Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Sponsorship and Role of Consensus Development Programs within National Health Care Systems Itzhak Jacoby In 1977, under the leadership of Donald Fredrickson, the U.S. National Institutes of Health (NIH) began sponsoring consensus development conferences. This new program was NIH's innovative response to the widely perceived need for assessment of the safety and efficacy of technologies. In the U.S. Congress and elsewhere, debates were raging about health technology assessment in general and the role of the federal government in particular, including the impact of NTH's research effort on clinical practice. Senator Ed- ward Kennedy asked in a 1976 speech: Shouldn't some institution in our society have an ongoing function of reviewing not just new knowledge that might be transferred into clinical practice but also old knowledge that underpins current procedures involving risks, high costs, or simply great inconvenience to millions of patients in order to determine what needs to be changed, updated, or further researched? (Perry, 1988~. Further impetus for developing the program came from rapidly escalating health care costs, which were linked by the public and policymakers to the uncontrolled diffusion of expensive, but not necessarily cost-effective, technologies. The program also fulfilled Dr. Frederickson's desire for NTH to bolster information transfer at the interface between biomedical science and clinical practice. With more than 75 consensus conferences to its credit, the NIH program has served as a mode} for the United States and other coun- tries in the development of health technology assessment. Similar 7
8 CONSENSUS DE:VELOPMENT programs that have emerged in several European countries and Can- ada differ markedly from the NIH model in their goals, scope of consideration, and mode of disseminating results. This divergence has resulted from differences among the countries' health care deliv- ery systems and the nature of the organizations that sponsor the conferences. These differences in the context of technology assess- ment play a critical role in determining the scope, format, and im- pact of each country consensus development program. The consensus development process is designed for conflict reso- lution. It borrows from several other conflict resolution modes: the judicial process, collegiate peer review, democratic debate, and col- lective bargaining. As pointed out by Kosecoff et al. (1987), for the process to have impact and meaning, topics must involve clinical practice-related controversies and a gap between knowledge and practice. The process has its greatest effect in the resolution of controversies about technologies or issues on the cutting edge of medical practice, and therefore, it should be reserved for this pur- pose. Broad participation of multiple factions is necessary in sponsor- ing, planning, and implementing the conferences in order to increase the likelihood of proper selection of topics, encourage development of conclusions applicable to practice decisions and policy-making, and produce a significant impact on health care delivery. Partici- pants should include the biomedical research community, those in medical practice, payer organizations, and organizations concerned with the administration and delivery of services. To meet the needs of these factions, consensus conferences should aim to resolve con- flicts about safety, efficacy, cost and cost-effectiveness, indications and contraindications, societal acceptance, and other ethical issues. Any compromise in this agenda, especially given the urgent need for consensus results and the high cost of these conferences, can only be considered a missed opportunity. The context within which health technology assessment- and consensus development in particular- takes place in each country has determined the extent to which this ideal is realized. THE U.S. EXPERIENCE IN CONSENSUS DEVELOPMENT The focus of the NIH Consensus Development Program is nar- rowly defined, in line with the agency's mission as a biomedical
SPONSORSHIP AND ROLE 9 research institution. The program limits assessment to issues of safety and efficacy, excluding social, economic, and ethical consid- erations. Topics are proposed by one of NIH's many component institutes when sufficient data are available to conduct such an as- sessment and when some element of controversy exists. Participants in the process are broadly recognized leaders in their respective fields. Technical and research experts present information to an appropriate pane} of physician specialists, informed public represen- tatives, and others with specialized expertise. This panel reviews the scientific data presented to develop the consensus statement. Representatives of physician specialty groups, third-party payers, or health administrators are not usually participants in the planning or execution of the conference. From the beginning, the NIH process was supposed to seek a "technical consensus" on the clinical significance of new findings, the adequacy of validation for efficacy and safety, and the need for further research. NIH recognized that further consensus develop- ment on the broader health care delivery issues should complement its technical consensus before a particular clinical modality is rec- ommended for adoption. This second process was then referred to as interface consensus and was to deal not only with safety and efficacy, but also with cost; cost-effectiveness; and legal, ethical, and other societal issues (Perry, 1988~. To provide this broader assessment, the U.S. Congress established the National Center for Health Care Technology (NCHCT). The center's responsibilities included consensus development on the full range of clinical practice-related issues. The one conference that the center sponsored as a follow-up to the NIH consensus conference on coronary artery bypass surgery complemented the NTH conclusions with considerations of broad societal issues. The NCHCT succumbed to budgetary and political pressures in 1981 after only two years of operation. While it existed, the NIH added the word technical to the title of its consensus development program to distinguish NTH's re- sponsibilities from those of the center. Although the word technical has since been dropped, NIH-sponsored consensus conferences still conform to the restricted model. The broad goal of the NIH program is to facilitate the appropriate and timely application of biomedical research findings to clinical practice. In evaluating the extent to which the program has attained that goal, analysis both of conferences that have produced an impact
10 CONSENSUS DEVELOPMENT and those that have not point to the advantages of broad sponsor- ship. For the most part, NIH relies upon topic selection criteria concerning the adequacy of the evidence for resolving controversies and the thoroughness with which the clinical implications of scien- tific or technologic developments have been explored and discussed in the literature. Although these criteria are reasonable in the con- text of scientific peer review, they do not address the needs of the intended recipients of technology transfer: physicians, hospitals, and other providers of health care, as well as their patients. To increase the impact of consensus development conference results, those who choose topics should systematically consider the current state of practice along with the current state of science in determin- ing the potential usefulness of a conference. The NTH consensus development conference on the treatment of primary breast cancer illustrated the importance of considering the state of practice when planning a conference. This conference failed to produce a change in practice because there was widespread pre- conference conformity- with the major consensus recommendation (Kanouse et al., 1989~. An inability to effect change may also result from the panel's avoidance of the areas of greatest controversy in generating recom- mendations. The U.S. conference on coronary artery bypass surgery is a case in point. In the consensus statement from that conference, the pane} confined its recommendations to the relatively clear cases at the extremes, where there was less controversy, instead of provid- ing unequivocal recommendations regarding the most common clini- cal scenarios leading to bypass surgery, where clinicians might wel- come some assistance in making choices (Kanouse et al., 19891. In general, sole custodianship of any process encourages a natural ten- dency for a group to avoid controversy and to safely promote the view of the sponsor, which, in this case, was the biomedical re- search community. Greater involvement of physician specialty so- cieties and payer organizations, including health insurance groups and agencies such as the Health Care Financing Administration that reimburse providers and beneficiaries with government funds, might help to produce the most urgently needed consensus development information. When participation in NTH consensus development conferences extends beyond the biomedical research community, the benefits are clear. A recent evaluation of the NIH Consensus Development Pro- /
SPONSORSHIP AND ROW 11 gram singled out the 1980 conference on cesarean childbirth as a success in both information transfer and practice impact. Since both the obstetrical community and the lay public viewed sharply rising cesarean section rates with alarm, NIH's sponsorship of this confer- ence was spurred by a suggestion from the American College of Obstetricians and Gynecologists. The conference thereby capital- ized on the existing physician awareness and concern that formed a basis for motivating change in hospital policies and physician prac- tices. The conference was oriented to this well-defined interest and performed a unique service in summarizing the accumulated knowI- edge on this topic, pointing out the practice implications of this information, and developing clear recommendations about changing the management of certain deliveries. All of these activities had never before occurred with the clinician as the intended audience. Finally, NIH arranged for publication of the consensus statement in two major obstetrical journals, and the American College of Obste- tr~cians and Gynecologists took an active interest in helping to dis- seminate the findings (Kanouse et al., 19891. Given these elementsa scientifically based and clinically rele- vant message delivered for the first time to a receptive audience that recognized the need for change it is perhaps not surprising that the conference on cesarean sections was especially successful in chang- ing physicians' and hospitals' practices (Kanouse et al., 19891. Such a combination of favorable conditions may be the exception rawer than the rule for consensus development topics. The likelihood of identifying motivation for change in a controversial area may be enhanced by seeking out the active participation of the potential users of consensus statements. An analysis of the topics selected for the NTH Consensus Devel- opment Program over the last eight years would suggest a shift away from controversial subjects accompanied by the production of con- clusions that are not only less controversial but also more general and less helpful to the practicing health care professional commu- nity or patients. Two exceptions, in addition to the one on cesarean sections, are the conferences on liver transplantation and the use of ultrasound in prenatal care. In both cases, outside pressure was brought to bear on NIH to hold the conferences. To achieve the greatest impact with consensus development con- ferences in the United States, NIH should contribute, along with other participants from the broader health care community, to a con-
12 CONSENSUS DEVELOPMENT sensus process administered by an organization with a much broader mission. Topics should be selected that represent a significant chal- lenge from the perspective of clinicians and the public at large. The influence, resources, and insight of the biomedical research commu- nity, combined with those of physician specialty societies, third- party payers, and individuals involved in health care policy and administration, would constitute a formidable force for change in health care. One additional change in the current procedure should also be considered. Occasionally, a consensus development pane} has pro- duced a statement with assertions that are not supported by evidence or a statement that does not adequately answer one or more of the questions posed. It should be routine for all consensus statements to be reviewed by an independent group, as are papers in peer-reviewed journals and reports from the National Academy of Sciences. This peer review should result in the production of more valid statements. CONSENSUS DEVELOPMENT OUTSIDE THE UNITED STATES As the NIH Consensus Development Program in the United States matured, international interest in the process began to emerge. In late 1981, Sweden's consensus development program was established. The program was the result of numerous discussions and visits by representatives of the Swedish Medical Research Council, a sister agency of NIH, and the Swedish Planning and Rationalization Insti- tute for the Health and Social Services (Spri), which is responsible for health planning in Sweden. Although a great deal of discussion has occurred concerning the similarities between the U.S. and Swed- ish programs, the cosponsorship of the Swedish program by a health planning organization has received little attention. Spri's orientation has had a distinctive impact on the consensus development program's activities. This arrangement was particularly important considering the decentralization of the national health service delivery system in Sweden, where most decisions are made by regional councils. The first Swedish consensus development conference addressed total hipjoint replacement. It followed a conference on the same topic in the United States. The Swedish conference, in line with the sponsors' interests, added a question on the cost implications of the procedure to questions shared with the U.S. conference. Other top-
SPONSORSHIP AND ROLE 13 ice have been selected primarily on the basis of their importance to health service delivery and the practice of medicine, in addition to the willingness of the scientific community to deal with the issue. An examination of the impact of the Swedish conference on hip- joint replacement indicated an increase in the number of procedures immediately following the conference (Calitorp, 1988~. In the United Kingdom, the consensus development program be- gan in 1984. Unlike other programs, this endeavor stemmed not from legislative or governmental impetus but from individual inter- est and belief in the process. The program found a base in The King's Fund Centre for Service Development, an academic and phil- anthropic institution prominent in the British medical community for training health service managers. The organizers of the consensus development program nevertheless sought the participation of gov- ernment agencies responsible for the National Health Service and participation of physicians' groups. These organizations, however, while supporting the purpose of the program, chose to participate as observers and adopt a "wait and see" attitude. After lengthy negotiations, the first topic chosen was coronary artery bypass grafting. At the time, only slightly more than 10 procedures per 100,000 population were performed in the United Kingdom, as opposed to more than 80 procedures per 100,000 popu- ilul1 ;11 flu; Ullii`;U attires and approximately 4U per 100,000 in other European countries. Since most health services utilization experts agreed that the European rate was appropriate, it was not surprising that the consensus process recommended raising the rate for the United Kingdom. In recognition of Britain's limited health care resources, the consensus statement provided criteria for clini- cians to use in the selection of patients who could benefit most from the medical process for receipt of treatment. Compared with the U.S. conference on this topic in 1981, the King's Fund conference produced very detailed clinical advice as to indications and contrain- dications for the coronary artery bypass procedure. In addition, as discussed in a detailed comparison of the U.K. and U.S. conferences on bypass surgery (Stocking, 1985), the questions considered by the panel went beyond scientific issues to encompass the costs and implications of increased use of coronary artery bypass grafting on the British National Health Service. In line with these interests, and especially because of the capped budget of the service, the U.K. pane} consisted of people who were not experts on the 1~_ -_ I_ IT_: Cat ~ _ _ · . ~ ~~
4 CONSENSUS DEVELOPMENT procedure; half of the panel members were not clinicians. Subse- quent conferences in the United Kingdom have dealt both with nar- row issues in medicine as well as broad, societal issues. Increased emphasis has been placed on the impact on the health care delivery system of adopting new technologies, as well as social, economic, and cultural implications, in an effort to make the consensus state- ments as useful as possible in the public policy arena. This broad scope was feasible because the sponsoring institution was an aca- demic center rather than an active participant in the health delivery system. In The Netherlands, a technology assessment program has been initiated by the National Organization for Quality Assurance in Hospitals (Centraal Begeleidingsorgaan voor de Intercollegiate Toets- ing), a physician organization. The program began as a physician- initiated quality assurance program, similar to the U.S. professional standards review organizations and, later, the professional review organizations. Only after its emergence was it recognized as an ana- log to the consensus development program. This program is signif~- cantly different from the previous two examples. Each assessment is a long process, involving many internal review cycles. The main feature that should be highlighted is that, since the consensus devel- opment process is sponsored by a body of physicians, there is little opportunity for contributions by other participants in health care delivery. Because it is sponsored by physicians, the program has a high potential for making a significant impact on practice behavior. In fact, the conclusions reached by this consensus development pro- gram serve as the basis for the quality assessment program in place in Dutch hospitals (Klazinga et al., 1987~. In France, the Institut National de la Sante et de la Recherche Medicale (INSEAM), a sister organization of NIH, attempted to es- tablish a consensus development program in 1987. The effort was unsuccessful because of the research organization's inability to ob- tain the active participation of the country's social security admin- istration and physician organizations in sponsoring consensus con- ferences. This experience emphasizes the importance, in some quar- ters, of bringing the major participants in health services delivery into the assessment process in order to ensure the usefulness of results. The Israel Ministry of Health has sponsored jointly with the World Health Organization (WHO) and the U.S. NIH a consensus confer-
SPONSORSHIP AND ROLE 1 r ence on magnetic resonance imaging and positron emission tomogra- phy (PET). Israel's need for this conference developed from a con- troversy that arose between the philanthropic organization Hadas- sah, the Ministry of Health, and the Sick Fund, which provides care to the majority of Israelis, over the uses of the two technologies. Lacking the expertise to resolve the issue, Israel called upon WHO, which in turn called upon NIH. This broad sponsorship resulted in the acquisition of the needed expertise. The conference concluded that, although PET has great scientific importance, it was clearly not ready for clinical applications. As is characteristic of a small coun- try with a pluralistic health care system, the Israeli conference cov- ered practical aspects of economic and human resources in addition to assessment of the science. The conclusions drawn at the confer- ence had a direct impact on the provision of imaging services in Israel. Interestingly, the United States had begun to plan a conference on PET' but when it became clear to the planners that the recommenda- tion was going to be negative for clinical applications, the confer- ence was canceled. In April 1989, the Wall Street Journal carried an article entitled "Debate Grows Over Clinics Pushing Costly PET Scans," which examined the movement of PET into clinical settings with questionable and costly applications (Wai] Street Journal, 1989~. Israel's definitive action has enabled that country to deal effectively with PET before it emerged as a problem, as now appears to be the case in the United States. CONCLUSIONS This paper described five models for consensus development sponsorship and examined them in the context of the health service delivery systems of the respective countries. Other consensus devel- opment programs worldwide can be identified as belonging to one of these five models (see Table I). It is critical for program sponsors to consciously consider the mode} they intend to apply in light of their desired objectives. The goal of consensus development programs should be to re- solve conflicts in as broad a range of factors as is relevant, including safety, efficacy, cost and cost-effectiveness, indications and contra- indications, acceptability, and other societal and ethical issues. The likelihood of realizing this goal will be improved by expanding the
16 CONSENSUS DEVELOPMENT TABLE ~ Sponsorship and Scope of Consensus Conferences in Five Countries Country Sponsors Scope United States National Institutes of Health Safety and efficacy The Netherlands National Organization for Quality Quality assurance Assurance in Hospitals Sweden Swedish Medical Research Council Safety, efficacy, and and Swedish Planning and cost-effectiveness Rationalization Institute for the Health and Social Services United Kingdom The King's Fund College Broad societal Israel All parmers in national health, plus Broad consensus the World Health Organization and for application the U.S. National Institutes of Health sponsorship of the process to include all major participants in the biomedical research and health care communities. These sponsors should make every effort to select topics that represent a significant challenge from the perspective of clinicians and the public at large. In planning conferences, sponsors should consider both the current state of science and the current state of clinical practice to assure that an impact on clinical practice is possible. In particular, prefer- ence should be given to topics for which motivation to change pol- icy and clinical practice already exists among clinicians. Finally, consensus development programs should adopt a procedure to re- view consensus statements before they are published in final form. By marshaling the resources of the research, practice, health fi- nancing, and health policy communities, the consensus development process can contribute significantly to any nation's health agenda. REFERENCES Calltorp, J. 1988. Consensus development conferences in Sweden: Effects on health policy and administration. International Journal of Technology Assess- ment in Health Care 4:75-88. Kanouse, D.E., R.H. Brook, J.D. Winkler, J. Kosecoff, S.H. Berry, G.M. Carter, J.P. Kahan, L. McCloskey, W.H. Rogers, C.M. Winslow, G.M. Anderson, L. Brodsley, A. Fink, and L. Meredith. 1989. Changing medical practice through technology assessment: An evaluation of the NIH consensus development program. Santa Monica, Calif.: RAND Corporation.
SPONSORSHIP AND ROLE 17 Klazinga, N.S., A.F. Casparie, and J.J.E. van Everdingen. 1987. Contribution of medical decision-making to consensus development conferences. Health Policy 8:339-346. Kosecoff, J., D.E. Kanouse, W.H. Rogers, L. McCloskey, C.M. Winslow, and R.H. Brook. 1987. Effects of the National Institutes of Health consensus development program on physician practice. Journal of the American Medical Association 258(19):2708-2713. Perry, S. 1988. Consensus development: An historical note. International Journal of Technology Assessment in Health Care 4:481-484. Stocking, B. 1985. First consensus development conference in the United Kingdom: On coronary artery bypass grafting, views of audience, panel, and speakers. Brit- ish Medical Journal 291:713-716. Wall Street Journal. April 3, 1989. Debate grows over clinics pushing costly PET scans. B1.
