P-wave and QRS complex measurements in patients undergoing hemodialysis
Abbdenasser Drighil, MD,
aJohn E. Madias, MD,
c,d,* Asma Yazidi, MD,
bMohammed Bennani, MD,
aAhmed Bennis, MD,
aBeenyouness Ramdan, MD,
bAzzeddine Tahiri, MD
aaDepartment of Cardiology, Ibn Rochd University Hospital, Casablanca, Morocco
bDepartment of Nephrology, Ibn Rochd University Hospital, Casablanca, Morocco
cMount Sinai School of Medicine, of the New York University, New York, NY, USA
dDivision of Cardiology, Elmhurst Hospital Center, Elmhurst, NY, USA
Abstract Hemodialysis (HD) has been associated with an increase in the amplitude of QRS complexes.
Experience in a single patient with multiple measurements has shown that HD leads also to augmentation of P-wave amplitude. The objective of this investigation was to evaluate electrocardiogram (ECG) changes with HD in a cohort of patients undergoing this procedure with particular emphasis on the P-wave and QRS complex changes. The sum of amplitudes of P waves (AP) and QRS complexes (AQRS) in millimeters in the 12 leads of the ECG, along with a host of other ECG parameters, body weight, blood pressure, heart rate, electrolytes, and hemoglobin/
hematocrit were measured before and immediately after HD in 47 patients. Hemodialysis resulted in a loss of a mean of 3 kg of weight and an increase in theAP,AQRS, mean QRS duration, maximum P-wave duration, and P-wave duration measured in lead II, whereas the changes in mean P-wave and corrected QT interval durations were not statistically significant. Percentage change (D%) inAP and AQRS correlated poorly withD% in electrolytes, hematocrit, blood pressure, heart rate, and weight.
Values forAP andAQRS vs weight werer = 0.105,P = .48 andr = 0.09,P = .51, respectively.
The D% in AP correlated well with D% inAQRS (r = 0.42, P = .003). Alterations in P-wave amplitudes and duration commensurate with the ones affecting the corresponding QRS complexes occur in patients undergoing HD and indicate that evaluation of measurements in serial ECGs should take this into account. The mechanisms of these phenomena continue to be elusive, and whether they represent cardiac and/or extracardiac influences has not as yet been unraveled.
D2008 Elsevier Inc. All rights reserved.
Keywords: Electrocardiology; Electrophysiology; P waves; Low-voltage ECG; P-wave duration; P-wave dispersion;
Peripheral edema; Hemodialysis
Introduction
Augmentation of QRS complexes in patients undergoing hemodialysis (HD) has been previously reported and attributed to many causes, including myocardial ischemia, heart volume changes, and changes in the extracardiac conductivity1-8; the pathophysiologic mechanisms invoked were mostly poorly supported and often at odds with the
study results.1,6,8 More recently, the increase in the QRS voltage after HD has been attributed to an increase in the transfer impedance of the body volume conductor, resulting from the HD-induced fluid removal.5,7,8 Reduction of the fluid content of the volume conductor results in an overall decrease in conductivity of the transfer medium because the removed fluid has the lowest resistivity (special resistance) of all body constituents.9Thus, alleviation of fluid overload in general or HD leads to an increase in the transfer impedance with resultant increase in the ECG voltage, as per Ohm’s law. This mechanism, being extracardiac (not electrophysiologically based) as per this conceptualization, is expected to impact not only the QRS complexes, but also the entire ECG curve. Accordingly, the amplitude of
0022-0736/$ – see front matterD2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jelectrocard.2006.03.002
4 Corresponding author. Division of Cardiology, Elmhurst Hospital Center, Elmhurst, NY 11373, USA. Tel.: +1 718 334 5005; fax: +1 718 334 5990.
E-mail address:madiasj@nychhc.org
Journal of Electrocardiology 41 (2008) 60.e1 – 60.e7 Available online at www.sciencedirect.com
www.jecgonline.com
P waves is also expected to increase after HD, and indeed this was shown recently in a single patient studied during 26 HD sessions.10However, even a clear change in a vari- able, shown reproducibly in repeated assessments of a single subject, is not confirmatory of an association. Thus, the present study of a cohort of patients undergoing one of their routine HD sessions was designed with the objective to evaluate the response of P waves, R waves, and a host of other ECGs to this procedure.
Materials and methods Patients
Forty-seven subjects with end-stage renal failure attend- ing a routine midweek HD session were recruited in this study after informed consent was obtained. All were receiving bicarbonate-based HD sessions twice weekly lasting between 3 and 4 1/2 hours, using polysulfone capillaries and bicarbonate dialysate containing 138 Na+,
2.0 K+, 1.75 Ca2+, and 0.5 Mg2+mmol/L. All HD sessions were uncomplicated.
Study variables
Information pertaining to history, demographic data, and routine drug regimen were considered as study variables.
Before and after HD, the subjects were weighed, had their blood pressure and heart rate obtained, had a 12-lead ECG recorded, and a blood specimen was drawn for measurement of plasma electrolytes and hemoglobin and hematocrit.
Electrocardiography
Electrocardiograms were recorded using an Agilent (HP, currently Philips, Andover, Mass) M1771A Page writer 200i electrocardiograph at a paper speed of 25 mm/s and at standardization of 10 mm = 1.0 mV. Filter configurations were the following: for automatic report, filter/baseline 0.15 Hz and noise filter 40 Hz; for manual report, filter/
baseline 0.05 Hz and noise filter 40 Hz. We obtained leads V1 through V6 from fixed chest landmarks made, using a skin marker, to ensure reproducibility of the ECGs before and after HD.
Electrocardiogram measurements
The amplitude of P waves in millimeters was measured from peak to nadir using a magnifying glass; P waves less than 1.0 mm were set at the fixed levels of 0.25, 0.5, and 0.75 mm. The 0.25-mm amplitude designation was consid- ered when a P wave consisted of a tiny perturbation in the isoelectric line before the QRS complexes; the 0.5-mm measurement was considered for the amplitude of P wave estimated as such; the 0.75-mm measurement was considered when the amplitude of P wave was more than 0.5 mm but less than 1.0 mm. The sum of the P waves from all 12 ECG leads (AP) was then calculated, reflecting the changes in the amplitude of P waves in all 12 ECG leads. The QRS complexes in all 12 ECG leads were measured from peak to nadir in millimeters to the nearest 0.5 mm using a magnifying glass and calipers, and the sum (AQRS) was calculated.11All measurements were done by one of us (AD). Percentage
Table 1
Baseline characteristics of 47 subjects who underwent HD
Characteristics Frequency
Sex (male/female) 18/29
Age (y) 42.4F13 (21-70)
Duration of chronic HD (mo) 101F41 (36-216)
Coronary heart disease 1 (2%)
Hypertension 21 (42%)
Diabetes 5 (10%)
Dyslipidemia 2 (4%)
Drugs
Angiotensin-converting enzyme inhibitors 17 (34%)
Calcium blockers 26 (53%)
b-Blockers 3 (6%)
Cause of end-stage renal failure 7 (15%)
Glomerulonephritis 5 (10%)
Diabetes
Nephrosclerosis 7 (15%)
Tubulointerstitial nephritis 4 (8.5%)
Hereditary 1 (2%)
Autosomal dominant polycystic 4 (8.5%)
Vasculitis 7 (15%)
Unknown 12 (26%)
Table 2
Study variables before and after HD
Variable Pre-HD Post-HD Change % Change P
Potassium (mmol/L) 5.7F0.8 3.7F0.7 .0001
Calcium (mg/L) 83F8 114F10 .01
Phosphate (mg/L) 58.3F9.5 54.7F9.4 .0001
Bicarbonate (mmol/L) 16.3F1.8 21.9F1.7 .0001
Hemoglobin (g/dL) 9.0F1.36 9.8F1.3 .0001
Hematocrit (%) 30.6F3.4 36.5F4 .0001
Systolic BP (mm Hg) 136F23.8 124F22.5 .0001
Diastolic BP (mm Hg) 77F13 71F11 .0001
Heart rate (beats per minute) 74F12 78F12 .004
Body weight (kg) 56F9 53F9 3F1.1 6.4F2.4 .0001
AP (mm) 9.4F2.7 12.9F4.1 3.4F3 40F34 .0001
AQRS (mm) 139.1F50.7 177.2F64.8 38F32 28F22 .0001
Sokolow (mm) 27F12 34F15 .0001
Cornell (mm) 18F8 20F9 .002
P duration (ms) 66.1F13.4 69F10.5 .09
QRS duration (ms) 77.9F11.5 80.5F12.2 .003
QTc 406F30 405F32 .8
change (D%) in theAPs andAQRSs from the pre-HD values was also used as variables. Sokolow-Lyon voltage (sum of the amplitudes of S wave in V1and R wave in V5, or V6
N3.5 mV)12 and sex-specific Cornell voltage (sum of the amplitudes of S wave in V3and R wave in aVLN2.0 mV in women andN2.8 mV in men)13were used as ECG criteria for left ventricular hypertrophy (LVH). Because the development and subsequent alleviation of edematous states has been found to impact the ECG diagnosis of LVH, ECG evidence for LVH was used as a variable herein.14A tall P wave (height N2.5 mm) in leads II, III, and aVF was defined as P pulmonale, and the P duration more than 120 milliseconds in lead II or a negative component more than 1.0 mm deep and more than 40 milliseconds in duration was defined as P mitrale. These ECG parameters were used because development and subsequent alleviation of edematous states have been found to impact such diagnoses.15Because of the
low amplitude of P waves particularly before HD, repeat measurements of AP in 10 randomly selected ECGs from before and after HD were remeasured. The intraobserver variation ofAP was 5.2%F1.3% before HD and 2%F1.5%
after HD. The P-wave duration was measured in all leads from the first electrical activity to the offset at the junction between the end of P-wave deflection and the isoelectric line;
this variable was used here because it has been previously used in other HD studies, and prolongation of the P-wave duration has been considered a predictor of future atrial fibrillation.15 The QRS duration was measured in all leads because this ECG parameter has been altered when peripheral edema develops, or with amelioration of fluid overload.16,17 The mean corrected QT interval (QTc), which was calculated as the arithmetic mean of all corrected QT intervals measurable on a single ECG from all leads, and the heart rate in beats per minute was used as a variable because the QTc is impacted by perturbations of edematous states and HD.18,19The P-wave dispersion was defined as the difference between the maximum and minimum value of P-wave duration and was used herein because it has been considered in previous HD studies and is thought to represent a predictor of atrial fibrillation.20-23The QTc dispersion was calculated as the maximum QTc minus the minimum QTc and was used as a variable herein because of its role as an index of arrhythmogenicity in the context of HD and the increase noted with this procedure.24
Statistical analysis
Continuous variables are reported as mean F SD and were analyzed with the Student t test for paired data;
Table 3
P-wave duration before and after HD
P-wave duration before HD P-wave duration after HD P
V1 57.6F18.8 64.8F18.6 .023
V2 55.4F18.6 58.6F17.6 .27
V3 60.0F19.0 60.9F17.9 .6
V4 61.5F18.1 67.8F17.8 .008
V5 65.6F17.0 67.9F15.2 .16
V6 65.4F16.7 67.8F14.7 .13
I 67.1F16.4 67.1F18.1 1
II 71.3F15 73.9F12.5 .013
III 69.8F14.7 67.9F15.5 .3
aVR 67.0F1.2 68.8F15.2 .4
aVL 62.9F18.8 65.1F18.7 .5
aVF 70.9F15.2 71.6F12.0 .73
Fig. 1. Weight,AP, andAQRS measured before and after HD. C1through C6indicate V1through V6leads.
Pearson correlation for linear regression analysis ofD% of the study variables resulting from HD was used. All statistical operations were 2-tailed, and a P value of less than .05 was considered as statistically significant. The statistical package, SPSS version 11.5 (SPSS, Inc, Chicago, Ill), was used.
Results
The characteristics of the subjects are shown inTable 1.
The results of the changes in measured variables are shown in Table 2. After HD, significant changes in electrolytes, blood pressure, and heart rate were observed in association with a fall of the patients’ weight by a mean of 3 kg. Significant ECG changes (Tables 2 and 3) (Figs. 1 and 2) precipitated by HD included an increase in theAP, AQRS, mean QRS duration, maximum P-wave duration, and P-wave duration measured in lead II; lead II was the lead with the longest P-wave duration in 36 patients (76.5%). However, the changes in mean P-wave and QTc durations (Table 4) were not statistically significant. The average of P-wave dispersion was 21.3 F 17.8 milli- seconds before HD and 21.7F19.0 milliseconds after HD (P = .83). The QTc dispersion increased from 34F17 to 48 F 21 milliseconds, and this increase was statistically significant (P = .0001).
P mitrale was diagnosed in 3 patients before and 8 patients after HD. There was no P pulmonale diagnosed in patients before HD and only appeared in 1 after HD. In addition, both the Sokolow (Fig. 3) and Cornell (Fig. 4) voltages rose significantly post-HD. Left ventricular hyper- trophy was diagnosed in 10 patients before and 18 patients after HD based on Sokolow criteria and 7 patients before and 13 after HD based on Cornell criteria.
Percentage change inAP correlated poorly with D% in Ca2+ (r = 0.115, P = .43), K+ (r = 0.024, P = .87), phosphate (r = 0.079,P = .59), bicarbonate (r = 0.098,
Fig. 2. Enlarged leads II and V1from patient no. 29 before and after HD.
Table 4
Corrected QT interval before and after HD
QTc before HD QTc after HD P
V1 387.54F23.9 380.58F26.66 .28
V2 401.20F18.83 385.82F23.92 .024
V3 401.20F18.83 389.82F22.42 .077
V4 401.20F18.83 389.82F22.42 .077
V5 401.20F18.83 389.82F22.42 .077
V6 401.20F18.83 394.41F21.93 .305
I 403.27F25.42 389.00F33.71 .229
II 403.27F25.42 386.45F32.52 .204
III 406.76F24.75 387.85F34.47 .207
aVR 403.27F25.42 386.45F32.52 .204
aVL 399.77F15.03 371.57F30.39 .06
aVF 403.27F25.42 386.45F32.52 .204
P = .5), hemoglobin (r = 0.128,P = .39), hematocrit (r = 0.14, P = .39), systolic blood pressure (r = 0.22, P = .12), diastolic blood pressure (r = 0.117, P = .43), and weight (r = 0.105,P = .48); however, it correlated well withD% in heart rate (r = 0.532,P = .0001).
Also,D% inAQRS correlated poorly withD% in K+(r= 0.04, P = .78), phosphate (r = 0.093, P = .9), bicarbonate (r = 0.1,P = .5), hemoglobin (r = 0.117,P = .43), hematocrit (r= 0.018,P= .9), systolic blood pressure (r = 0.21,P = .15), diastolic blood pressure (r = 0.14,
Fig. 4. Weight and Cornell index measured before and after HD.
Fig. 3. Weight and Sokolow LVH voltage index measured before and after HD.
P = .32), and weight (r = 0.09, P = .51); however, it correlated withD% in Ca2+(r = 0.36,P = .013) andD% in heart rate (r= 0.294,P= .045). Finally,D% inAP correlated well withD% inAQRS (r = 0.42,P = .003).
Discussion
The new information in this communication is that HD leads to the enhancement of the amplitude of the P waves;
this had been shown previously in a single patient after multiple HD sessions.10 In addition, an increase in the amplitude of the QRS complexes was noted (Table 2), something that has been already shown previously.1-8There was a correlation of amplitude D% between P waves and QRS complexes, suggesting that the mechanism affecting these 2 components of the ECG curve may be common. Such a correlation was not found in the data from the multiple measurements of the single patient.8,10TheAP increased by a mean of 40% (close to the mean of 34%, noted in the single patient), whereas the AQRS increased by a mean of 28%
after HD (Table 2). That the correlation was not stronger could be because of amplitude measurements in P waves, being less accurate because they are much smaller in voltage than the QRS complexes and, on occasion, barely visible, particularly in the pre-HD stage (Figs. 1-4), although other mechanisms (vide infra) could also be implicated. As in the previously studied single patient,8no correlation was found between percentage of increase of AP and percentage of decrease in weight. Also as in the single patient, HD precipitated a reduction in blood pressure and a faster heart rate (Table 2). Similarly, a mean of 126% increase inAP in patients with peripheral edema of varying etiologies who subsequently lost weight did not correlate with percentage of decrease in weight or increase inAQRS.10In contrast, a 64% decrease in AP noted in the same patients after development of peripheral edema correlated well with percentage of increase in weight (r = 0.70, P = .0028) and percentage of decrease inAQRS (r = 0.62,P = .01).10 Augmentation of AP and AQRS may be only due to alterations on the electrical properties of the passive volume conductor resulting from HD, although one cannot discard the possibility that other parameters (eg, cardiac volume changes) are also contributory. The matter of mechanism of these HD-triggered ECG changes needs further scrutiny. The observed changes in the amplitude of the QRS complexes and P waves have clinical implications. For example, the observed increased rate of the ECG-based diagnosis of LVH after HD (Figs. 3 and 4) is the expected opposite noted previously in patients with marked peripheral edema of varying etiologies,14 where patients with ECG LVH in the admission ECG lost this feature in ECGs obtained after the development of peripheral edema. In addition, the increase rate of diagnosis of P mitrale with HD is the expected opposite of what was noted in patients with marked peripheral edema of varying etiologies,15 where such a diagnosis present in the admission ECG was not seen in ECGs recorded after the development of peripheral edema.
The increase in mean QRS duration with HD is in keeping with the opposite effect on the QRS duration observed with
marked peripheral edema of varying etiologies.16Hemodial- ysis led to some mild but statistically significant prolongation of the QRS duration in this study (Table 2), similar to a larger one noted with marked weight loss in patients who initially had large fluid accumulations17and opposite (as expected) to the QRS duration shortening resulting from fluid overload.16 Different QRS duration values before and after HD may impact differentiation of incomplete from complete intraven- tricular conduction delays25or decisions about cardioverter defibrillation implantation or implementation of cardiac resynchronization therapy.26
No changes in QTc (Table 2) or QTc dispersion were noted with HD, unlike the decrease (with fluid overload) and subsequent increase with weight loss found in patients with peripheral edema.18 It is possible that the modest change in weight, in contrast to the changes in patients with peripheral edema, was the reason for the lack of effect of HD on the QTc.
The increase in the maximum P-wave duration and P-wave duration measured in lead II, although not associ- ated with changes in the mean P-wave duration, is in keeping with observations made previously in the single patient who underwent repeated HD sessions20 and more systematic measurements of the P-wave duration made by other investigators.21-23The postulated mechanism produc- ing these changes is the noted alteration in electrolytes with HD,21,22although this was not corroborated by our findings (vide infra).
The lack of correlations ofD% between P waves or QRS complexes and the D% in the electrolytes and hematocrit consequent to HD are in agreement with work by others.6 The findings of the previously cited studies, which noted increase in P-wave duration with HD,10,22are at odds with the decrease in the P-wave duration observed when volume overload was alleviated by diuresis in patients treated for congestive heart failure.23 Another physiologic directional inconsistency is the hemoglobin/hematocrit increase that is always detected with HD and the associated ECG changes (Table 2); thus, hemoconcentration should lead to a decrease in surface ECG potentials, as documented in animal and clinical studies,1,27,28whereas an increase has been noted in this and in previous studies.1-9The previously cited studies suggest that the expected findings from HD changes in electrolytes and hemoglobin/hematocrit do not influence the amplitude of ECG potentials, being rather merely laboratory accompaniments or occasionally even correlates not etio- logically linked.20
The mechanism of the HD-based increases in the P-wave amplitude most probably is similar to the one invoked previously to explain similar directional changes in the amplitude of QRS complexes.5,7,8,11Because the change in theAQRS has been previously ascribed to alteration of the body electrical conduction properties due to fluid reten- tion,1,11measurements of the body electrical resistance (R) and reactance (Ro) were carried out immediately before and after HD, as previously described8; impedance (Z) was calculated using the formulaZ2 = R2+ Ro2.29Percentage of increase in AP correlated with percentage of increase in body electricalR (r = 0.44,P = .029), Ro (r = 0.46,
P = .02), and Z (r = 0.44, P = .029); this correlation suggests a role for the electrical properties of body volume conductor in mediating the increase in the amplitude of P waves. In reference to the mechanism of the increase in mean QRS and/or P duration, along with the changes in the amplitude of P waves and QRS complexes with HD, one of possible methodological interpretations of the findings is that the onset and offset of P and QRS is easier to detect when their amplitude increases, as stated previously.16 Although the previously cited ECG changes after HD are ascribed primarily to an extracardiac mechanism, it is possible that a cardiac alteration may be responsible for these changes, either singly or in combination with the influence of the body volume conductor.
Most probably, there are differences between diuresis in congestive heart failure, alleviation of peripheral edema, and HD, and these physiologic differences include the amount of fluid removed, time intervals during which such relief of volume overload takes place, location of the fluid overload (intravascular, interstitial, and intracellular), its distribution during alleviation, and other factors not yet considered.
These different factors need to be evaluated in concert in future investigative ventures on this topic.
The findings of this study on the behavior of the amplitude and duration of P waves and the duration of QRS complexes should be considered in concert with the previously described changes in the amplitude of QRS complexes during HD and should be kept in mind when serial ECGs are being evaluated. Because these ECG changes resulting from HD in comparison with the pre- HD status may affect various ECG diagnostic considerations and decision making (vide supra), perhaps the post-HD values of ECG parameters should be considered represen- tative for a patient with end-stage chronic renal failure.
References
1. Ishikawa K, Nagasawa T, Shimadu H. Influence of hemodialysis on electrocardiographic wave forms. Am Heart J 1979;97:5.
2. Diskin CJ, Salzsieder KH, Solomon RJ, et al. Electrocardiographic changes following dialysis. Nephron 1981;27:94.
3. Ono K, Hisasue Y, Kikawa K. Effects of dialysate sodium concentra- tion on the appearance of ischemic EKG pattern. Nephron 1983;
33:182.
4. Vitolo E, Madoi S, Palvarini M, et al. Relationship between changes in R wave voltage and cardiac volumes. A vectorcardiographic study during hemodialysis. J Electrocardiol 1987;20:138.
5. Kinoshita O, Kimura G, Kamakura S, et al. Effects of hemodialysis on body surface maps in patients with chronic renal failure. Nephron 1993;
64:580.
6. Fuenmayor AJ, Vasquez CJ, Fuenmayor AM, et al. Hemodialysis changes the QRS amplitude in the electrocardiogram. Int J Cardiol 1993;41:141.
7. Ojanen S, Koobi T, Korhonen P, et al. QRS amplitude and volume changes during hemodialysis. Am J Nephrol 1999;19:423.
8. Madias JE, Narayan V. Augmentation of the amplitude of electrocar- diographic QRS complexes immediately following hemodialysis: a study of 26 hemodialysis sessions of a single patient, aided by
measurements of resistance/reactance/impedance. J Electrocardiol 2003;36:263.
9. Geddes LA, Baker LE. The specific resistance of biological materials—a compendium of data for the biomedical engineer and physiologist. Med Biol Eng 1967;5:271.
10. Madias JE. P waves in patients with changing edematous states:
implications on interpreting repeat P wave measurements in patients developing anasarca or undergoing hemodialysis. Pacing Clin Electro- physiol 2004;27:749.
11. Madias JE, Bazaz R, Agarwal H, et al. Anasarca-mediated attenuation of the amplitude of electrocardiogram complexes: a description of a heretofore unrecognized phenomenon. J Am Coll Cardiol 2001;
38:756.
12. Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and limb leads.
Am Heart J 1949;37:161.
13. Casale PN, Devereux RB, Kligfield P, et al. Electrocardio- graphic detection of left ventricular hypertrophy: development and prospective validation of improved criteria. J Am Coll Cardiol 1985;
6:572.
14. Madias JE. Concealment of electrocardiographically based diagnosis of left ventricular hypertrophy by anasarca. Am J Hypertens 2004;
17:897.
15. Madias JE. Peripheral edema masks the diagnoses of P pulmonale, P mitrale, and biatrial abnormality: clinical implications for patients with heart failure. Congest Heart Fail 2006;12:20.
16. Madias JE. Significance of shortening of the mean QRS duration of the standard electrocardiogram in patients developing peripheral edema.
Am J Cardiol 2002;89:1444.
17. Madias JE. Unpublished data.
18. Madias JE. QTc interval in patients with changing edematous states:
implications on interpreting repeat QTc interval measurements in patients with anasarca of varying etiology, and those undergoing hemodialysis. Pacing Clin Electrophysiol 2005;28:54.
19. Naˆppi SE, Virtanen VK, Saha HH, Mustonen JT, Pasternack AI. QTc dispersion increases during hemodialysis with low-calcium dialysate.
Kidney Intern 2000;57:2117.
20. Madias JE. Increases in P-wave duration and dispersion after hemodi- alysis are totally (or partially) due to the procedure-induced alleviation of the body fluid overload: a hypothesis with strong experimental support.
Ann Noninvasive Electrocardiol 2005;10:129.
21. Szabo´ Z, Kakuk G, Fqlfp T, et al. Effects of haemodialysis on maximum P wave duration and P wave dispersion. Nephrol Dial Transplant 2002;17:1634.
22. Tezcan UK, Amasyali B, Can I, et al. Increased P wave dispersion and maximum P wave duration after hemodialysis. Ann Noninvasive Electrocardiol 2004;9:34.
23. Song J, Kalus JS, Caron MF, Kluger J, White CM. Effects of diuresis on P-wave duration and dispersion. Pharmacotherapy 2002;22:564.
24. Cupusti A, Galetta F, Morelli E, et al. Effect of hemodialysis on the dispersion of the QTc interval. Nephron 1998;78:425.
25. Madias JE. Apparent amelioration of bundle branch blocks and intraventricular conduction delays mediated by anasarca. J Electro- cardiol 2005;38:160.
26. Madias JE, Macfarlane PW. Artificial attenuation of ECG voltage produces shortening of the corresponding QRS duration: clinical implications for patients with edema. Pacing Clin Electrophysiol 2005;28:1060.
27. Nelson CV, Rand PW, Angelakos ET, Hugenholtz PG. Effect of intracardiac blood on the spatial vectorcardiogram. Results in the dog.
Circ Res 1972;31:95.
28. Rosenthal A, Restieaux NJ, Feig SA. Influence of acute variations in hematocrit on the QRS complex of the Frank electrocardiogram.
Circulation 1971;44:456.
29. Talluri A, Mapia G. Bioimpedance analysis (BIA) in hemodialysis:
technical aspects. Int J Artif Organs 1995;18:687.
