Nephrol Dial Transplant (2001) 16: 1387±1394
Original Article
Self-assessed physical and mental function of haemodialysis patients
Sanjeev K. Mittal, Lori Ahern, Edith Flaster, John K. Maesaka and Steven Fishbane
Winthrop-University Hospital, Mineola, New York, USA
Abstract
Background. Physical (PCS) and mental (MCS) component summary scales of the Short Form 36 (SF-36)
health survey are validated measures of quality of life
(QOL) and functional status. We sought to evaluate
the PCS and MCS in haemodialyis patients as
compared to the general population and other chronic
diseases.
Methods. A cohort of 134 haemodialysis patients
(mean age 60.9"14.3 years, males 63.4%, Caucasians
66.4%) was followed from January 1996 to December
1998 (mean follow up 14.5"5.7 months). SF-36 questionnaires were administered every 3 months and
PCS and MCS were calculated. Results were compared to the general population and other chronic
diseases. Correlators of PCS and MCS, change in
QOL over time, and the correlators of this change were
determined.
Results. Mean PCS was 36.9"8.8 and mean MCS was
47"10.7. Compared to the general US population,
these represent a decline of 8.7"0.8 for PCS
(P-0.0001) and 2.7"0.8 for MCS (P-0.001). PCS
and MCS in end-stage renal disease (ESRD) were
lower than in most other chronic diseases studied.
Univariate correlators of PCS in haemodialysis
patients included age, male sex, haematocrit, serum
albumin, and severity of comorbid cardiac and
pulmonary illnesses. Multivariate analysis demonstrated independent correlators of PCS to be male
sex, serum albumin and severity of comorbid cardiac
and pulmonary diseases. Univariate as well as multivariate correlators of MCS included: serum albumin,
KTuVurea, and status living alone. A trend analysis
revealed that both PCS and MCS tended to decline
in the initial months of dialysis but stabilized over
time. Status living alone was a signi®cant predictor
of improvement in MCS by univariate as well as
multivariate analysis.
Conclusions. Self assessed physical and mental health
of haemodialysis patients is markedly diminished
Correspondence and offprints requests to: Steven Fishbane MD,
222 Station Plaza N, Suite 510, Mineola, NY 11501, USA.
#
compared to the general population and other chronic
diseases.
Keywords: quality of life; SF-36; end-stage renal
disease; haemodialysis; depression; co-morbidity
Introduction
Recently there has been growing recognition of healthrelated quality of life (QOL) as an important indicator
of the quality of care for patients with various illnesses.
For patients with end-stage renal disease (ESRD),
chronic maintenance haemodialysis successfully prolongs life span. A broader goal, however, is to optimize
the patient's self perceived sense of well being and
quality of life. Studies comparing QOL between dialysis patients and the general population generally note
the negative impact of renal disease and its treatment
on patients' lives w1±6x, although this is not con®rmed
by all studies w7,8x.
The SF-36 health survey questionnaire is a self
administered standardized assessment of QOL and
functional status. It was developed and extensively
evaluated as part of the Medical Outcomes Study w9x,
and contains essential psychometric criteria that have
been shown to be both reliable and valid in a variety of
chronic disease states w10,11x. It has recently been used
in the dialysis population to evaluate self-perceived
health status w1,3,5,6,12x. We chose to use this tool
because of an ample literature supporting its validity,
and its ease of administration and interpretation. The
responses to items on the SF-36 questionnaire are
summed into eight scales of health: physical function
(PF), social function (SF), limitation in role due to
physical health (RP), limitation in role due to mental
health (RE), mental health (MH), vitality (VT), bodily
pain (BP), general health (GH). In clinical practice
we have found it dif®cult to use eight separate measures to meaningfully guide the detection and treatment of health status problems. We were intrigued by
the development and validation by John E. Ware of
summary scores based on the SF-36 to aggregate the
eight scale scores into a physical component summary
2001 European Renal Association±European Dialysis and Transplant Association
1388
score (PCS) and a mental component summary score
(MCS) w10x. The purpose of this study was to compare the PCS and MCS in haemodialysis patients to
the general population and other chronic diseases, to
determine correlators and to evaluate the change
in scores over time.
Subjects and methods
Patient population
The study was performed at Winthrop-University Hospital,
Mineola, NY, outpatient dialysis centre. At the time of study,
the centre treated approximately 191 haemodialysis patients,
with most on a thrice-weekly dialysis regimen. The SF-36
questionnaire was administered to all patients every 3 months
starting in January 1996. For the purpose of this study we
have extracted data on all patients dialysed in the period
between January 1996 and February 1998, who neither died
nor were transplanted. The recruitment rate was 100% at the
start of the study. All patients were dialysed with conventional, single use, high-ef®ciency polysulfone dialysers using
bicarbonate dialysate. The study protocol and informed
consent were approved by the institutional review board of
Winthrop-University Hospital.
Quality of life assessment and study design
Renal social workers were instructed and trained on the use
and purpose of the SF-36. They distributed the SF-36
questionnaire every 3 months to all haemodialysis patients.
Patients were given a brief explanation of the questionnaire
and were asked to complete it during regularly scheduled
dialysis treatments. Assistance was given for patients who
were illiterate or had a language problem. For those with
diminished vision, a large font version of the SF-36 was
given. The surveys were scored by hand, and the data were
entered on a Microsoft Excel spreadsheet which was updated
monthly.
The SF-36 is a 36-question instrument consisting of eight
scales to measure physical and mental dimensions of health
status. The eight scales are: PF, RP, BP, GH, VT, SF,
RE, MH. The scales are scored on a 0±100 possible range.
The above scales can be summarized into two component
summary scores which aggregate the physical and mental
components of the eight scale pro®le of SF-36 into PCS and
MCS scores. The PCS re¯ects physical morbidity and adaptation to disease, whereas the MCS re¯ects psychological or
mental morbidity and adaptation. The component summary
scores are positively scored and normalized to a general
population mean of 50 and a standard deviation of 10.
We calculated the PCS and MCS scores for each patient
with the help of SAS code for scoring algorithms provided by
the Medical Outcome Trust w10x.
Comorbidity information and demographics
Demographic data (age, sex, ethnicity), duration of haemodialysis, prescribed dialysis time, underlying aetiology of
ESRD, and marital status were recorded for each patient by
a blinded investigator. Level of education was graded as: 1,
illiterate; 2, can only read and write; 3, completed primary
studies; 4, high school; and 5, university studies or higher.
S. K. Mittal et al.
Presence of comorbid illness (cardiac, pulmonary and
peripheral vascular disease) was also graded and recorded
by disease type as: grade 0, no past or present illness;
grade 1, no current symptoms but positive disease prior to
6 months; grade 2, presence of mild symptoms within the
past 6 months; and grade 3, moderate to severe symptoms
within the past 6 months. Total number of hospitalizations
and hospital days during the study period were recorded.
Dialysis indicators
The mean (during the study period) serum albumin, haematocrit, KTuVurea, normalized protein catabolic rate (nPCR)
and urea reduction ratio (URR) were followed for all
patients.
Statistical methods
Univariate analysis of mean PCS and MCS was performed
with Spearman correlation coef®cients. Multivariate analysis
used stepwise regression to obtain the most ef®cient equation. Comparisons between this study cohort and the general
population, and in relation to other chronic diseases, were
performed using algorithms of the Medical Outcomes Trust
w10x. Rate of change in PCS and MCS over time was
computed in two ways: ®rst by obtaining the slope with time
using all observations for each patient; second by subtracting
the baseline from the ®nal value, and dividing by the time
between those readings. The results from both analyses were
the same, and, for simplicity, the later analysis is reported
here. All results are reported as mean"standard deviation.
Statistical signi®cance was considered to be a P value
of -0.05.
Results
One hundred and ninety-one haemodialysis patients
were entered into the initial dataset. Of these, 45
patients who had incomplete or single SF-36 scores
were excluded. The reasons for incomplete record
were: inability to speak English in 12 patients; unwillingness to answer the questionnaire in 20 patients;
anduor unsuitability due to cognitive impairment in
17 patients. Patients who were excluded did not differ
signi®cantly from those included in the study except
for a higher number of diabetes (23 out of 45) in the
excluded patients. Of the remaining 146 patients with
at least two complete scored questionnaires, three were
transplanted and nine patients died during the study
period, and were excluded from the analysis. This was
done to include only those patients who were stable
and completed the 2-year study period. Thus, we
report on 134 patients with at least two scored questionnaires (range 2±8 per patient). The mean time
interval between ®rst and last SF-36 scores was
14.5"5.7 (range 3±24) months. Descriptive statistics
including demographics and clinical information are
presented in Table 1. The mean PCS was 36.9"8.8
(range 18.1±56.7, median 36.3) and MCS was 48.7"9.3
(range 20.3±67.5, median 50.7).
Self-assessed physical and mental function of haemodialysis patients
Impact of ESRD on physical and mental health
as compared to the general population
There was a signi®cant reduction in both PCS and
MCS of ESRD patients as compared to that expected
of healthy subjects (Table 2). For PCS, the age and sex
adjusted score was 8.7"0.8 points lower than that of
the general US population (P-0.0001) and for MCS
the score was 2.7"0.8 points lower (P-0.001). Study
patients of all ages rated their PCS to be signi®cantly
Table 1. Descriptive statistics for study patients
Patient demographics
Age in years (range)
Males (%)
Married (%)
Living alone (%)
Education high school or more (%)
Ethnicity (%)
Caucasians
African-American
Hispanic
Others
Primary cause of ESRD (%)
Diabetes mellitus
Hypertension
Glomerulonephritis
Polycystic kidney disease
Other
Comorbidity
Diabetes (%)
Cardiac disease (%)
Pulmonary disease (%)
Peripheral vascular disease (%)
Number of hospitalizations (range)
Hospital days (range)
Dialysis adequacy
Dialysis KTuVurea
URR (%)
NPCR
Haematocrit
Serum albumin (gudl)
Months on dialysis
Prescribed dialysis time
SF-36 summary scores
Mean PCS
Mean MCS
60.9"14.3 (24±88)
63.4
33.6
16.4
76.9
66.4
26.1
3.7
3.7
30.6
50.7
16.4
20.9
1.5"1.9 (0±9)
12.2"21.2 (0 124)
36.9"8.8
48.7"9.3
ns134; data are mean"SD.
worse than their healthy counterparts, except for those
)75 years of age. In general, there was a trend
towards reduced PCS scores with age. Male patients
were found to have PCS scores 4.6 points above their
female counterparts (P-0.05), while MCS scores were
not signi®cantly different between the sexes.
Impact of ESRD on physical and mental health
as compared to other chronic diseases
PCS scores were signi®cantly lower in haemodialysis
patients as compared to patients with chronic angina,
diabetes mellitus, chronic lung disease, arthritis, hypertension, cancer and depression (Figure 1). The MCS
score was also signi®cantly lower in haemodialysis
patients than in several chronic diseases including
diabetes, chronic angina, limitation in the use of a
limb, and hypertension. In contrast, the MCS scores
for patients with chronic depression were signi®cantly
lower than that found in haemodialysis patients
(Figure 2).
Correlators of the physical and mental component
summary scores
24.6
29.9
12.7
8.2
24.6
1.5"0.3
69.4"6
1.2"0.3
32.9"2.3
3.7"0.3
29.8"44
212"24
1389
On univariate analysis, variables which signi®cantly
predicted a better PCS score were male sex (rs0.25,
Ps0.004), higher haematocrit (rs0.19, Ps0.03), and
a higher level of serum albumin (rs0.3, Ps0.0005). In
contrast, older age (rs 0.17, Ps0.05), diabetes as the
cause of ESRD (rs0.18, Ps0.036), severe cardiac
(rs 0.21, Ps0.013) and pulmonary disease (rs 0.29,
Ps0.0007) were inversely correlated with PCS. Higher
number of hospitalizations and hospital days trended
towards predicting lower PCS scores, but did not
achieve statistical signi®cance. Dialysis adequacy
measured by KTuVurea (rs0.16, Ps0.06) and a
higher serum albumin (rs0.15, Ps0.08) trended
towards predicting a better MCS score but without
reaching statistical signi®cance. Individuals living
alone had a higher MCS (rs0.18, Ps0.042) than
those living with others.
Serum level of albumin was the strongest correlate of
PCS on multivariate analysis (Table 3) with 8.8% of
the variation in PCS explained by the serum albumin.
Table 2. PCS and MCS scores as compared to general population by sex and age group
Total patients
Sex
Males
Females
Age groups
18±34
35±54
55±74
P75
n
PCS Study
134
36.9*"8.8
50"10
48.7"9.3
50"10
85
49
38.6*"8.3
34.0*"9.0
51.1"9.4
49.1"10.4
49.4"9.1
47.4"9.6
50.7"9.6
49.3"10.3
8
34
74
19
39.5*"8.5
37.9*"6.2
36.7*"8.8
35.5"6.2
52.6"6.9
50.2"5.8
44.8"10.5
37.9"11.2
49.7"9.1
46.9"10.1
47.5"8.2
51.7"9.6
49.1"8.3
48.5"9.0
51.1"9.7
50.4"11.7
Values are given as mean"SD. *P-0.05.
General population
MCS Study
General population
1390
Fig. 1. Effect of chronic diseases on PCS.
Fig. 2. Effect of chronic diseases on MCS.
S. K. Mittal et al.
Self-assessed physical and mental function of haemodialysis patients
We found that every 1 mgudl increase in serum
albumin is re¯ected by a 1.2 point increase in PCS.
Hospitalizations
Over the 2-year study period, 89 out of 134 patients
(66.4%) were hospitalized. The average length of
hospital stay was 12.2"21.1 days. Patients who were
hospitalized had a signi®cantly lower mean baseline
PCS and MCS then patients who were never admitted
(35.8"8.6 vs 39.2"8.8 PCS respectively, P-0.05; and
MCS 47.3"9.4 vs 51.3"8.7, P-0.05). The odds ratio
for hospitalization was 1.92 (95% con®dence interval
0.91±4.01, Ps0.08, x2) when the PCS score was lower
than the median value of 36.3.
1391
Rate of change of PCS and MCS over time
The mean rate of change of PCS (0.08"0.89umonth,
range 2.13 to 3.92, median 0.08) and MCS (0.1"
0.94umonth, range 2.49 to 3.11, median 0.09) scores
over time were not signi®cantly different from zero
(Figure 3). Rate of change scores in individual patients
were categorized as improved, unchanged, or worsened
by determining whether the patient's rate of change
was within or outside of the 95% con®dence bands for
group rate of change scores. Approximately one-half
of patients stayed the same physically (50.7%) or
mentally (53.7%) over the study period. A slightly
greater percentage declined (26.1%) in physical health
status than improved (23.1%) while the reverse was
true for mental health (27.6% improved and 18.7%
Table 3. Correlators of PCS and MCS (multivariate analysis)
Variable
PCS Slope (95% CL)
P
MCS Slope (95% CL)
P
Male sex
Living alone
Cause of ESRD (not diabetes)
Severity of cardiac disease
Severity of pulmonary disease
KTuVurea
Serum albumin
3.2 (0.3,6.1)
NS
2.2 (0.4,4.0)
1.4 ( 2.6, 0.2)
4.9 ( 8.7, 1.1)
NS
5.3 (1.1,9.5)
0.022
NS
0.043
0.026
0.003
NS
0.0005
NS
5.3 (1.1,9.5)
NS
NS
NS
5.6 (0.4,10.8)
5.3 (0.7,9.9)
NS
0.042
NS
NS
NS
0.037
0.033
Total r squared 0.24 (PKCS); 0.10 (MCS); CL, con®dence limit.
Fig. 3. Change in PCS and MCS over time.
1392
S. K. Mittal et al.
Table 4. Correlators of change in PkCS and MCS over time (multivariate analysis)
PCS Slope
(beta) (95% CL)
Months on dialysis
Living alone
0.0043 ( 0.0077,
NS
P
0.0009)
0.01
NS
MCS Slope
(beta) (95% CL)
0.0048 ( 0.0083,
0.66 (0.26,0.86)
P
0.0013)
0.007
0.0003
Total r squared 0.045 (PCS); 0.178 (MCS); CL, con®dence limit.
declined). None of these categorical changes were
statistically signi®cant.
Correlators of change of PCS and
MCS over time
By univariate analysis, number of months on dialysis
had an inverse relationship with the change in
PCS (rs 0.21, Ps0.01) as well as change in MCS
(rs 0.19, Ps0.03) scores. When the PCS and MCS
scores were plotted against time on dialysis, both
showed a regression to the mean (Figure 3). By multivariate analysis, months on dialysis proved to be the
only independent predictor of change in both PCS and
MCS over time (Table 4). A living status of living
alone was an independent predictor of improvement in
MCS with time (Table 4).
Discussion
The metrics used in the study, the PCS and MCS
scores of the Short Form-36 pro®le, proved easy to
work with and analyse. In clinical practice, we have
found the summary scores to be of greater use compared to the eight individual SF-36 scales. The PCS
and MCS scores allow the physician to rapidly assess
whether the health needs of importance to the patient
are being met. If either score is low, the individual
SF-36 scales can be used to focus further investigation.
For the purpose of research, utilizing the summary
scores in contrast to using the eight individual SF-36
measures, makes it possible to reduce the number of
statistical comparisons and thereby the role of chance
in testing hypotheses w10x. Validation studies make it
clear that little information is lost when aggregating
the eight component scores into the PCS and MCS
w10,11x. We sought, therefore, to establish the range of
values of the PCS and MCS in haemodialysis patients,
to evaluate how these scores change over time, and to
determine the importance of various factors that may
impact on patients' perceived health status.
The haemodialysis patients studied had a signi®cant
reduction in self-assessed physical and mental health
compared to the general population. Interestingly, the
reduction in scores compared to normals was greater
for physical rather than mental health. This ®nding
is in agreement with other reports of studies of functional health status of haemodialysis patients w6,12x.
A greater effect on self assessed physical compared
to mental health has also been found in other chronic
diseases w13x. It may be that with chronic disease, the
impact on aspects of self-assessed mental health may
become blunted with time, as a useful psychological
adaptation.
The disease states previously reported to cause the
greatest reduction in measured PCS (6±8 points range)
have been congestive heart failure (CHF) and limitation in the use of an arm or leg w10x. The effect of
CRF and haemodialysis on perceived physical health
were found in our study to exceed the effect of both
of these chronic diseases (Figure 1). We found several
factors that explained this reduction in PCS scores in
haemodialysis patients, including female sex, diabetes
as the cause of ESRD, severe comorbidity due to pulmonary or cardiac disease, and the strongest predictor
was low levels of serum albumin. The importance of
female sex will be discussed in a later section. The association of diabetic kidney disease and severe cardiac
with pulmonary disease, points to the important detrimental effect of comorbid factors superimposed on
chronic renal failure in reducing the patients' perceived
physical health w14x. The serum albumin remains a
powerful, yet enigmatic predictor of poor outcomes
in ESRD w15x. In recent years, a body of evidence
has accumulated suggesting that a low serum albumin
concentration re¯ects the presence of in¯ammation
and malnutrition w16x. The association we have found
between serum albumin and patients' perception of
their physical health is consistent with the general
perception that low serum albumin levels may re¯ect
poor overall physical health.
As was true for PCS scores, we found MCS scores
in haemodialysis patients to be decreased compared
to the general population and certain chronic diseases
such as angina pectoris and diabetes mellitus. The need
for haemodialysis treatment imposes a signi®cant
psychosocial burden on patients. Aside from the time
commitment, the increased dependence on family
members, and the anxiety that the treatment causes,
many patients feel tired or depressed after treatments.
We found that a lower KTuV and serum albumin
concentration independently predicted a reduction
in MCS. The complexity of interpreting the albumin effect has already been discussed. The effect of
lower levels of KTuV is interesting, and suggests
the value of including MCS as an outcome variable
in studies of dialysis adequacy.
An MCS score of O42 has a sensitivity and
speci®city of 73.7% and 80.6% respectively to diagnose
clinical depression w10x. We found that one out of four
haemodialysis patients met this criterion, a prevalence
Self-assessed physical and mental function of haemodialysis patients
1393
of depression similar to that found in other reports
in haemodialysis patients w4,6x. Depression impacts
on loss of well being to the same degree as chronic
medical illness w13x and may also contribute to
mortality in haemodialysis patients w17x. The MCS
may provide an easy screen to identify haemodialysis
patients at risk for clinical depression who require
further evaluation.
We found lower PCS scores in female compared to
male patients on haemodialysis. It is of interest that
this pattern has also been found in studies of the
general US population, where the mean PCS score is
2.0 points higher in men than women. This difference
by sex does not emerge until after age 24 w10,18x. Other
investigators have noted that female sex is associated
with a reduced quality of life in ESRD w19,20x. In fact,
these investigators found the impact of female sex on
physical function to be roughly equivalent to the effect
of having diabetes mellitus. The reason for this gender
difference remains speculative. Possible explanations
could include biological factors or cultural conditioning, biases in the provision of care according to sex
w21x, or the effect of differences in clinician's attitude
toward female patients w22x. This is an area that is
poorly understood and one that needs additional
study.
The effect of age on PCS and MCS in the general
population indicates that PCS begins to decline signi®cantly in the 5th decade, whereas MCS stays stable
throughout all age groups w10x. Our study of haemodialysis patients con®rms a similar pattern. There was
a roughly linear decline in PCS with age, but MCS
scales remained relatively stable in older subjects. In a
previous study, DeOreo found that MCS increased
signi®cantly in older dialysis patients w6x. It may be that
with older age, patients are better able to adapt
emotionally for their chronic disease, based on their
wealth of life experience. It should be noted, however,
that other investigators have found that self-assessed
mental function of dialysis patients may actually
decrease with age w19,23,24x.
The availability of social support can affect both
survival and health-related QOL of dialysis patients.
Perceived social support has been independently and
positively associated with a better perception of illness,
life satisfaction, and feeling about life in general
w25,26,27x. Surprisingly, we found that living alone
independently predicted a better MCS score in the
present investigation. None of the other social factors
studied, including level of education and marital
status, correlated with physical or mental health.
Better self-assessed mental status of a patient on maintenance dialysis living alone may partially be due to
dif®culties in coping with family responsibilities and
increased dependence w28x. Patients' autonomy and
control associated with aspects of self-care process
might be an additional factor w29x.
We could not ®nd any correlation between measures
of dialysis adequacy (KTuVurea, URR, and nPCR) or
prescribed dialysis time and PCS. MCS, however, had a
mild positive correlation with KTuVurea on multivariate
analysis. Similarly Kurtin et al. w1x and Morton et al.
w30x, using the RAND 36 item health survey, failed
to show any relationship between dialysis adequacy
and any of the domains of HRQOL. In another
study which used the SF-36 to measure HRQOL, no
signi®cant association was observed between dialysis
KTuVurea and scale scores w5x. Another important variable, haematocrit, has recently been demonstrated to
positively in¯uence QOL in dialysis patients w5,12,19x.
Our study con®rmed a positive impact of haematocrit
on PCS. This effect, however, was not evident on
multivariate analysis.
A practical use of functional health status measures
is to estimate changes over time and evaluate the
impact of various treatment interventions. As an
example, signi®cant improvement in various physical
and emotional scales have been reported as a result
of treatment with rHuEPO w12,31x. Overall, we found
no signi®cant mean change in self-assessed physical
or mental health of our patients during the 2-year
study period (Figure 3). Both PCS and MCS tended
to decline in the initial months of dialysis but stabilized
over time. Norms for 1-year change scores for various
chronic conditions also follow a similar pattern w10x.
Despite not ®nding a change in mean scores, we did
®nd that 23.1% (PCS) and 27.6% (MCS) of patients
had improvement in scores over time. Months on
dialysis and status living alone were the two most
important variables that independently predicted an
improvement. Of note, none of the other important
clinical variables like KTuV, URR, nPCR, haematocrit, serum albumin, level of education, severity of
comorbidity or number of hospitalizations predicted
a change in self-assessed physical or mental health.
In summary, we have found the component summary scales of SF-36 health survey to be user friendly
and easy to interpret. The results indicate the important negative impact that chronic renal failure and
haemodialysis treatments have on self-assessed mental
and physical health. We suggest that research in ESRD
should focus more on patient-assessed health outcomes
such as those provided by the SF-36 questionnaire.
References
1. Kurtin PS, Davies AR, Meyers KB, DeGiacomo JM, Kantz ME.
Patient-based health status measures in out-patient dialysis: early
experiences in developing an outcome assessment program.
Med Care 1992; 30: MS136±MS149
2. Wilson IB, Cleary PD. Linking clinical variables with healthrelated quality of life. A conceptual model of patient outcomes.
JAMA 1995; 273: 59±65
3. Khan IH, Garratt AM, Kumar A et al. Patients' perception
of health on renal replacement therapy: evaluation using a new
instrument. Nephrol Dial Transplant 1995; 10: 684±689
4. Kimmel PL, Peterson RA, Weihs LK et al. Behavioural
compliance in hemodialysis patients. J Am Soc Nephrol 1995;
5: 1826±1834
5. Merkus MP, Jager KJ, Dekker FW, Boeschoten EW, Stevens P,
Krediet RT, NECOSAD study group. Quality of life in patients
on chronic dialysis: self-assessment 3 months after the start of
treatment. Am J Kid Dis 1997; 29: 584±592
1394
6. DeOreo PB. Hemodialysis patient-assessed functional health
status predicts continued survival, hospitalization and dialysisattendance compliance. Am J Kid Dis 1997; 30: 204±212
7. Johnson JP, McCauley CR, Copley JB. The quality of life
of hemodialysis and transplant patients. Kidney Int 1982;
22: 286±291
8. Auer J, Gokal R, Stout JP et al. The OxforduManchester study
of dialysis patients. Scand J Urol Nephrol 1990; 131: 31±37
9. Tarlov AR, Ware JE, Green®eld S, Nelson EC, Perrin E,
Zubkoff M. The medical outcome study: an application
of methods for monitoring the results of medical care. JAMA
1989; 262: 907±913
10. Ware JE, Kosinki M, Keller SD. SF-36 physical and mental
heath summary scales: a user's manual (4th printing, revised)
Boston, MA. Health institute, 1994
11. Ware JE, Kosinki M, Bayliss MS, McHorney CA, Rogers WH,
Raczek A. Comparison of methods for the scoring and statistical
analysis of SF-36 health pro®le and summary measures:
summary of results from the Medical Outcome Study. Med
Care 1995, 33: AS264±AS279
12. Beusterien KM, Nissenson AR, Port FK, Kally M, Steinwald B,
Ware JE. The effects of recombinant human erythropoietin
on functional health and well being in chronic dialysis patients.
J Am Soc Nephrol 1996; 7: 763±773
13. Hayes RD, Wells KB, Sherbourne CD, Rogers W, Spritzer K.
Functioning and well-being outcomes of patients with depression
compared with chronic general medical illness. Arch Gen
Psychiatry 1995; 52: 11±19
14. Harris LE, Luft FC, Rudy DW, Tierney WM. Clinical
correlations of functional status in patients with chronic renal
insuf®ciency. Am J Kid Dis 1993; 21: 161±166
15. Lowrie EG, Lew NL. Death risk in hemodialysis patients.
The predictive value of commonly measured variables and
an evaluation of death rate differences between facilities.
Am J Kidney Dis 1990; 15: 458±482
16. Bistrian BR. Role of the systemic in¯ammatory response
syndrome in the development of protein-calorie malnutrition in
ESRD. Am J Kidney Dis 1998; 32 (Suppl. 4): S113±7
17. Kimmel PL, Weihs LK, Peterson RA. Survival in hemodialysis
patients. The role of depression. J Am Soc Nephrol 1993;
13: 12±27
18. Thalji L, Haggerty CC, Rubin R, Berckmans TR, Pardee BL.
1990 National Survey of Functional Health Status: Final Report.
Chicago, IL: NORC, 1991
S. K. Mittal et al.
19. Moreno F, Lopezgomez JM, Sanz GD, Jofre R, Valderrabano F.
Quality of life in dialysis patients. A Spanish multi centric study.
Nephrol Dial Transplant 1996; 11 (Suppl. 2): 125±129
20. Lindqvist R, Carlsson M, Sjoden PO. Coping strategies
and quality of life among patients on hemodialysis and
continuous ambulatory peritoneal dialysis. Scand J Caring Sci
1998; 12: 223±30
21. Mustard CA, Kaufert P, Kozyrskyj A, Mayer T. Sex differences
in the use of health care services. N Engl J Med 1998;
338: 1678±1683
22. Safran DG, Rogers WH, Taylor AR, Mchorney CA,
Ware JE Jr. Gender differences in medical treatment: the case
of physician-prescribed activity restriction. Soc Sci Med
1997; 45: 711±722
23. Evans RW, Manninen DL, Garrison LP. The quality of life
of patients with end stage renal disease. N Engl J Med 1985;
312: 553±559
24. Mckevitt PM, Jones JF, Marion RR. The elderly on dialysis:
physical and psychosocial functioning. Dial Transplant 1986;
15: 130±137
25. Tell GS, Mittlemark MB, Hylander B, Shumaker SA,
Russel G, Bukart JM. Social support and health-related quality
of life in black and white dialysis patients. ANNA J 1995;
22: 301±308
26. Kimmel PL, Peterson RA, Weihs KL et al. Aspects of quality
of life in hemodialysis patients. J Am Soc Nephrol 1995;
6: 1418±1426
27. Julius M, Hawthorne VM, Carpenter AP et al. Independence
in activities of daily living for end-stage renal disease patients:
biomedical and demographic correlates. Am J Kid Dis 1989;
13: 61±69
28. Bremer BA, McCauley CR, Wrona RM, Johnson JP. Quality
of life in end-stage renal disease: a reexamination. Am J Kid Dis
1989; 13: 200±209
29. Killingworth A. Psychosocial impact of end-stage renal disease.
Br J Nurs 1993; 2: 905±908
30. Morton AR, Meers C, Singer MA, et al. Quantity of dialysis:
quality of lifeÐwhat is the relationship. ASAIO J 1996;
42: M713±M717
31. Moreno F, Aracil FJ, Perez R, Valderrabano F. Controlled
study on the improvement of quality of life in elderly
hemodialysis patients after correcting end-stage renal disease
related anemia with erythropoietin. Am J Kid Dis 1996;
27: 548±556
Received for publication: 28.10.99
Accepted in revised form: 19.2.01