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.
1 Introduction The need to set national priorities for the assessment of medical prac- tice is becoming undeniable. The increasing interest of both the private and public sectors in evaluating the effectiveness of medical practice is coupled with finite economic and human resources for this purpose. A national prionty-setting process is needed to identify assessment areas in which investments in medical practice evaluation would make the most impact. Pr~onties must be set among He clinical conditions Cat affect Be U.S. population, and among the medical technologies and practices used to prevent, diagnose, and treat those conditions. The goal of this project is to demonstrate a pilot approach for setting national priorities for the as- sessment of clinical conditions and medical technologies. A growing national consensus among patients, provided, payers, health care administrators, manufacturers, and legislators to find out "what works in the practice of medicine" (Roper et al., 1988) characterizes this "era of assessment and accountability" in health care (Relman, 1988~. Patients are better informed consumers of heath care and are, appropriately, taking increased responsibility for decisions about their health. Heath care providers strive to integrate advances in medical technology in order to offer the most effective diagnostic and therapeutic options to Heir ~ In this report the term electiveness is used as it is formally defined in the assessment literature (see the Glossary). However, Me term has been used to refer to "what works in the practice of medicine" and broadened to include the efficacy, effectiveness, and appro- priateness of health services and their relationship to health care costs (Roper et al., 1988~. 9
10 NATIONAL ASSESSMENT PRIORITIES patients. As Me demand for health care sentences continues to increase, reports of variations and possible overudlization are frequently cited (Wennberg, 1984; Chassin et al., 1986; Roos et al., 1988~. As health care costs continue to risefimm 6.0 percent of the U.S. gross national product in 1966, to 8.6 percent in 1976, and to 10.9 percent in 1986 Rational Center for Health Statistics, 1989b~both private and public payers of health care become Increasingly interested In the cost-effec~veness of medical practices and technologies. The Department of Health and Human Services, responding to this growing need for the assessment of medical practice, embarked on a national Medical Treatment Effectiveness Program in fiscal year 1990. This national program incorporated two preceding federal programs, the Patient Outcome Assessment Research Program of He National Center for Health Services Research2 and the Effectiveness Initiative of He Health Care Financing Administration (HCFA)3 (Clinton, 19891. These activities have been further expanded and reorganized under a new Agency for Health Care Policy and Research (U.S. Congress, House, 1989~. These recent expansions in federal research programs and resources have focused national attention on medical effectiveness and patient outcomes research The broad domain of"patient outcomes'$4 and "effec- tiveness research' builds upon He work of related fields of He clinical 2 Ckigmating m 1986 legislation, the Patient Outcome Assessment Research Program was expended in 1989 to sponsor four assessment teams that will investigate the sources of emanation m medical care that result m adverse outcomes or inappropriate resource utiliza- tion. Lee ultunate goal of the research activities is to feed back the infonnation about outcomes and the costs of alternative practice patterns to practicing physicians (National Cent" for Health Services Research. 1988~. 3 ~ 1988 the Health Care Financing Administration proposed an Effectiveness Initiative that would include monitonug the outcomes associated with alternative treatments using Urge data bases, icing variations in the outcomes and patterns of care, assessing attentive interventions to reduce variations, and feeding this infonnation back to provid- e" and patients to educate them about the results of these effectiveness studies (Roper et aL, 1988; Instate of Medicme, 1989). 4 The focus of patient outcomes research is on the evaluation of medical interventions hom the patia~t's perspective to consider, systematically, all the outcomes Mat maw to patients, such as mortality, morbidity, complications of treatments, induction of symptoms, improvement of fimcdonal status, and quality of life (Wennberg, 1990~. ~ Effectivaless research ~dwsses He evaluation of patient care delivered in the every- day practice of medians. This is in contrast to studying He efficacy of medical technolo- gies and practices under controlled experimental conditions such as randomized clinical As.
I~RODUCrlON 11 evaluative sciences, such as clinical research, epidemiology, health serv- ices research, health status and functional assessment, medical decision- analysis, medical technology assessment, and quality assessment. As activities In the assessment of medical practice continue to develop at the national level, Be emphasis on patient outcomes and effectiveness re- search is bringing together the complementary perspectives and me~od- olog~es of these fields. In this report, the Institute of Medicine and its Council on Health Care Technology demonstrate a pilot approach to setting national priorities for the evaluation of medical practice that addresses the diverse assessment interests that arise In a changing research, health care delivery, and policy environment. This approach to setting national pnor~ties is based on the broad-based participation of diverse parties and on a conceptual frame- wow that integrates their perspectives. Although individual patients, pro- viders, payers, and others continue to set priorities based on their respec- tive needs and resources, a national agenda may stimulate evaluations of broader interest or capture assessment topics of national importance Cat are not being addressed. The pnonty-seuing group that chose the 20 national assessment priority areas identified in this pilot study represents major health care constituencies, including academia, govemment, health care providers, industry, and third-parLy payers. In addition, the concep- tual framework for sewing priorities was developed after considenng background information provided by 14 major heath care organizations which span He field of health care delivery (see Appendix C). During this pilot project, He council reviewed previous priority-sethng approaches and identified two major types of pnonty-setting methods consensus and modeling. The consensus approach relies on the synthesis of the expert opinion of a group of individuals. Consensus processes for setting priorities may vary with respect to He formality of their approach to identifying potential assessment topics, their reliance upon quantitative information, and their use of implicit as well as explicit criteria to rank me topics. These consensus approaches reflect the judgment of the individu- als who form the prionty-setting group. The priorities identified by this approach are influenced by the individuals' areas of expertise and their role within broader hemp care systems. In spite of these limitadons, consensus processes that involve wet/-balanced committees of experts may be He only way to address problems when primary data are not available or readily accessible. This consensus approach was used, for instance, in 1988 by the Institute of Medicine when it appointed a com- mittee of clinicians to provide advice to HCFA on initial pr~onties for He
12 NATIONAL ASSESSMENT PRIORITIES agency's Effectiveness Initiative. The priorities identified in that study were limited to clinical conditions affecting the Medicare population.6 Modeling approaches involving quantitative methods may be used to identify priority assessments or to generate estimates of their expected value. To explore the feasibility of this approach, the council commis- sioned two investigators to develop priority-setting models. Charles Phelps of the University of Rochester developed an econometric model, using a large hospice discharge data base, that estimates the annual wel- fare loss nationally fin dollars) that may result from variations in medical practice associated with the use of inpatient procedures for some cI=cal conditions (Phelps and Parente, 1989~. The pnority-setting mode] of David Eddy of Duke University uses evidence or, if necessary, expert opinion to generate quantitative estimates of the expected value of an assessment according to a defined set of health and economic outcomes fly, in press). These modeling approaches add rigor to the pnonty- setting process by defining objective cr~tena that facilitate quantitative ranking of potential assessment topics. However, quantitative models are necessanly limited by the degree to which assumptions are explicit and reasonable, the availability and quality of data, and the difficulty of accounting for important, but difficult to quantify, environmental and other factors. An ideal pnonty-setting methodology would account explicitly and quantitatively for aU factors significantly influencing choices among po- tential assessment topics. Such a pnonty-seuing process would be sub- mitted to periodic review and modification so that developments In emerg- ing technologies, the management of clinical conditions, or other factors that affect decision-makers' perspectives may be taken into account. The complexity of pnority-setting decision-making processes makes this ideal difficult to attain. The judgment of decision-makers with different roles in health care delivery may differ concerning which topics are of greater pnonty, what standards or criteria to use for selecting them, and what reladve importance to give these cntena. The pnma~y data for making these decisions are often not available or readily accessible. As a result, pnonty-se~ng decisions may involve implicit perceptions and judgments 6 The report of the Institute of Medicine Committee on the Effectiveness Initiative of He Heals Care Financing Adrninistraiion was published as Effectiveness Initiative: Setting Priorities for Clinical Conditions in April 1989 (Institute of Medicine, 1989). Although the recommendations of that committee were among die sources for the pilot review of priority-setting approaches for this report (see Appendix C), the process of determining the national assessment priorities for this report was undertaken separately.
I~RODUCHON 13 that may dominate these decisions (Weinstein and Statson, 1977~. To the extent that such approaches for setting assessment priorities are Convex by intuition or chance, they may not lead to the optimal use of limited resources. Recognizing the difficulties inherent in establishing national medical assessment pnonties, the council can provide a national focus for pnonty- seeting through its ability to convene experts from Me pertinent health care sectors. The pilot study and proposal for future efforts described in this report may evolve into national pnonty-seuing mechanisms that are better and more efficient than the current ones. By combining consensus and group judgment methods with a model- ing approach, the national pnority-setting methodology proposed here takes into account both objective and subjective factors Mat may be important in pnority-setting decisions. The nationally applicable criteria defined herein provide explicit standards by which different pnondes may be ranked. During the pilot effort, the pnonty-sefflng group of the council used the criteria outlined in Chapter 2 to choose the 20 high- pnonty clinical condition and technology assessment areas described in Chapter 3. In selecting this preliminary set of priorities from a larger list of candidates (see Appendix A), the group considered seconder, data pro- vided by a selection of health care organizations (see Appendix C). The two-round modified Delphi process used to choose the pnondes is dis- cussed in more detail In Chapter 2. For a foBow-up effort of this pilot project, the council would like to develop and apply a decision-analysis model to estimate, In a more quantitative way, the potential impact of candidate assessment topics by all Me criteria outlined. This mode} would draw on state-of-the-art information about Me potential assessment topics and on a fuller review of the preferences that representatives of different sectors, including patients and the public, may have for certain pnorities. The national pnonty-sefflng approach descnbed here, therefore, is distinct from previous approaches in two import ways. First, the approach attempts to recognize the assessment needs of venous pardes interested in the evaluation of medical practice and consolidate them unto a set of criteria that will be nationally applicable to many patient popula- tions. Second, Me conceptual framework used here accommodates Me different manners in which assessment needs may arise. For instance, a comparative evaluation of the available altemative medical technologies or practices may be needed to determine the best or most appropn ate approach for managing a clinical condition such as myocardial infarction or hip fracture. On Me other hand, Me in~ducdon of a promising new
14 NATIONAL A-SSE5SME2VTPRIORmES technology, such as a genetic ally en~nee~d clot-l ysing agent, a compu - tenzed radiolo~c system, a fonn of laser surgery, or a home diagnostic test kit raises questions about safer, cost, or over effects of an individual technology. Thus, the national priority-setting approach proposed here uses bow a clerical condition approach and a technology approach In considering national assessment pnonties. The set of 20 priority areas listed in Chapter 3 provides a starting point for the evaluation of medical practice. The hnstitute of Medicine and He Council on Health Care Technology look forward to continued coopera- tion with rep~senmives of pnv ate and public sector interests, in efforts ~ furler define He specific questions to be addressed about these general priority areas and to improve the methodology for sewing pnondes. This cooperation would help ~ ensure that this and filture sets of priority topics can be established and addressed at He national level.
