12
Conclusions and Recommendations

The charge to this committee, and the basis for the workshop reported in the previous sections, entails the following questions:

  1. Does the model of managed care or integrated delivery system (e.g., fully capitated managed care, gatekeeper-only model of managed care, discounted fee for service) influence (a) the types of interventions provided to patients with chronic conditions such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), and (b) the clinical and health status outcomes of those interventions?
  2. If so, are these effects quantitatively and clinically significant, compared to the effects that other variables (such as income, education, or ethnicity) have on patient outcomes?
  3. If the mode of health care delivery system appears to be related to patient care and outcomes, can specific organizational, financial, or other variables be identified to account for the relationships?
  4. If not, what research agenda should be pursued to provide critical information about the relationship between types of health care systems and the processes and outcomes of care delivered to populations with serious chronic conditions?

As noted in the introduction to this report, the published evidence for differences in treatment received or outcomes achieved by RA and SLE patients in various health care delivery systems is practically nonexistent. The committee was unable to locate any such studies involving SLE. The most extensive comparison of delivery models and RA is that of Yelin and his



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 213
--> 12 Conclusions and Recommendations The charge to this committee, and the basis for the workshop reported in the previous sections, entails the following questions: Does the model of managed care or integrated delivery system (e.g., fully capitated managed care, gatekeeper-only model of managed care, discounted fee for service) influence (a) the types of interventions provided to patients with chronic conditions such as rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE), and (b) the clinical and health status outcomes of those interventions? If so, are these effects quantitatively and clinically significant, compared to the effects that other variables (such as income, education, or ethnicity) have on patient outcomes? If the mode of health care delivery system appears to be related to patient care and outcomes, can specific organizational, financial, or other variables be identified to account for the relationships? If not, what research agenda should be pursued to provide critical information about the relationship between types of health care systems and the processes and outcomes of care delivered to populations with serious chronic conditions? As noted in the introduction to this report, the published evidence for differences in treatment received or outcomes achieved by RA and SLE patients in various health care delivery systems is practically nonexistent. The committee was unable to locate any such studies involving SLE. The most extensive comparison of delivery models and RA is that of Yelin and his

OCR for page 213
--> colleagues, which he summarized for the committee at the workshop. They were unable to demonstrate significant differences between RA patients treated in fee-for-service (FFS) relationships and those treated in prepaid group practice health maintenance organizations (HMOs). All patients received some care from a specialist (rheumatologist). This finding is consistent with a larger body of data from studies of chronic diseases other than RA and SLE, although the data presented by Tarlov suggest that this may not hold true for certain vulnerable subsets of the chronically ill. The committee notes that lack of evidence for an effect does not constitute evidence for the absence of an effect, but the essentially negative answer to the first of the four questions in the charge would seem to rule out meaningful responses to questions 2 and 3, since they presuppose a positive answer to question 1. The rest of this chapter is therefore organized as a series of conclusions that the committee feels are justified by the current state of knowledge, and as recommendations for research that flow from these conclusions. The reader is referred to relevant portions of the workshop proceedings as appropriate. However, the committee was selected to bring a wide spectrum of experience and knowledge to bear on the task of answering these questions, and its conclusions and recommendations reflect the committee's own expertise as well as the data and opinions provided to it by the workshop's invited speakers and other participants. The committee alone bears full responsibility for the conclusions and recommendations. Conclusion 1 Rheumatoid arthritis and systemic lupus erythematosus are very likely to be representative of a large number of chronic diseases that will increasingly come to dominate U.S health care. The study of systems of care for chronic diseases such as RA and SLE demands integrated research and longitudinal studies, despite the inherent limitations imposed by changing health care practices, physician and patient characteristics, population demographics, and limited funds. Recommendation: The committee recommends that the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) seek additional funding to expand its research to include substantial support for high-quality studies that would allow a broad approach more closely linking scientific and technological advances to clinical trials, outcomes research, and health services research more generally. Although some of the required research is of the sort currently funded by the Agency for Health Care Policy and Research (AHCPR), there is much to be said for

OCR for page 213
--> bringing fundamental experimental research, clinical trials, and services or outcomes research on rheumatic disease under one roof, so that the benefits of a more coherent and unified approach might be realized. For example, studies might look at the way new technology is assessed and implemented in various types of managed care organizations and the rate of penetration and prevalence of application of medically accepted new technology in various types of managed care organizations. Conclusion 2 Differences in the incidence and severity of chronic diseases such as RA and SLE among individuals are more likely to be due to patient factors such as genetics, age, education, and socioeconomic status than to medical care organization and financing, but outcomes are likely to depend on all of these factors. The committee's review of the literature and the workshop presentation by Davis and Schoen provided convincing evidence that income, age, education, and ethnicity are strongly related to morbidity and mortality from chronic disease in general and rheumatoid arthritis and systemic lupus in particular. The populations of individuals with RA or SLE are far from homogeneous. Just as managed care is not a useful term because it encompasses such a wide variety of care delivery variations, people with RA and SLE vary widely. Disease severity, comorbid conditions, ethnicity, income, age, and education have already been identified as variables that are strongly associated with the course of these diseases and the response to some treatments. It is not unreasonable to suppose that other socioeconomic variables will be discovered to be relevant, and it is almost certain to be the case that each of these variables will render some interventions highly desirable in certain cases and inappropriate in others. Recommendation: Studies of clinical interventions, health care delivery systems, and clinical course and outcomes should examine clinical, demographic, ethnic, and other subsets of patients with RA and SLE. Conclusion 3 Differences in delivery systems may well impact patient satisfaction and the types and intensity of interventions provided to RA and SLE patients, but to date there has been no clear and compelling demonstration of differential impact on the outcome or course of these diseases. As several speakers

OCR for page 213
--> pointed out, the continuing evolution of health care delivery has already made simplistic comparisons such as managed care versus fee for service pointless or even misleading. Potentially relevant variables discussed at length at the workshop and often confounded in the existing literature assessing managed care are the method of payment (capitation versus fee for service, for both patient-to-plan and plan-to-provider payments), specialty of the provider (generalist versus subspecialist), and health and socioeconomic background of the enrolled population. Yelin, for example, pointed out that although his study of RA patients controlled for specialty type by drawing all of the subjects from the practices of California rheumatologists, this very consistency may well be the reason the study failed to reveal differences in treatment of patients in FFS and fully capitated HMOs. Recommendation: Method of payment, medical specialty of the provider, and initial health and socioeconomic background of the patient subjects should be carefully controlled in future studies or, preferably, studied in their own right, and measures of health status and function should be included in addition to simply noting the interventions provided. Conclusion 4 In addition to the pharmacologic and surgical interventions which are important elements of care for RA and SLE patients, several clinical and social interventions that are especially important for patient satisfaction and compliance, particularly strategies emphasizing the role of the patient in managing chronic diseases (i.e., self-management) were identified by our speakers. Holman's address and the associated commentary by Lappin draw together a variety of excellent management practices (subscriber and patient education programs, remote monitoring and support via telephone or visits, and aerobic exercise and strength conditioning programs, for example). Recommendations: The incorporation of these clinical and social interventions into different health care delivery systems should be another area of research for NIAMS. In fact, the availability of such interventions would have to be considered in any evaluation of quality of care. Other research should focus on how these interventions might be improved and extended. For example, how can poorer, less educated patients and patients from different ethnic or cultural backgrounds be persuaded to take a more active role in the management of their rheumatic disease? Would ''patient report cards'' or more explicit planned feedback from

OCR for page 213
--> provider to patient make RA or SLE patients more effective team members in management of these diseases? Conclusion 5 The elderly, the poor, and the chronically ill may well be differentially and adversely affected by enrollment in prepaid health care plans or plans with strong incentives for providers to limit services. Some of the data presented by Davis and Schoen and the data presented by Tarlov point to a significant interaction between the delivery system and age, income, and health status. Davis and Schoen, for example, reported that respondents to a Commonwealth Fund managed care survey who rated their health as fair or poor or who reported that they or a family member had a serious illness in the past year, rated their plans quite differently depending on the type of plan providing their care. Figure 2 in the paper by Davis and Schoen, for example, shows that patients in managed care systems of all types were more likely to rate their plan fair or poor in providing access to specialists than were unmanaged fee-for-service patients. Tarlov's longitudinal data from the Medical Outcomes Study highlighted precisely this type of interaction among patients with chronic conditions (some patients had a rheumatic disease, but the study entry criteria specified diabetes, hypertension, congestive heart failure, postacute myocardial infarction, or clinical depression). Although a comparison of FFS and HMO coverage using the study's entire sample revealed no differences in self-reported health over the course of the study, comparisons using subsamples of elderly, poor, or initially sicker subjects all yielded significant differences favoring fee-for-service systems. Tarlov suggested that this pattern in the data may be attributable to the relative cost of care for these groups. It may be that excessive medical costs have largely been incurred by a relatively healthy segment of the general population whereas expenditures for the aged, the poor, racial minorities, and those with chronic diseases have been more reflective of actual medical needs. Under such conditions, a general reduction of expenditures for health care would differentially impact these vulnerable subgroups even if care for the relatively healthy majority was not seriously compromised. Recommendation: Future research should examine interactions of patient factors and system factors, and their effects on costs, clinical course, and outcomes rather than attempt straightforward univariate comparisons of the sort suggested by question 2 above.

OCR for page 213
--> Conclusion 6 It is generally believed that as many as 40 million Americans may be without health insurance of any kind. The charge to the committee assumes access to a health plan or health insurance by all persons with RA or SLE, but such access may be the most important determinant of care and outcome for all but the most affluent. Recommendation: Community-based samples that include nonmembers of health plans should be included in longitudinal studies of differential disease course and functional status of RA and SLE patients. Conclusion 7 The potential impact of "carve-outs" on persons with RA or SLE is not yet clear. This term refers to the practice of some health plans of treating a specific group of patients or services with special policies not applicable to the majority of the plan's enrollees. Mental health services, for example, may be provided at a different location, by a different group of providers, and/or at a different premium, fee, or capitation rate. The committee is concerned about access to plans, as well as services for enrolled patients once in a plan, and a potential advantage of the carve-out strategy is that it can remove or diminish the incentive for prepaid plans and fully capitated practices to avoid or discourage enrollment by individuals likely to require far more services than the average enrollee. Plans could, for example, carve enrollees with preexisting RA out of their standard agreement, assign a rheumatologist as their primary care provider, charge a premium more in line with their expected outlay, and provide the rheumatologist with a higher than average capitation payment. Recommendation: NIAMS should encourage research investigating the possibility of increased access to health care plans by persons with RA and SLE where plans opt for, and states allow, such carve-outs. Conclusion 8 Checklists and health plan report cards developed for purposes of accreditation or consumer education are useful, but still primitive developments and by their nature are unlikely to provide answers to the questions posed to

OCR for page 213
--> the committee. Quality-of-care measures need further development if delivery models (health care plans) are to be usefully evaluated. The National Committee for Quality Assurance (NCQA) has emerged as the primary source of accreditation for health care plans, in a rough parallel with the Joint Commission on Accreditation of Healthcare Organizations (JCAHO) and hospital accreditation. The NCQA evaluation relies heavily on the Health Plan Employer Data and Information Set (HEDIS), which includes specific quality measures (e.g., immunization rates) as well as data on enrollee access to care, utilization of specific services, and satisfaction, along with financial indicators. The aim of this effort is to provide a standard set of data that will be useful to consumers (especially employers) faced with choosing a health plan. The implicit assumption that potential enrollees are most likely relatively healthy, full-time employees dictates that the measures be generic and/or focused on commonly used services. The current version of HEDIS, for example, has no data to help a person with RA or SLE find a health plan that is especially well equipped for managing these diseases. Future versions of HEDIS may contain an element addressing RA or even SLE, but given the intent of the evaluation it will certainly not contain more than one or two relevant items. Thus, it appears unlikely that report cards on health care plans will provide the type and quantity of data on specific chronic diseases that would allow valid comparisons of individual plans or plan types. Recommendations: Reliable answers to the sort of generic questions posed by NIAMS will continue to require specific research projects using multivariate disease-specific measures of the sort listed by Matthew Liang in his reaction to Alvin Tarlov's presentation (the Arthritis Impact Measurement Scales, the Functional Status Index, and the SLE-Disease Activity Index, for example). The committee's pessimism about the utility of accreditation data for research purposes does not extend to the very large databases maintained by many health plans. With their defined populations and standardized treatment regimens, managed care plans represent a valuable research resource largely untapped to date. Privacy concerns must be addressed and resolved, but the committee recommends that NIAMS explore means of providing qualified researchers access to some of these databases as well as those associated with the Medicare and Medicaid experiments in managed care, with primary data from randomly selected member-patients to confirm the validity of the secondary data and fill in gaps in knowledge.

OCR for page 213
--> Conclusion 9 The question of the optimal utilization of subspecialists is still unresolved. A considerable portion of workshop discussion concerned the role of the rheumatologist in the management of RA and SLE, the stimulus being the increasingly common use of a generalist "gatekeeper" by managed care plans to control access to subspecialists. Rheumatologists were the focus of the discussion at the workshop, but it was recognized by all participants that the arguments are by no means unique to rheumatology or even chronic disease. The committee was divided over the utility of further studies comparing treatment and outcomes of RA or SLE patients who have rheumatologists as their sole or primary provider with patients managed primarily by nonrheumatologists. On the one hand it was pointed out that there is already some evidence, including the work reported by Yelin at the workshop, that RA patients may fare somewhat better under rheumatologists, and that the literature as a whole probably supports the conclusion that, all other things being equal, patients of subspecialists generally fare better. The counterargument was that the general literature seems to contain almost as many studies showing no advantage for patients of subspecialists as studies showing an advantage, and perhaps more importantly, in practice the question is not either/or, but when and how a rheumatologist should be involved. Recommendations: A starting point for research might be an analysis of how specialty referral clinics and affiliated ancillary care providers achieve better outcomes when they do. Is this due to better diagnosis, more appropriate or more timely interventions, or better education and empowerment of patients and their families (to name only a few possibilities)? The committee agreed that some research on the nature of the referral process is certainly merited. Also deserving consideration would be studies investigating the cost-effectiveness of increased clinical training in chronic disease management for general internal medicine, family medicine, pediatrics, and geriatric specialists vis-a-vis continued investment in subspecialist training. A Final Note Managed care is a powerful and still growing element of U.S. health care, although it is a heterogeneous movement the final form or forms of which are still evolving. Chronic disease is responsible for a large and growing proportion of health care utilization in the United States today, but those suffering from

OCR for page 213
--> these diseases are also highly heterogeneous. The growth of both managed care and chronic disease have cast work force issues into bold relief, demanding reanalysis of the optimal roles of generalists and subspecialists. The interaction of managed care and chronic disease is a complex nexus that requires new research paradigms, which should be as integrative as possible.

OCR for page 213
This page in the original is blank.