President's Note: End stage renal disease (ESRD) was selected as the focus of this issue of the Monitor because of the long-standing attention among nephrologists to the disabling decrements in functional capacity in ESRD. This concern has resulted in the development of many functional status assessment instruments and their application in numerous clinical trials. Our original intent was to interview Klemens Meyer, M.D. for the feature article and the interview. Dr. Meyer is Director of Dialysis Services at the New England Medical Center and Associate Professor of Medicine at Tufts University. An interview was conducted and provided such a wealth of interesting material that we went back to Dr Meyer for further discussions. This article is a synthesis of the interview and the discussions. We are grateful to Dr. Meyer for directing our attention to exciting developments in the field and to the remaining challenges in the management of patients living with ESRD. The Editors of the Monitor bear sole responsibility for the readability and accuracy of the text.

The Problem of End-Stage Renal Disease

Chronic renal failure is the late phase consequence of progressive kidney damage inflicted by a variety of diseases. Chronic renal failure was universally fatal until the advent of kidney transplantation in the 1950s and of chronic dialysis treatment in the 1960s. In the United States Medicare benefits are provided to pay for dialysis or transplantation in individuals who reach "end-stage renal disease", or ESRD. Treatment by dialysis or transplantation is instituted when the patient develops symptoms such as fatigue, nausea and anorexia, when physiologic complications such as fluid overload and elevated blood potassium levels are intractable to drug and diet treatment, or when the physician judges that the long term morbidity of inadequate kidney function outweighs the risks and morbidity of dialysis or transplantation.

At the end of 1997, there were 304,083 ESRD patients in the United States, 1,105 per million Americans1. In about a third of these patients, kidney failure was ascribed to diabetes, in just over a fifth to hypertension, and in about a sixth to glomerulonephritis. Minorities and the elderly are disproportionately affected. The prevalence of ESRD among African-Americans was more than triple, and that among Native Americans more than double the prevalence in the population at large. Similarly, the prevalence among those 65 years of age and over was more than triple the overall prevalence. Half of the patients beginning ESRD treatment are older than 65. The ESRD population is currently growing at 7%/year. It has doubled in a decade.

Physiologic status, functional status and well-being, access to appropriate services, satisfaction with services, and the cost of services are useful outcomes measures in the management of all chronic diseases. ESRD drastically affects how patients feel and perform in almost all dimensions of their lives. ESRD sharply reduces physical functioning and work performance, substantially lowers perception of one's own health, deprives the patient of energy and vitality, reduces or limits social interactions, and can erode mental health. ESRD can be debilitating, confining and isolating, and can deprive the afflicted of participation in the things that are most meaningful to him/her.

In the context of these profound personal and social limitations, functional status and well-being are increasingly being recognized as key parameters in monitoring and guidance of treatment modifications in this condition. For ESRD patients, essential treatment decisions have historically been made predominantly based on physiological measurements that do not necessarily reflect functional capacity in everyday living. Treatment goals should aim not only at normalizing serum potassium and creatinine but also with equal priority at improving quality of life or minimizing its decline.

Treatments and Choices

Dialysis and transplantation have prolonged the life expectancy of all patients with advanced renal failure, and made the timing of the fatal prognosis less predictable. For example, over half of the best-matched kidney transplants are still working after twenty-five years; patient survival is even longer for those patients. Dialysis patient survival can also exceed two decades for some individuals. But for the ESRD population as a whole, average life expectancy from the time of onset of renal failure is only 4 to 5 years. For most patients renal failure remains a terminal illness.

ESRD is important not only because it is common, but also because of the extraordinary limitations it imposes, the remarkable effectiveness of dialysis and transplantation, and the extraordinary costs of these treatments. The treatments of ESRD entail extraordinary intrusion into the patient's life in order to prolong that life. Of the 304,083 ESRD patients at the end of 1997 only 86,371 had working kidney transplants, while 221,596 were treated by dialysis 1. Dialysis represents the first long-term organ replacement therapy, the first real instance of outpatient "life support."

Despite thrice weekly hemodialysis treatments or daily peritoneal dialysis treatments, dialysis does not restore full health. There is a prevalent sense among the general public that life on dialysis is very difficult, that it is painful and depressing, that dialysis patients are weak and tired, and that for very old and debilitated patients perhaps life on dialysis is not worth living. "What's the quality of life? I'd never want to live like that, hooked up to a machine."

When chronic dialysis was first introduced in the 1960s, the extreme nature of the treatment prompted investigation of its emotional impact on the patients. There is now an annual meeting on "psychonephrology." Psychosocial counselling is an integral part of ESRD treatment, although in the U.S. the social worker's counselling role has in most settings been progressively diluted by administrative roles and financial constraints. Patients often consider withdrawing from treatment, thinking that death might be a preferable option. Approximately 20% of patients on dialysis elect to discontinue the treatment, preferring death1, p.98.

ESRD treatment involves making choices among options, and these options entail real tradeoffs. First, there is the decision whether to take on the burdens of treatment to achieve longer survival. This decision depends in turn on what those burdens are and how they can be expected to affect the person. For younger patients this may be straight-forward. But older individuals might interpret renal failure as a natural conclusion to their lives. Second, there is the timing of treatment. Should one have a pre-emptive transplant to avoid dialysis altogether? If so, at what level of kidney function? (How much longer could one have waited?) Should one begin dialysis only when there are intractable physiologic and metabolic complications? Alternatively, will an early or "healthy" beginning of dialysis prolong survival? How does one balance the psychological loss and the morbidity of earlier treatment against longer survival? How does one choose between hemodialysis and peritoneal dialysis? One may face choices about dialysis dosing. For many patients, longer hemodialysis treatments and more frequent or larger infusions of peritoneal dialysate will prolong survival. But at what discomfort, at the cost of what change in life-style? What will the cost be in the quality of the patient's life?

Studying physicians' and patients' choices regarding hemo versus peritoneal dialysis and dose is the purpose of the CHOICE study, a patient outcomes research team (PORT) on dialysis care2. One of the team's projects is a cohort study of incident dialysis patients at Dialysis Clinic, Inc. (DCI) and at New Haven Continuous Ambulatory Peritoneal Dialysis and St. Raphael's Hospital in New Haven, CT. To examine the choice between hemodialysis and peritoneal dialysis, the team constructed a dialysis-specific questionnaire, the Choice Health Experience Questionnaire, or CHEQ3. The CHEQ uses the Short Form-36 Health Survey (SF-36) as a generic core, adding questions on issues important to dialysis patients. The additional questions were chosen by a rigorous process, beginning with a structured review of prior instruments, and including focus groups among both patients and providers, importance-frequency ranking of candidate items, and pilot testing3.

Patients who have received kidney transplants almost universally report that they feel better than they did when on dialysis. However, transplantation does not restore full health either. The transplant is a treatment, not a cure; the patient is freed from the constraints on activities imposed by the dialysis treatment, and usually feels better. But there are the new problems of immunosuppression and vulnerability to infections, and steroid complications that can result in salt retention, weight gain, and hypertension.

In the late 1980s, an approximately 25% annual mortality rate in ESRD prompted scrutiny of the quality of ESRD care in the United States. A panel convened by the Institute of Medicine conducted a congressionally mandated review of the ESRD program. At about the same time maturation of the new field of health services research led to the development of techniques for obtaining directly from patients their perceptions of their health and well-being and their functional capacity in the home, neighborhood and workplace4,5. This made possible among patients having ESRD quantitative assessment of their health-related quality-of-life and their outcomes from treatment.

Functional Status and Well-Being: Measures and Use

As the ESRD population grew, a series of investigators used a variety of psychometric instruments to define various aspects of patient experience in the context of research studies. Evans6, Kimmel7, and Shidler8 applied generic instruments to renal failure patients. Laupacis9 and Parfrey10 developed ESRD-specific tools. The instruments applied to ESRD patients were reviewed in 1996 by Edgell11.

Three scientifically and technically driven developments in the care of patients on chronic long-term dialysis in the 1980s and early 1990s brought a surge of interest in the use of valid measures to assess changes in patient-reported health-related quality-of-life before and after the application of new treatment innovations.

The first was the development of erythropoictin. Correction of severe anemia has been shown to have clear effects on health status (Beusterien12). The effects represent a measurement of the contribution of anemia to the uremic syndrome. More recently, a study by Basarab13 in patients receiving hemodialysis but who also had cardiac disease compared hematocrit 42% to 30%. The study was stopped early because of excess mortality in the higher hematocrit group. However, the study also showed that the higher hematocrit level was associated with a 7.2 point enhancement of scores on the physical functioning scale.

The second development was the emergence of evidence that technical changes in hemodialysis treatment might reduce morbidity and mortality. The Hemodialysis Study (HEMO)14,15 is a 15 center 2x2 factorial study of the effects of two levels of hemodialysis doses and two levels of membrane flux on morbidity and mortality. Quality of life assessment using the Index of Well Being and the long form version of the Kidney Disease Quality of Life Questionnaire (KDQOL) is performed at baseline and annually. Health status will be a covariate in analysis of the study's primary outcomes and changes in quality of life during the study will be treated as secondary outcomes. KDQOL, developed at the RAND Corporation with sponsorship from Amgen, is a renal-failure specific extension of the SF-3616.

The third recent impetus to quality-of-life measurement in ESRD has been the finding that among some patients peritoneal dialysis appears to be associated with shorter survival than hemodialysis. Many people suspect that the issue is dialysis dose, and that with a sufficient dose of peritoneal dialysis the two therapies would be equivalent. The dose of peritoneal dialysis is increased by putting more fluid in the patient's abdomen at a time, or exchanging the fluid more frequently. There are two quality-of-life issues here. The first is that if the shorter survival among peritoneal dialysis patients is not related to dialysis dose, and even with sufficient dose peritoneal dialysis survival is not equivalent to hemodialysis survival, there will be a tradeoff of life duration for life quality for those patients for whom peritoneal dialysis is associated with a higher quality-of-life than hemodialysis. The second point is that if peritoneal dialysis dose is the issue, there will be tradeoffs around convenience and comfort.

Dialysis Clinic Incorporated (DCI), a chain of 140 not-for-profit dialysis centers, took the lead in the routine measurement of functional status and outcomes in their patients. In 1990, with help from the New England Medical Center's Health Status Monitoring and Management Program, DCI began quarterly SF-36 administrations as part of routine clinical care in its Boston unit. Beginning in 1995, DCI implemented a similar program across its national chain. Patients at all units now respond to the SF-36 two to four times a year. The results are used to identify changes in the patients' functioning and well-being and to reassess the treatment program accordingly, and to be alert to the appearance of complications or the emergence of a second disease. Follow-up duration on some of the original patients approaches a decade17.

A second early participant in dialysis health status assessment was the Centers for Dialysis Care (CDC) in Cleveland, which also instituted regular SF-36 administration18. Their data show a number of important clinical correlations with self-reported health status. For example, physical component summary scores below the median (<34) were twice as likely to die and 1.5 times more likely to be hospitalized compared to patients whose PCS scores were above the median.

The Independent Dialysis Foundation (Sadler19) contributed to the functional health status assessment movement by developing and pilot testing educational material to support the use of the Dartmouth COOP charts among dialysis patients. The COOP charts have the obvious attraction of simplicity and brevity. This is an important consideration among elderly and very ill patients, many of whom have poor vision, and some of whom may not be able to follow the SF-36. The follow-up educational materials that accompany the COOP charts are also important, because they reduce the burden on staff of interpreting the results for patients.

The DCI Outcomes Monitoring Program has yielded a number of insights, Dr. Meyer said. First, the changes in health status reported by ESRD patients are frequently so large that they exceed the confidence intervals around previous responses. Thus, even though the SF-36 is a short instrument designed for use at the group level, it can be informative, at least in this group, at the individual level. Second, changes in health status sometimes correlate with clinical events already observed by the dialysis team, but not always. That is, measurement of health status assessment not infrequently brings to clinical attention aspects of patient experience that had been overlooked or underestimated. Third, systematic health status assessment extends the formal history, and improves on casual conversation as a source of information about patients' functioning and well-being. Observations about an individual's health status are most valuable if serial measurements are taken. Finally, DCI's data collection has allowed the definition of health status norms for dialysis patients20.

Formal quality-of-life assessment has brought a new precision and specificity to the language of ESRD. Dr. Klemens Meyer has said, "We have always known that patients with ESRD were very sick, that they have kidney failure, and that they often suffer life-threatening elevations of the serum potassium. In the past the level of kidney function could be specified precisely. We knew exactly what the serum potassium and creatinine were. But we did not have a way of expressing, reproducibly and unambiguously how sick our patients were, and in what specific ways they were and were not sick. We can now begin to give an objective and circumstantial account of the patient's experience. We can differentiate one sick patient from another, and one aspect of an individual patient's experience from another aspect. We can make a reproducible record of these observations. We can say something much more meaningful about what their life is like. We can say that for most patients, ESRD is not accompanied by much more pain than the general population experiences, or much more depression. On the other hand, limitations on physical functioning and role limitations imposed by physical health are typically substantial. This precision represents an important advance."

Federally sponsored activities have clearly been important to the professional community's recognition of quality of life as a measurable outcome of renal failure and its treatment. The SF-36 is also being used by the African American Study of Kidney Disease (AASK)21 , and in special studies by the United States Renal Data System (USRDS)1. AASK is a multicenter clinical research trial of hypertensive renal disease in African-Americans. Because the KDQOL being used in the HEMO Study contains the SF-36 questions, the combination of AASK and HEMO data sets hold promise of gaining a more complete understanding of the long-term effect of progressive kidney failure on health status21.

The dialysis industry has also played a major role. In the mid-1990s, representatives of DCI, of the Independent Dialysis Foundation, the Northwest Kidney Center and of Centers for Dialysis Care (CDC) in Cleveland, four of the largest not-for-profit dialysis providers in the United States, participated in the ESRD Health Status Outcomes Group, which was convened to continue work begun at two conferences sponsored by the Institute of Medicine (Rettig22, and Sadler23).

A review of the use of functional status measurement in ESRD has recently been completed by Curtin24.

In summarizing the state of quality-of-life assessment in ESRD, the most important concept is technology. It is change in the technology of quality-of-life assessment that has brought the field to this point. Short forms with a pedigree of rigorous scientific standards have been essential. Brevity is essential to maintaining patient interest, or at least cooperation. Rigor in measurement, reliability, validity and precision has been essential to win and maintain physician tolerance for what many still perceive as very soft science. Easy scoring and reporting have been essential to make possible repeated administration and clinical use of the information. Population norms and ESRD specific norms have added greatly to interpretability. Standardization of instruments, scoring rules, and interpretation greatly facilitate comparison across centers and regions.

REMAINING CHALLENGES

Major issues for the future include 1) integration of the quality-of-life data into the electronic medical record, 2) the development of data displays that allow clinicians easily to recognize temporal relationships of patient scores to treatment variations and physiologic tests, 3) making quality-of-life assessment generally accessible to poorly educated and illiterate or marginally literate patients, 4) collection of far more data about patients' courses as they approach renal failure, and about the course of quality-of-life changes extending from transplantation to graft failure or death; 5) flexible administration, synchronized to the patient's own history rather than to an institutional schedule; 6) dynamic questionnaires using item response theory and computer adaptation which promise to add sufficient precision required for even further applications in individual patient monitoring. Dynamic technology will also permit a more refined assessment of patients functioning when their scores are at the floor or ceiling of the measurement scale.

There are also challenges that technology cannot meet. Physicians, nurses, pharmacists, rehabilitation experts and other clinicians should be taught more about the science of instrument development, validation, and use. Ignorance is not the clinician's fault; this body of knowledge is not readily accessible, and with some exceptions, its exposition is not easily understood. We need to figure out what disease-specific items mean, and how to use them. It has been virtually an article of faith that generic instruments alone were not enough. In kidney failure, the symptoms of which are so pervasive, generic questionnaires may in fact capture morbidity very adequately. However, it is also possible that as more observations are collected, the value of renal failure-specific items will become apparent.

In conclusion, outcomes measures have the potential to elevate the effectiveness of clinical management of patients with ESRD, perhaps substantially.