Click for next page ( 256


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 255
APPENDIX A Profiles of 20 Technology Assessment Programs Clifford S. Goodman* The profiles of 20 medical technology as- sessment programs in the United States con- tained in this appendix illustrate the great variety in a common frameworkof assess- ment activities in major sectors of the Ameri- can health care system. The profiles repre- sent selected assessment activities of medical societies, medical product makers, govern- ment assessment organizations, health care provider organizations, third-party payers, universities, and independent evaluators and policy research organizations. The informa- tion assembled for these profiles provided much of the basis for preparing Chapter 2, "The Scope of Medical Technology Assess- ment." The 20 profiled programs are the fol- lowing: Joint American College of Cardiology/Amer- ican Heart Association Task Force on As- sessment of Cardiovascular Procedures American College of Physicians Clinical Effi- cacy Assessment Project American Hospital Association Hospital Technology Series Program American Medical Association Diagnostic * National Research Council Fellow, National Academy of Sciences, Washington, D.C. 255 and Therapeutic Technology Assessment Program Battelle Memorial Institute Human Affairs Research Centers Blue Cross and Blue Shield Association Medi- cal Necessity Program Blue Cross and Blue Shield Association Tech- nology Evaluation and Coverage Program ECRI Institute of Society, Ethics and the Life Sci- ences (Hastings Center) The Permanente Medical Group, Inc., Divi- sion of Health Services Research Medtronic Inc. National Center for Health Services Research and Health Care Technology Assessment Office of Health Technology Assessment National Heart, Lung, and Blood Institute National Institutes of Health Office of Medi- cal Applications of Research Consensus Development Program National Library of Medicine Congressional Office of Technology Assess- ment Health Program Prospective Payment Assessment Commis- s~on Smith Kline & French Laboratories Cost- Benefit Studies Program University of California at San Francisco In- stitute for Health Policy Studies

OCR for page 255
256 Veterans Administration Cooperative Stud- ies Program The profiles were prepared by assembling information from personnel of the profiled programs, program publications, reports in the literature, and other sources. in each case, at least two drafts were provided to pro- gram personnel to be reviewed for matters of fact, for updating since previous drafts, and for generating suggestions for content. The profiles could not have been written without the gracious assistance of the individuals shown at the end of this introductory section; however, any errors or inadequate represen- tations are solely the responsibility of the pre- parer of this appendix. Although the programs are profiled ac- cording to the same 16 information catego- ries, the attempt was made to retain the per- spective and flavor of the individual programs. Profile narratives are arranged as follows. Introduction Purpose Subjects of Assessment Stage of Diffusion Concerns Requests Selection Process Assessors Turnaround Reporting Impact Reassessment Funding/Budget Examples Sources These profiles are a start at systematically characterizing current technology assessment activities in the United States. The nrofile.s are intended to answer the who, what, where, why, how, and how much of technol- ogy assessment of the 20 selected programs. Profiles such as these may be useful for the type of technology assessment clearinghouse recommended in this report. They could be published or made available on-line and up- dated periodically. Organizations already profiled could make additions and modifica- tions annually, or perhaps continuously, as might be the case with an on-line system. Or- ganizations wishing to be included could work with clearinghouse staff to develop a new profile, much in the manner in which these were prepared. The sources of financial support for such an ongoing activity could in- ASSESSING MEDICAL TECHNOLOGY elude publication fees, subscriptions, or other types of access fees. The possibilities for valuable cross-cuts of the profile data are many. Distributions of assessment programs such as those shown in Chapter 2 may be helpful in portraying gaps in and other characteristics of overall assess- ment activity. Listings of assessments by spe- cific technologies would be useful for organi- zations contemplating their own assessments of these technologies, or seeking information concerning procurement or use of technolo- gies. Each profile begins with a summary sec- tion portraying the program's major concerns and technologies assessed in matrix form, and checklists summarizing the stage of technolo- gies assessed, application of technologies, as- sessment methods, and approximate annual budget. The summary sections cannot substi- tute for the profile narratives. The summary sections necessarily make categorical distinc- tions where these may not be so clear in prac- tice. The following discussion and definitions may be helpful in understanding the matrix and checklists of assessment activities in the summary sections of each profile. TECHNOLOGY As noted and discussed in Chapter 1, the usage of the terms medical technology and assessment are those of the congressional Of- fice of Technology Assessment. For the pur- poses of the summary sheets, OTA's usage is expanded upon as follows. Drug: any chemical or biological substance that may be applied to, ingested by, or in- jected in order to prevent, treat, or diag- nose disease or other medical conditions. Included are biologicals such as vaccines and blood products, medicinals and botan- icals, and pharmaceutical preparations. Device: any physical item, excluding drugs, used in medical care. Included are diag- nostic and therapeutic equipment, pros- theses, surgical and medical instruments and supplies, dental equipment and sup- plies, ophthalmic goods, and in vitro diag- nostic products reagents, instruments, and systems used in the collection, prepa-

OCR for page 255
APPENDIX A: PROFILES ration, and examination of specimens taken from the human body to determine the state of a patient's health. Medical or surgical procedure: a practice of a health care provider that generally in- volves a combination, often quite complex, of special skills or abilities with drugs, de- vices, or both. In some cases, the drugs or devices involved are not predominant fac- tors in a procedure. Instead, the technique of the provider performing the procedure is most important, such as in the perfor- mance of a surgical procedure facilitated by the use of scalpels, clamps, and drugs against infection. Psychotherapy or pre- scription of a special diet are examples of procedures which may not involve drugs or devices. Support system: a system that provides the environment for and otherwise facilitates the provision of health care, but is not the focal technology in a medical regimen, sur- gical procedure, or other form of health care. Examples are laboratory and radiol- ogy services, medical information systems, blood banking services, hospital infection control programs, food services, laundry, hospital facilities, and physical plant. Many of these are often referred to as ancil- lary services, and some might be said to comprise the infrastructure of health care delivery. Organizational/administrative system: used in management and administration to en- sure that health care is delivered as effec- tively as possible. Included are alternative delivery modes or settings, e.g., health maintenance organizations (HMOs), area- wide emergency care systems, and home health delivery, and payment systems, e.g., prepayment using diagnosis-related groups. The last two categories support and organi- zational/administrative systemsare often considered to be subjects of health services re- search. This classification of technologies recog- nizes that a given technology may be com- prised of as well as part of other technologies. A drug may be a concoction of multiple chemical entities packaged in a capsule; a 257 medical device may be made of valves, biomaterials, and microchips; a surgical pro- cedure may involve drugs and medical de- vices as well as the surgeon's skilled hand. An area-wide emergency medical care system may encompass all of these, plus ambu- lances, helicopters, communications systems, and more. An organization such as the Pro- spective Payment Assessment Commission may have to evaluate aspects of drugs, de- vices, medical and surgical procedures, and support systems in order to adjust an admin- istrative technology prospective payment using diagnosis-related groups. In portraying the technologies assessed by the programs, Xs are placed across from drugs and medical devices/supplies/equip- ment when program assessments directly ad- dress the properties of the medical products themselves. Where a medical product is not the predominant factor in a procedure, or where the properties of the medical product are taken as given and the emphasis of the as- sessment is on the concerns of a product- embodied procedure relative to another, Xs are placed across from medical/surgical pro- cedure only. For instance, for the purpose of the summary section charts, an ECRI assess- ment of mechanical ventilators to measure and compare various technical properties of several brands of these devices would be con- sidered an equipment assessment. On the other hand, an assessment by a third-party payer of the circumstances under which in- termittent positive-pressure breathing using mechanical ventilators is medically necessary and therefore reimbursable would be shown as an assessment of a medical procedure. The unusual type of assessment, e. g., where a third-party payer primarily assesses medical and surgical procedures but has in an instance assessed ambulance services (a support technology), may not be noted as a major emphasis in the summary section charts. CONCERNS An assessment may address one or more of many concerns, attributes, or properties of a technology. The summary section groups a number of these into four categories.

OCR for page 255
258 Safety: a judgment of the acceptability of risk in a specified situation' e.g., for a given medical problem, by a provider with speci- fied training, at a specified type of facility. Efficacy: benefit for a given medical problem under ideal conditions of use. Effectiveness: benefit for a given medical problem under average conditions of use. Cost/cost-effectiveness/cost-benefit: includes costs, charges, pricing, cost-benefit, cost- effectiveness, and related concerns. Specif- ically: Cost-benefit: the costs of a project or technological application compared to the resultant benefits, with both costs and benefits expressed in the same units. This unit is nearly always monetary. Cost-effectiveness: the costs of a project or of alternative projects compared to the resultant benefits, with cost and benefits/effectiveness not expressed by the same unit. Costs are usually ex- pressed in dollars, but benefits/effective- ness are ordinarily expressed in terms such as lives saved, disability avoided, quality-adjusted life years saved, or other relevant objectives. Ethical/legal/social: includes implications of technology for societal norms, morals, in- stitutions, and relationships; and eco- nomic, medical, legal, and cultural values. Effectiveness in the summary section charts refers not only to the absolute benefit of the technology taken alone, but also to the marginal benefits to be gained from use of a technology under particular circumstances, considering a given patient's status and the use of and information gained from other technologies. Thus, appropriateness, a pri- mary concern of several of the profiled pro- grams, is categorized in the summary sections under effectiveness and/or cost-effectiveness, where cost is an explicit consideration in determining appropriateness. Evaluations of the sensitivity, specificity, and other operat- ing characteristics of diagnostic technologies are efficacy/effectiveness concerns. Again, the summary section matrices cite the major program emphases. Thus, al- though many of the profiled programs have legal departments or various legal require- ASSESSING MEDICAL TECHNOLOGY meets in connection with their assessment ac- tivities, Xs are placed under the ethical/legal/ social concerns column only for those programs having these as central concerns of their assessments. STAGE OF TECHNOLOGIES ASSESSED Emerging: in the applied research stage, about the time of initial clinical testing, e.g., monoclonal antibodies for immuno- therapy of cancer. New: past the stage of clinical trials but not yet in widespread use, e.g., extracorporeal lithotripsy for treatment of kidney stones. Established: considered by providers to be a standard approach to a particular condi- tion and diffused into general use. Obsolete/outmoded: superseded by another technology and/or demonstrated to be in- effective or harmful, e.g., gastric freezing for peptic ulcer. Technologies may be assessed at different stages of diffusion. The point in a technolo- gy's life cycle at which it is assessed may de- pend upon the purposes of an assessment pro- gram, and the course of the life cycle may be affected by the assessment itself. For new drugs and certain devices, FDA regulatory requirements may mediate technological dif- fusion before, during, and after assessment benchmarks such as initiation of clinical tri- als, approval for marketing, and removal from the market in the case of a product found to pose an imminent health hazard. Indeed, as discussed in Chapter 2, technolo- gies may be assessed for the very purpose of determining their stage of diffusion. APPLICATION OF TECHNOLOGIES Prevention: protects an individual from dis- ease, e.g., vaccination. Diagnosis: helps in determining what disease processes occur in a patient, e.g., upper gastrointestinal endoscopy. Screening: detects disease or abnormality, or potential for these, often in asymptomatic patients, e.g., Pap smear for cervical can- cer.

OCR for page 255
APPENDIX A: PROFILES Treatment: relieves an individual from dis- ease and its effects, including technologies that cure disease and those that give symp- tomatic relief but do not alter the underly- ing disease process, e.g., drug therapy for depression. Rehabilitation: to restore to a condition of health or useful and constructive activity, e. g., assistive devices for severe speech im- pairment. ASSESSMENT METHODS Laboratory testing: nonclinical (in vitro) testing of medical technology, e.g., for drug, device, or equipment performance. Clinical trials: prospective clinical experi- ments designed to test the safety and effi- cacy of a medical technology in which peo- ple are assigned to experimental or control groups and outcomes are compared. In- cludes randomized controlled clinical tri- als, in which people are randomly assigned to experimental and control groups. Epidemiological and other observational methods: excludes the more rigorous ex- perimental design studies such as random- ized clinical trials. Included are such stud- ies known as quasiexperiments; series; case studies; cohort studies; natural experi- ments; and certain cross-sectional, case control, and longitudinal methods. Cost analyses: analyses, including cost-bene- fit and cost-effectiveness analyses, that enumerate, measure, and compare both the benefits and costs of medical technolo- gies. Analyses may vary in terms of per- spective (i.e., the parties to whom the ben- efits and costs accrue) and the choice and valuation of the benefits and costs consid- ered. Simulation/modeling: use of models repre- sentations of real-world phenomena to test or evaluate proposed interventions, of- ten undertaken when evaluation of the ac- tual intervention would be impractical. Simulations may involve manipulation of iconic, analog, or symbolic (often mathe- matical) models. Group judgment: a process in which a group of experts interact in assessing a technology 259 and formulate findings by vote or other process of reaching general agreement. The findings may note minority opinions; the group may determine that there is no consensus of opinion. The process may be informal, or it may be a formal one such as the nominal group or Delphi technique. Members of the group may be involved in drafting, editing, reviewing, and/or com- menting upon the findings. To be catego- rized as a group judgment it is necessary that group members have the opportunity to interact in formulating and reviewing each other's and the group's observations and findings. Expert opinion: consultation with individual experts who may be involved in drafting, editing, reviewing, or commenting upon assessments, but who do not interact as a group. Literature syntheses: summarizing, integrat- ing, and interpreting research findings re- ported in the literature. May include un- structured literature reviews as well as various systematic and quantitative proce- dures such as meta-analysis. ACKNOWLEDGMENTS I wish to thank the following individuals for providing source material, reviewing drafts of the profiles, and sharing important insights into the assessment programs of which they are a part. John R. Ball, American College of Physicians Clyde I. Behney, Congressional Office of Tech- nology Assessment Nancy E. Cahill, American Medical Associa- tion Arthur L. Caplan, Hastings Center Enrique D. Carter, National Center for Health Services Research Susan M. Clark, National Institutes of Health Office of Medical Applications of Research Morris F. Collen, Kaiser Permanente Medical Care Program Dennis I. Cotter, Prospective Payment Assess- ment Commission Martin Erlichman, locational Center for Health Services Research

OCR for page 255
260 David I. Feild, American College of Cardiol- ogy Robert C. Flink, Medtronic Peter L. Frommer, National Heart, Lung, and Blood Institute Susan Gleeson, Blue Cross and Blue Shield As- sociation Mark D. Goodhart, American Hospital Associ- ation Jerome G. Green, National Heart, Lung, and Blood Institute Ping Huang, Veterans Administration Itzhak Jacoby, National Institutes of Health Office of Medical Applications of Research Richard J. Jones, American Medical Associa- tion Bryan R. Luce, Battelle Memorial Institute Harold Margulies, National Center for Health Services Research Judith D. Moore, Prospective Payment Assess- ment Commission ASSESSING MEDICAL TECHNOLOGY Lawrence C. Morris, Jr., Blue Cross and Blue Shield Association Robert Mosenkis, ECRI jay Moskowitz, National Heart, Lung, and Blood Institute Joel I. Nobel, ECRI Thomas D. Overcast, Battelle Human Affairs Research Centers Morton L. Paterson, Smith Kline & French Laboratories Jonathan A. Showstack, University of Califor- . ~ ~ . nova, ban r ranc~sco Elliot Siegel, National Library of Medicine Kent A. Smith, National Library of Medicine David Tennenbaum, Blue Cross and Blue Shield Association Malin VanAntwerp, ECRI Linda Johnson White, American College of Physicians Donald A. Young, Prospective Payment Assess- ment Commission

OCR for page 255
loins American College of Carcliology/American Heart Association Task Force on Assessment of Carcliovascular Procedures Heart House, 9111 Old Georgetown Road Bethesda, MD 20814 (301) 897 5400 Major Emphases of Technology Assessment Activities Concerns Technology Drugs Safety . Efficacy/ Cost/ Cost-Effect/ Effectiveness Cost-Benefit i _ X Ethical/Legal/ Social Medical Devices/Equipment/Supplies Medical/Surgical Procedures Support Systems Organizational/Administrative X Stage of Technologies Assessed X Emerging/new X Accepted use Possibly obsolete, outmoded Application of Technologies Prevention X Diagnosis/screening X Treatment Rehabilitation Assessment Methods Laboratory testing Clinical trials Epidemiological and other observational methods Cost analyses Simulation/modeling X Group judgment X Expert opinion X Literature syntheses Approximate 1985 budget for technology assessment: $12,000* * This is a rough estimate of the joint ACC/AHA task force budget only, and does not include budgets for other ACC or AHA activities. JOINT AMERICAN COLLEGE OF CARDIOLOGY/AMERICAN HEART ASSOCIATION TASK FORCE ON ASSESSMENT OF CARDIOVASCULAR PROCEDURES Introduction The Joint American College of Cardiol- ogy/American Heart Association Task Force on Assessment of Cardiovascular Procedures is a cooperative assessment effort of the two parent organizations, each of which conducts activities related to cardiovascular health. The American Heart Association (AMA) is a voluntary health agency devoted to the re- 26 auction of premature death and disability caused by heart and blood vessel diseases. The association has 55 affiliates nationwide, and its headquarters are in Dallas. In 1984, AHA provided $43.7 million for biomedical research, $24.6 million for public health edu- cation, $14.8 million for professional educa- tion and training, and $22.6 million for com- . . munlty services. The American College of Cardiology (ACC) is a 13,500-member, nonprofit profes- sional medical society. The mission of the col- lege is to ensure optimal care for persons with cardiovascular disease or the potential for de- veloping it and, ultimately, through appro-

OCR for page 255
262 priate educational and socioeconomic activi- ties, to contribute to the prevention of cardiovascular disease. Members are physi- cians and scientists concerned with clinical and basic science disciplines related to the cardiovascular system. Among its major activities, the ACC con- ducts a comprehensive program of continu- ing education and establishes standards of cardiovascular care. The college has three mechanisms for the assessment of new medi- cal knowledge and technology. Two of these, the Cardiovascular Procedures Committee and the Cardiovascular Norms Committee, are described briefly in this introductory sec- tion, along with other ACC activities related to assessment. The third, the Joint ACC/ AHA Task Force on Assessment of Cardiovas- cular Procedures, is the main subject of this profile. The ACC Cardiovascular Procedures Committee reviews requests from federal agencies as well as from the private sector (hospitals, clinics, third-party carriers) rela- tive to standards, criteria, appropriateness, etc., of procedures normally performed in a hospital setting by physicians treating cardio- vascular disease. Recommendations are for- warded from this committee to the president of the college. This committee consists of 12 members of the ACC. The ACC Cardiovascular Norms Commit- tee is a new activity begun in 1983. This com- mittee reviews and assists the Executive Committee of ACC and the president in re- sponding to requests concerning standards of care and in assessing proposed standards or norms of particular interest to the college membership. In addition, this committee ob- tains consensus on dynamic norms defined by ACC as factors essential for quality care- for the diagnosis and management of the most common cardiac disorders, including considerations of the cost-effectiveness of al- ternative management plans or diagnostic techniques. This committee consists of eight members of the ACC. Through these committees and related ac- tivities, the ACC has developed positions on such technologies as applicability of and indi- cations for phonocardiography, cardioky- mography, ergonovine testing, percutaneous ASSESSING MEDICAL TECHNOLOGY transluminal coronary angioplasty, pro- grammable pacemakers, transtelephonic pacemaker monitoring, and the training re- quirements for the safe handling of radioiso- topes utilized for cardiovascular diagnostic testing. Opinions have been rendered on au- tomated blood pressure monitoring, heparin infusion pumps, diagnostic endocardial elec- trical stimulation, intraoperative ventricular mapping, Doppler ultrasound, hyperbaric oxygen therapy, photoplethysmography, dig- ital subtraction angiography, and rapid se- quence pyelograms. The ACC has initiated conferences to iden- tify the state of the art on the relative sensitiv- ity, specificity, and indications for diverse techniques in the assessment of ventricular functions. For instance, The Twelfth Be- thesda Conference, held in 1981, on noninva- sive technology in the assessment of ventricu- lar function was a state-of-the-art conference to develop diagnostic strategies for using var- ious types of echocardiography, nuclear car- diologic techniques, cardiac computed to- mography, and digital subtraction angiog- raphy. ACC has testified in support of the Na- tional Heart, Lung, and Blood Institute ap- propriations and authorizations and has sup- ported increased warnings on cigarette packages, the desirability of sodium content labeling for prepared foods, special consid- eration for orphan drugs, and extension of patent protection time for drugs requiring prolonged clinical testing periods. ACC par- ticipates in the Medical Necessity Program of the Blue Cross and Blue Shield Association. The Joint ACC/AHA Task Force on Assess- ment of Cardiovascular Procedures (joint task force) first met in November 1981. Purpose The purpose of the Joint ACC/AHA Task Force on Assessment of Cardiovascular Pro- cedures (joint task force) is to define the role of noninvasive and invasive procedures in the diagnosis and management of cardiovascular disease. As opposed to the ACC Cardiovascu- lar Procedures and Cardiovascular Norms Committees, which respond to inquiries

OCR for page 255
APPENDIX A: PROFILES made by outside parties the joint task force ... . ' Process initiates its own assessments. Subjects of Assessment The first assessment to be completed was of cardiac pacemaker implantation, in 1984. Assessments ongoing in 1985 were of exercise stress testing and nuclear imaging proce- dures. Among other procedures that are un- der consideration for assessment topics are angiography, Hotter monitoring, echocar- diography, and intracardiac electrophysio- logical studies for management of arrhyth- mias. Stage of Diffusion The program is primarily concerned with new and emerging technologies. Concerns The joint task force is most concerned with the safety and efficacy of cardiovascular pro- cedures. Specifically, it may address the con- tribution, uniqueness, sensitivity, specificity, indications, and contraindications of cardio- vascular procedures. Although the charge of the joint task force also includes cost-effec- tiveness as a potential concern, this has not yet been explicitly addressed in deliberations to date. Requests It is anticipated that the joint task force will develop most of its assessment topics, al- though suggestions may come from either the ACC or the AHA, for instance from the ACC Cardiovascular Procedures Committee via the president. The topic of the cardiac pace- maker implantation was generated by the ACC president, and the topics of exercise stress testing and nuclear imaging were gen- erated by the joint task force. Selection The topics for assessment are chosen by consensus of joint task force members. 263 Assessment reports are written by ad hoc subcommittees designated by the joint task force, e.g., the Subcommittee on Pacemaker Implantation. Portions of the reports are drafted first by individual subcommittee members. In addition to sharing the initial drafts with other subcommittee members, the subcommittee members may also seek ad- vice and information from other experts. Subcommittee members then consolidate their section drafts into a single document, which is reviewed by all subcommittee mem- bers and which may be shared with other ex- perts for their opinion. In the case of the car- diac pacemaker implantation report, the subcommittee met four times. The final subcommittee draft is then for- warded to the joint task force for approval. The joint task force meets at least twice a year; its small size enables much of its work to be conducted by telephone and through the mail. Once approved by the joint task force, reports are forwarded to the presidents of ACC and AHA for approval by the organi- zations' respective ruling bodies. Final ap- proval is given in a letter signed jointly by the presidents of the two organizations. This ap- proval makes the report an official, jointly supported policy statement of the two orga- nizations. Assessors The joint task force has a chairman and four other members, two representing ACC and two representing AHA and designated by their respective organizations. The chair- person of the joint task force is selected by agreement of the two organizations. A1- though it may not always be the case, the chairperson is likely to be a member of one or both organizations. No terms of office have been set for joint task force members. The joint task force selects the subcommittee chairpersons and other subcommittee mem- bers. loins task force members may serve on the subcommittees. Subcommittee members are not necessarily members of either organi- zation and are not necessarily physicians, al-

OCR for page 255
264 though all subcommittee members appointed to date have been members of both the ACC and the AHA. Thus far, the number of mem- bers on the subcommittees has ranged from 6 to 11. Upon completion of their tasks, the ad hoc subcommittees are dissolved. Turnaround Turnaround time for joint task force re- ports will be variable, depending on the mag- nitude and complexity of the subject. It is anticipated that for most reports full turn- around time will be approximately 18 months. The time from adoption of cardiac pacemaker implantation as an assessment topic to publication of the report was 21 months. This included a year for appoint- ment and work of the subcommittee and ap- proval by the joint task force, 6 months for final approval by the two organizations, and publication 3 months thereafter. Reporting By agreement of ACC and AHA, final as- sessment reports are published simultane- ously in the Journal of the American College of Cardiology and Circulation, which is pub- lished by the AHA. There are 20,000 world- wide individual and organizational sub- scribers to the Journal of the American College of Cardiology, including all ACC members. Circulation goes out to 13,000 U.S. and 10,000 overseas subscribers. The re- ports are also distributed to the other domes- tic and overseas journals and members of the press, policymakers, and other parties that ASSESSING MEDICAL TECHNOLOGY may be interested in a given topic. Copies are also available on request from the ACC and the AHA. Impact There are currently no plans to study the impact of the assessment reports, other than through noting individual reactions to the re- ports and requests for reprints. Reassessment The joint task force will reassess a technol- ogy as warranted by new evidence regarding safety, efficacy, and appropriate use. Funding/Budget Expenses of the joint task force and staffing are shared equally by the two organizations. The annual budget of the joint task force is approximately $10,000 to $15,000. This amount covers direct costs only, and does not include estimates of indirect costs, the cost of staff time, the value of time provided by the committee members, or publication costs. Example On the following pages is the full text of the 1984 report of the Joint ACC/AHA Task Force on Assessment of Cardiovascular Pro- cedures on guidelines for permanent cardiac pacemaker implantation, published in the Journal of the American College of Cardiol- ogy. It is reproduced here with permission.

OCR for page 255
APPENDIX A: PROFILES Sources American College of Cardiology. 1982. Twelfth Bethesda Conference: Noninvasive technology in the assessment of ventricular function. American Journal of Cardiology 49:1309-1374. American College of Cardiology. 1985. Statements of charge: Joint ACC/AHA Task Force on Assessment of Cardiovascular Procedures (and Subcommittees), Cardiovascular Procedures Committee, and Cardio- vascular Norms Committee. American Heart Association. 1984. American Heart Association 1984 Annual Report. Dallas. Feild, D. J., Director, Special Projects, American College of Cardiology. 1985. Personal communica- tion. Joint American College of Cardiology/American Heart Association Task Force on Assessment of Car- 265 diovascular Procedures (Subcommittee on Pacemaker Implantation). 1984a. Guidelines for permanent car- diac pacemaker implantation, May 1984. Journal of the American College of Cardiology 4~2~: 434-442. Joint American College of Cardiology/American Heart Association Task Force on Assessment of Car- diovascular Procedures (Subcommittee on Pacemaker Implantation). 1984b. Special report: Guidelines for permanent cardiac pacemaker implantation, May 1984. Circulation 70:331A-339A. Knoebel, S. B. 1983a. President's page: The next challenge and balancing individual quality care with community resources. Journal of the American Col- lege of Cardiology 3:972-974. Knoebel, S. B. 1983b. Presentation to the Depart- ment of Health and Human Services Technology Co- ordinating Committee.

OCR for page 255
APPENDIX A: PROFILES TABLE A-16 Continued 479 Cooperative Studies in Final Analysis Hepatitis and Dentistry Patient Compliance and Its Role in Dental Plaque Control A Comparison of Hospital & Home Treatment Programs for Aphasic Patients Evaluation of Anti-Epileptic Drugs (Phenobarb vs. Phenytoin vs. Primidone vs. Carbamazepine) Drugs & Sleep Phase IIIComparison of Phenobarb & Dalmane in Insomnia EEG Antabuse in Treatment of Alcoholism A Randomized Comparison of the Peritoneo-Venous Shunt (LeVeen) & Conventional Medical Treatment Alone for Ascites in Patients with Alcoholic Cirrhosis The Treatment & Prevention of Infection-Induced Urinary Stones in Spinal Cord Injury Recently Completed Studies Renal Failure Self Care Dialysis (Hemo vs. Peritoneal Dialysis) Nafeillin Therapy of Staphylocoeeal Baeteremia Platelet Aggregation in Diabetes (Use of Aspirin & Persantine) Community vs. VA Nursing Home Care vs. Hospitalization in Psychiatric Patients Alcoholic Hepatitis (Steroid Therapy) Aspirin in Unstable Angina Vasodilators in Acute MI Characteristics of Psychiatric Programs & Their Relationship to Treatment Effectiveness Alcoholic Hepatitis (Steroid Therapy) Bowel Prep for Elective Colon Surgery SOURCE: Veterans Administration (1984~. Concerns The primary concerns of cooperative stud- ies are safety, efficacy, and cost-effective- ness. Although cost-effectiveness is not a con- cern for all studies, it is receiving greater attention, and its appropriateness will be considered for all future studies. Of the 14 studies in active planning (see Table A-16), at least 6 have an explicit cost-effectiveness component, e.g., in studies of hepatitis B vaccine, percutaneous transluminal coronary angioplasty, and continuous peritoneal dialy- sis. At least three ongoing studies have ex- plicit cost-effectiveness components, includ- ing studies of pneumococcal vaccine immunity in high-risk patients, percutaneous transluminal angioplasty in the lower ex- tremities, and parenteral nutrition in mal- nourished surgical patients. Requests The origination, selection, conduct, and reporting of VA cooperative studies follow well-defined guidelines. A cooperative study usually begins with the submission by a VA researcher- physicians and investigators in VA centers around the United States of a planning request, i.e., an initial study pro- posal, to the chief of the Cooperative Studies Program. Recently, the CSP office has begun to encourage studies in certain areas of spe- cial interest to the VA. Selection Planning requests are reviewed by VA pro- gram specialists who provide written cri- tiques and are then evaluated by a Triage Re- view Committee. This committee comprises primarily representatives of the Office of the Associate Chief Medical Director for Re- search and Development, Medical Research Service and Professional Services. They may decide to reject a request, assign a priority rating, or ask for additional information. The requests with a priority rating are put on a waiting list, and those with the highest pri- ority are chosen for planning. At the time that a study is approved for planning, and again when a study is ap-

OCR for page 255
480 proved for funding, VA medical centers are invited to participate. Medical centers seek- ing to participate are considered based on a number of criteria, including level of inter- est, availability of appropriate staff re- sources, and availability of eligible study pa- tients. A medical center's participation in a CSP is voluntary, but an agreement to partic- ipate means acceptance of the study protocol without change and acceptance of the prac- tices and guidelines of the CSP program. When a study is funded for planning, the principal investigator is notified by the CSP. Biostatisticians and, where appropriate, clin- ical research pharmacists are assigned to the study planning process and the production of a detailed final proposal. The principle inves- tigator is also informed as to which of the four CSP coordinating centers the study has been assigned. The four special centers estab- lished by the VA to support and coordinate the CSP are located at VA medical centers in Hines, Illinois; Palo Alto, California, Perry Point, Maryland; and West Haven, Connect- icut. These provide the biostatistical and data processing support and administrative coordination for the cooperative studies and ensure their compliance with program guide- lines. A CSP clinical research pharmacy coor- dinating center in Albuquerque, New Mex- ico, provides additional support and coordination for those studies involving drugs and devices, such as dispensing and monitor- ing drugs, and liaison with the FDA and pharmaceutical companies. Ethical, scientific, professional, budget- ary, and administrative aspects of the final proposal are evaluated by three groups, a hu- man rights committee, a cooperative studies evaluation committee, and a budget review group, and at least three independent re- viewers who provide written critiques of these same areas. Based on the judgment of these groups, proposed studies may be com- pletely rejected, rejected with recommenda- tions for resubmission, given conditional ap- proval, or given unconditional approval. Process Almost all VA cooperative studies are ran- domized clinical trials. A few have been non- ASSESSING MEDICAL TECHNOLOGY randomized trials and observational studies. Five groups share the responsibility for con- ducting or monitoring a cooperative study: the study group, the executive committee, the operations committee, the CSP coordi- nating center human rights committee, and the cooperative studies evaluation commit- tee. In general, the current schedule of meet- ings for the study group, executive commit- tee, and operations committee consists of an initial meeting for organizational, informa- tional, and training purposes prior to patient intake, a meeting 9 months after the initia- tion of patient intake, and annual meetings thereafter. If drugs or devices used in a study require FDA approval, an investigational new drug application or investigational device exemp- tion must be filed with the FDA before the study can begin. The study group, composed of all partici- pating investigators and permanent consul- tants to the study and chaired by the study chairperson, reviews the progress of the study, discusses problems encountered, and provides suggestions for improving the study. Results of blind data related to study end points are not discussed with this group. The executive committee is the manage- ment group and major decision-making body for the operational aspects of the study. It in- cludes the study chairperson, the study bio- statistician, the clinical research pharmacists, the hearths) of any special central support unitsts), two or three participating investiga- tors, and selected consultants. This committee decides on all changes in the study and on any subprotocols or other use of the study data and on publications of study results, and takes actions on medical centers whose performance is unsatisfactory. As with the study group, the results of the blind portions of the study are not presented to this group. The operations committee usually consists of experts in the subject matter of the study, an independent biostatistician, and other technical or scientific consultants. Nonvoting members include the study chairperson and the study biostatistician. This committee considers from meeting to meeting whether the study should continue, based on study performance, patient accrual, treatment effi-

OCR for page 255
APPENDIX A: PROFILES cacy, adverse effects, and other factors; as- sesses the performance of each participating center and makes recommendations regarding continuation, termination, or change in sup- port; and reviews and provides recommenda- tions regarding protocol changes. The human rights committee, besides re- viewing the protocol for human rights issues prior to initiating the study, is responsible for ensuring that patients' rights are protected during the course of the study. This commit- tee meets at least once a year for the duration of the study with the operations committee. Each year, the human rights committee con- ducts three site visits to participating medical centers to ensure that the human rights as- pects of the studies are being observed. . . . . All cooperative studies undergo an in- depth review by the cooperative studies eval- uation committee at 3-year intervals during their active phase. Assessors The assessors include physicians and other health care providers, biostatisticians, clini- cal pharmacists, and others noted above. Turnaround The time from submission of a planning re- quest to approval and initiation of formal planning is generally 4 months. Another 12 months are taken between formal planning of a study and its approval and initiation. Once begun, cooperative studies may range from 1 to 8 years. Reporting All VA medical centers conducting medi- cal research must report regularly to the VA Medical Research Service. Summaries of co- operative studies are included in the annual report to Congress of medical research in the VA. Where applicable, sponsors of investiga- tional new drugs and investigational devices are required to submit annual reports to the FDA. The primary means of keeping cooper- ative study participants informed between meetings are study newsletters prepared and issued regularly by the study chairperson and 481 the study biostatistician or by the executive committee. The presentation or publication of data collected by investigators on patients entered into VA cooperative studies is under the con- trol of the study's executive committee. The results of cooperative studies are published in major refereed journals as the New England Journal of Medicine, Circulation, Journal of the American Medical Association, Archives of General Psychiatry, and others. Impact Results from VA cooperative studies have made significant impacts on clinical practice in the VA systems, as well as on clinical prac- tice at large. Among the most significant studies have been those on aspirin therapy of unstable angina, drug treatments for moder- ate and severe hypertension, chemotherapy for schizophrenia, and coronary artery by- pass surgery in chronic stable angina. Reassessment There have been several instances of reas- sessment of drugs, such as reserpine for hy- pertension and the beta-blocker propranolol for hypertension and for angina. These reas- sessments occur when new drugs and/or regi- mens are compared to prevailing treatments in clinical trials. Funding/Budget Although most VA cooperative studies are supported by the Medical Research Service, occasionally studies are funded by other VA sources or by outside sources such as NIH or the pharmaceutical industry. The funding level for each cooperative study is deter- mined by the Budget Review Group. The 1985 budget for the Cooperative Stud- ies Program is approximately $12 million. However, this is deceptively small. These funds are primarily for support of the CSP coordinating centers and other nonpatient care aspects of the cooperative studies. The clinical costs of these trials is met entirely through VA medical benefits, and is not re- flected in the CSP budget.

OCR for page 255
482 Example On the following pages is a report by the VA Coronary Artery Bypass Surgery Cooper- ative Study Group, published in the New En- gland Journal of Medicine, November 22, 1984, and is reproduced here with permis- sion. Sources Blue Sheet. 1984. VA spending $4.3 million in EY 1984 on prosthetic/amputation R&D. May 30:P&R-8. Hagans, J., Chief, Cooperative Studies Program, Veterans Administration, Washington, D.C. 1984. Personal communication. Huang, P., Staff Assistant, Cooperative Studies Program, Veterans Administration, Washington, D.C. 1985. Personal communication. ASSESSING MEDICAL TECHNOLOGY Congressional Budget Office. 1984. Veterans Ad- ministration Health Care: Planning for Future Years. Washington, I).C. Goldschmidt, P., Director, Health Services Re- search and Development Service, Veterans Adminis- tration, Washington, D.C. 1984. Personal communi- cation. Office of Technology Assessment. 1983. The Im- pact of Randomized Clinical Trials on Health Policy and Medical Practice. Washington, D.C.: U.S. Gov- ernment Printing Office. Veterans Administration. 1983. Administrator of Veterans Affairs Annual Report 1982. Washington, D.C. Veterans Administration. 1983. Guidelines for the Planning and Conduct of Cooperative Studies in the Veterans Administration. Sixth Edition. Washington, r).c. Veterans Administration. 1984. Cooperative Stud- ies Program Status Report. Washington, D.C.

OCR for page 255
APPENDIX A: PROFILES 483 ELEVEN-YEAR SURVIVAL IN THE VETERANS ADMINISTRATION RANDOMIZED TRIAL OF CORONARY BYPASS SURGERY FOR STABLE ANGINA THE VETERANS ADMINISTRATION CORONARY ARTERY BYPASS SURGERY COOPERATIVE STUDY GROUP Abstract We evaluated long-term survival after coro- nary-artery bypass grafting in 686 patients with stable an- gina who were randomly assigned to medical or surgical treatment at 13 hospitals and followed for an average of 11.2 years. For all patients and for the 595 without left main coronary-artery disease, cumulative survival did not differ significantly at 11 years according to treatment. The 7-year survival rates for all patients were 70 per cent with medical treatment and 77 per cent with surgery (P = 0.043), and the 11-year rates were 57 and 58 per cent, respectively. For patients without left main coronary- artery disease, the 7-year rates were 72 and 77 per cent in medically and surgically treated patients, respectively (P = 0.267), and the 11-year rates were 58 per cent in both groups. A statistically significant difference in survival suggest- ing a benefit from surgical treatment was found in patients without left main coronary-artery disease who were sub- divided into high-risk subgroups defined angiographical- ly, clinically, or by a combination of angiographic and clinical factors: (1) high angiographic risk (three-vessel disease and impaired left ventricular function) at 7 years, 52 per cent in medically treated patients versus 76 per cent in surgically treated patients (P = 0.002); at 11 years, 38 and 50 per cent, respectively (P = 0.026); TN 1975 the Veterans Administration Cooperative 1 Study of Surgery for Coronary Arterial Occlusive Disease first reported a statistically significant surviv- al difference in favor of surgery in the subgroup of patients with left main coronary-artery disease. Two years later, in a preliminary reports on patients with- out disease in the left main artery who were followed for a minimum of 21 months, no significant difference in survival was found between medical and surgical treatment groups either overall or in angiographically defined subgroups. Subsequently, a high-risk sub- group of patients without left main coronary-artery disease, defined on the basis of clinical risk factors alone, was reported to have a significantly reduced five-year cumulative mortality with surgery.3 This report compares 7-year and 1 1-year survival after assignment to medical and surgical treatment in patients who were followed for a minimum of 107 months. Survival results for the entire group as well as for risk groups defined by angiographic and clinical measures are also presented for patients without left main coronary-artery disease. Updated survival re- sults for patients with such disease have been reported previously.4 Report prepared by Katherine M. Detre, M.D., D.P.H., Peter Peduzzi, Ph.D., Timothy Takaro, M.D., Herbert N. Hultgren, M.D., Marvin L. Murphy, M.D., and George Kroncke, M.D. Address repent requests to Dr. Detre at the Veterans Administration Medical Center, West Haven, CI 06516. For a complete listing of participants, members of the Operations and Executive Committees, Coordi- nat~ng Center staff, and consultants, refer to Circulation 1981, 63:1329 (Ap- pendix C). Supported by the Veterans Administration Cooperative Studies Pro- gram, Medical Research Service, Veterans Adminis~anon Central Office, Wash- ington, D.C. (2) clinically defined high risk (at least two of the fol- lowing: resting ST depression, history of myocardial in- farction, or history of hypertension)at 7 years, 52 per cent in the medical group versus 72 per cent in the surgi- cal group (P = 0.003); at 11 years, 36 versus 49 per cent, respectively (P = 0.015); and (3) combined angiographic and clinical high risk at 7 years, 36 per cent in the medical group versus 76 per cent in the surgical group (P = 0.002); at 11 years, 24 versus 54 per cent, respec- tively (P = 0.005). Survival among patients with impaired left ventricular function differed significantly at 7 years (63 per cent in the medical group versus 74 per cent in the surgical group [P = 0.049]) but not at 11 years (49 versus 53 per cent). The surgical treatment policy resulted in a nonsignifi- cant survival disadvantage throughout the 11 years in sub- groups with normal left ventricular function, low angio- graphic risk, and low clinical risk, and a statistically significant disadvantage at 11 years in patients with two- vessel disease. We conclude that among patients with stable ischemic heart disease, those with a high risk of dying benefit from surgical treatment, but beyond seven years the sur- vival benefit gradually diminishes. (N Engl J Med 1984; 311 :1333-9.) METHODS The Veterans Administration cooperative study of coronary-ar- tery bypass grafting is a randomized controlled trial of medical therapy versus medical plus surgical therapy for the treatment of patients with stable angina pectoris and angiographically confirmed coronary-artery disease. The study design, entry criteria, and base- line characteristics of the patient population have been described previously.S Briefly, between 1972 and 1974, 686 patients with sta- ble angina pectoris of more than six months' duration who had been receiving medical therapy for three months and who had resting or exercise electrocardiographic evidence of myocardial ischemia were randomly assigned to medical or surgical therapy. Patients were excluded from randomization if they had had a myocardial infarc- tion within six months or if they had refractory systemic diastolic hypertension (>100 mm Hg), left ventricular aneurysm or other serious cardiac disease, other organ-system disease making surgery inadvisable or limiting life expectancy to less than five years, unsta- ble angina, or uncompensated congestive heart failure. In the 1972-1974 cohort, 354 patients were randomly assigned to medical therapy, and 332 to surgical therapy, at a total of 13 clinical sites. The base-line distribution of risk factors (history, angiograph- ic findings, electrocardiographic findings, and severity of angina) was comparable in the two treatment groups.6 Twenty patients randomly assigned to bypass surgery did not have an operation. Ninety-four per cent of those who underwent surgery did so within three months after random assignment. The average number of diseased vessels in surgically treated patients was 2.4, and the average number of grafts placed was 2.0. All 45 patients with single-vessel disease received at least one graft, and one fourth received multiple grafts. Of the 102 patients with two- vessel disease, 80 per cent received two or more grafts. Of the 163 patients with triple-vessel disease, 90 per cent received two or more grafts, and 37 per cent received three or more. The overall 30-day operative mortality rate was 5.8 per cent. The incidence of perioperative myocardial infarction, calculated on the basis of the development of new Q waves, was 9.9 per cent. Vein- graft angiography was performed in 79 per cent of surgical patients (247 of 312) between 10 and 15 months after surgery, and 353 of 503 grafts placed (70 per cent) were patent at one year; 87 per cent of

OCR for page 255
484 ASSESSING MEDICAL TECHNOLOGY patients (214 of 247) had at least one patent graft. Five years after random assignment, 52 of the 312 patients had died, and 9 had had a second operation. Graft angiography was done at five years in 156 of the 251 remaining eligible patients (62 per cent), and the patency rate was 67 per cent. The one-year and five-year graft-patency rates for the 143 patients evaluated at both times were 74 and 67 per cent, respectively. Of 354 patients randomly assigned to medical treatment, l 33 (38 per cent) had bypass surgery during an average follow-up of 11.2 years. Of the 133, 22 had left main coronary-artery disease and crossed over to surgery on an elective basis in accordance with a protocol amendment.7 8 Thirty-five ( l l per cent) of the 312 patients randomly assigned to surgery who had coronary-artery bypass grafting have had repeat grafting. Medical therapy consisted of nitrates, beta-blockers, and other medications administered to achieve symptomatic relief of angina. At one year, 26 per cent of medically treated patients were taking nitroglycerin or nitrates only, 3 per cent were using propranolol only, and 65 per cent were taking both types of medication. The corresponding rates at five years were 27, 4, and 57 per cent. At one year and five years, 6 and 12 per cent of patients, respectively, were not taking any medication. The surgical patients took less medica- tion than the medical patients, but their use of medication increased between one and five years. At one year, 42 per cent took no medica- tion, 45 per cent were taking nitroglycerin or nitrates only, and the remaining 13 per cent were taking both types of medication. The corresponding rates at five years were 25, 36, and 36 per cent. At five years 3 per cent of surgical patients were taking propranolol alone. The average follow-up time was calculated as the average time from the date of random assignment to the date of the analysis. Included in the calculation were patients who died and those known to be alive but without current follow-up data except for survival status. Cumulative survival rates were determined by the actuarial life- table method, with death from all causes used as the end point. The survival status for patients who did not return for scheduled follow- up visits was ascertained by a retrieval system known as BIRLS (Beneficiary Identification and Records Locator Subsystem, Austin Tex.), which is unique to the Veterans Administration network. Telephone contact was used in the few cases in which BIRLS had no patient record. The survival status was known for all but one pa- tient. At present, 99.9 per cent of study patients have completed 9 years of follow-up, 91 per cent have com- pleted 10 years, and 73 per cent have com- pleted 11 years. Life-table cumulative sur- vival rates were calculated according to the original treatment assignment (treatment policy) from the date of randomization. Dif- ferences in cumulative survival between the p=.124 p=.043 p=.45 p=.527 p=.267 p=.813 two treatment groups were assessed by the Mantel-Haenszel test. Thus, the 7-year sta- 1.00 fistic represents the cumulative survival ex- ~ ~ ~~ perience up to 7 years, and the 11-year sta- A 90 - _ 83 - ~ .82 talc represents the cumulative experience > .80 ~'~ _ 77 - ~ .77 up to 1 1 years. All P values reported are two- > 78~ ~ .80 ` tailed and uncorrected for multiple compar- ~ 70 ~ ~~ - - 72'~'-~~ isons. z .60 `%o. 58 ~8 _ In addition to the overall treatment re- O 57 .58 suits, survival was compared in angiographi- ~ 50 cally and clinically defined subgroups of 0 .40 - patients without left main coronary-artery ~ disease. The angiographically defined sub- ~ 30 groups were identified on the basis of the ~ .20 ~ SURGICAL SURGICAL number of vessels diseased (one, two, or >-~ MEDICAL on oMEDICAL three) and left ventricular function. The .10 - 332 S 479 S - - 284 S ,52 S presence or absence of left ventricular func- 354 M ~ 84 M 34 ~ M ~ 68 M tional impairment was determined accord- ing to a central reading of base-line left ventriculograms performed at the Seattle Veterans Administration Medical Center under the supervision of Dr. Karl Hammer- meister. Global ejection fractions and seg- mental contraction abnormalities were measured. Contraction abnormalities were graded as follows: 1, no abnormality; 2, minimal hypokinesis or akinesis involving less than 25 per cent of the heart border; 3, moderate hypokinesis or alcinesis of 23 to 75 per cent of the heart border; 4, dyskinesia, left ventricular aneurysm, or paradoxical wall motion; and 5, severe generalized hypokinesis or akinesis of the entire heart border. Left ventricular function was considered to be impaired if the global ejection fraction was less than 50 per cent or if the contraction grade was over 2; otherwise, left ventricular function was defined as normal. By this definition, 55 per cent of the patients had impaired function. In previous reports-5 9 left ventricular function was defined as abnor- mal if there was cardiac enlargement, elevated end-diastolic pres- sure (>14 mm Hg), an ejection fraction under 45 per cent, or any degree of contraction abnormality, as evaluated by the individual participants. Central readings on left ventriculograms were done in 75 per cent of the patients. Base-line readings by the individual clinics were used to provide data on left ventricular function in the other patients, according to the new definition. The combination of three-vessel disease and impaired left ventricular function was clas- sified as a high angiographic risk. All other combinations of one-, two-, or three-vessel disease and normal or impaired left ventricular function constituted a low angiographic risk. Subgroups with a low, middle, or high clinical risk were defined on the basis of a multivariate risk function in order to predict five- year mortality,3 ~ using four established clinical-risk variables measured at base line: the New York Heart Association classifica- tion, a history of hypertension, a history of myocardial infarction, and an ST-segment depression on the resting electrocardiogram. Other risk factors, which were not uniformly measured in all pa- tients at base line (e.g., positive exercise test) or which had a low prevalence in the study population (e.g., congestive heart failure), could not be considered for inclusion in this risk function. Patients in the low-risk subgroup included those with none or only one of the four risk factors except for ST depression. The high-risk subgroup consisted of patients with combinations of two or three of the strong- est predictors (ST depression, a history of myocardial infarction and a history of hypertension)i.e., those with multiple clinical risk factors. The validity of this method for the classification of patients into clinical-risk groups has been established in an inde- pendent population.~ Reviewers and others ~2 have criticized such "post hoc" subgroups. However, the original protocol for the 1972-1974 Veterans Administration study clearly outlined the ana- CUMULATIVE SURVIVAL RATES ALL HOSPITALS ALL PATIENTS ALL NON-LMD .00 .90 80 70 .60 .50 .40 .30 .20 .10 1 2 3 4 5 6 i 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 YEAR Figure 1. Eleven-Year Cumulative Survival for All Patients and for Those without Left Main Coronary-Artery Disease (non-LMD), According to Treatment Assignment. Numbers of patients at risk are given at bottom of figure. M denotes medical, and S surgical.

OCR for page 255
APPENDIX A: PROFILES 1yncal steps for development of both the ang~ographic- and clinical- nsk groups: "Regression analyses will be used to determine the effect [of base-line variables] on mortality and to determine groups with low and high mortality Ask.... Analyses [life table] will be done for patients in low and high risk groups." Thus, the clinical- nsk groups, defined by noninvasive base-line measures, were nei- ther more nor less post hoc than the ang~ographic-nsk groups, de- fined by such factors as left main coronary-artery disease or the combination of vessel disease and left ventricular function. Survival analyses are based on data from all 13 hospitals, by treatment policy. In the Results we focus on the 7-year and 11-year cumulative survival rates. The five-year survival rates (early expe- r~ence) are shown in the figures but are not discussed in the text, except in comparing this study with the European and the Nanona1 Heart, Lung, and Blood Institute studies. All treatment comparisons are presented in reference to the medical treatment policy. RESULTS Cumulative survival rates are shown in Figures 1 through 5. Patients with left main coronary-artery dis- ease are excluded from the subgroup analyses shown in Figures 2 through 5. Overall Results Overall, excluding patients with left main coronary- artery disease, the treatment difference was not signifi- cant at seven years (72 per cent in the medical group versus 77 per cent in the surgical group, P = 0.267; Fig. 1~. The medical and surgical survival curves con- verge when follow-up data are extended to 11 years (58 per cent in both groups, P = 0.813~. During the first 7 years of follow-up the average annual mortality rates were 4.0 per cent for medical therapy and 3.3 per 1.00 .90 ,>, .80 is, . 70 CO z .60 i_ .50 O .40 O .30 .20 .10 485 cent for surgery, including operative mortality, as compared with 3.5 and 4.8 per cent, respectively, be- tween 7 and 1 1 years. If patients with left main coronary-artery disease are included, the trends are similar, although the treatment difference in all patients was statistically significant at 7 but not at 11 years. Vessel Disease There was a nonsignificant trend toward improved survival with surgery at seven years in the subgroup of patients with three-vessel disease: 63 per cent with medical treatment vs. 75 per cent with surgical treat- ment, P0.061 (Fig. 2~. The difference in the cumu- lative survival rates diminished after 7 years, resulting in only a 6 per cent difference at 11 years. At 7 years neither patients with single-vessel disease nor those with double-vessel disease had a significant difference in survival associated with treatment, although at 11 years surgically treated patients with two-vessel dis- ease had a marginally significant disadvantage in sur- vival (P = 0.045). Left Ventricular Function At 7 but not at 11 years, there was a significant difference in survival between medically and surgical- ly treated patients with impaired left ventricular func- tion (63 vs. 74 per cent, respectively; P = 0.049); sur- vival rates at 11 years were 49 and 53 per cent, respectively (P = 0.249, Fig. 3~. Among patients with normal left ventricular function, survival was 84 per CUMULATIVE SURVIVAL RATES ALL HOSPITALS 1 VESSEL DISEASE 2 VESSEL DISEASE 3 VESSEL DISEASE I I I I 1 1 1 1 1 1 1 _' ~ ~ ~ ~ ~ ~ I 1 1 1 I I I I 1 1 1 p= 291 p= 522 p= 589 p= 267 p= 213 p= 045 p= 305 p= 061 p= 16 ~-~z p..9 9'S7~3 .. I ~ SURGICAL ~ SURGICAL SURGICAL ~ oMEDlcAL of oMEDlcAL of o MEDICAL 1 - 49 S 31 S - - 98 S 49 S - - 135 S 70 S - 48 M 30 M 105 M 67 M 156 M 71 M ~ I ~ 1. 1 1 1 ~ I I I ~ 1 1 1 1 1 1 1 1 1 1 1 1 2 3 4 5 6 7 8 9 1011 1 2 3 4 5 6 7 8 9 1011 1 2 3 4 5 6 7 8 9 1011 YEAR Figure 2. Eleven-Year Cumulative Survival for Patients without Left Main Coronary-Artery Disease Who Had Single-, Double-, or Triple-Vessel Disease. Numbers of patients at risk are given at bottom of figure. M denotes medical, and S surgical.

OCR for page 255
486 1.00t oz to } .80 ~ .70 _ an z .60 O- .50 G O .40 _ O .30 .20 _ 10 IMPAIRED LVF p= 130 p= 049 63~>~53 . 49 SURGICAL -~oMEDICAL CUMULATIVE SURVIVAL RATES ALL HOSPITALS l p= .249 _ - 150S 77S 1 75 M 79 M 1 NORMAL LVF p= 229 p=.350 p= 249 84 85 ~~ ~71 64 _ - ~ SURGICAL on-"oMEDICAL 133 S 74 S - 135 M 88 M I 1 1 1 1 1 1 1 1 1 1 I I ~ 1 1 1 1 1 1 1 1 2 3 4 5 6 7 8 9 10 11 1 2 3 4 5 6 7 8 9 10 11 YEAR Figure 3. Eleven-Year Cumulative Survival for Patients without Left Main Coro- nary-Arte~ Disease, According to Whether Left Ventricular Function (LVF) Was Im- paired or Normal. Numbers of patients at risk are given at bottom of figure. M denotes medical, and S surgical. cent in the medical group and 80 per cent in the surgical group at 7 years, and 71 and 64 per cent, respectively, at 11 years. These differences were not . . s~gn~hcant. Risk Analysis There was a statistically significant improvement in survival at seven years in surgically treated patients with a high angiographic risk (three-vessel disease with impaired left ventricular func- tion); the rates were 52 per cent for medical treatment versus 76 per cent for surgical treatment (P = 0.002, Fig. 4~. Although from 7 to 11 years, the marked difference in survival rates diminished from 24 to 12 per cent, the difference in cumulative survival up to 11 years remained significant (P= 0.026~. In contrast, by 11 years, patients with a low angiographic risk had a survival rate of 68 per cent with medical therapy and 61 per cent with surgery (P = 0.105~. The Veterans Administration study reported a significant surviv- al benefit with surgery at five years in patients with a high clinical risk and a significant benefit with medi- cal therapy in patients at low risky The seven-year survival rates in the high-risk tercile were 52 per cent for the medical group and 72 per cent for the surgical group (P = 0.003, Fig. 5). Beyond seven years, surviv- 1 00 \ .80 ~ .70 U) id o - o o .60 .50 .40 .30 at ASSESSING MEDICAL TECHNOLOGY al remained higher in the surgical group, but the difference gradually diminished from 20 to 13 per cent by 11 years. The cumulative surviv- al experience up to 11 years differed significantly between the two treat- ment groups (P = 0.015~. In con- trast, for patients in the low-risk tercile there was a 7 per cent surviv- al disadvantage with surgical ther- apy at seven years (88 per cent in the medical group vs. 81 per cent in the surgical group, P = 0.093), which increased to 10 per cent at 1 1 years (73 per cent in the medical group vs. 63 per cent in the surgical group, P= 0.066~. In the middle- risk tercile, there was no significant difference in survival at any time. When survival in the subgroup of patients with a high angiographic risk was studied separately in the patients at low, middle, and high clinical risk, a statistically signifi- cant surgical benefit was observed only in patients who were at high risk not only angio- graphically but also clinically (Table 1~. No signifi- cant difference in survival was observed in patients with a high angiographic risk who had a low or middle clinical risk. Although the survival experience was sim- ilar in all clinical-risk subgroups of surgically treated patients with a high angiographic risk, survival in medi- cally treated patients decreased with increasing clini- cal risk at both 7 and 11 years, reflecting the strong CUMULATIVE SURVIVAL RATES ALL HOSPITALS ANGIOGRAPHIC HIGH RISK ANGIOGRAPHIC LOW RISK l l l 4 p = 018 p= 002 p= 026 1 - - 3 ) - Abe `~ N ok" ~50 52' - ~` 38 ~ SURGICAL .__ on oMEDlcAL .10 - 71 S 97 M 1 1 1 1 1 1 1 1 ~ 1 ~ 1 2 3 4 5 6 7 8 9 10 11 p= 156 p= 156 p= 105 O W_~ ~ 35 S 35 M SU RG I CAL ~ MEDICAL - 210 S 114 S - 211 M 132M _ 1 1 1 1 1 1 1 1 1 1 1 2 3 4 5 6 7 8 9 10 11 YEAR Figure 4. Eleven-Year Cumulative Survival for Patients without Left Main Coronary- Artery Disease, According to Angiographic Risk. High risk was defined as three-vessel disease plus impaired left ventricular function, and low risk as one-, two-, or three-vessel disease plus normal left ventricular function or one- or two-vessel disease plus impaired left ventricular function. Numbers of patients at risk are given at bottom of figure. M denotes medical, and S surgical.

OCR for page 255
APPENDIX A: PROFILES additional effect of clinical risk factors on the natural history of patients with a high angiographic risk. Conversely, when survival in patients with a low angiographic risk was studied separately in the three clinical risk groups (Table 1), a disadvantage with surgery was observed at 7 and 11 years in patients who were at low risk by both measures; however, the differences were not significant. Again, the strong ef- fect of the clinical risk factors on the natural history was evident. Left Main Coronary-Artery Disease The mortality rate for 48 patients with left main coronary-artery disease who were randomly assigned to surgical treatment increased after the seventh year of follow-up (data not shown). During the first seven years, the average annual mortality rate was approxi- mately 3 per cent; thereafter, the rate increased to nearly 5 per cent. Comparison with the assigned medi- cal group was futile, since 47 per cent of the original medical group crossed over and 44 per cent died, leav- ing only four patients with the original treatment as- signment at seven years. DISCUSSION Current survival results by treatment policy for all 13 hospitals with an operative mortality rate of 5.8 per cent (Fig. 1 ) indicate that for all patients and for those without left main coronary-artery disease, the cumula- tive survival experience up to 11 years did not differ significantly (at a two-tailed alpha level of 5 per cent) between medical and surgical treatment groups. 1.00 .10 487 This overall result disregards the heterogeneous natural history of the subgroups. Consistent with the hypothesis that surgery could be advantageous for pa- tients whose natural history was expected to be poor but would offer little or no advantage for those with a good prognosis, we found a range of treatment effects in subgroups, from a significant advantage to a bor- derline or nonsignificant advantage and even a disad- vantage, with surgical therapy. In particular, the small group of patients who were at high risk both an- giographically and by noninvasive clinical risk meas- ures derived the greatest survival benefit from surgery (Table 1), second only to the benefit in patients with left main coronary-artery disease reported previous- ly. l'4 A statistically significant difference was also found when subgroups were defined by angiographic predictors alone i.e., three-vessel disease and mod- erate to severe impairment of left ventricular function (Fig. 4) or by the high-clinical-risk measure alone (Fig. 5~. The surgical benefit was not significant at 11 vears in the subgroups with three-vessel disease alone (Fig. 2) or with moderate to severe impairment of left ventricular function (Fig. 3), although a bene- fit of borderline significance appeared at 7 years (P = 0.061 and 0.049, respectively). At the other end of the spectrum were the subgroups of patients with a good prognosis: those with one- or two-vessel disease (Fig. 2), normal left ventricular function (Fig. 3), a low angiographic risk (Fig. 4), and a low clinical risk (Fig. 5~. With the exception of the group with two-vessel disease, for which the surgical survival rate was significantly worse (P = 0.045), the CUMULATIVE SURVIVAL RATES ALL HOSPITALS H I GH RISK TERCILE MID RISK TERCILE LOW RISK TERCILE l l l l l l l ~ l l l l l l l ,,,,,,,,, r, p=.0006 p=.003 p=.015 p=.818 p=.34s p=.737 p=.012 p=.093 p=.066 ~90 ~04~ ~ `~37 ~7'~ O .30 .36 _ 20 ~ SURGICAL ~ SURGICAL -- SURGICAL _ ~ MEDICAL on MEDICAL on MEDICAL - 83 S 37 S - - 84 S 48 S - - 113 S 66 S - 94 M 33 M 91 M 53 M 123 M 81 M I I 1 1 1 1 1 1 1 1 1 l l ~ I I I I I I I I I 1 1 1 2 3 4 5 6 7 8 9 1011 1 2 3 4 5 6 7 8 91011 1 2 3 4 5 6 7 8 9 1011 YEAR Figure 5. Eleven-Year Cumulative Survival for Patients without Left Main Coronary-Artery Disease, According to Clinical Risk. See text for definitions of high, middle, and low clinical risk. Numbers of patients at risk are given at bottom of figure. M denotes medical, and S surgical.

OCR for page 255
488 Table 1. Cumulative Survival Rates at 7 and 11 Years In Patients without Left Main Coronary-Artery Disease, According to Angio- graphic and Clinical Risk.* 7-YEAR RATE (%) 1 I-YEAR RATE (%) NO. MED t SURG t P MED t SURG t P rare +S.E. rare +S.E. High sagiographic risk High clinical risk 67 36+7 76+9 0.002 24+7 54+10 0.005 Middle clinical risk 49 SO+9 79+9 0.069 40+9 i3+13 0.345 50 79+8 77+8 0.818 62+10 52+10 0.586 Low clinical risk Low ang~ographic risk High clinical risk 108 67+7 70+6 0.531 46+7 46+7 0.732 Middle clinical risk 124 82+5 78+5 0.569 73+6 68+6 0.521 Low clinical risk 184 90+3 81+4 0.079 76+5 66+5 0.092 *A total of 13 pacts could not be classified: 4 had missing data for the number of diseased vessels, 2 for Ich ventricular function, and 7 for the cl~nical-nsk subgroup. tMed denotes medical group, and Surg surgical group. 11-year survival disadvantage with surgical treatment was not statistically significant in the low-risk sub- groups. The small survival disadvantage with surgery in these subgroups can probably be explained by the initial mortality associated with surgery. Although the individual subgroup results reported here are weakened by the multiplicity of comparisons and the loss of power in some strata with small num- bers of patients, the survival differences during the first seven years can generally be characterized by what is known about the natural history of chronic stable angina. Depending on the presence or absence of known prognostic indicators, the seven-year cumu- lative mortality in the medical cohort varied widely, from 10 to 64 per cent, yet mortality in the surgical cohort ranged only from 12 to 30 per cent. This result- ed in a significant reduction of mortality for surgically treated patients with combinations of risk factors. Two other large-scale randomized studies of coro- nary-artery bypass grafting have reported survival re- sults: the European Coronary Surgery Studyi3 and the National Heart, Lung, and Blood Institute's Coronary Artery Surgery Study (CASS).~4 Comparison of the three studies is difficult, because each enrolled differ- ent types of patients, and the periods of patient enroll- ment were different. For example, unlike the Veterans Administration study, CASS did not enroll patients with serious left main coronary-artery disease or with severe angina,~4 and the European study did not en- roll patients with an ejection fraction under 50 per cent or with single-vessel disease.~3 Probably in part because of differences in patient characteristics and in part because of improvement in both therapies, the five-year survival rates in medi- cally and surgically treated patients without left main coronary-artery disease differed among the three stud- ies. For medically treated patients the rates were 80 per cent in the Veterans Administration study, 85 per cent in the European study (calculated from reported data'3), and 92 per cent in CASS; for sur- gically treated patients the rates were 82, 93 (cal- ASSESSING MEDICAL TECHNOLOGY culated from reported data~3), and 95 per cent, respec- tively. The corresponding percentage reductions in overall mortality for the surgical group as compared with the medical group were 10 per cent in the Veter- ans Administration study, 53 per cent in the European study, and 38 per cent in CASS. Neither the Veterans Administration result nor the CASS result was statis- tically significant. The subgroups with the largest benefit from surgi- cal treatment in the Veterans Administration study have no direct counterparts in the other two studies. The subgroup of patients in CASS that resembles most closely the Veterans Administration high-angio- graphic-risk group, the subgroup with three-vessel disease and an ejection fraction under 50 per cent, had a nearly significant benefit from surgery at five years (P = 0.063~. On the other hand, the European study, which examined the joint effect of clinical and angio- graphic risk factors, concluded that "in the absence of "clinically defined] prognostic variables in patients with either two- or three-vessel disease the outlook is so good that early surgery is unlikely to increase the prospect of survival." Thus, all three studies point toward a possible surgical benefit in the presence of high risk i.e., multiple risk factors that indicate a poor prognosis measured by angiography or by other means. With data now available from the extended follow- up in the Veterans Administration study, long-term results of therapy can be studied. So far the most , important observation is that the mortality rate in all surgical subgroups increased between 7 and 11 years. During the first seven years of follow-up, the average annual mortality rate was 3.3 per cent for all surgi- cally treated patients without left main coronary-ar- tery disease, as compared with a rate of 4.8 per cent during the next four years. For medically treated pa- tients without left main coronary-artery disease, the rates were 4.0 and 3.5 per cent, respectively. The in- creased mortality in the surgical group is consistent with the findings of the Montreal Heart Institute in- vestigators who followed a series of patients for 12 years. They found that although angina had improved in 80 per cent of patients at 6 years, it remained improved in only 47 per cent of the 12-year survi- vors. '5 The annual graft-closure rate at 7 to 12 years was 5.2 per cent more than double the 2.1 per cent rate between 1 and 7 years. The Montreal Heart Institute predicted that because of late graft changes, long-term relief of symptoms and survival may be compromised. Our observation of accelerated mortality after seven years in surgically treated pa- tients but not in those receiving medical treatment supports this prediction. In conclusion, we found that in the Veterans Ad- ministration study population bypass surgery did not significantly improve overall survival among patients without left main coronary-artery disease. However, a survival benefit with surgery was observed at five to seven years in subgroups of patients with multiple

OCR for page 255
APPENDIX A: PROFILES clinical and angiographic risk factors. The observed benefit with surgery diminished gradually when fol- low-up was extended to 11 years. REFERENCES 1. 489 7. Takaro T. Hultgren HN, Lipton MJ, Detre KM, et al. The VA Cooperative Randomized Study of Surgery for Coronary Arterial Occlusive Disease. II. Subgroup with significant left main lesions. Circulation 1976; S4: Suppl 3:III-107-17. 2. Murphy ML, Hultgren HN, Detre K, Thomsen J. Talcaro T. et al. Treatment of chronic stable angina: a preliminary report of survival data of the random- ized Veterans Administration cooperative study. N Engl J Med 1977; 297:621-7. 3. Detre K, Peduzzi P. Murphy M, et al. Effect of bypass surgery on sur- vival of patients in low- and high-risk subgroups delineated by the use of simple clinical variables: Veterans Administration Cooperative Study of Surgery for Coronary Arterial Occlusive Disease. Circulation 1981; 63: 1329-38. 4. Takaro T. Peduzzi P. Detre KM, et al. Survival in subgroups of patients with left main coronary artery disease. Circulation 1982; 66:14-22. S. Detre KM, Hultgren HN, Takaro T. et al. Veterans Administration Cooper- ative Study of Surgery for Coronary Arterial Occlusive Disease. III. Meth- ods and baseline characteristics, including experience with medical treat- ment. Am J Cardiol 1977; 40:212-25. 6. Takaro T. Hultgren HN, Detre KM, Peduzzi P. Murphy M. Results of the VA randomized study of medical and surgical management of angina pec- toris. In: Hammermeister KE, ed. Coronary bypass surgery: the late results. New York: Praeger, 1983:19. Detre K, Peduzzi P. The problem of attributing deaths of nonadherers: the VA coronary bypass experience. Controlled Clin Trials 1982; 3:355~4. 8. Parisi AF, Peduzzi P. Detre K, Shugoll G. Hultgren HN, Talcaro T. Character- istics and outcome of medical Is in the Veterans Adminimabon Costive Study of Coronary Artery Surgery. Am J Cardiol 1984; 53:23-8. 9. Talcaro T. Hultgren HN, Detre KM, Peduzzi P. The Veterans Admin~stra- tion Cooperative Study of stable angina: current status. Circulation 1982; 65: Suppl 2:II~7. 10. Peduzzi PN, Detre KM, Chan YK, Oberman A, Cutter GR. Validation of a risk function to predict mortality in a VA population with coronary artery disease. Controlled Clin Trials 1982; 3:47 60. I 1. Braunwald E. Effects of coronary-artery bypass goofing on survival: impli- cations of the randomized coronary-artery surgery study. N Engl J Med 13. 1983; 309:1181-4. 12. Idem. The treatment of coronary artery disease: lessons from clinical trials. Presented at the f~fty-sixth annual meeting of the American Heart Associ- ation, Anaheim, Calif., November 16, 1983. European Coronary Surgery Study Group. Long-tenn results of a prospec- tive randomised study of coronary artery bypass surgery in stable angina pectoris. Lancet 1982; 2:1173-80. 14. CASS Principal Investigators. Coronary Artery Surgery Study (CASS): a randomized trial of coronary artery bypass surgery: survival data. Circula- 15. 16. don 1983; 68:939-50. Enjalbert M, Vaislic C, Ignorance J. Grondin CM, Bourassa MG, Cam- peau L. Relief of angina and survival 12 years after XOrtocoronary saphe- nous vein graft bypass surgery. Circulation 1983; 68: Suppl 3 m-116. ab- stract. Campeau L, Enjalbert M, Vance J. Va~slic C, Grondin CM, Bou~assa MG. Atherosclerosis and late closure of XOrtocoronary saphenous vein grafts: sequential angiog~aphic studies at 2 weeks, 1 year, 5 to 7 years, and 10 to 12 years after surgery. Circulation 1983; 68: Suppl 2:II-1-7. Reprinted from The New England Journal of Medicine 311:1333-1339 (November 22), 1984 ~Copyright, 1984, by the Massachusetts Medical Society Printed in the U.S.A.