Kidney International, Vol. 57 (2000), pp.
351–365
Optimization of pre-ESRD care: The key to improved
dialysis outcomes
Principal discussant: Brian J.G. Pereira
New England Medical Center and Tufts University School of Medicine, Boston, Massachusetts, USA
CASE PRESENTATIONS
Patient 1
A 56-year-old man presented to the emergency room at New
England Medical Center in May 1998 with nausea, vomiting,
hemoptysis, pulmonary edema, and altered sensorium. A pri-
mary care physician had seen him for hypertension and renal
insufficiency on a few occasions in the previous five years, but
he was lost to follow-up in the last year. At presentation, he
was somnolent and his blood pressure was 170/120 mm Hg.
Positive findings on examination included a pericardial friction
rub, diffuse pulmonary crackles, mild anasarca, and asterixis.
Laboratory tests revealed a hematocrit of 29%; blood urea
nitrogen, 236 mg/dL; serum creatinine, 15.4 mg/dL; albumin,
3.3 g/dL; calcium, 6.3 mg/dL; and phosphorus, 7.4 mg/dL. The
arterial blood pH was 7.14; pO2, 133; pCO2, 12; and bicarbonate,
4 mEq/liter. He was admitted to the Intensive Care Unit (ICU),
where a subclavian dialysis catheter was placed and hemodialy-
sis started. After three days in the ICU, he was transferred to
the ward. He remained hospitalized for 15 days, during which
his cardiovascular and fluid status were stabilized, permanent
access was placed, and he underwent education regarding dial-
ysis and the management of uremic complications.
Patient 2
A 42-year-old woman with the established diagnosis of au-
tosomal-dominant polycystic kidney disease was referred by
The Nephrology Forum is funded in part by grants from Amgen,
Incorporated; Merck & Co., Incorporated; Dialysis Clinic, Incorpo-
rated; AstraZeneca LP; and R & D Laboratories.
Key words: renal replacement therapy, dialysis, osteodystrophy, ESRD
costs and outcomes.
 2000 by the International Society of Nephrology
351
NEPHROLOGY FORUM
her primary care physician to the Nephrology Clinic of the
New England Medical Center 11 years ago for control of hyper-
tension. At her initial visit, her blood pressure was 190/120
mm Hg and the serum creatinine was 1.3 mg/dL. Treatment
with enalapril was begun and the patient was followed several
times per year. Seven years ago, she was enrolled in the Modifi-
cation of Diet in Renal Disease (MDRD) Study and was ran-
domly assigned to the high-protein group (estimated protein
intake, 1.1 to 1.2 g/kg/day). At enrollment, her weight was 76
kg; BUN, 30 mg/dL; and serum creatinine, 2.1 mg/dL. Six years
ago, the serum calcium was 9.7 mg/dL and phosphorus 5.4
mg/dL; calcium carbonate therapy was started. Three years
ago, the BUN had risen to 76 mg/dL, the serum creatinine to
5.4 mg/dL, and the serum albumin was 4.1 mg/dL. Four months
later, iron supplementation and erythropoietin were begun
when her hematocrit was noted to be 30%. After reading litera-
ture provided by the nephrology team on modalities of renal
replacement therapy, she chose peritoneal dialysis. Mild uremic
symptoms appeared four months later. The blood pressure
was 130/82 mm Hg; hematocrit, 36%; BUN, 83 mg/dL; serum
creatinine, 6.9 mg/dL; albumin, 4.2 g/dL; calcium, 8.9 mg/dL;
and phosphorus, 4.2 mg/dL. A peritoneal dialysis catheter was
placed, and dialysis was initiated one month later. The patient
initially performed continuous ambulatory peritoneal dialysis
and subsequently received a cadaveric renal transplant.
DISCUSSION
Dr. Brian J. G. Pereira (Vice Chairman, Department
of Medicine, New England Medical Center, and Professor
of Medicine, Tufts University School of Medicine, Boston,
Massachusetts, USA): These two cases exemplify the ex-
tremes in caring for patients who are approaching end-
stage renal disease (ESRD). The first patient essentially
received no pre-ESRD care because he was lost to fol-
low-up. There is no indication that strategies for delaying
progression of renal disease were employed. Also, con-
trol of his blood pressure was poor, and he received no
treatment for anemia, malnutrition, or renal osteodystro-
phy. The patient was never seen by a nephrologist and
thus did not have the opportunity to make an informed
choice about the modality of renal replacement therapy.
Dialysis access could not be placed in a timely manner,
and dialysis needed to be started urgently with a tempo-
rary catheter. The second patient’s case represents opti-
mal pre-ESRD care by current standards.
352 Nephrology Forum: Optimization of pre-ESRD care
Worldwide, the number of patients with end-stage re-
nal disease receiving renal replacement therapy (RRT)
is growing. In the United States, the number of patients
enrolled in the Medicare-funded program has grown sub-
stantially, from approximately 10,000 beneficiaries in
1973 to 86,354 in 1983, and to 304,083 as of December 31,
1997 [1]. These figures represent a more than doubling
in the prevalence of ESRD during the last decade and
correspond to an annualized increase of approximately
10%. The rising prevalence of treated ESRD can be
attributed primarily to the increase in the number of
individuals added to the rolls of treated ESRD patients
each year. In fact, 79,102 new patients entered the U.S.
ESRD program in 1997 alone [1]. Patients with ESRD
consume a disproportionate share of national health care
resources. In 1997, the total cost of care of U.S. patients
with ESRD was estimated to be $15.62 billion [1]. De-
spite the magnitude of the resources committed to the
treatment of ESRD patients, these patients continue to
experience significant morbidity and a reduced quality
of life. Hospitalization rates for patients with ESRD
are much higher than those for age- and risk-adjusted
comparative cohorts. In addition, the mortality rate
among dialysis patients remains high. The life expectancy
of dialysis patients in the U.S. is only 16% to 37% of
the age-, gender-, and race-matched population [2].
Analyses of historical data from databases maintained
by single centers, dialysis provider “chains,” and govern-
mental organizations have demonstrated a strong associ-
ation between the risk of death and factors such as age,
gender, race, non-renal co-morbidity, nutrition, and dose
of dialysis [3–5]. However, despite increasing attention
to modifiable factors, the mortality rate among ESRD
patients remains high. Hence, other factors might be
operative. The quality of care prior to initiation of dial-
ysis (“pre-ESRD care”) is a factor that thus far has re-
ceived scant attention, but which theoretically could sig-
nificantly affect morbidity and mortality in dialysis
patients. A successful strategy for improving pre-ESRD
care requires: (1) establishing the limits of renal function
that define “pre-ESRD”; (2) identifying indices of pre-
ESRD care, based on a demonstrated association be-
tween these indices and clinical and economic outcomes;
(3) elucidating the prevalence and predictors of subopti-
mal pre-ESRD care; (4) determining and correcting the
causes of suboptimal pre-ESRD care; and (5) estimating
the size of the pre-ESRD population and the resources
required to provide optimal pre-ESRD care.
The limits of renal function that constitute the pre-
ESRD stage and the size of the pre-ESRD population
are currently not well established. While pre-ESRD un-
questionably ends with the initiation of RRT, the renal
function at which a patient with chronic renal failure
(CRF) could be classified as pre-ESRD has not been
agreed upon. A 1994 National Institutes of Health (NIH)
Consensus Statement on Morbidity and Mortality of Di-
alysis recommended that referral of a patient to a
nephrologist should occur when the serum creatinine
level reaches 1.5 mg/dL in women and 2.0 mg/dL in men
[6]. These levels of renal function can be used to define
entry into the pre-ESRD phase. A recent analysis of the
population-based National Health and Nutrition Exami-
nation Survey III has estimated that the number of indi-
viduals in the U.S. with serum creatinine levels above
1.5 mg/dL, 1.7 mg/dL, and 2.0 mg/dL are 10.9 million,
3.0 million, and 0.8 million, respectively [7].
Indices of “optimal pre-ESRD care” have been nei-
ther clearly defined nor validated. This deficiency stands
in contrast to the well-established indices of “optimal
dialysis care,” such as those developed by the U.S. Health
Care Financing Administration (HCFA) Health Care
Quality Improvement Program (ESRD Core Indicators
Project) in 1994 [8] and, more recently, the National
Kidney Foundation’s Dialysis Outcomes Quality Initia-
tive Guidelines (NKF-DOQIe) [9]. The identification
of well-defined indices of optimal pre-ESRD care has
been handicapped by the paucity of studies exploring the
link between measures of pre-ESRD care and outcomes
after the start of RRT. Nonetheless, several factors could
be used to define the quality of pre-ESRD care (Fig. 1)
[10]. I will restrict this review to selected indices of qual-
ity of pre-ESRD care. For each index, I will address the
rationale for its selection as an index of quality of care,
examine its current status in the pre-ESRD population,
and make recommendations for optimization of care.
Early detection of renal insufficiency and interventions
for slowing the rate of progression of renal disease
Renal function progressively declines in most patients
with renal disease after sufficient damage has raised the
serum creatinine level to 1.5 to 2.0 mg/dL. Indeed, in the
Modification of Diet in Renal Disease (MDRD) study of
840 patients with a variety of non-diabetic renal diseases
and an initial glomerular filtration rate (GFR) ranging
from 13 to 55 mL/min, GFR declined in 85% of patients
and was stable in the remainder [11]. The average rate
of decline was 4 mL/min/year. The benefit of angiotensin-
converting-enzyme (ACE) inhibitors [12–14], blood pres-
sure control [15–18], and blood glucose control (among
diabetics) [19, 20] in retarding the progression of renal
failure in patients with renal insufficiency and protein-
uria has been well established. The benefit of low-protein
diets is controversial [21–23], but a meta-analysis of clini-
cal trials suggests that dietary protein restriction retards
the development and progression of diabetic and nondia-
betic renal disease [24].
Limited information is available in the U.S. regarding
early detection of renal disease and interventions to de-
lay its progression. McClellan et al conducted a retro-
spective chart review of 587 Medicare patients admitted
 Nephrology Forum: Optimization of pre-ESRD care 353

Fig. 1. Components of optimal pre-end-stage renal disease (ESRD) care. Abbreviations are: CRF, chronic renal failure; RRT, renal replacement
therapy; ACE, angiotensin-converting enzyme; BP, blood pressure [32].

Table 1. Early detection of renal disease and use of interventions to delay its progression in the population at risk for chronic renal failurea

Diabetes mellitus Hypertension

% %
Tests for evidence of renal disease
Microalbuminuria 1.6 0.6
Urinalysis 68 58
Serum creatinine 91 97
Renal disease documented in the discharge summary among patients with
Proteinuria $11 7.8 12.5
Serum creatinine $1.5 mg/dL 9.8 11.4
ACE inhibitors prescribed at discharge among patients with
Proteinuria $11 33.3 12.5
Serum creatinine $1.5 mg/dL 31.7 25.7
Nonsteroidal anti-inflammatory drugs prescribed at discharge 6.0 8.8
a
Adapted from Ref. 24. Data comprise a retrospective chart review of 587 Medicare patients younger than 75 years

to the hospital with a primary diagnosis of diabetes melli-
tus or hypertension in the state of Georgia [25]. The
salient findings of this study were that: (1) testing for
renal disease among individuals at high risk for renal
disease was not universal; (2) medical records of patients
with abnormal renal function tests documented neither
awareness of the condition nor plans for further evalua-
tion of the possible underlying renal disease; (3) strate-
gies for delaying progression were infrequently employed;
and (4) nephrotoxic medications such as nonsteroidal anti-
inflammatory drugs continued to be prescribed to pa-
tients with abnormal renal function (Table 1). The results
of this study probably reflect widespread suboptimal clin-
ical practice and are consistent with a survey of physi-
cians’ “self-reported behaviors,” which revealed limited
awareness of and compliance with published guidelines
for the care of patients with diabetes [26].
Prevention or attenuation of uremic complications
Hypoalbuminemia. Hypoalbuminemia at initiation of
dialysis is a strong predictor of early death on dialysis.
Studies involving U.S. and Canadian patients have re-
vealed a strong association of low serum albumin level
at the start of dialysis with subsequent increased risk of
death on dialysis (Fig. 2) [27, 28]. These results are not
surprising, as several studies have shown that hypoal-
buminemia is an important predictor of mortality among
patients who are already undergoing dialysis [29]. Hypo-
albuminemia also has been associated with increased
rates of hospitalization among patients with ESRD [30].
Hypoalbuminemia among patients with CRF accrues
from multiple causes [31, 32]. Declining renal function
is associated with multiple derangements in protein me-
tabolism and spontaneous dietary protein restriction due
to loss of appetite. These alterations lead to loss of lean
354 Nephrology Forum: Optimization of pre-ESRD care
Fig. 2. Malnutrition at the start of dialysis is associated with increased
mortality. Based on data from the Case-Mix Adequacy Study of the
United States Renal Data System, which included 2897 hemodialysis
(h) and 666 peritoneal dialysis ( ) patients who began dialysis in
1986–87 [27].
Fig. 3. Serum albumin levels at initiation of dialysis. Based on data
from 110,843 patients who began dialysis in the United States between
1995 and 1997 [36]. Mean, 3.2 g/dL; median, 3.3 g/dL.
body mass and to increased essential amino acid and
nitrogen requirements. Decreased production of albu-
min due to liver dysfunction, excessive losses due to
nephrotic syndrome, or fluid shifts between the intracel-
lular and the extracellular compartments also can influ-
ence serum albumin levels. Finally, acute and chronic
inflammatory states (perhaps including CRF) can mark-
edly influence the serum albumin concentration [33, 34].
The prevalence of pre-ESRD hypoalbuminemia is
high among patients beginning RRT in the U.S. The U.S.
Renal Data System Case Mix Severity Study reported a
50% prevalence of hypoalbuminemia among patients
who began dialysis in the U.S. in 1986 and 1987 [35].
We recently examined the prevalence of, and factors
associated with, hypoalbuminemia at initiation of dialysis
among patients who began dialysis in the U.S. between
1995 and 1997 [36]. The serum albumin was below the
recorded lower limit of normal in 60% and # 3.5 g/dL
in 67% (Fig. 3). The odds of having hypoalbuminemia
were higher among patients who were older than 65
years, female, non-white, diabetic, covered by non-pri-
vate insurance or uninsured, unemployed, starting hemo-
dialysis (as opposed to peritoneal dialysis), and treated
in certain geographic regions. Similarly, in a recent study
of patients who began hemodialysis at the New England
Medical Center, Boston, between 1992 and 1997, the
adjusted odds of having hypoalbuminemia were greater
among African-Americans, patients with diabetic ne-
phropathy, and patients whose predicted GFR was less
than 5 mL/min/1.73 m2 at the initiation of dialysis [37].
This association between hypoalbuminemia and demo-
graphic, socioeconomic, and geographic factors suggests
that differences in access to care and practice patterns
contribute to the high prevalence of hypoalbuminemia
among patients beginning dialysis.
Practice patterns in the U.S., which are potentially
modifiable, could affect the high prevalence of malnutri-
tion in patients beginning dialysis. First, patients who
are referred late to a nephrologist are more likely to
present with hypoalbuminemia at the start of dialysis
than are patients referred earlier [37]. Whether a differ-
ence exists in long-term outcomes between patients re-
ferred early versus late is not yet known. Second, an
analysis of the Dialysis Morbidity and Mortality Study
(DMMS) Wave 2 demonstrated that only one-half of
the patients who began dialysis were evaluated by a
trained dietitian during the pre-ESRD phase of CRF [2].
Adequate pre-ESRD nutritional care, as recommended
in the 1993 NIH Consensus Statement, involves an early
nutritional assessment by a trained dietitian; specific di-
etary recommendations for protein, energy, carbohy-
drate, lipid, fluid, and sodium and phosphate intake; and
continued monitoring of nutritional and fluid status [6].
A protein intake of 0.7 to 0.8 g/kg/day, the minimum
required dietary allowance for adults, is recommended
unless malnutrition is present, when protein intake
should be 1.0 to 1.2 g/kg/day. Urinary urea nitrogen
appearance can be calculated to estimate protein cata-
bolic rate and to monitor protein intake [38].
These considerations notwithstanding, hypoalbumi-
nemia in CRF might not merely reflect inadequate pro-
tein intake and consequently might not be corrected by
nutritional intervention alone [34]. Malnutrition, cardio-
vascular disease, and inflammation can be linked in the
pathogenesis of hypoalbuminemia. Research is needed
to: (1) further elucidate the mechanism of hypoalbumi-
nemia in patients with CRF; (2) determine whether pre-
ESRD dietary intervention improves serum albumin;
and (3) evaluate whether improved serum albumin is
associated with improved outcomes in patients with
ESRD. Meanwhile, timely use of interventions such as
modification of cardiovascular risk factors, adequate
control of diabetes, and optimal control of uremic com-
plications could ameliorate co-morbid conditions and
 Nephrology Forum: Optimization of pre-ESRD care 355

potentially reduce the prevalence of hypoalbuminemia

prior to the initiation of RRT.
Severe anemia. Several issues regarding the manage-
ment of anemia in pre-dialysis patients require clarifica-
tion. These include determining the following: (1) long-
term effect of correction of anemia with recombinant
human erythropoietin (rHuEPO), including the effect
of partial regression of left-ventricular hypertrophy
(LVH) on the frequency of cardiac complications; (2)
optimal hematocrit level; (3) appropriate patient selec-
tion for rHuEPO therapy; (4) cost-effective rHuEPO
dosing strategies; and (5) clinical predictors of develop-
ment or exacerbation of hypertension.
Heart disease is the leading cause of death among
patients with ESRD. Left-ventricular hypertrophy, pres-
ent in as many as 74% of patients at initiation of RRT,
and cardiac failure are important predictors of cardiac
morbidity and mortality among patients undergoing dial-
ysis [39, 40]. Foley and colleagues reported that among
patients with a left-ventricular mass greater than the
mean for the study population at initiation of dialysis,
the mortality rate was higher than that in patients with
a left-ventricular mass below the mean (Fig. 4A) [41].
Patients with cardiac failure at the initiation of dialysis
had a 79% higher relative risk of mortality compared
with patients without cardiac failure (Fig. 4B). Anemia is
independently associated with the development of LVH
and other cardiac complications among patients with
CRF [40, 42]. Treatment of anemia with rHuEPO results
in partial regression of LVH, both among dialysis and
pre-ESRD patients [43]. Further, rHuEPO has reduced
the frequency of cardiac complications such as CHF and Fig. 4. (A) Predicted survival (Cox proportional hazards) for patients
at initiation of dialysis with left-ventricular mass index above and below
the number of days of hospitalization among dialysis the gender-adjusted mean (reprinted with permission [41]). (B) Esti-
patients [44]. mated survival (Cox proportional hazards) for patients at initiation of
We recently examined hematocrit levels immediately dialysis with and without cardiac failure at baseline (reprinted with
permission [41]).
prior to initiation of chronic dialysis among patients who
began dialysis between 1995 and 1997 in the United
States and found that 51% of patients had a pre-dialysis
hematocrit of ,28% (Fig. 5) [36]. The odds of having a declines below 80% of the mean level for the gender- and
hematocrit ,28% at the initiation of dialysis were higher age-matched population [45]. Oral iron supplementation
among females, non-whites, those covered by non-private and correction of nutritional deficiencies are appropriate
insurance or uninsured patients, and those whose initial initial steps; worsening of anemia should prompt rHuEPO
modality was hemodialysis (as opposed to peritoneal therapy. The hematocrit threshold at which rHuEPO ther-
dialysis). Overall, only 23% of patients had received apy should be started in pre-ESRD patients and the
rHuEPO prior to initiation of dialysis. The association ideal maintenance level have not been established. A
between anemia and demographic and socioeconomic hematocrit level under 30% clearly justifies rHuEPO
factors, and the low prevalence of rHuEPO use suggest therapy. The DOQI recommendation of a target hema-
that differences in access to care and practice patterns tocrit of 33% to 36% for patients on dialysis [45] appears
contribute to the high prevalence of anemia among pa- appealing. However, the efficacy and safety of this target
tients beginning RRT. Whether anemia prior to initia- have not been established. The recommendation of
tion of dialysis is associated with an increased mortality transferrin saturation .20% and ferritin .100 ng/ml for
rate remains to be established. dialysis patients could be applied to pre-ESRD patients,
Patients with CRF should be monitored regularly for as adequate iron stores are necessary to sustain erythro-
declining hematocrit, and a thorough investigation for poiesis in response to rHuEPO.
non-renal causes of anemia is warranted if the hematocrit Concerns have been raised regarding the safety of pre-
356 Nephrology Forum: Optimization of pre-ESRD care
ESRD rHuEPO therapy, specifically, the risk of worsen-
ing hypertension and accelerated deterioration of renal
function. Treatment with rHuEPO can elevate blood
pressure or require an increase in antihypertensive medi-
cation [46]. However, clinical trials have shown that renal
failure does not progress when blood pressure control
is maintained during rHuEPO therapy [47]. Another
important consideration regarding the widespread use
of pre-ESRD rHuEPO therapy is its cost [48]. The addi-
tion of pre-dialysis patients to the pool of those receiving
rHuEPO therapy might significantly raise total health
care costs. Consequently, policies aimed at maximizing
the cost-effectiveness of this therapy are needed.
Renal osteodystrophy. Reduced renal function is asso-
ciated with a variety of bone disorders including osteitis
fibrosa (the hallmark lesion of secondary hyperparathy-
roidism), osteomalacia (due to vitamin D deficiency or
excess aluminum), aplastic bone disease (excess alumi-
num or oversuppression of parathyroid hormone pro-
duction with calcitriol), or mild and mixed lesions [49].
Metabolic acidosis also can contribute to renal osteodys-
trophy by stimulating osteoclastic activity or by causing
direct physicochemical dissolution of calcium [50]. Renal
osteodystrophy, although rarely symptomatic, can cause
bone pain and an increased risk of fractures.
The abnormalities that lead to renal osteodystrophy
occur early in renal insufficiency. Increased parathyroid
hormone (PTH) levels have been demonstrated when
the GFR falls below 60 to 80 mL/min, and levels continue
to increase as renal function declines [51, 52]. Histologic
changes of renal osteodystrophy are manifest early in
renal insufficiency. Among patients with creatinine clear-
ances of 20 to 59 mL/min, 87% of them had abnormal
bone histology [53].
The skeletal changes as well as parathyroid hyperpla-
sia are not easily reversed with currently available thera-
pies. Thus, early interventions to prevent secondary hy-
perparathyroidism and control metabolic acidosis are
Fig. 5. Hematocrit levels at initiation of dial-
ysis. Based on data from 131,484 patients who
began dialysis in the United States between
1995 and 1997 [36]. Mean, 27.9%; median,
27.9%.
critical. Hyperphosphatemia and hypocalcemia must be
corrected by modifying the diet, utilizing phosphate
binders, and administering calcitriol. Dietary phospho-
rus should be restricted at the earliest manifestations of
azotemia, at the latest before the creatinine clearance
falls below 40 mL/min, without awaiting the develop-
ment of hyperphosphatemia [49, 54, 55]. Therapy with
calcium-containing phosphate binders should be initi-
ated when dietary manipulation alone is insufficient to
maintain the serum phosphate at a normal level. Alumi-
num-containing binders should be reserved for extreme
hyperphosphatemia despite calcium and calcitriol ther-
apy, or if hypercalcemia is present. There is little experi-
ence with the newer calcium-free phosphate binders, but
these agents are likely to be useful in the management
of such patients.
The use of vitamin D supplements during the pre-
ESRD phase is controversial because of concerns that
progression of renal disease will be hastened [56].
Calcitriol should be used with appropriate vigilance to
prevent the development of an elevated calcium-phos-
phorus product or hypercalcemia and oversuppression
of PTH. Doses of calcitriol lower than 0.25 to 0.5 mg/day
should provide safe but adequate suppression of PTH
(goal PTH level, 100 to 300 pg/mL) [57, 58].
Optimal control of cardiovascular complications
The mortality rates due to cardiovascular disease are
10 to 20 times higher among ESRD patients treated with
dialysis compared to the general population [39], and car-
diovascular disease accounts for nearly one-half of the
deaths among ESRD patients [3]. Morbidity due to cardio-
vascular disease also is substantial. The Canadian Hemo-
dialysis Morbidity Study reported a probability of hospital
admission for nonfatal myocardial infarction (MI), angina,
or pulmonary edema among incident patients of 8% to
10% per year [30], and in the Hemodialysis (HEMO)
 Nephrology Forum: Optimization of pre-ESRD care 357

Table 2. Prevalence of cardiovascular disease among patients beginning renal replacement therapy a

ESRD Registry studies Cohort study

USRDSb[35] USRDSc[2] CORRd[62] ANZDATAe[63] Foley[41]
Year 1986–87 1996 1988–1996 1994–1997 1982–1991
Number of patients Population 3,399 3,468 24,497 5602 433
Coronary artery disease All 40.8% 36% (PD)f 37.2% 14.1%
Myocardial infarction 42% (HD)
Diabetic ,45 years 9.2%
Diabetic .45 years 24.0%
Non-diabetic ,45 years 2.2%
Non-diabetic .45 years 19.2%
Angina All 18.8%
Congestive heart failure All 41.1% 31% (PD) 30.8%
40% (HD)
Left-ventricular hypertrophy All 30.9% 18% (PD) 74.0%
22% (HD)
Cardiomegaly All 38.1%
a
Table modified from Ref. 64
b
USRDS[35], United States Renal Data System (USRDS) Case Mix-Severity Study
c
USRDS[2], Dialysis Morbidity and Mortality Study Wave 2
d
CORR[62], Canadian Organ Replacement Register
e
ANZDATA[63], Australia and New Zealand Dialysis and Transplant Registry
f
PD, peritoneal dialysis; HD, hemodialysis

Study, approximately one-third of first hospitalizations
were related to cardiovascular complications [39].
Clinically apparent cardiovascular disease at onset of
RRT is a strong predictor of subsequent cardiac morbid-
ity and mortality. Patients with CHF at initiation of RRT
have a higher likelihood of subsequent episodes of CHF
compared to patients initially free of CHF, and the former
also are at increased risk of all-cause mortality [35, 59].
Abnormalities in left-ventricular geometry on echocardi-
ogram at initiation of RRT also are strong predictors
of subsequent cardiac morbidity and mortality among
ESRD patients [18, 60].
Cardiovascular disease begins early in the course of
CRF and is present at the onset of dialysis in a substantial
fraction of patients (Table 2). The prevalence of echocar-
diographic evidence of LVH appears to be directly re-
lated to the severity of CRF. Among patients with creati-
nine clearances less than 25 to 30 mL/min, LVH has
been reported to be present in 38% to 45% of patients,
compared to a prevalence of 16% to 31% among patients
with creatinine clearances above that level [40, 61].
Traditional risk factors for cardiovascular disease—
older age, male gender, diabetes mellitus, family history,
smoking, high blood pressure, LVH, high total choles-
terol levels, low high-density lipoprotein (HDL) levels,
physical inactivity, and menopause [65]—are common
among patients with CRF. Diabetes mellitus is the most
common cause of ESRD in the U.S., accounting for 42%
of new cases in 1996 [3]. At the onset of RRT, 75% to
100% of patients have hypertension [66]. Furthermore,
even among treated patients, blood pressure control is
often suboptimal [67]. Lipid abnormalities also are com-
mon, particularly among patients with nephrotic syn-
drome [68]. In addition, factors specific to the uremic
state (anemia, arteriovenous fistula, hypervolemia, hy-
perparathyroidism, and malnutrition) contribute to the
risk for cardiovascular disease among patients with CRF
[68]. Other non-traditional risk factors—elevated homo-
cysteine, triglyceride, and Lp(a) levels—have a less well-
defined role in the pathogenesis of cardiovascular disease
[68, 69].
In summary, patients with CRF constitute a high-risk
group for cardiovascular morbidity and mortality. Con-
sequently, aggressive risk factor modification in the pre-
ESRD stage could reduce the prevalence of cardiovascu-
lar disease among patients reaching ESRD and could
improve outcomes among ESRD patients. The National
Kidney Foundation Task Force on Cardiovascular Dis-
ease recently published recommendations regarding car-
diovascular risk interventions in patients with CRF and
ESRD; these are summarized in Table 3 [70].
Adequate preparation for renal replacement therapy
Modality selection. Patient noncompliance with the
dialysis prescription is common in the U.S. and contrib-
utes to suboptimal dialysis delivery [2]. Education and
informed choice of dialysis modality likely will enhance
patient compliance, reduce the need for subsequent
modality changes, and facilitate the patient’s continued
participation in the work force.
An informed decision on dialysis modality made under
non-emergency circumstances reduces the likelihood of
a subsequent change in dialysis modality. This process
is generally facilitated by early referral to the nephrolo-
gist [32, 71]. Patients referred to the nephrologist for
pre-ESRD care also are more likely to select peritoneal
dialysis compared to patients referred late [2]. Other
important determinants of modality selection include fi-
358 Nephrology Forum: Optimization of pre-ESRD care

Table 3. Primary and secondary prevention of cardiovascular disease (CVD) in chronic renal failurea

Risk factor Rationale/target Intervention

Hypertension Volume control (“dry weight”) Salt restriction
Blood pressure control goals Diuretics
,125/75 mm Hg in patients with proteinuria Antihypertensive agents (ACE inhibitors are suggested)
,130/85 mm Hg in patients without proteinuria
Hyperlipidemia LDL #100 mg/dL Diet
HMG CoA reductase inhibitors (“statins”)
Hyperglycemia Normal or nearly normal blood glucose levels Diet
Oral agents
Insulin
Tobacco use Discontinue use Counseling
Nicotine replacement therapy
Physical activity Moderate physical activity (30 minutes per day, Regular exercise program
most days of the week)
Menopause Relationship between menopause and CVD in CRF Hormone replacement therapy
is poorly understood
Individual decision-making recommended
Homocysteine Effect of dietary fortification with folic acid on B vitamins (folic acid, B6, B12)
homocysteine levels, and the effect of lowering
homocysteine on cardiac risk are unknown
Thrombogenic factors Decreased platelet function and elevated procoagu- Aspirin (75–325 mg/day) is reasonable, but can worsen
lant activity platelet function
Revascularization Relief of ischemia, especially in high-risk patients CABG, PTCA (with stent)
LVH Hypertension and anemia are associated with the Diuretics and antihypertensive agents
development of LVH. Treatment of each con- Erythropoietin
dition causes regression of LVH, but the effect
of clinical cardiac events is yet unknown
Blood pressure target as above
Hematocrit target of 33%–36% as recommended
by the NKF-DOQIe
Screening
Stress tests Insufficient evidence to recommend routine stress Stress tests to detect inducible ischemia
tests. Pre-operative screening for non-cardiac
surgery and renal transplantation in high-risk
patients
Echocardiography Insufficient evidence to recommend for or against Echocardiography to assess left-ventricular mass
routine echocardiography for detection of
LVH
a
Abbreviations are: ACE, angiotensin-converting enzyme; LVH, left-ventricular hypertrophy; CVD, cardiovascular disease; CRF, chronic renal failure; CABG,
coronary artery bypass graft; PTCA, percutaneous transluminal coronary angioplasty; NKF-DOQIe, National Kidney Foundation Dialysis Outcomes Quality
Initiative. Modified from the recommendations of the National Kidney Foundation Task Force on Cardiovascular Disease [70].

nancial and reimbursement policies, physician prefer-
ence, resource availability, and the social background
and cultural habits of the patient.
Pre-ESRD education and intervention can play an
important role in modality selection and ensuring contin-
uation of employment. A significantly higher proportion
of peritoneal dialysis patients, a greater proportion of
whom are under the care of a nephrologist prior to
ESRD, take the lead in selecting dialysis modality com-
pared to hemodialysis patients (Fig. 6) [2]. In a study of
an educational intervention program, participants who
received pre-ESRD education were more than twice as
likely to continue employment after the start of dialysis
than were non-participants [72].
Discussions regarding the impending need for RRT
and choice of modality should be initiated at least six
months prior to the anticipated start of RRT. The renal
team should assess the patient’s suitability for dialysis
and transplantation and provide realistic advice regard-
ing therapeutic options. A detailed evaluation should
assess social aspects of RRT, including the patient’s mo-
bility, housing conditions (especially if home dialysis is
chosen), family support, continuation of employment,
transportation to the dialysis unit, and financial issues.
Dialysis access placement. Patients who select perito-
neal dialysis should have placement of a cuffed perito-
neal dialysis catheter two to four weeks prior to the
anticipated need for dialysis to reduce the possibility of
temporary hemodialysis [73]. Patients who select hemo-
dialysis should have a functioning permanent vascular
access at the initiation of RRT. A native arteriovenous
fistula (AVF) is the preferred type of permanent vascular
access. Patients with AVFs have fewer complications
related to infection and thrombosis, and possibly better
survival rates compared to patients with arteriovenous
grafts (AVGs) [30, 73–76]. Percutaneous catheters have
the highest complication rates and also result in a lower
delivered dose of dialysis [77, 78]. Elective outpatient
placement of a permanent dialysis access obviates the
 Nephrology Forum: Optimization of pre-ESRD care
Fig. 6. Physician and patient roles in selection of dialysis modality.
Choice was made by: the patient ( ); medical team, (h); or a collabora-
tion of both (j). Based on data from 2400 patients in the Dialysis
Morbidity and Mortality Study of the United States Renal Data
System [2].
need for hospital admission for initiation of dialysis, and
reduces patient discomfort and cost [79].
Several recent studies have documented that fewer
than one-half of patients beginning dialysis in the U.S.
have a permanent vascular access in place for their first
treatment [2, 80]. In addition, despite the recognized
superiority of AVF, there has been a trend towards more
frequent placement of AVGs in the past decade, and
more than 50% of patients required the use of a tempo-
rary (uncuffed) catheter within the first 60 days of hemo-
dialysis [2]. There are several possible explanations for
the popularity of AVGs. These include: (1) ease of place-
ment in numerous anatomic locations; (2) late referral
to a nephrologist, which precludes planned placement
of an AVF; and (3) preferences and technical skills of
access surgeons. The fact that many patients do not recall
receiving instructions to “save” an arm for hemodialysis
access is further evidence of suboptimal planning for
permanent vascular access in the U.S. [2].
Education regarding vascular access should begin
when the serum creatinine exceeds 3 mg/dL [73]. Patients
and staff should be instructed to avoid phlebotomies
and intravenous lines in the cephalic veins as well as
subclavian vein cannulation. The DOQI guidelines rec-
ommend placement of an AVF when the serum creati-
nine exceeds 4 mg/dL, creatinine clearance falls below
25 mL/min, or ESRD is anticipated within one year. A
maturation time of at least three months is recommended
[73]. Only patients who are not candidates for an AVF
or in whom an attempt has failed should have placement
of an AVG, and these grafts ideally should not be used
for one month after surgery.
Timing of initiation of dialysis. The rationale for
timely initiation of RRT in patients with CRF has re-
cently been reviewed [32]. Several investigators have
argued that patients who begin dialysis at relatively
higher levels of residual renal function (early start) have
359
less morbidity and later mortality compared with pa-
tients who begin dialysis at more traditional low levels
of renal function (late start). Although evidence from
controlled clinical trials is lacking, several observations
support this hypothesis. First, early initiation of dialysis
protects against malnutrition during the pre-ESRD pe-
riod and might prevent the increased morbidity and early
mortality related to malnutrition among dialysis patients.
Second, patients in whom dialysis is begun at a lower
GFR than the optimal solute clearance levels for dialysis
patients might suffer the same increased morbidity and
earlier mortality as the dialysis patients with poor solute
clearances [32, 81]. Third, suboptimal control of volume
overload as well as decreased clearance of phosphate,
advanced glycation end products, b2-microglobulin, and
other uremic toxins can lead to increased complication
rates. Finally, uncontrolled clinical studies from Italy in
the mid-1970s and the U.S. in the late 1980s suggest
that the severity of certain uremic complications and
mortality and hospitalization rates correlate inversely
with residual renal function at the initiation of dialysis
[81–83].
Limited data are available regarding the level of resid-
ual renal function at initiation of RRT in the U.S. Most
reports are based on levels of serum creatinine and/or
estimated creatinine clearance and are limited to selected
populations. In the MDRD study, GFR was measured
using iothalamate clearance in 94 patients who reached
ESRD during the study period. The mean GFR at ini-
tiation of dialysis was 9.0 mL/min/1.73 m2 [23]. In the
CANUSA study, the mean average of renal creatinine
and urea clearance at initiation of dialysis was 3.8 mL/
min [84]. There has been a steady trend in the U.S.
toward beginning dialysis at lower levels of serum creati-
nine. In the late 1970s, the mean serum creatinine at the
start of dialysis was 14.5 mg/dL [85]; in the 1980s, the
mean serum creatinine concentration was 11.7 mg/dL
[86]. In contrast, the mean serum creatinine was 9.0 mg/
dL in 1993 and 8.5 mg/dL in 1995 to 1997 [32]. We
recently examined GFR levels immediately prior to initi-
ation of chronic dialysis among patients who began dial-
ysis between 1995 and 1997. Glomerular filtration rate
was predicted by a formula derived from the MDRD
study [87]. The mean (SD) predicted GFR was 7.1 mL/
min/1.73 m2 (SD 3.1) with a range from 1 to 42 mL/min/
1.73 m2. The proportion of patients with predicted GFR
.10, 5 to 10, and ,5 mL/min/1.73 m2 was 14%, 63%,
and 23%, respectively. The mean predicted GFR was
significantly lower among younger patients, women, Af-
rican-Americans, patients with higher body weight, pa-
tients with ESRD due to diseases other than diabetes,
uninsured patients, patients who were employed, home-
makers, students, and patients selecting hemodialysis
[88].
The Peritoneal Dialysis Adequacy Work Group of
360 Nephrology Forum: Optimization of pre-ESRD care
the DOQI recommends initiation of dialysis when the
weekly renal Kt/Vurea decreases to below 2.0, unless all
three of the following criteria are fulfilled: (1) stable
or increased edema-free body weight, (2) normalized
protein-equivalent of total nitrogen appearance greater
than 0.8 g/kg/day, and (3) absence of clinical symptoms
and signs attributable to uremia [9]. Potential problems
related to earlier initiation of hemodialysis include hemo-
dynamic instability, vascular-access-related infections,
potential adverse effects on longevity of the vascular
access, and faster decline in residual GFR. Concerns
regarding earlier initiation of peritoneal dialysis include
morbidity associated with peritonitis and potential ad-
verse effects on peritoneal membrane longevity. Finally,
the economic consequences of earlier initiation of RRT
are a major concern in an era of rising health care costs.
QUESTIONS AND ANSWERS
Dr. Nicolaos E. Madias (Chief, Division of Nephrol-
ogy, New England Medical Center, Boston, Massachu-
setts): Thank you, Brian, for this wonderful presentation.
You spoke eloquently about the anticipated benefits to
patients from early referral to nephrologists. Can you
reflect a bit more on the overall impact of early referral
on the health care system in terms of cost and staffing?
Dr. Pereira: You bring up a very important issue,
which is one of the reasons why this debate continues.
Certainly, referral to the nephrologist is likely to increase
cost over the pre-ESRD phase because of the use of
more advanced diagnostic and therapeutic interventions.
However, the incremental resources used over the pre-
ESRD phase could be counterbalanced by: (1) prolonga-
tion of the pre-ESRD phase, with a resultant deferral of
dialysis with its cost of $60,000 per year and (2) lower
first year (and potentially later) cost of ESRD due to
better preparation for dialysis and less morbidity. If pa-
tients were to live significantly longer because of these
interventions, the total cost would increase, but that is
a societal issue and one that arises with any intervention
aimed at keeping patients alive and healthy. Unfortu-
nately, rigorous economic analyses are not available.
The issue of who is going to look after these patients
when the number of available nephrologists is limited is
a major concern. I don’t believe that we have enough
nephrologists to care for the patients who likely will
flood the system as the benefits of early referral become
clear. A possible strategy is for nephrologists to develop
a partnership with primary care physicians and provide
a complete set of recommended medical interventions.
Initially, the nephrologist would develop an individual-
ized management plan. Thereafter the patient would be
seen by the nephrologist approximately once yearly.
When the serum creatinine level reached 4 mg/dL or so,
the nephrologist would assume complete responsibility
for care.
Dr. Madias: I found the evidence you presented about
grossly suboptimal clinical practice at prominent medical
institutions very disturbing and suspect that the situation
is likely worse at institutions of lesser prominence. What
strategies can be used to improve this situation?
Dr. Pereira: Yes, the level of care is quite worrisome.
Several recent reports have shown that pre-ESRD care
at some inner-city hospitals is even worse than the data
presented here suggest. What needs to be done? Some
organizations such as the National Kidney Foundation
have put together an Early Intervention and Prevention
Task Force, which includes representatives of the target
physician groups, to develop a curriculum for primary
care physicians and family practitioners. This approach
is likely to become a necessity not just for nephrology,
but also for every other specialty, beginning at the medi-
cal school level and extending to residency programs.
Dr. John T. Harrington (Dean, Tufts University
School of Medicine, Boston): Brian, thank you for a
superb review. My question relates to the issue of timely
initiation of dialysis. One can’t disagree with the notion
that there should be timely initiation of dialysis. My fear,
however, is that we’re now being pushed into starting
dialysis earlier than needed. If we start dialysis six
months sooner, and patients live longer on dialysis by
six months, then we are faced with the problem of “lead-
time bias,” and not a true benefit in terms of mortality.
All we’ve done is increase the cost to the system. How
do you respond to this question?
Dr. Pereira: John, that is a valid point. The same issue
arose in the breast and prostate cancer literature. If a
condition is diagnosed earlier and the patient lives longer
from the time of diagnosis, lead-time bias must be consid-
ered. There is no water-tight way of dealing with lead-
time bias. One mechanism might be to adjust any analy-
ses of effects of early initiation of dialysis for the esti-
mated duration of time the patient would have remained
off dialysis. This interval could be predicted using an
estimated rate of decline of renal function. The estimate
derived from the MDRD study is a rate of decrease of
GFR of 4 mL/min/year, and is assumed to be constant
[11]. Lead-time bias is a serious concern that deserves
careful attention.
Dr. Harrington: You’ve presented many correlations
of individual variables with outcomes, particularly serum
albumin. Could multivariate analyses of several factors
tell us when we should start dialysis with a degree of
validity stronger than analyses of individual factors?
Dr. Pereira: No data are available that would allow
us to make a prediction regarding who needs to start on
dialysis. Again, it comes back to the issue of lead-time
bias. We have yet to prove that if a patient starts dialysis
when the GFR is 12 mL/min versus a GFR of 8 mL/min,
 Nephrology Forum: Optimization of pre-ESRD care
the survival advantage will be more than the one year
of “lead time,” assuming a rate of decline of 4 mL/min
per year. If by starting with a GFR of 12 mL/min versus
8 mL/min, survival is increased by significantly more
than one year, then it would seem worthwhile. If you’re
buying only the same one year or less, then it’s not worth
it to the patient, and the process actually might result in
greater societal cost. This issue needs to be fully resolved
before widespread recommendations can be made.
Dr. Ronald D. Perrone (Division of Nephrology,
New England Medical Center): Can you dissect out the
components of cost in the first months following the
initiation of dialysis? My hunch is that the increased
costs during this period are related to access issues.
Dr. Pereira: I suspect you are absolutely correct. We
are in the process of ascertaining the cost of inpatient
and outpatient access procedures as well as hospitaliza-
tions and total hospital days attributable to access, in-
cluding access procedures performed during hospitaliza-
tions for another reason. Our preliminary analyses do
reveal that access-related hospitalizations are the major
cause of resource utilization in the first six months of
dialysis.
Dr. Perrone: Could you share your thoughts about
prospective clinical trials in which dialysis would be initi-
ated at different levels of GFR and all else would be
equal, that is, access is in place, erythropoietin is appro-
priately used, etc.?
Dr. Pereira: I am doubtful that there ever will be a
prospective trial. Years ago, when Bloembergen and col-
leagues showed that early mortality was higher in perito-
neal dialysis patients [89], a prospective “randomized”
trial was initiated. However, it was noted that when the
options of dialytic therapy were presented by a person
committed to peritoneal dialysis, a larger proportion of
patients chose peritoneal dialysis. Conversely, in areas
where people are strongly opposed to peritoneal dialysis,
the opposite was seen. Thus, the randomization process
failed, and I believe that biases regarding the correct level
of GFR for initiation of dialysis would lead to a similar
outcome. Dr. Levey might have an opinion on this.
Dr. Andrew S. Levey (Division of Nephrology, New
England Medical Center): I agree that this would be a
hard clinical trial to do. Patients would be eligible for
entry only when their level of renal function was within
a certain range, assuming the level of renal function
could be estimated accurately and conveniently. During
that time, they would have to be randomized into the
study, and early intervention would have to begin before
the renal function fell below the threshold level for inter-
vention in the control group. I would expect a large
number of “drop-ins” and “drop-outs” from the early
intervention group, which would minimize the difference
in outcomes between the randomized groups.
I wanted to ask you two questions about economics
361
and health policy. First, who should take care of patients
with chronic renal insufficiency? What can we learn
about costs and outcomes from our colleagues in other
specialties, such as cardiology or diabetes? Second, I’m
afraid we might be naı̈ve to think that we can avoid the
costs that occur at the time of initiation of dialysis by
starting it earlier. It simply might be expensive to begin
dialysis with temporary access. Of course it would be
desirable to avoid temporary access and hospitalization
for dialysis initiation. But in many cases, the onset of
renal failure is associated with an episode of illness that
requires hospitalization.
Dr. Pereira: Andy, your analogy to the cardiac patient
population is logical, but a myocardial infarction cannot
be predicted, and it is an acute event. At the New Eng-
land Medical Center, only 12% of patients who started
RRT developed ESRD following irreversible acute renal
failure. A typical example of this is the patient who has
diabetes and chest pain and who undergoes catheteriza-
tion, which results in irreversible renal failure. However,
in the vast majority of patients, initiation of dialysis can
be planned. Among these patients in whom renal re-
placement therapy could have been planned, the cost of
initiating dialysis probably could be diminished by early
nephrology referral, and by timely vascular access place-
ment and initiation of dialysis. I strongly believe that
planned initiation would decrease the excessive costs
incurred in the first six months of dialysis.
The first issue you raised is much harder to answer—
who should take care of patients with chronic renal fail-
ure? There are conflicting results on outcomes compar-
ing care between specialists and non-specialists. In the
area of cardiology, one study showed that the one-year
mortality rate was 12% lower among patients with acute
myocardial infarction who were cared for by cardiolo-
gists compared to patients cared for by generalists [90].
In contrast, in another study of patients with unstable
angina, rates of death or myocardial infarction were simi-
lar among patients cared for by non-cardiologists or
cardiologists [91]. In this latter study, cost of care was
$2,000 lower among patients cared for by non-cardiolo-
gists (a non-significant difference) and the length of hos-
pital stay did not differ.
Dr. Andrew J. King (Division of Nephrology, New
England Medical Center): You alluded to the increasing
use of ACE inhibitors before ESRD to control blood
pressure and to retard the progression of renal disease.
Do you have any information regarding the use of other
known preventive measures for cardiovascular disease,
in particular, the use of aspirin or cholesterol-lowering
agents, in the pre-ESRD patient population?
Dr. Pereira: Little data are available in the ESRD
population regarding these other interventions, and we
have even less data in pre-ESRD patients. The National
Kidney Foundation Task Force has published recom-
362 Nephrology Forum: Optimization of pre-ESRD care
mendations regarding the treatment of cardiovascular
disease among patients with chronic renal insufficiency
and ESRD (dialysis and transplantation), generally ex-
trapolating data from patients without renal failure [70].
The recommendation for management of hyperlipidemia
is that physicians follow the National Cholesterol Educa-
tion Program Adult Treatment Panel guidelines for the
general population. These include obtaining a fasting lipid
profile on patients with HDL cholesterol #35 mg/dL or
total cholesterol $200 mg/dL. The target is an LDL
cholesterol level of #100 mg/dL, to be achieved with
dietary or drug intervention, depending on initial level.
Given the increased risk of platelet dysfunction with
lower levels of renal function, the use of aspirin in low
doses is recommended only for patients with mild renal
insufficiency or those with a high risk for cardiovascular
disease. We don’t know the impact of these interventions
and outcomes in patients with renal failure.
Dr. Madias: You referred to malnutrition in uremia
but, as you know, heart disease itself is a cause of malnu-
trition. Could you please comment on the association
between malnutrition and cardiac and renal disease?
Dr. Pereira: The link between inflammation in ure-
mia, malnutrition, and cardiac disease has been an area
of focus in our laboratory. We among others strongly
believe that these factors are interrelated, with inflam-
mation having a strong effect on heart disease and mal-
nutrition, and these latter two factors having an effect
on each other. The possible cause of inflammation in the
pre-ESRD stage is not defined. In the ESRD setting, we
have more information available; dialyzer membranes,
contaminants in the dialysate, and dialysis-related in-
terventions, such as the use of intravenous iron, could
trigger inflammation. Some of us believe that pro-
inflammatory cytokines might be important mediators
of inflammation, with the C-reactive protein being simply
a marker of the inflammatory state. This “inflammatory
state” might suppress albumin gene expression in the
liver and lead to hypoalbuminemia. Inflammation also
can trigger inflammatory and endothelial cell reactions
and can induce atherogenesis in experimental models.
Heart failure itself can reduce appetite, thus contributing
to malnutrition. The role of inflammation in uremia is
an area of active investigation.
Dr. Madias: Is the protein malnutrition incurred in
uremia sufficient to cause depletion of L-arginine, the
substrate of nitric oxide, with resultant cardiovascular
consequences?
Dr. Pereira: I am not certain about this. Perhaps Dr.
Levey has the answer.
Dr. Levey: l-arginine levels were nearly normal in the
MDRD study, except among patients with the lowest
level of renal function, that is, those with a creatinine
clearance below 15 mL/min/1.73 m2, who have l-arginine
levels approximately 20% below normal [92]. In general,
it appeared in that study that the lower the level of
renal function, the more amino acid abnormalities were
present.
Dr. Klemens B. Meyer (Division of Nephrology, New
England Medical Center): As a practicing nephrologist,
I am a bit confused. Only a few years ago the Federal
government accused nephrologists of starting patients
on dialysis too early. I still regularly sign government
forms explaining why we started dialysis in a patient with
an estimated GFR exceeding 10 mL/min. Now you are
suggesting that dialysis be started earlier. Could you
comment on this reversal?
Dr. Pereira: I had the opportunity to be involved in
a case a few years ago in which it was alleged that, to gain
financially from the system, unscrupulous nephrologists
were starting patients on dialysis too early. The USRDS,
which was asked to analyze the centers believed to be
exploiting the system, found that patients who were “be-
ing started on dialysis earlier” (and who had been denied
eligibility by the Medical Advisory Board of the regional
ESRD network) had a mortality rate that was threefold
higher than those who had been approved by the Medical
Advisory Board. An example is the 90-year-old lady who
weighs 40 kilograms and has intractable heart failure but
a serum creatinine of only 2.5 to 3 mg/dL. The Medical
Advisory Board would not be able to decide whether
this patient should be on dialysis. The truth of the matter
is that there was no exploitation of the system. Such
patients were justifiably on dialysis irrespective of their
serum creatinine levels.
Dr. Meyer: This leads to another observation: the
definition of ESRD is becoming less clear. You implicitly
raise questions about how one defines “end-stage” and
whether one should even wait until the “end stage” to
initiate dialysis. Also, there’s the question of whether
we only initiate chronic dialysis for end-stage renal fail-
ure. Isn’t it helpful to think of chronic dialysis as a treat-
ment for a certain level of renal function given the pa-
tient’s cardiac function?
Dr. Pereira: I agree with you. Many times patients
have not reached ESRD. We have no firm definition of
ESRD. We arbitrarily define it as reaching the point at
which one requires renal replacement therapy. When
this end-point is reached is subjective, and it depends
on the clinical characteristics of the patient. For example,
a patient with heart failure and a GFR of 22 mL/min
who requires dialysis would be considered to have ESRD
because RRT would be necessary to manage severe
edema, not because of a particular level of renal function.
Dr. Meyer: One final question. What about the pa-
tients who shouldn’t receive or be subjected to renal
replacement therapy? With all the emphasis on starting
early, is anyone thinking about counseling patients and
families regarding conservative options?
 Nephrology Forum: Optimization of pre-ESRD care
Dr. Pereira: Are you talking about patients who
shouldn’t be on renal replacement therapy at all?
Dr. Meyer: That’s correct.
Dr. Pereira: The Renal Physicians Association (RPA)
and the American Society of Nephrology (ASN) have
formed a Task Force to develop recommendations re-
garding who shouldn’t start on renal replacement ther-
apy and who should be withdrawn. Hopefully these rec-
ommendations will soon be finalized and disseminated.
Dr. Harrington: I have a brief comment. One of the
major reasons general internists haven’t sent patients
earlier to nephrologists over the last 30 years is that
nephrologists have not been successful at retarding pro-
gression. Patient 2 is a good example: the system worked
well, but the patient went into renal failure anyway. The
general internist knows that this happens. I also have a
question. I am surprised by the very low utilization of
erythropoietin; two-thirds of patients start on dialysis with
hematocrits of less than 30%. What is the outcome of
patients when you simply isolate anemia as the variable?
Dr. Pereira: We are investigating that issue. In April
1995, the new Medical Evidence Form went into use,
which required information on laboratory data prior to
the initiation of dialysis. We have recently obtained the
files on outcomes and cost and will be analyzing the
relationship of pre-ESRD hematocrit with mortality and
resource utilization.
Dr. Madias: According to the best available data, the
annual incidence of ESRD is growing by 6% to 7%.
What do we know about the factors that contribute to
this growth?
Dr. Pereira: The causes are speculative. One of the
causes is that patients with diabetes and hypertension
are now living long enough to develop renal failure.
Second, people who previously might not have been
referred by their primary physicians for dialysis are now
being referred. It is becoming more acceptable to dialyze
“marginal” patients, that is, those with poor expected
outcomes. This has prompted the ASN and the RPA to
develop guidelines regarding this issue.
Dr. Harrington: My own bias is that virtually nobody
with severe renal failure escapes dialysis in the U.S.. Has
anyone looked at all patients with a serum creatinine
over 5 mg/dL in this hospital or others to determine
whether they were all started on dialysis?
Dr. Pereira: The major problem in answering this
question is that the population at risk is not well defined.
That is, we don’t know what the denominator is, either
at the New England Medical Center or in the U.S. Re-
searchers have looked at this in the U.K. and Canada,
and they have found that a surprisingly high number of
patients were not considered by the general practitioners
as candidates for dialysis. In the U.S., there is probably
more of an incentive for putting even marginal patients
363
on dialysis, and there is less societal acceptance of non-
intervention.
ACKNOWLEDGMENTS
The Principal Discussant wishes to acknowledge the major contribu-
tions of Drs. Gregorio T. Obrador, Pradeep Arora, and Annamaria
T. Kausz in the development of the concepts reviewed in this Forum
and in the preparation of this manuscript. The help and guidance of
Dr. Andrew S. Levey, Dr. Klemens B. Meyer, Connie Jenuleson, and
Robin Ruthazer are sincerely appreciated. Drs. Philip J. Held, Friedrich
K. Port, Robert A. Wolfe, and the members of the USRDS Data
Coordinating Center, and Drs. Lawrence Y.C. Agodoa and Camille A.
Jones from the National Institutes of Health have generously provided
technical and scientific assistance.
Reprint requests to Dr. B.J.G. Pereira, New England Medical Center,
Division of Nephrology, 750 Washington Street, Box 5224, Boston,
Massachusetts 02111, USA.
E-mail: bpereira@lifespan.org
REFERENCES
1. U.S. Renal Data System: USRDS 1999 Annual Data Report.
Bethesda, National Institutes of Health, National Institute of Dia-
betes and Digestive and Kidney Diseases, 1999
2. U.S. Renal Data System: USRDS 1997 Annual Data Report.
Bethesda, National Institutes of Health, National Institute of Dia-
betes and Digestive and Kidney Diseases, 1997
3. U.S. Renal Data System: USRDS 1998 Annual Data Report.
Bethesda, National Institutes of Health, National Institute of Dia-
betes and Digestive and Kidney Diseases, 1998
4. Owen WF, Lew NL, Liu Y, Lowrie EG, Lazarus JM: The urea
reduction ratio and serum albumin concentration as predictors
of mortality in patients undergoing hemodialysis. N Engl J Med
329:1001–1006, 1993
5. Keane WF, Collins AJ: Influence of co-morbidity on mortality
and morbidity in patients treated with hemodialysis. Am J Kidney
Dis 24:1010–1018, 1994
6. NIH Consensus Statement: Morbidity and mortality of dialysis.
Ann Intern Med 121:62–70, 1994
7. Jones C, McQuillan G, Kusek J, Eberhardt M, Herman W,
Coresh J, Salive M, Jones C, Agadoa L: Serum creatinine levels
in the U.S. population: Third National Health and Nutrition Exami-
nation Survey. Am J Kidney Dis 32:992–999, 1998
8. Health Care Financing Administration: 1994 ESRD Core Indi-
cators Project initial results: Opportunities to improve care for adult
in-center hemodialysis patients. Baltimore, Department of Health
and Human Services, Health Care Financing Administration,
Health Standards and Quality Bureau, 1994
9. NKF-DOQI: Clinical practice guidelines for peritoneal dialysis
adequacy. Am J Kidney Dis 30(Suppl 2):S67–S136, 1997
10. Obrador GT, Arora P, Kausz A, Pereira BJG: Pre-ESRD care
in the United States: A state of disrepair. J Am Soc Nephrol
9:544–554, 1998
11. Hunsicker LG, Adler S, Caggiula A, England BK, Green T,
Kusek JW, Rogers NL, Teschan PE, the Modification of Diet
in Renal Disease Study Group: Predictors of the progression of
renal disease in the Modification of Diet in Renal Disease Study.
Kidney Int 51:1908–1919, 1997
12. Lewis EJ, Hunsicker LG, Bain RP, Rohde RD, Collaborative
Study Group: The effect of angiotensin-converting enzyme inhibi-
tion on diabetic nephropathy. N Engl J Med 329:1456–1462, 1993
13. Giatras I, Lau J, Levey AS: Effect of angiotensin-converting-
enzyme inhibitors on the progression of non-diabetic renal disease:
A meta-analysis of randomized trials. Ann Intern Med 127:337–345,
1997
14. Laffel LMB, McGill JB, Gans D, on behalf of the Northern
American Microalbuminuria Study Group: The beneficial effect
of angiotensin-converting enzyme inhibition with captopril on dia-
betic nephropathy in normotensive IDDM patients with microal-
buminuria. Am J Med 99:497–504, 1995
364 Nephrology Forum: Optimization of pre-ESRD care
15. Peterson JC, Adler S, Burkart JM, Greene T, Hebert LA,
Hunsicker LG, King AJ, Klahr S, Massry SG, Seifter JL: Blood
pressure, proteinuria, and the progression of renal disease: The
Modification of Diet in Renal Disease Study. Ann Intern Med
123:754–762, 1995
16. The Sixth Report of the Joint National Committee on Prevention,
Evaluation, and Treatment of High Blood Pressure. Bethesda, Na-
tional Institutes of Health, National Heart, Lung, and Blood Insti-
tute, 1997, pp 47–48
17. Jacobson HR, Striker GE: Report of a workshop to develop
management recommendations for the prevention of progression
in chronic renal failure. Am J Kidney Dis 25:103–106, 1995
18. Lazarus JM, Bourgoignie JJ, Buckalew VM, Greene T, Levey
AS, Milas NC, Parandini L, Peterson JC, Porush JG, Rauch
S, Soucie JM, Stollar C, Modification of Diet in Renal Disease
Study Group: Achievement and safety of a low blood pressure
goal in chronic renal disease. Hypertension 29:641–650, 1997
19. The Diabetes Control and Complications Trial Research
Group: The effect of intensive treatment of diabetes on the devel-
opment and progression of long-term complications in insulin-
dependent diabetes mellitus. N Engl J Med 329:977–986, 1993
20. The Diabetes Control and Complications Trial Research
Group: Effect of intensive therapy on the development and pro-
gression of diabetic nephropathy in the Diabetes Control and Com-
plications Trial. Kidney Int 47:1703–1720, 1995
21. Klahr S, Levey AS, Beck GJ, Caggiula AW, Hunsicker L, Kusek
JW, Striker G, Modification of Diet in Renal Disease Study
Group: The effects of dietary protein restriction and blood-pres-
sure control on the progression of chronic renal disease. N Engl
J Med 330:877–884, 1994
22. Klahr S: Is there still a role for a diet very low in protein, with
or without supplements, in the management of patients with end-
stage renal disease? Curr Opin Nephrol Hypertens 5:384–387, 1996
23. Levey AS, Adler S, Caggiula AW, England BK, Greene T,
Hunsicker LG, Kusek JW, Rogers NL, Teschan PE, Modifica-
tion of Diet in Renal Disease Study Group: Effects of dietary
protein restriction on the progression of advanced renal disease
in the MDRD Study. Am J Kidney Dis 27:652–663, 1996
24. Pedrini MT, Levey AS, Lau J, Chalmers TC, Wang PH: The
effect of dietary protein restriction on the progression of diabetic
and non-diabetic renal diseases: A meta-analysis. Ann Intern Med
124:627–632, 1996
25. McClellan WM, Knight DF, Karp H, Brown WW: Early detec-
tion and treatment of renal disease in hospitalized diabetic and
hypertensive patients: Important differences between practiced
and published guidelines. Am J Kidney Dis 29:368–375, 1997
26. Kenny S, Smith P, Goldschmid M, Newman J, Herman W: Survey
of physician practice behaviors related to diabetes mellitus in the
U.S.. Diabetes Care 16:1507–1510, 1993
27. Port F: Morbidity and mortality in dialysis patients. Kidney Int
46:1728–1737, 1994
28. McCusker FX, Teehan BP, Thorpe KE, Keshaviah PR,
Churchill DN, Canada-USA (CANUSA) Peritoneal Dialysis
Study Group: How much peritoneal dialysis is required for the
maintenance of a good nutritional state? Kidney Int 50(Suppl
56):S56–S61, 1996
29. Lowrie EG, Lew LN: Death risk in hemodialysis patients: The
predictive value of commonly used measured variables and an
evaluation of death rate differences between facilities. Am J Kidney
Dis 5:458–482, 1990
30. Churchill D, Taylor D, Cook R, Laplante P, Barre P, Cartier
P, Fay W, Goldstein M, Jindal K, Mandin H: Canadian hemodi-
alysis morbidity study. Am J Kidney Dis 19:214–234, 1992
31. Hakim RM, Levin N: Malnutrition in hemodialysis patients. Am
J Kidney Dis 21:125–137, 1993
32. Obrador GT, Pereira BJG: Early referral to the nephrologist and
timely initiation of renal replacement therapy: A paradigm shift
in the management of patients with chronic renal failure. Am J
Kidney Dis 31:398–417, 1998
33. Owen WF, Lowrie EG: C-reactive protein as an outcome predictor
for maintenance hemodialysis patients. Kidney Int 54:627–636, 1998
34. Bergström J, Lindholm B: Malnutrition, cardiac disease, and mor-
tality: An integrated point of view. Am J Kidney Dis 32:834–841,
1998
35. United States Renal Data System: Comorbid conditions and
correlations with mortality risk among 3,399 incident hemodialysis
patients. Am J Kidney Dis 20(Suppl 2):32–38, 1992
36. Obrador GT, Ruthazer R, Arora P, Kausz A, Pereira BJG:
Prevalence of and factors associated with sub-optimal care prior
to initiation of dialysis in the United States. J Am Soc Nephrol
10:1793–1800, 1999
37. Arora P, Obrador G, Ruthazer R, Kausz A, Meyer K, Jenule-
son C, Pereira BJG: Prevalence, predictors and consequences of
late nephrology referral at a tertiary care center. J Am Soc Nephrol
6:1281–1286, 1999
38. Mitch WE: Dietary protein restriction in patients with chronic
renal failure. Kidney Int 40:326–341, 1991
39. Foley R, Levey A, Sarnak M: Clinical epidemiology of cardiovas-
cular disease in chronic renal failure. Am J Kidney Dis 32(Suppl
3):S112–S119, 1998
40. Levin A, Singer J, Thompson CR, Ross H, Lewis M: Prevalent left
ventricular hypertrophy in the predialysis population: Identifying
opportunities for intervention. Am J Kidney Dis 27:347–354, 1996
41. Foley RN, Parfrey PS, Harnett JD, Kent GM, Martin CJ, Mur-
ray DC, Barre PE: Clinical and echocardiographic disease in pa-
tients starting end-stage renal disease therapy. Kidney Int 47:186–
192, 1995
42. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC,
Barre PE: The impact of anemia on cardiomyopathy, morbidity
and mortality in end-stage renal disease. Am J Kidney Dis 28:53–61,
1996
43. Portoles J, Torralbo A, Martin P, Rodrigo J, Herrero J, Bar-
rientos A: Cardiovascular effects of recombinant human erythro-
poietin in predialysis patients. Am J Kidney Dis 29:541–548, 1997
44. Eschbach JW, Aquiling T, Haley NR, Fan MH, Blagg CR: The
long-term effects of recombinant human erythropoietin on the
cardiovascular system. Clin Nephrol 38:S98–S103, 1992
45. NKF-DOQI: Clinical practice guidelines for the treatment of ane-
mia of chronic renal failure. Am J Kidney Dis 30(Suppl 3):S192–
S240, 1997
46. Schwartz A, Prior J, Mintz G, Kim K, Kahn S: Cardiovascular
hemodynamic effects of correction of anemia of chronic renal
failure with recombinant human erythropoietin. Transplant Proc
23:1827–1830, 1991
47. Anonymous: Effectiveness and safety of recombinant human
erythropoietin in predialysis patients: Austrian Multicenter Study
Group of r-HuEPO in predialysis patients. Nephron 61:399–403,
1992
48. Powe N, Griffiths R, Watson A, Anderson G, de Lissovoy G,
Greer J, Herbert R, Milam R, Whelton P: Effect of recombinant
human erythropoietin on hospital admissions, readmissions, length
of stay, and costs of dialysis patients. J Am Soc Nephrol 4:1455–
1465, 1994
49. Sherrard DJ, Hercz G, Pei Y, Maloney NA, Greenwood C,
Manuel A, Siphoo C, Fenton SS, Segre GV: The spectrum of
bone disease in end-stage renal failure—An evolving disorder.
Kidney Int 43:436–442, 1993
50. Bushinsky D: The contribution of acidosis to renal osteodystrophy.
Kidney Int 47:1816–1832, 1995
51. Wilson L, Felsenfeld A, Drezner MK, Llach F: Altered diva-
lent ion metabolism in early renal failure: Role of 1,25(OH)2D.
Kidney Int 27:565–573, 1985
52. Martinez I, Saracho R, Montenegro J, Llach F: The importance
of dietary calcium and phosphorus in the secondary hyperparathy-
roidism of patients with early renal failure. Am J Kidney Dis
29:496–502, 1997
53. Coen G, Mazzaferro S, Ballant P, Sardella D, Chicca S,
Manni M, Bonucci E, Taggi F: Renal bone disease in 76 patients
with varying degrees of predialysis chronic renal failure: A cross-
sectional study. Nephrol Dial Transplant 11:813–819, 1996
54. Slatopolsky E, Bricker NS: The role of phosphorus restriction
in the prevention of secondary hyperparathyroidism in chronic
renal disease. Kidney Int 4:141–145, 1993
55. Slatopolsky E, Finch J, Denda M, Ritter C, Zhong M, Dusso
 Nephrology Forum: Optimization of pre-ESRD care
A, MacDonald PN: Phosphorus restriction prevents parathyroid
gland growth. J Clin Invest 97:2534–2540, 1996
56. Christiansen C, Rodbro P, Christensen MS, Hartnack B,
Transbol I: Deterioration of renal function during treatment of
chronic renal failure with 1-25 vit D. Lancet 2:700–703, 1978
57. Torres A, Lorenzo V, Hernandez D, Rodriguez JC, Concepcion
AM, Rodriguez AP, Hernandez A, deBonis E, Darias E, Gon-
zales-Posada JM, Losada M, Rufino M, Felsenfeld AJ, Rodri-
guez M: Bone disease in pre-dialysis, hemodialysis, and CAPD
patients: Evidence of a better bone response to PTH. Kidney Int
47:1434–1442, 1995
58. Wang M, Hercz G, Sherrard DJ, Maloney NA, Segre GV, Pei
Y: Relationship between intact 1-84 parathyroid hormone and
bone histomorphometric parameters in dialysis patients without
aluminum toxicity. Am J Kidney Dis 26:836–844, 1995
59. Harnett JD, Foley RN, Kent GM, Barre PE, Murray D, Parfrey
PS: Congestive heart failure in dialysis patients: Prevalence, inci-
dence, prognosis and risk factors. Kidney Int 47:884–890, 1995
60. Silberberg JS, Barre PE, Prichard SS, Sniderman AD: Impact of
left ventricular hypertrophy on survival in end-stage renal disease.
Kidney Int 36:286–290, 1989
61. Tucker B, Fabbian F, Giles M, Thuraisingham RC, Raine AE,
Baker LR: Left ventricular hypertrophy and ambulatory blood
pressure monitoring in chronic renal failure. Nephrol Dial Trans-
plant 12:724–728, 1997
62. Annual Report 1998. Don Mills, Ontario: Dialysis and Transplanta-
tion, Canadian Organ Replacement Register, Canadian Institute
for Health Information, Ottawa, 1998
63. Disney A, Russ G, Walker R, Sheil AE: ANZDATA Report
1998. Adelaide, Australia and New Zealand Dialysis and Trans-
plant Registry, 1996
64. Bloembergen W: Cardiac disease in chronic uremia: Epidemiol-
ogy. Adv Ren Replace Ther 4:185–193, 1997
65. Wilson PW: Established risk factors and coronary artery disease:
The Framingham study. Am J Hypertens 7:7S–12S, 1994
66. Vendemia F, Forniasieri A, Velis O, Baroni M, Scarduelli B,
D’Amico G: Different prevalence rates of hypertension in various
reno-parenchymal diseases, in Secondary Forms of Hypertension:
Current Diagnosis and Management, edited by Balufox MD, Bian-
chi C, New York, Raven, 1990, pp 1583–1599
67. Buckalew VM, Berg RL, Wang S-R, Porush JG, Rauch S,
Schulman G: Prevalence of hypertension in 1,795 subjects with
chronic renal disease: The Modification of Diet in Renal Disease
Study cohort. Am J Kidney Dis 28:811–821, 1996
68. Foley RN, Parfrey PS: Cardiac disease in chronic uremia: Clinical
outcomes and risk factors. Adv Ren Replace Ther 4:234–248, 1997
69. Bostom A, Lathrop L: Hyperhomocysteinemia in end-stage renal
disease: Prevalence, etiology and potential relationship to arterio-
sclerotic outcomes. Kidney Int 52:10–20, 1997
70. National Kidney Foundation Task Force on Cardiovascular
Disease: Controlling the epidemic of cardiovascular disease in
chronic renal disease: What do we know? What do we need to
learn? Where do we go from here? Am J Kidney Dis 32(Suppl):S1–
S199, 1998
71. Ratcliffe PJ, Phillips RE, Oliver DO: Late referral for mainte-
nance dialysis. Br Med J 288:441–443, 1984
72. Rasgon S, Schwankovsky L, James-Rogers A, Widrow L, Glick
J, Butts E: An intervention for employment maintenance among
blue-collar workers with end-stage renal disease. Am J Kidney Dis
22:403–412, 1993
73. NKF-DOQI: Clinical Practice Guidelines for Vascular Access. Am
J Kidney Dis 30(Suppl 3):S150–S191, 1997
74. Woods JD, Turenne MN, Strawderman RL, Young EW, Hirth
365
RA, Port FK, Held PJ: Vascular access survival among incident
hemodialysis patients in the United States. Am J Kidney Dis 30:50–
57, 1997
75. Schwab SJ: Vascular access for hemodialysis. Kidney Int 55:2078–
2090, 1999
76. U.S. Renal Data System: USRDS 1995 Annual Data Report.
Bethesda, National Institutes of Health, National Institute of Dia-
betes and Digestive and Kidney Diseases, 1995
77. Schillinger F, Schillinger D, Montagnac R, Milcent T: Post-
catheterization vein stenosis in haemodialysis patients: Compara-
tive angiographic study of 50 subclavian and 50 internal jugular
accesses. Nephrol Dial Transplant 6:722–724, 1991
78. Held PJ, Port FK, Wolfe RA, Stannard DC, Carrol CE, Daug-
irdas JT, Bloembergen WE, Greer JW, Hakim RM: The dose of
hemodialysis and patient mortality. Kidney Int 50:550–556, 1996
79. Bleyer AJ, Rocco MV, Burkart JM: The cost of hospitalizations
due to hemodialysis access management. Nephrol News Issues 9:
19–22, 1995
80. Hood SA, Schillo B, Beane E, Rozas V, Sondheimer JH, Mem-
bers of the Michigan Renal Plan Task Force of the Michigan
Public Health Institute: An analysis of the adequacy of prepara-
tion for end-stage renal disease care in Michigan. ASAIO J
41:M422–M426, 1995
81. Tattersall J, Greenwood R, Farrington K: Urea kinetics and
when to commence dialysis. Am J Nephrol 15:283–289, 1995
82. Bonomini V, Albertazzi A, Vangelista A, Bortolotti GC, Stef-
oni S, Scolari MP: Residual renal function and effective rehabilita-
tion in chronic dialysis. Nephron 16:89–99, 1976
83. Bonomini V, Feletti C, Scolari MP, Stefoni S: Benefits of early
initiation of dialysis. Kidney Int 28:S57–S59, 1985
84. Churchill DN, Taylor DW, Keshaviah PR, the Canada USA
(CANUSA) Peritoneal Dialysis Study Group: Adequacy of dial-
ysis and nutrition in continuous peritoneal dialysis: Association
with clinical outcomes. J Am Soc Nephrol 7:198–207, 1996
85. Johnson WJ, Kurtz SB, Mitchell JCI, Van Den Berg CJ: Results
of treatment of center hemodialysis patients. Mayo Clin Proc
59:669–671, 1984
86. Malangone JM, Abuelo JG, Pezzullo JC, Lund K, McGloin
CA: Clinical and laboratory features of patients with chronic renal
disease at the start of dialysis. Clin Nephrol 31:77–87, 1989
87. Levey A, Bosch J, Lewis J, Greene T, Rogers N, Roth D: A
more accurate method to estimate glomerular filtration rate from
serum creatinine: A new prediction equation. Ann Intern Med
130:461–470, 1999
88. Obrador G, Arora P, Kausz A, Ruthazer R, Pereira B, Levey
A: Level of renal function at initiation of dialysis in the U.S. end-
stage renal disease population. Kidney Int (in press)
89. Bloembergen WE, Port FK, Mauger EA, Wolfe RA: A compari-
son of mortality between patients treated with hemodialysis and
peritoneal dialysis. J Am Soc Nephrol 6:177–183, 1995
90. Jollis J, Delong E, Peterson E, Muhlbaier L, Fortin D, Califf
R, Mark D: Outcome of acute myocardial infarction according to
the specialty of the admitting physician. N Engl J Med 335:1880–
1887, 1996
91. Schreiber T, Elkhatib A, Grines C, O’Neill W: Cardiologist
versus internist management of patients with unstable angina:
Treatment patterns and outcomes. J Am Coll Cardiol 26:577–582,
1995
92. Laidlaw S, Berg R, Kopple J, Naito H, Walker W, Walser M,
Modification of Diet in Renal Disease Study Group: Patterns
of fasting plasma amino acid levels in chronic renal insufficiency:
Results from the feasibility phase of the modification of diet in
renal disease study. Am J Kidney Dis 23:504–513, 1993