www.lejacq.com
DIFFICULT CASES
IN
ID:4484
HEART FAILURE
Mandeep R. Mehra, MD; Hector O. Ventura, MD Editors
Standard Electrocardiographic and Signal-Averaged Electrocardiographic Changes in Congestive Heart Failure A 54-year-old man with congestive heart failure and peripheral edema was monitored during treatment with serial electrocardiograms (ECGs) and two signal-averaged ECGs. Weights and a host of ECG parameters were monitored. ECGs recorded in the 5 years before his index admission were also considered in the quantitative ECG analysis. Amelioration of his peripheral edema was associated with reduction of his weight and increase in the QRS and P-wave amplitudes, and duration in QRS complexes and QT intervals in the ECGs, but unchanged signal-averaged ECGs. This case report provides insight into the mechanism of the change in QRS duration in ECGs during changing edematous states resulting from clinical deterioration or improved compensation in patients with congestive heart failure. (CHF. 2005;11:266–271) ©2005 CHF, Inc.
P
erusal of serial electrocardiograms (ECGs) in patients admitted with congestive heart failure (CHF) reveals marked augmentation in the amplitude of QRS complexes as the patients are being diuresed, and the admission ECGs of such patients show marked attenuation in the QRS complexes in comparison with older tracings taken when the patients did not have CHF, or were well compensated.1–3 The changes seen in patients with CHF are similar to those observed in patients undergoing hemodialysis,4–6 or in patients with peripheral edema (PED) of varying etiologies experiencing amelioration of their edematous state.7 Since changes imparted on the ECG are not unique to the QRS complex, the P waves show similar attenuation/augmentation as patients with PED gain/lose weight.8 Interestingly, changes in the QRS and QT durations have also been reported in association with alterations in the amplitude of QRS complexes in patients with PED.9,10
Signal-averaged ECGs (SAECGs) are occasionally recorded in patients with CHF or undergoing hemodialysis, since such patients are prone to malignant arrhythmias and sudden cardiac death,11–13 and changes in repeat SAECG are thought to be of some prognostic value.14–18 Apart from the possible electrophysiologically-mediated (i.e., real) changes in repeat SAECG, one wonders whether the loss of PED with diuresis in patients with CHF would enhance the amplitude and the duration of the late potentials and thus would compromise the reproducibility of repeat SAECG by resulting in apparent changes. An opportunity arose to evaluate the issue in a patient under our care.
Case Report
A 54-year old Pakistani man with a history of hypertension; diabetes mellitus; diabetic nephropathy, retinopathy, and neuropathy; CHF; cerebrovascular accidents; peripheral vascular disease;
coronary artery disease (for which he had undergone aortocoronary bypass grafting 5 years earlier); and depression was admitted with complaints of worsening of his exercise and resting dyspnea and fatigue. On physical examination, his blood pressure was 150/83 mm Hg, his heart rate was 104 bpm, and on auscultation he had rales halfway up both sides of his posterior and lateral chest walls, jugular venous distension, an S3 gallop, and marked PED to his thighs. His weight on admission was 188 lb. A portable chest radiograph showed cardiomegaly with moderate pulmonary congestion, without infiltrates or pleural effusions. Due to the enormous engorgement in his peripheral veins, the possibility of deep vein thrombosis was entertained, but the patient had normal duplex Doppler compression sonograms with normal compressibility and augmentation of spontaneous Doppler signal in all peripheral veins, excluding the diagnosis of deep vein
John E. Madias, MD From the Mount Sinai School of Medicine of New York University, New York, NY; and the Division of Cardiology, Elmhurst Hospital Center, Elmhurst, NY Address for correspondence: John E. Madias, MD, Professor of Medicine (Cardiology), Division of Cardiology, Elmhurst Hospital Center, 79-01 Broadway, Elmhurst, NY 11373 E-mail:
[email protected] 266
difficult cases
september . october 2005
Congestive Heart Failure® (ISSN 1527-5299) is published bimonthly (Feb., April, June, Aug., Oct., Dec.) by CHF, Inc., Three Parklands Drive, Darien, CT 06820-3652. Copyright ©2005 by CHF, Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in Congestive Heart Failure® do not necessarily reflect those of the Editor and Publisher. For copies in excess of 25 or for commercial purposes, please contact Sarah Howell at
[email protected] or 203.656.1711 x106.
Table I. Weights and Serial Electrocardiographic (ECG) Variables in a Patient With Congestive Heart Failure VARIABLES Day (1 = admission)
ECG-1
ECG-2
ECG-3
ECG-4
ECG-5
ECG-6
–1906
–120
1
8
15
Weight (lb)
N/A
N/A
188
174
140
ΣQRS12 (mm)*
218
142
99
131.5
182.5
136.5
237.6
160.1
130
145.9
200.5
151
ΣQRS6 (mm)*
53.5
39.5
21
17.5
39.5
18.5
13
5.5
12.5
ΣQRSV5&V6 (mm)*
32
36
12
10.5
17.5
106
115
103
106
112
44
26
45
30
45
30
QRS horizontal axis (°)
–52
–29
–4
–76
–60
–65
QTc (ms)
431
391
438
460
518
484
QTc minimum (ms)
344
390
328
347
568
452
QTc maximum (ms)
446
585
543
629
815
634
QTc dispersion (ms)
102
195
215
283
250
187
79
68
107
104
79
70
165
166
168
166
190
175
ΣQRS12 (mm)
ΣQRS1&2 (mm)*
QRS duration (ms) QRS frontal axis (°)
Heart rate (bpm) PR interval (ms) ΣP12 (mm)*
5.5
30 (clinic) N/A
24.5 7.5 17 117
15.8
10.1
5
5.6
10.2
ΣP12 (mm)
18
14.4
7.3
8.4
12
P mean duration (ms)
79.1
85
68
71.1
74.5
70.8
P minimum duration (ms)
36
32
26
23
34
39
P maximum duration (ms)
104
112
90
94
106
97
P duration dispersion (ms)
68
80
64
71
72
58
P duration dispersion SD (ms)
23.3
22.3
18.2
20.5
24.5
17.8
P mean area (Ashman U)
20.2
14.8
7.6
9.5
13.9
P frontal axis (°)
70
84
73
P horizontal axis (°)
59
68
7.7 9.2
9.5 76
15
7
21
27
3
42
T frontal axis (°)
250
241
19
131
84
234
T horizontal axis (°)
–45
–40
–51
–77
–59
–63
N/A=not available; ΣQRSx=sum of QRS amplitudeleads; ΣP12=sum of P-wave amplitudes in all 12 leads; *manually measured or calculated; all other ECG variables are automated measurements or calculations
thrombosis. Hemoglobin ranged from 9.4–10.3 g/dL; hematocrit, 30.9%– 33.3%; K+, 3.4–4.3 mEq/L; and Mg++ was 2.0 mg/dL. There was evidence of mild pleural effusions, dilation of the inferior vena cava, and mild ascites on an abdominal sonogram, all attributed to CHF. The patient was placed on IV furosemide and also given aspirin, digoxin, fosinopril, spironolactone, and atorvastatin; a β blocker was to be administered soon after completion of his IV diuresis. The patient was hospitalized for 15 days, during which he had 10 EGGs recorded. We also analyzed the tracings recorded at 6, 30, and 34 days after discharge, and all 24 ECGs in his ECG file, starting more than 5 years ago. Six of the 37 ECGs are shown in Figures 1–6, corresponding to a variety difficult cases
of clinical states. The patient’s ECG on admission (Figure 3) displayed attenuated QRS and P-wave amplitudes in comparison with a tracing recorded 4 months earlier (Figure 2) when the patient was seen in the cardiac clinic. It was striking from just glancing at his serial ECGs that when he was in good compensation of his CHF (Figure 1) the amplitudes of the QRS complexes were increased, while when admitted for exacerbation of CHF the amplitudes were decreased, but less so than noted on Figure 3 from his admission, with the worst fluid accumulation witnessed in this patient. The same phenomenon could be observed during his previous clinic follow-up visits, where the corresponding ECGs with comparatively lower QRS complexes were the ones associated with clinic notes stating that
his PED had increased in comparison with the previous visit, or that PED had been observed, while the note of the previous visit indicated no PED. Measurements on 15 ECGs (10 from his current admission; ECGs of Figures 1, 2, 5, and 6; and the last ECG on postdischarge Day 34) were carried out. The QRS complexes were measured to the nearest 0.5 mm from nadir to zenith, using a magnifying glass and calipers under strong overhead light. The sum of QRS complexes of all 12 leads (ΣQRS12), the six limb leads (ΣQRS6), and leads I and II (ΣQRS1&2) were calculated. The employment of a separate sum of the limb leads was intended to avoid the inherent variability of serial recordings of the precordial leads;19 and since many contemporary ECGs measure only leads I and II, and calculate september . october 2005
267
Congestive Heart Failure® (ISSN 1527-5299) is published bimonthly (Feb., April, June, Aug., Oct., Dec.) by CHF, Inc., Three Parklands Drive, Darien, CT 06820-3652. Copyright ©2005 by CHF, Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in Congestive Heart Failure® do not necessarily reflect those of the Editor and Publisher. For copies in excess of 25 or for commercial purposes, please contact Sarah Howell at
[email protected] or 203.656.1711 x106.
Table II. Correlation of Weights (N=10) With Electrocardiographic Variables in a Patient With Congestive Heart Failure VARIABLE
R
P
ΣQRS12
0.91
0.0002
0.91
0.0003
ΣQRS6
0.93
0.0001
0.97
0.00005
ΣQRSV5&V6
0.73
0.02
QRS duration
0.73
0.02
QTc
0.69
0.03
QTc minimum
0.68
0.03
QTc maximum
0.84
0.002
QTc dispersion
0.40
0.25
ΣP12
0.96
0.00005
ΣP12 (automated)
0.94
0.0001
P mean duration
0.35
0.32
P minimum duration
0.44
0.2
P maximum duration
0.38
0.28
P duration dispersion
0.12
0.74
P duration dispersion SD
0.64
0.05
P mean area
0.85
0.002
ΣQRS12 (automated) ΣQRS1&2
ΣQRSx=sum of QRS amplitudeleads; ΣP12=sum of P-wave amplitudes in all 12 leads
Table III. Correlation of QRS and P-Wave Amplitudes Measured Manually and by the Automated Program in a Patient With Congestive Heart Failure VARIABLES MANUAL
AUTOMATED
ΣQRS12
ΣQRS12
ΣP12
ΣP12
R
P
1.0
0.00005
0.91
0.00005
ΣQRS12=sum of QRS amplitude in all 12 leads; ΣP12=sum of P-wave amplitudes in all 12 leads
Figure 1. Electrocardiograph from a period of good compensation of congestive heart failure
the other four limb leads, the ΣQRS1&2 could be representative of ΣQRS6.20 Since previous experience in patients with anasarca showed a disproportionate reduction in the amplitude of leads 268
difficult cases
V5 and V6 in comparison with the other precordial leads due to gravity-mediated fluid retention in the dorsal aspect of the body,7 the sum of QRS amplitude in these two leads (ΣQRSV5&V6) was
also calculated. In addition, the amplitude of P waves was measured manually, and the sum of the amplitudes of P waves from all 12 leads (ΣP12) was calculated.8 The QRS duration, heart rate, PR interval, P axis, QRS axis, and QT interval corrected by the Bazzett’s formula (QTc), as provided in the automated ECG interpretation hard copy form,21 was also considered. Other measurements provided by the extended automated ECG interpretation program21 for our ECGs included amplitude (mm) and duration (ms) of P waves, and amplitude of R waves and S waves of QRS complexes of all 12 ECG leads, from which another calculation of the ΣQRS12 and ΣP12 was derived. In addition, this automated system provided QT in all 12 leads, QT minimum, QT maximum, QT dispersion, and QT dispersion SD, all in milliseconds, and P area of all 12 leads in Ashman units (one Ashman unit = 40 ms × 0.1 mV),21 and P minimum, P maximum, and dispersion of P-wave duration in ms. For the above QT interval variables, the corresponding QTc in ms was also provided. The objective was to evaluate serially all these ECG parameters, and for those available by both manual and automated derivation, to compare the values from these two measuring approaches. Data of all the foregoing 27 parameters corresponding to six ECGs are shown in Table I. Pearson correlation of weights over the course of hospitalization (N=10)22,23 as the independent variable, with ECG variables as the dependent variables, are shown in Table II. Weights obtained in the clinic were not included in this analysis, since they were not obtained under the strict conditions of weights obtained in the hospital (the same scale used, weights obtained by the author, always in the morning, and with the patient wearing only his gown). Loss of weight was associated with increase in the ΣQRS12, ΣQRS6, ΣQRS1&2, ΣQRSV5&V6, QRS duration, QTc, and ΣP. Correlations of QRS and P-wave amplitudes measured manually and by the automated program (Table III) were excellent. The patient’s ECG 5 years before his most recent admission september . october 2005
Congestive Heart Failure® (ISSN 1527-5299) is published bimonthly (Feb., April, June, Aug., Oct., Dec.) by CHF, Inc., Three Parklands Drive, Darien, CT 06820-3652. Copyright ©2005 by CHF, Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publishers. The opinions and ideas expressed in Congestive Heart Failure® do not necessarily reflect those of the Editor and Publisher. For copies in excess of 25 or for commercial purposes, please contact Sarah Howell at
[email protected] or 203.656.1711 x106.
(Figure 1) satisfied criteria for left ventricular hypertrophy by the RomhiltEstes, Sokolow-Lyon, and Cornell voltage systems, while the ECG on admission (Figure 3) did not satisfy criteria for left ventricular hypertrophy by any of these three systems.24 At discharge, the ECG (Figure 4) again satisfied criteria for left ventricular hypertrophy by the RomhiltEstes and Cornell criteria. The patient had two SAECGs, on Day 8 (Figure 7A) and on Day 15 (Figure 7B) of hospitalization, before his discharge, which did not disclose any changes, while the corresponding ECGs revealed an increase of ΣQRS12 by 38.8%, associated with a weight loss of 37 lb. When the patient was seen in the clinic after his discharge he had clear lungs, no S3 gallop, and only trace peripheral edema in his lower extremities, with impressive skin wrinkling from the receded anasarca; but the subsequent clinic visit disclosed fluid reaccumulation. Correlation of ΣQRS12, ΣQRS6, ΣQRS1&2, and ΣQRSV5&V6 with ΣP was excellent, with r (0.92, 0.94, 0.95, and 0.68, respectively) and with p=0.00005 for ΣQRS12, ΣQRS6, and ΣQRS1&2, and p=0.005 for ΣQRSV5&V6. The correlation of ΣQRS12, ΣQRS6, ΣQRS1&2, and ΣQRSV5&V6 with the QRS duration and QTc was poor, with p values >0.05. This also applied to similar correlations with the automated ΣQRS12. The SAECG recorded on Day 8 and repeated on Day 15 (discharge) of hospitalization (Figure 7) did not change. On both occasions the SAECG was indicative of the presence of abnormal late potentials. The Day 8 study showed a filtered QRS (fQRS) of 126 ms, a root mean square voltage in the last 40 ms (RMS40) of 14.7 µV, and a duration of signal