Narrative Review

Sudden Cardiac Death in Hemodialysis Patients:

An In-Depth Review

Darren Green, MBChB,1 Paul R. Roberts, MD,2 David I. New, PhD,1 and
Philip A. Kalra, MD1

Sudden cardiac death (SCD) is the leading cause of death in hemodialysis patients, accounting for death in
up to one-quarter of this population. Unlike in the general population, coronary artery disease and heart failure
often are not the underlying pathologic processes for SCD; accordingly, current risk stratification tools are
inadequate when assessing these patients. Factors assuming greater importance in hemodialysis patients
may include left ventricular hypertrophy, electrolyte shift, and vascular calcification. Knowledge regarding SCD
in hemodialysis patients is insufficient, in part reflecting the lack of an agreed-on definition of SCD in this
population, although epidemiologic studies suggest the most common times for SCD to occur are toward the
end of the long 72-hour weekend interval between dialysis sessions and in the 12 hours immediately after
hemodialysis. Accordingly, it is hypothesized that the dialysis procedure itself may have important implications
for SCD. Supporting this is recognition that hemodialysis is associated with both ventricular arrhythmias and
dynamic electrocardiographic changes. Importantly, echocardiography and electrocardiography may show
changes that are modifiable by alterations to dialysis prescription. The most effective preventative strategy in
the general population, implanted cardioverter-defibrillator devices, are less effective in the presence of chronic
kidney disease and have not been studied adequately in dialysis patients. Last, many dialysis patients
experience SCD despite not fulfilling current criteria for implantation, making appropriate allocation of
defibrillators uncertain.
Am J Kidney Dis. 57(6):921-929. © 2011 by the National Kidney Foundation, Inc.

INDEX WORDS: Arrhythmia; dialysis; electrocardiogram; sudden cardiac death.

T he study of arrhythmia during dialysis is well

established, but only recently has the size of the
problem of sudden cardiac death (SCD) in dialy-
EPIDEMIOLOGY AND DEFINITION OF SCD IN
DIALYSIS PATIENTS

sis patients become apparent. Improved understand-
ing of the pathogenesis and epidemiologic characteris-
tics of SCD is vital to improve the long-term outcome
of dialysis patients. Sudden and unexpected death
accounts for 1 in 4 deaths in patients with end-stage
renal disease (ESRD).1 These deaths do not appear to
be caused by coronary artery disease (CAD), as is the
case for most SCDs in the general population. This is
reflected in SHARP (Study of Heart and Renal Protec-
tion), which recently showed that statins do not
improve overall survival in patients with chronic
kidney disease (CKD) despite a decrease in cardio-
vascular events.2 Quantification of the problem is
hindered further by variation in definitions of SCD
across studies in dialysis patients. This means there
currently is neither consensus on the relative impor-
tance of pathologic factors nor accurate tools for
risk stratification and management. This review
compares how current guidelines and risk factors
for SCD in the general population apply to dialysis
patients and how pathologic processes relating to
CKD and dialysis may manifest as SCD, arrhyth-
mia, or dynamic electrocardiographic changes. Such
changes may prove useful in risk assessment in this
high-risk group.
SCD is the unexpected natural death from a cardiac
cause within 1 hour of onset of symptoms in a person
not known to have a potentially fatal condition.3 In the
general population, this occurs at a rate of ⬃1 death/
1,000 patient-years and accounts for 6%-13% of all
deaths.4,5 US Renal Data System (USRDS) reports
include deaths from confirmed arrhythmia and deaths
from cardiac arrest of unknown cause, but without
specifying time frame or whether the death is of
cardiac origin. These SCDs account for 26.1% of
deaths in patients with ESRD and occur at a rate of 59
deaths/1,000 patient-years in dialysis patients.1 Most,
51.5 deaths/1,000 patient-years, are “cardiac arrest
cause unknown.” Although this definition overesti-
From 1Salford Royal Hospital, Salford; and 2Southampton Uni-
versity Hospital, Southampton, United Kingdom.
Received September 28, 2010. Accepted in revised form Febru-
ary 22, 2011. Originally published online April 18, 2011.
Address correspondence to Philip A. Kalra, MD, Department of
Renal Medicine, Salford Royal Hospital, Stott Lane, M6 8HD, UK.
E-mail: philip.kalra@srft.nhs.uk
© 2011 by the National Kidney Foundation, Inc.
0272-6386/$36.00
doi:10.1053/j.ajkd.2011.02.376

Am J Kidney Dis. 2011;57(6):921-929 921


mates the rate of SCD,6 the data still emphasize the
importance of CKD as an independent risk factor for
SCD in patients with coexistent CAD.7
In a report of 167 deaths in a prevalent dialysis
population of 476 patients, sudden death was defined
as unexpected natural death occurring within 1 hour
of the onset of symptoms.8 Unsurprisingly, this report
showed a lower SCD event rate than USRDS data,
with a 3-year cumulative incidence of 6.9% (⬃23
deaths/1,000 patient-years), accounting for 19% of
deaths. Parekh et al9 defined SCD as an unexpected
out-of-hospital death with an underlying cardiac cause.
In this cohort of 1,041 patients, 22% of deaths were
due to SCD. If “suddenly” is not defined as occurring
within 1 hour of onset of symptoms, arrhythmia
becomes a less common cause of death because this
definition allows inclusion of “unexpected” as well as
sudden deaths.9 These studies show that variability in
the use of the words sudden and cardiac in the
definition of SCD can lead to greater heterogeneity of
included disease processes, as well as differences in
studied end points.
A consensus on definitions of death would benefit
the nephrology community and distinction should be
made between deaths that truly are sudden (within 1
hour of onset of symptoms) and those that are unex-
pected or unwitnessed. The terms SCD and sudden
death should apply to deaths that are of cardiac and
noncardiac origin, respectively. Clarifying the defini-
tions will allow more focused investigation into the
mechanisms of both SCD and noncardiac sudden
death.
COMPARATIVE PATHOLOGIC PROCESSES OF
SCD IN CKD PATIENTS AND THE
GENERAL POPULATION
In the general population, up to 80% of SCD cases
will have post mortem evidence of abnormal coronary
vessels,10,11 and those at highest risk of SCD are
patients with heart failure. However, these pathologic
processes do not appear to be as important in SCD
events in dialysis patients. This section compares the
underlying disease processes in SCD between dialysis
patients and the general population.
Coronary Artery Disease
Although CAD is present in 38% of the prevalent
dialysis population,12 it is not clearly implicated in
SCD in these patients. Although 1 study found CAD
in 56.3% of dialysis patients who experienced sudden
or unexpected death,13 the importance of this finding
is unclear because there was no significant difference
in CAD prevalence between patients who experienced
sudden and nonsudden deaths in the study discussed.8
Only 1 study has analyzed post mortem data from
922
Green et al
dialysis patients who experienced sudden death; 93
patients were divided according to whether death
occurred within or after 24 hours of onset of symp-
toms.14 The most common finding in sudden deaths
was stroke (25.7%), not cardiac disease (14.3%).
Only 11.8% of cardiac sudden deaths were due to
CAD. Aortic dissection was the only pathologic pro-
cess statistically more common in sudden than later
death (14.3% vs 5.4%; P ⬍ 0.05). Hence, SCD in
dialysis patients is less likely to be ischemic in origin
compared with the general population, a view sup-
ported by 4D (Die Deutsche Diabetes Dialyse Studie)
and the AURORA (Assessment of Survival and Car-
diovascular Events) Study, which showed a lack of
benefit of statins in decreasing cardiovascular mortal-
ity in dialysis patients despite significant decreases in
cholesterol levels.15,16 Furthermore, the pathologic
process of CAD is different in patients with CKD
compared with the general population, with calcifica-
tion becoming pre-eminent in the coronary vascula-
ture at lower glomerular filtration rates and atheroma-
tous plaque formation becoming less prominent.17
This not only helps explain the limited benefit of
statins, but determines that the mechanism of sudden
death is less likely to be catastrophic infarction associ-
ated with plaque rupture.
Left Ventricular Failure
In the general population, poor left ventricular (LV)
function is one of the strongest predictors of SCD and
the only bedside marker that has underpinned routine
clinical practice because it is of major importance in
the current guidelines for the use of implanted cardio-
verter-defibrillator (ICD) devices, which currently are
the most effective preventative strategy for SCD risk
reduction.
In a cohort of 111 hemodialysis patients, Saravanan
et al18 found that 6.3% fulfilled the criteria for ICD
implantation. In a study of a prevalent hemodialysis
population (n ⫽ 243), Mangrum et al19 found that
only 17% of SCDs in dialysis patients occurred
in patients with LV ejection fraction ⬍30%, but these
patients had a 5-year SCD risk of 60%. The risk for
patients with normal LV function was 25%,19 which
would still be classified as high risk in the general
population. Therefore, although dialysis patients with
LV ejection fraction ⬍30% are a particularly concern-
ing group, poor ejection fraction alone cannot account
for the high risk of SCD associated with advanced
CKD. This finding is supported further by the study of
Genovesi et al,8 in which there was no significant
difference in sudden deaths between patients with LV
ejection fraction ⬎40% or ⬍40%.
Am J Kidney Dis. 2011;57(6):921-929
Investigation of Sudden Cardiac Death in Hemodialysis
LV Hypertrophy
Data from the Framingham study show that LV
hypertrophy (LVH) is associated with a greater risk of
sudden death in the general population,20 although
this information has not translated into a clinically
applicable tool. Although 74% of hemodialysis pa-
tients have LVH,21 which is associated with intradia-
lytic ectopy,22 studies fail to show either a difference
in sudden death rates between patients with and with-
out LVH8 or a difference in LV mass between patients
who experience sudden death and those who die of
other causes. In a study of 16 sudden death events in a
cohort of 123 dialysis patients, a progressive increase
in LV mass index was more predictive of sudden
death than both hypertension and CAD, but LVH
alone was not predictive.23 However, a substudy of
4D showed that electrocardiographic evidence of LVH
was associated with a higher rate of sudden death in
these diabetic hemodialysis patients.24 This latter study
included 1,253 patients; therefore, the difference in
findings may reflect how the studies were powered.
In patients with uremic cardiomyopathy, LVH may
occur in response to hypertension, LV dilatation asso-
ciated with volume overload, chronic inflammation,
and overactivation of the renin-angiotension pathway,
clinical factors associated with CKD and that may be
responsible for progression of increase in LV mass
index. However, of these, only chronic inflammation
is linked directly to sudden death in dialysis patients.9
It still is uncertain whether halting the progression of
LVH or modification of the underlying risk factors
can decrease the risk of sudden death.
Hyperkalemia
The “tented” or peaked T wave seen on an electro-
cardiogram (ECG) in patients with hyperkalemia is
due to overactivation of potassium-gated channels
and shortening of ventricular repolarization. Hyperka-
lemia also causes inactivation of sodium-gated chan-
nels, leading to slower depolarization and broadening
of QRS complexes. These abnormalities are associ-
ated with a risk of ventricular arrhythmia. Genovesi et
al8 found that serum potassium level ⬎6.0 mmol/L
was associated with a significant increase in risk of
sudden death in dialysis patients (hazard ratio, 2.74;
P ⫽ 0.009), although this was not reproduced in
another study of 123 patients followed up for 10
years.23 In a large cohort of 81,013 dialysis patients,
Kovesdy et al25 noted that a predialysis serum potas-
sium level of 4.6-5.3 mmol/L was associated with the
lowest all-cause mortality.
Diabetes Mellitus and Cardiac Autonomic Neuropathy
As well as the effect on CAD, diabetes mellitus is
associated with an independent risk of SCD, known as
Am J Kidney Dis. 2011;57(6):921-929
the “dead in bed” syndrome. This is believed to be
arrhythmic, caused by nocturnal hypoglycemia and
secondary QT-segment prolongation.26 However, in
the only published review of post mortem analyses of
sudden death in patients with diabetes, CAD was
evident in all cases. Diabetic nephropathy also was an
independent risk factor for SCD.27
Cardiac autonomic neuropathy also may contribute
to SCD, and both diabetes and uremia are potential
causes. Diabetic autonomic neuropathy is a microvas-
cular phenomenon28 associated with electrocardio-
graphic changes that may be a precursor to SCD,29
although causality has never been proved. Electrocar-
diographic changes that reflect autonomic function
are discussed in more detail next.
In 4D, patients with hemoglobin A1c levels ⱖ8%
had an increased risk of sudden death compared with
those with hemoglobin A1c levels within the reference
range (⬍6%; hazard ratio, 2.25; 95% confidence
interval, 1.32-3.81).30 It is uncertain whether this
association is caused by macrovascular or autonomic
complications of poor glycemic control.
Vascular Calcification
Nakano et al17 showed a high rate of coronary
vascular calcification in patients with advanced CKD
(adjusted odds ratio, 4.71 for calcification with glomer-
ular filtration rate ⬍30 vs ⬎60 mL/min/1.73 m2).
Vascular calcification is linked to greater cardiovascu-
lar morbidity and worse all-cause mortality in dialysis
patients, but this has not been extended to a proven
link to sudden death. In a study of 32 dialysis patients
who underwent noncontrast spiral computed tomogra-
phy and concurrent electrocardiography, there was an
independent link between coronary calcification score
and prolongation of QT interval31; however, causality
was not shown. The pathogenesis of vascular calcifi-
cation is complex and no study has shown a link
between hyperphosphatemia or hyperparathyroidism
and sudden death.
DIALYSIS AS A RISK FACTOR FOR SCD
With the exception of CKD itself, all previously
described pathologic factors shown to predict SCD in
dialysis patients are associated with SCD in the gen-
eral population. Hence, it may be that the high SCD
risk associated with dialysis reflects the relative preva-
lence of these factors rather than dialysis-specific
SCD. However, there is evidence that the process of
dialysis itself incurs an additional major risk, and this
is discussed in the following section.
Hemodialysis patients die of cardiac arrest and
arrhythmia at a higher rate than peritoneal dialysis
patients (62.2 vs 42.8 events/1,000 patient-years),1
with death most likely to occur in the last 12 hours of
923

Box 1. Lown Classification of Ventricular Arrhythmia in the
General Population
Grade 0: no VPB
Grade 1: ⬍30 unifocal VPBs/h
Grade 2: ⬎30 unifocal VPBs/h
Grade 3: multiform VPBs
Grade 4a: 2 consecutive VPBs
Grade 4b: ⬎2 consecutive VPBs
Grade 5: R-on-T phenomenon: QRS complex occurring over a
preceding T-wave
Note: Based on the grading system presented in Lown and
Wolf.32
Abbreviation: VPB, ventricular premature beat.
the longer weekend (3-day) interdialytic period.15 The
consequences of a proportional increase in fluid, elec-
trolyte, and metabolite accumulation during this pe-
riod will account for some of this excess. The next
most common time for SCD to occur is in the 12 hours
after the start of a dialysis session.8,13 This intradia-
lytic risk of arrhythmia and the potential for subse-
quent fatal events has been studied using pre- and
postdialysis 12-lead ECGs or predialysis ambulatory
ECGs. This section explores the findings of these
studies and their application to clinical practice.
Ectopy and Arrhythmic Events
One study of nonsustained ventricular tachycardia
(ⱖ3 successive ventricular beats terminating sponta-
neously within 30 seconds) used Holter monitoring
for a 48-hour period that included 1 dialysis session in
127 patients. It found that 35 (28%) patients had
either Lown class 4A or B ventricular activity (Lown
classification is shown in Box 1) or 3 or more consecu-
tive beats of ventricular tachycardia. These patients
fared no worse during a 4-year follow-up period than
patients who did not experience such events, and there
was no correlation between levels of serum electro-
lytes or metabolites and frequency of ventricular
arrhythmias.33,34 Two studies have compared arrhyth-
mias in the hemodialysis population with other CKD
groups. One found that Lown classes 3-4 cardiac
arrhythmias35 occurred in 2 of 27 continuous ambula-
tory peritoneal dialysis patients versus 9 of 27 hemo-
dialysis patients (P ⫽ 0.0149), but increased LVH
prevalence in hemodialysis patients (93% vs 52%;
P ⫽ 0.0008) was a confounder. The second study
compared 120 patients divided into 3 categories of
CKD. Although arrhythmias were more frequent in
hemodialysis patients, this did not reach statistical
significance (transplant recipients, 16%; CKD stage 4,
17%; hemodialysis, 34%; P ⫽ 0.086).36
Although the Lown classification of ventricular
arrhythmia has been used in these studies, it has
significant limitations for SCD risk assessment in
patients with CKD. Class 4 arrhythmia, for ex-
924
Green et al
ample, does not differentiate ventricular tachycar-
dia beyond the presence of couplets or triplets of
ventricular beats, and no importance is attached to
duration or frequency of these events. The classifi-
cation will not differentiate patients who carry
varying risk within the same class, and it cannot
reliably be used to stratify risk. Despite having
been in use for many years, this classification has
not effectively translated from experimental use
into clinical practice. Future studies may be im-
proved by providing better quantitative descrip-
tions of complex arrhythmias beyond the limita-
tions of Lown classes 3 and 4 descriptions.37
Table 1 lists findings from studies of more than 50
hemodialysis patients in which 24-hour Holter moni-
toring included at least 1 dialysis session. All studies
show a link between intradialytic arrhythmia and
underlying ischemic or structural heart disease, but
not with the alternative CKD pathologic states dis-
cussed previously (with the exception of LVH).38,39
QT Interval Changes
Changes in QT interval relate to ventricular myo-
cardial repolarization, a phase during which the risk
of ventricular arrhythmia is high. QT dispersion
(QTd) describes the difference in QT interval of
leads in the same ECG. Small studies have failed to
prove this to be a predictor of SCD in the general
population,40 although in a substudy of a landmark
ICD trial, patients with high QTd were at higher
risk.41
In dialysis patients, a number of small studies of
ambulatory electrocardiography concur that both time-
corrected QT interval (QTc) and QTd are increased in
the hemodialysis population compared with healthy
controls, and QT interval lengthens during dialy-
sis,42-44 although study findings and inclusion criteria
have varied. No difference in in-hospital mortality
was found in acute dialysis patients with a range of
QTc intervals. Studies using ambulatory or 12-lead
ECGs have consistently shown that QTc and QTd are
affected by dialysate calcium and potassium concen-
trations.42,45-47 Published data fail to show a link
between QT interval changes and sudden death, but
there is potential for an ECG to aid in risk assessment
for SCD by drawing attention to optimal dialysate
prescription.
T-Wave Alternans
Microvoltage T-wave alternans (mTWA) describes
the change in T-wave amplitude from beat to beat
following an A, B, A, B pattern. Changes are not
noticeable to the naked eye. An increase in amplitude
of mTWA is a direct precursor to ventricular fibrilla-
tion in ischemic myocardium.48 It is reported to have
Am J Kidney Dis. 2011;57(6):921-929
Investigation of Sudden Cardiac Death in Hemodialysis

Table 1. Holter Studies of Ventricular Ectopy and Lown Class Arrhythmia During Dialysis

Study No. Method Findings

Kitano et al39 150 HD patients with ischemic changes
on exercise test. Comparison of
those with/without CAD on
angiography.
de Lima et al36 95 Comparison of arrhythmias (Lown
class 2, 3, or 4) in HD,
transplant, and CKD4 patients.
Patients with significant CAD
were excluded.
Abe et al38 72 HD patients with history of
arrhythmia or CAD. Frequency
of VPBs was recorded.
Canziani et al35 54 Comparison of Lown classes 3 and
4 events in PD and HD patients.
Radaelli et al33 127 Multicenter selection of
maintenance dialysis patients
without exclusion criteria based
on cardiovascular history.
Note: Studies are limited to those with at least 50 patients and including at least 1 dialysis session.
Abbreviations: CAD, coronary artery disease; CI, confidence interval; CKD4, chronic kidney disease stage 4; HD, hemodialysis; OR,
odds ratio; PD, peritoneal dialysis; VPB, ventricular premature beat.
Lown class 4 ventricular arrhythmias were significantly
more frequent in the stenotic than nonstenotic group
during dialysis and for 12 h after (P ⬍ 0.03).
Arrhythmias were not statistically more frequent in
dialysis patients (transplant, 16%; CKD4, 17%; HD,
34%; ␹2 ⫽ 4.9; P ⫽ 0.086). Dialysis did not
influence frequency of arrhythmias. Systolic blood
pressure (OR, 1.015; 95% CI, 1.001-1.027; P ⫽
0.03) and left ventricular systolic dysfunction (OR,
7.04; 95% CI, 1.3-36.7; P ⫽ 0.02) were the only
parameters associated with arrhythmia.
Frequency of VPBs was related to change in
potassium levels during dialysis.
Arrhythmia occurred in 2/27 PD and 9/27 HD patients
(P ⫽ 0.0149). However, left ventricular hypertrophy
was present in 52% of PD and 93% of HD patients
(P ⫽ 0.0008).
27/127 had Lown class 4A or B; 8/127 had ⱖ3
consecutive beats of ventricular tachycardia.
Frequency of ventricular arrhythmias increased
significantly during hour 3 of HD. Effect lasted for at
least 5 h after dialysis and was associated with age
ⱖ55 y, left ventricular dysfunction, and presence of
ventricular arrhythmias before dialysis. No
difference in outcome after 4-y follow-up.

a strong negative predictive value for SCD in the
general population.49,50 It may be that patients who do
not develop mTWA during cardiovascular strain are
less likely to develop ventricular fibrillation. It is
mentioned in the current guidelines for the manage-
ment of ventricular arrhythmia.3
mTWA rarely has been investigated in dialysis
patients. In 1 study, 115 of 200 dialysis patients had
non-negative test results. This correlated with a his-
tory of CAD and increased LV mass and volume.51
Standard mTWA assessment is performed using exer-
cise stress testing and thus would not be useful in
assessing time-dependent changes, such as those dur-
ing dialysis. In a small study of patients who under-
went pre- and postdialysis exercise tests, 2 of 4
patients with a negative predialysis mTWA test result
became non-negative after dialysis.52 mTWA can be
measured in a time domain using the modified moving
average system. This devises an average reading across
successive 15-second periods of Holter recording,
allowing the clinician to see dynamic changes in
mTWA, such as what potentially be may seen during
dialysis53 and without the need for exercise. Modified
moving average analysis has not yet been studied in
dialysis patients.
Heart Rate Variability
Heart rate variability (HRV) is an electrocardio-
graphic marker of cardiac autonomic response to
stresses. HRV is obtained by measuring RR-
interval variation on successive cardiac cycles. De-
creased HRV is associated with increased risk of
ventricular arrhythmia and cardiac death in the
general population.54 Decreased HRV is seen in
both hemodialysis and peritoneal dialysis patients.
Hemodialysis causes a decrease in HRV that recov-
ers 2 hours postdialysis; these changes improve
with better dialysis adequacy.55 It has been associ-
ated with SCD in a cohort of 174 hemodialysis
patients with LVH (SCD-free survival, 29.4% and
98.1% in patients with and without significant HRV,
respectively).56 Although HRV has been associated
with all-cause and cardiovascular mortality in an
unselected dialysis cohort,57 this assessment has
not translated into an independent risk stratification
tool for SCD in dialysis patients because its exact
role is unclear. For example, Nishimura et al58
noted that HRV and therefore diabetic autonomic
neuropathy correlated with LVH in diabetic dialysis
patients. Therefore, it may be that abnormal HRV is

Am J Kidney Dis. 2011;57(6):921-929 925

 Green et al

Table 2. Studies of Ischemic Changes During HD Measured Using Holter Monitoring

Study No. Initial Findings Follow-up

Conlon et al60 67 No relationship between baseline variables and likelihood No difference in cardiac mortality, MI,
of ST depression. or CABG in follow-up (500-750 d)
Abe et al38 72 Frequency of ventricular premature beats related to None
change in potassium during HD. Cardiac symptoms
were more frequent if ST changes occurred during HD.
Mohi-ud-din et al62 23 Known CAD patients were excluded. Only 20% of None
patients with HD-induced ST changes had
angiographically shown CAD. No link between
arrhythmias and biochemical changes. Patients with
arrhythmias were more likely to have shown ST
changes.
Narula et al61 38 Arrhythmias more frequent in patients with ST changes None
(72% vs 33%) but no statistical significance. No
correlation between metabolic changes and
arrhythmias.
Shapira et al43 39 Age was the only predictive factor for arrhythmias. None
Ischemic-looking changes were not more common in
patients with a history of CAD.
Burton et al22 40 Patients with regional wall motion abnormalities (27/40) None
had higher rate of arrhythmia (P ⬍ 0.01), as did
patients with established structural heart disease (P ⬍
0.05).
Abbreviations: CABG, coronary artery bypass grafting; CAD, coronary artery disease; HD, hemodialysis; MI, myocardial
infarction.

reflective of underlying cardiac disease, rather than Summary: Intradialytic Changes

an independent marker of risk of SCD.
Evidence of Intradialytic Myocardial Ischemia
Although CAD cannot be responsible for the ex-
cess of SCD in dialysis patients, there is some evi-
dence of an association between the 2 during dialysis.
Importantly, myocardial ischemia, assessed using se-
rial echocardiographic measurement of LV regional
wall motion abnormalities during dialysis, occurs in
patients with and without underlying CAD. Ischemic
changes are increased in patients with cardiac struc-
tural changes. Dialysis-associated ischemia can lead
to a long-term decrease in LV function and worse
prognosis at 12 months.59 It also is associated with
increased frequency of Lown-classified ventricular
arrhythmia.22 However, these arrhythmic changes are
not necessarily precursors to fatal arrhythmic events
because they resolve or become less frequent after
dialysis. Intradialytic myocardial ischemia also is
manifest by ST-segment depression ⬎1 mm occurring
in 20%-46% of all patients, and arrhythmia is more
common in patients who show these intradialytic
ST-segment changes.60,61 Studies do not agree about
whether these events are more common in patients
with CAD39,60,62 or whether intradialytic ST-segment
changes translate into mortality and cardiovascular
events (summarized in Table 2).
Although Genovesi et al8 have shown that the
overall rate of sudden death is not more common in
dialysis patients with CAD and heart failure, 2 studies
have specifically examined cardiac arrests that occur
during dialysis, showing that CAD and heart failure
are more common in these patients compared with the
rest of the dialysis population.63,64 This suggests that
the relative importance of particular pathophysiologic
factors in SCD during dialysis may be different com-
pared with sudden death at other times. Routine
investigation for dialysis-associated ischemia or elec-
trolyte-associated electrocardiographic changes may
become advisable, although alternative risk assess-
ment for sudden death at other times will still be
needed.
ICD DEVICES IN DIALYSIS PATIENTS
The most effective therapy for SCD in the general
population is the use of ICD devices. This section
discusses their use in dialysis patients. The current
international guidelines for ICD implantation are listed
in Box 2. A meta-analysis of 6 cohort studies65-72
assessed the outcome of ICD therapy in patients with
CKD according to conventional guidelines. This com-
pared mortality between patients with and without
ESRD. Dialysis patients had 2.7-fold higher mortal-
ity.25 This suggests that advanced CKD is an indepen-
dent risk factor for all-cause mortality, which is well

926 Am J Kidney Dis. 2011;57(6):921-929

Investigation of Sudden Cardiac Death in Hemodialysis
Box 2. Current Indications for ICD Device Therapy in the
General Population
Survival of cardiac arrest due to VT or ventricular fibrillation
Episode of sustained VT causing severe hemodynamic
compromise
Episode of sustained VT without hemodynamic compromise ⫹
EF ⬍35%
MI ⫹ EF ⬍35% ⫹ nonsustained VT on 24-h ECG ⫹ inducible
VT on electrophysiologic testing
MI ⫹ EF ⬍30% ⫹ QRS duration ⱖ120 ms on ECG
Note: Excludes familial cardiac conditions.
Abbreviations: EF, ejection fraction; ECG, electrocardiogram;
ICD, implantable cardioverter-defibrillator; MI, myocardial infarc-
tion; VT, ventricular tachycardia.
Source: Zipes et al.3
established. Seventy-one percent of dialysis patients
who experience SCD do not fulfill the current criteria
for ICD implantation,19 which strengthens the view
that further investigation into the potential for use of
ICD in dialysis patients outside of current guidelines
is warranted. Herzog et al73 studied outcome after
cardiac arrest in 6,173 dialysis patients from the
Medicare database. They noted a 42% risk reduction
for death in patients with ICDs implanted according to
conventional guidelines.73 Although mortality was
significantly higher than in patients without ESRD,
the decrease in relative risk for death was similar. This
suggests a potential role of ICD therapy in patients
with ESRD, even using only conventional criteria for
implantation. The current recommendation is that
dialysis patients should not be excluded from ICD
therapy, although this conclusion is not drawn from
prospective studies because patients with CKD have
been excluded from all major ICD trials. However,
there is a significantly greater risk of device complica-
tions in dialysis patients compared with the general
population, with hematoma, thrombosis, and in-
creased defibrillation threshold more common,74 al-
though there has been no statistically proven increase
in either infection or fistula thrombosis in ICD recipi-
ents, even though the latter may represent only an
effect of study size. A limitation of the current data
concerning ICD recipients on dialysis therapy is that
quality of life after cardiac arrest has not been evalu-
ated. This should be assessed in any outcome studies
of populations with very extensive comorbid condi-
tions.
As stated, the landmark trials of ICD therapy have
excluded patients with advanced CKD. The ICD2
(Implantable Cardioverter Defibrillators in Dialysis
Patients) Trial75 will prospectively evaluate the use of
ICDs for a decrease in sudden arrhythmic death in
maintenance dialysis patients during a 4-year period.
This pilot study includes adult long-term hemodialy-
sis patients older than 55 years who do not fulfill
current criteria for ICD implantation. Two arms of
Am J Kidney Dis. 2011;57(6):921-929
100 patients will compare “conventional” therapy
with the addition of an ICD. Although the diverse
pattern of risk factors in dialysis patients will compli-
cate effective post hoc analysis of risk stratification
because large numbers of variables will need to be
accommodated, ICD2 has the potential to determine
whether dialysis alone provides adequate indication
for ICD therapy, a question that needs to be addressed.
CONCLUSION
Risk factors for sudden death in the general popula-
tion also apply to people on dialysis therapy; however,
the relative frequencies of these factors differ between
the 2 groups. Furthermore, comparisons between di-
alysis patients and the general population are hindered
by differences in definitions of SCD used. Even com-
parisons of studies of sudden death within nephrology
are limited by the lack of a common term when
classifying a death as sudden or unexpected.
CAD and heart failure, pre-eminent risk factors for
SCD in the general population, do not fully explain
the excess of SCD in dialysis patients despite being
highly prevalent. Instead, the high prevalence of LVH,
volume changes, and electrolyte level abnormalities
are more important factors than in other patient groups.
Until definitive evidence is available, noninvasive
measurement and management of these should be
concertedly pursued with the hope of decreasing SCD
risk. This will include frequent echocardiographic
assessment of LVH and chamber size and optimal
control of blood pressure and volume status. Regular
pre- and postdialysis electrocardiographic measure-
ment of QT-interval and ST-segment changes may
indicate the need for dialysate and medication prescrip-
tion changes and consideration of referral to cardiac
specialties.
Dialysis patients should not be precluded from ICD
therapy. However, although the relative risk reduction
for SCD is good compared with dialysis patients
without ICD, actual increase in life expectancy is
limited. Current guidelines do not account for the
increase in device complications in dialysis patients,
and quality of life has not been considered in those
who have experienced cardiac arrest. Evidence to
support the more widespread use of ICDs with more
effective risk stratification is required and may derive
from ICD2 and additional trials.
Importantly, because to date there is no consistent
definition of what constitutes SCD in the dialysis
population, current data will include many deaths
from noncardiac causes. Until this is addressed, re-
sults of studies that use electrocardiography, echocar-
diography, and cardiovascular-related serum biomark-
ers to singularly assess deaths that are of heterogeneous
cause must be interpreted with caution. Equally, even
927

if the rate of SCD can be improved, it is likely that an
excess of sudden or unexpected deaths from other
causes will still prevail. Therefore, it is vital that
improved epidemiologic and post mortem data for the
likely cause of all unexpected out-of-hospital deaths
are collected prospectively.
ACKNOWLEDGEMENTS
Support: None.
Financial Disclosure: The authors declare that they have no
relevant financial interests.
REFERENCES
1. Rich A. Morbidity and mortality. In: USRDS 2008 Annual
Data Report. Bethesda, MD: National Institutes of Health, Na-
tional Institute of Diabetes and Digestive and Kidney Diseases;
2008:129-146.
2. The Sharp Collaborative Group. Should we reduce LDL
cholesterol in patients with chronic kidney disease? The results of
the Study of Heart and Renal Protection (SHARP) [abstract
LB-FC6]. ASN Renal Week, Denver, CO, November 20, 2010.
3. Zipes DP, Camm AJ, Borggrefe M, Buxton AE. ACC/AHA/
ESC 2006 Guidelines for Management of Patients with Ventricular
Arrhythmias and the Prevention of Sudden Cardiac Death. Circula-
tion. 2006;114:e385-484.
4. Chugh SS, Jui J, Gunson K, et al. Current burden of sudden
cardiac death: multiple source surveillance versus retrospective
death certificate-based review in a large U.S. community. J Am
Coll Cardiol. 2004;44(6):1268-1275.
5. Priori SG, Aliot E, Bossaert L, et al. Task Force Report Task
Force on Sudden Cardiac Death of the European Society of
Cardiology. Eur Heart J. 2001;22:1374-1450.
6. Kuller LH. Sudden death—definition and epidemiological
considerations. Prog Cardiovasc Dis. 1980;23(1):1-12.
7. Goldenberg A, Moss AJ, McNitt S, et al. Relations among
renal function, risk of sudden cardiac death, and benefit of the
implanted cardiac defibrillator in patients with ischemic left ven-
tricular dysfunction. J Cardiol. 2006;98:485-490.
8. Genovesi S, Valsecchi MG, Rossi E, et al. Sudden death and
associated factors in a historical cohort of chronic haemodialysis
patients. Nephrol Dial Transplant. 2009;24(8):2529-2536.
9. Parekh RS, Plantinga LC, Kao WHL, et al. The association
of sudden cardiac death with inflammation and other traditional
risk factors. Kidney Int. 2008;74(10):1335-1342.
10. Thomas AC, Knapman PA, Krikler DM, Davies MJ. Com-
munity study of the causes of “natural” sudden death. BMJ.
1988;297:1453-1456.
11. Leach IH, Blundell JW, Rowleyf JM, Turner DR. Acute
ischaemic lesions in death due to ischaemic heart disease. Eur
Heart J. 1995;16:1181-1185.
12. Stack AG, Bloembergen WE. Prevalence and clinical corre-
lates of coronary artery disease among new dialysis patients in the
United States: a cross-sectional study. J Am Soc Nephrol. 2001;
12(7):1516-1523.
13. Bleyer A, Hartman J, Brannon PC, Satko SG. Characteris-
tics of sudden death in hemodialysis patients. Kidney Int. 2006;88:
2268-2273.
14. Takeda K, Harada A, Okuda S, Fujimi S, Oh Y. Sudden
death in chronic dialysis patients. Nephrol Dial Transplant. 1997;
12:952-955.
15. Wanner C, Krane V, März W, et al. Atorvastatin in patients
with type 2 diabetes mellitus undergoing hemodialysis. N Engl
J Med. 2005;353(3):238-248.
928
Green et al
16. Fellström BC, Jardine AG, Schmieder RE, et al. Rosuvasta-
tin and cardiovascular events in patients undergoing hemodialysis.
N Engl J Med. 2009;360(14):1395-1407.
17. Nakano T, Ninomiya T, Sumiyoshi S, et al. Association of
kidney function with coronary atherosclerosis and calcification in
autopsy samples from Japanese elders: the Hisayama Study. Am J
Kidney Dis. 2010;55(1):21-30.
18. Saravanan P, Freeman G, Davidson NC. Risk assessment
for sudden cardiac death in dialysis patients: How relevant are
conventional cardiac risk factors? Int J Cardiol. 2010;144(3):431-
432.
19. Mangrum J, Lin D, Dimarco J, Lake D. Prognostic value of
left ventricular systolic function in renal dialysis patients. Heart
Rhythm. 2006;3(5s):154.
20. Haider AW, Larson MG, Benjamin EJ, Levy D. Increased
left ventricular mass and hypertrophy are associated with in-
creased risk for sudden death. J Am Coll Cardiol. 1998;32(5):1454-
1459.
21. Foley RN, Parfrey PS, Harnett JD, Kent GM. The prognos-
tic importance of left ventricular geometry in uremic cardiomyop-
athy. J Am Coll Cardiol. 1995;5(12):2024-2031.
22. Burton JO, Korsheed S, Grundy BJ, McIntyre CW. Hemodi-
alysis-induced left ventricular dysfunction is associated with an
increase in ventricular arrhythmias. Ren Fail. 2008;30(7):701-709.
23. Paoletti E, Specchia C, Di Maio G, et al. The worsening of
left ventricular hypertrophy is the strongest predictor of sudden
cardiac death in haemodialysis patients: a 10 year survey. Nephrol
Dial Transplant. 2004;19(7):1829-1834.
24. Krane V, Heinrich F, Meesmann M, et al. Electrocardiogra-
phy and outcome in patients with diabetes mellitus on maintenance
hemodialysis. Clin J Am Soc Nephrol. 2009;4:394-400.
25. Kovesdy C, Regidor D, Mehrotra R, Jing J. Serum and
dialysate potassium concentrations and survival in hemodialysis
patients. Clin J Am Soc Nephrol. 2004;2:999-1007.
26. Gill GV, Woodward A, Casson IF, Weston PJ. Cardiac
arrhythmia and nocturnal hypoglycaemia in type 1 diabetes—the
“dead in bed” syndrome revisited. Diabetologia. 2009;52:42-45.
27. Suarez GA, Clark VM, Norell JE, Kottke TE. Sudden
cardiac death in diabetes mellitus: risk factors in the Rochester
diabetic neuropathy study. J Neurol Neurosurg Psychiatry. 2005;76:
240-245.
28. Hadi HAR, Suwaidi JA. Endothelial dysfunction in diabe-
tes mellitus. Vasc Health Risk Manag. 2007;3(6):853-876.
29. Balcioğlu S, Arslan U, Türkoğlu S, Ozdemir M, Cengel A.
Heart rate variability and heart rate turbulence in patients with type
2 diabetes mellitus with versus without cardiac autonomic neurop-
athy. Am J Cardiol. 2007;100(5):890-893.
30. Drechsler C, Krane V, Ritz E, März W, Wanner C. Glycemic
control and cardiovascular events in diabetic hemodialysis pa-
tients. Circulation. 2009;120(24):2421-2428.
31. Di Iorio BR, Bortone S, Piscopo C, et al. Cardiac vascular
calcification and QT interval in ESRD patients: is there a link?
Blood Purif. 2006;24(5-6):451-459.
32. Lown B, Wolf M. Approaches to sudden death from coro-
nary heart disease. Circulation. 1971;44:130-142.
33. Radaelli B, Cavalli A, Latini R, Maggioni A. Multicentre
cross-sectional study of ventricular arrhythmias in chronically
haemodialysed patients. Lancet. 1988;332(8606):305-309.
34. Sforzini S, Latini R, Vincenti A, Redaelli B. Ventricular
arrhythmias and four-year mortality in haemodialysis patients.
Lancet. 1992;339:212-213.
35. Canziani ME, Cendoroglo Neto M, Saragoça M, et al.
Hemodialysis versus continuous ambulatory peritoneal dialysis:
effects on the heart. Artif Organs. 1995;19(3):241-244.
36. de Lima JJ, Vieira ML, Lopes HF, et al. Blood pressure and
the risk of complex arrhythmia in renal insufficiency, hemodialy-
Am J Kidney Dis. 2011;57(6):921-929
Investigation of Sudden Cardiac Death in Hemodialysis
sis, and renal transplant patients. Am J Hypertens. 1999;12(2, pt
1):204-208.
37. Bethge K-P. Classification of arrhythmias. J Cardiovasc
Pharmacol. 1991;17(suppl 1):S20.
38. Abe S, Yazawa T, Sloman G. Electrocardiographic abnor-
malities in patients receiving hemodialysis. Am Heart J. 1991;56:
1137-1144.
39. Kitano Y, Kasuga H, Watanabe M, et al. Severe coronary
stenosis is an important factor for induction and lengthy persis-
tence of ventricular arrhythmias during and after hemodialysis.
Am J Kidney Dis. 2004;44(2):328-336.
40. Zabel M, Klingenheben T, Franz M, Hohnloser S. Assessment
of QT dispersion for prediction of mortality or arrhythmic events after
myocardial infarction. Circulation. 1998;97:2543-2550.
41. Haigney MC, Zareba W, Gentlesk PJ, et al. QT interval
variability and spontaneous ventricular tachycardia or fibrilla-
tion in the Multicenter Automatic Defibrillator Implantation Trial
(MADIT) II patients. J Am Coll Cardiol. 2004;44(7):1481-1487.
42. Genovesi S, Rivera R, Fabbrini P, et al. Dynamic QT
interval analysis in uraemic patients receiving chronic haemodialy-
sis. J Hypertens. 2003;21(10):1921-1926.
43. Shapira OM, Bar-Khayim Y. ECG changes and cardiac
arrhythmias in chronic renal failure patients on hemodialysis. J
Electrocardiol. 1992;25(4):273-279.
44. Suzuki R, Tsumura K, Inoue T, Kishimoto H, Morii H. QT
interval prolongation in the patients receiving maintenance hemo-
dialysis. Clin Nephrol. 1998;49(4):240-244.
45. Genovesi S, Dossi C, Viganò MR, et al. Electrolyte concen-
tration during haemodialysis and QT interval prolongation in
uraemic patients. Europace. 2008;10(6):771-777.
46. Severi S, Grandi E, Pes C, et al. Calcium and potassium
changes during haemodialysis alter ventricular repolarization dura-
tion: in vivo and in silico analysis. Nephrol Dial Transplant.
2008;23(4):1378-1386.
47. Näppi SE, Virtanen VK, Saha HH, Mustonen JT, Pasternack
AI. QTc dispersion increases during hemodialysis with low-
calcium dialysate. Kidney Int. 2000;57(5):2117-2122.
48. Nearing BD. Progressive increases in complexity of T-wave
oscillations herald ischemia-induced ventricular fibrillation. Circ
Res. 2002;91(8):727-732.
49. Gehi AK, Stein RH, Metz LD, Gomes JA. Microvolt
T-wave alternans for the risk stratification of ventricular tachyar-
rhythmic events. J Am Coll Cardiol. 2005;46:75-82.
50. Chow T, Kereiakes DJ, Bartone C, et al. Prognostic utility of
microvolt T-wave alternans in risk stratification of patients with
ischemic cardiomyopathy. J Am Coll Cardiol. 2006;47(9):1820-1827.
51. Patel R, Mark P, Halliday C, et al. Microvolt T-wave
alternans in end-stage renal disease patients–associations with
uremic cardiomyopathy [published online ahead of print Novem-
ber 18, 2010]. Clin J Am Soc Nephrol. 2010 doi: 10.2215/
CJN.06370710.
52. Friedman A, Groh W, Das M. A pilot study in hemodialysis
of an electrophysiological tool to measure sudden cardiac death
risk. Clin Nephrol. 2007;68(3):159-164.
53. Nearing BD, Verrier RL. Modified moving average analysis
of T-wave alternans to predict ventricular fibrillation with high
accuracy. J Appl Physiol. 2002;92:541-549.
54. Task Force of the European Society of Cardiology, the
North American Society of Pacing Electrophysiology. Heart rate
variability, standards of measurement, physiological interpreta-
tion, and clinical use. Circulation. 1996;93:1043-1065.
55. Coquet I, Mousson C, Rifle G, et al. Influence of ischemia
on heart-rate variability in chronic hemodialysis patients. Ren Fail.
2005;27(1):7-12.
56. Nishimura M, Tokoro T, Nishida M, et al. Sympathetic
overactivity and sudden cardiac death among hemodialysis pa-
Am J Kidney Dis. 2011;57(6):921-929
tients with left ventricular hypertrophy. Int J Cardiol. 2010;
25;142(1):80-86.
57. Oikawa K, Ishihara R, Maeda T, et al. Prognostic value of
heart rate variability in patients with renal failure on hemodialysis.
Int J Cardiol. 2009;131(3):370-377.
58. Nishimura M, Tashimoto T, Kobyashi H, Fukuda T. Associa-
tion between cardiovascular autonomic neuropathy and left ventric-
ular hypertrophy in diabetic haemodialysis patients. Nephrol Dial
Transplant. 2004;19(10):2532-2538.
59. Burton JO, Jefferies HJ, Selby NM, McIntyre CW. Hemodi-
alysis-induced cardiac injury: determinants and associated out-
comes. Clin J Am Soc Nephrol. 2009;4(5):914-920.
60. Conlon P, Krucoff MW, Minda S, Schumm D, Schwab SJ.
Incidence and long-term significance of transient ST segment devia-
tion in hemodialysis patients. Clin Nephrol. 1998;49(4):236-239.
61. Narula AS, Jha V, Bali HK, et al. Cardiac arrhythmias and
silent myocardial ischaemia during hemodialysis. Ren Fail. 2000;
22(3):355-368.
62. Mohi-ud-din K, Bali HK, Banerjee S, Sakhuja V, Jha V. Silent
myocardial ischemia and high-grade ventricular arrhythmias in pa-
tients on maintenance hemodialysis. Ren Fail. 2005;27(2):171-175.
63. Pun PH, Lehrich RW, Honeycutt EF, Herzog CA, Middleton
JP. Modifiable risk factors associated with sudden cardiac arrest
within hemodialysis clinics. Kidney Int. 2011;79(2):218-227.
64. Lafrance J-P, Nolin L, Senécal L, Leblanc M. Predictors
and outcome of cardiopulmonary resuscitation (CPR) calls in a
large haemodialysis unit over a seven-year period. Nephrol Dial
Transplant. 2006;21(4):1006-1012.
65. Levy R, Dellavalle A, Atav S, Sklar AH, Stamato NJ. The
relationship between glomerular filtration rate and survival in
patients treated with an implantable cardioverter defibrillator. Clin
Cardiol. 2008;269:265-269.
66. Cuculich PS, Kerzner R, Greenberg SL, et al. Poor progno-
sis for patients with chronic kidney disease despite ICD therapy for
the primary prevention of sudden death. PACE. 2007;30:207-213.
67. Turakhia MP, Varosy PD, Tseng ZH, et al. Impact of renal
function on survival in patients with implantable cardioverter-
defibrillators. PACE. 2007;30:377-384.
68. Hreybe H, Razak E, Saba S. Effect of end-stage renal failure
and hemodialysis on mortality rates in implantable cardioverter-
defibrillator recipients. PACE. 2007;30(9):1091-1095.
69. Eckart RE, Gula LJ, Reynolds MR, et al. Mortality follow-
ing defibrillator implantation in patients with renal insufficiency.
J Cardiovasc Electrophysiol. 2004;17(9):940-943.
70. Robin J, Weinberg K, Tiongson J, et al. Renal dialysis as a
risk factor for appropriate therapies and mortality in implantable
cardioverter-defibrillator recipients. Heart Rhythm. 2006;3(10);
1197-1201.
71. Wase A, Basit A, Nazir R, et al. Impact of chronic kidney
disease upon survival among implantable cardioverter-defibrillator
recipients. J Interv Card Electrophysiol. 2004;(7):199-204.
72. Sakhuja R, Keebler M, Lai T-S, et al. Meta-analysis of
mortality in dialysis patients with an implantable cardioverter
defibrillator. Am J Cardiol. 2009;103(5):735-741.
73. Herzog CA, Li S, Weinhandl E, Strieff JW, Collins AJ. Sur-
vival of dialysis patients after cardiac arrest and the impact of
implantable cardioverter defibrillators. Kidney Int. 2005;68:818-825.
74. Dasgupta A, Montalvo J, Medendorp S, et al. Increased
complication rates of cardiac rhythm management devices in
ESRD patients. Am J Kidney Dis. 2007;49(5):656-663.
75. de Bie MK, Lekkerkerker JC, van Dam B, et al. Prevention
of sudden cardiac death: rationale and design of the Implantable
Cardioverter Defibrillators in Dialysis patients (ICD2) Trial—a
prospective pilot study. Curr Med Res Opin. 2008;24(8):2151-
2157.
929