training services to buyers

may become a distinguishing feature of somemanufacturers’ marketing and sales strate-gies. The relative importance of this factorto future sales, though, is not expected to behigh.

————. -.— .—

Overall, product differentiation is emerging asan important part of each NMR manufacturer’snonprice competition strategy. Fonar Corp., forexample, is placing great emphasis on its perma-nent magnet design. ADAC Laboratories, an ac-knowledged leader in “add-on” computer systemsfor diagnostic imaging modalities (185), intends

to emphasize the company’s strengths in imageprocessing, data communication, and radiofre-quency coil design as part of its marketing strat-egy, in addition to pursuing proprietary perma-nent magnet designs. General Electric has adopteda different tack, developing a 15 kilogauss super-conducting magnet prototype, which the com-pany believes will appeal to hospitals concernedabout buying “adequate” magn et strength. NaloracCryogenics expects to differentiate its product byoffering superconducting magnet systems that canoperate within 10 to 20 kilogauss, but which canalso be upgraded to 40 kilogauss for high-resolu-

tion animal studies and NMR spectroscopy. Re-gardless of the specific strategy employed, it seemsclear that product differentiation will be impor-tant to each manufacturer’s success and, in somecases, corporate survival.

Vertical Integration. In the earlier discussion of industry structure and corporate acquisition andmerger activity, vertical integration in the NMRimaging device industry appeared to have impor-tant implications for production costs and, hence,for product pricing policies. Vertical integrationcan also be used by manufacturers to coerce rivals

and influence market entry. For instance, in theNMR imaging device industry, the forward in-tegration of distributorship networks could im-pede other firms from selling their products insome areas. Similarly, backward integration of magnet manufacturers could conceivably bar en-try to potential competitors who are not capableof producing their own magnets and, therefore,must depend on outside suppliers.

Although at least one case of backward integra-tion involving magnets has taken place in recentyears (see Technicare in table 12), it is not likely

that NMR manufacturers will gain control of either of the two major worldwide magnet sup-pliers (Oxford Instruments and IntermagneticsGeneral Corp.). Instead, the net effect of manyNMR manufacturers’ plans to develop in-house

8/14/2019 8420


.Ch. 4—The NMR Im aging Device Industry 57

— .

magnet manufacturing capabilities will likely beto achieve greater independence from the magnetsuppliers. An individual NMR imaging firm thatchooses this strategy could gain a competitive edgeonly if its rivals did n ot vertically integrate in simi-lar fashion, or, assuming that its rivals did fol-low suit, if its magnet operations were more effi-

INDUSTRY PERFORMANCE

Industry performance is most frequently evalu-ated in terms of the efficiency and profitabilityof its firms (31). Common measures of efficiencyinclude costs-to-sales ratios and percent of adver-tising or promotional costs-to-sales ratios (9).Data on advertising and sales in the NMR imag-ing device industry are nonexistent because the

FDA prohibits promotion and profitmaking salesduring the premarket approval stage of develop-ment. Thus, the relative efficiency with whichvarious firms allocate their resources to buildNMR imagers cannot be determined at this time.It is expected that, following FDA approval, pro-motional activities will abound in the industry,largely for product differentiation purposes.

cient or produced higher quality magnets thanthose of its competitors. Vertical integration inthe NMR imaging device industry, therefore, ismore likely over the long run to influence indus-try conduct (i. e., product pricing and product dif-ferentiation) than industry structure (i.e., marketentry by newcomers).

Profitability has been measured as the rate of return on investment (or assets) or the price-costmargin (i. e., the gap between price and marginalcost). As with the previous case of measuring effi-ciency, FDA prohibition on making a profit fromthe placement of an investigational device haspreclud ed th e quantitative assessment of NMR in-

dustry profitability. Available data on the X-rayand electromedical industry show returns onassets ranging from 5.6 percent for the larger firmsto 11.4 percent for companies with smaller assets(9). It is expected that NMR imaging sales willlikely become an important source of companyrevenues for many manufacturers over the nextfew years (60).

THE FUTURE OF NMR IMAGING IN RELATION TO OTHERDIAGNOSTIC IMAGING MODALITIES

Given the uncertainties regarding the natureand impact of future health care regulations, aswell as the extent to which the clinical potentialof NMR imaging will be realized, it is difficult tomake estimates of future sales of NMR imagerswith any degree of certainty.

Table 13 provides data on estimated sales of various diagnostic imaging modalities projectedby F. Eberstadt & Co., Inc. (60). As can be seenfrom the table, in mid-1983, worldwide sales of the diagnostic imaging industry were estimated

at $4 billion per year, and this worldwide mar-ket is currently projected to continue expansionat a rate of 15 percent per year (60). Sales of ultra-sound, digital X-ray equipment, and NMR im-

agers are expected to grow more rapidly thanother segments of the diagnostic imaging market,primarily due to the reduction in or eliminationof ionizing radiation associated with their use.

The table also shows that, despite a projected21-percent increase in aggregate sales of X-raymodalities over the next 5 years, the percentageof all diagnostic imaging industry sales that canbe attributed to X-ray modalities is expected todecrease by 41 percent (from a 72 percent to a 43percent share) between 1983 and 1988. X-ray CT

unit sales are projected to decrease dur ing thattime from $1 billion per year to $0.5 billion peryear, a 76-percent decrease (from 25 percent downto 6 percent) .

8/14/2019 8420


58 Health Technology Case Study 27: Nuclear Magnetic Resonance Im aging Technology: A Clinical, Industrial, and Policy Analysis—. —— ————— —-———

Table 13.—Diagnostic Imaging industry Sales Growth Projections

1983(E) 1988(E) 1983 to 1988

Overall Annual OverallPercentage Percentage percentage percentage change inof industry of industry change in change in fraction of

Market size sales Market size sales market size market size industry salesModality ($ millions)

— ( % ) ($ millions) (%) (%) (%) ( “ / 0 )

All X-ray modalities ., . . $2,900 72.5% $3,500 43% +21% + 4 % – 4 1 %

Conventional X-ray . . . . . . (1 ,300) (32.5) (500) (6) (-61) (– 17) (-82)Digital X-raya. . . . . . . . . . . . . . (600) (15,0) (2,500) (30) (+317) (+33) (+100)CT . . . . . . . . . . . . . (1 ,000) (25,0) (500) (6) ( -50) (- 13) (-76)

Ultrasound . . . . . . . . . . . . . . . . . . 750 19,0 1,900 23 + 153 +20 +21Nuclear medicine ... . . . . . . 250 6.0 300 4 +20 + 5 –33NMR . . . . . . . . . 100 2,5 2,500 30 +2,500 +90 + 1,100

Total . . . . . . . . . . . 4,000 100,0 8,200 100 + 105 + 15.0 —

alncludes both digital acid-on and full. syslems with a dlwtal capablhty —.

SOURCE R B Emm!tt and J W Lasersohn, “Company Report on Dlasonics, ” F. Eberstadt & Co , Inc , New York, May 26, 1983

NMR sales, in contrast, are expected to increasefrom $100 million per year in 1983 to $2.5 bil-lion per year in 1988 (see tables 13 and 14 andfig. 11), an annual rate of growth in market size

of 90 percent. According to this estimate, the per-centage of industry sales attributable to sales of NMR imaging systems will increase from 2.5 per-cent in 1983 to 30 percent by 1988. The estimatedrate of growth in worldw ide NMR sales displayedin table 14 can be compared to a worldwidegrowth of X-ray CT unit sales of approximately60 0 units per year over the first 5 years of X-rayCT availability (59). Most manufacturers withwhom we spoke believed that the 50-50 percentsplit between U.S. and non-U. S. sales currentlyexisting for X-ray CT systems will be observedfor NMR sales as well.

It is useful to consider the assumptions on whichthe estimates are based. First, the estimates assumethat, given the expected change to prospective sys-tems of hospital payment, the hospital industrywill be unable to bear a major increment in capi-

tal expenditures over the next several years. It isassumed, however, that although the rate of in-crease of hospital expenditures on imaging equip-ment may slow over the next several years, the

slowing will be offset by increases in purchasesby private radiology groups and that the recentgrowth in sales of diagnostic imaging equipmentof 15 percent annually will remain constant overthe next 5 years.

The second major assumption is that, at leastin the near future, sales of NMR imaging systemswill compete primarily with sales of X-ray CT sys-tems. This situation is thought to be the case be-cause both systems provide cross-sectional tomo-graphic images with what is expected to be similarspatial resolution in the near future. The estimatestherefore assume th at man y hospitals will be mak-ing decisions about purchasing either NMR im-agers or X-ray CT scanners.16

10ThiS wi]l a]so be th e Case for hospitals that have one or more

X-ray CT scanners and are contemplating buying additional ones,In 1980, more than 100 U.S. hospitals had more than one X-ray CT

Table 14.— Estimated Worldwide NMR Market

Annual unit Cumulative unit Average A n n u a l s a l e sYear a deliveries deliveries sales price ($ million)

1982 . . . . . . . . . . . . . . . . . 15 15 $1,300,000 $ 201983E . . . . . . . . . . . . . . . . 75 90 1,300,000 100

1984E . . . . . . . . . . . . . . . . 200 290 1,500,000 300

1985E . . . . . . . . . . . . . . . . 400 690 1,600,000 6501986E . . . . . . . . . . . . . . . . 650 1,340 1,700,000 1,1001987E . . . . . . . . . . . . . . . . 905 2,290 1,850,000 1,7501988E . . . . . . . . . . . . . . . . 1,250 3,540 2,000,000 2,500aE ~ Est!mated

SOURCE R B Emmitt and J W Lasersohn, “Company Report on D!asontcs, ” F Eberstadt & Co , Inc , New York, May 26, 1983

8/14/2019 8420


8/14/2019 8420


60 • Health Technology Case Study 27 : Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis

noted, though, that while the model assumes thatsales of NMR imagers will exceed sales of X-rayCT imagers in 1988, it does not assume that useof NMR will replace use of X-ray CT clinicallyin the near future. 17

Finally, the estimates assume an annual infla-tion rate of 5 percent.

In addition to it being difficult to predict themagnitude and nature of future diagnostic imag-ing equipment sales, it seems equally hazardousat this time to project what the character of the

’71t seems very likely that NMR will not replace X-ray CT in thenear future. There may always be a role, for example, for X-rayCT in patients with metallic implants who will not be consideredcandidates for NMR, and for guiding biops y or surgical proceduresthat employ metallic instruments. Furthermore, it is difficult topredict the extent to which X-ray CT scanning techniques will im-prove in the future. In the 10 years since X-ray CT scanners wereintroduced, their scanning speed has increased 300 times, their spatialresolution has increased 8 times, their densit y resolution has increased

3 times, and their radiation dosage to the patient has decreasedmarkedly (15). In addition, new X-ray CT scanners are being de-veloped that are capable of completing a scan in about 30 milli-seconds, permitting performance of real-time cardiac X-ray CTimaging (15).

NMR imager manufacturing industry will be 5years hence. Although no predictions can be m aderegarding how many or which of the currentNMR imaging manufacturers will be involved inthe field in 1988, two general comments can bemade. First, to the extent that “turf battles” be-tween radiologists, nuclear medicine specialists,

pathologists, neurologists, and cardiologists de-velop over control of NMR imaging and spec-troscopy, the market shares of large X-ray man-ufacturing companies that are currently based onradiolog y franchises may decrease (60). Second,although the emergence of new im aging mod alitieswas thought to be the primary challenge to major X-ray manufacturers in the 1970s, it is an-ticipated that the rapidly expanding role of dataand image processing across all imaging modalitieswill become the major challenge to X-ray equip-ment manufacturers in the 1980s (60). To the ex-tent that this does, in fact, become the case, future

concentration or fragmentation in the diagnosticimaging market may, in large part, be determinedby the responsiveness of large diagnostic imag-ing conglomerates to this anticipated trend (60).

8/14/2019 8420


5.Hospital Costs and Strategies

8/14/2019 8420


Hospital Costs and Strategies

INTRODUCTION

One of the major concerns regarding NMR im-aging relates to the impact this new technologywill have on health care costs. This concernderives in part from the high costs anticipated forthe purchase and installation of an NMR imag-ing system and uncertainties regarding the extentto which NMR imagers will be used in additionto, rather than instead of, other diagnostic mo-dalities already in hospitals. The purpose of thischapter is to present a framework for addressing

CAPITAL AND OPERATING COSTS

Initial capital expenses and routine operatingexpenses for an NMR imaging system are primar-ily determined by the type of magnet used to pro-duce the static magnetic field. Although insuffi-cient experience has been accumulated to permitaccurate predictions of the likely expenses of ac-quiring and operating an NMR imaging system,tables 15 to 17 attempt to provide the best avail-able estimates of such expenses for four differenttypes of NMR systems. It should be emphasized

that the values provided in the following tablesare estimates, and nothing more. They shouldtherefore be interpreted in that spirit. Our pur-pose in presenting these estimates is to providesome indication of the factors that will contrib-ute to the cost of operating NMR imaging sys-tems and what total costs are likely to be, givenreasonable assumptions.

In determining annualized capital costs, wehave made the assumption that the useful life of the NMR imaging system itself is 5 years, whilethat of the site renovation is 10 years. A 10-percent interest rate has been employed. Physi-cian costs (i. e., professional fees) have not beenincluded.

supply costs,dure, vary with

estimated to be $15 per proce-the number of procedures per-

the cost issue. The chapter is organized into threesections. The first section presents data regardingthe likely capital and operating costs of differenttypes of NMR systems. The second section ad-dresses other factors that will influence the effectof NMR imaging on patient care costs, and thefinal section describes the NMR acquisition strat-egies and decisions of different segments of thehospital industry.

formed per day. In the near future, while NMRremains investigational, it is probably reasonableto assume that no more than 10 procedures willbe performed per day. By 1985 or 1986, it is rea-sonable to assume that 20 procedures per day willbe performed on a single machine. If paramagneticcontrast agents come into use, supply costs willincrease.

Maintenance costs are estimated to be either 5or 7 percent of the purchase price. The former esti-

mate is more reflective of the cost of maintenanceperformed by the hospital; the latter of the costof maintenance provided by a manufacturer aspart of a service contract. Given the absence of moving parts in NMR imaging systems, actualmaintenance costs may be less than the estimatesprovided. Finally, overhead, which has been esti-mated to be 25 percent of operating costs, willvary from institution to institution.

As can be seen in table 15, purchase prices forpermanent and superconducting NMR units tendto be higher than those for current generation

X-ray CT scanners (estimated to be $600,000 to$1,200,000) (180,181). Based on the assumptionsimplicit in this analysis, including the perform-ance of 10 procedures per day, the cost of an NMRimaging study in 1983 exclusive of professional

63

8/14/2019 8420


8/14/2019 8420


Ch. 5—Hospit al Costs and Strat egies 65

Table 16.—Estimated Annual Costs for NMR Imaging Systems by Type of Magnet, 1983

Resistive Permanent(0.15 T) (0.3 T)

Annual capital costs: a

Purchase b . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Facility modification . . . . . . . . . . . . . . . . . . . . . . . .

Total annual capital costs . . . . . . . . . . . . . . . . . . . .

Annual operating costs.d

# proceduresShift maintenance per day

SingleSingleSingleSingleDoubleDoubleDoubleDouble

5 %5 %7 %7 %5 %5 %7 %7 %

10 ..., . . . . . . .20 . . . . . . . . . . .10 . . . . . . . . . . .20 . . . . . . . . . . .10 . . . . . . . . . . .20 . . . . . . . . . . .10 . . . . . . . . . . .20 .., . . . . . . . .

Range of operating costs . . . . . . . . . . . . . . . . . .

Annual overhead: 25°/0 of operating costs . . . . . . . . . . . . . . . . . .

Total annual costs. . . . . . . . . . . . . . . . . . . . . .

$203,972(800,000)

23,787 (150,000)

$227,759

$177,500215,000193,500231,000247,600285,000263,500301,000

177,500

to301,000

44,375to

75,250

449,634to

604,009

$ 3 8 2 , 4 4 7(1,500,000)

11,894(75,000)

$ 394,341

$ 190,700228,200220,700258,200260,700298,200290,700328,200

190,700

to328,200

47,675to

82,050

632,716to

804,591

Superconductive Superconductive(0.5 T) (1,5 T)

$ 382,447(1 ,500,000)

103,078(650,000)

$ 485,525

$ 232,500270,000262,500300,000302,500340,000332,500370,000

232,500

to370,000

58,125to

92,500

776,150to

948,025

$ 5 0 9 , 9 2 9(2,000,000)

111,007(700,000)

$ 620,936

$ 267,500305,000307,500345,000337,500375,000377,500415,000

267,500

to415,000

66,875t o

103,750

955,311t o

1,139,686aTotal costs on which annualized costs are based are in parenthesesbAssumlng 5.year useful life, IO percent interest, and 60 equal monthly PaYmentscAssumlng lo.year useful ilfe, IO p e r c e n t interest, and 120 equal monthly PaYmentsd T he ~lectrlclty and Cwogens costs assumed (~ryogens apply only to superconductive systems) were as follows (S00 table 15)

Resistive (O 15 T) &30,c#o

Permanent (O 3 T) $ 8,200Superconductive (O 5 T) ~ ~ $50,000Superconductive (1 5 T ) $60,000

SOURCE Table 15

Table 17.—1983 Estimated Costs per NMR Imaging Study by Type of Magnet

Resistive Permanent Superconductive Superconductive(0.15 T) (0.3 T) (0.5 T) (1.5 T)

Single shift, 5% maintenance,—

10 procedures per day: Annual cost ... ... . . . . . . . . . . . . . . . . . . . . . . $449,634 $632,716 $776,150 $ 955,311Cost per procedure. ., . . . . . . . . . . . . . . . . . . . . . . $ 180 $ 253 $ 310 $ 382Double shift, 7% maintenance,

20 procedures per day: Annual cost ... . . . . . . . . . . . . . . . . . . . . . . . . $604,009 $804,591 $948,025 $1,139,686Cost per procedure ... . . . . . . . . . . . . . . . . . . . . $ 121 $ 161 $ 190 $ 228

SOURCE Table16

8/14/2019 8420


66 Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy An alysi s—.—— — —

fees will likely range between $180 and $382,depend ing on the type of magnet system emp loyed(see tables 16 and 17). However, Joseph P. Whalen,Chief of Radiology at The New York Hospital-Cornell Medical Center, New York City, recentlyestimated the cost of an NMR study as $700 (111),and recent industry estimates range from $382-

$632 (66) to $500-$700 (159). This range of esti-

PATIENT-CARE COSTS

To consider only capital and operating expensesin a discussion of the fiscal impact of NMR im-aging on hospital costs (or health care costs ingeneral) ignores the effect of NMR imaging on p at-terns of patient management. Although physi-cians, hospital administrators, and health careresearchers have alluded in the past to the impor-

tance of technology’s effect on patterns of patientmanagement, techniques for estimating the mag-nitude of a technology’s effect on health costs arefairly primitive. With the advent of prospective,per-case systems of hospital payment, however,it has become increasingly important, particularlywith regard to decisions to acquire new technol-ogy, for hospital managers to be able to explicitlyassess the expected marginal cost of new servicesin relation to projected marginal benefits (102).

Regardless of the potential attractiveness of NMR imaging (or spectroscopy) as a diagnostic

or research tool and the potential of NMR to bea cost-saving addition to physicians’ diagnosticarmamentarium, the actual impact of NMR onhealth care costs will depend not only on its diag-nostic efficacy, but also on how it is employedby physicians in actual practice situations. Sev-eral factors should be considered in this regard.The first is the extent to which NMR imaging isperformed instead of, as opposed to in additionto, other diagnostic modalities in the managementof specific patient complaints or disease entities.It is possible, for example, that NMR will be usedto assess the existence of lumbar disc protrusion

in the evaluation of patients with low back pain,since it can provide excellent images of the verte-bral column. To the extent that NMR imagingsubstitutes for the more invasive and risky tech-nique of myelography (in which a radiopaque

mates derives from differences in underlying as-sump tions and suggests that it is too early to makecost-per-stud y estimates with much precision. Inparticular, lower estimates appear to reflect per-sonnel and maintenance costs more typical of routine operations of a settled and defined tech-nology, not of one still in an uncertain and de-

velopmental phase.

substance is injected into the spinal arachnoidspace), NMR may decrease the cost of managingand increase the quality of care of such patients.To the extent that NMR is used in addition tomyelography, however, NMR might improve pa-tient care, but at additional expense.

A second determinant of NMR’s impact onhealth care costs is the extent to which it will beused in situations in w hich n o diagnostic modalityis currently used.1 Since NMR use does not in-volve radiation risk, such “newly induced” testusage may occur frequently. Consider, for exam-ple, the patient with low back pain alluded to pre-viously. Patients suspected of having low backstrain might in the future undergo NMR scans toconfirm the clinical assessment of strain, ratherthan being treated with bed rest, heat, and anal-gesics without the use of any diagnostic imagingmodality. Such use of NMR is likely to increase

health care costs.Two other potential newly induced test uses can

be foreseen with the introduction of NMR. Thefirst is “sequential NMR scanning” (see ch. 2) tomonitor the natural history of disease in patientswith atherosclerosis, multiple sclerosis, cysts, etc.,in whom symptoms have either increased, de-creased, or even not changed. The second is “se-quential NMR scanning” to monitor therapeuticprogress in patients being treated for cancer, in-fections, etc. The extent to which NMR will, infact, be used in such a fashion should be deter-

mined by the sensitivity and specificity of the tech-nique in each clinical application. The impact such

‘This could be considered a special case of “add-on, ” with thepatient history and physical examination being construed as diag-nostic tests,

8/14/2019 8420


.

Ch. 5—Hospital Costs and Strategies 67 — . —

usage will have on health care costs will dependon the as yet u ndemonstrated extent to wh ich su chusage decreases or increases total patient manage-ment costs in add ition to imp roving diagnostic in-formation or the quality of patient care.

Recent analyses have suggested that over the

past decade there have been striking increases inthe amount of real inputs employed both per pa-tient-day and per admission in U.S. hospitals (48).It is much too early to determine the aggregateeffect of NMR imaging on patient care costs.Much will depend on such factors as how muchsurgery is avoided, whether hospital lengths of stay are shortened, and whether diagnostic work-ups that were previously performed in the hospi-tal are shifted to the outpatient setting.

With the advent of prospective, per-case sys-tems of hospital payment and increasing competi-

HOSPITAL

Introduction

STRATEGIES

Different segments of the hospital industry haveemployed different strategies for determiningwhether, when, and what type of NMR imagingequipment to buy. Each strategy and subsequentacquisition decision reflects the priorities of thehospital-industr y segment or of particular hospi-tals within a segment and provides insights not

only into technology assessment as practiced byhospitals, but also into hospitals’ perceptions of the state of development of NMR imaging tech-nology. An attempt has been made in this sec-tion to describe the acquisition strategies and deci-sions of three different segments of the hospitalindustry: university teaching hospitals and ma-

jor medical centers; the Veterans Administration;and investor-owned hospital chains.

University Teaching Hospitals andMajor Medical Centers

The Acquisition Decision

Most of the early NMR units acquired by hos-pitals have been installed in university teachinghospitals or major medical centers. This is not sur-

tion in health care, it is likely that those vestedwith the responsibility for making decisions re-garding the acquisition of new technology suchas NMR for hospitals will increasingly feel theforce of two conflicting incentives. On the onehand, there will be the already mentioned fiscalpressure to be increasingly discerning of the pa-

tient care benefits compared to costs associatedwith acquisition of new technology. On the otherhand , there will be pressure to offer the “best” andmost recently available services in order to pro-tect (or increase) an individual hospital’s share of patients in the increasingly competitive market forpatients. Whether and how hospital directors willobtain the type of information necessary to makesuch decisions may d etermine not only which hos-pitals survive in the current economic climate, butalso the rate at w hich they acquire pr omising newtechnology such as NMR imagers.

prising given the interest such hospitals have inperforming research and being at the “cuttingedge” of medical developments, the manufac-turers’ need to have research performed in orderto obtain FDA premarket approval, and the ten-dency of such hospitals to have large numbers of beds and a complicated mix of patients.

In addition to these forces driving university

hospitals to acquire NMR imaging technologyearly on, several benefits that have accrued tothose university hospitals that were among theearliest to acquire NMR imaging systems may helpexplain their acquisition decisions.

First, university centers have been able to usetheir special strengths to obtain NMR imaging sys-tems from manufacturers at decreased or even nocost, Among the assets that university hospitalsoffer to manufacturers are: 1) their ability to pro-vide a “laboratory setting” in which clinical data,needed by manufacturers for preparation of anFDA premarket approval application (PMAA),can be collected; 2) their special research talentsin basic science, engineering, and clinical trial de-sign, from w hich m anufacturers have derived ben-efits in the form of improved system design, helpwith PMAAs, and publicity from publications in

8/14/2019 8420


68 Ž Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, industrial, and Policy Analysis

professional journals or presentations at profes-sional meetings; and 3) the prestige associatedwith their institutions, which manufacturers’ mar-keting divisions can convert into an effective formof advertising.

Second, because many of the university hospi-

tals that were first to obtain NMR systems didso at little or no charge, they have ironically pro-tected themselves from much of the cost associ-ated with technological obsolescence.2

Third, the “price” and operating costs of theseexperimental systems are often partly subsidizedby research grants provided to the hospital by themanufacturers.

Fourth, because many university hospitals haveshared their NMR imaging systems with nonhos-pital university researchers, some of the acquisi-tion expenses were often shared with the univer-

sity. Finally, those hospitals and universities thatobtained NMR imaging systems early are now ina position to capitalize on any research funds thatwill be awarded in early 1984 by the NationalCancer Institute as part of its “Comparative NMRImaging Studies” program (see ch. 6).

These observations suggest that many of theuniversity hospitals that have obtained NMR im-aging systems to date may have done so partlybecause they did not have to be so concerned withacquisition and early operating costs as other hos-pitals have to be.

Interestingly, in the case of NMR, a second op-portunity to capitalize on university teaching hos-pital assets is now emerging for those hospitalsthat did not benefit the first time around. This

‘Although NMR imaging systems will undoubtedly change overtime, some experts believe that the changes in the hardware willbe much less dramatic than those that have occurred with X-rayCT scanners. Aside from the possibil i ty that low field strengthresistive systems will become obsolete compared to higher fieldstrength superconducting systems (which could be a concern for those“early bird” hospitals that acquired resistive systems), NMR sys-tems may simply evolve through a continuing series of upgradesin software and minor changes in electronics (90). With X-ray CTscanners, in contrast, the entire set of electronics as well as the

reconstruction algorithm, and other parts of the system are specificto the particular gantry being used. Thus, improvements have comein generations rather than through a process of simple upgrades (90).One other issue to be considered in this regard is that “early bird”hospitals might also have to redesign their facilities in the futureto accommodate changes in magnet design or field strengths.

second opportunity derives from the fact that in-creasing numbers of manufacturers are beginningto offer high field strength (1.5 T) NMR systemson which spectroscopic applications (an area inwhich many universities are replete with talent)need to be explored. These second-round hospi-tals can be expected to be fewer in number than

those in the first round and are likely to obtaintheir benefits in the form of “two-for-one” bar-gains in which a 1.5 T system and a lower fieldstrength system are obtained for close to the priceof the lower field strength systems. Second-roundbuyers will benefit from the experience in site-planning gained by first-round hospitals.

Choosing a Manufacturer (or University)

Manufacturers have tended initially to installequipment in prestigious university centers. Thereis little information available on whether these

first-round hospitals sought ou t the m anufacturersor whether the manufacturers courted the hospi-tals, In some instances (such as Siemens andWashington University in St. Louis), installationshave been a natural consequence of longstandingbusiness relationships and research collaborations.

Several factors can be expected to influenceuniversity teaching hospitals’ choices of manufac-turers in futu re round s of buying. Potentially mostimportant are the hospitals’ perceptions of a man-ufacturer’s survivability in the NMR industry, theNMR system’s features and capabilities, imagequality, system reliability (up-time), manufac-turers’ interest in collaborative research, and theeffect of a hospital’s choice on its existing rela-tionships with radiology equipment manufactur-ers. Potentially of lesser importance to universityhospitals interested in performing research areprice, protection against early obsolescence, anddelivery time. To the extent that a research col-laboration evolves, good service and technologyupgrades (software and hardware) can be ex-pected.

The Veterans Administration3

The Veterans Administration operates 172 med-ical centers n ationwide in 6 regions and 28 districts

3The information in this section was drawn from a personal com-munication with S. Smith (171).

8/14/2019 8420


.- .-

Ch. 5—Hospital Costs an d Strategies 69

on an annual medical service budget of $7 billionto $8 billion. Of these 172 medical centers 130have onsite nuclear medicine services, and 80 to90 have X-ray CT scanners.

The VA central office must approve acquisitionof technology costing more than $100,000 (a sort

of “certification-of-need” (CON) analog). Thesehigh-cost items are ap parently not considered p artof the budgets allocated to individual hospitals,districts, or regions by the VA central office.

The NMR Decision

The VA’s interest in acquiring an NMR imageroriginated in the VA central office rather than inone of its hospitals. In December 1981, after apresentation by an NMR manufacturer, VA electedto defer acquis i t ion of an NMR sys tem.

In early 1983, the VA decided to initiate whatcould become a program of s taged acquis i t ionwith a s ingle NMR demonstration and evaluationproject. This decision derived from an interest i n“helping the VA march into the future” (171). Noestimates of the fiscal impact of NMR on the costof patient care were made. The decision to restrictthe initial purchase to a single unit emanated froma concern about the rapid rate at which NMRtechnology was changing and the desire to avoidinstalling a large number of systems that mightsoon become obsolete.

Choosing a Manufacturer

In early 1983, the VA solicited bids from man-ufacturers for a single system. No specificationswere given regarding the type of magnet desired.Two bids, both for 0.15 T resistive magnet sys-tems, were received.

Three factors guided the VA in its final choiceof a system and manufacturer. First was a con-cern about manufacturer “corporate durability. ”(The VA regrets having bought six to eight X-rayCT scanners from Pfizer, which subsequently

stopped manufacturing X-ray CT systems. ) Sec-ond was evidence of a manu facturer’s pr oven rec-ord of reliability in its already existing installa-tions. Third was price.

Choosing a Site

The 0.15 T system obtained by the VA was tohave been installed in the Cochran VA Hospitalin St. Louis in October 1983. This decision wasagain made centrally with interest expressed bythe Cochran VA. Three factors were considered

in the choice of an installation site: the site hadto have all other major diagnostic imaging mo-dalities, a proven ability in high technology, anda good working relationship with the universitywith which it was affiliated. NMR expertise wasdesirable, but not necessary. CON controls werenot a consideration because they do not apply toVA installations.

Site Operations

The NMR imaging system will be under thecontrol of the hospital Chief of Staff, rather thanbeing placed in either Radiology or Nuclear Medi-

cine, This administrative decision was made tohelp foster the multidisciplinary team effort thatthe VA would like in its NMR program. Researchprotocols will be developed with input from theSt. Louis staff, the VA central office, and outsideconsultants. The VA has not yet allocated moniesspecifically for NMR research. It should be re-called that the VA does not charge its patients.The VA will be getting a small research grant (ap-proximately $75,000 per year for 2 years) fromTechnicare, the manufacturer of the VA’s unit.

Future NMR Acquisitions

In June 1983, Dr. Donald Custis, the Chief Medical Director of the VA, formed a High Tech-nology Assessment Group to determine whatcourse the VA should follow with respect to ac-quisition of major new technology such as NMRimagers (e. g., what type, how many, over whattime period).

lnvestor-Owned Hospitals

Humana 4

Background.—Humana, based in Louisville,KY, owns or operates 92 hospitals. Humana has

4The information in this section was drawn from a personal com-munication with F. D. Rollo ( 153).

8/14/2019 8420



invested considerable effort in an assessment of NMR technology over the past 2 years. It has ac-tively monitored NMR developments and dis-cussed the technology frequently with manufac-turers to help it decide what and when to buy.In addition, Humana has undertaken an interest-ing joint venture with Vanderbilt University,

Through this arrangement, Humana obtains de-tailed information from Vanderbilt regardingNMR installation and operating costs, advicefrom Vanderbilt personnel regarding importantquestions to pose to manu facturers, and data fromclinical studies. In return, Humana helps Vander-bilt obtain special consideration regarding price,software, and access to scientific and engineeringexpertise from the manufacturer(s) hoping to ob-tain a high-volume purchase agreement from Hu-mana. (Humana conducted a similar joint ven-ture with Vanderbilt before purchasing digitalradiography equipment for Humana hospitals. )

Humana is also providing a grant to Vanderbiltto assess the value of NMR in community hos-pitals.

The NMR Decision.—Although Humana hasnot made final decisions regarding what type(s)of and how many NMR systems to buy, it willprobably buy in a phased approach, beginningwith a purchase of three systems in the nearfuture. Humana feels that such an approach willenable it to conduct an in-house evaluation of NMR, yet take advantage of future progress inthe development of NMR imagers, particularly

the possibility that p ermanent m agnet systems willbecome more practical for smaller hospitals.

Humana’s decision to acquire an NMR sys-tem(s) in the near future is based more on strate-gic considerations than on a belief that NMR’sclinical role has been p roven. These considerationsare that Humana should not depend on eithermanu facturers or university hospitals to d eterminethe optimal type of NMR system and NMR clini-cal applications in community hospitals.

Choosing a Manufacturer.—Humana identifiedseven criteria that it would employ in choosinga manufacturer: corporate durability, systemquality, system reliability, how NMR informa-tion is related to the user, manufacturer agree-ments related to upgrading of a purchased sys-

.- —

tern, quality of a service program, manufacturer’sinterest in collaborating with Humana’s researchinterests, and price. Of these, price was consid-ered to be the least important and corporatedurability the most important.

Site Selection.--Humana considered three maincriteria in determining which of its hospitals wereappropriate for installation of NMR imaging sys-tems. Appropriate hospitals were considered tobe those with: 1) multispecialty practices withheavy emphasis on the neuroscience, oncology,and cardiovascular diseases; 2) high-volume out-patient services; and 3) adequate land for crea-tion of an outpatient diagnostic center that wouldinclude, but not be limited to, NMR imagingequipment. In addition to these primary criteria,consideration was also given to the existence of NMR expertise among hospital staff and to wh eth-er NMR facilities would enhance the Preferred

Provider Organization (PPO) and Health Main-tenance Organization (HMO) programs Humanais developing. Finally, since Humana intends itsNMR facilities to serve as community resources,it has sought to place them in areas with large pa-tient populations.

Using these criteria in conjunction with in-depthfinancial analyses, Humana has identified threeof its hospitals as appropriate for NMR installa-tions: one each in St. Petersburg, FL (300 beds),Louisville, KY (484 beds), and Dallas, TX (600beds). CON applications have been filed for twoof these installations (the Louisville application

was approved in September 1983), and a letterof intent has been filed for the third.

Site Operations.—Humana plans to undertakean educational program for the administratorsand medical staff of the hospitals in which theNMR systems will be installed. These programswill deal with NMR in general and with physi-cians’ use of NMR in diagnostic strategies, Ac-cess to an NMR system within a hospital will begoverned by that hospital. Hospital administra-tors may undertake studies to evaluate the impactof NMR on the cost of managing various types

of patients.Future.—Humana could purchase as many as

12 NMR imaging sys tems over the next 3 to 5years ,

8/14/2019 8420


AMI Diagnostic Services, Inc.5 6

Background.—AMI owns 90 hospitals a n dplans to build 50 to 100 freestanding diagnosticcenters that will be affiliated with physiciangroups or hospitals. When considering the acqui-sition of new technology, AMI generally tries toassess whether the new technology will replaceexisting technology, whether it will do so at lesscost, and whether it will shorten the length of in-patient stays. AMI requires an expectation of 20percent return (pre-tax) on any of its investments.

The NMR Decision.—AMI started its strategicplanning related to NMR technology in October1982. It views NMR as safe and effective, espe-cially in necrologic applications. It has questions,however, regarding the potential applicability of NMR to body imaging. AMI es t imates that in i-tial patient throughput is likely to be 10 to 15 pa-tients per day per machine. Because of continued

uncer ta in ty r egard ing w hen r e imburs emen t fo rN M R imag ing w i l l be app roved , w he ther r e im-bursement for NMR will be sufficient to cover itscosts, the safety regulations regarding siting re-quirements that S tate and Federal agencies wil limpose, and the impact a decision not to acquireNMR imaging technology will have on AMI’s pro-fessional staff, AMI has not yet decided whetheror when to acquire an NMR imaging system.

Choosing a Manufacturer.—The three mostimportant criteria identified by AMI for choos-ing a manufacturer were perceived corporate lon-

gevity, service quality, and reliability (up-time)based on experience in existing installations. Aswas the case with other hospital chains, price wasa less important consideration. Once the field isnarrowed to manufacturers satisfying these con-cerns, AMI will leave the final decision to indi-vidual hospitals and physicians. AMI does estab-lish national contracts for maintenance of itsequipment, however.

Site Selection.—AMI identified four character-istics for determining which of its hospitals wouldbe appropriate sites for installation of NMR im-aging equipm ent: bed-size (greater than 25 0 beds),

‘An American Medical International, Inc , health care subsicllary.

“ T h e lntt~rmati{~n ]n this sect]on was dra~jn trom a p(,rw>nal conv

munlcatl(~n w] th 1 Atk]n\ ( 10 I

Ch. 5—Hospit al Costs and Strat egies 71 —

patient mix (heavy emphasis on necrologic andcardiac disease), large outpatient volume, and adominant position in the community. Using thesecriteria, AMI currently considers 12 of its 90 hos-pitals to be appropriate for NMR installations andis applying for a CON in each of these cases.

Site Operations .—AMI does not intend to im-pose control over physician decisionmaking re-garding use of NMR. It does intend, however, toimplement a physician-education program per-taining to NMR and diagnostic-test-ordering strat-egy in general.

Future.—If AMI decides to purchase NMRequipment, it could purchase 50 to 100 units forits planned diagnostic centers and 12 units for hos-pitals, over a 24- to 36-month period.

National Medical Enterprises, Inc. (NME)7

NME owns, operates, or manages 339 acute,psychiatric, and long-term hospitals.

NMR.—NME began its strategic planning forNMR in October 1982. At the present time, NMEdoes not plan to budget for NMR equipment u n -til fiscal year 1986. It is maintaining close com-munication with manufacturers and with institu-t ions that have already acquired NMR devices ,however, to be aware of developments that mightlead to a change in plans.

NME decided to defer acquisition of NMR tech-nology because of its uncertainty regarding which

magnet types and field strengths would prove tobe most effective and whether separate systemswould be required to perform proton imaging andspectroscopic analysis.

Site Selection.—NME has not decided whichof its facilities would be appropriate sites forplacement of NMR technology. It did believe thatit would tend to put NMR imagers in its largerfacilities, however.

Site Operations.–NME believes that NMRtechnology in the near future will be complemen-tary to, rather than competitive with, X-ray CT.

8/14/2019 8420



Lifemark 8

NMR.—Lifemark owns or operates 30 hospi-tal facilities. Since it began assessing NMR imag-ing technology in early 1983, Lifemark has at-tempted to keep abreast of NMR developmentsand to assess the instruments manufactured byvarious companies. It has made no decision re-

garding whether or when to acquire NMR imag-ing equipment, because it would like to be fairlycertain that third-party payment is forthcomingbefore deciding to acquire NMR technology. Ithas as yet made no assessment of the likely im-pact of NMR on total patient-management costs.

Choosing a Manufacturer.—The major factorsconsidered by Lifemark in any major equipmentpurchase are corporate durability, service com-mitment, and protection against technological ob-solescence (as evidenced by manufacturers’ ongo-ing R&D programs and willingness to supply

software or hardware updates).Site Selection. --Lifemark believes that the three

hospitals it owns that have more than 300 bedsand the one 300-bed hospital it has under con-struction will be the most likely early candidatesfor NMR imaging technology. Smaller hospitalsthat have a strong neurological or neurosurgicalorientation would also be potential candidates,Lifemark’s Columbia Regional Hospital in Mis-souri, a 300-bed general medical-surgical, multi-specialty referral hospital, expressed an interestin obtaining an NMR unit over a year ago. Al-

though the hospital received CON approval inMarch 1983, no definite purchase decision hasbeen made.

Site Operations.—No definite decisions havebeen made regarding how NMR units would beutilized in hospitals. Any restrictions on NMR usewould depend on the type of payment that is ap-proved by third-parties,

Future.—Lifemark anticipates the possibility of purchasing four NMR imaging systems over thenext 3 to 4 years.

Hospital Corp. of America’

Background. -Hospital Corp. of America (HCA)owns 150 hospitals and manages 150 others. It hasan internal diagnostic-imaging technology advi-sory board that has traditionally taken a cautiousapproach to acquiring new technology. This ap-proach has often resulted in HCA’s getting new

equipment up to 18 months after other hospitals.Recently, HCA decided that it would like to beginevaluating new technology such as NMR at anearlier point in the technology’s evolution. Thisdecision is based on the need to generate infor-mation regarding the likely role, operating costs,and patient throughput for new imaging technol-ogy in community hospital settings. (HCA hasconcluded that data emanating from universityhospitals are not always applicable to their com-munity hospitals. ) In addition to recognizing theneed for this type of information, HCA believesthat it has sufficient numbers of 300- to 400-bedhospitals to be able to generate this informationinternally and that such information could helpmanufacturers obtain FDA and third-party pay-ment approval.

The NMR Decision.—With this strategy inmind, HCA has decided to purchase five NMRsystems—one 0.15 T resistive system, three 0.5 Tsuperconductive systems, and one permanent-magnet system—from four manufacturers (Tech-nicare, Picker, Philips, and Fonar), enabling HCAto evaluate several different magnets and manu-facturers simultaneously. Each of the five hospi-

tals earmarked to receive an NMR unit has ana-lyzed the likely financial impact of introducingNMR on its patient care costs.

Choosing a Manufacturer.—HCA consideredseveral factors in choosing the manufacturers: per-ceived corporate durability; maintenance capabili-ties; expected delivery time; and interest in HCA’sresearch programs, as manifested by a willingnessto supply onsite product specialists to help HCAevaluate instruments in community hospitals andto ensure that HCA obtains software updates. In

‘The information in this section was drawn from a personal com- ‘The information in this section was drawn from a personal com-munication with K. Harville (82). munication with R. Bird (16).

8/14/2019 8420


Ch. 5–Hospital Costs and Strategies 7 3

general, HCA likes to obtain equipment from atleast two, but not more than three, preferred man-ufacturers. HCA believes that such a strategyhelps ensur e against a manu facturer’s “getting lax”in service and being unable to accommodate allof HCA’s needs. Within this limited range of po-tential manufacturers, HCA permits each of its

hospitals to make its own acquisition decision.

Site Selection. --HCA considered four primarycriteria in choosing th e five sites for initial installa-tions: bed size (250 to 40 0 beds); type of hospital(acute-care hospital with a large nearby clinicalreferral base and a sophisticated emergency roomcapable of handling trauma patients); type of pa-tient-mix (with neurology, oncology, cardiology,and orthopedics emphasized); and degree of so-phistication of the hospital’s imaging department.On the basis of these criteria, HCA decided to in-stall three units in 400-bed hospitals and two units

in 250-bed hospitals. 10 Each of these hospitals haveeither applied for or are in the process of apply-ing for CON.

Site Operation.—Each NMR facility will beoperated as a separate cost center to improve thequality of financial information pertaining to theuse of NMR. NMR units will be installed withinimaging departments, which include both Radiol-ogy and Nuclear Medicine. Physician education

“’The five hospitals are Chippenham Hospital in Richmond, VA;Medical Center Hospital in Large, FL; West Florida Hospital in Pen-sacola, FL; Coliseum Park Hospital in Macon, GA; and ParkviewHospital in Nashville, TN.

programs will be prepared. The various NMR fa-cilities may have different clinical emphases,depending on manufacturer needs.

Future.—The first stage of HCA’s evaluationwill involve five or six installations. Over the next5 years, HCA could obtain as many as 25 to 50

NMR imaging systems.

Conclusions

Organizations that own or operate multiplehospitals seem to be employing two different strat-egies regarding acquisition of NMR imaging equip-ment. The first strategy is to obtain one or moreNMR imaging systems as part of an in-houseevaluation project to guide future decisions re-garding acquisition of the technology. The alter-native strategy is to defer any acquisition untiladditional information about NMR is available.

What is clear is that no one considers it advisableto make large-scale purchases of NMR imagingequipment at this time. Although all hospitals areconcerned about the impact of NMR on total pa-tient management costs, only Humana and HCAappear to have conducted a formal, patient-man-agement, cost-impact assessment. Finally, whilemany university teaching hospitals are able to usetheir prestige to obtain “favored status” from man-ufacturers, companies that operate chains of hos-pitals are able to elicit special consideration from‘manufacturers because of their buying power. TheVA could capitalize on its potential to make high-

volume purchases by following HCA’s approachof designating a small number of preferred man-ufacturers.

8/14/2019 8420


6.History of Funding for

NMR Research

8/14/2019 8420


6History of Funding for NMR Research

INTRODUCTION

Government policies related to funding of med-ical device research and development by univer-sities and manufacturers can have important im-pacts on the evolution of technology and theshape of p articular d evice indu stries. The pur poseof this chapter is to review the history of govern-ment funding for NMR research in the UnitedStates and in England and Scotland where much

UNITED STATES

National Institutes of HealthOver the past decade, the National Institutes

of Health (NIH) have supported research relatingto NMR imaging, biomedical applications of NMR relaxation-time parameters, and biomedicalapplications of NMR spectroscopy. Although NIHhas provided some funds for the development anduse of software and ancillary hardware, NIH hasnot provided, and does not plan to provide, sup-port to clinical or research institutions to be usedeither to develop or purchase NMR imaging ma-chines for use in human imaging.

Intramural Research

Conventional Spectroscopy.—NIH has had anactive intramural program of research involvingconventional chemical and physical applicationsof NMR spectroscopy for many years. A descrip-tion of these activities is beyond the scope of thisreport.

Instrumentation and Imaging Techniques. -Dr.David Hoult, a physicist and electronics engineerformerly involved with in vivo phosphorus spec-troscopic research at Oxford University, has been

at the Biomedical Engineering and Instrumenta-tion Branch of NIH since 1977. Over the past 6years, his research at NIH has focused on NMRimaging instrumentation and techniques. In ad-dition to developing the rotating frame technique

of the early work on NMR imaging was per-formed. Policy issues that emerge from this re-view are discussed in the final section of thischapter. 1

‘Readers who are not interested in the details of government fund-in g of NTMR research may want to read only the final section of

this chapter,

of NMR imaging (94) and systematically explor-ing the parameters affecting image quality, resolu-tion, contrast and signal-to-noise ratio, Hoult hasbuilt a small-bore (30 cm diameter) 0.117 tesla (T)NMR imager. The imager as yet has been usedonly to image phantoms, but may be used forstudies of animals or newborns in the future.

Physiology.--Dr. Robert Balaban, Senior Staff Fellow at the National Heart, Lung, and BloodInstitute, has been studying physiological applica-tions of phosphorus, sodium, and nitrogen NMRspectroscopy for the past 3½ years. In his researchat NIH Balaban has employed in vivo 31P NMR

spectroscopy to measure the phosphorus contentof cardiac muscle under varying physiologicalconditions, has measured the kinetics of metabolicreactions such as the transfer of phosphorus fromadenosine triphosphate to creatine phosphate, andhas studied sodium transport across plasma mem-branes and the concentration of various nitrogen-containing buffers in various tissues in the body.

Clinical Research. —Using funds contributedfrom many of its Institutes, NIH has purchaseda 0.5 T whole-body NMR imaging system fromPicker International. This system has been deliv-

ered to NIH and clinical studies were scheduledto begin in September 1983. This system, whichwas s la ted to be p laced in the Depar tment of Radiology, wil l be u t i l ized by several Ins t i tu testo car ry out research protocols approved by an

77

8/14/2019 8420


78 • Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology; A Clinical, Industrial, and Policy Ana lys is

NIH NMR Users Committee. Initial studies willbe done at a magnetic field strength of 0.15 T andwill include investigations of demyelinating dis-ease, the effects of chemotherapy and radiationtherapy on NMR parameters, and whether NMRcan be used to predict patients’ responses tochemotherapy and radiation therapy.

Cellular Metabolism.—A fifth NMR researchgroup has been formed at N IH by Charles Meyers,Chief of the Clinical Pharmacology Branch of theDivision of Cancer Treatment at the NationalCancer Institute (NCI). This group will be explor-ing the use of NMR in the study of the metabo-lism of both normal and cancer cells, as well asthe effect of various drugs on cellular metabolism.The group will also be exploring possible applica-tions of NMR to the study of the developmentof tumors (37).

Meyers’ group will be drawing on the exper-

tise of Dr. Jack Cohen, an NIH veteran of 15years, whose research has focused on biochemicalapplications of NMR. Over the past 3 years ,Cohen has used NMR spectroscopic techniquesto study cellular metabolism (67), proteins, DNAconformation (34), and drug binding, includingthe binding of the chemotherapeutic agent Adria-mycin to specific DNA sequences. Cohen has re-cently developed a method to perfuse living cellsin an NMR spectrometer (68), which he has usedto study ATP metabolism in mammalian cells(Chinese hamster lung fibroblasts).

Extramural Research: Past

NIH-Supported NMR Spectroscopy ResearchFacilities.—Although a complete description of NIH-supported, conventional NMR spectroscopicresearch is beyond the scope of this report, men-tion should be made of a number of NIH-sup-ported NMR-spectroscopy research facilitiesaround the United States that are devoted to thestudy of biological molecules. These include theMiddle Atlantic NMR Research Facility at theUniversity of Pennsylvania, the Western RegionalNMR Biomedical Facility at the University of

Utah, the Stanford Magnetic Resonance Labora-tory, the Purdue Biochemical Magnetic ResonanceLaboratory, the NMR Facility for BiomedicalStudies in Pittsburgh, and the Francis Bitter Na-

tional Magnet Laboratory at the MassachusettsInstitute of Technology (190). The MIT NationalMagnet Laboratory, run by Dr. Leo J. Neuringer,is a national resource that makes available highfie ld NMR spectrometers to biomedical in-vestigators. 2

NMR Imaging.–Although a few of the NMR-

related extramural grants that have been fundedby NIH over the past decade have been fundedby the National Heart, Lung, and Blood Institute(NHLBI) (e.g., research by Lauterbur in 1975related to the use of NMR imaging to study bloodflow and about $200,000 per year provided toLauterbur by NHLBI since 1978) and other Insti-tutes, most of them have been funded by NCI.The first extramural NCI grant related to NMRimaging was awarded to Lauterbur at SUNY-Stony Brook in 1973 after publication of his land-mark article (115). The award was made to helpLauterbur further develop his technique of NMRimaging and investigate its application to cancerresearch. His initial funding of approximately

$100,000 per year for 3 years has been renewedat an approximately constant level, without in-terruption, since 1973. NIH also supported earlywork on T1 measurements of surgically excisedhuman tumors (45,46) and tumors in mice andrats (85,92), as well as on the imaging of tumorsin live animals (44).

Extramural Research: Present

NIH is currently funding ap proximately $2 mil-

lion of research relating to NMR imaging and/ orin vivo spectroscopy in at least 10 different insti-tutions (204). A complete description of the con-tent of this research is beyond the scope of thisreport. The Department of Energy has awardedan add i t iona l $1.8 million for NMR-related re-search (204). It should also be noted that in 1983,NIH began providing support for innovative re-search performed by small businesses in a pro-gram similar to the one sponsored by the N ational

‘Of interest is the fact that IBM has recently supplied several mil-lion dollars to the Brigham & Women’s Hospital in Boston, the

University of California at Berkeley, and the MIT National MagnetLaboratory for a joint research program aimed at addressing basicbiological and medical questions and developin g a high-field, whole-body magnet with sufficient field homogeneit y to permit perform-ance of in vivo “P spectroscopy (1).

8/14/2019 8420


—. — —

Science Foundation (197). No information wasavailable regarding wh ether supp ort has been pro-vided for NMR research under this program.

Extramural Research: Future

The dramatic advances over the past decade indiagnostic imaging technologies, many of whichcan assist clinicians in diagnosing multiple typesof pathologies, have created the need for com-parative performance studies to clarify how ourexpanding d iagnostic imaging arsenal can be u sedmost effectively and efficiently. In October 1981,the NCI expanded its focus beyond supporting re-search directed principally at cancer detection anddiagnos is by forming a Diagnos t ic Imaging Re-search Branch to advance the art of imaging alltypes of morphologic and functional pathologies.

In October 1982, this Diagnostic Imaging Re-search Branch announced a solicitation of pro-

posals (191) for the performance of studies: 1) toexplore and define the current and potential use-fulness of present-day NMR imaging systems inclinical applications; 2) to establish optimal im-aging conditions for NMR use in specific clinicalproblems; and 3) to carry out comparative per-formance and evaluation studies to determine thecapabilities and limitations of NMR imaging sys-tems in comparison with other modalities for clin-ical applications in human subjects in the detec-tion, imaging, quantification, and diagnosis of morphological and functional pathology and inthe noninvasive characterization of tissue (191).

The other techniques with which NMR is to becompared could include, but are not limited to,conventional X-ray imaging, computed tomogra-phy (CT), ultrasonic imaging, and radioisotopeimaging, including single photon emission com-puted tomography (SPECT), and positron emis-sion tomography (PET). Awards are to be for 3years and are in the form of cost-reimbursementcontracts. Announcements of awards, originallyscheduled for June 1983, were made in mid-1984(144).

There were three minimum requirements forqualification for an award: award recipients mustown or have available a working NMR-imagingsystem of sufficient size and image quality formeaningful whole-body or head studies of human

Ch. 6-–History of Funding for NM R Research 79 — — — — — - - . . — . — — .

subjects; recipients must have access to equipmentto carry out comparative imaging with one ormore of the other techniques listed previously; re-cipients must be capable of imaging at least 150patients in the first year and at least 200 patientsin each of years 2 and 3 (191).

NIH is also planning to issue a Request forApplications for grants to study the physical,chemical, and biological bases for T1 and T2 relax-ation times to gain further insight into the infor-mation implicit in these parameters. Although a3-year program, with up to $1 million in grantsin the first year, is being considered, no funds haveactually been approved (144).

NIH is also considering the possibility of fund-ing a small nu mber of training fellowships in N MR(144).

National Science Foundation

The National Science Foundation (NSF), an in-depend ent agency of the Federal Government thatsupports basic research in 18 different scientificsubject areas, supported a pioneering researchproject in NMR imaging through its Research Ap-plications Directorate (Instrumentation Technol-ogy Program) during 1977-793 as well as a 3-year($95,000) project by Lauterbur related to micro-scopic NMR imaging. NSF currently has no pro-grams that support research in NMR imaging perse. However, NSF has provided, and continuesto provide considerable support for investigations

relating to digital signal processing, two- andthree-dimensional analyses and image reconstruc-tion, as well as other scientific principles on whichNMR imaging is based (49).

NSF Regional Instrumentation Facilities

In 1978, NSF initiated a program designed toimprove the quality and scope of research con-ducted in the United States by making sophisti-cated instruments broadly available to research-ers in both academia and industry (196). Thisprogram of Regional Instrumentation Facilities

was predicated on two beliefs. First, there wasthought to be a growing need for researcher ac-

‘There was $309,500 awarded under grant FApR-7708185, “In-L’ivo Nuclear Magnetic Resonance of Flow Patterns” to the Un]\ ’er-

sity of California at Berkeley in 1977.

8/14/2019 8420


80 • Health Technology Case Study 27: Nuclear Magnetic Resona nce imaging Technology : A Clinical, Industrial, and Policy Analysis —

cess to the powerful scientific instrumentation thathad evolved out of recent advances in electronics,but which unfortunately was affordable by onlya few research institutions. Second, it was thoughtthat a program of instrumentation sharing mightencourage interaction and cooperation betweenresearchers from scientific disciplines that do not

ordinarily collaborate with one another (150).By November of 1979, NSF had awarded a total

of $11,392,000 to eight universities in seven Statesfor the establishment of such regional instrumen-tation facilities (196). Of the 14 grants awardedduring the first 2 years of the program, 5 wentto universities to establish facilities for the per-formance of high-resolution NMR spectroscopy,at a cost of over $5 million.4 Researchers in theseinstrument facilities used the NMR spectrometersin a broad range of physical, chemical, biochem-ical, biophysical, and molecular biologic experi-ments. No NMR imaging instruments were in-stalled in the facilities and no imaging researchwas performed. No information is available re-garding the extent to which the spectroscopic re-search performed in these facilities was pertinentto the later development of NMR imaging tech-niques.

NSF also provides institutional support to theNational Magnet Laboratory at the MassachusettsInstitute of Technology.

Small Business Innovation Research Program

Although NSF directs most of its research sup-port to basic scientific and engineering projectsat academic institutions in the United States, since1977 it has also operated a Small Business Innova-tion Research (SBIR) program. This program en-courages science-based and high-technology smallbusinesses with strong research capabilities in ap-plied science, basic science, or engineering to sub-mit proposals pertaining to research in scientificor engineering problems that could lead to sig-nificant public benefit (197). Three primary objec-tives of this program are to stimulate technologi-.———

4

NMR spectroscopy centers were established at Colorado StateUniversity, the University of South Carolina, the California Insti-tute of Technology, the University of Illinois, and Yale University.

cal innovation in the private sector, to increasethe commercial application of NSF-supported re-search results, and to increase the national eco-nomic and social benefits derivable from Federalresearch investments (197).

Awards under the most recent SBIR programsolicitation were to have been announced in Jan-uary 1984. Research proposals pertaining to:1) the application of superconductivit y to elec-tronically oriented industries; 2) improved NMRprobes; and 3) new procedures for NMR datadisplay have been particularly encouraged. NSFexpects to award about 100 Phase I (feasibilitystudy) grants of up to $35,000 each. One-thirdto one-half of Phase I awardees will receive PhaseII grants, which will cover 2 years of research andare expected to average $200,000 each.

NSF and ACR Research Workshop

In November 1982, NSF and the American Col-lege of Radiology (ACR) sponsored an engineer-ing research workshop for engineers, physicists,computer experts, physicians, product developers,and business managers to identify critical gaps inimaging knowledge that require engineering sup-port and to examine how engineering research-ers, medical scientists, and clinicians could de-velop effective collaborative relationships out of which advances in radiological imaging mightemerge (49).

In addition to identifying areas in which basic

imaging research is needed, the conference at-tendees addressed the current research and devel-opment roles of various elements along the re-search-development-production-application con-tinuum. They concluded that the medical imag-ing-system manufacturers perform little of thebasic research on which advances in imaging tech-nologies depend:

. . . The instrument producer functions largely asa designer/ integrator of high technologies intofunctional systems, taking the knowledge result-ing from research done elsewhere and convertingit into clinically useful imaging systems (49).

The need for additional sources of support forbasic research was underscored.

8/14/2019 8420


.

——— .-

—

ENGLAND AND SCOTLAND

Between 1973 and 1983, at least three differentnoncommercial entities provided financial supportfor NMR research in England and Scotland. Be-tween 1974 and 1979, the Medical Research Coun-cil, the British analog of our NIH, provided sup-port to three different groups. Two of these groupswere based at the University of Nottingham, oneunder the direction of Professor Peter Mansfield,the other under Professor Raymond Andrew. Thethird group was based at the University of Aber-deen in Scotland, under Professor John Mallard.Andrew’s and Mallard’s groups received supportfrom the Medical Research Council between 1975an d 1978; Mansfield’s group received support be-tween 1976 and 1979.

In late 1977, the Wolfson Foundation, a privatephilanthropic organization in England, announced

the availability of funds for medical research. An-drew’s group at the University of Nottinghamused funds received from the Wolf son Founda-tion between 1978 and 1980 to do NMR imagingresearch, including construction of a whole-bodyNMR imaging sys tem.

Final ly , the Depar tment of Heal th and SocialSecurity (DHSS) in England, through a program

des igned to suppor t development of technology

that might be of use in hospitals, also providedsignificant financial support for the developmentof NMR imaging technology. In 1976, DHSS pro-vided funds to help establish an NMR researchteam at the London-based company EMI (29). Aresistive magnet-based machine was built at EMIand head images were produced in 1978 (29). Fol-lowing 2 additional years of developmental re-search, the EMI team constructed a superconduct-ing, whole-body NMR imaging system, whichwas installed in Hammersmith Hospital in Lon-don in March 1981 (29). EMI also contributedfunds to this development project. ’

In 1981, NMR imaging began at the Hammer-smith Hospital using the resistive NMR unit de-veloped by EMI. DHSS apparently has also pro-vided support to the clinical imaging program at

Hammersmith , as wel l as suppor t for NMR re-search to Professor Mansf ie ld a t the Univers i tyof Nottingham (177).

——

‘In approximately 1977, GEC of England independently beganwork on the development of an NMR imaging system. In April 1Q81 ~€

GEC acquired the Picker Corp. and formed Picker International byconsolidating the Picker Corp., GEC Medical, and Cambridge Med-ical Instruments. In October 1Q81 Picker Intern atlona] I.td. Divi-sion in England purchased EMI’s NTMR technology and program,]ncluding its program at Hammersmith Hospital

RESEARCH AND DEVELOPMENT POLICY ISSUESCertain differences between the history of the

development of NMR imaging in the United Statesand Great Britain should be considered. In con-trast to the NSF program’s supporting basic re-search by small businesses on a broad spectrumof scientific and engineering problems (as opposedto product development research), the BritishGovernment attempted to develop technologythat might specifically be of use in hospitals. Thiseffort was focused through a program funded byDHSS that lent considerable financial support to

the development of NMR imaging techniques.Comparatively little support for the specific de-velopment of NMR imaging technology was pro-

vided by the U.S. Government. However, onceit became apparent that the development of NMRimaging systems was not only commercially vi-able, but also potentially extremely profitable,U.S. manufacturers rapidly and intensively beganinvesting in N MR imaging d evelopment programs.In addition to applying their considerable elec-tronics and engineering resources to this effort,U.S. manufacturers aggressively (and successfully)recruited scientists who had been actively in-volved in the early NMR development programs

in British universities. Similarly, U.S. universitiesand hospitals have recruited British NMR scien-tists to help initiate and promote active NMR im-

8/14/2019 8420


82 • Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy A n a l y s i s

aging research and development programs, con-tributing to what has become somewhat of a“brain drain” from Britain.

Finally, in Britain there seem to have been sev-eral groups of physicists based in university set-tings who had background and interest in the de-

velopment of NMR imaging techniques and whointeracted and collaborated with medical research-ers, clinicians, and each other. In the UnitedStates, in contrast, most of the early work onNMR imaging was done by Lauterbur and Dama-dian with apparently little, if any, interaction be-tween the two, despite the fact that both were at

SUNY campuses. In addition, there seem to havebeen fewer centers in the United States in whichscientists with varied backgrounds collaboratedon the type of interdisciplinary research thatresulted in the NMR imaging advances in Britain.Perhaps the collaborations between physicists andphysicians that are now developing as part of

hospital- and university-based NMR imaging pro-grams, as well as the recognized need for inter-disciplinary research collaborations that emergedfrom the 1982 NSF and ACR research workshop,will change this situation in the future.

8/14/2019 8420


7.Regulation by the Food and

Drug Administration

8/14/2019 8420


7.

Regulation by the Food andDrug Administration

INTRODUCTION

NMR imaging devices are the first imaging de-vices for which the Food and Drug Administra-tion (FDA) premarket approval has been requiredunder the Medical Device Amendments of 1976.(Previous imaging devices, such as X-ray CT scan-ners, were introduced before the amendmentswere passed. ) Because the FDA approval processcan have such an important effect on the rate atwhich new technology is introduced into thehealth care system, it is important to examine how

the FDA regulatory process operates and how apromising technology such as NMR imaging hasfared in its encounter with it. This chapter isdevoted to those two tasks,

SOURCES OF FDA AUTHORITY

Radiation Control forHealth and Safety Act

FDA authority over NMR imaging devices de-

rives from two Federal acts. The first is the Radi-ation Control for Health and Safety Act of 1968,established to protect the public from hazardousradiat ion emit ted by electronic devices . Becauseno hazards deriving from electromagnetic fieldshave been identified with current NMR devices,FDA has not established any radiation emissionperformance standards for NMR devices underthe authority of the Radiation Control Act.According to the Director of Electronic Productsin FDA’s Center for Devices and RadiologicalHealth, “. . . it is not likely that the RadiationControl Act will have any significant impact on

the development of NMR imaging as a medicaldiagnostic modality” (164). However, if defectswere found in NMR devices that rendered themunsafe , FDA could use the author i ty of the act

The chapter is divided into three sections. Thefirst section describes the statutory sources of FDAauthority over NMR imaging devices. The secondsection describes the process through which newdevices such as NMR imagers are brought to m ar-ket. The section includes a flow diagram (fig. 12)that illustrates the p rocess, and a summ ary of howNMR imagers have fared in each stage of the ap-proval process. The final section assesses thepremarket approval process as a whole and raises

a nu mber of policy issues that should be add ressedin evaluating it.

to recall them, even thou gh no performance stand-ards have been developed.

Food, Drug, and Cosmetic ActThe second source of FDA authority over NMR

imaging devices is the Food, Drug, and CosmeticAct as amended in 1976, which controls the intro-duction of medical devices into commerce. In con-trast to the Radiation Control Act, the Food,Drug, and Cosmetic Act has had, and continuesto have, a significant impact on the developmentof NMR imaging devices. The 1976 Medical De-vice Amendments require that all medical devicesin tended for human use be class i f ied in to threeregulatory categories (classes) based on the extentof control necessary to provide reasonable assur-ance of the safety and effectiveness of each device.

Class I devices are those for which generalcontrols relating to adulteration, misbranding,

85

8/14/2019 8420


86 Health Technology Case St udy 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy A naly sis

banning, notification, reporting, registration,restrictions on sale or distribution, and good man-ufacturing practices are considered sufficient toprovide reasonable assurance of safety and effec-tiveness.

Class II devices are th ose for wh ich general con-

trols are considered insufficient to provide rea-sonable assurance of safety and effectiveness, andfor which there is sufficient information to es-tablish a performance standard to provide suchassurance.

Class III is reserved for devices: 1) that are u sedin supp orting or sustaining hum an life, are of sub-stantial importance in preventing impairment of

human health, or present a potential unreasonablerisk of illness or injury; and 2) for which ClassI and Class II controls are either insufficient toprovide reasonable assurance of safety and effec-tiveness or for which insufficient information ex-ists to establish a performance standard thatwould provide this assurance. Class III devices re-

quire premarket approval (PMA) from FDA.

Figure 12 provides a flow diagram that illus-trates how a new m edical device, such as an NMRimager, finds its way through the FDA processinto commercial distribution. The following sec-tions describe the illustrated process in moredetail.

REGULATION OF NEW MEDICAL DEVICES UNDER THE

FOOD, DRUG, AND COSMETIC ACTPremarket Notification of Intent ToMarket a New Device

The Medical Device Amendments of 1976 in-clude a provision titled Premarket Notification(sec. 510(k)), which was designed to ensure thatmanufacturers did not begin marketing new de-vices until such devices had either receivedpremarket approval or had been reclassified intoClass I or II. Under this provision, a manufac-turer must notify the FDA 90 days before it in-

tends to begin marketing a device that was notsold prior to May 28, 1976. At the time of thisPremarket Notification, the manufacturer mustalso specify the class into which the device hasbeen classified or state the fact that the device hasnot yet been classified.

Under the 1976 amendments, any new deviceis automatically classified into Class III unless itis deemed to be “substantially equivalent” to eithera preenactment device (i. e., one introduced priorto May 28, 1976) or a postenactment device thathas already been classified into either Class I orClass 11. ’

‘If the new device is deemed to be “substantially equivalent” toa preenactment device, then the new device automatically assumesthe classification of that preenactment device. Of significant impor-tance is the fact that if a new device is deemed to be substantiall y

equivalent to a preenactment Class III device, it may immediatel y

be marketed without a premarket approval application.

To our knowledge, no NMR imaging manufac-turer has su bmitted a petition to FDA arguing thatNMR imaging devices are substantially equivalentto either a preenactment device or a postenact-ment Class I or Class II device. NMR imagingdevices thus are Class III devices that unlessreclassified, cannot be marketed prior to approvalof a premarket approval application (PMAA) ora Product Development Protocol before market-ing (see below).

Getting a Class Ill Device to MarketAs ind icated in figur e 12, a Class III device such

as NMR imagers can be brought to marketthrough one of several pathways: reclassificationinto Class I or II, initiated either by a petition orFDA; premarket approval; or Product Develop-ment Protocol.2 Although NMR manufacturersconsidered reclassification, they have used the pre-market approval approach.-——

‘Under a Product Development Protocol, a manufacturer and FDAwould agree on a plan of stud y to demonstrate reasonable assuranceof the safety and effectiveness of a device. After receipt of a noticeof completion of an approved protocol, FDA may declare the pro-

tocol completed or find that the results of the trials performed underthe protocol differ substantially from the results required by the pro-tocol, or that the results do not provide reasonable assurance of the safety and effectiveness of the device under the conditions of use in the proposed labeling. At least until December 1983, no man-ufacturer had elected the approach of a Product DevelopmentProtocol.

8/14/2019 8420


m

n

UJ

8/14/2019 8420


88 Ž Health Technology Case St udy 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial , and Policy Ana lys is —

Reclassification

According to the Medical Device Amendmentsof 1976, a manufacturer may petition FDA toreclassify a Class III device into either Class I orII. Reclassification petitions are referred to an ex-pert advisory panel that within 90 days must rec-

ommend to FDA whether classification of the de-vice in Class 111 is required to provide reasonableassurance of the device’s safety and effectiveness.Within 90 days of receipt of the panel’s recom-mendation, FDA must either approve or deny thepetition.

On July 6, 1982, the National Electrical Man-ufacturers Association (NEMA), a trade associa-tion representing 13 companies involved in thedevelopment of NMR imaging systems and mag-nets, requested a meeting with FDA to discuss thepossibility of initiating the reclassification proc-ess. At a meeting in December 1982, NEMA of-

fered the view that NMR was an anatomical im-aging modality whose safety and effectivenesswere adequately assured by the General Controlsof Class 1. FDA expressed concern that NMR wasa rap idly d eveloping technology w hose safety andeffectiveness had yet to be demonstrated. Accord-ing to the Director of the Division of ElectronicProducts in the Center for Devices and Radiologi-cal Health:

The clinical possibilities for NMR imaging andthe immaturity of its current applications werefactors behind the FDA’s opinion that Class III

is appropriate for the modality. NMR’s promise,while immense, is still unrealized. Clinical experi-ence is still inadequate to establish effectivenessof specific NMR applications and to permit thedevelopment of adequate labeling, indications,techniques, and instructions. Each area of re-search will have to be studied scientifically andclinically to develop this information (164).

The official minutes of the December 8, 1982,meeting state:

There were a nu mber of concerns that the da tapresented left a clear impression that industry has

not substantiated a general claim th at NMR was

an effective device that could be ut ilized acrossthe spectrum as an imaging m odality. It was fur-ther stated that it w ould be ad visable for industry

to start with a limited claim on the effectivenessof NMR with supporting scientific documentation. . . (It was further stated) that the Panel wouldreview a reclassification petition if it were sub-mitted in the appropriate legal manner (140).

No reclassification petition had been submittedas of December 1, 1983.

The NMR imager manufacturers that we sur-veyed were divided over whether it was redun-dant or wasteful to require that all manufacturersobtain PMA (see below). About half of the man-ufacturers felt that NMR imaging was sufficientlygeneric that once a PMAA was approved, FDAshould set performance standards rather than con-tinue the PMA requirement for each device. Theother half of the manufacturers felt that NMR im-aging is not “generic” because important differ-ences exist between available NMR imaging sys-tems, and that, consequently, each manufacturer

should be required to obtain PMA (see below),These manufacturers argued in addition that thePMA process serves an important quality-assur-ance function and should be applied to all manu-facturers. Although we think it is important toidentify these two viewpoints, we lack sufficientinformation to comment on either the extent towhich different manufacturers’ NMR imaging sys-tems do, in fact, differ, or the extent to whichmanufacturers’ opinions reflect, in part, how closethey are to obtaining PMA.

The Premarket Approval Process

In order to obtain premarket approval for a de-vice, a manufacturer must provide reasonable as-surance that the device is safe and effective underthe conditions of use prescribed, recommended,or suggested in the proposed labeling. NMR im-aging devices are the first imaging devices to haveencountered this process. The following sectionsexplain how the PMA process works and describehow FDA has applied the process to NMR imag-ing devices.

Significant Risk Devices

A m anufacturer may p lace a device at an inves-tigational site to collect data pertaining to safety

8/14/2019 8420


— .

Ch. 7—Regulatio n by the Food and Drug Adm inist ration 89

and effectiveness. Such data can be collected ac-cording to a plan approved either by a local In-stitutional Review Board (IRB) (a local commit-tee that reviews proposed scientific studies) orFDA. If the IRB approves the investigation as in-volving a nonsignificant risk device (one that doesnot pose serious risk to experimental subjects), theinvestigation is deemed to have an approved In-vestigational Device Exemption (IDE) and an ap-plication to FDA is not necessary. A sponsor needapply to FDA for approval to conduct clinicalstudies under an IDE only if an IRB has deter-mined a device to be a “significant risk device, ”i.e., a device that presents a potential for seriousrisk to the health, safety, or welfare of experi-mental subjects. FDA does not maintain recordsof nonsignificant risk investigations and may noteven be aware that such investigations are beingconducted.

On February 25, 1982, FDA issued “Guidelinesfor Evaluating Electromagnetic Risks for Trials of Clinical N MR Systems” (188). The guidelines wereissued by FDA to “provide assistance to sponsorsof clinical investigations, researchers, and IRBsin d etermining when a clinical stud y involving anNMR device might represent a ‘significant risk’under the Investigational Device Exemption . . .and to prevent submission of IDE applications [toFDA] when they are not necessary” (188). Theguidelines related to details of NMR imaging thatthe FDA believed would be least familiar to theIRBs:

On th e basis of current information the bureaubelieves that a study which does not exceed these

guidelines probably d oes not present an unaccept-able

1.

2.

3.

risk in these ‘three areas:

Static (Direct Current) Magnetic Fields—wh ole or partial body exposu res of 2 tesla(T).

Time Varying Magnetic Fields—whole orpartial body exposures of 3 T/ second.Radiofrequency Electromagnetic Fields—exposu re to RF fields th at resu lt in a sp ecific

absorption rate that exceeds 0.4 W/ kg asaveraged over the whole body, or 2 W/ kgas averaged over one gram of tissue, Studiesthat expose pa tients above these gu idelinesshould be considered to pose “significantrisk” (188).

It should be emphasized that in issuing theseguidelines, FDA did not declare that NMR imag-ing was “safe. ” Rather, it stated that it was rea-sonable to proceed with investigations thatadhered to these guidelines.

Ten month s later in a December 28, 1982, mem o

to NMR manufacturers, the Director of FDA’s Di-vision of Compliance stated that “. . . over thepast few months it has become clear that the in-tention of the guidelines has been widely mis-understood, ” and that “. . . it has been ratherwidely reported to us that the guidelines have beeninterpreted as limits for p atient exposures in N MRimaging investigations . . . This is not our intent”(189).

In an effort to clarify this misunderstanding,FDA stated that:

It continu es to be our view that w ithin the con-

text of the p resent un derstand ing of the biologiceffects of electric and magnetic fields, the med i-cal NMR imaging devices currently in investiga-tional use span a range from those which requireno d etailed an alysis to demon strate that they donot m eet the definition of significant risk to thosewh ich d o require analysis to make such a d eter-mination. It is the purpose of the guidelines toprovid e some simple criteria for u se in establish-ing that dem arcation. No implication shou ld betaken that a device which exceeds the gu idelinesshould necessarily be considered a significant risk device (189).

Our survey of NMR imager manufacturers re-vealed that most manufacturers found FDA’s pro-mulgation of “Significant Risk Guidelines” to havebeen quite helpful, but expressed two strong con-cerns. First, although FDA has clearly describedits guidelines as an aid to IRBs in making signifi-cant risk determinations and not as limits govern-ing patient exposures, simply by virtue of theirexistence, the guidelines have inevitably tendedto become product specifications with which man-ufacturers are loath not to comply. Second, themanufacturers felt that the FDA guidelines were“poorly conceived. ” For example, manufacturers

suggested that the 3 T/ second gu ideline pertain-ing to time-varying magnetic fields was uninter-pretable in the absence of a specified pulse dura-tion. (The National Radiologic Protection Board

8/14/2019 8420


90 • Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and P o l i c y A n a l y s i s

in Great Britain, for example, suggested a guide-line of 20 T/ second for p ulses more than or equalto 10 millisecond in its 1980 guidelines. )

These two concerns relate more to a disagree-ment over the content of FDA’s guidelines thanto their issuance per se. It seems appropriate for

FDA to have issued the guidelines. If their con-tent is deficient (and we are not in a position toevaluate that issue), scientific experts could helpto change them, and the process through whichthey w ere established could be reviewed to assurethat it provides for adequate expert scientificinput.

Regulations Pertaining to Investigational Devices

During the period in which a device is consid-ered to be investigational (i. e., while it is beingassessed under an IDE), manufacturers and inves-

tigators must comply with four regulatory pro-hibitions:

1.

2.

3.

4.

I n

t w o

they may not engage in promotion or testmarketing of the investigational device (21CFR 812.7(a));they may not commercialize an investiga-tional device by charging the subjects orinvestigators more for it than the amountnecessary to recover costs of manufacture,research, development, and handling (21CFR 812.7(b));3

they may not prolong an investigation of aClass III device beyond the point where it

has become apparent that premarket approv-al cannot be justified (21 CFR 812.7(c)); andthey may not represent an investigational de-vice as being safe or effective for the pur-poses for which it is being investigated (21CFR 812.7(d)).

our survey of NMR imager manufacturers,issues related to IDE regulations were iden-

3According to an undated policy statement issued by the Officeof Radiological Health’s Division of Compliance, “. . . investigatorsmay charge a patient their normal physician’s fee and the cost of scanning the patient, provided the scanning costs do not include

a profit. ” The letter continues, however, that sec. 50.25(b)(3) of theinformed consent regulation requires that “any additional cost tothe subject that may result from participation in the research be in-cluded in the consent form where appropriate. ” FDA agrees thatit is appropriate in this situation, since third-part y re imbursemen tmay not occur. (141).

tified. First, a number of manufacturers com-plained that other manufacturers were not ad-hering to the spirit of the IDE prohibitions onproapproval promotion and test-marketing of NMR imaging devices. Such behavior, these man-ufacturers asserted, created a situation in whichall manufacturers were forced to either test-marketand promote, or suffer while following the law.One manufacturer said that to avoid this situa-tion in the future, FDA should limit the numberof research sites in which manufacturers are per-mitted to install investigational devices. In a fewinstances, such as Neodymium YAG (YttriumAluminum Garnet) lasers, FDA has establishedguidelines for the numbers of patients and thelength of followup required in studies of investiga-tional devices (50). Other manufacturers, how-ever, were pleased that FDA was not being morestrict in its enforcement activities. They consid-ered the existing situation to be an acceptablecompromise between prohibitions and no pro-hibitions.

The second IDE issue raised by our survey per-tains to the prohibition on making a profit frominvestigational devices. Although manufacturersvoiced a preference for being able to make a profitduring the IDE stage, most thought the existingprohibition was logical and reasonable in con-cept. 4 Furthermore, they suggested that, in thecase of “high R&D cost” devices such as NMR im-agers, it is difficult to recoup R&D expenses dur-ing the IDE stage because of the small number in-

stalled.

In a recent article, Anthony Young, a Wash-ington, DC, attorney, concurred with the viewthat the IDE regulations should not present aproblem to device manufacturers:

Existing regulations allow a manufacturer tocharge for investigational devices and thus to de-fray a portion of the expense involved in bring-

40ne manufacturer felt that the existing prohibition on profitmak-ing from proapproval devices disproportionatel y hurt small manu-facturers. Small manufacturers, it was argued, generally are depend-ent on external sources of capital to fund research and development

and do not fare well in their quest for funds when they are pro-hibited from demonstrating a profit. Others argued, in contrast, thatwhat is required for success in external capital markets is profit-making potential, rather than profits themselves. It would seem thatwithout the prohibition on profitmaking during the IDE stage man-ufacturers would have less incentive to apply for PMA.

8/14/2019 8420


Ch. 7—Regulation by the Food and Drug Adm inist ration 91 —

ing a new device to market. There is sufficientlatitude in the regulations concerning publicity of

availability of the device that the man ufacturercan reach those p ractitioners who will eventuallybecome customers. A manufacturer who is straight-forward in h is claims for his device and w ho does

not attemp t to circum vent regulations in an at-

tempt to get a jum p on comp etitors should hav eno problems with FDA (206). 5

Clinical Studies

With an IDE in hand, manufacturers may con-duct clinical studies to substantiate the safety andeffectiveness of the devices they propose to mar-ket, Under Federal regulations, well-controlled in-vestigations are the p rincipal m eans used to estab-lish the effectiveness of a device. However,according to the committee report accompany-ing the Medical Device Amendments, FDA isauthorized to accept meaningful data developed

u n d e r p ro c e d u re s l e s s r ig o ro u s th a n we l l -controlled investigations when well-documentedcase histories assure protection of the public healthor when well-controlled investigations would pre-sent undue risks for subjects or patients. This pro-vision is not intended to authorize approval onthe basis of anecdotal medical experience with thedevice or unsubstantiated opinion as proof of ef-fectiveness (183).

During the proapproval period, FDA realizedthat manufacturers were concerned about how toestablish the safety of NMR. FDA responded to

this concern by exempting manufacturers fromresponsibility for submitting data on electromag-netic interactions in their PMA applications.FDA’s actions and rationale were summarized byMr. Schneider of the Center for Devices andRadiological Health:

After considering th e biological interactions of the fields used in NMR imaging, [FDA] conclud edthat the existing fundamental scientific uncertain-

ties could not be resolved by experiments nor-mally associated with d evice evaluation.

In fact, it would probably be economically im-practical for any individu al sponsor to assum e the

financial burden of supp orting the research nec-essary to make significant progress in eliminat-

5Mr. Young’s statement should in no way be construed as repre-sentative of FDA’s viewpoint.

ing these uncertainties. Further, it w ould seem un-wise, with resp ect to societal benefits, to suspen dthe development and dep loyment of NMR imag-ing as a medical diagnostic modality pendingsubstantial improvement in the understanding of the biological interactions of radiofrequency elec-tromagnetic fields and static magnetic fields.

From available information, no immediateacute effects are expected from exposure condi-tions prevailing in the devices under investigation.Further, it seems that whatever risks maybe asso-

ciated with these exposures will be small com-pared to the potential medical benefits of themodality. [Therefore], potential sponsors havebeen advised that they need not subm it experi-mental data on electromagnetic biological interac-tions as part of the safety component of a pre-market approval application. Each sponsor wasasked to provide an assessment of the physical ex-

posure conditions in its device. The FDA will con-du ct a continu ing review of the risk potential of

these exposures in light of developing scientificknowledge (164).

The Premarket Approval Application

When a manufacturer believes it has collectedsufficient data to establish the safety and effec-tiveness of its device, it submits a premarket ap-proval application (PMAA) to FDA. The PMAAmust include:

1. a statement of the components of the device;2. a statement of the principles of operation of

the device;

3. a description of the methods u sed in the m an-ufacture of the device;4. a summary of investigations and informa-

tion bearing on the safety and effectivenessof the device under the proposed conditionsof use; and

5. specification of the claims, indications, andinstructions with which the manufacturerproposes to label the device.

The type and breadth of the claims made in theproposed device label determine, in part, the scopeof the research that must be performed prior tosubmission of a PMAA. As Schneider of FDA ex-plains:

Each claim that is mad e mu st be supp orted by

adequate scientific and clinical research. Thismeans th at in broad ening the ran ge of claims for

8/14/2019 8420



a device, a sponsor increases the expense and ef-fort necessary to secure premarket approval (164).

With regard to NMR, Schneider has stated that:

. . . there is a natural temptation to be enthusi-astic about all possible applications of the modal-ity. Under the premarket approval process, thiscan be expensive when a d evice is as new as N MRimaging . . . Under th ese circum stances, a spon-sor may wish to make claims that insure commer-cial viability of a system but that are not inor-dinately costly (164).

In November 1980, FDA published a “Guidelinefor the Arrangement and Content of a PremarketApproval Application” to aid sponsors in thepreparation of such applications (187). Accord-ing to those guidelines, a PMA application shouldinclude a description of the disease(s) or condi-tion(s) that the device will diagnose and the pa-tient population for whom the device is intended.

About half of the manufacturers surveyedstated that they would have liked more preciseguidelines from FDA regarding the required con-tent for a PMAA (e.g., how many patients needto be studied, whether studies need to be blinded,etc. ); the others felt that sufficient guidance hadbeen provided by FDA. Manufacturers who hadreceived feedback from FDA on submitted PMAAsfelt that FDA officials had been extremely helpful,fair, and reasonable in their review of PMAAs,particularly since NMR was the first Class 111 im-aging device to go through the PMA process. Afairly common complaint from manufacturers,however, was that the PMAA format was unnec-essarily tedious and complicated.

FDA Review

FDA is allotted 180 days to review and eitherapprove or disapprove a PMAA that satisfies allregulatory requirements. During this review p roc-ess, FDA customarily provides feedback to spon-sors regarding possible deficiencies in theirPMAAs. The underlined qualifier can thus takeon significant importance, since FDA can stop the180-day clock while a sponsor responds to or

remedies the possible deficiencies that have beenidentified by FDA.

Panel Review

The Medical Device Amendments require thatFDA refer each PMAA to an appropriate expertadvisory panel which, after considering all dataprovided, makes a nonbinding recommendationto FDA regarding whether the PMAA at issueshould be disapproved, approved, subject to cer-tain modifications, or approved. On July 6 an d7, 1983, FDA conducted an open hearing of theRadiologic Devices Panel on three NMR devicePMAAs submitted by Diasonics, Picker Interna-tional, and Technicare. Picker’s PMAA pertainedto NMR imaging of the head and neck only, whileDiasonics’ and Technicare’s pertained to NMR im-aging of both the head and body, The panel con-sidered all the applications “approvable” andvoted unanimously to recommend approval of allthree PMAAs, subject to various contingencies,such as making specified modifications in Site

Planning Guides or labeling.

FDA Approval

The Medical Device Amendments state that aPMAA is to be denied if:

1.

2.

3.

4.

reasonable assurance is lacking that the de-vice is both safe and effective under the con-ditions of use prescribed, recommended, orsuggested in the proposed labeling;the methods used in the manufacture of thedevice do not conform to Good Manufac-

turing Practices;the proposed labeling is false or misleadingin any particular; orthe device does not conform to a perform-ance standard with which it is to comply.

In evaluating the safety and effectiveness of adevice for PMA of a Class III device, FDA con-siders, among other relevant factors:

1.

2.

the persons for whom the device is repre-sented or intended;the conditions of use for th e device, includingconditions prescribed, recommended, or sug-

gested in the labeling or advertising of thedevice, and other intended conditions of use;

8/14/2019 8420


Ch. 7—Regulation by the Food and Drug Adm inist ration 93

3.

4.

the probable benefit to health from the useof the device weighed against any probableinjury or illness from such use; andthe r eliability of the dev ice (21 CFR 860.7).

After having considered these factors, FDA reg-ulations specify that:

1. There is reasonable assurance that a deviceis safe when it can be determined, basedupon valid scientific evidence, that the prob-able benefits to health from use of the de-vice for its intended uses and conditions of use, when accompanied by adequate direc-tions and warnings against unsafe use, out-weigh any probable risks. The valid scien-tific evidence used to determine the safety of a device shall adequately demonstrate theabsence of unreasonable risk of illness or injury associated with the use of the device forits intended uses and conditions of use (21CFR 860.7).

2. There is reasonable assurance that a deviceis effective when it can be determined, basedon valid scientific evidence, that in a signif-icant portion of the target population, theuse of the device for its intended uses andconditions of use, when accompanied byadequate directions for use and warningsagainst unsafe use, will provide clinicall y sig-

nificant results (21 CFR 860.7).

The safety and effectiveness of a device mustthus be considered in conjunction with oneanother, since assurance of safety depends on anevaluation of effectiveness.

FDA issued formal premarket approval forNMR imaging devices manufactured by Diasonicsand Technicare on March 30, 1984, and for headand neck imaging devices by Picker on May 10,1984.

CONCLUSIONS: THE PMA PROCESS AS A WHOLE

The application of the PMA process to NMRimaging devices raises several issues. First iswhether there should be a PMA process at all,and, if so, what benefits derive from it. Congressestablished the PMA process in 1976 in responseto a perceived need for greater protection fromunsafe, unproven, ineffective, and experimentalmedical devices. At least with regard to NMR im-

agers, the PMA process seems to have successfullyadd ressed that p erceived need. Although disagree-ments may exist over how much data should berequired before PMA is granted, there seems tobe a general consensus that the PMA processserves a useful function in assuring the safety andeffectiveness of marketed devices. As one manu-facturer stated, “The PMA process provides thediscipline required to force manufacturers to de-velop information they ought to have. ”

The second general PMA issue re la tes towhether a separate PMA should be required foreach clinical application of NMR or whether PMAshould be granted for the technology as a whole.No clear consensus emerges on this issue. On theone hand, it seems possible that an imaging tech-nology such as NMR may well prove to be effec-

tive in some but not all potential applications, sug-gesting that it would be reasonable for FDA togrant PMA on a clinical application, by clinicalapplication basis, much as it does with drugs. Onthe other hand, it can be argued that as long asthere is reasonable assurance that NMR imagingis safe and that NMR is effective, in the sense thatit gives a fairly accurate representation of inter-

nal anatomy, pathology, or function, it shouldbe up to physicians, rather than FDA, to decidewhich NMR applications are appropriate. Givensome threshold level of demonstrated effective-ness, it would seem that the latter viewpoint isnot only reasonable, but also may be the only fea-sible one for FDA to adopt, since FDA cannotcontrol each application once NMR devices areinstalled. b How FDA resolves this issue may de-pend on the breadth of the claims made by man-ufacturers in their proposed labels,

bAlthough i t would not be feasible for FDA itself to enforce a re-striction on the use of NMR to certain clinical applications, theabsence of third-party coverage for uses not approved by FDA mighteffectively curtail such uses.

8/14/2019 8420


94 Ž Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis ——.

A third general issue relates to the manufac-turers’ costs for data collection and PMAA prep-aration and FDA’s costs for reviewing the applica-tions. The central question regarding the cost of the PMA process for the manufacturer relates tothe amount of money that would not otherwisehave been spent on the assessment of safety and

effectiveness if the PMA process did not exist.Most manufacturers said the difference was “anegligible amount, ” with most of it associatedwith employment of study design consultants andclerical preparation of the PMAA itself. FDA esti-mated that by July 1983 it had expended about800 person-hours of effort on reviewing the firstthree NMR PMAs submitted to it (163). Theseestimates do not suggest that FDA regulation of NMR devices has entailed high direct costs. Tothe extent that these assessments are accurate,there seems to be little at issue other than the pos-sibility of streamlining the PMAA itself. To the

extent that pertinent, well-designed clinical studiesare performed that wou ld not otherwise be fundedby manufacturers in the absence of the PMA proc-ess, it would seem that the PMA process is serv-ing a useful function.

Fourth, the question arises as to how much thePMA process has constrained development andearly placement of NMR imagers. There is no in-dication that the PMA process has restrained de-velopment of the prototypes themselves. In ad-dition, the great majority of NMR manufacturersthat we surveyed in the summer of 1983, statedthat i f the PMA process had not exis ted , theywould have placed few, if any more NMR imag-

ing systems in hospitals than they had already be-cause many manufacturers were still developingand refining prototype systems and had not yetbegun full “assembly-line” production capable of meeting existing demand. FDA thus does not ap-pear to be significantly delaying the introductionof experimental model NMR imaging devices into

hospitals. In addition, it should be realized thatmanufacturers use the experience they gain dur-ing the IDE period to refine system designs beforeembarking on full-scale production.

The actual and potential impact of the PMAprocess may well change in the near future, how-ever, as manufacturers emerge from the prototypedevelopment stage. Manufacturers have stated,for example, that many existing “orders” are con-tingent on the manufacturers’ receiving PMA. If PMA is not granted in a timely fashion, thesemanufacturers may begin to experience delays inreceiving revenues to cover their developmentcosts.

Perhaps the greatest potential impact of thePMA process—stemming from its ability to con-fer a competitive advantage on manufacturerswho have received PMA first—is yet to be seen.How much of a financial benefit, in both the shortand long run, will accrue to NMR manufacturerswho are first to obtain PMA may well help deter-mine not only the future of the NMR manufac-turing industry, but also the speed with whichmanufacturers pursue development of other newtechnologies that emerge in the future.

8/14/2019 8420


8.Third-Party Payment Policies

9

8/14/2019 8420


8Third-Party Payment Policies

INTRODUCTION

As one of several important economic andsocial forces influencing the adoption and use of medical technologies in recent years, third-partypayment policies have increasingly become a major focus of attention (12,100,157,200). Of greatconcern to many policy makers have been the in-centives engendered by payment based on coststhat have already been incurred. The general fail-ure of such payment mechanisms to distinguishbetween cost-saving and cost-raising technologieshas offered little incentive for hospitals to includeefficiency among their capital investment objec-tives (100,200). Consequently, hospitals’ decisionsto acquire new technolog y have placed little em-phasis on the comparative cost effectiveness of technology use in clinical practice,

The cost-based system of payment has alsotended to reward institutions that increase theirdefinable costs without necessarily improvingquality of care, while simultaneously penalizingthose that improve care with a concomitant reduc-

THIRD-PARTY PAYER DECISIONS

In setting policies for the payment of hospitaland medical services involving the use of newmedical technologies, third-party payers mustwrestle with three questions:

1. Should they pay for such services?2. If so, under what conditions or circum-

stances should they pay for them?3. How much should they pay under specified

conditions?

The first and second questions relate to policydecisions regarding coverage of new services and

the specific conditions that may apply, The thirdquestion pertains to policy decisions involving thereimbursement or payment level permiss ib leun der specified cond itions of service coverage. To-gether, these three questions represent a sequence

tion in operating expense (151,198). In view of these concerns, the Medicare program is begin-ning to pay hospitals according to a prospectivepayment system, with rates set in advance of theperiod during which they apply. This change hasnew implications for technology adoption anduse.

This chapter ad dresses third-party paym ent pol-icies and how they apply to NMR imaging de-vices. The first section addresses the types of pol-icy decisions made by the major third-partypayers and the processes by which they determine

coverage and payment levels for new technol-ogies. The second section d iscusses th e history an dcurrent status of third-party payer decisions re-garding NMR imaging. Readers familiar with theoperations and policymaking processes of the major third-party payers may wish to read only thesecond section. Otherwise, the first section pro-vides a foundation for understanding how pol-icy decisions are made.

through which all third-party payers must passwhen formulating comprehensive policies towardmedical technologies. Payers tend to differ, how-ever, in their general procedures, methods of as-sessment, and decision criteria. Nevertheless, theend produ ct in each case is a determination or p ol-icy statement intended to guide policymakingwithin the program or within member plans orcompanies.

Coverage Policies

When a new medical technology moves fromthe laboratory into the hospital, third-party pay-ers must decide whether or not to pay for its use.For a device such as NMR that requires the Food

8/14/2019 8420


and Drug Administration’s (FDA) approval before nology assessments may also originate from out-marketing, the coverage question often arises side groups, such as medical specialty societies andbefore or concurrently with FDA review, as man- manufacturers (192).ufacturers or providers contact insurers aboutcoverage policy before reimbursement claims aresubmitted. In the period prior to FDA premarketapproval, the technology or device is considered

“investigational,” and third-party payers tend notto reimburse for clinical services performed withit. The critical decisionmaking period, therefore,is the time just after FDA premarket approval hasbeen granted, when hospitals and other providersanxiously await third-party payers’ decisions onboth coverage and reimbursement level.

Within HCFA, a coverage question is directedfirst to the Bureau of Eligibility, Reimbursement,and Coverage and its Office of Coverage Policy.

HCFA may in turn seek the advice of the PublicHealth Service. This advisory role rests with theOffice of Health Technology Assessment (OHTA)in the National Center for Health Services Re-search. The full assessment process generally re-quires 8 to 18 months to complete (106,107). T h eassessment process frequently coincides with theFDA premarket approval process and the two

Medicareagencies often share available data and informa-tion. In making its coverage policy decision,

The Medicare program may reimburse for only HCFA is not bound by the Public Health Servicethose devices, services, or procedures that are recommendations.determined to be both “reasonable and necessary.”

In making this determination, the Health CareFinancing Administration (HCFA), which admin-isters Medicare, first considers whether FDA hasfound the device “safe and effective. ” In practice,HCFA generally does not approve coverage of anew device unless FDA has already approved it.HCFA considers it to be a necessary but not suf-ficient condition that a technology be “safe andeffective” in order for it to be “reasonable and nec-essary.” HCFA, however, will not necessarily ap-prove coverage for all devices that FDA has ap-proved, largely because the two agencies differin their respective definitions of “effectiveness. ”

FDA deems a technology “effective” if it does whatthe m anufacturer claims it w ill do, whereas HCFAconsiders the effectiveness of the technology withrespect to health outcome. Another important con-sideration in HCFA’s decision is the stage or level

HCFA does not give consideration to the costeffectiveness of a technology when formulatingits coverage policy decision, but may do so laterwhen making policy decisions regarding reim-bursement or payment levels. ’ At its discretion,HCFA may place restrictions on the coverage of a technology rather than gr ant “blanket approval”for the technology as a whole. In the past, restric-tions have sometimes been application-specific—i.e., reimbursement is provided only when thetechnology is u sed for sp ecific clinical ap plicationsor diseases. In other instances, coverage restric-tions have centered on specific service settings

(e.g., only inpatient) or providers or practitioners(e.g., physicians only), or even manufacturers andtheir devices. For example, coverage of CT scan-ners was, at first, limited to only specific modelsproduced by certain manufacturers (106,107).

of acceptance of the innovation by the medicalMedicaid

community—i.e., the extent to which the tech-nology has become an accepted part of clinical The HCFA decisionmaking process, as describedpractice. above for Medicare coverage policy, generally

In the absence of a centrally established HCFAdoes not apply to the Medicaid program. Respon-

coverage policy for a particular medical technol-sibility for making Medicaid coverage policy deci-

ogy, the fiscal contractors for the Medicare pro-sions rests with the individual States, which, attheir own discretion, may choose to cover cer-

gram may make their own coverage policy deci-sions. When questions regarding the safety andclinical effectiveness of a technology arise in thefield, however, fiscal contractors may request thatHCFA perform an assessment. Requests for tech-

‘HCFA sometimes sets charges or allowable rates for a new tech-nology based on previous charge experience with technologies thatare clinical alternatives to the innovation in question (see discus-sion on reimbursement level decisions).

8/14/2019 8420


tain technologies or services not covered by Medi-care. States may also devise their own coveragerestrictions on technology use. In cases where aState program extends Medicaid coverage to tech-nologies not covered by Medicare, the programmay initiate its own internal technology assess-ment, utilizing its own staff and possibly a panel

of outside experts. Alternatively, States may optto refer technology-related inquiries directly toHCFA for potential assessment.

As with current Medicare policies, Medicaidwill pay only for services and technologies judged“reasonable and necessary. ” Noncovered technol-ogies receive no reimbursement under Medicaid.

Blue Cross and Blue Shield Plans

Although the national Blue Cross and BlueShield (BC/ BS) Association p lays an imp ortantrole in assessing new technologies on behalf of its

member plans, each plan reserves the right tomake its own coverage policy decisions regardingspecific medical technologies or services. In mak-ing such determinations, individual plans gener-ally require that a technology receive FDApremarket appr oval before considering reimbu rse-ment coverage for its use (47). FDA approval, onthe other hand, does not a u t o m a t i c a l l y e n s u r eBC/ BS coverage; nor does HCFA approval of thet e ch n o lo g y a s “reasonable and necessary” forMedicare beneficiaries. HCFA decisions, never-theless , are scrut in ized careful ly b y t h e p l a n s .

When questions arise regarding coverage of anew technology ( i .e., an individu al BC/ BS planreceives an inquiry, claim, or letter of intent froma participating hospital or physician), the plan willoften contact the national Association and requestassistance. An internal technology assessmentprocess then begins to develop recommended cov-erage policy for plans to consider when makingtheir respective policy decisions. The assessmentprocess generally requires 4 months to 1 year tocomplete (47).

Unlike the policy decisions of HCFA regardingMedicare coverage, the policy statements issued

by the n ational BC/ BS Association are str ictly rec-o m m e n d a t i o n s or guidelines. The thrust of theBC/ BS assessment process, therefore, is not to for-mulate policy with carefully delineated conditions

of coverage, but rather to identify the importantissues that plans must address and to presentuseful information that will aid local decisionmak-ing. An important aspect of this “informationclearinghouse” function is the clarification of tech-nical and clinical details relating to a specific tech-nology’s safety, effectiveness, clinical status, and

appropriate use.BC/ BS employs a th ree-level scale of clinical

status: 1) experimental—the technology has beenused only in animal studies, 2) investigational—the technology has entered preliminary clinicaluse, and 3) accepted medical practice—the tech-nology has gained general use in medical prac-tice. Some BC/ BS plans w rite broad exclusionaryclauses in their contracts for “experimental/ in-vestigational” devices. Others prefer to deal withemerging medical technologies on a case-by-casebasis, making it possible, but not likely, that someinvestigational devices may receive coverage,albeit for specific clinical applications or uses.

When formulating policy recommendations oncoverage of new technologies, the internal reviewcommittees of th e BC/ BS Association take cost-effectiveness information into account. The infor-mation does not affect the recommendationsdirectly, but rather is transmitted along with thepolicy statement to member plans. Plans are thenfree to weigh the information accordingly in theirrespective coverage policy decisions. Some plansengage in sophisticated assessment activities of their own. BC/ BS of Massachusetts, for example,

convenes an Interdisciplinary Medical AdvisoryCommittee to assist it in making coverage deci-sions for new technologies (47). Other plans mayinstead conduct their own assessments or surveysof available information on new technologies. Inaddition, since many plans serve as Medicarefiscal contractors (i. e., ad minister Medicare claimsfor HCFA), they may either closely observe orparticipate in the HCFA assessment process.

Commercial Insurance Companies

Commercial insurance companies operate inde-

pendently of one another and, therefore, makeindependent decisions regarding coverage of newor emerging medical technologies. When a ques-tion arises concerning payment for a technology

8/14/2019 8420


100 Health Technology Case Study 27: Nuclear Magnetic Resonance imaging Technology: A Clinical, industrial, and Policy Analysis

whose safety and effectiveness may not be known,individual companies contact the Health Insur-ance Association of America (HIAA), a privateorganization that represents and serves the com-mercial insurance industry. HIAA membershipincludes 338 companies, which collectively pro-vide approximately 85 percent of all non-Blue

Cross, private health insurance coverage in theNation (112).

The HIAA inquiry and assessment process isfrequently conducted concurrently with the FDApremarket approval process, as well as with thetechnology assessment activities of other third-party p ayers, most notably HCFA and the BC/ BSAssociation. Individual commercial insurancecompanies may, of course, supplement the in-formation obtained from the HIAA process byundertaking their own assessment activities. Suchindepend ent efforts tend to be of a limited natu re,often involving direct solicitation of expert opin-ion from the most relevant medical specialtygroups.

In making coverage policy decisions, companiesface the same choices encountered by other third-party payers—i.e., coverage without restrictions,coverage with restrictions, or no coverage at all.As with HCFA and BC/ BS, comm ercial insurersview FDA premarket approval as a necessary butnot sufficient condition for coverage of a newtechnology. Companies examine closely the pol-icy decisions rendered by HCFA and by variousBC/ BS plans, but w ill not necessarily adop t them.

Reimbursement or PaymentLevel Policies

Once the question of coverage policy has beendecided by a third-party payer, attention is fo-cused next on the issue of appropriate or “reason-able” payment for provision of covered services.Although “reasonableness” is a concept with in-trinsic meaning to all third-party payers, its oper-ating definition in practice will vary among pay-ers. Complicating the picture is the need for eachpayer 2 to set separate reimbursement or payment

‘Medicare does not set technology-specific payment rates for pro-spectively paid inpatient hospital services under Part A, but doesset rates for physician services under Part B.

rates for the hospital or facility in which the tech-nology will be employed and for the professionalservice fee of the physician. Both rates can varyby geographic area, by service setting (e.g., hos-pital versus physician’s office), by physician spe-cialty, by clinical application of the technology,and by the past experience and fee history of the

individual practitioner.

Medicare

HCFA sets reimbursement rates for covered physician services based on what it considers tobe “customary, prevail ing, and reasonable”charges. In making these determinations, HCFAstaff in the Office of Reimbursement Policy con-sider such factors as:

ŽWhere the technology will be used—in thehospital, in the physician’s office, or in someother setting?

• How the technology will be used—for whatclinical applications or disease conditions? ŽBy whom the technology will be used—by

physicians with what type or level of spe-cialty training and / or experience?

“Customary and prevailing” charges imply thatconsideration is also given to: 1) the customaryor usual fees charged by a given practitioner forsimilar or related services in the past, and 2) theprevailing fee (or market price) charged by physi-cians in the same geographic area and with simi-lar training/ experience for similar or related

services. In the case of many new or emergingtechnologies, customary and prevailing chargesare impossible to document since little clinical ex-perience, if any, has been gained in the generalmedical community by the time an HCFA policydecision is made. In such instances, HCFA staff or individual Medicare contractors look to simi-lar technologies and base their reimbursement orpayment rates, at least in part, on past experiencewith established services. For example, assumingMedicare coverage is granted for NMR imaging,policy decisions on reimbursement level will likelybe based, in part, on the previous history of

charges for X-ray CT scanning.Since October 1, 1983, HCFA has begun to pay

hospitals prospectively for inpatient services onthe basis of diagnosis related groups (DRGs).

8/14/2019 8420


.—

Ch. 8— Third-Party Paym ent Policies 101 — . — .—— —

The payment for utilizing a given technology fora specific DRG is thus embedded within the estab-lished DRG cost per case. For the time being,Medicare is continuing to pay for the capital costs(e.g., acquisition expense, interest, rent, landcosts, and other expenses) incurred for majorequipment. The capital allowance will be based

on the proportion of total hospital charges in somebase year that is attributable to Medicare patients.Capital expenses will be treated in this “pass-through” manner for the next 3 years, or until cap-ital costs are brought into the DRG payment rates.For uncovered services, HCFA would refuse topay such pass-through capital costs.

Thus, a hospital that chooses to invest in newtechnology under the prospective payment sys-tem is at risk that the added patient-managementcosts induced by the new technology will resultin financial losses. More specifically, if use of anew technology (either as a substitute for, or asan add-on to, some other modality) increases theaverage operating cost of a given DRG (or DRGs)to a level above the prospective payment rateestablished for that DRG (or those DRGs), thehospital will not recover its costs.3 In addition,hospitals adopting and using an uncovered tech-nology would not be reimbursed for associatedcapital costs, and would also stand to lose shouldtheir average operating costs in the DRGs that usethe new technology exceed the approved DRGpayment rates.4 A decision to acquire and to usenew technology under the Medicare prospectivepayment systems, therefore, may have serious im-plications for the financial well-being of a hosp ital.

Although HCFA will not need to establish areimbursement rate for use of a new technologyin the inpatient setting, it will need to establishpayment rates for outpatient usage and for theprofessional fee associated with both inpatient andoutpatient usage. The level(s) at which thesepayments are established could have a tremend ousimpact on the rate at which new technology isadopted in both inpatient and outpatient settings.

3Conversely, adoption of cost-saving technology that decreasesthe average operating cost of a given DRG (or DRGs) relative tothe prospective payment rate would benefit the hospital, which isentitled to keep the savings that would be generated,

‘Costs could potentially be recovered if payments exceeded costsfor other DRGs or if costs were shifted to non-Medicare payers,

Medicaid

Medicaid reimbursement policy for new tech-nologies generally is not tied to that of the Medi-care program (107). Although Medicaid rates forboth hospital care and physicians’ services are setusing many of the same criteria described abovefor Medicare, the relative weights of such factors

will differ by State program, resulting in consid-erable variation in payment levels across theNation.


Once coverage for a new technology has beenapp roved by a BC/ BS plan, the criteria of “usual,customary, and reasonable” (UCR) fees are em-ployed to set physician charges for services. Underthis approach, considerable weight is given to pasthistory and to the plan’s experience with particularphysicians (47), Individual plans differ, however,

in their approach to payment for hospital serv-ices; some plans reimburse hospital charges;others pay only for costs. With emerging tech-nologies, the lack of relevant technology-specificdata or past experience often requires the plan toexamine charges or costs for related technologiesor services from which they can impute likelycosts and set “reasonable” charges for the newservice.


The general procedures and criteria used by in-dependent commercial insurance companies inestablishing allowable hospital rates and physi-cian fee schedules are essentially similar to thosedescribed above for Blue Cross and Blue ShieldPlans. Commercial insurers, however, tend todeal directly with the insured rather than withproviders. ’ Within a given company, therefore,the reimbursement for a specific service is morelikely to be standard ized than it is within a BC/ BSplan (112).

General Observations

These cost-based reimbursement policies for

hospitals and fee-for-service payments to physi-5This may be changing, as more commercial insurers begin to par-

ticipate in the development of preferred provider arrangements withhospitals and physician groups.

8/14/2019 8420


cians have been criticized for their retrospectivenature and for their inherent biases toward in-creased technology adoption (198,200). Becausephysician fees for new technologies cannot easilybe tied to historic or prevailing charges, and be-cause UCR fee sched ules are n ot likely to be estab-lished until after such technologies have been in-

troduced, payment for new procedures is oftenset high and rewards technology adoption. Ret-rospective cost- or charge-based reimbursementsystems also give providers little incentive to dis-

tinguish between cost-saving and cost-raising tech-nologies and may influence private physiciangroups to acquire new technologies without re-gard to their cost effectiveness. The widespreadacquisition of X-ray CT scanners by private radi-ology groups, for example, may find its parallelin large-scale purchases of NMR imaging devices

by such groups. In addition, the continuation of historical reimbursement policies for physicianfees may provide physicians with incentives tooverutilize technology even in the hospital setting.

HISTORY AND STATUS OF COVERAGEPOLICY DECISIONS FOR NMR IMAGING DEVICES

During early 1984, increasing numbers of thirdparties began paying for NMR. By June 1984, atleast 10 commercial insurers were paying for NMR

as part of “generally accepted practice, ” and atleast three Blue Cross plans had accepted NMRfor payment.’ Assessments of NMR imaging arebeing und ertaken by HCFA/ OHTA and by theBC/ BS Association. The statu s of each p ayer’spolicy regarding NMR imaging is summarizedbelow.

Medicare

HCFA became involved in the assessment of NMR imaging as early as January 1982, when theagency received literature pertaining to NMR

from the General Electric Co., together with a re-quest for comments, but no request for an assess-ment (17). In May 1982, HCFA received a formalquery regarding Medicare coverage policy forNMR imaging from a Blue Cross plan in Califor-nia, which had itself received an inquiry from aneurosurgeon (17). Acting on this inquiry, staff from the Office of Coverage Policy performed aliterature review, contacted other Federal agen-cies (including FDA), surveyed NMR imagingmanufacturers for information, and prepared apresentation to the HCFA Physician Panel inAugust 1982 (17). Later that month, the Physi-

cian Panel requested that OHTA perform a full

“Nfoblle Techni.~l[>gy~ [nc , unpublished data, Lo \ ingeles, C A,ILlnt’ 10! 1Q84.

assessment of NMR imaging. In September 1982,OHTA began to look at NMR but delayed a com-plete assessment pending FDA premarket ap-

proval. Recently, following FDA approval of thefirst manufacturers’ applications, OHTA initiatedefforts to assess NMR imaging. A decision byHCFA on Medicare coverage of NMR imaging isnot expected during 1984.

Medicaid

Presently, it is not known whether State Med-icaid agencies have conducted their own assess-ments of NMR imaging. It is possible that thecurrent HCFA/ OHTA assessment process maysatisfy the information needs of the individual

State Medicaid programs.


In October 1982, national Association staff for-mally requested an assessment of NMR imaging.Later that month, the Medical Advisory Subcom-mittee reviewed th e request and decided to initiatean assessment. At that time, the Subcommitteealso designated NMR imaging as an “investiga-tional device” for which no reimbursement shouldbe provided. The staff then performed an assess-ment and reported its findings to the Subcommit-

tee in March 1983. The Subcommittee reviewedthe report and approved its submission to mem-ber plans as a “Medical Policy Newsletter. ” Thenewsletter provides information on the current

8/14/2019 8420


. .

status of NMR imaging as an investigational de-vice, with a view toward clinical applications,technical considerations, and charges (18). Thenewsletter also offers advice to plans based on cur-rent evidence from the literature, from the FDA,and from the American College of Radiology. Itis not intended, however, as a uniform medical

policy statement, since the issue of NMR imag-ing is still under consideration by the Association.


Through December 1983, the HIAA had not re-ceived any inquiries from member organizationsregarding NMR imaging (112). HIAA staff werealso not aware of any claims for NMR servicesthat might have been received by member com-panies. Therefore, the staff had not solicited an

CONCLUSIONS

Since FDA has granted premarket approval toan NMR device, third-party payers now have major influence over the rate at which NMR imagersare acquired by hospitals. This influence derivesnot only from their decisions regarding whetherto cover use of NMR, but also from their deci-sions regarding th e circumstances in wh ich u se will

be covered. Third-party payers such as HCFA,for example, will first need to decide whether theywill reimburse for use of only those manufac-

turers' NMR devices that have gained PMA, orwhether they will reimburse for use of any man-ufacturer’s NMR device. Such a decision couldhave a m ajor imp act on the market share achievedby manufacturers in the short run.

Third-party payers will also have to decidewhether to make a broad or narrow coverage de-cision. In the case of NMR, at least in the shortrun, this will come down to deciding whether toapprove reimbursement for some applications of NMR imaging of the head only (the applicationsin which NMR has so far proved most effica-

cious ), or whether to approve reimbursement forall uses of NMR, regardless of their stage of de-velopment. To the extent that the narrow strat-egy is followed, hospitals may be restrained in thespeed with which they acquire NMR devices. To

opinion on NMR imaging from the Council onMedical Specialty Societies.

According to a survey of 30 commercial com-panies that provide health insurance, by February1984, five had determined that NMR was part of “generally accepted practice” and were paying for

its use. In February, six other companies had ac-knowledged the clinical usefulness of NMR, butwere reviewing each case before payment. By mid-June 1984, 10 of the 30 companies deemed NMRgeneral ly accepted and were paying for proce-dures, and 11 other companies had provisionallyaccepted NMR and were paying after review of each case. 7

“H?ld.

the extent that the latter strategy is followed,HCFA and other third-party payers will likely besubsidizing research on N MR applications that areless well developed.

The third major decision to be made by HCFAand other third-party payers is the monetary levelat which outpatient use of the NMR imager willbe reimbursed. How mu ch of a difference in reim-bursement is established for outpatient use of

NMR as compared to outpatient use of X-ray CTwill have a major impact on the rate at whichNMR imaging systems diffuse into the outpatientsetting. HCFA is beginning to pay hospitals pro-spectively on the basis of inpatient diagnosis, butwill need to set an inpatient fee for physicians.Other third-party payers, such as Blue Cross andcommercial insurers, may set inpatient rates,depending on their payment methods. Because itis likely that the outpatient rates set by HCFA willinfluence the inpatient rates established by BlueCross and commercial insurers, HCFA’s out-patient rate takes on even greater importance.

Another important decision to be made bythird-party payers is the level at which profes-sional fees for NMR imaging are set. At least inthe near future, it can be expected that more pro-

8/14/2019 8420


104 Ž Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis———.

fessional time will be required for NMR scans thanfor X-ray CT scans. The level at which profes-sional fees are established, therefore, may wellhave a significant impact on the level of interestthat radiologists and other potential users mani-fest with regard to acquisition of NMR imagingdevices.

Prospective systems of payment may have amajor influence on the rate at which NMR dif-fuses throughout the medical system. With the in-troduction of DRG-based prospective paymentunder Medicare, hospitals will have to respondto different incentives from those to which theyhave been accustomed. Technology acquisitionis one area of hospital operations in which changeis likely to result. Hospitals, in theory, will haveto weigh financial considerations against patient-care benefits more carefully when acquiring tech-nologies. The constraints imposed by prospectivepayment on hospital budgets will likely detersome institutions from acquiring medical technol-ogies that raise operating costs. In such instances,prospective payment may supersede State certifi-cate-of-need regulation as a constraining influenceon hospital investment decisions. In addition, hos-pitals may need to become more discriminatingabout deciding how acquired technology is used.Whether and how such rationing decisions willbe made remain uncertain.

One potential concern about the advent of pro-spective systems of payment is whether some hos-pitals will be so financially constrained that they

—

will be unable to acquire valuable new technol-ogy. If or when capital costs become included inthe Medicare DRG payment rates, hospitals maybe further constrained in their technology acqui-sition decisions. It is also important to realize thatMedicare’s DRG payment may vary among hos-pitals in its effect on their financial condition and

their ability to acquire and use new technology.For some institutions, such as municipal hospi-tals serving large Medicare and Medicaid popula-tions, the DRG payment system may exacerbatean already financially troubled state, impedinghospital capital formation necessary for the ac-quisition of high-cost but beneficial new technol-ogies. The net effect may be to weaken furtherthose institutions that are the primary sources of care for disadvantaged populations.

How much of an impact the prospective pay-ment system will have on technology acquisitionis likely to depend on HCFA decisions regardingupdating and recalibration of DRG payment rateswhen new technologies become available (186). New technologies such as NMR are likely to beused across multiple DRGs. If NMR proves to bebeneficial but not cost-saving in certain DRG ap-plications, hospitals will be confronted by con-flicting patient care and financial considerations.Periodic recalibration of DRG payment rates maythus be required as technological change in medi-cine occurs, In the absence of such recalibration,patients may be restricted from access to poten-tially beneficial new technologies.

8/14/2019 8420


9.State Certificate-of-Need Programs

8/14/2019 8420


8/14/2019 8420



important quality. CON agencies frequently playpivotal roles in determining which institutionsmay acquire new technologies. Determinationsbased on broad concepts of “need,” including therelative need demonstrated by competing CONapplicants, are intended to ensure equitable alloca-tion of new technology among hospitals. CON

efforts to achieve distributional planning goals,however, have sometimes conflicted with pro-gram objectives involving cost containment. Forexample, as some observers (11) suggest, themisdirection of cost containment goals in the earlyyears of X-ray CT scanner diffusion produced amaldistribution among hospitals that disenfran-chised whole segments of the hospital industry—e.g., the municipal hospitals serving disadvan-taged populations. Avoidance of this “franchis-ing effect” is important if CON regulation is tohave an even-handed impact on future diffusionof new technologies, such as NMR imaging d evices.

In the past, CON programs have employed dif-ferent policy orientations to address the issuesassociated with technology adoption and distri-bution (35,135). At various times, health plannershave used strategies such as:

Pro forma denial—denial of all CON applica-tions for an indefinite period of time as ameans of strict cost control; usually stemsfrom serious concerns over the safety, effica-cy, and cost of a technology.Formalized strategy of delay—temporarylimitation of all CON applications, pending

future availability of better data for CON re-view; often achieved through moratoria andapplication review deferrals.Predetermined limits on diffusion—limitationof CON approvals to specific sites or pro-viders; often conditional on the provision of clinical data that can aid future evaluationof the technology.Uncontested approval–approval of CONapplications for - new technology in the ab-sence of data on which to base sound CONdecisions or in the face of statutory require-ments that dictate approval unless need canbe shown not to exist.

Of these four strategies, only the second andthird have been used to advantage by CON agen-

cies. All, however, suffer from their reliance onthe high capital-cost “trigger” that is the hallmarkof CON programs, and from their inability to re-view technologies in the premarket stages of de-velopment (36). For these reasons, CON programshave not been successful in either controlling theintroduction of new technology or assuring equi-

table distribution of equipment among hospitals(135). A further problem is that CON review of innovative change places health planners on lessfamiliar ground where they lack the requisite tech-nical, medical, and analytic skills needed to an-swer important questions about safety and effec-tiveness in the absence of FDA findings (36).

Newly emerging technologies are especially dif-ficult to review since the information required forassessment is usually unavailable.

At present, State CON laws generally apply tothe acquisition by hospitals of medical equipment

and devices that exceed specified Federal dollarthresholds: 3 $400,000 for major medical equip-ment and $250,000 for new institutional services(Public Law 97-35, 1981). In order to receive Fed-eral funds for various health programs, Statesmust comply with Federal law that requires theirCON statutes to contain provisions for review of acquisitions, by anyone, of major medical equip-ment that will be used to provide services to hos-pital inpatients (182). This requirement is intendedt o p r e v e n t c i r c u m v e n t i o n o f S t a t e C O N l a w seither by hospitals that have been denied plan-ning agency approval for a specific technology or

by physician groups seeking to acquire and in-stall major medical equipment in a facility out-side the hospital (such as a medical arts building)where technology acquisitions may otherwiseescape CON review. The precise coverage policiesgoverning CON review vary by State program,but most do not cover equipment acquisition inphysicians’ offices. Only eight States (Colorado,Connecticut, Hawaii, Iowa, New Hampshire,Rhode Island, Virginia, and Wisconsin) plus theDistrict of Columbia currently have CON lawsthat provide more stringent coverage of equip-ment acquisitions, such as in physicians’ offices,

than the minimum Federal requirements (182).3States must use these thresholds in order to comply with Fed-

eral lan”; they may, however, use more stringent thresholds, at theirdiscretion.

8/14/2019 8420


Ch. 9—State Certificate-of-Need Programs • 109

RELATIONSHIP OF CON REVIEW TO THEFDA PREMARKET APPROVAL PROCESS

In theory, the FDA premarket approval proc-ess should precede the CON review process, butin practice, the two often coincide. In most States,“investigational devices” are exempt from full

CON review, but a notice of intent to acquire sucha device for research or experimental purposesmu st nevertheless be filed by the h ospital with theappropriate health planning agency or agencies.Thus, a hospital may acquire an investigationaldevice without passing through formal CON re-view while the device is undergoing FDA reviewfor possible premarket approval. Once FDA ap-proval is granted to a medical device, all subse-quent acquisitions by other providers must under-go CON scrutiny, provided that the acquisitioninvolves a setting (e. g., hospital, ambulatory care

center, etc. ) that is specifically covered by appli-cable State law. FDA premarket approval, there-fore, is generally a prerequisite for widespread dif-fusion of a new technology but i t does notnecessarily guarantee broad adoption, since CONreview is based on criteria that differ from thoseused by the FDA.

Whereas FDA review examines the safety andeffectiveness of a medical device, CON review isconcerned with demonstration of “need.” The def-inition of “need” varies greatly by State CON pro-gram and may involve such diverse factors or cri-teria as: consistency of the proposed project with

State health plans, consistency of the project withthe institutional applicant’s long-range plan, sys-temwide effects, financial feasibility of the proj-

ect, access to care, quality of care, availability of services and personnel, construction and archi-tectural considerations, effects on competition,competence and character of institutional manage-

ment, and selection of the best alternative meansof providing the proposed service (143). FDAassurance that a device is safe and effective is notsufficient to demonstrate need for the device.

The ability of some hospitals to acquire devicesin the “investigational” stage of their developmentwithout having to undergo full CON review con-tains potential for abuse. As was the case withCT scanners—and as we are now seeing withNMR imagers—some hospitals tend to engage in“anticipatory behavior, “ i.e., they file applicationsfor CON exemption early in the diffusion proc-

ess in the hope of securing the technology beforecompetition for the device leads to limited CONapprovals. Once obtained, CON exemption be-stows coveted status on a hospital relative to thatof its competitors and establishes a “franchise”which, in practical terms, is not likely to berevoked by the CON program once FDA statusof the device changes following premarket ap-proval. Thus, the hospital that acquires NMR im-aging as an investigational device is likely to keepthe technology later when competing CON ap-plications may be filed with the review agency.This “franchising effect” could , in the case of NMR

imaging, work to the detriment of some segmentsof the hospital industry.

LESSONS FROM THE EXPERIENCE WITH X-RAY CT SCANNING

Several important lessons have emerged fromthe CON experience with X-ray CT scanning dur-ing the 1970s, the most important of which wasthat slowed technological diffusion has its advan-tages and disadvantages. In the case of X-ray CTscanning, early CON moratoria in some areas en-abled health planners and hospitals to delay crit-ical decis ions pending fur ther information. Thisdelay tact ic a l lowed society to “buy t ime” unt i l

better decisions regarding technology acquisitionand “need” could be made. On the other hand,the inability of planners to evaluate the technol-ogy constrained its diffusion into medical prac-tice more severely than may have been wise. Thelack of available evaluative mechanisms and cri-teria for review made it difficult for planners todispel the uncertainty surrounding X-ray CT scan-ning, thereby leading to many controversial and,

8/14/2019 8420


110 Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology A Clinical Industrial, and Policy Analysis

at times, seemingly arbitrary decisions on indi-vidual CON applications. The net effect was a lossof credibility by the planners, as evidence of thetruly revolutionary nature of X-ray CT scanningaccumulated over time. The approach of selec-tively controlled diffusion now being used in someStates (e.g., New York, Illinois, New Jersey) with

regard to NMR imaging devices is more rational,by comparison (see next section).

A second an d equally important lesson was thatshared CT services among h ospitals pr oved u nsat-isfactory for many institutions. Practical con-siderations, such as access and service volumeneeds, worked against the basic principles of shar-ing, causing many hospitals to abandon theirshared-service arrangements and to acquire theirown X-ray CT scanners. Moreover, some hospi-tals found that multiple X-ray CT units were re-quired to meet service demand, making shared

services even less enticing. The same potentialproblem exists for the shared-service arrangementsnow being proposed for NMR imaging in someareas of the country.

A third lesson involved the unusual behaviorexhibited by hospitals in response to the incen-tives created by the CON process. Anticipatorybehavior of the type described earlier was by nomeans unexpected. In some States, hospitals wereable to acquire CT scanners before CON lawstook effect. In other States, problems arose whenCON programs failed to recognize the inherentinequities that were created by the nature of the

process itself, i.e. hospitals that obtained in-vestigational devices became “grandfathered” oncediffusion of the technology accelerated and insti-tutions that “played by the rules” were effectivel y

penalized for not having taken action sooner. Inadd ition, circumven tion of CON au thority occurred

in many States where physicians’ offices were notcovered by State law. The ability of private ra-diology groups to make large capital purchasesenabled these circumventions around the CONlaws to succeed.

In the case of NMR imaging, it will be difficultto control these observed behaviors. A numberof hospitals have already acquired NMR imagersat significant cost, including siting and construc-tion costs for placement of the equipment. In prac-tical terms, it will be extremely difficult for ahealth planning agency to dislodge an NMR unit

from an existing site. Therefore, “franchising” hasalready begun and is likely to continue in someareas, at least in the near term. Without CONcoverage of physicians’ offices and other nonhos-pital settings, it will be virtu ally impossible to con-trol the diffusion of NMR imagers to privategroups who can raise the necessary capital. Thus,continued circumvention of CON regulation isequally likely to occur.

Finally, there is the lesson regarding the clini-cal utilization of X-ray CT scanning. Since itsearly diffusion, the clinical use of X-ray CT scan-

ning has evolved and matured. Over the years,physicians have experimented with the technol-ogy, compared it with alternative modalities, andonly now are beginning to understand its optimalapplication— i.e., when and how to use it as adiagnostic tool. NMR imaging is more complexand requires considerable expertise and skill onthe p art of the ph ysician. It will be som e timebefore the technology’s optimal clinical applica-tion will be understood even among the experts.The next few years will be a period of clinical ex-perimentation and learning, as physicians famil-iarize themselves with the technology and com-

pare it to other diagnostic imaging modalities,including X-ray CT scanning. The potential im-pact of NMR imaging on the future practice of medicine may prove to be as far-reaching as wasthe case with X-ray CT scanning in the past dec-ade. It may, thus, be appropriate to limit diffu-sion of the technology to selected sites—e.g.,clinical research or teaching centers or a limitednumber of community hospitals, where evalua-tion of its proper place in clinical medicine maybe conducted.

8/14/2019 8420


.

Ch. 9--Stat e Certif icate-of-Need Programs 111

CON ACTIVITIES RELATED TO NMR IMAGING DEVICES

NMR imaging is emerging as an issue of greatinterest and concern to many Health SystemsAgencies (HSAs) and State Health Planning andDevelopment Agencies (SHPDAs). There are in-dications that these agencies, which hold respon-

sibility for local and State CON review, respec-tively, are beginning to see increasing CONactivity related to NMR imaging. Consequently,individual agencies in several States are taking ag-gressive action toward convening expert taskforces, developing criteria and standards for CONreview, and conducting reviews of CON applica-tions already submitted by hospitals.

As a means of gathering information on CONactivities involving NMR imaging, the NationalHealth Planning Information Center (NH PIC) sentout “Program Information Letter 83-15” on July

8, 1983, to all health planning agencies requestingthat they provide information on the number of actual proposals already reviewed, the numberof anticipated reviews, criteria or guidelines forreview of NMR imagers, and enacted or pendingState legislation governing the placement of NMRunits (194).

By mid-September 1983, 27 SHPDAs and 30HSAs had responded to the NHPIC Program In-formation Letter, yielding a cross-sectional viewof CON activities currently under way in manyStates (195). Program activities relating to NMRimaging fall into three main categories: applica-tions review, criteria or standards development,and legislation or regulations adoption.

CON Reviews

As of September 12, 1983, 12 SHPDAs and 5HSAs in the sample had conducted a total of 33CON reviews for NMR imaging devices (195).SHPDAs reported review of 28 proposals with 16 approvals and HSAs reported 5 reviews with 3approvals. The total of 19 approvals excludeswaivers and exemptions for research applications.In addition, the agencies reported the receipt of 43 letters of intent or new prop osals: 18 b ySHPDAs and 25 by HSAs (195). Capsule s u m -maries of selected CON reviews appear below.

Missouri

Missouri became the first State to approveNMR imaging in a nonresearch, nonuniversity -affiliated hospital setting when its CON agency,the Health Facilities Review Board, approved theapplications of two community hospitals locatedin Columbia (64). The review board made thesecontroversial decisions despite SHPDA recom-mendations to the contrary (64). The reviewboard also rejected SHPDA recommendations forlimited diffusion of NMR imaging to only univer-sity hospital settings and for the formation of atask force to develop criteria and standards forCON review of the technology (99).

The two hospitals receiving CON approval areColumbia Regional Hospital, a 301-bed facilitythat is part of the Lifemark investor-owned hos-

pital chain, and Boone Hospital Center, a county-owned, nonprofit facility with 344 beds (64).

Illinois

In the absence of NMR criteria for CON re-view, the Illinois CON agency (the Health Facil-ities Planning Board) has invoked the technolog-ically innovative equipment clause of the StateCON law to limit the diffusion of NMR imagingto medical school affiliated hospitals (see later dis-cussion under legislation or regulations). As of August 1983, tw o hospitals affiliated with medi-

cal schools (Rush-Presbyterian-St. Luke’s Medi-cal Center in Chicago, and St. Francis’ Hospitalin Peoria) had applied for and received CON ap-proval for NMR imaging devices (72).

Nebraska

Beginning in late 1982 and early 1983, the StateCON agency received multiple applications forNMR from individual hospitals in Omaha. In aneffort to encourage cooperative planning, theState agency announced in the spring of 1983 thatit would “batch” NMR applications for simultane-ous review (205). Three private, nonprofit hos-pitals (Nebraska Methodist, Archbishop BerganMercy, and Children’s Hospitals) responded byforming a private corporation, NMR Inc., which

8/14/2019 8420


112 Health Technology Case Stud y 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, industrial, and Policy Ana lys is

submitted a single CON application to place anNMR imager in a freestanding facility where allthree hospitals would share access. In July 1983,NMR Inc. received CON approval for the acqui-sition of a superconducting NMR system. Alsoreceiving CON approval in July for NMR imag-ing was the University of Nebraska Hospital,

which has referral agreements with two other fa-cilities, Omaha Veterans’ Hospital and BishopClarkson Hospital (205). One other CON applica-tion for NMR was reviewed and recommendedfor denial; the hospital subsequently withdrew theapplication and is now seeking a cooperative ar-rangement with a second hospital (205).

Kentucky

Albert B. Chandler Hospital, a teaching affiliateof the University of Kentucky, currently has anNMR imaging unit, which was granted exemp-

tion from CON r eview as a research/ experimentaldevice (62). In May 1983, the State CON agencyreviewed and disapproved an application forNMR from Audubon Hospital, a Louisville facil-ity that is part of the Humana investor-ownedhospital chain. In making its decision, the SHPDAinvoked the Kentucky State Health Plan, signedby the Governor, which states that NMR tech-nology “. . . shall be considered a tertiary levelservice and approval of one unit will be consid-ered for each of the two designated tertiarycenters” 4 in the State (195). The CON decision wasappealed by the hospital and granted reconsidera-

tion by the SHPDA (62). Since Humana leases andmanages the tertiary center in Louisville (Hum anaHospital-University), the corporation argued thatplacement of an NMR imager at Audubon Hos-pital (a Humana-owned facility) would still per-mit patients from the university hospital to haveaccess to the technology. Following a public hear-ing, Audubon Hospital’s application was ap-proved by the SHPDA.

The University of Kentucky’s Albert B. Chand-ler Hospital has applied for CON approval to useits previously installed NMR unit for clinical, aswell as research, purposes.

“’Centers” refers to the two university hospitals in the state: AlbertB. Chandler Hospital in Lexington and Humana Hospital-Universityin Louisville (62).

Other jurisdictions in which CON applicationsfor NMR have been reviewed include: SHPDAsin Georgia, Tennessee, Texas, Ohio, Iowa, Ari-zona, Kansas, and California; and HSAs in Mid-dle Tennessee, New York City, North CentralGeorgia, Chicago, and Southeast Kansas.

CON Criteria or Standards DevelopmentAs of September 12, 1983, 10 health planning

agencies (HSAs and SHPDAs) had reported to theNational Health Planning Information Center thatthey had established NMR-specific review criteriaor guidelines. An additional 15 agencies are in theprocess of developing review criteria (195). Twoof these reported that they were using CON re-view criteria or standards for CT scanners as thebasis for their efforts in NMR imaging (158). Sev-eral agencies have also formed or are beginningto form expert task forces or advisory panels. A

brief summary of current State and local effortsin th is regard appears below.

Nebraska

The Statewide Health Coordinating Council(SHCC) in April 1983 authorized the formationof a 12-member Task Force on New Develop-ments in Diagnostic Radiology to develop NMRguidelines. The Task Force consists of sevenradiologists, one internist, one neurosurgeon, andone consumer member of the SHCC (205). In Sep-tember 1983, the Task Force submitted for review

a set of draft guidelines for NMR scanners. TheSHCC also created a separate Task Force on NewTechnological Developments, which prepared andsubmitted in September 1983 draft guidelines forreview of emerging technologies.

Massachusetts

The SHPDA in Massachusetts is working withthe State CON agency (the Determination of Needprogram) to develop criteria and guidelines forCON review of NMR (28). An Advisory Commit-tee on NMR is being formed, with representativesdrawn from the State Rate Setting Commission,the hospital industry, the professional medicalsocieties, and consumers. It is anticipated that theState may move toward limited diffusion of NMRimaging dur ing an initial research/ experimenta-tion phase (28).

8/14/2019 8420


—

Ch. 9—Stat e Certif icate-of-N eed Programs 113

In an apparently independent effort, the HealthPlanning Council for Greater Boston (the State’slargest HSA) developed proposed guidelines forNMR, which were expected to be adopted in finalform in December 1983 (79). The proposed guide-lines would allow NMR units to be placed in clin-ical, nonteaching settings under certain conditions(79).

Georgia

The Georgia SHPDA, with the aid of medicalspecialty societies and other professional groups,has convened a “blue ribbon committee” of ex-perts to develop NMR-specific criteria and guide-lines for review (75). The committee, which iscomposed of radiologists, nuclear medicine spe-cialists, internists, and hospital administrators,was expected to release a draft final report in thefall of 1983. The anticipated recommendations are

likely to urge caution, with NMR diffusion tem-porarily restricted to two medical schools pendingFDA approval and the articulation of reimburse-ment policies for NMR (75).

Oklahoma

The State CON agency in Oklahoma is nowin the process of assembling a Select Committeeon Technology to recommend criteria and stand-ards for NMR (25). Two avenues that will likelybe explored are the limited diffusion strategy of the Illinois CON program and the group applica-

tion/ shared-service model encouraged by theNebraska CON program (see the earlier discus-sions of both States’ experiences with CONreviews).

In addition to these CON programs, otheragencies involved in either task force developmentor criteria/ standard s developm ent include SHPDAs

in the District of Columbia, New Jersey, NorthCarolina, Illinois, Ohio, Maryland, H awaii, Florida,and Pennsylvania, and HSAs in SoutheasternMassachuset ts, Cen tral New Je rsey , EasternVirginia, North Central Georgia, Western Mich-igan, Central Arizona, and N orthwest Oregon(69,134,172,195). Several other agencies have de-veloped either NMR plans (Newark HSA, South-

east Kansas HSA) or position papers (SoutheasternPennsylvania HSA, Southwes tern Pennsylvania

HSA). Developmental activities of this nature areexpected to continue and expand in other areasof the Nation (13).

The American College of Radiology (ACR), amedical specialty society, has been contacted bymany State CON agencies requesting information

on NMR imaging. In response to these requests,the ACR (through its Commission on NMR, Sub-committee on Government Relations) prepared adocument, “Guidelines for Preparation of CONApplications, “ intended to assist State CON agen-cies in performing reviews of NMR-related CONapplications.

State CON Legislation/Regulations

During 1983, significant developments occurredin several States regarding CON legislation or reg-ulations that affect the review of NMR imag-

ing devices. A sampling of major developmentsfollows.

New York

The New York State Hospital Review and Plan-ning Council, the CON body in the State, draftedregulations that call for a 2-year demonstrationperiod in which NMR imaging will be restrictedto a select number of hospitals (127). During thisperiod, data on the technology’s safety, efficacy,and cos t ef fect iveness wil l be gathered and an-alyzed. Upon complet ion of the demons trat ion ,

a determination of need will be made and, pro-vided that nei ther cos t ef fect iveness nor qual i tyof care is at issue, all participants in the demon-stration as well as any other hospitals in the Statemay then apply for CON approval. The proposedregulations were expected to be reviewed and ap-proved by the council in the fall of 1983. Thecouncil made the decision in April 1984 to per-mit placement of NMR imagers in no more than13 teaching hospitals during the demonstrationperiod (19).

The application for NMR submitted by NewYork Hospital is generally credited with havingprecipitated this regulatory process (127). Th ehospital’s application was approved with the un-derstanding that it could not receive reimburse-ment for NMR unless it was selected to partici-

8/14/2019 8420


114 Health Technology Case Study 27: Nuclear Magnetic Resonance imaging Technology: A Clinical, Industrial, and Policy Anal ysis —

pate in the planned demonstration. The MajorMedical Equipment Committee of the Council willcommence development of NMR review criteriaor standards once the demonstration gets underway and preliminary data are produced.

Illinois

As alluded to earlier in the discussion of CONreviews, the Illinois Department of Public Healthand the Illinois Health Facilities Planning Boardadopted regulations on March 1, 1983, that setforth specific “Standards and Criteria for Reviewof Applications for Permit for Technologically In -novative Equipment or Innovative Programs”(195). These regulations stipulate that any suchequipment or programs must be restricted to onlymedical school settings until FDA premarket ap-proval for the technology is granted and CON

guidelines or criteria for review have been estab-lished. This effectively limits NMR diffusion toonly 11 hospitals in the State (one hospital permedical school). NMR imaging is among the firsttechnologies to be regulated u nd er these ru les (72).

District of Columbia

In September 1982, the CON law in the Districtof Columbia was amended to permit the designated

CON agency the right to declare a 180-day “hold-ing period” on the review of any new technology

whose safety, efficacy, and clinical use are notclearly understood or are in question (134). TheCON agency has, since that time, drafted generalcriteria and standards for CON review. Twoclauses specifically relate to new technology. Onerequires that app licants d emonstrate to the SHPDAdirector’s satisfaction that the technology isbeneficial in controlled trials. The second clauserequires applicants to demonstrate need for tech-nology acquisition relative to actual or potentialneed for such technology at other institutions inthe District. The agency received a CON applica-tion for NMR imaging from an area hospital in

February 1984 and immediately invoked the 180-day moratorium provision pending developmentof review criteria (203). The agency has conveneda technical advisory panel and expected to developcriteria and standard s for review of NMR imagersby August 1984.

New Jersey

In November 1983, the New Jersey SHCC ap-p roved a p ropos a l tha t w ou ld r es t r i c t CO N ap-proval of NMR imagers to no more than four sitesover an evaluation period of 2 years (69). In se-lecting the four sites, preference will be given to

the State’s 16 major teaching hospitals. Datagathered from the four installations wou ld be usedto guide decisionmaking on future NMR diffusionin New Jersey. In an unprecedented move, theState Commissioner of Health asserted that thesenew NMR regulations would apply to physicians’offices as well as other health care facilities (69).Under present New Jersey law, physicians in pri-vate practice are exempt from CON review (195).However, the State Department of Health viewsthe purchase of NMR by physicians’ groups as go-ing “far beyond the private practice of medicine”(69).

Utah

The CON law in Utah was amended in May1983 to exempt all medical equipment from CONreview (74). NMR imaging devices, therefore, willnot be subject to CON review in Utah, as the Stateappears to be pursuing a strategy of uncontestedapproval for all medical equipment purchases.

California

The State of California has also passed legisla-tion that eliminates the dollar thresholds for CON

review of major medical equipment purchases(201). As in Utah, NMR imaging devices will notbe subject to CON review in California.

Future Prospects

The general consensus among health plannersand CON agency staff members who were con-tacted for this study was that the level of CONactivity related to NMR imaging is likely to in-crease dramatically over the next year, OnceHCFA renders a policy decision regarding Medi-care coverage, planners expect to see a rapid in-

crease in the number of CON applications filedby hospitals around the Nation. s In anticipation

‘Several proposals are now before the Congress that would raisethe CON thresholds for equipment review to $1 million or higher.If such legislation was enacted in the next few months, some Statesmight follow suit and amend their statutes, effectively exemptingthe less expensive NMR systems from CON review.

8/14/2019 8420


Ch. 9—St ate C’ert~ficat e-of-Need Programs • 115 —

of this onslaught of paperwork, CON agencies atboth State and local levels in the national healthplanning structure are continuing to p ush forwardin the creation of special task forces and in thedevelopment of criteria and standards for CONreview.

Of the four strategies previously described forCON treatment of medical technologies, at leastthree appear to be operating with respect to NMRimaging. The New York, Illinois, Ohio, Kentucky,and New Jersey SHPDAs; and the Eastern Vir-ginia HSA all appear to be employing predeter-mined limits on diffusion, whereas the SoutheastKansas HSA and the District of Columbia SHPDAhave adopted moratoria on NMR pending furtherstudy and planning (38,69,117,195). Nebraska, bycontrast, is encouraging shared-service arrange-ments. Both Utah and California, at this time, ap-pear to be using a strategy of uncontested app rov-

al. No CON program, on the other hand, hasadopted a policy of pro forma denial.

are the site considerations that are unique to N MRimaging. Unlike the placement of CT scanners,NMR installation is costly and is likely to haveconsiderable impact both on hospital plant con-figuration and on the organization of staff. NMRplacement can be disruptive to the hospital, mak-ing internal management of the technology and

its use far more difficult than was (or is) the casewith CT. Shared service arrangements amonghospitals may prove fragile, owing to the fact thathost institutions may experience difficulty in ra-tioning the use of NMR imaging among partici-pants. Utilization of NMR units may increase tre-mendously as physicians discover new clinicalapplications and perform “sequential scanning. ”Should NMR imaging come to be used in thisway, hospital administrators will find it difficultto ration NMR use among medical staff members,let alone among other hospitals. Interspecialtydisputes over the use of NMR imaging may fur-

ther cloud the issues of approp riate utilization andrationing.

NMR imaging is likely to differ from the CTexperience in several ways. First, and foremost,

8/14/2019 8420


Appendixes

8/14/2019 8420


Appendix A.— NMR: Technical Background

NMR Spectroscopy—TechnicalBackground

Ordinarily the magnetic moments of hydrogen atoms(the vector representations of the net magnetic prop-erties of hydrogen atoms) will point in random direc-

tions.

1

When exposed to an externally applied mag-netic field, however, these magnetic moments tend toalign themselves either parallel to or antiparallel to themagnetic field.2 Because the energy state of a hydro-gen nucleus is lower when its magnetic moment ispointing in the direction of the applied magnetic fieldthan w hen it is pointing in the opp osite direction, moreof the magnetic moments will point in the directionof the applied field than in the opposite direction. Inquantum mechanical terms, hydrogen nuclei exposedto an externally applied magnetic field, B, will residein one of two possible magnetic energy levels (see fig.A-l). For hydrogen nuclei to go from the lower energylevel, E1, to the higher energy level, E2, an E2 – E1

amount of energy needs to be added to the system.When hydrogen nuclei go from the higher energy levelto the lower energy level, in contrast, E2 – E1 energyis emitted from the system.

) For slmpl]c it y, the remainder of this d]scuss]t]n WIII perta]n to hydrogen,which has a spin of one-half ( a quant urn nuclear spin number ot one-half

to describe the rotation of the hydrogen nucleus I‘Other forces, such as thermal agltatlon, will Influence the ortentat ion of

these magnetic moments. but cons] c!erat]on of such forces IS beyond the scope

o f th]~ case study

Figure A-1 .—Energy Levels of Hydrogen Atoms in

Applied Magnetic Fields

BI

According to quantum theory, hydrogen nuclei willmove from El to E2 only when exactly the right amountof energy (namely E2 — E1) is applied to the system.In NMR experiments, excitational energy is providedin the form of radiofrequency (RF) waves. In order forhydrogen nuclei residing in energy level E 1 to be ex-cited into level E2 by such radiofrequency energy, the

RF waves must be applied at exactly the same rota-tional frequency as that at which the magnetic nucleiin El are processing.3 This frequency is known as theLarmor frequency.

When an NMR system is appropriately constructed(see fig. A-2), the energy emitted by excited hydrogennuclei during the process of relaxation can be detectedas an NMR signal. The intensity of the signal will beproportional to the number of hydrogen nuclei in thesample being studied.

Variations in the local magnetic field give rise tovariations in the frequency at which nuclei spin. Thesespin differences depend on the exact molecular envi-ronment in which each nucleus exists. This observed

“shift” in nuclear precessional frequency, and there-fore the radiofrequency at which NMR signals are de-tected, has been termed “chemical shift. ” These slightvariations in NMR signals induced by variations inmolecular environment provide the basis for NMRspectroscopy and its use in acquisition of informationabout molecular structure and conformation.

Application of NMR Principles toImaging—Technical Background

A decade ago, scientists realized that exciting pos-sibilities emerged if, instead of applying a uniform

magnetic field across an experimental sample, theyestablished a magnetic field gradient across it (see fig.A-3). In so doing, the strength of the magnetic field,B, varies from one line (L1) to another (L2). Becauseof the one-to-one relationship between the magneticfield strength in which magnetic nuclei exist and thefrequency of radiofrequency energy that will excitethem, a pulse of radiofrequency energy applied to thesystem depicted in figure A-3 will be absorbed (andsubsequently re-emitted) only by nuclei that are lo-cated in the vertical line LX of magnetic field strengthBx that corresponds to the Larmor frequency of theenergy being applied. By sequentially varying the fre-quency of the energy being supplied, one can thus

selectively excite nuclei, line by line. The establishmentof a magnetic field gradient across a sample thus “spa-

‘ [ ’ reces s Io n IS a term descrlbln~ the mot](ln of a proton ]n an ex te rn a lSOURCE E P Steinberg, Johns Hopkins Medical Instltutlons, Baltlmore. MD ,

1983magnet]c t leld or of a top In th e Earth s gra~’]tat Ional t]eld

119

25-341 0 - 84 - 9 : QL 3

8/14/2019 8420



Figure A-2.—Schematic Diagram of NMR System

The radiotransmitter provides radiowaves at the appropriaterotational frequency to excite the proton. When theradiowaves are turned off, the excited protons precess inphase, thereby producing radiowaves that are detected with

a radio receiver and then displayed.SOURCE E. P Steinberg, Johns Hopkins Medical Institutions, Baltimore. MD, 1983

tially encodes” the NMR information inherent in a sys-tem. This type of spatial encoding of the origin of aNMR signal forms the basis of NMR imaging.

Radiofrequency Pulse Sequences

As explained earlier, NMR signals are obtainedthrough the excitation of magnetic nuclei with RFwaves. In a typical NMR experiment, the nuclei to beexcited (and imaged) are repeatedly exposed to RFwaves in a pulsed fashion. Specific “pulse sequences”are characterized by the duration and intensity of each

pulse used, as well as by the interval between repeatedpulses.

Several different types of pulse sequences can be uti-lized to produce NMR images. Probably the three mostcommon types of sequences currently employed inNMR imaging are Saturation Recovery, InversionRecovery, and Spin-Echo sequences (27,55,77). Al-though the exact technical details of how these se-quences are produced are beyond the scope of this re-port, it should be mentioned that Saturation Recoveryand Inversion Recovery sequences result in NMRsignals (and therefore images) that tend to reflect

8/14/2019 8420


— .

App . A–NMR: Techn ica l Background 121 —

Figure A-3.— Experimental Sample in Magnetic Field Gradient

Experimentalsample/

Bo B1B2

Magnetic field strength (B)

(Magnetic field strength in line LO is Bo; in line L1 is B1; andin line L2 is B2.)

SOURCE E P Steinberg, Johns Hopkins Medical Instltutlons, Baltlmore, MD, 1983

predominantly the T1 character of tissues (i. e., theyare T1 weighted), whereas Spin-Echo sequences tendto reflect m ore T2 information (i.e., they are T2 weighted).NMR images will thus vary depending upon the par-ticular pulse sequences used to produce them. Con-siderable effort is being expended on trying to definepulse sequences that optimally demonstrate differenttypes of morphologic and pathologic abnormalities.This task is formidable, given the infinite number of pulse sequences that can be used.

Techniques for Spatial Encoding

and Image Acquisition

Several different techniques have been developedthrough which NMR information can be spatially en-coded, acquired, and transformed into an image. Thefirst such technique, adapted from use in X-ray com-

puted tomography (CT) reconstructions, was the pro- jection-reconstruction method of NMR imaging usedby Lauterbur (115). In this technique, magnetic gra-dients are electronically rotated to produce multipleprojections of the sample being imaged. These projec-tions are then “pieced together” with th e help of a com-puter to construct an NMR image. Imaging methodsdeveloped su bsequently employ a variety of techniquesto “selectively excite” nuclei. These techniques haveacquired names based on whether they excite nucleion a point-by-point (sequential point methods), line-by-line (sequential line methods), plane-by-plane(planar imaging) or whole-volume (three-dimensionalimaging) basis (8). The latter techniques require lesstime to form images (8). The recently developed echo-planar method of imaging permits scans to be obtainedin milliseconds, raising the possibility of dynamic, orreal-time, NMR imaging (133).

8/14/2019 8420


Appendix B. —Survey of Manufacturers: Methods

To gather the information required for this report,a su rvey of NMR-imaging-device manu facturers w ascondu cted involving th e following steps:

1.

2.

3.

With the assistance of the National ElectricalManufacturer’s Association (NEMA), appropri-ate contacts (either the director of marketing or

the technical director for NMR imaging) wereidentified in each of the firms represented byNEMA. Potential contact p ersons in non-NEMA

companies were identified through other sources.A total of 20 companies were targeted for survey.Letters of introduction describing the study andits purpose w ere mailed to all identified man ufac-turers in June 1983. Manufacturers were notifiedof our intent to contact them by telephone to ar-range a mutually convenient time for a telephoneinterview.Between June and August, 15 manufacturers wereinterviewed. Interviews varied in length, but theaverage d iscussion lasted ap proximately 1 to 1½

hours. In each case, the discussion was structuredaround the following key issues:

History of the firm’s program in NMR imag-ing—its genesis and development to the pre-sent dayCurrent status of the manufacturer’s NMRimaging systems—including magnet type,magnet field strength, and imaging capa-bilitiesC l in ica l p lacemen ts o f N M R imag ingsystems—by site, system capabilities, anddate of installation

4.

5.

Collaborative relationships with universitiesand/ or med ical schools for NMR-imagingR&DCompany characteristics—size, ownership,staff composition, and product linesThe future market for NMR imaging sys-

tems—projected growth, competitive tech-nologies (e.g., CT, ultrasound, etc.), and keyfactors influencing NMR diffusionCosts of NMR-imaging devices—likely cap-ital acquisition costs, annual operating ex-penses once installedFDA policies—premarket approval for ClassIII devicesThird party payment policies—HCFA/ Med-icare coverage d ecisions, Blue Cross/ BlueShield decisions, commercial insurers’ deci-sions, prospective payment systemsState certificate-of-need policiesFederal role in funding applied R&DPatentsFuture p lans for NMR-imaging development

Follow-up telephone calls were made for clarifica-tion of responses, as necessary.Upon submission of the draft final report to OTAin early September, each of the 15 participatingmanufacturers was invited by OTA to review andcomment on its respective company descriptioncontained in appendix C to this report. In addi-tion, representatives of NEMA were invited byOTA to review the full draft report.

122

8/14/2019 8420


Appendix C.— Manufacturers of NMR Imaging Devices

The following company descriptions of NMR-imaging-device manufacturers are based on interviews withrepresentatives from each firm. The initial interviews took place in 1983 and information was updated in August1984. Magnet field strength is stated in kilogauss (kG). Conversion of kilogauss to Tesla units is as follows: 10kG= 1.0T.

ADAC LABORATORIES

4747 Hellyer AvenueSan Jose, CA 95138

Background: ADAC Laboratories is an independent publicly owned company. The company decided to investin R&D efforts for NMR in 1982. ADAC is in the early stages of NMR-imaging-system development. It expectsto have an engineering model available in early 1986 and a commercial prototype system in the second half of 1986.

Current NMR-imaging models: None (permanent magnet prototype is in development)Collaborative arrangements with universities or medical centers: Negotiated, yet to be announced.Clinical placement sites: NoneNumber of employees engaged in NMR imaging: 40 (including magnet developers)Other diagnostic imaging products: Nuclear medicine, digital radiography; conventional X-ray; and fluoroscopeOther medical products: Radiation-therapy planning; special procedures room; clinical information systems;medical linear accelerators

Non-health-care-related products: Instruments for nondestructive testing

BRUKER MEDICAL INSTRUMENTS, INC.Manning Park

Billerica, MA 01821(Head Office: West Germany)

Background: Bruker Instruments is a privately owned subsidiary of Bruker A.M. of West Germany. The com-pany began work on NMR spectroscopy in 1961 and developed extensive magnet technology and experiencewith pulse and spectroscopy techniques. Bruker made its first commitment to NMR imaging in 1977, and by1979, had completed its first experimental prototypes. In 1982, Bruker placed its first NMR-imaging unit in anoutside clinical setting. In 1983, the company acquired Oxford Research Systems, which specializes in animalresearch systems, from Oxford Instruments and plans to build superconducting magnets with its new subsidiary.Bruker had its first marketing prototype available for placement in 1983.

Current NMR-imaging models:

Year firstMagnet type Field strength Bore size available

Resistive 1.3 kG Whole bod y 1979Superconducting 47 kG Anim al 1979Superconducting 19 kG Animal or head only 1982Resistive (self-shielded) 2.4 kG Whole bod y 1984

Collaborative arrangements with universities or medical centers: 21. Baylor College of Medicine, Houston, TX2. Yale University, New Haven, CT

Clinical placement sites:

Date of installationHospital or clinic NMR imaging system (E = expected)

1. Baylor College of Medicine, Resistive, 1.3 kG, whole body 1982Houston, TX

2. Baylor College of Medicine, Superconducting, 47 kG, animal 1982Houston, TX

3. Japan Resistive, 1.3 kG, whole body 1982

123

8/14/2019 8420


124 Health Technology Case Study 27: Nuclear Magnetic Resonance imaging Technology: A Clinical, Industrial, and Policy Analysis — ———

4, Brigh am & Wom en ’s H osp ital, Resistive, 1.3 kG, wh ole bod y 1983Boston

5. Br igh am & Wom en ’s H osp ital, Su p er co nd u ct in g, 19 k G, a nim al 1983Boston

6. D.K.D. H osp ita l, W iesba den , Resistive, 1.3 kG, w hole bod y 1983West Germany

7. Yale University, New Haven, CT Su percond ucting, 15 kG, w hole 1985EbodyNumber of employees engaged in NMR imaging: 50Other diagnostic imaging products: NoneOther medical products: Parent firm m akes ECG monitors, mobile d efibrillator, and patient m onitoring systemsNon -health-care-related prod ucts: NMR sp ectrometers

CGR MEDICAL CORP.2519 Wilkins Avenue

Baltimore, MD 21203

(Head office: Paris, France)

Background: CGR Medical Corp. is a private, wholly owned subsidiary of Thompson-Brandt of France. Thecompany was created in 1971 by the acquisition of Westinghouse Medical X-Ray Division by CGR of France,

which merged in 1979 with Thompson-CSF to form Thompson-Brandt. CGR Medical Corp. decided to investin R&D efforts for NMR imaging in 1979. Its first engineering models were available in 1982. The company ex-pects to place its first NMR imaging unit in a clinical setting and to have available for placement its first commer-cial system in 1984.



Resistive 1.5 kG Whole bod y 1982Su percondu cting 3,5 kG Whole bod y 1983Superconducting 5 kG Whole bod y 1983

Collaborative arrangements with universities or medical centers: None in USA (nu mber in Europe not available)

Clinical placement sites: NoneNumber of employees engaged in NMR imaging: approximately 150

Other diagnostic imaging products: Computed tomography; ultrasonography; nuclear medicine; digitalradiography; conventional X-ray and fluoroscope

Other medical products: NoneNon-health-care-related products: Parent firm makes assorted electrical appliances and equipment

DIASONICS INC.

NMR Division

533 Cabot Road

South San Francisco, CA 94080

Background: Diasonics is an independent, publicly owned company. Initial R&D on NMR imaging began in1975 as a University of California, San Francisco (UCSF) project with outside funding. In 1976 Pfizer Corp. began

funding the work. In 1981, Diasonics purchased the rights to all patentable NMR technology developed underthe UCSF-Pfizer agreement. Diasonics had its engineering model available in 1981, and the company made itsfirst clinical placement of an NMR imaging unit the same year. Its first commercial prototype system becameavailable for placement in 1983.

8/14/2019 8420


App. C—Manu facturers of NM R Imaging Devices 125



Superconducting 5 kG a Whole bod y 1981a

System Operating at 3.5 kilogauss; probable commercial prototype system; commitment to upgrade to higher fields if Diasonics believes

clinical relevance demonstrated.

Collaborative arrangements with universities or medical centers: 3

1. University of California, San Francisco2. University of Texas, Dallas3. University of Michigan, Ann Arbor

Clinical placement sites:Date of installation

Hospital or clinic NMR imaging system (E = expected )

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

University of California, SanFranciscoHu ntington Med ical Research

Institutes, Pasadena, CAUniversity of Texas Health ScienceCenter, DallasPrivate radiology clinic, NJ

University of Michigan, Ann Arbor

St. Anthony’s Professional BuildingSt. Petersburg, FLUCSF-Radiation InstrumentationLaboratory, San Francisco, CAMontclair Radiological Association,PA, Montclair, NJUniversity of Texas Health ScienceCenter, DallasNMR Associates, Houston, TX

Private clinic, Wuppertal, West

GermanyInstitute of Radiology, Geneva,SwitzerlandRoentgen Institut, Dusseldorf, WestGermanyNMR Imaging, Torrance, CA

Long Island MRI, New Hyde Park,N YNortheast Medical Center, Ft.Lauderdale, FLMagnetic Resonance Images, Inc., St.Petersburg, FLDiagnostic Imaging Center,

Lausanne, SwitzerlandHeart to Heart, Phoenix, AZ

San Jose MRI, San Jose, CA

Superconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, whole

bodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 20 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 20 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, whole


bodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebody

1981

April 1983

September 1983

1983

1983

1984

1984

1983

1984

1984

1984

1984

1984

1984

1984

1984

1984

1984

1984

1984

8/14/2019 8420


126 Health Technolog y Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analysis

21. NMR Scan Center, Ft. Lauderdale, Sup ercond ucting, 5 kG, w hole 1984FL body

22. NMR Imaging, Santa Ana, CA Sup ercond ucting, 5 kG, 1984whole body

23. Magnetic Resonance Center of San Su percond ucting, 5 kG, w hole 1984Diego, San Diego, CA body

Number of employees engaged in NMR imaging: 152Other diagnostic imaging products: Ultrasonography and surgical C-arm imaging equipmentOther medical products: NoneNon-health-care-related products: None

ELSCINT LTD.Head Office: Haifa, Israel

U.S. Subsidiary: Elscint Inc.

930 Commonwealth Avenue

Boston, MA 02215

Background: Elscint is an independent, publicly owned company in Israel. The company decided to invest inR&D efforts for NMR imaging in 1981, and by 1982, had developed its first engineering model. Elscint producedits first prototype NMR imaging system in 1983 and expects to have a second prototype in late 1984. The com-pany made its first clinical placement outside the company’s plant in November 1983.



Superconducting 5 kG

Collaborative arrangements with universities or1. Hebrew University, Jerusalem, Israel2. Weitzman Institute, Rehovoth, Israel


Hospital or clinic

Whole body 1982

medical centers: 2

Date of installationNMR Imaging system (E = expected)

1. Skokie Valley Imaging, Skokie, IL Superconducting, 5 kG, whole 1983body

2. Fondren Imaging, Houston, TX Superconducting, 5 kG, whole 1984body

3. Private clinic, Freiburg, West Superconducting, 5 kG, whole 1984Germany body

4, Herzlyia MRI Clinic, Herzlyia, Israel Superconducting, 5 kG, whole 1983body

Number of employees engaged in NMR imaging: Not availableOther diagnostic imaging products: Computed tomography; ultrasonography; nuclear medicine; digital

radiography; conventional X-ray; and fluoroscopeOther medical products: N oneNon-health-care-related products: None

8/14/2019 8420


FONAR

App. C—Manufacturers of NMR imaging Devices 12 7

CORP.

110 Marcus DriveMelville, NY 11747

Background: Fonar has been an independent, publicly owned corporation since 1981. Founded originally as theRAANEX Corp., it has invested in R&D efforts for NMR imaging since 1978. In 1980, Fonar completed its firstexperimental prototype and made its first clinical placement of a unit outside the plant. The firm’s first commer-cial prototype system became available for placement in 1983.



Permanent 0.4 kG Whole bod y 1980Permanent 3 kG a Whole body 1983Permanent 3 kG a Whole body (mobile) 1983

aProbable Commercial prototype system(s).

Collaborative arrangements with universities or medical centers: 11. University of California at Los Angeles (UCLA)


Hospital or clinic N MR imaging system (E = expected )1.

2 .

3.

4.

5.

6.

7.

8.

9.

10.11.12.

13.14.15.

16.

17.

Diagnostic Imaging Associates,Cleveland, OH

Hospital Universitario, Monterey,Nuevo Leon, MexicoSan Raffaele Hospital, Milan, ItalyNakatsugawa Hospital, Nagoya,JapanBrunswick Memorial Hospital,Amityville, NYUniversal NMR, Inc. (mobilescanner )

UCLAMontvale Diagnostic Imaging Center,Montvale, NJNeurodiagnostic Center, New York,N YChicago Medical School, Chicago, ILNMR Centers, Inc., Los Angeles, CAParkview Hospital, Nashville, TN(Hospital Corporation of America)Loyola University, Chicago, ILMercy Hospital, Altoona, PAAMD (Advanced MedicalDiagnostics), Melbourne, FLNMR Investors, Inc., Santa Monica,CAOdessa Diagnostic Imaging Center,

Permanent, 0.4 kG, whole body

Permanent, 0.4 kG, whole body

Permanent, 0.4 kG, whole bodyPermanent, 0.4 kG, whole body

Permanent, 3 kG, whole body


Permanent, 3 kG, whole bodyPermanent, 3 kG, whole body


Permanent, 3 kG, whole bodyPermanent, 3 kG, whole bodyPermanent, 3 kG, whole body

Not availablePermanent, 3 kG, whole bodyPermanent, 3 kG, whole body


Permanent, 3 kG, whole bodyOdessa, TX

Number of employees engaged in NMR imaging: 100

December 1980 (Thisunit is no longer in

place)April 1981

March 1982July 1982

October 1983

October 1983

19841984

1984E

19841984E1984E

Not available1984E1984E

1984E

1984E

Other diagnostic imaging products: NoneOther medical products: NoneNon-health-care-related products: None

25-341 0 - 84 - 10 : QI, 3

8/14/2019 8420


128 Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analy sis

GENERAL ELECTRIC CO.

Medical Systems Business Group

P. O. Box 414

3000 Grandview Avenue

Milwaukee, WI 53201

Background: General Electric is a publicly owned, multiproduct company. The Medical Systems Business Groupis responsible for NMR imaging R&D. Early R&D work in phosphorus spectroscopy began in 1978, but firm

corporate commitment to NMR imaging was not made until 1980. In 1982, General Electric completed its firstengineering model and made its first clinical placement of an NMR imaging unit outside the company’s plant.General Electric expects to have its first commercial prototype available for placement in late 1984.


Year first available

Magnet type Field strength Bore size (E = expected)

Su percondu cting 15 kG a Whole body 1984Ea

Probable commercial prototype.

Collaborative arrangements with universities or medical centers: 41. University of Pennsylvania, Philadelphia, PA2. Medical College of Wisconsin, Milwaukee, WI

3. Yale University, New Haven, CT4. Duke University, Durham, NC


Hospital or clinic NMR imaging system (E = expected)

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

University of Pennsylvania Hospital,Philadelphia, PAYale University School of Medicine,New Haven, CTDuke University, Durham, NC

Memorial Sloan Kettering CancerCenter, New York, NYMedical College of Wisconsin,Milwaukee, WIPittsburgh NMR Institute, Pittsburgh,PAUniversity of Illinois, Chicago, IL

University of Nebraska, Omaha, NE

Henry Ford Hospital, Detroit, MI

Ohio State University, Columbus,OHMichigan State University, Ann Ar-bor, MINew York Hospital—Cornell Univer-sity, New York, NYRochester Consortium, Rochester,NY

Resistive, 1.2 kG, whole body


Superconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebody

October 1982

August 1983

February 1984

August 1985E

October 1984E

November 1984E

February 1984E

April 1985E

January 1985E

November 1984E

February 1985E

December 1984E

November 1984E

8/14/2019 8420


App. C—Manufacturers of NMR Imaging Devices 12 9.

14.

15.

16.

1 7 .

18.

19.

20.

State University of New York atAlbany Medical School, Albany, NYUniversity of Texas Health ScienceCenter, San Antonio, TXUniversity of Texas/ HermanHospital, Houston, TX

Driscoll Children’s Hospital, CorpusChristi, TXUniversity of Washington, Seattle,WAStanford University, Palo Alto, CA

University of Western Ontario, Lon-don, Ontario, Canada

Su percon du ctin g, 15 kG, w hole March 1985EbodySu percon du ctin g, 15 kG, w hole March 1985EbodySu percon du ctin g, 15 kG, w hole March 1985Ebody

Su percon du ctin g, 15 kG, w hole March 1985EbodySuperconducting, 15 kG, whole December 1984EbodySu p er co nd u ct in g, 15 k G, w h ole D ece mb er 1984EbodySu p er co nd u ct in g, 15 k G, w h ole Febr uar y 1985Ebody

Number of employees engaged in NMR imaging: in excess of 500Other diagnostic imaging products: Computed tomography, ultrasonography, nuclear medicine, digital

radiography, conventional X-ray, and fluoroscopeOther medical products: Assorted electromedical equipmentNon-health-care-related produ cts: Assorted electrical appliances and equipment

JEOL USA, INC.

235 Birchwood Avenue

Cranford, NJ 07016

(Head office: Japan)

Background: JEOL USA is a publicly owned subsidiary of JEOL of Japan. JEOL has been manufacturing NMRspectrometers since 1960. In 1973, the parent firm was acquired by Mitsubishi. In 1982, the firm began investingin NMR imaging and spectrometry R&D. JEOL has decided not to pursue the clinical NMR market, and willinstead focus on the research and experimentation market. The firm expects to have its first engineering modelavailable in 1984.

Current NMR-imaging models: NoneCollaborative arrangements with universities or medical centers: N one

Clinical placement sites: N oneNumber of employees engaged in NMR imaging: 5

Other diagnostic imaging products: NoneOther medical products: Radioimmunoassay equipment, blood gas analyzers; fluid analyzersNon -health-care-related produ cts: NMR spectrometers, Parent firm also makes electron m icroscopes.

M&D TECHNOLOGY LTD. *

Unit 1, Whitemyres Avenue

Aberdeen, Scotland, U.S. AB2-6HQ

Background: M&D Technology is an independent, private company in Scotland. M&D was formed in 1982 tocommercially develop the NMR imaging system that had evolved from the work of Professor Mallard at Aber-deen, Scotland since 1974. M&D’s first engineering model became available in 1982. In 1983 the company madeits first clinical placemen t of a NMR imaging unit. Also, for a short time during 1983, M&D had entered into

a marketing agreement with Fischer Imaging corporation. M&D expected to have a marketing prototype systemavailable in 1984.

8/14/2019 8420





Resistive 0.4 kG Whole bod y 197?Resistive 0.8 kG Whole bod y 1982


1. University of Aberdeen, Scotland, U.K.Clinical placement sites:

Date of installationHospital or clinic N MR imaging system (E = expected )

1. Edinburgh Royal Infirmary, Resistive, 0.8 kG, whole body 1983Edinburgh, U.K.

2. Private Clinic, Geneva Resistive, 0.8 kG, w hole bod y 19833. Two additional sites End of 1983E

Number of employees engaged in NMR imaging: Not availableOther diagnostic imaging products: N oneOther medical products: N oneNon-health-care-related products: None

*Information on M&D Technology Ltd. is as of October 1983.

NALORAC CRYOGENICS CORP.1717 Solano Way, Suite #37

Concord, CA 94520

Background: Nalorac Cryogenics is an independent, privately owned company, which was founded in 1975 byDr. James Carolan to develop and manufacture superconducting magnets, cryogenic devices, and NMR systems.In 1977, the company was acquired by Nicolet Instrument Corp. In 1981, Dr. Carolan purchased the companyback from Nicolet and reaffirmed its commitment to developing NMR-imaging magnets and systems. The com-pany currently manufactures superconducting high resolution NMR magnet systems with bore diameters from50 to 320 mm. and field strengths from 20 kG to 70 kG. The company is presently developing a complete imag-ing spectrometer system which is scheduled for introduction in early 1985,

Projected NMR-imaging models (spectrometer systems):


Superconducting 20-40 kGSuperconducting 20 kGSuperconducting 10 kG

Collaborative arrangements with universities or medicalprior to late 1984.


Animal (330 mm) 1985EPed iatric (450 mm) 1986EHead/ appendage (600 mm) 1987E

centers: No formal arrangements will be announced

Date of installationHospital or clinic NMR imaging system (E = expected)

1. University of California, San Su percond ucting, 60 kG, 100 January 1984Francisco mm bore

2. University of Texas Health Science Supercond u cting, 20 kG, 320 June 1984Center, Houston, TX mm bore


Other diagnostic imaging products: N oneOther medical products: None

Non-health -care-related p roducts: Sup ercondu cting high resolution an alytical NMR magn ets, gradient coils, pow ersupplies, dewars, NMR probeheads

8/14/2019 8420


8/14/2019 8420


132 Health Technology Case Stu dy 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, industrial, and Policy Analysis

12.

13.

14.

15.

16.17.

18.

Casa di cura “Pio X,” Milano, Italy

Erasmus Ziekenhu is/ Free University,Brussels, BelgiumUniversitaetsklinik Koeln Cologne,West Germany

Akademisch Ziekenhuis Leiden,Leiden, The NetherlandsNeuro Besta, Milano, ItalyMontreal Neurological Institute,Montreal, CanadaCentro Diagnostic Immagini Com-

Superconducting, 5 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 5 kG, wholebody

Superconducting, 5 kG, wholebodyResistive, 1.5 kG, whole bodySuperconducting, 15 kG, wholebodySuperconducting, 5 kG, whole

puterizzate, Catania, Italy bodyNumber of employees engaged in NMR imaging: 30–North American Philips (U.S.);

Netherlands)

September 1984E

October 1984E

October 1984E

November 1984E

November 1984EDecember 1984E

December 1984E

80—N.V. Philips (The

Other diagnostic imaging products: Computed tomography; ultrasonography; digital radiography; conventional

X-ray and fluoroscope; nuclear medicinec

Other medical products: Assorted electromedical equipment. Parent firm also produces surgical supplies and

dental equipment

Non-health-care-related products: NMR spectrometers. Parent company produces assorted electrical appliances

and equipmentc

Nuclear medicine imaging products distributed by ADAC in the United States and by N.V. Philips in the rest of the world.

PICKER INTERNATIONAL, INC.

595 Miner Road

Highland Heights, OH 44143

(Head office: United Kingdom)

(Corporate headquarters: Ohio)

Background: Picker International is a U.S. corporation operating as an 80-percent-owned subsidiary of the GeneralElectric Company (P. L.C.) of England (GEC). The company was formed in April 1981 through the acquisitionof Picker from RCA in combination with GEC Medical and Cambridge Medical Instruments. GEC had earlieracquired the NMR technology of EMI of England and, by the end of 1981, the first Picker International NMRunit was clinically operating. In 1983, the first commercial units were shipped.


Year firstMagnet type Field Strength Bore size available

Resistive 1.5 k Ga Whole body 1978Superconducting 3 k G Whole bod y 1981Sup ercond ucting 5 k Ga Whole bod y 1983

a

Probable commercial prototype.


1.

2.3.4.

5.6.7.

8.9.

10.11.

University of Nottingham, Nottingham, U.K.Royal Postgraduate Medical School and Hammersmith Hospital, London, U.K.Mount Sinai Hospital, Cleveland, OHMayo Clinic, Rochester, MN

Bowman Gray Medical School, Winston-Salem, NCUniversity of British Columbia, Vancouver, CanadaCity of Faith Medical and Research Center, Tulsa, OKNational Institutes of Health, Bethesda, MDUniversity of Iowa, Iowa City, IAQueens Square Hospital, London, U.K.National Heart Institute, London, U.K.

8/14/2019 8420


App. C–Manufacturers of NM R Imaging Devices 133 —


Date of installation

Hospital or clinic NMR imaging system (E = expected )

1.

2.

3.

4.

5.6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.24.25.26.

27.

University of Nottingham Hospital,Nottingham, U.K.HIRST Research Center, London,U.K.Queens Medical Center, Nottingham,U.K.Hammersmith Hospital, London,U.K.Mount Sinai Hospital, Cleveland, OHMayo Clinic, Rochester, MNUniversity of British Columbia, Van-couver, British Columbia, CanadaUniversity of Manchester,Manchester, U.K.Bowman Gray Medical School,Winston-Salem, NCDr. Wallnhofer, Private Clinic, 6500Mainz 1, Mun ich, West German yDr. Assheuer, Private Clinic, 500Koln 80, Cologne, West GermanyCity of Faith Medical and ResearchCenter, Tulsa, OKNational Institutes of Health (NIH),Bethesda, MDUniversity of Alabama, Birmingham,ALDuarte CT, Duarte, CA

Neurology Center, Washington, DC

University of Iowa, Iowa City, IA

Queens Square Hospital, London,U.K.National Heart Institute, London,U.K.Private clinic, Cologne, WestGermanyPrivate clinic, Frankfurt, WestGermanyChiba University, Chiba City, Japan

First Hill Diagnostic, Seattle, WAGlasgow Hospital, Glasgow, ScotlandShinsuma University, Cobe, JapanPicker Clinical Research Center,Cleveland, Ohio

Picker Clinical Research Center,Cleveland, OH




Superconducting, 3.5 kG, wholebody (operating at 1.5 kG)Resistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole bodySuperconducting, 3 kG, wholebodySuperconducting, 3 kG, wholebodyResistive, 1.5 kG, whole body

Sup erconducting, 3 kG, wh olebodySup erconducting, 3 kG, wh olebodySup ercondu cting, 5 kG, wh olebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodySup ercondu cting, 5 kG, wh olebodySuperconducting, 5 kG, wholebodySup ercondu cting, 5 kG, wh olebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodyResistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole body

Superconducting, 5 kG, wholebody

1978

November 1978

February 1981

March 1981

October 1982December 1982

March 1983

March 1983

June 1983

July 1983

October 1983

1983

1983

1984E

1984

1984

1984

1984

1984

1984

1984

1984

1984198419841984

1984

8/14/2019 8420


134 Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Analy sis

28. Picker Clin ical Research Center, Sup ercond u cting, 15 kG, w hole 1984Cleveland, OH body

29. H IRST Research Center, Lond on, Supercond ucting, 20 kG, w hole 1984U.K. body


Other diagnostic imaging products: Computed tomography; ultrasonography; nuclear medicine; digital

radiography; conventional X-ray; and fluoroscope

Other medical products: Electrocardiogram equipment. Parent company makes other electromedical equipment

Non-health-care-related products: None

SIEMENS MEDICAL SYSTEMS, INC.

186 Wood Avenue South

Iselin, NJ 08830(Head office: West Germany)

Background: Siemens Medical Systems is a publicly owned subsidiary of Siemens A.G. of West Germany. Asearly as 1965, Siemens had a research group in NMR working on blood flow and blood viscosity. In 1978, thecompany made a commitment to develop NMR-imaging systems. The first engineering model was put into opera-tion in 1980. A year later, Siemens made its first clinical placement of an NMR-imaging unit outside the com-pany’s plant. The company had a commercial prototype system available for placement in 1983.



Resistive 1.2 kG Whole body 1980Resistive 2 kG Whole body 1981Superconducting 5 kG a Whole body 1983Superconducting 15 kG a Whole body 1983

a

Probable commercial prototypes.

Collaborative arrangements with universities or medical centers: 61. Wash ington Univer sity, St. Louis, MO2. University of Hanover Medical Center, Hanover, West Germany

3. Radiological Institute, Frankfurt, West Germany4. Radiological Institute, Munich, West Germany5. Mount Sinai Medical Center, Miami, FL6. Allegheny General Hospital, Pittsburgh, PA



1.

2.

3.

4.

5.

University of Hanover MedicalCenter, Hanover, West GermanyRadiological Institute, Munich, WestGermany (Dr. Heller)Mallinckrodt Institute, WashingtonUniversity, St. Louis, MORadiological Institute, Frankfurt,West Germany (Dr. Kuehnert)Allegheny General Hospital,Pittsburgh, PA

Resistive, 2 kG, whole body 1982

Su percondu cting, 5 kG, w hole June 1983bodySu percondu cting, 5 kG, w hole July 1983bodySu p er con du ctin g, 5 kG, w hole Sep tem ber 1983bodySuperconducting, 5 kG, whole October 1983body

8/14/2019 8420


App. C—Manufacturers of NMR Imaging Devices 13 5

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

Mount Sinai Medical Center, MiamiFL

Mallinckrodt Institute, WashingtonUniversity, St. Louis, MOSt. Vincent Medical Center, LosAngeles, CAPomona Valley Community Hospital,Pomona, CALoma Linda University MedicalCenter, Loma Linda, CAMemorial Hospital Medical Center of Long Beach, Long Beach, CAHershey Medical Center, Hershey,PABoone County Medical Center, Col-umbia, MORadiology Associates, Little Rock,ARSt. Francis Medical Center, Peoria,IL

Digital Diagnostics, Baton Rouge, LA

Magnetic Resonance Imaging, Inc.,Brooklyn, NYWendover Park Associates,Greensboro, NCUniversity of Minnesota, Min-neapolis, MNOchsner Foundation, New Orleans,LA

University of Virginia, Charlottes-ville, VAAmerican Shared Hospital Services,San Francisco, CAPacific Medical Center, San Fran-cisco, CASiemens Headquarters, Iselin, NJ

Flower Hospital, Toledo, OH

Magnetic Imaging Associates, LosAngeles, CASouthwest Texas Methodist Hospital,San Antonio, TXUniversity Diagnostic Institute of Tampa, Tampa, FLAmerican Shared Services, SanFrancisco, CAFlorida Medical Association, Tampa,FL

Medical College of Virginia, Rich-mond, VA


Superconducting, 20 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 15 kg, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 5 kG, wholebodySup ercondu cting, 10 kG, wh olebodySuperconducting, 5 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySupercondu cting, 10 kG, wh olebodySupercondu cting, 10 kG, wh olebodySupercondu cting, 20 kG, wh olebodySuperconducting, 10 kG, wholebodySuperconducting, 15 kG, wholebodySup erconducting, 10 kG, wholebodySup erconducting, 20 kG, wholebodySup erconducting, 10 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySup erconducting, 10 kG, wholebodySuperconducting, 10 kG, wholebodySupercondu cting, 10 kG, wh olebodySuperconducting, 10 kG, wholebody

November 1983

May 1984

April 1984

1984E

1984E

April 1984

NA

April 1984

May 1984

1984E

1984E

1984E

1984E

1984E

1984E

1984E

April 1984

1984E

March 1984

1984E

1984E

1984E

1984E

five systems expectedin 1984-85

1984E

1984E

8/14/2019 8420


136 . Health Technology Case Study 27: Nuclear Magnetic Resonance imaging Technology: A Clinical, industrial, and Policy Analysis

32.

33,

34.

35.

36.

37.

38.

39.

40.

41,

42.

43.

44.

45.

46.

47.

48.

49.

50.

51.

Magnetic Resonance, Inc., Rockville,MDMagnetic Resonance of Williamsport,Williamsport, VAMid Miami Assets MRI, Miami, FL

Harper Hospital, Detroit, MI

Faculty Medical Practice, Memphis,TNNew England Medical Center,Boston, MALong Island Jewish Hospital, NewHyde Park, NYMethodist Hospital, Houston, TX

Nebraska Methodist Hospital,Omaha, NEMount Sinai Medical Center, Miami,FL

Methodist Hospital, Houston, TX

Fox Chase Medical Center,Philadelphia, PAPrivate clinic, New York, NY

New Rochelle Radiology, NewRochelle, NYPrivate clinic, Munich, WestGermanyUniversity of Berlin, Berlin, WestGermanyUniversity of Hiedelberg, Heidelberg,West GermanyUniversity of Hiedelberg, Heidelberg,

West GermanyUniversity of Upsala, Upsala, Sweden

University of Tokyo, Tokyo, Japan

Superconducting, 10 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, whole

bodySuperconducting, 20 kG, wholebodySuperconducting, 10 kG, wholebodySup ercondu cting, 20 kG, wh olebodySuperconducting, 5 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 15 kG, wholebodySuperconducting, 20 kG, whole


bodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, whole

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984E

1984

1984

1984

1984

1984

1984body


Other diagnostic imaging products: Computed tomography; ultrasonography; nuclear medicine; digital

radiography; conventional X-ray; and fluoroscopeOther medical products: Assorted electromedical equipmentNon-health-care-related products: Parent company produces assorted electrical appliances and equipment

8/14/2019 8420


App. C—Manufacturers of N MR Imaging Devices 137

TECHNICARE CORP.

29100 Aurora Road

Solon, OH 44139

Background: Technicare is a publicly owned subsidiary of Johnson & Johnson. Johnson & Johnson made an in-itial commitment to NMR imaging as early as 1977, but major R&D effort did not begin until the acquisitionof Technicare in 1979 from Ohio Nuclear. In 1980, the company completed its first engineering model. The follow-

ing year Technicare made its first clinical placement of an NMR-imaging system. In 1982, the Magnet Corpora-tion of America was acquired to build superconducting magnets for Technicare’s NMR-imaging systems. Thatsame year, Technicare had its first commercial prototype available for placement.



Superconducting 15 kG a Animal 1980Resistive 1.5 kG a H ead only 1981Su percondu cting 3 kGb Whole body 1982Su percondu cting 5 k Ga Whole body 1983Su percondu cting 6 kG Whole body 1983Su percondu cting 15 k G Whole bod y 1983

aprobab]e commercla] prototype system(s).bNo ]Onger available.

C o l l a b o r a t i v e a r r a n g e m e n t s w i t h u n i v e r s i t i e s o r m e d i c a l c e n t e r s : 1 7

1 .

2.

3.

4.

5.6.

7.

8.

9.

10.

11.

12.

13.14.

15.16.17.

Massachusetts General Hospital, Boston, MAUniversity Hospital, Cleveland, OHCleveland Clinic Foundation, Cleveland, OHUniversity of Kentucky, Lexington, KYIndiana University, Indianapolis, INHershey Medical Center, Hershey, PAMillard Fillmore Hospital, Buffalo, NYSt. Joseph’s Hospital, London, Ontario, CanadaJohns Hopkins University, Baltimore, MDOntario Cancer Institute, Toronto, CanadaCharlotte Memorial Hospital, Charlotte, NCNew York Hospital, New York, NY

Vanderbilt University, Nashville, TNDefalquu Clinic, Charleroi, BelgiumUniversity of Florida, Gainesville, FLBaylor University Medical Center, Dallas, TXRush Presbyterian-St. Luke’s Medical Center, Chicago, IL


Date of installation


1. Massachusetts General Hospital, Superconducting, 15 kG, animal 1981Boston, MA

2. Massachusetts General Hospital, Resistive, 1.5 kG, head only 1981B o s t o n , M A

3 . U n i v e r s i t y H o s p i t a l , C l e v e l a n d , O H S u p e r c o n d u c t i n g , 3 . 0 k G , w h o l e October 1982body

4. Cleveland Clinic Foundation, Resistive, 1.5 kG, w hole bod y N ov em ber 1982Cleveland, OH

5. University of Kentucky, Lexington, Resistive, 1.5 kG, wh ole bod y Decem ber 1982KY

8/14/2019 8420


138 Health Technology Case Study 27: Nuclear Magnetic Resonance Imaging Technology: A Clinical, Industrial, and Policy Anal ysis

6.7.

8.

9.

10.

11.

12.

13.14.

15.

16.

17.

18.

19.

20.

21.

22.23.

24.

25.

26.

27.

28.29.30.

31.

32.33.

34.

Indiana University, Indianapolis, INHershey Medical Center, Hershey,PAMillard Fillmore Hospital, Buffalo,N YSt. Joseph’s Hospital, London,

Ontario, CanadaOntario Cancer Institute, Toronto,CanadaNew York Hospital, New York, NY

Charlotte Memorial Hospital,Charlotte, NCDefalque Clinic, Charleroi, BelgiumVanderbilt University, Nashville, TN

Shands Teaching Hospital, Universityof Florida, Gainesville, FLHouston Imaging Center, Houston,TX

Cleveland Clinic Foundation,Cleveland, OHCleveland Clinic Foundation,Cleveland, OHMassachusetts General Hospital,Boston, MAScottsdale Memorial Hospital, Scott-sdale, AZ

AMC Cancer Research Center andHospital, Lakewood, COSt. Luke’s Hospital, Jacksonville, FLRush Presbyterian-St. Luke’s MedicalCenter, Chicago, ILGreenberg Radiology Clinic,

Highland Park, ILNorth Shore University Hospital,Manhasset, NYUniversity Park Imaging, Urbana, IL

Veterans Administration MedicalCenter, St. Louis, MONMR SA, Barcelona, SpainML and Associates, New York, NYUniversity Hospitals of Cleveland/ Case Western Reserve University,Cleveland, OHBaylor University Medical Center,Dallas, TX

Temple Radiology, New Haven, CTAlbert Einstein Medical Center,Philadelphia, PANuclear Facilities, Brooklyn, NY

Resistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole body




Superconducting, 5 kG, wholebodyResistive, 1.5 kG, whole body

Resistive, 1.5 kG, whole bodySuperconducting, 5 kG, wholebodyResistive, 1.5 kG, whole body


Superconducting, 6 kG, wholebodySuperconducting, 15 kG, wholebodySup ercondu cting, 6 kG, wh olebodySup ercondu cting, 6 kG, wh olebody


Resistive, 1.5 kG, whole bodySuperconducting, 5 kG, wholebodyResistive, 1.5 kG, whole body

Sup ercondu cting, 6 kG, wh olebodySup ercondu cting, 6 kG, wh olebodyResistive, 1.5 kG, whole body

Resistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole bodySuperconducting, 15 kG, wholebody

Sup ercondu cting, 6 kG, wh olebody

Resistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole body

Sup ercondu cting, 5 kG, wholebody

February 1983February 1983

March 1983

March 1983

March 1983

June 1983

February 1983

April 1983July 1983

October 1983

July 1983

August 1983

May 1984

February 1984

September 1983

June 1983

December 1983November 1983

June 1983

December 1983

June 1984

February 1984

January 1984October 1983

July 1984

January 1984

February 1984March 1984

May 1984

8/14/2019 8420


.—

App . C—Manufac turers of NMR imaging Devices 139

35.

36.

37.

38.

39.

40.

41.42.

43.44.

NMR Diagnostic Center, Sun City,AZGarden State Medical Center,Marlton, NJMagnetic Imaging of Bellville,Bellville, IL

Private clinic, Union, NJ

Ft. Worth Magnetic Imaging In-stitute, Ft. Worth, TXBroward NMR, Ft. Lauderdale, FL

Clairval Hospital, Marseille, FrancePrivate clinic, Hanover, WestGermanyClinique du Park, Paris, FrancePrivate clinic, Antwerp, Belgium


Superconducting, 6 kG, wholebodySup ercondu cting, 6 kG, wholebody

Sup ercondu cting, 6 kG, wholebodySuperconducting, 5 kG, wholebodySuperconducting, 5 kG, wholebodyResistive, 1.5 kG, whole bodySuperconducting, 5 kG, wholebodyResistive, 1.5 kG, whole bodyResistive, 1.5 kG, whole body


January 1984

June 1984

March 1984

August 1984

August 1984

January 1984

December 1983November 1983

April 1984April 1984

Other diagnostic imaging products: Computed tomography; ultrasonography; nuclear medicine;

radiographyOther medical products: Parent firm makes surgical instruments and supplies; dental equipment

Non-health-care-related products: N one

digital

8/14/2019 8420


Appendix D. —Glossary of Terms and Acronyms

Glossary of Terms

Field strength: The strength of the magnetic field of a magnet.

Inhomogeneities in magnetic field: Lack of uniform-ity in magnetic field strength.

Ionizing radiation: A form of radiant energy within

the electromagnetic spectrum that has the capabil-ity of penetrating solid objects and altering the elec-trical charge of their atoms. High-energy radiation,such as X-rays and gamma rays, is ionizing ra-diation.

Kilogauss: A unit of measurement of the magneticforce per unit area that can be generated within adefined region. (See tesla. )

Magnetic field gradien t: A magnetic field that increasesor decreases in strength in a given direction alonga sample.

Magnetic moments: The vector representations of thenet magnetic properties of hydrogen atoms.

Medical technology: The drugs, devices, medical and

surgical procedures used in medical care, and theorganizational and supportive systems within whichsuch care is provided.

Nuclear: Pertaining to the nucleus, the positivelycharged central portion of an atom that consists of protons and neutrons, except in hydrogen, whichhas only one proton.

Paramagnetic: A substance with a small but positivemagnetic susceptibility (magnetizability) that mayincrease the contrast between tissues and NMR im-ages (4).

Prospective payment: Payment for medical care ac-cording to rates set in advance of the period dur-ing which they apply.

Pulse sequen ce: The pattern of radiofrequency energyused to excite protons.

Radiation: Emission of or exposure to radiant energy,which travels as a wave motion. Radiant energyranges from low-frequency, nonionizing radiofre-quency waves used in NMR to high-frequency,ionizing waves used in X-rays.

Radiofrequency waves: Low-energy, electromagneticwaves that do not emit ionizing radiation and thatare used in NMR imaging.

Rate of loss of coherence: The rate at which protonsstop rotating in phase with each other.

Relaxation time characteristics: The rate at whichtissue hydrogen atoms that have been excited byradiofrequency energy return to their equilibriumstates.

Resonance: The oscillation of nuclei between higherand lower energy levels as radiofrequency energyis applied and withdrawn.

140

Shimming: Adjustments, such as addition of specialcoils, made to eliminate inhomogeneities in themagnetic field.

Spatial resolution: The extent to which two adjacentstructures can be distinguished.

Spectrogram: Graphic depiction of the individual com-ponents of NMR signals from phosphorus-con-taining compounds arranged according to fre-quency.

Spectroscopy: A technique in which the individualcomponents of the NMR signals from compounds,such as phosphorus-contain in g compounds , a r eanalyzed according to frequency.

T 1: “Spin-lattice” relaxation time. A time constant thatreflects the rate at which excited protons exchangeenergy with the surrounding environment.

T 2: “Spin-spin” relaxation time. A time constant thatreflects the rate at which protons stop rotating inphase with each other because of the local magneticfields of adjacent nuclei.

Tesla: A unit of measurement of the magnetic forceper unit area that can be generated within a definedregion; 1 tesla = 10,000 gauss (10 kilogauss), Forperspective, the magnetic field strength of the Earthis approximately half a gauss.

Tomographic scan: The image of an individual sliceor plane.

Glossary of Acronyms

ACR –AMI –A TPBC / BS - -

CON –CTD H H S - -

D H S S –

DRG –ECG –EMI –FDA –F D C A –F O N A R –GE –GEC –

HCA –H C F A –

H I A A –H M O –

American College of RadiologyAmerican Medical Internationaladenosine triphosphateBlue Cross and Blue Shield Association

certificate of needcomputed tomograph y

Department of Health and Human Serv-ices (United States)Department of Health and Social Security(United Kingdom)diagnostic related grou pelectrocardiogramEnglish Mu sic Ind ustryFood and Drug Adm inistration, DHH SFood, Drug, and Cosmetic Actfield focusing nuclear magnetic resonance

General Electric Co. (United States)General Electric Co. (United Kingdom )

Hospital Corporation of AmericaHealth Care Financing Administration,DHHS

Health Insurance Association of Americahealth ma intenance organization

8/14/2019 8420


App. D—Glossary of Terms and Acronyms Ž 141

HSA –IDE –IGC –IRB –NCI –N E M A –

N H L B I –

N H P I C –

NIH –NME –NMR –N R P B –

NSF –O H T A –

PDP –

Health Systems Agencyinvestigational device exemptionIntermagnetics General Corp.Institutional Review BoardNational Cancer Institute, NIHNational Electrical Manufacturers Asso-ciationNational Heart, Lung, and Blood Insti-

tute, NIHNational Health Planning InformationCenterNational Institutes of Health, DHHSNational Medical Enterprises, Inc.nuclear magnetic resonanceNational Radiological Protection Board(United Kingdom)National Science FoundationOffice of Health Technology Assessment,DHHSProduct Development Protocol

PET –PMA –P M A A –Pro –R C H S A –

R&D –RF –

SBIR –

S H C C –S H P D A –

S P E C T –

S U N Y –UCR –U C S F –VA –YAG –

positron emission tomographypremarket approvalpremarket approval applicationPreferred Provider OrganizationRadiation Control for Health and SafetyActresearch and developmentradiofrequency waves

Small Business Innovation ResearchprogramStatewide Health Coordinating CouncilState Health Planning and DevelopmentAgencysing le pho ton emiss ion computed to -mographyState University of New Yorkusual, customary, and reasonable chargesUniversity of California, San FranciscoVeterans Administrationyttrium aluminum garnet laser

8/14/2019 8420


Appendix E. —Acknowledgments andHealth Program Advisory Committee

The authors are grateful to Mark Farfel at the Johns Hopkins School of Hygiene and Public Health for hiscontributions as a research assistant and are indebted to Charlotte Chalmers, Connie Pollack, and Amy Bohnfor their expert assistance in manuscript preparation.

In addition, we wish to thank the many physician experts, scientists, hospital administrators, government of-ficials, health planners, industry trade association officials and company representatives for the help and infor-mation they provided to us.

The following people, in addition to the advisory panel, provided valuable guidance.

Herbert AbramsBrigham & Women’s Hospital

J. AlwenMedical Research Council, United Kingdom

Prem K. AnandPicker International

Mary P. AndersonNational Center for Devices and Radiological Health,U.S. Food and Drug Administration

Raymond AndrewUniversity of Florida

David ArmourSiemens Medical Systems, Inc.

Larry AtkinsAMI Diagnostic Services, Inc.

G. BarenzU.S. Department of Health and Human Services

I. M. Bar-NirElscint Ltd.

James BensonNational Center for Devices and Radiological HealthU.S. Food and Drug Administration

Howard BerkeADAC Laboratories

Katherine L. BickNational Institute of Neurological and

Communicative Disorders and StrokeNational Institutes of Health

Robert BirdHospital Corp. of America

Joel L. BlankDiasonics, Inc.

M. BlankHealth Care Financing Administration

Merry BlumenfeldGeneral Electric Co.

Robert M. BradenAmerican College of Radiology

Thomas BradyMassachusetts General Hospital

Robert J. BritainU.S. Food and Drug Administration

Paul C. BrownOklahoma Health Planning Commission

R. Burns

Massachusetts State Health Planning andDevelopment Agency

James CarolanNalorac Cryogenics Corp.

John CasseseFonar Corp.

J. S. CohenNational Institute of Child Health and Human

DevelopmentNational Institutes of Health

Donald R. CohodesBlue Cross and Blue Shield Association

Morris Cohen

Kaiser-Permanente Medical Care ProgramJohn P. ConomyCleveland Clinic Foundation

Dennis CotterInstitute for Health Policy AnalysisGeorgetown University Medical Center

D. DeCoriolesBlue Cross and Blue Shield Association

Gilbert B. DeveyTechnology consultant

Richard diMondaAmerican Hospital Association

Richard B. EmmittF. Eberstadt & Co., Inc.

142

8/14/2019 8420


App. E—Acknow ledgments and Health Program Advisory Commit tee 143

J. FairchildKentucky Department for Health Services

James W. FletcherU.S. Veterans Administration Medical Center,

St. Louis

P. GarnerIllinois Certificate-of-Need Program,

Health Facilities Planning BoardB. Goff Utah Department of Health

A. GoldmanGeorgia State Health Planning and Development

Agency

Samuel C. GoldmanElscint Inc.

Richard J. GreeneU.S. Veterans Administration

Friedrich GuddenSiemens AG, Federal Republic of Germany

Ralph Halpern,

Health Planning Council for Greater Boston, Inc.Timothy HansenPicker International

Keith HarvilleLifemark Corp.

John D. HaytaianFonar Corp.

Mel HellstromCGR Medical Corp.

David HelmsAlpha Center for Health Planning

Waldo S. Hinshaw

Technicare Corp.Dick HullihenPicker International

R. HutchersonState Health Planning and Development AgencyState of Missouri

Cynthia HutchinsJEOL USA, Inc.

John HuttonInstitute for Research in the Social SciencesUniversity of York, United Kingdom

Herbert JacobsHealth Care Financing Administration

John JacobyGeneral Electric Company

A. Everette James, Jr.Vanderbilt University

S. KatzHealth Care Financing Administration

Michael KellyJEOL USA, Inc.

A, J. Korsak

Health Insurance Association of AmericaMichael LandaU.S. Food and Drug Administration

Paul C. LauterburState University of New York at Stony Brook

I. LeeState Health Planning and Development AgencyGovernment of the District of Columbia

Otha W. LintonAmerican College of Radiology

John MallardUniversity of Aberdeen

Peter MansfieldUniversity of Nottingham, United Kingdom

Alexander MargulisUniversity of California School of Medicine, Sa n

Francisco

Robert C. McCuneNational Electrical Manufacturers Association

Thomas F. MeaneyCleveland Clinic Foundation

Edward MillerNational Center for Devices and Radiological HealthU.S. Food and Drug Administration

Glenn MillerDiasonics NMR

J. MillirenState of New YorkDepartment of Health

S .N . Mo h ap a t r aPicker International

W. Gorden MonahanAmerican Medical International, Inc.

Laura Lee MurphyNational Electrical Manufacturers Association

R. OshelState Health Planning and Development AgencyGovernment of the District of Columbia

Anthony M. PaceNational Medical Enterprises, Inc.

25-341 0 - f34 - 11 : QL 3

8/14/2019 8420


144 • Health Technology Case Study 27: Nuclear Magnetic Resonance Technology: A Clinical, Indust rial, and Policy Analy sis.—

C. Leon PartainVanderbilt University

Robert PhillipsNational Center for Devices and Radiological HealthU.S. Food and Drug Administration

Gerald PohostUniversity of Alabama Medical Center

Roger PowellNational Cancer InstituteNational Institutes of Health

Stanley ReiserUniversity of Texas Health Science Center

Daryl ReynoldsNational Medical Enterprises, Inc.

F. David RolloHumana, Inc.

M. J. RossIntermagnetics General Corp.

Richard RussellNational Health Planning Information Center

Janet SamuelPhilips Medical Systems, Inc.

Richard SanoPhilips Medical Systems, Inc.

Sheldon SchafferTechnicare Corporation

Roger SchneiderNational Center for Devices and Radiological HealthU.S. Food and Drug Administration

Aldona A. SiczekFischer Imaging Corp.

Norman SlarkTHD Department of Health and

Social Security, United Kingdom

James S. SmithU.S. Veterans Administration

E. StammFlor ida Sta te Heal th Plann ing and Development

Agency

Robert E. SteinerHammersmith Hospital, United Kingdom

S. StilesAlpha Center for Health Planning

Robert A. StreimerHealth Care Financing Administration

Christian TanzerBruker Instruments, Inc.

J. M. TaverasMassachusetts General Hospital

Peter TaylorTechnicare Corp.

Guenter WestermanSiemens Medical Systems, Inc.

C. WilsonState Health Planning and Development AgencyGovernment of the District of Columbia

Gabriel H. WilsonUniversity of California at Los Angeles

Medical Center

Robert WoodwardCGR Medical Corp.

HEALTH PROGRAM ADVISORY COMMITTEE

Stuart H. Altman*DeanFlorence Heller SchoolBrandeis UniversityWaltham, MA

H. David Banta

Deputy DirectorPan American Health OrganizationWashington, DC

Sidney S. Lee, Committee Chair President, Michael Reese Hospital and Medical Center

Chicago, IL

Carroll L. Estes**ChairDepartment of Social and Behavioral SciencesSchool of NursingUniversity of California, San FranciscoSan Francisco, CA

8/14/2019 8420


App. E—Acknow ledgments and Health Program Advisory Comm itt ee 145 —

Robert EvansProfessorDepartment of EconomicsUniversity of British Colum biaVancouver , BC

Rashi FeinProfessorDepartment of Social Medicine and Health PolicyHarvard Medical SchoolBoston, MA

Harvey V. FinebergDean

School of Public HealthHarvard UniversityBoston, MA

Melvin A. Glasser* * *DirectorHealth Security Action Council-Committee for National Health InsuranceWashington, DC

Patricia King

ProfessorGeorgetown Law CenterWashington, DC

Joyce C. Lashof DeanSchool of Public HealthUniversity of California, BerkeleyBerkeley, CA

Alexander Leaf Professor of MedicineHarvard Medical SchoolMassachusetts General HospitalBoston, MA

Margaret Mahoney* * * *PresidentThe Commonwealth FundNew York, NY

Frederick MostellerProfessor and ChairDepartment of Health Policy and ManagementSchool of Public HealthHarvard UniversityBoston, MA

Norton NelsonProfessorDepartment of Environmental MedicineNew York UniversityMedical SchoolNew York, NY

Robert OseasohnAssociate Dean

University of Texas, San AntonioSan Antonio, TX

Nora PioreSenior AdvisorThe Commonwealth FundNew York, NY

Mitchell Rabkin*PresidentBeth Israel HospitalBoston, MA

Dorothy P. RiceRegents LecturerDepartment of Social and Behavioral Sciences

School of NursingUniversity of California, San FranciscoSan Francisco, CA

Richard K. RiegelmanAssociate ProfessorGeorge Washington UniversitySchool of MedicineWashington, DC

Walter L. RobbVice President and General ManagerMedical Systems OperationsGeneral ElectricMilwaukee, WI

Frederick C. RobbinsPresidentInstitute of MedicineWashington, DC

Rosemary StevensProfessorDepartment of History and Sociology of ScienceUniversity of PennsylvaniaPhiladelphia, PA

“Unt]l April 1983

* “Until March 1984

* “Until October 1983

‘Until August 1~83.

8/14/2019 8420


8/14/2019 8420


8/14/2019 8420


8/14/2019 8420


8/14/2019 8420


8/14/2019 8420


8/14/2019 8420


8/14/2019 8420


References 155 —

177.

178.

179.

180.

181.

182.

183.

184.

185

186.

187.

188.

“NMR of the Pericardium: Normal and Patho-logic Findings, “ Radiology 150(2):469-474, February

1984.Steiner, R. E., Hammersmith Hospital, London,personal communication, June 11, 1983.Steiner, R. E., “NMR of the Abdomen, ” paperpresented at a Symposium on NMR Imaging

sponsored by the Commission on Nuclear Mag-netic Resonance, American College of Radiology,Chicago, June 11, 1983.Steiner, R. E., and Worthington, B., testimonypresented at FDA Radiologic Device Panel Meet-ing, Rockville, MD, July 6, 1983.Stephens, W. H., Patton, J. A., Lagan, J. E., etal., “Certain Economic Considerations in NMR

Imaging, ” in Nuclear Magnetic Resonance and Correlative Imaging Modalities, C. L. Partain(cd. ) (New York: The Society of Nuclear Medi-cine, Inc., 1983).Stephens, W. H., James, A. E., Winfield, A. C.,et al., “Financial Implications of NMR Imaging, ”

in Nuclear Magnetic Resonance (NMR) Imaging,C. L. Par t a in , A. E. James, F. D. Rollo, et al.(eds. ) (Philadelphia: W. B. Saunders, 1983).Stiles, S., Alpha Center for Health Planning,Bethesd a, MD, personal comm un ication, July 26,1983.U.S. Congress, House Com mittee on Interstat eand Foreign Comm erce, Medical Device Amend-

ments of 1976, Report No. 94-853 to accompanyH.R. 21124 (Washington, DC: U.S. GovernmentPrinting Office, 1976).Taveras, J. M., Chairman, Department of Radi-ology, Massachusetts General Hospital, Boston,MA, personal communication, June 1, 1983.

U.S. Congress, Office of Technology Assessment,“Case Study #13: Radionuclide Imaging and CostEffect iveness, ” The Implications of Cost-Effec-tiveness Analysis of Medical Technology, OTA-

BP-H-9(13) (Washington, DC: U.S. GovernmentPrinting Office, 1982).U.S. Congress, Office of Technology Assessment,

Diagnosis Related Groups (DRGs) and the Medi-care Program: Implications for Medical Technol-ogy—A Technical Memorandum, OTA-TM-H-

17 (Washington, DC: U.S. Government PrintingOffice, 1983).U.S. Department of Health and Human Services,Food and Drug Administration, Bureau of Med-

ical Devices, Guidelines for the Arrangement and Content of a Prem arket Approval Application(PMA), Silver Spring, MD, November 1980.U.S. Department of Health and Human Services,

189

190.

191.

192

193

194.

195.

196<

Food and Drug Administration, “Guidelines forEvaluating Electromagnetic Risk for Trials of Clinical NMR Systems, ” memorandum to spon-sors, manufacturers, and researchers of clinicalNMR devices, Feb. 25, 1982.U.S. Department of Health and Human Services,Food and Drug Administration, “Guidelines for

Evaluating Electromagnetic Risk for Trials of Clinical NMR Systems, ” memorandum to spon-sors, manufacturers, and researchers of clinicalNMR devices, Dec. 28, 1982.U.S. Department of Health and Human Services,National Institutes of Health, Division of Re-search Resources, Biotechnology Resources: Re-search Resources Directory, NIH publication No.83-1430, Bethesda, MD, revised September 1983.U.S. Department of Health and Human Services,National Institutes of Health, National Cancer In-stitute, Diagnostic Imaging Research Branch, “Re-quest for Proposal #NCI-CM-375624-27: Com-parative NMR Imaging Studies, ” 1982.

U.S. Department of Health and Human Services,Office of Health Technology Assessment, Pub-lic Health Service Procedures for EvaluatingHealth Care Technologies for Purposes of Medi-care Coverage (Washington, DC: U.S. Govern-ment Printing Office, May 1983).U.S. Department of Health and Human Services,Office of Health Maintenance Organizations, Na-tional HMO Census of Prepaid Plans, D H H Spublication-1 No. (PHS) 81-50159 (Washington,

DC: U.S. Goverment Printing Office, 1981).U.S. Department of Health and Human Services,Public Health Service, Health Resources andServices Administration, Bureau of Health Main-

tenance Organizations and Resources Develop-ment, National Health Planning InformationCenter, “Program Information Letter 83-15, ”Washington, DC, July 8, 1983.U.S. Department of Health and Human Services,Public Health Service, Health Resources andServices Administration, Bureau of Health Main-tenance Organizations and Resources Develop-ment, National Health Planning InformationCenter, “Summary Report of Health PlanningAgency Experiences Related to Nuclear MagneticResonance Devices: Responses to Program Infor-mation Letter 83-15, ” Washington, DC, Sept. 29,1983.

U.S. National Science Foundation, “NSF AwardsGrants To Estab}ish Regional Instrumentation Fa-cilities, ” National S cience Foun dation N ews, NSFPR79-86, Nov . 8, 1979.

8/14/2019 8420



197.

198.

199.

200,

201.

202.

U.S. National Science Foundation, “Small Busi- 203.ness Innovation Research Program Solicitation, ”NSF 83-26, closing date July 15, 1983.Wagner, J. L., “Reimbu rsement Shapes Marketfor Technology,” Hospitals, J . A.H.A. 53:91-94, 204,

1979.Walters, G. K., and Fairbank, W. M., “PhaseSeparation in H e 3-H e 4 Solutions, ” Z%ys. Rev.103(1):262, July 1, 1956.Warner, K. E., “Effects of Hosp ital Cost Conta in- 205.

ment on the Development and Use of MedicalTechnology,” Milbank Mern . Fun d Q. 56(2):187-211, Spring 1978. 206.Washington Report on Medicine and Health,“Perspectives: Health Planning Looks to theFuture, ” vol. 37, No. 48, Dec. 5, 1983. 207.Weisman, I. D., Bennett, L. H., Maxwell, L. R.,et al., “Recognition of Cancer In Vivo by NuclearMagnetic Resonance, ” Science 178:1288, Decem-ber 1972.

Wilson, C., State Health Planning and Develop-ment Agency, Government of the District of Col-umbia, Washington, DC, personal communica-tion, Mar. 22, 1984.Wilson, G. H., Special Assistant for DiagnosticImaging Research, Radiation Research Program,Division of Cancer Treatment, National CancerInstitute, National Institutes of Health, Bethesda,MD, personal communication, June 24, 1983,Wilson, V., “NMR in Nebraska: Playing Catch-up Ball, ” internal paper, Nebraska State Depart-ment of Health, Lincoln, NE, July 19, 1983.Young, A. L. “Commercialization of Investiga-tional Medical Devices, ” Med. Device Diag. In -dustry 5(3):66-69, 82, 1983.Young, I. R., Hall, A. S., Pallis, C. A., et al.,“Nuclear Magnetic Resonance Imaging of theBrain in Multiple SclerosisJ” L.ancet 2(8255 ):1063-

1066, Nov. 14, 1981.

U . S . G O V E R N M E N T P R I N T I N G O F F I C E : 1 9 8 4 0 - 2 5 - 3 4 1 : Q L 3