5
Managed Care and Rheumatoid Arthritis: Utilization and Outcomes Over 11 Years



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 73
--> 5 Managed Care and Rheumatoid Arthritis: Utilization and Outcomes Over 11 Years

OCR for page 73
This page in the original is blank.

OCR for page 73
--> Introduction Bevra Hahn This portion of the discussion focuses on data from Dr. Edward Yelin and colleagues comparing the outcomes of individuals with rheumatoid arthritis (RA) who received care (1) in a fee-for-service system compared to a prepaid group HMO (predominantly Kaiser Permanente), and (2) from nonrheumatologist physicians compared to rheumatologists. It is important to note that the approximately 1,200 individuals enrolled in the study were identified by rheurnatologists. Thus, all the persons with RA had at least one visit with a specialist. Outcomes might be expected to differ from those of individuals who never received advice or care from a specialist. Results of these studies may be summarized as follows: With regard to outcomes of individuals whose care providers at some time were rheumatologists, there were no significant differences in outcomes whether care was delivered in a fee-for-service or an HMO setting. With regard to the impact of rheumatologists as the main physician care giver, persons with a rheumatologist were significantly more likely to receive treatment with a disease modifying arthritis drug (DMARD), including methotrexate; furthermore their functional outcomes (measured by the Health Assessment Questionnaire [HAQ]) and their perception of improvement were significantly better than those of patients having a nonrheumatologist as their main physician care giver. Additional measures of improvement favored the rheumatologist, although they did not reach statistical significance. These included number of swollen joints, number of painful joints, and duration of morning stiffness. Surgical interventions and use of nonsteroidal anti-inflammatory drugs (NSAIDs) were nearly identical in the two groups.

OCR for page 73
--> These data indicate that a person with RA may have a better outcome if she or he can interact with a physician with experience and knowledge relevant to rheumatic diseases—in this case, a rheumatologist. The health care structure by which this access is provided probably has less impact on outcome than does the ability of the care giver. Such a statement presumes that the interventions recommended by the expert are available to the consumer/ patient/client with the chronic disease. This conclusion is further supported by the preliminary survey conducted by MacLean and colleagues,78 which showed that among some 12,000 persons with RA in a large health care database (1.7 million covered lives per year), the majority did not receive regular therapy with anti-inflammatory drugs or DMARDS, and only a minority ever consulted with a rheumatologist. It is clear that knowledgeable health care providers impact positively on disease outcome in a chronic condition such as RA. Educators and managed care plans must produce knowledgeable physicians and ensure patient access to these skilled care givers. 78   MacLean, CH, KK Knight, PG Shekelle, HE Paulus, and RE Brook (1996). Drug use in rheumatoid arthritis (abstract) Arthritis and Rheumatism, 39: R5.

OCR for page 73
--> Invited Address Edward Yelin This paper presents the results of a study designed to assess the health care used and medical outcomes experienced over an 11-year period by persons with rheumatoid arthritis receiving care in fee for service (FFS) and one form of managed care, the prepaid group practice (PGP) form of HMO.79 Because these two sectors of care differ in their propensity to use subspecialists in the treatment of conditions such as RA, I will also present some data on whether in fact subspecialists and generalists differ in outcomes. In order to place these results in a proper context, I will summarize what studies in the literature say about how fee for service and managed care differ in the process of care and in outcomes. Subsequently, I will describe the design of the University of California, San Francisco (UCSF) RA panel, the data set used in the analyses reported on here, and will present the study findings. Finally, because the PGP form of HMO is the slowest-growing form of managed care and because the structure of the faster-growing forms is not conducive to the conduct of research, I will close by advocating that we graft outcomes studies onto currently mandated quality assurance mechanisms in order to assess the impact of managed care on persons with chronic disease in general and rheumatoid arthritis in particular. 79   For a more complete report of this work see Yelin E, L Criswell, and P Feigenbaum (1996). Health care utilization and outcomes among persons with rheumatoid arthritis in fee-for-service and pre-paid group practices. Journal of the American Medical Association, 276: 1048–1053. The study was supported by NIAMS Grant AR-20684 and the Arthritis Foundation.

OCR for page 73
--> Literature on Health Care Utilization and Outcomes in FFS and HMOs Most of the literature on managed care concerns the PGP form of HMO, and most of these studies—including one randomized trial, the RAND Health Insurance Experiment—concern healthy populations. Results of these studies indicate that the PGP form of HMO and fee for service use similar amounts of ambulatory care. However, PGPs use the hospital 25 percent less often on average, and this results in about 25 percent savings in total medical care costs. Studies of other forms of HMO, such as the independent practice association, are less numerous, and their results in terms of health care utilization, costs, and outcomes are inconclusive. Results from the general literature indicate that PGPs save costs by treating the healthy and acutely ill less intensively. However, there are few studies of persons with discrete chronic conditions, and their results show no consistent difference in the process or outcome of care. The largest such endeavor, Medical Outcomes Study (MOS), found no meaningful difference in outcome between fee for service and managed care for persons with hypertension or non-insulin dependent diabetes.80 With respect to rheumatic diseases, we published results from the first two waves of the UCSF panel and reported that utilization and outcomes did not differ for persons with RA in the FFS and PGP sectors.81 Similarly, Holman and his colleagues from Stanford studied utilization and costs for osteoarthritis and found that persons with this condition receiving care in the FFS and HMO sectors did not differ in these parameters. Interestingly, they reported that both systems were more expensive than an experimental plan designed expressly to provide low-cost care by engaging the patient as an active partner in care.82 Study Design Data for this report derive from the UCSF Rheumatoid Arthritis Panel. The RA panel was begun in 1982 and 1983 when we drew a random sample 80   Greenfield, S, W Rogers, M Mangotich, MF Carney, and AR Tarlov (1995). Outcomes of patients with hypertension and non-insulin dependent diabetes mellitus treated by different systems and specialties: Results from the Medical Outcomes Study. Journal of the American Medical Association, 274: 1436–1444. 81   Yelin, EH, MA Sheam, and WV Epstein (1985). Health outcomes for a chronic disease in prepaid group practice and fee-for-service settings. New England Journal of Medicine, 312: 962–967. 82   Lubeck, D, B Brown, and H Holman (1985). Chronic disease and health system performance: Care of osteoarthritis across three health services. Medical Care, 23: 266–277.

OCR for page 73
--> of 40 rheumatologists from among all then practicing in Northern California. In 1989, we replenished the panel by adding a smaller random sample of another 10 rheumatologists. Participating rheumatologists maintained logs of all persons meeting strict criteria for RA presenting to their offices over a one-month period and provided standardized measures of severity. We successfully enrolled 97 percent of all the persons with RA listed in the physician logs, 1,025 in all. Each year, a trained interviewer conducts an hour-long survey covering demographic characteristics; signs and symptoms of RA; number and kind of comorbid conditions; a complete inventory of all health care used for RA and other purposes, including the content of ambulatory encounters and hospital admissions; and generic and disease-specific measures of health and functional status. The survey also collects information about the kind and amount of health insurance. We categorize persons with RA into those receiving care in HMOs or FFS on the basis of their responses to these items. Overall, 227 of the 1,025 patients in the panel reported that an HMO was their principal health insurer. Of these, 196 (86 percent) reported receiving care in Kaiser Permanente. Accordingly, the results of this study are especially germane to the pre paid group practice form of HMO. As of the end of 1994, we had accumulated more than 7,700 person-years of follow-up with RA. Results Health Care Utilization There were no systematic differences in health care utilization for RA between persons in the FFS and PGP sectors over the 11 years of the study. With respect to ambulatory care, both groups made about 90 visits to rheumatologists for RA during this time. Similarly, both groups made about 20 visits to all physicians other than rheumatologists for RA, or roughly 10 visits a year. The two groups did not differ in the distribution of visits to various medical specialties. Of course, persons with RA have physician visits for reasons other than RA. When we evaluated the number of non-RA visits over the 11-year period among persons with RA in the FFS and PGP sectors, the difference almost reached the traditional criterion for statistical significance, with the former group reporting an additional 5 visits, for a total of 39. PGP physicians in our study attributed this finding to the fact that in most PGPs, rheumatologists provide primary care for rheumatoid arthritis patients during their visits for RA. In contrast, in the FFS sector, the physicians receive additional compensation when their patients have non-RA visits.

OCR for page 73
--> In RA treatment, the hospital accounts for the majority of all medical care costs. Persons with RA receiving care in the FFS and PGP settings did not differ in the number of RA-related hospital admissions over the 11 years of the study. They also did not differ significantly in the proportion with one or more hospitalizations; 55 percent among persons with RA in PGP settings and 54 percent among those in FFS. On the hospital side of the cost ledger, total joint replacement is the single most expensive item. Persons with RA receiving care in the FFS and PGP sectors did not differ in the proportion receiving total joint replacement surgery, although there was a slight trend for the former group to have higher rates of joint replacement surgery. In contrast, there was a slight trend for those in the PGP sector to have higher rates for all kinds of surgeries. However, neither trend was sufficient to render invalid the overall conclusion that medical care utilization did not differ systematically in the two sectors. In addition to evaluating the total number of ambulatory encounters, we analyzed differences in the content of these encounters in the two sectors of care. Persons with RA receiving care in PGPs were significantly more likely than those receiving care in FFS settings to receive an x-ray; those in FFS were significantly more likely to receive a complete physical exam. The two groups did not differ in the other eight characteristics of ambulatory encounters analyzed, including the proportion receiving gold injections, or urine and blood tests, the highest-cost items. Health Outcomes The RA panel study includes measures of symptomatology, including extent of pain, stiffness, and swelling; functional status; and global health status. In interpreting the outcomes for those receiving care in FFS and PGP settings, it is important to note that all analyses take into account differences in the baseline characteristics of the two groups, with the results reported presumably representing the long-term effects of the two systems of care. With respect to symptomatology, the two groups did not differ significantly in the number of painful or swollen joints, the way in which they rated their overall pain, the duration of morning stiffness, and the proportion reporting improvement in the overall status of their RA. Persons with RA have traditionally rated functional capacity as the most important outcome of care. Of note, persons with RA receiving care in FFS and PGP settings also did not differ in the disease-specific functional status measure or the global measure of function. Finally, the two groups did not differ in the proportion

OCR for page 73
--> reporting simultaneous improvement in all of the outcome measures, although there was a slight trend in the global measure of function favoring those in PGP settings. Summary of FFS-PGP Comparisons Thus, to summarize, the results for the comparison of FFS and PGP settings are broadly consistent with the MOS and other studies of discrete chronic conditions in finding no significant differences in ambulatory utilization, hospital admissions, and surgery, including joint replacement. There also were no clinically or statistically significant differences in outcome measures over 11 years of the study. Comparisons of Care Provided by Rheumatologists and Nonrheumatologists Over the course of the study, roughly 17 percent of the RA patients have migrated away from the practices of rheumatologists because they moved to a different area, experienced a change in health plan, were referred back to a primary care physician, or simply grew dissatisfied with the care they were receiving. As a result of these moves, we are able to evaluate differences in the amount and kind of content of care and outcomes for persons with rheumatologists and nonrheumatologists as the main physicians for the RA. In contrast to the comparison between FFS and PGP, we could detect statistically significant differences in functional status, number of painful joints, rating of overall pain, proportion reporting improvement in overall RA status, and proportion experiencing simultaneous improvement in all measures. In each case, persons with RA in the care of rheumatologists fared better in the outcome measure than those in the care of nonrheumatologists. In addition, for outcome differences that did not meet the traditional criterion of statistical significance, the point estimate favored those in the care of rheumatologists. When we evaluated the treatments used by rheumatologists and nonrheumatologists used in providing care for persons with RA, we found that the former were 50 percent more likely to use any form of disease-modifying agent as well as the current such agent of choice, methotrexate. In contrast, there was no difference in the proportion of persons with RA in the care of rheumatologists and nonrheumatologists who used nonsteroidal NSAIDs or the proportion receiving surgery in general and total joint replacement in particular.

OCR for page 73
--> Thus, to summarize results for the comparison of subspecialty versus generalist care, we found that persons with a rheumatologist as their main RA physician fared significantly better on several measures and that trends favored them on all other measures. We also found that these persons were more likely to receive a disease-modifying agent, especially methotrexate, the current therapy of choice for RA. Conclusions All of the large-scale studies of how persons with discrete chronic conditions fare in FFS and PGP settings fail to detect appreciable differences in outcome. The present study is no exception. However, we did find that persons with rheumatologists as their main RA physicians fared better on several measures of outcome. A contrast of the results for FFS versus PGP and subspecialist versus generalist suggests that the absence of a difference between FFS and PGP in this study may be due to the preponderance of rheumatologists in both settings. After all, we began the RA panel study with a random sample of rheumatologists. After rheumatologists have completed their fellowship training, they may be so highly socialized that the setting in which they practice does not affect the care they provide. If so, the study of Feinglass and colleagues becomes especially germane to understanding the care of rheumatic diseases because they reported that newer forms of managed care, the ones with the greatest relative growth, are much less likely to use subspecialists,83 In these plans the ratio of subspecialists to beneficiaries is much lower than in FFS or in the traditional PGP form of HMO such as Kaiser Permanente. Finally, it is necessary to point out that few studies of the impact of these new and fast-growing forms of managed care exist. We experienced substantial difficulty when we approached several of these organizations to solicit their participation in our studies. First of all, they have not been around long enough to have sufficient numbers of persons with discrete chronic conditions for study and, just as importantly, because they are new, they have a rapid turnover in beneficiaries. Secondly, they do not have an infrastructure for supporting such studies. When we began our studies, we were able to solicit individual physicians because they were not part of larger plans. Now that most have contracts with plans, the individual physician has to receive the approval of the entire plan to participate in any studies. This suggests that if 83   Feinglass J, J Schroeder, B Gifford, and L Manheim (1992). Gatekeepers and the medical specialist: The impact of managed care on rheumatologists, Journal of the American Association of Preferred Provider Organizations, 2: 13–36.

OCR for page 73
--> we are going to do a systematic evaluation of the newer forms of managed care, we may have to graft studies onto ongoing forms of quality assurance that either are mandated by the Health Care Finance Agency or are necessary if the health plan wants to market its services to large employers. The notion of grafting these kinds of studies onto quality assurance mechanisms makessense for another reason: the cost of replicating the current study comparing FFS and PGPs in multiple forms of managed care would be prohibitive, both because the individual plans may not have sufficient numbers of cases and because the plans change so quickly that it is difficult to follow persons with a discrete chronic condition and their physicians through time.

OCR for page 73
This page in the original is blank.

OCR for page 73
--> Invited Reaction Elizabeth Badley I was not really surprised that there were no differences in rheumatological care between the two payment systems. As a professional observer of rheumatological care, particularly in the United Kingdom and Canada, and in looking at the literature, I have observed that rheumatologists, on the whole, know how to do their job and do it well. The big difference when you go into clinics in different countries is consultation time. In Britain, it can be amazingly fast—a few minutes per patient. Consultation is more leisurely on this side of the Atlantic. Dr. Yelin's study raises a question of other outcomes. Most of the outcome measures here and in other studies are much more measures of disease than of illness, if I can hark back to what Drs. Barondess and Holman said this morning. We must include the less tangible outcomes for patients namely, the ability to live a full daily life, to carry on working, to feel confident, and to have self-efficacy in the management of their diseases. I think we have to take an integrated view, particularly when we are talking about changing the health care system. The patients Ed Yelin studied were very privileged, in that they were seen mainly by rheumatologists. It is interesting to note that there were, in fact, differences between those who were treated by rheumatologists and nonrheumatologists. The question that we have to ask is where the patients came from (i.e., what was the referral pathway). We have talked a lot about gatekeepers, but we need to talk about appropriate referral. We have to realize that referral is not only governed by economic costs. There are a lot of reasons why a physician might refer a patient to a rheumatologist or decide not to refer. These might relate to continuity of care, a wish to do best by the patient, or a wish not to have to

OCR for page 73
--> treat something as complicated as rheumatoid arthritis or lupus. In looking at the literature on primary care and our studies of primary care physicians in Canada, one of the big barriers to referral is lack of knowledge. Other than rheumatologists, doctors are not, in general, well trained in the examination of the musculoskeletal system or in the diagnosis and management of treatment of musculoskeletal diseases. Consider triage: if family doctors cannot even recognize lupus or rheumatoid disease when patients walk into their clinics, they are not going to refer—whatever the incentives or disincentives are. I think this points to the issue of continuing medical education for physicians, particularly those in a primary care situation.

OCR for page 73
--> Invited Reaction Bradford H. Gray The theme of a number of the papers and a lot of the commentary today has been that managed care has both promises and pitfalls. Dr. Yelin's data suggest that insofar as costs are driven by utilization, the cost savings of managed care—at least of the model that he was studying here, which was Kaiser—may be quite limited for rheumatoid arthritis. However, as he noted, the world has changed. All plans are now under cost containment pressures, and the differences across plans may occur in areas other than cost. At least three other areas seem worthy of brief mention in considering the possible effects of different models of managed care. One is that managed care might provide a means for rapidly and effectively implementing state-of-the-art knowledge regarding diagnosis, treatment, and management as these things are reflected in professionally developed practice guidelines and standards. Another hypothesis is that managed care might improve the coordination of services. In fact, this is one of the hopes and ideals of the prepaid group practice model. Still another possibility is that managed care might steer patients to the proper sequence of high-quality providers. As we think about topics like this, it is clear that there may be important differences across managed care models. The basic challenge that faces HMOs is that they must serve a population of enrollees for whom they are receiving capitated payments and must do so in a way that is satisfactory to purchasers. Let me comment briefly on both of these topics. First, an elementary point is that the costs providing care to enrollees must be kept below the capitated payment rate. There are a lot of ways in which managed care organizations can pursue this. I will not go through them all,

OCR for page 73
--> but let me just mention four or five. One is that they can remove the perverse incentives from fee-for-service practice and rely on medical professionalism. That, it seems to me, was the idea behind group and staff models, and it was enough to produce substantially lower utilization rates than traditional FFS medicine. It was not that many years ago that the lower costs of HMOs were described as something of a mystery, because the Kaisers and Group Health of Puget Sound didn't seem to have any formal mechanisms by which they achieved their cost savings. They just did. This kind of a model may be ideal in many ways. It emphasizes professional processes of decision making. It allows for the use of practice guidelines. It has high potential for integration of services in ways that make sense and provide sound referral patterns. However, the problem is that these plans are losing market share, as we all have heard. Other models have been developed that are able to contain costs just as successfully. Large numbers of potential enrollees prefer the convenience and freedom of choice that are difficult to achieve in facility-based plans. A second approach toward reducing costs is to employ various forms of utilization management. This doesn't require much in the way of collegial processes, but it does provide opportunities to move state-of-the-art knowledge into the forums in which it can be applied. However, insofar as utilization management is triggered by a particular high-cost event, it may not be particularly satisfactory as the solution to the problem of the patient with a chronic illness. I will mention the third and fourth methods together: controlling costs by reducing the fees in FFS payments—that is a really popular one—and shifting risk downward. These are fairly easy for managed care organizations to do, but they have a number of negatives. They can create conflicts of interest, which have come up several times today. For purposes of this discussion, however, the important factor seems to be that by themselves, these methods do nothing to improve management of the care of the chronically ill, either by transmitting knowledge and practice guidelines, by improving the coordination or continuity of services, or by steering patients to high-quality providers. They may in fact excessively deter referrals to specialists for management or consultation. The fifth method of controlling costs and trying to keep costs in line with premium is to control them through favorable enrollee selection. This has come up several times today, and Karen Davis will also discuss how the high costs are concentrated in a relatively small number of patients in any insured population. From the standpoint of care of the chronically ill, this is potentially the most dangerous cost containment option, because it implies that the organization will benefit from providing poor care to, and allocating resources away from, the chronically ill. I think this is a valid concern because plans may be able to exercise influence on services to the seriously

OCR for page 73
--> ill through the strategies used in creating provider networks or through the structuring of incentives. All in all, however, I think that it is not as easy to control costs by this method as one might assume. For one thing, chronic conditions are really pretty widespread in the population. I had a postdoctoral student analyze some data that had been collected from the Employee Health Value Survey, which was done a couple of years ago among employees of Xerox, GTE, and Digital. All of the people in the sample were active employees and were surveyed about their satisfaction with their health benefit programs. A couple of questions asked about illness and disability: 80 percent of the people who were surveyed—all of whom were employed—reported that they had at least one chronic condition or illness, and 36 percent reported having at least three such conditions. Almost 50 percent of the employees reported moderate or severe problems with at least one activity of daily living, and 15 percent reported having problems with at least three activities. I would not contend that these are all disabled people who are having a hard time getting along in life. The point is simply that even a fully employed population is not just a bunch of healthy people. There are many people with serious ongoing illness in an employed population. A managed care organization that tries to follow a strategy of not providing service to people who get sick will soon acquire a pretty bad reputation. It is not as simple as it may seem to steer resources away from the people who need them in the hope that they will disenroll. That brings us to the role of the purchaser. Purchasers do have a strong interest in seeing that they are getting what they pay for. They are paying a premium for services for their employees, and they expect the employees to be able to obtain the services paid for. Employers have led the demand for performance measurement and reporting by health plans. The initial versions of Health Plan Employer Data and Information Set (HEDIS) had only a few measures of performance relating to chronic illness, but there is no doubt that future versions of HEDIS will move in that direction. All the pressure is in that direction, and the purchasers are interested in doing so as well. I will conclude with the observation that purchasers are the natural allies of providers who are committed to the development and provision of cost-effective services for people with chronic illness. They are a voice that belongs in discussions such as those being held today, and they are certainly a voice that belongs in discussions about the future of managed care and chronic illness.

OCR for page 73
This page in the original is blank.

OCR for page 73
--> Discussion RONALD MACKENZIE: As an internist who practices at a rheumatic disease hospital, perhaps it did not disturb me as much as it did some of my colleagues to see this lack of differential between rheumatologists and internists in outcomes for people with rheumatoid arthritis. Still, it does run a little bit against what one would expect. Did you have any opportunity to look at subgroups of patients based on severity or complexity of disease? A second question is whether we are to conclude from this that all one really has to do to negate the benefit of rheumatologists in the care of patients with rheumatoid arthritis is to increase the comfort level of the general internist in the use of methotrexate? EDWARD YELIN: I will answer the second question first. We don't know the answer, but that is a testable hypothesis. As to the first question, we did not have the statistical power to detect a small difference between rheumatologists and nonrheumatologists. We did find some differences that favored rheumatologists, and all the point estimates favored the them. My take on the data is a little different from what you were implying. It isn't much of a difference, but it is something. If you are a person with a chronic disease and you don't have much that you can control but could join a system with for example, a 10 percent increase in the proportion of people who say their RA has improved, or a 5 percent difference in function, or even a 2 percent difference in joint count—if that is something that you could control by the choice of a physician, it appears to me to be something you would do. Capturing that little difference may not be cost-effective from the perspective of plan administrators, but from the perspective of a person with the condition, it may be important to them.

OCR for page 73
--> BEVRA HAHN: I want to share another experience with the panel and get your reaction to it. At the University of California, Los Angeles (UCLA) multipurpose arthritis center we have started a different type of project, one in which we take a big population of people with rheumatoid arthritis and start defining them on the basis of whether or not they have seen a specialist, rather than starting with a specialist's identifying patients with rheumatoid arthritis. Catherine McLean led this project, which has published data on RA cost in abstract form.84 She looked at a large managed care organization database, where she found some 12,000 individuals coded in the International Classification of Diseases (ICD-9) as rheumatoid arthritis patients. Of that 12,000 people, more than 80 percent had never seen a rheumatologist. More than 50 percent were not receiving any anti-inflammatory drug therapy (salicylates or NSAIDs) and were being managed with analgesics, usually narcotics. Only about 15 percent had ever received a disease-modifying antirheumatic drug during their care. So, if we take what rheumatologists think is a good approach to the care of an individual with rheumatoid arthritis, we would have to say that this large proportion of people was not receiving such care, which bring up some of the issues Dr. Badley raised in terms of education within our groups. BRADFORD GRAY: There is an important point to be made about sample selection and research. The way in which a sample is selected does have consequences for the kinds of things you will find in the data. I am really sort of struck by those findings. ELIZABETH BADLEY: The findings of our recent survey of family physicians in Ontario suggest that almost half would not refer an early rheumatoid arthritis patient to a rheumatologist in a situation where referral was indicated in the view of a current practice panel. Similarly, another study of Ontario family doctors with regard to referral for knee replacement surgery showed a reluctance to refer, often on grounds of old age or obesity, when such reluctance was not justified. This work was done as part of the Patient Outcome Research Team program, with parallel work in the United States. We need to do more in medical schools to make sure that all who come through them are trained in the examination of the musculoskeletal system. Our study showed that family physicians have a similar level of confidence in examining the musculoskeletal system as they do for neurological exams. They are more confident with the cardiovascular system. 84   MacLean et al., op. cit.

OCR for page 73
--> I think we have to start with our medical schools, trying to ensure that qualifying physicians have basic musculoskeletal examination skills. These probably do not have to be terribly sophisticated but should be sufficient to examine the major joints relatively comprehensively. We also have to give higher priority to continuing medical education in this area. JEREMIAH BARONDESS: It is difficult for me to believe that this problem affects only rheumatic disorders. As a matter of fact, I think that education for clinical management of chronic disease is weak. It is not lacking. It is just lost in the enormous amount of biologically organized material that people are asked to absorb as the summum bonum of the educational process. If you are able to understand the cardiac cycle, it seems to me that you ought to be able to figure out the nuances of clinical care and how to stay conversant with literature that will alert you to ways to alter a disease with some kind of reasonable intervention. So, I think the issues are broader, although you are right, of course, in suggesting that all new physicians should be able to examine joints. BRADFORD GRAY: In theory, this is the sort of problem that should be easier to deal with in managed care. Presumably the McLean study that Bevra Hahn described used a managed care database. So, this gets back to the comments I made about certain forms of managed care not having the capability within them to do very much about this sort of problem at all, insofar as they rely on networks, discounts and so forth. BEVRA HAHN: The study looks at a mix of all kinds of care structures and of people shuffling between one care structure and another. Only about 15 percent of the individuals were in fee for service. The others were all in some kind of managed care plan. THEODORE FIELDS: Dr. Yelin, were there any practice guidelines at Kaiser at the time of the study? If not, will the newly published practice guidelines for rheumatoid arthritis be relevant to new studies, and how can they be incorporated looking at the kind of questions you examined? EDWARD YELIN: I actually raised that issue with my colleagues in the Kaiser system. Rheumatologists there see so much RA, they say, that they would read the guidelines the day they came out and either find out that they were already pretty much adhering to them or shift their practices slightly to adhere to them. It is more an issue for early RA in the system of care there. In the Kaiser system, RA patients get referred quite early because these people are not very appealing patients for the general internist, who doesn't see that much RA.

OCR for page 73
--> There, in that closed system, they have an opportunity to refer it, which is exactly the opposite of the incentive system in the fee for service, where they could get more money by taking care of these people. This conference was organized around RA and SLE, but it would have been interesting to add osteoarthritis (OA) to the mix. OA, I think, is an example of where the promulgation of practice guidelines would have a far more profound impact in the Kaiser system, because it is a much more common disease and one that general internists, by and large, follow throughout care. I don't think rheumatologists see any but a handful of these cases and then mainly on a referral basis. WILLIAM HAZZARD: There appears to me to be one sector of the health care industry that really understands the problem, and that is pharmaceutical companies. Rarely a day goes by that I don't have one of them offering to help me educate my residents on how to give sophisticated care for osteoarthritis, osteoporosis, dislipoproteinemia, and other sorts of problems that are underdiagnosed and undertreated until you get to a subspecialist. LAURA ROBBINS: There are two studies, one done by the Arthritis Foundation in San Mateo and one done at Cornell, that have looked at patients' beliefs and attitudes in terms of the seriousness of arthritis and whether or not anything could be done about it. In summary, the findings were that it was seen as an old person's disease about which nothing could be done. One wonders if primary care physicians, based on their orientation and training and their own personal belief, do not see arthritis as something that is not a serious disease, and therefore referral patterns and treatments lag relative to diseases such as cancer. EDWARD YELIN: Hal Holman and I were on the board at the San Mateo project. Indeed, a significant fraction of primary care physicians felt that osteoarthritis and most musculoskeletal conditions were not that difficult or important a problem, so they really didn't pay much attention to these diseases. A minority of physicians thought that these were important problems but weren't going to refer them because they were the bread and butter of their business. ELIZABETH BADLEY: We have to go all the way to the general public with the message that there are different kinds of arthritis, some of them more serious than others, some requiring a specialist and some not. We need to tell them that there are things you can do for arthritis and that nobody should ever tell you to go home because nothing can be done. There are things we can do, even if only in self-management. I think if we had a more educated public, it would also act as pressure on doctors for referral and better services.