[Downloadedfreefromhttp://www.sjkdt.orgonTuesday,January14,2014,IP:50.175.24.

210]||ClickheretodownloadfreeAndroidapplicationforthisjournal

Saudi J Kidney Dis Transpl 2010;21(3):471-477

2010 Saudi Center for Organ Transplantation Saudi Journal
of Kidney Diseases
and Transplantation

Original Article

Electrocardiographic Manifestations of Hyperkalemia

in Hemodialysis Patients
Eghlim Nemati1,2, Saeed Taheri3
1
Baqiyatallah Research Center for Gastroenterology and Liver Disease, 2Baqiyatallah University
of Medical Sciences, 3Dr. Taheri Medical Research Group, Tehran, Iran

ABSTRACT. This study was performed to evaluate whether any electrocardiogram (ECG) para-
meter can predict the presence of hyperkalemia in patients on maintenance hemodialysis (HD). In
January 2006, we conducted a cross-sectional study of 80 stable patients with end-stage renal
disease from four university-based HD units of Tehran, Iran, receiving conventional thrice-weekly
HD. Pre-HD serum electrolyte values and conventional 12-lead ECG were obtained from each pa-
tient. Bivariate linear regression was used for assessing relationship of the study variables with
hyperkalemia (K+ > 5.2 mg/dL). Multivariable logistic regression was used for evaluating inde-
pendent relationship between decreased T wave duration ( 170 ms) and other variables. Bivariate
correlation analysis showed a significant inverse correlation between serum potassium concentra-
tion and T wave duration (P< 0.05). None of the patients with serum potassium of 5.6 mg/dL
had T wave duration > 200 ms. Multivariate logistic analysis, after adjustment for other factors,
also showed a significant relationship between decreased T wave duration ( 170 ms) and hyper-
kalemia. We conclude that although hyperkalemia does not induce the usual ECG changes in HD
patients, decreased T wave duration was found to be a good indicator of this lethal condition.

Introduction sium excretion.3 Typical electrocardiographic

In patients with renal failure, sudden death is
a major reason for mortality and hyperkalemia
with cardiac arrhythmia, is one of the main
causes.1,2 In patients on hemodialysis (HD), hy-
perkalemia is known to be a consequence of
tissue break down and decreased renal potas-
Correspondence to:
Dr. Saeed Taheri
Dr. Taheri Medical Research Group,
Tehran, Iran
E-mail: taherimd@gmail.com
(ECG) manifestations include peaked T-waves
in the precordial leads, and widening of the
QRS-complex, both abnormalities of altered
cardiac conduction. Flattening or absence of the
P-wave, and a sine-wave appearance is asso-
ciated with severe hyperkalemia.4
Despite the critical importance of detecting
hyperkalemia quickly, our knowledge about the
condition is limited in patients with end-stage
renal disease (ESRD). Although, ECG is gene-
rally speculated as a reliable rapid method for
detection of potentially lethal hyperkalemia, in-
vestigators have reported that even in the pre-
[Downloadedfreefromhttp://www.sjkdt.orgonTuesday,January14,2014,IP:50.175.24.210]||ClickheretodownloadfreeAndroidapplicationforthisjournal
472
sence of profound hyperkalemia, there may be
only minimal or non-specific ECG changes, or
even no changes in HD patients, suggesting that
one cannot entirely rely on the ECG features
of hyperkalemia in renal failure.5-11
This study was performed to evaluate the pre-
sence of expected ECG features in HD patients
with hyperkalemia, and to determine whether
any of the ECG parameters could predict the
existence of hyperkalemia in this population.
Materials and Methods
The patients were selected from a pool of 111
patients with ESRD receiving HD at four out-
patient HD units in Tehran, Iran. Exclusion
criteria included:
a) medically unstable patients;
b) atrial fibrillation;
c) under 18-years of age;
d) less than three months on HD;
e) pregnant women;
f) mentally disabled; and
g) immeasurable T waves.
Verbal consent was taken from all patients.
Eighty patients were included in the study. All
patients underwent conventional HD for an
average of four hours, three times a week. The
blood flow rate during HD was 300-350 mL/
min. The dialysate composition was: sodium 140
mEql/L, potassium 2.5 mEq/L, bicarbonate 34
mEq/L, acetate 3.0 mEq/L, chloride 109 mEq/L,
calcium 3.0 mEq/L, magnesium 1.0 mEq/L, glu-
cose 1 g/L. Patients were categorized into four
groups on the basis of their serum potassium
concentration (Group-I: < 4.4 mg/dL; Group-II:
4.4-4.9 mg/dL; Group-III: 4.9-5.2 mg/dL; and
Group-IV: > 5.2 mg/dL).
The serum albumin, serum potassium, total
serum calcium, and serum sodium levels were
measured in a blood sample obtained just be-
fore a routine dialysis session. Laboratory mea-
surements were performed at Baqiyatallah hos-
pitals central laboratory without any delay af-
ter sample collection; all samples were imme-
diately analyzed using standard laboratory tech-
niques. Hyperkalemia was defined as a serum
K+ concentration > 5.2 mg/dL.
A conventional 12-lead ECG was recorded
Nemati E, Taheri S
ten minutes before every HD at a paper speed
of 25 mm/s for all patients. All ECGs were ob-
tained after a five-minute resting period, with
the patient lying comfortably in the supine po-
sition. For the evaluation of ECG parameters,
all ECGs were enlarged twice. Three consecu-
tive cardiac cycles were measured and ave-
raged. All ECGs were read and analyzed by
one investigator without knowledge of patients
laboratory results. ECG parameters records
included P, T and S amplitudes, heart rate, PR
interval, ST duration, QRS duration, T dura-
tion, and QT interval. Each QT interval and T
wave duration were corrected for patient heart
rate: QTc = QTint/RR and Tdc = Tdur/RR
(in milliseconds [ms]). Wave amplitudes and
durations were measured from the first deflec-
tion to the point of offset regarding the isoelec-
tric TP baseline.
Data were expressed as mean SD. The se-
rum potassium concentration was sorted into
quartiles and the relationships of mean diffe-
rences between Groups were analyzed using
one way ANOVA and Chi-square methods.
Post hoc differences between patient-Groups
were examined using Tukeys test for multiple
comparisons. T-test was used to assess the
difference between serum potassium values in
patients with T wave duration (ms) 170 and
> 170. P< 0.05 was considered significant. The
independent association of S wave amplitude,
T wave duration, and corrected T wave dura-
tion (Tdc) with all other variables were ana-
lyzed using multiple linear regression. Multi-
ple logistic regression was used for evaluating
independent relationship between T wave du-
ration and other variables. Variables with a P
value < 0.1 were entered to multivariate eva-
luations. Computations were done in SPSS (Sta-
tistics Program for Social Sciences version
11.5, SPSS Inc., Chicago, IL, USA).
Results
There were 44 males and 36 females with
mean age of 52.6 17.7 years (range 19-82
years). The mean duration of ESRD was 47
47 months (range 3-229 months). Causes of re-
nal failure included diabetes mellitus in 20 (25%)
 [Downloadedfreefromhttp://www.sjkdt.orgonTuesday,January14,2014,IP:50.175.24.210]||ClickheretodownloadfreeAndroidapplicationforthisjournal
ECG of hyperkalemia in HD patients
Table 1. Comparison of subjects electrocardiogram parameters by quartiles of serum potassium (K) concentration using one way ANOVA method
Variables Serum potassium concentration (mg/dL) Total P-value
< 4.4 4.4-4.9 4.9-5.2 > 5.2
PR int (sec) 0.18 0.04 0.17 0.03 0.19 0.03 0.18 0.03 0.18 0.03 0.575
(0.15, 0.21)* (0.15, 0.20) (0.16-0.22) (0.17-0.20) (0.17-0.19)
ST elevation -0.27 1.16 -0.31 0.71 0.05 1.18 0.01 0.73 -0.71 0.85 0.556
(-1.05, 0.51) (-0.80-0.16) (-0.85-0.96) (-0.20-0.22) (-0.26-0.12)
QRS dur (sec) 0.080 0.02 0.080 0.007 0.085 0.028 0.082 0.015 0.083 0.017 0.891
(0.064, 0.103) (0.077, 0.088) (0.064, 0.106) (0.077, 0.088) (0.072, 0.092)
QT (sec) 0.39 0.12 0.37 0.05 0.37 0.037 0.40 0.048 0.40 0.05 0.852
(0.27, 0.48) (0.25, 0.48) (0.33, 0.41) (0.38, 0.41) (0.22, 0.54)
QTc (sec) 0.36 0.12 0.33 0.06 0.34 0.06 0.35 0.07 0.35 0.08 0.781
(0.28, 0.44) (0.29, 0.37) (0.29, 0.38) (0.33, 0.38) (0.33, 0.37)
Tdur (sec) 0.20 0.04 0.19 0.036 0.18 0.02 0.17 0.03 0.18 0.03 0.062
(0.16, 0.24) (0.16, 0.22) (0.17, 0.20) (0.16, 0.18) (0.16, 0.21)
Tdc (sec) 0.18 0.05 0.16 0.03 0.17 0.03 0.15 0.03 0.16 0.03 0.104
ECG findings (0.15,0.22) (0.14, 0.19) (0.15-0.19) (0.15, 0.17) (0.15, 0.17)
T amplitude 4.50 2.95 4.55 3.57 3.89 5.04 5.23 3.19 4.88 3.41 0.682
(2.52-6.48) (2.25-6.85) (0.01-7.76) (4.30-6.17) (4.11-5.65)
P amplitude 1.50 0.45 1.58 0.50 1.53 0.52 1.40 0.50 1.46 0.50 0.896
(1.17, 1.84) (1.23, 1.86) (1.08, 1.87) (1.28, 1.57) (1.35, 1.57)
R amplitude 11.6 6.73 11.55 4.36 12.33 8.38 9.93 6.70 10.66 6.58 0.675
(6.78, 16.42) (8.61, 14.48) (7.12, 18.67) (7.96, 11.89) (9.16, 12.15)
S amplitude** 16.20 10.17 15.09 7.23 8.78 6.74 14.41 6.62 14.23 7.76 0.104
(8.90, 23.04) (11.45, 20.94) (3.67, 14.03) (12.34, 16.82) (12.78, 15.23)
Heart Rate (/min) 71.86 19.56 82.43 17.93 76.40 18.80 73.40 15.98 74.89 16.72 0.664
(54.57, 85.78) (67.84, 94.80) (59.87, 94.39) (60.31, 86.90) (62.34, 87.45)
*95% confidence interval for mean, **Negative values were considered positive, Tdur: T duration, Tdc: T duration corrected

473
[Downloadedfreefromhttp://www.sjkdt.orgonTuesday,January14,2014,IP:50.175.24.210]||ClickheretodownloadfreeAndroidapplicationforthisjournal

474 Nemati E, Taheri S

Table 2. Results of multiple linear regression analysis with S wave amplitude as the outcome variable
Variable Standard regression coefficient SE P value
HD duration .229 .025 .178
Age -.065 0.032 .652
Serum calcium concentration -.417 1.200 .020
Serum sodium concentration .006 .245 .972
Serum albumin value .299 .150 .110
Serum potassium concentration (> 5.2) .056 2.255 .746
Heart rate .000 .069 .998
Sex (female) -.213 2.306 .228

and glomerulonephritis in 22 patients (28%).
Other causes included: hypertension in five
(6.5%); polycystic kidney disease, systemic
lupus erythematosus and vasculitis in one pa-
tient each (1.5%) and unknown etiology in 30
patients (37.5%). The mean pre-dialysis serum
potassium concentration was 5.2 0.77 mg/dL
(range 2.5-7; Group-I: 11 patients (13.5%);
Group-II: 11 patients (13.5%); Group-III: nine
patients (11.5%); and Group-IV: 47 patients
(59%).
Forty-seven patients (59%) had hyperkalemia
(serum K > 5.2 mg/dL). The pre-HD mean SD
(range) measurements of these patients were:
heart rate 77 14/min (range: 52 - 110); serum
albumin 5.5 7.1 mg/dL (2.85 - 6.23), sodium
(Na) 138.5 3.9 mg/dL (125 - 146) and serum
total calcium (Ca) was 8.9 1.0 mg/dL (6.1 -
11.8). None of these patients had arrhythmia
during the study period. The Groups were
equivalent in their serum electrolytes concen-
tration (P> 0.05). None of the ECG parame-
ters showed any significant differences between
quartiles of serum potassium concentration (P
> 0.05). Details are shown in Table 1.
However, when we divided the study sub-
jects, based on their serum K+ levels into two
groups (> 5.2 and 5.2 mg/dL), significant di-
fference in their T wave duration (Tdur) was
observed (P= 0.008). Results of multiple linear
regression analysis with S wave amplitude as
the outcome variable in relation to hyperkalemia
lemia is shown in Table 2.
Multivariate logistic analysis after adjustment

Table 3. Results of multivariate logistic regression with respect to T wave duration corrected (Tdc 120 ms)
Variables OR CI (95%) Sig
Age
> 55 yr 1.635 .542 4.737 0.623
55 yr 1
Serum calcium concentration (mg/dL)
> 11.5
< 8.5 1.188 0.300 4.707 0.806
8.5 & 11.5 1
Serum sodium concentration (mg/dL)
> 145 0.882 0.069 11.303 0.924
< 135 1.369 .216 8.674 0.739
135 & 145 1
Serum potassium concentration (mg/dL)
> 5.2 9.712 1.098 85.905 .041
5.2 1
Sex
Male 0.853 .219 3.314 .818
Female 1
HD duration
> 4 yr 2.430 .600 9.836 .213
4 yr 1
[Downloadedfreefromhttp://www.sjkdt.orgonTuesday,January14,2014,IP:50.175.24.210]||ClickheretodownloadfreeAndroidapplicationforthisjournal
ECG of hyperkalemia in HD patients
Figure 1. Relationship between ECG T-wave du-
ration and pre-dialysis serum potassium level in
HD patients
Figure 2. Relation between ECG corrected T-
wave duration and pre-dialysis serum potassium
level in HD patients
for other factors also revealed significant re-
lationship between decreased T wave duration
( 170 ms) and hyperkalemia (P = 0.009, OR:
5.98, 95% CI: 1.56-22.79). Decreased correc-
ted T wave duration (Tdc 120 ms) was also
associated with hyperkalemia (Table 3).
Bivariate correlation analysis found signifi-
cant inverse correlations between serum K+
and T wave duration (Figure 1), serum K and
corrected T wave duration (Figure 2), and serum
475
Figure 3. Relation between ECG S-wave ampli-
tude and pre-dialysis serum calcium level in HD
patients
Ca concentration and S amplitude (Figure 3).
Discussion
We found a prevalence of 59% of hyperka-
lemia among our HD patients, which is com-
parable with other studies.4 Although we did
not observe typical ECG manifestations asso-
ciated with hyperkalemia, we found a signifi-
cant association between decreased T wave
duration ( 170 ms) and pre-HD hyperkalemia
(> 5.2 mEq/L) in our HD patients. We also
found that there was a significant inverse re-
lation between serum Ca++ and S wave ampli-
tude with higher calcium concentration related
to deeper S waves.
Our findings confirmed that the typical ECG
manifestations attributed to hyperkalemia, are
rare, especially among severely hyperkalemic
HD patients.5-10 Despite this, none of the pre-
vious studies had investigated all ECG para-
meters and their potential relationship with
electrolyte disturbances in this population.
Aslam et al reported that patients on HD show
no specific ECG manifestations in relation with
hyperkalemia; however, in that survey the in-
vestigators had only assessed the T wave am-
plitude and T wave to R wave amplitudes ratio
and their association with serum potassium and
[Downloadedfreefromhttp://www.sjkdt.orgonTuesday,January14,2014,IP:50.175.24.210]||ClickheretodownloadfreeAndroidapplicationforthisjournal
476
calcium concentration.5 In contrast with our
findings, they reported an inverse relation bet-
ween T wave amplitude and serum calcium
concentration, whereas we found a reverse re-
lationship between serum calcium value and S
wave amplitude (Figure 2). On the other hand,
they also had excluded patients using agents
which may counteract cardiac effects of hyper-
kalemia. Nevertheless, when we also excluded
such subjects, our findings did not change con-
siderably and the found associations remained
significant as well (data not shown). Szerlip et
al also reported two HD patients with profound
hyperkalemia (serum K > 9 mg/dL) without
any of the expected ECG features. In another
case report, Dowod et al11 reported a HD pa-
tient with a serum potassium concentration of
10.3 mg/dL, who had only non-specific ST in-
terval and T wave duration changes.8
Previous data have suggested that the rate of
rise of serum potassium level is more relevant
compared to its actual serum concentration.
They explained that a slow elevation in serum
potassium value, like in ESRD patients, permits
compensatory mechanisms to minimize the ef-
fects of hyperkalemia.6 Despite this, the rele-
vance of the condition should not be underes-
timated. In a case series, Siddiqui et al12 re-
ported three deaths attributed to hyperkalemia
in HD patients with serum potassium concen-
trations from 7.4 to 8.0. The most important
relevance of the problem is when a HD patient
comes to the emergency department for any
medical problem. Without typical ECG features
of hyperkalemia among HD patients, the atten-
ding doctors may fail to diagnose; hence the
potential life threatening consequences of hy-
perkalemia may occur. Therefore, we should
consider hyperkalemia as a very critical condi-
tion to be aware of, especially in patients with
impaired renal function and must educate emer-
gency room staff and physicians for the rele-
vance and unusual ECG manifestations of hy-
perkalemia in HD patients.
Anyhow, although we did not detect classic
ECG features of hyperkalemia in our HD po-
pulation, there was a surprising new relation-
ship between decreased T wave duration and
the condition. Since, some of our patients with
Nemati E, Taheri S
serum potassium concentrations over 5.2 mEq/L
had T wave durations longer than 170 ms, one
may claim that complete reliance on this ECG
feature may lead to loss of some of patients
with lethal hyperkalemia; but, on the other hand,
none of the subjects with serum potassium >
5.6 had a T wave duration over 200 ms.
In summary, although hyperkalemia does not
induce usual ECG changes in all HD patients,
it significantly decreases T wave duration. Fur-
ther studies with larger study population seem
necessary to determine more precise T wave du-
ration point in association with hyperkalemia.
Acknowledgement
This work was supported by grants from
Baqiyatallah University of Medical Sciences.
The authors would gratefully acknowledge in-
valuable assistance of Professor Guy Neild in
editing the manuscript.
References
1. Chazan JA. Sudden death in patients with chro-
nic renal failure on hemodialysis. Dial Transplant
1987;16:447-8.
2. Morrison G, Michelson EL, Brown S,
Morganroth J. Mechanism and prevention of
cardiac arrhythmias in chronic hemodialysis
patients. Kidney Int 1980;17: 811-9.
3. Ahmed J, Weisberg LS. Hyperkalemia in dia-
lysis patients. Semin Dial 2001;14(5):348-56.
4. Mattu A, Brady WJ, Perron AD. Electrocardio-
graphic manifestations of hypothermia. Am J
Emerg Med 2002;20(4):314-26.
5. Morales MA, Gremigni C, Dattolo P, et al.
Signal-averaged ECG abnormalities in haemo-
dialysis patients. Role of dialysis. Nephrol Dial
Transplant 1998;13(3):668-73.
6. Aslam S, Friedman EA, Ifudu O. Electrocar-
diography is unreliable in detecting potentially
lethal hyperkalaemia in haemodialysis patients.
Nephrol Dial Transplant 2002;17:1639-42.
7. Wrenn KD, Slovis CM, Slovis BS. The ability
of physicians to predict hyperkalemia from the
ECG. Ann Emerg Med 1991;20(11):1229-32.
8. Dowod TA, Ali JI, Burud S. Profound
hyperkalemia and the electrocardiogram. Lack
of correlation: A case report. Kuwait Med J
2006;38(3):232-4.
[Downloadedfreefromhttp://www.sjkdt.orgonTuesday,January14,2014,IP:50.175.24.210]||ClickheretodownloadfreeAndroidapplicationforthisjournal
ECG of hyperkalemia in HD patients
9. Martinez-Vea A, Bardaji A, Garcia C, Oliver
JA. Severe hyperkalemia with minimal electro-
cardiographic manifestations: A report of
seven cases. J Electrocardiol 1999;32(1):45-9.
10. Esposito C, Bellotti N, Fasoli G, Foschi A,
Plati AR, Dal Canton A. Hyperkalemia induced
ECG abnormalities in patients with reduced
renal function. Clin Nephrol 2004;62(6):465-8.
477
11. Szerlip HM, Weiss J, Singer I. Profound hyper-
kalemia without electrocardiographic manifest-
tations. Am J Kidney Dis 1986;7(6):461-5.
12. Siddiqui JY, Fitz AE, Lawton RL, Kirkendall
WM. Causes of death in patients receiving
long-term hemodialysis. J Am Med Assoc
1970;212:1350-4.