Kidney International, Vol. 57, Suppl. 74 (2000), pp. S-113–S-120

QUALITY OF LIFE

Just whose quality of life is it anyway? Controversies andconsistencies in measurements of quality of life

PAUL L. KIMMEL

Division of Renal Diseases and Hypertension, Department of Medicine, George Washington University Medical Center,Washington DC, and Division of Kidney, Urologic and Hematologic Diseases, National Institute of Diabetes, andDigestive and Kidney Diseases, National Institutes of Health, Bethesda, Maryland, USA

Just whose quality of life is it anyway? Controversies and or differences between levels of coping in patients under-consistencies in measurements of quality of life. The proper going different medical treatments. Recent work has fo-means of measuring quality of life (QOL) in patients with renal cussed on using measures of QOL as outcomes in them-disease, including end-stage renal disease, have not been agreed

selves, as well as predictors of other outcomes. Orupon. QOL has health-related and non-health-related do-clinicians and providers may wish to improve patients’mains. QOL measures may be subjective or objective, function-

based or satisfaction-based, non-specific or applied to a particu- evaluations of their QOL. Interestingly, little attentionlar disease. Domains of health-related QOL typically include has been devoted to the development of interventionsmeasures of physical functioning, mental health, and social

designed primarily to improve perception of patients’relationships. There is increasing interest in measures of happi-QOL.ness or global satisfaction in patients and in assessing how

patients’ feelings reflect their relationships with caregivers. An additional issue has been the measurement ofThere are advantages and disadvantages to the use of older QOL in patients with chronic illness, rather than in gen-psychologically based measures as QOL indicators as well as eral populations. Original descriptions of QOL were de-newer questionnaires such as the Kidney Disease Quality of

veloped in an attempt to categorize the US experienceLife (KDQOL) Instrument. Challenges for the future includeas perceived by different groups of people, at differentcorrelation of QOL measures with patient biochemical and

disease measures, and with meaningful patient outcomes. Most stages of life [2]. Notions of QOL have extended fromimportant, we must create and apply methods to enhance our a general but comprehensive focus, to one related to thepatients’ QOL. experience of illness, sometimes categorized as Health-

related Quality of Life (HRQOL). This review will focuson some of the most common QOL evaluators used in

The conceptual and theoretical bases for, and the patients with end-stage renal disease (ESRD) treatedproper means of measuring, quality of life (QOL) in with chronic maintenance hemodialysis (HD). It shouldchronically ill patients remain incompletely defined [1]. not be taken, however, as comprehensive, and only aAlthough it has generally been agreed that such mea- subset of the large literature will be directly cited. Severalsures must assess several and diverse aspects of peoples’ previous reviews however, provide a comprehensive bib-experience, the practical issues regarding evaluation of liography through 1997 (cited in [3]). I will concentratepatients’ QOL remain controversial and undetermined. on aspects of the commonly used measures to assessThe use of such measures may also be quite variable. QOL in HD patients. I will not review in depth theClinicians may desire these assessments to gauge the literature on erythropoietin and assessment of QOL, aspresent status of a patient or patients. Payor organiza- this will be touched on by another contributor. I willtions may wish to evaluate patients’ satisfaction with make a case for the use of multiple, well developed scalescare, or differences in levels of perceived satisfaction

to capture psychologically validated and generalizablebetween patients treated in several medical organiza-

aspects of patients’ experiences, rather than the use oftions. Researchers may wish to evaluate patients’ percep-

disease-specific scales to standardize measurements oftions of the effects of a therapy. Physicians may want toHD patients’ QOL.quantify the adjustment of patients to a medical regimen,

In the United States, during the 1960s and early 1970s,chronic dialysis was perceived as a heroic but intrusivetechnological advance that allowed patients otherwiseKey words: hemodialysis, chronic kidney disease, depression, social

support, effects of illness. doomed to death to survive. Clinicians, policymakers,and patients could agree that whatever the problems 2000 by the International Society of Nephrology

S-113

Kimmel: Quality of life in HD patientsS-114

Table 3. Quality of life categoriesTable 1. Types of quality of life measures

Domains Satisfaction/functionObjective vs. SubjectiveFunction-based vs. Satisfaction-based Health related quality of lifeDisease-based vs. Non-specific Functional status F

Intrusive effects of illness S/FAffective functioning S/F

Social functioning F/STable 2. Domains of quality of life Global satisfaction S

Employment FPhysical functioning Leisure F/SMental health Marital satisfaction F/S

General affect (mood) Sexuality F/SPerception of well-being (illness effects) Spirituality S/F?Life satisfaction (happiness)

Social relationships

review of assessment of QOL in patients with ESRDwith the treatment, it provided a clear and attractive categorized QOL into measurements of functional sta-alternative to the inevitable. After the enactment of fed- tus, health status, well-being and patient satisfaction [3].eral legislation making ESRD therapy available on a Original assessments of QOL in normative populationswider basis, the number of patients in the U.S. ESRD included evaluations of housing conditions, employmentprogram rapidly burgeoned [4]. Pioneering studies of status, education and finances [2]. Although these areQOL of patients treated with different modalities of often considered measures of socioeconomic status, suchrenal replacement therapies by Simmons et al [5] deline- issues can impact upon patients assessments of QOL,ated that most patients had perceptions of QOL that and perhaps their HRQOL.were comparable with, if perhaps lower than normative

In previous studies, our group in Washington DCpopulations, but that a functioning renal transplant was(composed of George Washington University Medicalassociated with perception of a higher QOL than in theCenter, Howard University Medical Center and the Vet-general population. Conversely, a failed transplant waserans Affairs Medical Center) has favored the use of aassociated with worsened perceptions of QOL. ESRDwider gamut of measures, structuring the domains ofcare over the past decades has been extended from pa-HRQOL along several dimensions that have beentients with relatively few comorbid illnesses to an aginggauged by psychological tests that are in wide use incohort of patients in which almost half the membersboth general and ill populations (Table 3) [11].have diabetes mellitus [4]. As renal replacement therapy

It is important to start any discussion of QOL assess-has developed [6] and the patient population has ex-ments with a definition of terms. These definitions arepanded, questions regarding the relationship of type,taken in part (with slight modifications) from the usefulquality and duration of treatment to outcomes, including

survival and patient satisfaction, have been raised [3]. review of Gill and Feinstein [1]. An item is a singleConcern regarding the demographics, costs and mortal- question used to evaluate QOL. A scale is a way in whichity in the program, coupled with a perception that the to present the responses to questions. A domain is theelderly were not experiencing consonant benefits [7, 8], focus of attention of an item (such as happiness, or func-led to renewed and invigorated interest in the assessment tional capacity). An instrument or index is a collectionof QOL in patients with ESRD. of items, which can vary in number from one to many.

Questions that are fundamental to issues to relevance Instrument components may be cited individually, asof HRQOL measures include: 1) Does the way a person profiles or subscales; or, components can be cited in ag-feels about him- or herself, family and friends, or the gregate as a score. Alternatively instrument componentsway the illness affects him or her, have an impact on can be cited in both manners, as subscales and aggregateoutcome in patients with chronic medical illness? and, scores.2) Does the manner in which a person reacts to the

Gill and Feinstein [1] have emphasized the utility ofillness within the medical community have an impactassessing QOL, especially in research studies, by usingupon outcome in patients with chronic medical illness?more than one instrument simultaneously in the sameA plethora of measurement tools of various kinds andinvestigation, in order to provide a more comprehensiveutilities has been used in various populations to assessfocus to the meaning of a particular QOL indicator. TheyQOL [1]. QOL instruments may be objective or subjec-have also advocated the use of a particular instrument,tive, satisfaction-based of function-based, or be disease-possibly composed of a single item, in which the subject isor organ-based rather than non-specific (Table 1) [9].specifically asked to rate his or her perception of currentAnother categorization divides HRQOL into physical,

mental and social dimensions (Table 2) [10]. A recent quality of life, presented using that specific phrase [1, 12].

Kimmel: Quality of life in HD patients S-115

MEASUREMENT TOOLS Low correlations have been delineated between IWBand IGA and Sickness Impact Profile (SIP) Physical Di-Probably the most widely used measure of HRQOLmension and total scores (see below), and other func-in patients with ESRD treated with HD is the Karnofskytional scores, leading some to suggest these measures doPerformance Status Scale ([13–17] and reviewed in [10])not capture the totality of HRQOL [10]. We too haveto quantify an individual’s level of functioning [13]. It isshown that satisfaction with life scores (a global, subjec-composed of 11 qualifying statements, ranging from ative measures of QOL) correlated with advancing agescore of 100 for normal function, to zero for death, withand increased severity of illness, but were not correlatedmarkers for requiring assistance at different levels, need-with Karnofsky scores [11, 21, 22]. Global, subjectiveing custodial care, and needing hospital care. Highersatisfaction with life can be assessed in patients withscores indicate better functional status. Originally de-ESRD, and is related to subjective factors other thansigned as an objective scale for measuring function inobjective physical function as assessed by the medicalpatients receiving cancer chemotherapy, the Karnofskystaff on the Karnofsky scale. Rather, satisfaction with lifeScale has been used in many different disease settings.is significantly related to better subjective evaluations, asIt can be used as a subjective scale, when given to patientsassessed by the patient in a perception of illness scalefor their own assessments of their functional status. We(IEQ) (see below). Specifically, Karnofsky ratings, anhave used this scale in a modified format for subjectiveobjective functional measure of quality of life, correlatedassessments, truncating it at a level of 40 (Disabled; re-with relatively few parameters of patient perception wequires special care and assistance). We have also usedassessed. These results indicate to us the dissociationthe Karnofsky Scale to compare the assessments of pa-of feelings of well-being from functional assessments intients and spouses (cited in [11]), which were different.patients treated with HD [9], and emphasize the impor-Caveats have been raised regarding the reproducibilitytance of individual perception, rather than family or care-and interater reliability of the scores [18]. However,giver assessments in determining functional quality ofmany studies of chronic HD patients have demonstratedlife. Such data lend credence to the recent consensusmean scores in the range of 70–80, indicating a levelforming that the patients’ subjective assessments areof functioning between being able to perform normalmore important in evaluating care than “objective” mea-activity with effort (80) and maintaining the ability tosures [9]. Our data, however, would support the notioncare for oneself but being unable to carry on normalthat elderly patients treated with HD perceive them-activity or do active work (70). Comorbidity scales andselves as having a good quality of life.symptom checklists have also been used to enhance the

Interestingly, several domains of quality of life in-utility and the subjective dimensions of functional assess-crease dramatically with age in the general populationments [10].in the United States [2]. Satisfaction with health status,Campbell et al [2] pioneered the field by surveyingin contrast, decreases with age in the general population,the perception of quality of life regarding numerous do-as might be expected [2]. Our findings in a chronicallymains in large samples of Americans of different agesill population, however, are in agreement with the dataand backgrounds. Measures of general well-being havefrom a “well” population, suggesting that older HD pa-been derived from these scales, and used in patients withtients have greater perceived satisfaction with life regard-ESRD ([4, 11, 16, 19] and reviewed in [10]). The Indexless of the severity of illness, and with the work of Kutnerof General Affect measures how the subject feels aboutet al [23], who showed elderly African American HDlife, ranging from extremes of satisfaction to dissatisfac-patients, in particular, had greater satisfaction with lifetion. The Index of Life Satisfaction is a single item deal-compared with white patients. SWLS scores, regardlessing with satisfaction with life on a scale from 1 to 7. Theof age, in our studies were correlated with depression,Index of Well-being score is derived from a combinationperception of illness, social activities, presence of a rela-of the two scales. High scores signify high satisfaction.tionship and social support.The scores exhibited good psychometric properties in

Depressive affect should not be confused with theESRD patients [10, 19].diagnosis of clinical depression [24]. Measures of de-The Satisfaction with Life Scale (SWLS) [20] is a five-pressive affect are included in many QOL indices (re-item scale with a 1 to 7 satisfaction rating for each item,viewed in [10]). The prevalence of depression has variedwhich we have used extensively [11, 21, 22]. The itemsimpressively between studies of patients with ESRDask about ideal life, conditions of life and satisfactiontreated with HD at different times and in different loca-with present and past life. The scale exhibits good psy-tions [24]. One problem may be the confounding of thechometric properties [20] and correlations of 0.50–0.60somatic characteristics of depression with uremic symp-with other subjective well-being scales. Higher scorestoms [24]. A critical issue with adherents on both sidesindicate higher satisfaction. We have used the SWLS inis whether depression is independently associated withour studies as a general global, subjective quality of life

measure [11, 21, 22]. mortality in patients treated with HD [24].

Kimmel: Quality of life in HD patientsS-116

Probably the most widely used measure of depression and special persons) were averaged to form a total sup-port score (MSP) [11]. In studies in HD patients thesein patients with ESRD treated with HD is the Beck

Depression Inventory (BDI) [10, 24, 25]. Use of the social support scores correlated with other support mea-sures and depression [11].Center for Epidemiologic Studies Depression Scale [10]

has also been advocated in patients with ESRD and other The presence of a committed dyadic relationship andmarital satisfaction have been considered important fac-chronic medical illnesses. The BDI is a well-validated

measure of depression, which assesses somatic and cogni- tors is the assessment of QOL in the general population[2]. They may be in fact most related to the conceptstive aspects of depression, such as fatigue and quality of

sleep function [25]. The BDI is composed of 21 items, of social support and satisfaction with life. However,participating in a dyadic relationship and patients’ per-each scored on a 4-point Likert scale. Total scores can

range from zero to 63. Higher scores indicate greater ceptions of assessments of the quality of the relationshipmay markedly influence quantification of other QOLlevels of depressive effect. The BDI provides cutoffs for

the diagnosis of depression, and therefore can be used domains. Our data suggest dyadic adjustment should beviewed as a social support variable, rather than a qualityto estimate the incidence and prevalence of depressive

symptoms within populations. A score of 11–18 is associ- of life measure. The presence of a relationship correlatedwith social support and weakly with a measure of behav-ated with mild depression, 19–25 with moderate depres-

sion, and 26 or more with severe depression, in patients ioral compliance [11]. How such determinants interactwith treatment parameters and outcomes is a matter ofwithout chronic medical illness. Scores of less than 11

suggest absence of clinical depression. The meaning of great interest. These factors may have quite differentinterrelationships in men and women and in patients ofvery low scores (0–2) is not clear. It is necessary to

validate the results of an instrument measuring de- different ethnic backgrounds.Intrusive effects of illness comprise the patient’s per-pressive effect with a clinical interview or a structured

diagnostic tool such as the Diagnostic Interview Sched- ception of how the disease state interferes with thepatient’s life. Interestingly, two patients with similarule [26], in order to establish the diagnosis of depression

in a patient. However, when assessed in such a rigorous medical characteristics (for example, blind, amputateddiabetic patients with ESRD who have sustained myo-manner, a score of 15 produced excellent sensitivity and

negative predictive values in a study of dialysis patients, cardial infarctions and numerous vascular surgeries) mayhave very different perceptions of illness intrusiveness.distinguishing depressive symptoms from the presence

of a psychiatric disorder [27]. Such findings emphasize the dissociation of measures ofillness intrusiveness from functional status assessments,We derived and have used the Cognitive Depression

Index (CDI) [11, 21, 22, 24, 28, 29], a subset of the BDI, and their possible association with measurements of gen-eral well-being, happiness, depression and social support.which focuses on the thoughts and feelings related to

the diagnosis of depression, such as guilt, disappointment Intrusiveness indices can assess the illness, its treatmentor overall perceptions. Two instruments have been de-and failure, excluding the items regarding somatic func-

tion. The CDI has previously been highly correlated with veloped and used in the majority of studies of patientswith ESRD to assess patients’ perceptions of illness in-the BDI, and perception of illness effects (IEQ) (see

below) in ESRD patients [11, 21, 22). Previously docu- trusiveness.The Illness Intrusiveness Rating Scale (IRS) is a self-mented correlations suggested that the CDI is a better

measure of depression than the BDI in patients with report index which rates the extent to which the illnessinterferes with 13 domains related to QOL [32]. HigherESRD [11].

The Affect Balance Scale has also been frequently scores indicate worsened perception of illness intru-siveness. It has exhibited good psychometric propertiesused to assess depressive effect in patients with ESRD

(reviewed in [10]). It is composed of two scales: the in patients with ESRD [10].The Illness Effects Questionnaire (IEQ) [32, 33] as-Positive and Negative Feelings Scales, and has exhibited

good psychometric properties in studies of patients with sesses the individual’s perceptions of how the subject’sillness interferes with or affects personal, physical, andESRD.

Social support is multidimensional information that an social behavior. It is a subjective, generic instrument.Higher scores reflect a more negative perception. Theindividual is a member of a complex network of affection,

mutual aid, and obligation. The Multidimensional Scale IEQ has high test-retest reliability [32, 33] and high inter-nal reliability [32, 33]. The IEQ is strongly correlatedof Perceived Social Support (MSPSS) is a twelve item

inventory of overall perceived social support, divided with depression in medical patients and patients withESRD [11, 21, 22, 33]. Several groups have advocatedinto three factor groups: social support from a special

person, family, and friends [30]. The MSPSS scale has the use of the IEQ as a QOL measure [33, 34].Patients’ age, severity of illness, and serum albumindemonstrated good internal and test-retest reliability

[30]. In our studies the three subscales (family, friends, concentration did not correlate with our measured scale

Kimmel: Quality of life in HD patients S-117

Table 4. Kidney Disease Quality of Life (KDQOL)of patients’ assessment of the effects of illness, the IEQ.kidney-specific domains

IEQ scores reflecting worsened perceptions of the effectsSymptoms/problemsof the illness were also highly correlated with higherEffects of kidney disease on daily lifelevels of cognitive depression, lower levels of perceivedBurden of kidney disease

social support, and worsened perception of satisfaction Cognitive functionWork statuswith life. Worsened IEQ scores and Karnofsky scoresSexual functionwere weakly correlated in a large population of incidentQuality of social interaction

and prevalent HD patients [11]. Interestingly, this rela- SleepSocial supporttionship did not hold in patients who had recently startedDialysis staff encouragementESRD therapy [22]. The intercorrelations, however, sug-Patient satisfaction

gest that the IEQ is an excellent measure of HRQOL.The meaning of the IEQ may be slightly different inincident compared with prevalent patients.

The Sickness Impact Profile (SIP) [35] is a subjectivemanner to track ESRD patients’ perceptions over time

measure used to assess the effect of the illness on the[39]. The SF-36 has also been used in ESRD patients

patient, designed to be non-disease-specific. There are[40, 41] and to assess changes associated with provision136 statements regarding 12 activities. Subscale scoresof erythropoietin [42, 43]. Patients with ESRD treatedinclude a Physical Dimension, and a Psychosocial Di-with HD had lower SF-36 scores compared with patientsmension, the latter including mental health and socialwith other chronic diseases [44] and patients with Typerelations, and an overall score. Scores may range fromII diabetes mellitus in the Medical Outcomes Study [45],0–100 on subscale and total scales, with higher scoresbut higher scores than patients with chronic obstructiveindicating poorer functional or psychosocial status. Itpulmonary disease and asthma.has been used about as frequently as the Karnofsky Scale

The Kidney Disease Questionnaire [46] represents anin patients with ESRD [10], and has been used in Euro-early attempt to harness some of these disparate mea-pean studies of patients with ESRD [36, 37].sures into a disease-focussed scale for specific use inThe RAND 36-Item Health Survey 1.0 (SF-36) is apatients with renal disease. It is composed of five dimen-generic measure of HRQOL, designed by Dr. Ware andsions: Physical Symptoms, Fatigue, Depression, Rela-colleagues for use in the Medical Outcomes Study [38].tionships with Others and Frustration.It is a short form, composed of 36 items evaluating func-

The Kidney Disease Quality of Life (KDQOLe) In-tional status, well-being and perceptions of health status.strument (dialysis version) [44] is a self report measureThe eight scales are scored from 0 to 100 and are: Physicalspecifically developed for use with patients with ESRDFunctioning, Role Limitations due to Physical Health,treated with HD. It is composed of 134 generic andBodily Pain, General Health Perception, Vitality, Socialkidney disease/dialysis specific items. It was developedFunctioning, Role Limitations due to Emotional Prob-using focus groups consisting of patients and staff mem-lems, and Mental Health. Higher scores signify betterbers, and was tested on small samples of HD patientsperception of health. U.S. population means for the[44]. Functioning and well-being are evaluated using, asscales vary from 61 to 84, and the norms can be adjustedits basis, [47, 48] and additional scales to assess concernsfor age. A Physical Component Score (PCS) can be cal-specific to dialysis patients (Table 4). These latter in-culated, and is an appraisal of Physical Functioning, Roleclude: a symptom/problem scale, an effects of kidneyPhysical, Bodily Pain, and General Health scores, withdisease on daily life scale, and an employment statusa transformed mean of 50, and a standard deviation ofevaluation. Items to assess cognitive function were de-10. The Mental Component Score (MCS) (calculatedrived from the SIP [35, 49]. Other scales were derivedfrom the Vitality, Social Functioning, Role Emotional,from other instruments: a quality of social interactionand Mental Health subscales) has an identical trans-scale, a sexual function scale, and items to assess sleep,formed mean and standard deviation. Therefore, forsocial support and patient satisfaction [44]. Items createdthese summary scores of the SF-36, scores above 50 arefor the KDQOL assess patients’ perceptions of encour-above average and scores below 50 are below average.agement by staff and their evaluation of their healthInterestingly, the questions comprising the vitality scoresstatus. Scale scores can be transformed linearly intoassess issues regarding fatigue and energy, and like many0–100 point scores. Higher scores indicate more favor-depression scores, may be confounded by the medicalable perceptions. RAND scoring algorithms may be usedaspects of a patient’s illness [11].to score the KDQOLe SF-36 items. The scales showThe SF-36 has several potential uses. It can provideinternal consistency and reliability [44]. A shorter form,descriptive information regarding populations, and canthe KDQOL-SFe (composed of 79 questions) was de-be used to compare populations. In addition, it can be

used in individual patients and has been used in this veloped because of concerns regarding the time it took

Kimmel: Quality of life in HD patientsS-118

to administer the KDQOLe [44]. The number of ESRD- Interestingly, the relationship between depressive af-fect and survival has been extremely controversial inrelated and generic QOL items was markedly decreased.

Subsequent analyses suggested the scales comprise four studies of patients with ESRD. This may be a functionof the different definitions that have been used to assessdimensions. Physical health (composed of physical func-

tioning, work status, role limitations due to physical depression, and the variation in populations studied [11].However, the analytic strategies for assessing the rela-health, general health perceptions, pain, energy), mental

health (emotional well-being, quality of social interac- tionship between depression and outcome have in gen-eral been poorly controlled. Few studies have assessedtions, burden of kidney disease, social support, role limi-

tations due to emotional problems), kidney-disease is- the relationships of demographic factors, patient compli-ance, nutrition, dose of dialysis delivered and psycho-sues (cognitive function, symptoms/problems, effects of

kidney disease on daily life, sexual function and sleep) neuroimmunologic function to outcome in research ondepressive affect in ESRD patients [24]. Recent evi-and patient evaluation of care (patient satisfaction and

perceptions of staff encouragement). Social functioning dence suggests longitudinal analyses may provide supe-rior data [56].is related to both the domains of physical and mental

health [44]. Limited material in the KDQOLe is devoted Perceptions of social support have been linked to pa-tient survival in almost every chronic illness in which itto the assessment of global satisfaction, marital adjust-

ment or social support. has been assessed [57]. Several studies have establishedthis as the case for patients with ESRD [55, 58–60]. AsThe use of the KDQOLe has increased remarkably

over the last several years. The KDQOL is being used outlined in a review article, the challenge is to understandthe mechanisms by which the experience of social sup-in the NIDDK-supported HEMO study. The HEMO

study reported on its first assessment of QOL parameters port is translated into improved survival. Candidates formediators include better access to health care, betterin a large group of patients enrolled during the initial

phases of the study [50]. The mean physical components compliance and better neurophysiologic or psychologicfunctioning [57].score of the HEMO patients was lower than the national

We recently showed patients with lower perception ofaverage, but mean mental component scores were ap-the intrusiveness of illness using the IEQ had improvedproximately equal to the U.S. norm. Interestingly, in ourpatient survival in a population of incident and prevalentpopulation of HD patients, studied in Washington, DC,patients with ESRD treated with HD, when the variationprimarily composed of African-American men, total andin several demographic predictors, including age, medi-ESRD-related KDQOL scores were highly correlatedcal comorbidity, and serum albumin concentration waswith patients’ perceptions of both depressive symptomscontrolled [55]. The mechanisms underlying this associa-and the intrusiveness of illness effects, but not with socialtion are unclear, but may involve neurohumoral as wellsupport or Karnofsky scores [51].as behavioral pathways [55, 57]. This association withoutcome, as well as its intercorrelations and lack of corre-

DO THESE MEASURES MATTER? lation with severity of illness make the IEQ a very attrac-A critical issue for the measure of QOL is are these tive measure of QOL for use in patients with ESRD

assessments related to outcomes [52]? As the field ma- treated with HD.tures, it is becoming increasing clear that these measures Socioeconomic status (SES) has consistently been as-are related to important outcomes in patients with sociated with improved health outcomes in many popula-ESRD. tions [61]. Few studies have assessed the domains of SES,

Functional status parameters have been shown by sev- and their association with QOL indicators and outcomeeral investigators [17, 53, 54] to predict hospitalization measures in patients with ESRD [62]. Intriguing prelimi-and survival, when the variation in several demographic nary data suggest that these parameters operate differ-predictors is controlled. However, it is to be noted that ently in different patient groups. More studies are neces-functional scores are not correlated with satisfaction sary to assess these issues in the U.S. and in other ESRDscores, are often not subjective, and are insufficient for populations.making judgements regarding continuation of dialytic Finally, in a study of 79 ESRD patients treated withtherapy. HD, Julie Kovac in our group showed improved behav-

We were unable to show an association of our satisfac- ioral compliance and higher levels of serum albumintion with life scale and patient survival, although it was were associated with patients’ perceptions of their satis-highly correlated with what we consider many QOL faction with nephrologists [63]. These patient percep-scales [55]. Subjective, patient-centered satisfaction with tions of satisfaction with the physicians were not associ-life measures may however be most useful to families ated with the severity of patients’ illnesses. Such findingsand care-givers in evaluating decisions to discontinue suggest the perceptions of patients regarding their physi-

cians may be associated with improved nutrition anddialysis treatments.

Kimmel: Quality of life in HD patients S-119

11. Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Boyle DH,compliance with the dialysis prescription, both of whichCruz I, Umana WO, Alleyne S, Veis JH: Aspects of quality of

are associated with improved survival. Therefore our life in hemodialysis patients. J Am Soc Nephrol 6:1418–1426, 199512. Lara-Munoz C, Feinstein AR: How should quality of life berelationship with patients may be a critical part of their

measured? J Invest Med 47:17–24, 1999perceptions of the quality of their treatment, and there-13. Karnofsky DA, Burchenal JH: The clinical evaluation of chemo-

fore their HRQOL. therapeutic agents in cancer, in Evaluation of ChemotherapeuticAgents, edited by MacLeod CM, New York, Columbia UniversityAs studies become more refined, it seems probablePress, 1949 pp 191–205.that measurements of QOL will be related to many out-

14. Gombos EA, Lee TH, Harten MR, Cummings JW: One year’scome measures, such as compliance with medications, experience with an intermittent dialysis program. Ann Intern Med

61:462–469, 1964fluid and dietary regimens and the dialytic prescription,15. Gutman RA, Stead WW, Robinson RR: Physical activity andand with hospitalizations, and decisions regarding dis-

employment status of patients on maintenance dialysis. N Engl Jcontinuation of dialysis and death [52]. It will be interest- Med 304:309–313, 1981

16. Evans RW, Manninen DL, Garrison LP, Hart G, Blagg CR,ing to discover whether QOL measurements are relatedGutman RA, Hull AR, Lowrie EG: The quality of life of patientsto treatment decisions such as level of Kt/V and treat-with end-stage renal disease. N Engl J Med 312:553–559, 1985

ment time prescribed, over time, in longitudinal studies. 17. McClellan WM, Anson C, Birkeli K, Tuttle E: Functionalstatus and quality of life: Predictors of early mortality among pa-Perceptions of QOL in patients treated with HD musttients entering treatment for end stage renal disease. J Clin Epide-be measured in several ways, probably using a mix of miol 44:83–89, 1991

subjective disease-specific scales, as well as well-appreci- 18. Hutchinson TA, Boyd NF, Feinstein AR, Gonda A, HollombyD, Rowat B: Scientific problems in clinical scales, as demonstratedated and validated psychological measures. The chal-by the Karnofsky index of performance status. J Chronic Dislenge will be at least threefold: to identify meaningful 32:661–666, 1979

QOL indicators that can be easily appraised by both 19. Bremer BA, McCauley CR, Wrona RM, Johnson JP: Qualityof life in end-stage renal disease: A reexamination. Am J Kidneypatients and practitioners, to integrate such measures,Dis 13:200–209, 1989.including satisfaction scores, into routine ESRD care, 20. Diener E, Emmons RA, Larsen RJ, Griffin S: The satisfaction

and to define treatment methods associated with ensur- with life scale. J Personal Assessm 49:71–75, 198521. Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Boyle DH,ing and maximizing patient satisfaction and perception

Verme D, Umana WO, Veis JH, Alleyne S, Cruz I: Behavioralof QOL. compliance with dialysis prescription in hemodialysis patients. JAm Soc Nephrol 5:1826–1834, 1995

Reprint requests to Paul L. Kimmel, M.D., Division of Kidney, Uro- 22. Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Boyle DH,logic and Hematologic Diseases, National Institute of Diabetes and Umana WO, Kovac JA, Alleyne S, Cruz I, Veis JH: PsychologicDigestive and Kidney Diseases, National Institutes of Health, Room functioning, quality of life and behavioral compliance in patients6AS19B, Building 45, Bethesda, MD 20892, USA. beginning hemodialysis. J Am Soc Neprol 7:2152–2159, 1996

23. Kutner NG, Devins GM: A comparison of the quality of lifereported by elderly whites and elderly blacks on dialysis. GerREFERENCESNephrol Urol 8:77–83, 1998

24. Kimmel PL, Weihs K, Peterson RA: Survival in hemodialysis1. Gill TM, Feinstein AR: A critical appraisal of the quality of lifemeasurements. JAMA 272:619–626, 1994 patients: The role of depression. J Am Soc Nephrol 4:12–27, 1993

25. Beck AT, Steer RA, Garbin MG: Psychometric properties of the2. Campbell A, Converse P, Rodgers W: The Quality of AmericanLife. New York, Russell Sage, 1976 Beck Depression Inventory: Twenty five years of evaluation. Clin

Psychol Rev 8:77–100, 19883. Rettig RA, Sadler JH, Meyer KB, Wasson JH, Parkerson GR,Kantz B, Hays RD, Patrick DL: Assessing health and quality of 26. Frasure-Smith N, Lesperance F, Taljic M: Depression following

myocardial infarction. JAMA 270:1819–1825, 1993life outcomes in dialysis: A report on an Institute of Medicineworkshop. Am J Kidney Dis 30:140–155, 1997 27. Craven JL, Rodin GM, Littlefield C: The Beck Depression

Inventory as a screening device for major depression in renal4. United States Renal Data System (USRDS) 1998 Annual Report,Bethesda, MD, National Institutes of Health, National Institute dialysis patients. Int J Psychiatry Med 18:365–374, 1988

28. Sacks CR, Peterson RA, Kimmel PL: Perception of illness andof Diabetes and Digestive and Kidney Diseases5. Simmons RG, Anderson C, Kamstra L: Comparison of quality depression in chronic renal disease. Am J Kidney Dis 15:31–39,

1990of life of patients on continuous ambulatory peritoneal dialysis,hemodialysis, and after transplantation. Am J Kidney Dis 4:253– 29. Peterson RA, Kimmel PL, Sacks CR, Mesquita ML, Simmens

SJ, Reiss D: Depression, perception of illness and mortality in255, 19846. von Albertini B, Bosch JP: Short hemodialysis. Am J Nephrol patients with end-stage renal disease. Int J Psych Med 21:343–354,

199111:169–173, 19917. Husebye DG, Westlie L, Styrvoky TJ, Kjellstrand CM: Psycho- 30. Zimet GD, Dahlem NW, Zimet SG: The multidimensional scale

of perceived social support. J Personality Assess 52:30–41, 1988logical, social and somatic prognostic indicators in old patientsundergoing long-term dialysis. Arch Intern Med 147:1921–1924, 31. Devins GM, Mandin H, Hons RB, et al.: Illness-intrusiveness

and quality of life in end-stage renal disease: Comparison and19878. Ifudu O, Mayers J, Matthew J, Tan CC, Cambridge A, Friedman stability across modalities. Health Psychol 9:117–142, 1990

32. Peterson RA, Greenberg GD: The role of perception of illness.EA: Dismal rehabilitation in geriatric inner-city hemodialysis pa-tients. JAMA 271:29–33, 1994 Health Psychol 11:2–3, 1989

33. Greenberg GD, Peterson RA: Illness effects questionnaire. In:9. Kutner NG: Assessing end-stage renal disease patients’ function-ing and well-being: Measurement approaches and implications for Evaluating Stress: A Book of Resources, edited by Zalaguett CP,

Woods RJ, Lanham, MD, The Scarecrow Press Inc., pp 141–164clinical practice. Am J Kidney Dis 24:321–333, 199410. Edgell ET, Coons SJ, Carter WB, Kallich JD, Mapes D, Da- 34. Wagner MK, Armstrong D, Laughlin JE: Cognitive determi-

nants of quality of life after onset of cancer. Psychol Reports 77:147–mush TM, Hays RD: A review of health-related quality-of-lifemeasures used in end-stage renal disease. Clin Ther 18:887–938, 154, 1995

35. Bergner M, Bobbit RA, Carter WB, Gibson BS: The Sickness1996

Kimmel: Quality of life in HD patientsS-120

Impact Profile: Development and final revision of a health status cal distress/well-being and cognitive functioning measures in Mea-suring Functioning and Well-being: The Medical Outcomes Studymeasure. Med Care 19:787–805, 1981Approach, edited by Stewart AL, Ware JE, Durham NC, Univer-36. Moreno F, Aracil FJ, Perez R, Valderrabano F: Controlledsity Press, 1992, pp 102–142study on the improvement of quality of life in elderly hemodialysis

50. Meyer K, Paranandi L, Hays R, Benz R, Athienites N, Kusekpatients after correcting end-stage renal disease-related anemiaJ, Levey A, for the HEMO Study Group: Quality of life in thewith erythropoietin. Am J Kidney Dis 27:548–556, 1996HEMO study: An interim report. J Am Soc Nephrol 8:204A, 199737. Moreno F, Lopez Gomez JM, Sanz-Guajardo D, Jofre R, Val-

51. Kimmel PL, Peterson RA, Shidler NR: What does KDQOLderrabano F: Quality of life in dialysis patients. A Spanish multi-measure? J Am Soc Nephrol 9:1091A, 1998centre study. Nephrol Dial Transplant 11 (Suppl 2): 125–129, 1996

52. Wilson IB, Cleary PD: Linking clinical variables with health-38. Stewart AL, Greenfield S, Hays RD, Wells K, Rogers WH,related quality of life. JAMA 273:59–65, 1995Berry SD, McGlynn EA, Ware JE: Functional status and well

53. De Oreo P: Hemodialysis patient-assessed functional health statusbeing of patients with chronic conditions: Results from the Medicalpredicts continued survival, hospitalization, and dialysis-atten-Outcomes Study. JAMA 262:907–913, 1989dance compliance. Am J Kidney Dis 30:204–212, 199739. Meyer KB, Espindle DM, DeGiacomo JM, Jenuleson CS, Kurtin

54. Kimmel PL, Peterson RA, Weihs KL, Simmens S, Kovac J, VeisPS, Davies AR: Monitoring dialysis patients’ health status. Am JJ, Cruz I: Compliance with dialysis prescription, social support,Kidney Dis 24:267–279, 1994quality of life and decreased functional level are associated with40. Khan IH, Garratt AM, Kumar A, Cady RJ, Gatto GRD, Ed-enhanced survival in inner city hemodialysis patients. J Am Socward N, Macleod AM: Patients’ perception of health on renalNephrol 7:1451A, 1996replacement therapy: Evaluation using a new instrument. Nephrol

55. Kimmel PL, Peterson RA, Weihs KL, Simmens SJ, Boyle DH,Dial Transplant 10:684–689, 1995Umana WO, Kovac JA, Alleyne S, Cruz I, Veis JH: Psychosocial41. Merkus MP, Jager KJ, Dekker FW, Boeschoten EW, Stevensfactors, behavioral compliance and survival in urban hemodialysisP, Krediet RT: NECOSAD Study Group: Quality of life in patientspatients. Kidney Int 54:245–254, 1998on chronic dialysis: Self-assessment three months after the start

56. Kimmel PL, Peterson R, Weihs K, Simmens S, Cruz I, Veis J:of treatment. Am J Kidney Dis 29:584–592, 1997 Depression predicts Short-term mortality in chronic hemodialysis42. Levin NW, Lazarus JM, Nissenson AR: Maximizing patient bene- patients. J Am Soc Nephrol 10:1455A, 1999fits with epoietin alfa therapy: National cooperative rHu erythro- 57. House JS, Landis KR, Umberson D: Social relationships andpoietin study in patients with chronic renal failure—an interim health. Science 241:540–545, 1988report. Am J Kidney Dis 22:45–82, 1993 58. McClellan WM, Stanwyck DJ, Anson CA: Social support and

43. Beusterien KM, Nissenson AR, Port FK, Kally M, Steinwald subsequent mortality among patients with end-stage renal disease.B, Ware JE: The effects of recombinant human erythropoietin on J Am Soc Nephrol 4:1028–1034, 1993functional health and well being in chronic hemodialysis patients. 59. Christensen AJ, Wiebe JS, Smith TW, Turner CW: PredictorsJ Am Soc Nephrol 7:763–773, 1996 of survival among hemodialysis patients: Effect of perceived family

44. Hays RD, Kallich JD, Mapes DL, Coons SJ, Carter WB: Devel- support. Health Psychol 13:521–525, 1994opment of the Kidney Disease Quality of Life (KDQOLe) Instru- 60. Christensen AJ, Smith TW, Turner CW, Holman JM, Jr, Greg-ment. Quality Life Res 3:329–338, 1994 ory MC, Rich MA: Family support, physical impairment and,

45. Ware JE, Snow KK, Kosinski M, Gandek B: SF-36 Health Survey: adherence in hemodialysis: An investigation of main and bufferingManual and Interpretation Guide. Boston, MA, The Health Insti- effects. J Behavior Med 15:313–325, 1992tute, 1993 61. Adler NE, Boyce T, Chesney MA, Cohen S, Folkman S, Kahn

46. Laupacis A, Muirhead N, Keown P, Wong C: A disease-specific RL, Syme SL: Socioeconomic status and health: The challenge ofquestionnaire for assessing quality of life in patients on hemodialy- the gradient. Am Psychol 49:15–24, 1994sis. Nephron 60:302–306, 1992 62. Kadlubek P, Thamer M, Richard C, Kimmel PL: Socioeconomic

47. Hays RD, Sherbourne CD, Mazel RM: The RAND 36-Item status and mortality in US hemodialysis patients. J Am Soc NephrolHealth Survey 1.0. Health Econ 2:217–227, 1993 10:1244A, 1999

48. Ware JE, Sherbourne CD: The MOS 36-item short-form health 63. Kovac J, Shidler N, Peterson RA, Kimmel PL: Patients’ satisfac-survey (SF–36): I. Conceptual framework and item selection. Med tion with medical and nursing staff are associated with hemodialysisCare 30:473–483, 1992 patients’ compliance with dialysis prescription. J Am Soc Nephrol

9:214A, 199849. Stewart AL, Ware JE, Sherbourne CD, Wells KB: Psychologi-