J Am Soc Nephrol 10: 1297–1302, 1999 QT Dispersion in Patients with End-Stage Renal Failure and during Hemodialysis

Љ ISTVAN´ LORINCZ, JANOS´ MATYUS,´ ZSOLT ZILAHI, CSABA KUN, ZSOLT KARANYI,´ and GYORGY¨ KAKUK 1st Department of , University of Debrecen, Hungary.

Abstract. Interlead variability of the QT interval in surface ms postdialysis; the difference was not significant. The maxi- electrocardiogram (ECG), i.e., QT dispersion, reflects regional mal QT interval changed significantly from 449 Ϯ 43 to 469 Ϯ differences in ventricular recovery time, and it has been linked 41 ms (P Ͻ 0.01). The corrected maximal QT interval in- to the occurrence of malignant in different cardiac creased significantly from 482 Ϯ 42 to 519 Ϯ 33 ms (P Ͻ diseases. The purpose of the study was to assess the effect of 0.01). The QT dispersion changed from 56 Ϯ 15 to 85 Ϯ 12 ms hemodialysis on QT and corrected QT (QTc) interval and (P Ͻ 0.001) and the corrected QT interval dispersion from dispersion in chronic hemodialyzed patients. Data of 34 non- 62 Ϯ 18 to 95 Ϯ 17 ms (P Ͻ 0.001). During hemodialysis, the diabetic patients (male/female ϭ 21/13; mean age, 54 Ϯ 15 yr) serum potassium and phosphate levels decreased from 5.5 Ϯ on chronic hemodialysis were studied. Polysulfone capillaries 0.8 to 3.9 Ϯ 0.5 (mM) and from 2.3 Ϯ 0.5 to 1.6 Ϯ 0.4 (mM), and bicarbonate dialysate containing (in mEq/L) 135 Naϩ, 2.0 respectively, whereas calcium increased from 2.2 Ϯ 0.23 to Kϩ, 1.5 Ca2ϩ, and 1.0 Mg2ϩ were used. Simultaneous 12-lead 2.5 Ϯ 0.22 (mM). It is concluded that hemodialysis increases ECG were recorded before and after hemodialysis in a standard the QT and QTc interval and QT and QTc dispersion in setting. The QT intervals for each lead were measured manu- patients with end-stage renal failure. Thus, it may be stated that ally on enlarged (ϫ3) ECG by one observer using cali- the nonhomogeneity of regional ventricular repolarization in- pers. Each QT interval was corrected for patient creases during hemodialysis. Measurement of QT and QTc rate: QTc ϭ QT/͌RR (in milliseconds [ms]). The average dispersion is a simple bedside method that can be used for cycle intervals were 853 Ϯ 152 ms predialysis and 830 Ϯ 173 analyzing ventricular repolarization during hemodialysis.

Patients with end-stage renal failure commonly have different in conduction velocity and/or repolarization in the different cardiovascular diseases. Although a decline in cardiovascular parts of the could provide a substrate for tachyar- death has recently been observed in the general population, a rhythmias (8,9). Experimental data have demonstrated a strong similar trend has not been seen in dialysis patients (1,2). link between the vulnerability of the ventricular myocardium According to a number of recent reports, it is known that half and increased temporal dispersion of refractoriness (10). The of the patients receiving chronic hemodialysis die of pathogenic role of increased ventricular repolarization nonho- . The main causes are congestive heart mogeneity was confirmed by clinical and experimental studies failure, , and sudden death as a result of (10,11). The exact spatial dispersion of ventricular myocar- hyperkalemia or (3–5). Reported rates of sudden dium can be determined by invasive methods such as the death in these patients range from 1.4 to 25% (3,6). measurement of monophasic action potentials of the left and However, hemodialysis patients have a wide variety of elec- right ventricle, but it can also be measured by a noninvasive trocardiographic (ECG) abnormalities and, in certain instances, method like surface mapping. Both methods are time consum- hemodialysis itself seems to be a cause of ECG changes and ing and require special technical devices, which limit their use different kinds of dysrhythmias. Arrhythmias are often ob- in everyday practice (11). served after the start of hemodialysis and last at least 5 h after In everyday clinical practice, different methods of surface dialysis (5–7). ECG are being studied for their ability to predict either the The arrhythmogeneity depends partly on variable factors occurrence of these ventricular arrhythmias or their clinical such as autonomic tone, and partly on abnormalities in ven- relevance (6,11). In the conventional ECG, the prolonged QT tricular anatomical structure and metabolism. Nonhomogeneity interval has been reported to be associated with arrhythmogen- esis in a number of cardiac disorders (12). Recent studies have indicated that interlead variability of the QT interval in surface Received May 19, 1998. Accepted December 15, 1998. 12-lead ECG (i.e., the QT interval dispersion defined as the Preliminary data have been presented previously in a letter (see reference (1)). difference between maximal and minimal QT interval dura- Correspondence to Dr. Istva´n Lorincz,Љ Debrecen, Do´czy u. 24. 4032, Hungary. Phone: 36 52 411 600; Fax: 36 52 414 951; E-mail: [email protected] tion) reflects better the regional differences in ventricular re- 1046-6673/1006-1297 covery time. This QT dispersion has been linked to the occur- Journal of the American Society of rence of arrhythmias in patients with congenital long QT Copyright © 1999 by the American Society of Nephrology syndromes or with drug-induced , and sudden 1298 Journal of the American Society of Nephrology J Am Soc Nephrol 10: 1297–1302, 1999 death in patients with congestive , hypertrophic Table 1. Clinical data of patientsa , hypertensive heart disease, pro- lapse syndrome, etc. (9,11,13–19). Furthermore, there are data No. of patients 34 about the increased risk of intraoperative death in two patients Male 21 with renal and liver transplants who had increased QT disper- Female 13 Ϯ sions (128 ms and 125 ms separately) before (20). Average age (yr) 54 15 (20 to 77) Ϯ Moreover, in patients awaiting cardiac transplantation and who Average duration of chronic HD 24 18 Ϯ had QT dispersion longer than 140 ms, there was a fourfold ( SD, months) Ϯ greater risk of death than in patients with QT dispersion shorter Weight loss during dialysis (L) 1.9 1.3 Ϯ than 140 ms. QT dispersion better predicts the risk of sudden Systolic BP at the beginning of HD 150 26 Ϯ death in patients awaiting heart transplantation than the left ( SD, mmHg) Ϯ ventricular ejection fraction (21). Systolic BP at the end of HD 141 24 Ϯ The purpose of our study was to assess the effect of hemo- ( SD, mmHg) Ϯ dialysis on QT and corrected QT (QTc) interval and QT and Diastolic BP at the beginning of 82 11 Ϯ QTc interval dispersion in patients with end-stage renal failure HD ( SD, mmHg) Ϯ 10 min before (pre-HD) and 10 min after each hemodialysis Diastolic BP at the end of HD 80 11 Ϯ (post-HD). ( SD, mmHg) Cause of end-stage renal failure Materials and Methods glomerulonephritis 18 (53%) Patients tubulointerstitial nephritis 7 (20%) The data of 34 adult patients, consisting of 21 men and 13 women nephrosclerosis 5 (15%) on chronic hemodialysis with a mean age of 54 Ϯ 15 yr (range, 21 to autosomal dominant polycystic 2 (6%) 78), were studied. The causes of chronic renal failure were as follows: kidney disease glomerulonephritis (n ϭ 18), tubulointerstitial nephritis (n ϭ 7), renal tubular acidosis 1 (3%) nephrosclerosis (n ϭ 5), polycystic kidney disease (n ϭ 2), and other unknown origin 1 (3%) (n ϭ 2) (Table 1). Exclusion criteria were: (1) unmeasurable T waves; Cardiac complications (2) atrial ; (3) ; and (4) antiarrhythmic pectoris 7 (20%) drugs that lengthen the QT interval. The mean duration of hemodial- old 2 (6%) Ϯ ysis was 22 19 mo (range, 1 to 77). The dialyses were carried out 2 (6%) in a standard setting (Fresenius 2008 device; Fresenius Medical Care, cardiomyopathy 4 (12%) Bad Homburg, Germany) with F6 and F8 polysulfone capillaries (Fresenius) for 3.5 to4h2or3times per week. Bicarbonate dialysate sick sinus syndrome 1 (3%) containing (in mM) 135 Naϩ, 2.0 Kϩ, 1.5 Ca2ϩ, and 1.0 Mg2ϩ was supraventricular 1 (3%) used. During hemodialysis, no drug therapy was applied, except Medication isotonic NaCl or hypertonic (10%) NaCl infusions and sodium hep- digoxin 4 (12%) arin. The maintenance drug therapy including digitalis, antihyperten- antiarrhythmic medication (beta 8 (24%) sive, antianginal, and beta blocking agents was not changed. The blockers) following laboratory tests were performed before hemodialysis and at other (antihypertensive drugs) 12 (34%) the end of treatment: serum electrolyte Naϩ,Kϩ,Mg2ϩ,Ca2ϩ, phosphate, blood nitrogen, and creatinine. Arterial BP and heart a HD, hemodialysis. rate were recorded during the procedure. Intact parathyroid hormone level was determined by RIA (BioRad; normal value 1.2 to 6.8 for each lead were measured manually with calipers by one observer. pmol/L). Different data subgroups were created according to age; The QT interval was measured from the first deflection of the QRS gender; the duration of the hemodialysis program; the presence or complex to the point of offset, defined by the return of the absence of ischemic heart disease; the presence or absence of hyper- terminal T wave to the isoelectric TP baseline. In the presence of U tension; the use or nonuse of beta blocking agents; and the presence wave interrupting the T wave, the terminal portion of the visible T or absence of severe hyperparathyroidism (if the parathyroid hormone wave was extrapolated to the TP baseline to define the point of T [PTH] level was Ͼ40 pmol/L). wave offset. If the end of T wave could not be reliably determined, the lead was not included in the analysis. Each QT interval was corrected Methods of Measurement of QT and QTc Interval, and for patient heart rate using Bazett’s formula: QTc ϭ QT/͌RR (ms), QT and QTc Dispersion where QTc is the corrected QT interval. QT, QTc dispersions were Conventional 12-lead ECG were recorded for assessment of QT defined as differences between the minimal and maximal QT and QTc dispersion, 10 min before and 10 min after every hemodialysis. values in each of the 12 leads studied (22). Simultaneous 12-lead ECG were recorded by means of a 12-channel ECG recorder (Hewlett Packard Page Writer 200i) at a paper speed of Reproducibility 25 mm/s. On every occasion, the ECG were obtained after a 5-min To determine the intraobserver variability of QT interval and QT resting period, with the patients lying comfortably in the supine interval dispersion measurement, all ECG strips were evaluated by position. For the analysis of the QT interval, the 12-lead ECG were one investigator on two different occasions. The intraobserver vari- enlarged on the same photocopier by a factor of three. Three consec- ability was demonstrated by Bland and Altman plot methods (Figures utive cardiac cycles were measured and averaged. The QT intervals 1 and 2). The mean difference in QT interval between the two J Am Soc Nephrol 10: 1297–1302, 1999 QT Dispersion in Patients with ESRD 1299

Figure 1. Bland and Altman plots for reproducibility of QT interval between two separate measurements.

Figure 3. Relative risk (RR) intervals, QTmax, and maximum cor- rected QT (QTcmax) interval before (pre-HD) and after (post-HD) hemodialysis in patients with end-stage renal failure.

Figure 2. Bland and Altman plots for reproducibility of QT dispersion between two separate measurements.

measurements, the relative error, was 4.5% and in the QT dispersion it was 11%.

Statistical Analyses Ϯ Data were expressed as mean SD. The relationship of mean Figure 4. QT and QTc dispersion before (pre-HD) and after (post-HD) differences between intervals and dispersions in groups (pre-HD and hemodialysis in patients with end-stage renal failure. post-HD) and differences among subgroups (ischemic heart disease, hypertension, gender, parathyroid hormone level) were analyzed us- ing ANOVA. A linear regression was used for evaluation of correla- The maximal QT interval changed significantly from 449 Ϯ 43 tion between intervals and dispersions, and to compare ages and Ϯ Ͻ duration of HD program. P Ͻ 0.05 was considered significant. The to 469 41 ms (P 0.01). The corrected maximal QT interval Ϯ Ϯ intraobserver variabilities of ECG measurements were calculated as (QTc) increased significantly from 485 42 to 519 34 ms relative errors according to the formula [A Ϫ B]/[A ϩ B]/2, where A (P Ͻ 0.03). The QT interval dispersion changed from 57 Ϯ 15 represents the first and B the second measurement (23). Statistical to 85 Ϯ 13 ms (P Ͻ 0.001), and the corrected QT interval package SAS for Windows 6.12 version was used for the analysis. dispersion from 62 Ϯ 18 to 95 Ϯ 17 ms (P Ͻ 0.001). Both QT and QTc intervals and QT and QTc dispersions increased Results significantly at the end of hemodialysis (P Ͻ 0.001). There The data of patients were classified into two groups: Data at were no differences between the average values of the QT and the beginning of hemodialysis were grouped under pre-HD, QTc intervals and between the QT and QTc dispersions. and the data at the end of hemodialysis under post-HD. Our During hemodialysis, the serum potassium level decreased results are summarized in Figures 3 and 4. They show the mean from 5.5 Ϯ 0.8 to 3.97 Ϯ 0.55 mM and phosphate level from values with SD of measured relative risk (RR) intervals, the 2.37 Ϯ 0.51 to 1.58 Ϯ 0.38 mM, whereas calcium increased maximal QT interval and QT dispersion, and the maximal QTc from 2.2 Ϯ 0.23 to 2.5 Ϯ 0.23 mM. Both potassium and interval and QTc dispersion of groups pre-HD and post-HD. phosphate level changes were significant (P Ͻ 0.001), as was There were no significant differences between the RR inter- the change in calcium level, which increased significantly (P Ͻ vals. The average cycle intervals were 853 Ϯ 152 ms before 0.001). There were no changes in the levels of sodium and hemodialysis and 830 Ϯ 175 ms after hemodialysis (P ϭ 0.52). magnesium (Table 2). 1300 Journal of the American Society of Nephrology J Am Soc Nephrol 10: 1297–1302, 1999

Table 2. Laboratory findings before and after hemodialysisa on the autonomic tone, and partly on abnormal ventricular structure and metabolism (24). Increased dispersion of refrac- Parameter Pre-HD Post-HD P Value toriness may be due to the regional differences of ventricular Naϩ (mM) 139 Ϯ 3.9 138 Ϯ 2.07 0.29 wall stress (mechanoelectric or contraction-excitation feed- Kϩ (mM) 5.53 Ϯ 0.83 3.97 Ϯ 0.55 0.001 back), which is caused by ventricular dilation and fibrosis (e.g., Ca2ϩ (mM) 2.2 Ϯ 0.23 2.54 Ϯ 0.23 0.001 in chronic heart failure). Myocyte hypertrophy and increased Phosphate (mM) 2.37 Ϯ 0.51 1.58 Ϯ 0.38 0.001 interstitial matrix were demonstrated. Myocyte hyper- ϩ Mg2 (mM) 1.13 Ϯ 0.26 1.11 Ϯ 0.29 0.772 trophy may cause a lengthening of action potential duration, BUN (mM) 33.3 Ϯ 6.8 14.9 Ϯ 3.4 0.001 and increased interstitial fibrosis may be associated with re- Creatinine (␮M) 965 Ϯ 219 480 Ϯ 116 0.001 duced action potential amplitude and membrane potential, and shortened action potential (25). As seen in patients with cardiac a HD, hemodialysis; BUN, blood urea nitrogen. failure, the myocardial fibrosis, especially PTH-dependent in- termyocardiocytic fibrosis (26) and probably the myocardial calcification, also influence the homogeneity of ventricular As a result of the comparison of different subgroups, the repolarization, which might lead to increased QT dispersion in basal (pre-HD) maximal QT and QTc interval prolongation and our uremic patients. QT and QTc abnormalities in uremic QT and QTc interval dispersion were independent of gender, patients may be partly due to autonomic failure caused by age of patients, hypertension, and duration of chronic hemo- uremic (27). The impaired cardiac per- dialysis program and hyperparathyroidism, but not of concom- formance and impaired potassium, calcium, and/or phosphate itant cardiac disease (Tables 3 and 4). There was no correlation metabolism may be major contributors to the pathogenesis of between the increase of QT and QTc prolongation and max max arrhythmias in uremic patients receiving hemodialysis. There QT and QTc interval dispersion at the end of HD (post-HD) are some contributing factors according to publications in this and the patients’ age, gender, hypertension, duration of chronic dialysis treatment, and concomitant cardiac abnormalities, field: (1) the differences between the serum and intracellular namely, ischemic heart disease (Tables 3 and 4). potassium, or rapid decrease of serum potassium (5,28,29); (2) serum calcium and magnesium level changes (30,31); (3) de- Discussion crease of the circulatory volume (30); (4) fast correction of metabolic acidosis (29,31); (5) increased PTH levels (31,32); This study shows a slightly increased basal, pre-HD QTmax, and (6) increased free fatty acid level (33). In our cases, the QTcmax interval, and QT, QTc interval dispersions. At the end prolongation of QT and QTc interval and slightly increased of HD (post-HD), the data show significant increases in QTmax basal pre-HD QT and QTc dispersion correlated with a pres- and QTcmax interval prolongation and QT and QTc interval dispersion in patients with end-stage renal failure receiving ence of ischemic heart disease but not with hypertension, hemodialysis. gender, age of patients, or duration of chronic dialysis or beta The mechanism of QT dispersion prolongation in patients blocking therapy. According to several publications, the nor- with end-stage renal failure during hemodialysis is a matter of mal range for QT dispersion is 40 to 50 ms with a maximum speculation. In general, the arrhythmogeneity depends partly of 65 ms, and if the QT dispersion values are greater than 65

a Table 3. The results of QTmax interval and QT interval dispersion in subgroups of hemodialyzed patients

QTmax Interval (ms) QT Dispersion (ms) Subgroup No. of Patients Pre-HD Post-HD Pre-HD Post-HD

IHD 9 451.4 Ϯ 26.5 468.1 Ϯ 34.0 60.29 Ϯ 16.87b 89.14 Ϯ 12.73 No IHD 25 449.5 Ϯ 45.7 466.1 Ϯ 46.0 47.25 Ϯ 6.61 75.75 Ϯ 5.90 BB 8 444.6 Ϯ 41.8 469.1 Ϯ 32.0 57.00 Ϯ 20.00 87.29 Ϯ 10.36 No BB 26 451.7 Ϯ 41.6 465.9 Ϯ 45.9 56.59 Ϯ 14.77 84.86 Ϯ 13.53 HYP 29 447.8 Ϯ 38.1 462.4 Ϯ 40.2 55.88 Ϯ 16.32 86.04 Ϯ 13.57 No HYP 5 461.2 Ϯ 57.1 488.0 Ϯ 52.2 61.75 Ϯ 12.55 81.75 Ϯ 3.5 PTH Ͻ40 pmol/L 25 447.4 Ϯ 39.0 458.9 Ϯ 43.1 55.90 Ϯ 15.45 83.25 Ϯ 9.7 PTH Ͼ40 pmol/L 9 456.1 Ϯ 47.1 484.8 Ϯ 37.2 58.44 Ϯ 17.34 90.33 Ϯ 17.4 Male 21 456.0 Ϯ 36.9 468.3 Ϯ 46.7 57.87 Ϯ 16.40 83.79 Ϯ 12.21 Female 13 439.7 Ϯ 47.3 463.8 Ϯ 36.1 54.50 Ϯ 15.16 88.60 Ϯ 13.68

a HD, hemodialysis; IHD, ischemic heart disease; BB, therapy; HYP, hypertension; PTH, parathyroid hormone level. b P ϭ 0.011, IHD versus no IHD (pre-HD). J Am Soc Nephrol 10: 1297–1302, 1999 QT Dispersion in Patients with ESRD 1301

Table 4. The results of rate corrected QT, QTcmax interval, and QTc interval dispersion in subgroups of hemodialyzed patientsa

QTcmax Interval (ms) QTc Dispersion (ms) Subgroup No. of Patients Pre-HD Post-HD Pre-HD Post-HD

IHD 9 488.1 Ϯ 43.3 522.0 Ϯ 35.1 66.4 Ϯ 19.7b 100.6 Ϯ 16.4 No IHD 25 476.6 Ϯ 43.6 510.8 Ϯ 31.2 51.5 Ϯ 10.4 81.1 Ϯ 9.0 BB 8 487.1 Ϯ 46.0 517.4 Ϯ 35.6 62.4 Ϯ 17.6 94.4 Ϯ 17.9 No BB 26 478.3 Ϯ 33.6 523.7 Ϯ 29.9 62.0 Ϯ 23.4 98.0 Ϯ 15.2 HYP 29 484.3 Ϯ 76.2 531.3 Ϯ 28.3 66.5 Ϯ 10.8 90.3 Ϯ 18.1 No HYP 5 485.1 Ϯ 43.3 517.0 Ϯ 34.8 61.6 Ϯ 19.8 96.0 Ϯ 17.1 PTH Ͻ40 pmol/L 25 484.3 Ϯ 42.9 513.1 Ϯ 37.8 61.8 Ϯ 19.0 93.3 Ϯ 16.0 PTH Ͼ40 pmol/L 9 484.8 Ϯ 45.5 532.0 Ϯ 18.9 63.3 Ϯ 19.1 99.6 Ϯ 19.4 Male 21 484.8 Ϯ 43.9 518.7 Ϯ 33.2 63.0 Ϯ 19.6 92.4 Ϯ 14.8 Female 13 485.2 Ϯ 43.3 519.4 Ϯ 37.0 60.9 Ϯ 17.8 100.7 Ϯ 20.4

a Abbreviations as in Table 3. b P ϭ 0.007, IHD versus no IHD (pre-HD). ms, the patients are at risk for serious ventricular arrhythmias Acknowledgment or sudden death (18,34). In our patients, the average value QT We thank Dr. Ben Thomas for helpful assistance in preparing the dispersion was 85 ms and QTc was 95 ms at the end of HD. manuscript. The post-HD QT and QTc interval prolongation and QT and QTc interval dispersion lengthening were independent of gen- References der, patient age, hypertension, duration of chronic hemodialy- 1. LorinczЉ I, Zilahi Zs, Kun Cs, Ma´tyus J, Kakuk Gy: ECG abnor- sis program and hyperparathyroidism, and concomitant cardiac malities in hemodialysis [Letter]. Am Heart J 134: 1138–1140, disease (Tables 3 and 4). According to continuous ECG and BP 1997 monitoring, there were no major arrhythmias including non- 2. U.S. Renal Data System: USRDS 1990 Annual Data Report, U.S. sustained and sustained ventricular tachycardias, or ventricular Department of Health and Human Services, National Institutes of , and serious hypotensive episodes (Ͻ90 mmHg for Health, National Institutes of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, August 1990 systolic BP or decrease in systolic BP Ͼ30 mmHg) during HD. 3. Leier CV, Boudolulas H: Renal disorders and heart disease. In: Our results indicate that the nonhomogeneity of regional ven- Heart Disease: A Textbook of Cardiovascular Medicine, 5th Ed., tricular repolarization increases during hemodialysis, which is edited by Braunwald E, Philadelphia, Saunders, 1997, pp 1914– suggested by increased QTmax and QTcmax interval and QT and 1938 QTc interval dispersion. The results of this study may add a 4. Chazan JA: Sudden death in patients with chronic renal failure new dimension to recent reports indicating the usefulness of on hemodialysis. Dial Transplant 16: 447–448, 1987 QT dispersion as a predictor of sudden death after myocardial 5. Morrison G, Michelson EL, Brown S, Morganroth J: Mechanism and prevention of cardiac arrhythmias in chronic hemodialysis infarction, in heart failure of ischemic etiology, hypertrophic patients. Kidney Int 17: 811–819, 1980 cardiomyopathy, as well as the risk of arrhythmia in the long 6. Abe S, Yoshizawa M, Nakanishi N, Yazawa T, Yokota K, Honda QT syndrome (8,9,13,14,16–18). M, Sloman G: Electrocardiographic abnormalities in patients It is concluded that the nonhomogeneity of regional ventric- receiving hemodialysis. Am Heart J 131: 1137–1144, 1996 ular repolarization in patients with chronic end-stage renal 7. Gruppo Hemodialisie Patologie Cardiovascular: Multicentre failure receiving hemodialysis may be suggested by the in- cross-sectional study of ventricular arrhythmias in chronically crease in QT and QTc interval or increase in QT and QTc haemodialysed patients. Lancet 2: 305–309, 1988 dispersion. The prolongation of these parameters may be a 8. Tomaselli GF, Beuckelmann DJ, Calkins HG, Berger RD, Kessler PD, Lawrence JH, Kass D, Feldman AM, Marban E: further noninvasive marker of susceptibility to ventricular ar- Sudden cardiac death in heart failure: The role of abnormal rhythmias. Additional studies are needed to clarify whether repolarization. Circulation 90: 2534–2539, 1994 increased postdialysis QT dispersion results in an increased 9. Pye MP, Cobbe SM: Mechanism of ventricular arrhythmias in occurrence of arrhythmias. QT and QTc dispersion is an easily cardiac failure and hypertrophy. Cardiovasc Res 26: 740–750, obtainable, noninvasive, simple, inexpensive, and widely 1992 available method of risk stratification in uremic patients re- 10. Zabel M, Portonoy S, Franz MR: Electrocardiographic indexes ceiving chronic hemodialysis. of dispersion of ventricular repolarization: An isolated heart validation study. J Am Coll Cardiol 25: 746–752, 1995 1302 Journal of the American Society of Nephrology J Am Soc Nephrol 10: 1297–1302, 1999

11. Day CP, McComb JM, Cambell RWF: QT dispersion: An indi- mortality in an apparently healthy population. Circulation 84: cation of arrhythmia risk in patients with long QT intervals. Br 1516–1523, 1991 Heart J 63: 342–344, 1990 23. Bland JM, Altman DG: Statistical methods for assessing agree- 12. Schwartz PJ, Wolf S: QT interval prolongation as a predictor of ment between two methods of clinical measurement. Lancet 1: sudden death in patients with myocardial infarction. Circulation 307–310, 1986 57: 1074–1079, 1978 24. Weber KT, Pick R, Silver MA, Moe GW, Janicki JS, Zucker IH, 13. Barr CS, Naas A, Freeman M, Lang CC, Struthers AD: QT Armstrong PW: Fibrillar collagen and remodeling of dilated dispersion and sudden unexpected death in chronic heart failure. canine left ventricle. Circulation 82: 1387–1401, 1990 Lancet 343: 327–329, 1994 25. Weber KT, Brilla CG, Janieki JS: Myocardial fibrosis: Func- 14. Davey PP, Bateman J, Mulligan IP, Forfar C, Barlow C, Hart G: tional significance and regulatory factors. Cardiovasc Res 27: QT interval dispersion in chronic heart failure and left ventricular 341–348, 1993 hypertrophy: Relation to autonomic nervous system and Holter 26. Amann K, Ritz E, Weist G, Klaus G, Mall G: A role of para- tape normalities. Br Heart J 71: 268–273, 1994 thyroid hormone for the activation of cardiac fibroblast in ure- 15. Higham PD, Campbell RWF: QT dispersion. Br Heart J 71: mia. J Am Soc Nephrol 4: 1814–1819, 1994 508–510, 1994 27. Kirvela M, Yli-Hankala A, Lindgren L: QT dispersion and 16. Linker NJ, Colonna P, Kekwick CA, Till J, Camm AJ, Ward DE: autonomic function in diabetic and nondiabetic patients with Assessment of QT dispersion in symptomatic patients with con- renal failure. Br J Anaesth 73: 801–804, 1994 genital long QT syndromes. Am J Cardiol 69: 634–638, 1992 28. Rombola G, Colussi G, Elisabetta DeFerrari M, Frontini A, 17. Buja G, Miorelli M, Turrini P, Melacini P, Nava A: Comparison Minetti L: Cardiac arrhythmias and electrolyte changes during of QT dispersion in hypertrophic cardiomyopathy between pa- hemodialysis. Nephrol Dial Transplant 7: 318–322, 1992 tients with and without ventricular arrhythmias and sudden death. 29. Fantuzzi S, Caico G, Amatruda O, Cervini P, Abu-Turky H, Am J Cardiol 72: 973–976, 1993 Baratelli L, Donbnaty D, Gastaldi L: Hemodialysis-associated 18. van de Loo A, Arendts W, Hohnloser SH: Variability of QT cardiac arrhythmias: A lower risk with bicarbonate? Nephron 58: dispersion measurements in the surface electrocardiogram in 196–200, 1991 patients with acute myocardial infarction and in normal subjects. 30. Ramirez G, Bruggemeyer CD, Newton JL: Cardiac arrhythmias Am J Cardiol 74: 1113–1118, 1994 on hemodialysis in chronic renal failure patients. Nephron 36: 19. Priori SG, Napolitano C, Diehl L, Schwartz PJ: Dispersion of the 212–218, 1984 QT interval: A marker of therapeutic efficacy in the idiopathic 31. Harnet JD, Parfrey PS: Left ventricular dysfunction in dialysis long QT syndrome. Circulation 89: 1681–1689, 1994 subjects. In: Principles and Practice of Dialysis, edited by 20. Tallgren M, Ho¨ckerstedt K, Isoniemi H, Orko R, Lindgren L: Heinrich WL, Baltimore, Williams & Wilkins, 1994, pp 170– Intraoperative death in cardiac with increased QT 180 dispersion in the electrocardiogram. Anesth Analg 80: 1233– 32. Gafter U, Battler A, Eldar M, Zevin D, Neufeld HN, Levi J: 1235, 1995 Effect of hyperparathyroidism on cardiac function in patients 21. Pinsky D, Sciacca RR, Steinberg J: QT dispersion as a marker of with end-stage renal disease. Nephron 41: 30–33, 1985 risk in patients awaiting heart transplantation. J Am Coll Cardiol 33. Suzuki Y, Narita M, Yamaziki N: Effect of L-carnitine on ar- 29: 1576–1584, 1997 rhythmias during hemodialysis. Jpn Heart J 23: 349–359, 1982 22. Schouten EG, Dekker JM, Meppelink P, Kok FJ, Vandenbroucke 34. Antzelevitch C, Shimizu W, Yan GX, Sicouri S: Cellular basis JP, Pool J: QT interval prolongation predicts cardiovascular for QT dispersion. J Electrocardiol 30: 168–175, 1998

This article can be accessed in its entirety on the Internet at http://www.wwilkins.com/JASN along with related UpToDate topics.