THE CONGRESSIONAL CHARGE TO THE INSTITUTE OF MEDICINE
In the Social Security Amendments of 1972, Congress created an entitlement to Medicare for all persons with a diagnosis of permanent kidney failure who were fully or currently insured or eligible for benefits under Social Security, and for spouses or dependent children of such persons. The End-Stage Renal Disease (ESRD) program thus established has been very successful, saving several hundred thousand Americans from premature death and giving hope to individuals who once faced certain death. At present, over 150,000 ESRD patients receive Medicare services, including inpatient and outpatient care, from both physicians and treatment units, for dialysis, kidney transplantation, and other medical services.
The ESRD program is unique within Medicare. It is the only case in which the diagnosis of a categorical disease provides the basis for an entitlement for persons of all ages. In addition, both ESRD patients and providers depend on Medicare reimbursement policy to a greater extent than in any other domain of medicine. Its highly visible status has led Congress to request studies of various aspects of the program from time to time.
In the Omnibus Budget Reconciliation Act of 1987 (OBRA 1987),1 Congress asked the Institute of Medicine (IOM) to study the ESRD program with respect to the following issues:
the quality of care provided to ESRD beneficiaries, as measured by clinical indicators, functional status of patients, and patient satisfaction (Chapter 12),
the adequacy of existing data systems to monitor these matters on a continuing basis (Chapter 13).
The IOM convened an expert committee to conduct the study; this report constitutes its response.2 The committee also examined patient perspectives (Chapter 2), ethical issues (Chapter 3), the structure of the provider community (Chapter 6), and ESRD research needs (Chapter 14).
The committee believes that people with permanent renal failure should be the primary concern of Medicare ESRD policy. This emphasis is reflected in the chapters that deal with patient perspectives, ethical issues, and the epidemiology of the patient population, as well as in the committee's concern with access, quality, and reimbursement.
Some recommendations of this committee would entail increased funding. The committee recommends (in Chapters 7 and 8) that Congress extend the Medicare entitlement for ESRD treatment to all Americans, remove the three-year limit on Medicare eligibility of successful transplant recipients, and pay for immunosuppressive drugs for the period of a transplant patient's entitlement. It proposes these as ways to improve an already successful program by increasing the equitable treatment of all ESRD patients and promoting improved patient outcomes.
The committee was not asked to address, nor has it considered, the broader health needs of the American people or the relation of expenditures for ESRD patients relative to other deserving claimants. However, the committee recognizes that Congress must weigh the recommendations that involve increased funding against competing uses of resources and must make the appropriate and equitable decisions regarding the allocation of scarce resources. The committee is quite aware of the fact that the continuing annual federal budget deficits and increasing health care costs make such decisions very difficult.
The IOM committee, therefore, endorses the following objectives for the Medicare ESRD program: to guarantee access to treatment for all for whom it is medically appropriate; to provide care of high quality that achieves desirable health outcomes consistent with patient health status and current professional knowledge; to develop policies that steadily improve patient well-being and patient outcomes; and to manage the program prudently at the lowest cost compatible with adequate care.
ESRD PATIENTS AND THEIR TREATMENT
The number of patients under treatment for ESRD has grown substantially over time, increasing from approximately 10,000 beneficiaries when the entitlement became effective in 1973, to over 150,000 dialysis and transplant patients today. In the year 2000, it is projected that a quarter of a million patients will be enrolled in the ESRD program.
The ESRD patient population is highly diverse in age, sex, race, and primary diagnosis and has changed greatly over time. For example, those 65 years and older increased from 5 percent of the total Medicare ESRD population in 1973 to 27 percent in 1988, and accounted for 38 percent of all new patients in 1988. The major diseases causing kidney failure are glomerulonephritis, diabetes, and hypertension. The proportion of ESRD patients with a primary diagnosis of diabetes mellitus increased from about 13 percent in 1980 to 24 percent in 1988; this diagnosis accounted for 31 percent of all new patients in 1988. Those with a primary diagnosis of hypertension increased from about 20 percent in 1980 to 24 percent in 1988; they constituted 27 percent of new patients in 1988.
Kidney failure is nearly four times as frequent among blacks as in whites. In the late 1960s, black patients constituted 10 percent of the treated ESRD patient population; today they represent 29 percent of the total population and 28 percent of all new patients. Several studies suggest that renal failure in the growing Hispanic population also is substantially higher than in the white population. These trends—more elderly, more diabetic and hypertensive patients, and disproportionate representation of minorities—have characterized the ESRD program in the 1980s and will continue to do so in the 1990s.
The ESRD patient population, as of December 31, 1988, is distributed by treatment modality as follows: 60 percent, hemodialysis; 10 percent, peritoneal dialysis; 25 percent, transplantation; and 5 percent, unknown. Even though dialysis is the major ESRD treatment modality, kidney transplantation is clearly the treatment of choice for the great majority of renal failure patients in the committee's view. Kidney transplants increased from almost 3,200 in 1974 to almost 9,000 in 1986. The number of Medicare ESRD beneficiaries with functioning transplanted kidneys was 30,000 at the end of 1989; another 10,000 individuals with functioning transplants are estimated to have left the Medicare rolls after reaching the three-year limit on eligibility for benefits. The outcome of both living related and cadaver donor transplants, as measured by survival of the recipient and the transplanted kidney, has improved markedly during the past decade largely because of improved ways to prevent kidney rejection. The total number of transplants performed, however, has leveled off since 1986 due to a shortage of kidneys.
A major challenge of ESRD treatment, for both individuals and society, is its cost. Based on 1987 cost data provided by the Health Care Financing
Administration (HCFA), the annual Medicare expenditure for a dialysis patient was $32,000. First-year expenditures for a successful transplant patient were $56,000, but were only $6,000 in each subsequent year. These costs are not atypical for patients with serious, chronic diseases. Medicare, of course, does not cover all costs either for dialysis or transplant patients: premiums, deductibles, and copayments must be met from other sources.
Although total ESRD program costs are high, cost control for dialysis has been impressive. The initial reimbursement rates for outpatient dialysis were established in 1973 and remained unchanged until 1983. They were lowered in 1983 and again in 1986. During this time, no adjustment was made for inflation. In real dollars, payment rates per dialysis treatment have fallen steadily over the program's history. The outpatient dialysis reimbursement rate was $138 in 1974 and $125 in 1989. When adjusted for inflation by the GNP price deflator, the 1989 reimbursement rate was less than $54 in 1974 dollars, a 61 percent reduction over this period.
In 1988, Medicare spent $3.0 billion ($1.3 billion in 1974 dollars) for ESRD beneficiaries. (A recent estimate by HCFA, using a new methodology, places 1988 expenditure at $3.7 billion.) Total national expenditures for ESRD patients from all sources are estimated at $5.4 billion in 1988. The total cost of the program has grown primarily because the number of patients has increased steadily over time. It has been controlled by progressively lower real costs per dialysis treatment, by the beneficial cost effects of successful transplantation, and by some shifting of payment responsibility to private payers.
The kidney provision of the Social Security Amendments of 1972 removed most of the financial barriers to ESRD treatment for the vast majority of U.S. citizens. However, it did not establish a universal benefit, a fact not widely appreciated. From the outset of the program through 1986, Medicare-certified dialysis facilities reported 6–7 percent of the ESRD dialysis patients as not eligible for benefits; this proportion increased to 7.2 percent in 1987 and 7.5 percent in 1989. The absolute number of such patients has been growing, and substantial variation exists among states and major cities.
The major non-Medicare sources for financing ESRD treatment are not adequate to meet the needs of the unentitled patients. The Department of Veterans Affairs dialysis program is shrinking, the Indian Health Service faces many other demands on its limited resources, and the 19 state kidney disease programs have benefits that vary and budgets that are not increasing. State Medicaid programs, therefore, represent the only payer for many ESRD patients who are not eligible for Medicare. Although Medicaid ESRD benefits appear to be adequate in many states, in others they are not, as eligibility criteria and covered benefits vary widely according to state discretion. Moreover, as state Medic-
aid programs consume a rapidly increasing proportion of state government budgets, they will be closely scrutinized by governors and legislatures and are unlikely to receive more resources for ESRD services.
Practically speaking, eligibility for Medicare ESRD coverage is a function of Social Security insured status.3 Those ineligible include some federal, state, and local government employees,4 domestic, farm and all other workers in covered occupations but who may not have applied for benefits, and those who have never worked, such as young, nonworking mothers, many welfare mothers, and their children. Persons in most of these categories are disproportionately concentrated among the poor and minorities.
The available evidence on ESRD patients who are reported as not eligible for Medicare coverage is sparse. No information is available about their demographic, health status, and socioeconomic characteristics nor about the effects of their noneligible status on the receipt or denial of treatment. Clearly, information on these matters should be obtained. However, the committee does not believe that access to life-saving dialysis and transplantation should depend on the precise number of ESRD patients not eligible for Medicare. Limiting Medicare ESRD eligibility to those with Social Security insured status places some excluded persons at risk of death.
As a matter of equitable treatment of all individuals with end-stage renal disease, the committee finds no justifiable basis for restricting Medicare eligibility other than citizenship or resident alien status. It recommends that Congress modify the Medicare eligibility criteria for individuals with ESRD and extend the entitlement to all United States citizens and resident aliens.
Kidney transplant patients face two major restrictions on access. First, their Medicare eligibility is limited to a three-year period following a successful transplant. Second, they are eligible for only a single year of reimbursement for expensive immunosuppressive drugs which they must take on a life-long basis. Transplant patients are not cured: they remain at risk of rejecting the transplanted kidney throughout their lives; the medications they must take to prevent rejection place them at risk of other medical problems. Congress, in various pieces of legislation enacted in 1978, 1984, 1986, 1988, and 1990, has sought to encourage organ transplantation. However, the committee believes that the Medicare limits on eligibility of transplant patients and on payment for immunosuppressive drugs inadvertently tend to discourage kidney transplantation and should be changed to foster this treatment.
The committee recommends that Congress eliminate the three-year Medicare eligibility limit for successful transplant patients and thereby authorize a lifetime entitlement comparable to that of dialysis patients.
The committee also recommends that coverage for payment of immunosuppressive medications for kidney transplant patients be made coterminous with the period of entitlement.
The reasons for the shortage of transplantable kidneys are not well understood. Although several studies suggest that only a fraction of suitable kidneys is made available for renal transplantation, the extent of the gap between potential and actual donors is unclear and effective ways to close this gap are not well understood. When coupled with the rapid annual increase in new ESRD patients, the shortage of kidneys means that dialysis will continue to be the predominant form of ESRD treatment. Because transplantation is generally recognized as the preferred therapy for the great majority of ESRD patients, because kidneys are scarce, and because the long-term cost of transplantation is less than dialysis,
the committee recommends that increasing the donation of kidneys receive very high priority in the coming decade.
Dialysis and transplantation have raised ethical questions since their introduction to clinical practice in the early 1960s. The questions have changed over time as the patient population, treatment, and financing have changed. The committee focused on three major concerns—the acceptance of patients for treatment, the termination of treatment, and ethical questions arising for caregivers who deal with problem patients.
Some concern has been expressed that patient acceptance criteria have expanded since enactment of the Medicare ESRD entitlement, resulting in an increasing number of patients with limited survival possibilities and relatively poor quality of life. The committee, recognizing this concern, believes that patient acceptance criteria should be medical, not economic, and based on concern for the best interest of individual patients. Age was considered and explicitly rejected by the committee as a patient acceptance criterion, as it does not measure the ability of an individual to benefit from treatment. Comorbidities—at any age—are the primary determinants of quality of life and of survival.
Decision making about the initiation of treatment should result from informed discussion among the patient, the family, the physician, and other caregivers. ESRD patients usually rate their quality of life higher than do ''objective'' observers. This emphasizes the need to regard patient preferences very highly in decisions about their care. However, physicians should recognize that the existence of a public entitlement does not mean that they are obligated to treat all patients who present with kidney failure. Clinical judgment and patient-family preferences will sometimes indicate terminal palliative care rather than life-extend-
ing care. Thus, the choice is not between treatment and abandonment, but rather between different goals of treatment.
Withdrawal from treatment by competent patients is increasingly reported, especially among the elderly. This should be regarded as a rational decision by an autonomous person who has concluded that the burdens of continuing treatment outweigh the benefits. The termination of treatment of incompetent patients should occur only after full discussion between the patient's family, or other representatives, and the physician.
Three types of patients raise different ethical problems for clinicians today: the noncompliant, self-destructive dialysis patient; the hostile, abusive dialysis patient; and the self-destructive transplant patient. The physician's responsibility is to care for the patient with an understanding of human frailty and the complex psychology of living with chronic illness, to develop and maintain effective communication with them, and to assure continuity of care. The self-destructive transplant patient whose behavior threatens the survival of the transplanted kidney may disqualify himself as a candidate for a subsequent transplant if the current organ fails.
The committee recommends that patients, clinicians in adult and pediatric nephrology, and bioethicists develop guidelines for evaluating patients for whom the burdens of renal replacement therapy may substantially outweigh the benefits. These guidelines should be flexible and encourage the physician to use discretion in assessing an individual patient. Guidelines should be developed specifically for children and should describe the role of the parents in the decision-making process.
Nephrologists and other clinicians should discuss with all ESRD patients their wishes about dialysis, cardiopulmonary resuscitation, and other life-sustaining treatments and encourage documented advance directives.
ESRD clinicians should be encouraged to participate in continuing education in medical ethics and health law. Some specialists in the medical ethics of renal disease should be available to educate clinicians, to train members of ethics committees, and to do research on ethical issues in dialysis and transplantation.
THE PROVIDER COMMUNITY
The organization of the provider community has changed over time. During the past decade, the number of transplant centers increased from 150 to 219. Organ procurement organizations (OPOs) shifted from being hospital-based to being independent entities, and consolidation of OPOs occurred in response to the National Organ Transplant Act of 1986.
Treatment capacity for outpatient dialysis grew from 1980 to 1988: the number of units increased at an annual rate of 7.1 percent, from 1,004 to 1,740 units; the number of approved dialysis stations increased at an average annual rate of 8.1 percent from 12,216 to 22,803.5 The demand for treatment also grew during this time; outpatient hemodialysis treatments increased at an average annual rate of 9.7 percent from over 5,672,000 in 1980 to 11,866,000 in 1988. The increased demand for treatment outpaced the growth in treatment capacity and resulted in an increased utilization of units and stations at the national level. On empirical grounds, the committee concludes that the rate of growth in outpatient treatment capacity, measured by treatment stations, is not unreasonable in relation to the increased demand for treatment, measured by number of treatments. It further observes that normative criteria relating capacity to demand have not been advanced by either the provider community or HCFA.
Independent rather than hospital-based units accounted for most of the growth in outpatient dialysis facilities. They now account for 70 percent of all dialysis stations. The independents' share of dialysis treatments increased from 40 percent to more than 60 percent between 1980 and 1988. For-profit units, primarily independent, increased more rapidly than not-for-profit units. The committee is not aware of any evidence that shifts among providers from hospital-based to independent and from nonprofit to for-profit dialysis facilities have resulted in access or quality problems.
It recommends, however, that such changes be monitored closely to assess their implications for access and quality.
Differences in growth of dialysis capacity are apparent between states having certificate-of-need (CON) and those without CON. CON states have had much slower increases in the number of units and stations, and thus greater utilization in both treatments and patients per station.
Who benefits from state CON regulations as they are applied to ESRD units? In particular, what are the effects of higher utilization (or productivity) of treatment capacity in CON states? Existing data permit only a limited response. First, higher utilization does not reduce Medicare expenditures, nor does lower utilization in non-CON states involve any additional cost to Medicare. Second, CON confers local monopoly benefits on approved providers, allowing them to reap the economic benefits of increased productivity; the costs of the legal defense of monopoly, moreover, can be absorbed more easily by provider chains than by single-owner facilities. Third, by sanctioning local provider monopoly, CON decreases patient choice of providers and eliminates competition among facilities for patients on the basis of amenities. Fourth, CON provides no assurance of improved quality, for which it was not designed. Finally, in some states it has contributed to serious access problems.
The committee finds no persuasive reason for the application of state certificate-of-need regulations on dialysis treatment capacity that is regulated mainly by the Medicare payment level. The committee strongly favors the elimination of CON as applied to dialysis facilities, but recognizes that this requires state government rather than federal government action. It recommends that HCFA review with each of the relevant states the effect of CON regulations on ESRD patient access to care in light of national data on utilization.
REIMBURSEMENT AND QUALITY
The committee confronted several basic facts as it considered the effects of reimbursement on quality. First, reimbursement rates for outpatient dialysis, in real dollars, have steadily decreased over the 18-year history of the ESRD program, due both to a fixed rate (from 1973 to 1983) that was never adjusted for inflation as well as to explicit rate reductions in 1983 and 1986. No other part of the Medicare program has been subjected to a similar reimbursement policy.
Second, quality has not been systematically measured or monitored by the federal agencies responsible for the ESRD program. HCFA has initiated a quality assurance effort within the past several years, but it currently bears no relationship to reimbursement policy.
Third, HCFA argues that reimbursement rates should be cut still further, although no system exists to monitor the effects of such cuts on quality. On the other hand, many providers, patients, and researchers believe that quality has been adversely affected by the progressive reductions in reimbursement and that further cuts will erode quality dangerously. Regrettably, existing data are inadequate to document adequately whether or how quality has changed over time or to determine whether the present level of quality is reasonable in the context of the patient population and current professional knowledge. Evaluating these divergent views has been a major endeavor of this committee.
Reimbursement Effects on Quality
The committee examined several aspects of the effect of reimbursement on quality: mortality, morbidity (as measured by hospitalization), dialysis unit staffing patterns, and treatment innovation.
The trends in mortality in the ESRD program in the 1980s have received much attention. After a period of stable mortality rates from 1978 to 1982, an abrupt rise in unadjusted ESRD mortality occurred between 1982 and 1983, and gross mortality has increased since then. Interpretation of this rise is complicated
by changes in the patient population, treatment patterns, and methodological factors. For example, the incidence and prevalence of elderly and diabetic ESRD patients, who often have other comorbid conditions, have increased. These patients have higher mortality rates than younger patients or those without diabetes. ESRD mortality data, when adjusted for age and primary diagnosis, show stability over time.
The committee examined the question of the effects of reimbursement on mortality, which only two studies have addressed directly. In 1987, an Urban Institute report (Held et al., 1987) found no association between the introduction in 1983 of the composite rate, which reduced reimbursement to most providers, and a change in mortality. In 1990, this study requested the Urban Institute to update its earlier analysis. This more recent analysis, discussed in Chapter 10, suggests, by one approach (a price-level model), an inverse relation between patient mortality and the standardized price (reimbursement rate) received by a dialysis facility. This relation was not statistically supported by a second approach (a first-difference model). The committee regards these results as suggestive but insufficient to establish firmly a direct effect of reimbursement on quality.
An indirect effect of reimbursement changes on mortality is suggested. Treatment times of dialysis patients have become shorter in the past decade, due partly to the pressures to reduce costs by treating more patients in a given nursing shift. Shorter treatment time has been implicated in higher mortality. Although an inverse relation of reimbursement to mortality is not proved directly, decreased reimbursement is associated with shorter treatment time. The committee interprets this possible indirect effect cautiously, since factors such as prevailing clinical opinion and patient preferences appear to have influenced treatment time.
The use of mortality as an outcome measure represents a positive step toward the quantification of outcomes. However, the emphasis on the effect of the 1983 composite rate on mortality may be inappropriate. The changing patient population has been mentioned above, as has the effect of clinical opinion on shorter treatment time. In addition, the reimbursement reduction introduced by the 1983 composite rate was small compared with the unadjusted effects of inflation over a longer time. Finally, physicians and treatment units, confronting reduced reimbursement, tend to adapt in numerous ways that may protect the lives of patients but may result in diminished quality as reflected in increased morbidity, inadequate dialysis, reduced staffing patterns, and lower health status.
The committee examined hospitalization as a measure of morbidity. In 1987, the Urban Institute (Held et al., 1987) found that inpatient stays of a 1984 ESRD patient cohort were higher than those of a 1982 cohort, but did not find an association between the decrease in the payment rate in 1983 and increased inpatient stays in the following year. The 1990 Urban Institute study (Held et
al., 1990) provides some evidence that the level of the dialysis price may affect hospital use by ESRD patients. Using a price-level model, the study found a significant inverse correlation between the dialysis price level and hospital admission rates as well as hospital length of stay (days). These results were most conclusive for diabetic patients, who are clinically less stable and often require extra care in management on dialysis. However, when the data were analyzed by a different method in which each facility was used as its own control (first-difference model), there was no correlation between price changes and hospital use. As with mortality, there is the suggestion that there may be an indirect effect mediated by changes in dialysis treatment time. Shorter treatment time, which correlates with reduced reimbursement, also correlates with increased hospitalization. Thus, there is evidence that reimbursement reductions are associated with excess morbidity as measured by hospitalization, but no firm conclusions can be drawn.
Staffing in a treatment unit is a structural measure of quality. There is increasing evidence of staff changes as a result of reimbursement. One study found that units receiving higher payment had more total staffing hours per patient per week and more registered nurses (RNs) available per patient; the converse was true for units receiving a lower payment. Another study showed a reduction of 17 percent in total staff hours per patient per week for hospital-based outpatient units and 11 percent for independent units from 1982 to 1987. Data from several providers also show that staff-to-patient ratios decreased substantially during the 1980s.
The composition of dialysis unit staff has changed as well: in general, the proportion of RNs decreased, that of technicians increased, and licensed practical nurses and nursing assistants show a mixed pattern. Replacement of RNs with technicians decreases the clinical training and skill level of the personnel who directly treat patients.
Staffing by social workers and dietitians also decreased during the 1980s. These professionals are currently responsible for very large numbers of patients; for example, staff-to-patient ratios of 1 to 100 and 1 to 200, respectively, exist in many dialysis units. Consequently, social workers and dietitians may have been reduced to fulfilling minimal routine functions rather than the essential social or nutrition counseling considered optimal for patient care.
ESRD reimbursement policies for outpatient dialysis have resulted in reduced staffing patterns. Although optimal patterns are not known, the professional capability to treat a patient population of increasing complexity has been reduced. However, the effect of these staffing changes on patient outcomes has not been monitored and should be assessed. At minimum, HCFA should recognize the likely impact of any proposed reimbursement reductions on staffing and undertake to monitor the effect of decreased payments on patient management and outcomes.
The committee also considered the effects of reimbursement on innova-
tion. In general, economic discipline has encouraged cost-reducing, quality-enhancing technical change in the competitive equipment and supplies product market. Such economies have reduced the equipment and supply (nonlabor) component of the total cost per dialysis treatment from roughly one-third fifteen years ago to one-fifth or less at present. Further cost-saving technical change, therefore, will afford smaller economic benefits to providers and payers.
Major clinical innovations, such as cyclosporine for preventing rejection of the transplanted kidney and erythropoietin for treating anemia in dialysis patients, have resulted from progress in basic scientific research. The development of such innovations is not likely to be influenced by ESRD reimbursement policy, although their adoption and use by the medical community surely will be.
Some data suggest that decreased reimbursement may increase mortality, directly or indirectly. However, available studies are insufficient to establish either a direct or an indirect effect of reimbursement on mortality conclusively. Similarly, some studies suggest but do not prove a direct or indirect effect of decreased reimbursement on increased morbidity, assessed from hospitalization data.
Data strongly suggest that decreased reimbursement has led to decreased staffing in dialysis units, to shifts from nurses to technicians, and to important reductions in social worker and dietitian staffing. Although there is no evidence that these changes in staffing patterns have affected quality, professional opinion favors this contention.
Dialysis treatment time has decreased in the past decade. Studies indicate that decreased time may have adversely affected mortality and morbidity. The shorter times are attributable in part to economic pressure from decreased reimbursement. However, a clear relation between quality and reimbursement cannot be established because clinical judgment and patient preference also have influenced shortening of dialysis times.
The lack of progressive improvement in age-and diagnosis-adjusted dialysis patient mortality over the past decade suggests that providers may have reached the limits of increasing efficacy in the application of current technology. It is possible that age-and diagnosis-adjusted mortality would have improved over the past decade, as has happened with other medical conditions, had reimbursement not been eroded by reductions and by inflation. Even if prior reductions in reimbursement had had no effect, it does not follow logically that further decreases will not increase mortality. We may be at the edge of a "slippery slope," beyond which further cuts will have large effects because the limit of ability of providers to absorb the effects of decreased payment without dangerously eroding quality has been reached. Because dialysis is a life-sustaining treatment, the committee concluded that it must give some weight even to imperfect data that point in the direction of adverse effects.
Taken together, the suggestive evidence deserves attention because all results point in the same direction (decreased reimbursement may have eroded quality). Changes in quality are related temporally to changes in reimbursement, and the underlying behavioral and physiological hypotheses are plausible. Although none of these studies constitute conclusive proof of the adverse effects of prior reductions in reimbursement on patient outcomes, none rule out such an effect. None suggest that such reductions are contributing to improved quality of care.
Outpatient Dialysis Reimbursement Issues
The committee addressed the implications for dialysis reimbursement policy of the information about (1) the effect of reimbursement on quality of care and (2) future needs in the light of projections of growth in the size and diversity of the dialysis patient population over the next decade. It did so (in Chapter 11) with respect to covered services, the rate-setting process, and payment policy for facilities and physicians. Transplant reimbursement issues are treated in Chapter 9.
Two broad approaches to reimbursement policy were considered, which may be termed the "implicit" and the "explicit" approaches. Historical and current policy has taken the implicit approach. Since the inception of the ESRD program, Medicare has defined the set of covered services as "all necessary services" required for dialysis and has listed the items that are included within this framework (as well as separately billable services). These services implicitly reflect what providers do in the dialysis procedure. Rates are determined by what providers actually spend to deliver these dialysis services to patients.
This approach minimizes governmental involvement in clinical practice. Providers are free to use their judgment in deciding how to achieve quality within the limits of their reimbursement rate. Historically, this implicit rate-setting process succeeded in financing adequate care for patients, fostered progressive reduction in costs, promoted technical innovations, and developed a climate in which dialysis capacity grew appropriately to meet increasing patient need.
However, the committee came to believe on the basis of the data assembled in this report that the current reimbursement system needs modification if it is to serve dialysis patients and society well in the future. Concern about the implicit rate-setting method arose primarily because it makes no connection between reimbursement and quality.
In particular, covered services have changed over time without change in reimbursement: some services that were billed separately before 1983 have
since been incorporated into the composite rate; in addition, the composite rate now covers some services, such as the labor-intensive administration of complex medications, that were nonexistent in 1983. Although covered services have changed, no explicit analysis has been made to determine how reduced payment levels have influenced what providers do or whether the outcomes of care are altered. Nor have proposed reimbursement changes been made with a prior assessment of their likely quality impacts.
The committee's concern is that implicit rate-setting may have generated a vicious cycle in which rate reductions lead to unmonitored system-wide quality reductions; the lower cost of these inferior services then constitutes the basis in the rate-setting process for a cut in reimbursement.
Two policy options were considered by the committee. First, continue to rely on the implicit definition of covered services that is used by HCFA, i.e., defining covered services in terms of what providers actually furnish to dialysis patients.6 Second, as some providers recommend, develop a more explicit definition of covered services. Absent data on the effects of reimbursement on patient outcomes to guide an explicit approach, the committee favors continued reliance on the implicit standard. However, the committee is concerned that the services currently provided reflect not only professional judgments but also the effects of prior reductions in reimbursement that may have reduced services below desirable levels. Therefore, it believes that a technical advisory committee should examine covered services on a continuing basis and make judgments about the elements needed for appropriate care and outcomes, ideally based on patient outcomes research.
Rate-Setting and Payment Policy
ESRD providers operate in a system dominated by Medicare. HCFA determines the services that are to be included in the reimbursement rates, the reimbursement rates for these services, and the corresponding audited or allowable "costs" of providing these services. Allowable costs, in turn, provide the basis for rate revision.
Nominally, reimbursement rates for dialysis are based on the audited costs of providers. In fact, the rate-setting process for dialysis facilities, unlike that for the hospital sector under prospective payment, has made no provision for updating rates to reflect the effects of inflation on costs during the time between the year when cost data are collected and the adoption of a new rate. The current rate is based on 1977–79 cost data and has never been updated for inflation.
In principle, rates should be recalculated (rebased) periodically using timely data on audited costs. Rebasing, in effect, adjusts reimbursement to reflect the cost of the services providers actually furnish patients. As noted above, the committee believes that current expenditure levels by providers reflect not only
their professional judgment of services needed for adequate care of dialysis patients but also the effects of prior reductions in reimbursement that may have reduced services below desirable levels. Therefore, it seems imprudent to rebase payment now using recent audited provider cost data.
The committee proposes two stages for reimbursement policy. First, for the immediate future, rates should not be decreased. Moreover, until a quality assessment program is in place, annual rates should be adjusted for inflation by some appropriate factor. During this initial phase of the proposed rate-setting policy, the committee believes that a quality assessment and assurance program should be implemented. It notes the research opportunity inherent in the data showing wide variations in costs among dialysis units. Analysis of the relation between cost and quality may reveal that the least costly units have maintained quality or, on the contrary, that low costs have been achieved only at the price of unacceptable decreases in quality.
Second, once a quality assessment program is in place, HCFA may wish to revert to an "implicit" rate-setting process. Should that process lead to rate reductions, HCFA will be in a position to monitor its effects in patients. The committee discussed the alternative of an "explicit" rate-setting process that would specify a detailed "bundle of services" needed for the adequate care of dialysis patients. The sum of the prices of these individual components would constitute the basis for rate-setting. However, the committee decided against this alternative because it would unduly involve government agencies in setting detailed and potentially rigid standards of clinical practice. On the other hand, the committee felt strongly that some mechanism for ongoing review of rate-setting was appropriate. It recommends, therefore, that ProPAC, which has been mandated to review ambulatory care, also review rate-setting in the ESRD program.7 In addition, the committee recommends that HCFA establish an expert committee to advise it on potential additions to the "bundle of services" needed for dialysis patients as innovations arise and clinical practice changes.
Given a patient population of increasing complexity, suggestive but not conclusive evidence of the erosion of quality, the absence of a system to monitor quality, and the life-saving nature of dialysis treatment,
the committee recommends8that Congress and HCFA adopt the following payment policies for dialysis facilities:
Do not reduce the composite rate at this time.
Do not, at present, rebase (recalculate) the rate using recent cost report data because there is reason to believe that current costs reflect prior payment reductions rather than provider decisions about the services needed for appropriate medical care.
Follow general Medicare payment policies in setting dialysis payment policies:
Update the rate yearly, as is the practice for the rest of Medicare.
Rebase the rate only when HCFA rebases the other parts of Medicare governed by the Prospective Payment System.
Ultimately, predicate rebasing of outpatient dialysis reimbursement on efficacy and quality studies that determine the components needed for appropriate dialysis care. Because these components will change as clinical and technical knowledge advance, HCFA should establish an expert advisory body to review periodically the services that Medicare should reimburse.
The Prospective Payment Assessment Commission, consistent with its recently expanded charge to examine Medicare outpatient as well as inpatient reimbursement, should periodically review ESRD payment policy.
Adopt the following specific ESRD reimbursement policies:
Evaluate the justification for the rate differential between hospital-based and independent facilities, especially in terms of patient complexity, and retain or eliminate the differential based on that analysis.
Establish a separate rate for hospital backup units that treat both inpatients and outpatients and that provide support to independent units in the care of complex outpatient cases.
Establish a separate rate for ESRD pediatric patients.
Evaluate the need for a separate rate for rural facilities.
The committee also recommends that HCFA review the monthly capitation physician payment policy (MCP) in light of its exclusion from the Medicare Fee Schedule regarding the impact of this policy on the quality of care provided.
QUALITY ASSESSMENT AND ASSURANCE
In all areas of medicine, there is increasing emphasis on the need for effective systems for quality assessment and assurance (QA). This is especially true for the ESRD program, which provides life-extending treatment for a large group of patients at high cost. Both patients and policy would benefit from systematic evaluation of the quality of ESRD treatment. Although a number of quality indicators are widely used, no systematic guidelines for evaluating quality have been developed for the ESRD program. Moreover, support for quality assessment research in this area has been very limited.
The ESRD QA function within Medicare involves several elements. Conditions of coverage for ESRD providers, published as regulations in 1976, represent traditional quality criteria and standards. However, these have never been related to outcomes or to all relevant process measures.
The HCFA-funded state surveys of dialysis facilities have potential for effective quality assurance, especially for structural and some process measures of quality. However, this system at present is demoralized by poor support, inadequate training of surveyors, and inconsistent oversight. Its surveys vary so widely in approach, thoroughness, and fairness that they are not accepted as valid by clinicians. State surveys could be an important component, but not the centerpiece, of an integrated ESRD QA system.
The committee recommends that HCFA improve the Medicare ESRD state survey system by developing uniform training and certification requirements for surveyors and integrating the state survey system with other ESRD QA efforts.
The HCFA Health Standards and Quality Bureau (HSQB), working with the ESRD networks, is currently (1990-91) testing the ''National Medical Review Criteria Screens'' and "Medical Case Review Procedures" to implement certain provisions of OBRA 1986 regarding the quality and appropriateness of patient care. Although the HSQB effort is creditable, it has certain technical limitations, is costly, and implicitly assumes that quality is best assured by a top-down monitoring system.
The committee believes that better methods are available for QA. These include integrated clinical indicators of outcomes and process and the use of functional and health status outcome measures. The report includes an example of the former as applied to the treatment of anemia. Also discussed is the use of continuous quality improvement, which focuses on improving patient management at the treatment-unit level. The report presents three examples of providers using some of these modern QA methods.
The committee also believes that justifiable HCFA QA oversight of ESRD providers could be exercised by coordination within HCFA of all relevant bureaus, linking of existing data bases, and the systematic examination of the relations between cost of treatment and quality of care.
The committee recommends that HCFA:
Evaluate all policies, including reimbursement policies, for their quality impacts on patients.
Provide adequate financial support to facilities for QA by incorporating facility QA costs in reimbursement for both dialysis and transplantation.
Coordinate within the HCFA the efforts of Health Standards and Quality Bureau, Bureau of Policy Development, and Office of Research and Demonstrations; link existing data bases for the development and operation of ESRD QA oversight systems to relate cost of treatment and quality of care; and integrate the ESRD networks and state surveys in a coherent national QA strategy.
Establish an advisory group of nephrology professionals and experts in QA to design and develop ESRD-specific QA systems.
Support the regional and national data systems necessary for an effective QA system.
Support a continuing program of ESRD QA research.
The data available for evaluating the ESRD program and for monitoring epidemiology, access, quality, and reimbursement effects on quality are good in some respects and limited in others. The committee considered three major data systems: the HCFA ESRD Program Management and Medical Information System (PMMIS), the United States Renal Data System (USRDS), and the United Network for Organ Sharing (UNOS) data system.
In general, data about the ESRD program are currently more available and of higher quality than ever before, reflecting a decade of substantial accomplishment. HCFA is to be commended for its efforts, National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) for establishing the USRDS, and UNOS for its new data system.
In the coming decade, HCFA should address itself to these needs: improving data on primary diagnosis of renal failure, including data on insulin-dependent and non-insulin-dependent diabetes mellitus; developing overall measures of patient complexity, including the extent and severity of comorbid conditions; developing measures of functional and health status and other short-term outcomes; collecting information on ESRD patients who are not eligible for Medicare coverage; obtaining better ethnic data on ESRD patients, including those of Hispanic origin; and measuring patient mortality in the first three months of treatment (before Medicare eligibility is established) for patients under 65 years of age. A critical need is for HCFA to link its existing data systems in order to examine the relations among resources, treatment, and patient outcomes.
The committee recommends that HCFA maintain a strong ESRD data acquisition and analysis capability; modify the acquisition of patient information to include the periodic collection of data on comorbid conditions, risk factors, functional status, and other factors bearing on treatment and outcomes; identify data needs for systematically analyzing the relations among resource use, treatment, and patient outcomes; fund research pertinent to these questions; and provide public use tapes for research on all aspects of the ESRD program to the interested research community.
The decision by the NIDDK to develop the epidemiology of kidney disease as an integral part of its research efforts is a welcome step, one taken earlier
by the National Institutes of Health dealing with cancer, heart disease, and aging. The NIDDK establishment of the USRDS represents an extremely important initiative in this context. The committee commends NIDDK for this action.
The committee recommends that USRDS be authorized to conduct research linking epidemiologic and economic data on the ESRD patient population and that the special studies approach be exploited to its full potential.
The National End-Stage Renal Disease Registry was authorized by Congress in OBRA 1986. Its mission was (1) to collect and analyze data pertaining to the epidemiology of kidney disease, (2) to conduct cost and cost-effectiveness analyses of treatment modalities, and (3) to analyze trends in mortality and morbidity and develop other indices of quality. The committee finds that the first function is being performed by USRDS, the second function has been the subject of bureaucratic controversy and is not yet implemented, and the third is being implemented mainly in regard to mortality.
The committee recommends that the Secretary of Health and Human Services take all steps necessary to fully implement all functions of the National End-Stage Renal Disease Registry called for in OBRA 1986 within the next two years.
It also recommends that the renal data system be continued another five years when the current USRDS contract expires and that the contract be recompeted at that time.
The committee believes that a comprehensive program of basic and applied research related to end-stage renal disease is essential. This includes basic research on the diseases that cause ESRD, clinical studies, preventive and epidemiological research, and health services research. Basic research holds great potential for reducing the frequency and severity of ESRD, with both human and economic benefits. At present there is great need for clinical research on dialysis and transplantation, which has the potential to improve patient quality of life and to reduce costs by creating less expensive technology and techniques. Preventive and epidemiologic research also has high clinical and economic potential.
The committee recommends that:
NIDDK expand its support for basic research having the promise to prevent or to reduce the progression of kidney disease that leads to ESRD.
Clinical research on dialysis, supported by NIDDK or HCFA, be resumed.
Epidemiologic research be supported to identify risk factors for ESRD and its clinical outcomes, the value of various interventions to prevent and treat ESRD, and factors influencing access to ESRD care. Special attention should be given to racial and ethnic differences in the incidence and causes of renal failure and in transplantation (access, outcomes, organ donation).
Health services research be supported to determine which components of medical care lead to better outcomes and to identify and validate quality measures for structure, process, and outcomes.
Section 4036(d), Omnibus Budget Reconciliation Act of 1987 (Public Law 100–203), December 19, 1987.
The study was supported by the Health Care Financing Administration (Cooperative Agreement No. 14-C-99338/3-02).
Insured status is measured by "quarters of coverage." Fully insured status requires a minimum of 6 quarters of covered employment; 40 quarters of coverage grants permanent status. Currently insured status requires 6 quarters of covered employment in the past three years.
State and local government employees have been covered by Social Security on a voluntary basis since 1950 but on a mandatory basis only since 1984.
A dialysis station consists of a dialysis machine, a chair or bed, and the associated floor space required to dialyze a single patient.
Indeed, some members of the committee believe this is the preferable method in any case because no expert committee can equal or improve on the patterns of services revealed by provider actions.
The committee discussed this recommendation during 1989 and 1990. In the Omnibus Budget Reconciliation Act of 1990 [Public Law 101-508, Section 4201 (b)], Congress directed ProPAC to "conduct a study to determine the costs and services and profits associated with various modalities of dialysis services provided to end stage renal disease patients" under Medicare. ProPAC was also directed to consider the conclusions and recommendations of this study.
One committee member recommends a different payment policy; this is described in Chapter 11.
Held PJ, Bovbjerg RR, Pauly MV, Garcia JR, Newmann JM. 1987. Effects of the 1983 "Composite Rate" Changes on ESRD Patients, Providers, and Spending. Washington, D.C.: Urban Institute, December 21.
Held PJ, Garcia JR, Wolfe RA, Gaylin DS, Pauly MV, Cahn MA. 1990. Price of dialysis and hospitalization. Paper prepared for the Institute of Medicine ESRD Study Committee. Washington, D.C.: Urban Institute, June 19.