Electrocardiographic Voltage in Pericardial Effusion· Donald v. Unverferth, M.D.;·· Thomas E. WiUitJms, M.D.;t and Phillip K. Fulkerson, M.D.*

The usefulness of the electrocardiographic sign of ''low voltage" in the diagnosis of pericardia1 etfosion was investigated in 122 patients comprising three study groups. Sixty-four patients (group 1) had a pericardial effusion detected and measured by echocardiographic studies. The volume of the effusion showed no correlation with electroeardiographic voltage. A second group of 36 patients was identified as having low voltage on routine electrocardiograms. Only 13 (36 percent) had echocardiographically demonstrable pericardial effusion. Group 3

R educed electrocardiographic voltage has long been recognized as an abnormal finding. Sprague and White! in 1926 and Willius and Killins2 in 1927 commented on the poor prognosis of For editorial comment, see page 113

those patients who had less than a 5 mm deflection

in all of their standard limb leads. The low voltage

electrocardiogram has been associated with pericardial effusion since the work of Oppenheimer and Mann" in 1923. Although most electrocardiographic texts4 - 11 cite pericardia! effusion as one of the principal causes of low voltage, the diagnostic usefulness of this sign has not previously been evaluated. The purpose of this report is to examine the role of low voltage in the recognition of pericardial effusion and to discuss current standards for low voltage. MATERIALS AND METHODS

The patients in this study were hospitalized from April 1968 to June 1977. Echocardiographic studies were performed using one of two ultrasonoscopes ( Smith-Kline Ekoline 20A or Unirad ultrasonoscope) coupled with a recording system (Irex or Electronics for Medicine). Electrocardiograms were obtained on an electrocardiograph (Marquette 3000). Routine standardization was performed with each ECG, so that 1 mV was recorded at 10 mm of deflection. Six-foot chest roentgenograms were obtained in the °From the Departments of Medicine and Surgery, Ohio State University, Columbus. Supported by a grant from the Central Ohio Heart Chapter of the American Heart Association. ··Fellow in Cardiology. t Associate Professor of Surgery, Division of Thoracic Surgery. tAssistant Professor of Medicine, Division of Cardiology. Manuscript received May 22; revision accepted July 27. Reprint requests: Dr. Unverlerth, Room657 Means Hall, 410 West Tenth Avenue, Columbus 43210

CHEST, 75: 2, FEBRUARY, 1979

consisted of 22 patients who required pericardiocentesis. The ECGs obtained immediately after pericanllocentesls showed an increase in avel1lle voltage of 0.48 mm in the limb leads and 0.83 mm in the precordial leads for each 100 mI of IBid removed. This study demoDStrates that a single ECG with "low volfage" is not useful In the dialnosis of pericardial effusion but that a reduction in the voltage of serial ECGs may suggest the development 01 pericardia1 eftosion.

posteroanterior and lateral projections. The 122 patients in this investigation comprised three groups for study. Group 1 consisted of 64 patients with echocardiographically documented pericardia! effusion. Pericardial volume was estimated from the echocardiogram according to the method of Horowitz et al. l 2 Clinical records were reviewed, and patients with radiographically demonstrable pleural effusions were excluded. The patients' ECGs were evaluated for voltage amplitude by the following averaging technique. The QRS amplitude (apex of R wave to nadir of S wave) was summed for each of the six limb and precordial leads. This value was then divided by six (the number of leads used), and a mean value was obtained for either limb or precordial leads. In our study, we chose to define low voltage as an avef'age limb lead voltage under 5 mm or an average precordial voltage under 10 nun. Patients in group 2 (36 patients), who were selected because they had low voltage on routine ECGs, also underwent diagnostic echocardiographic studies. The patients' height, weight, blood pressure, and indices of thyroid function (thyroxine, triiodothyronine resin uptake, free thyroxine index), were obtained; and chest roentgenograms were reviewed for pleural fluid. The third group (group 3) consisted of 22 patients with pericardial effusion who required surgical drainage for therapy (11 patients) , diagnosis (one patient) , or both (ten patients). The volume of the pericardial aspirate was measured directly at inferior pericardiotomy. The electrocardiographic voltage before and after pericardial drainage was averaged for the precordial and limb leads, as discussed previously. All statistical procedures were considered significant at the level of P < 0.05 or were noted as not significant.

REsuLTS The echocardiographically determined volume of the pericardia! effusion in group 1 ranged from 87 ml to 1,016 ml (mean, 293 ml). The causes of the pericardia! fluid and the number of patients with each cause were as follows: congestive heart failure,

ECG VOLTAGE IN PERICARDIAL EFFUSION 157

•

1400.0

1

1200.0

~

~I ... ...

800.0

II

800.0

~ ~

•

1000.0

~

400.0 200.0

o

•

• • •• .1 •• ••••• ••• ,.): ••

. ,.-):' .. ... • . . •• +-----.......- - -.......- - -.....- - -.......---........

FiGURE 1. Plot of average voltages in limb leads against voltages of pericardial effusion o 5.0 IQO 15.0 20.0 25.0 in group 1 demonstrates no correlation AWRAGE .... LEAD VOLTAGE R+S WAVE IN ..WMETERS (r = -0.05).

29 (45 percent); hypothyroidism, ten (16 percent); hepatic disease, three (5 percent); congestive cardiomyopathy, three (5 percent); renal failure two (3 percent); and undetermined, 17 (27 percent). Low voltage was present in both limb and precordial leads in nine patients (14 percent), while 13 patients (20 percent) had decreased voltage in limb leads only and six (9 percent) in precordial leads only. In all, 56 percent (36) of the patients in group 1 with documented pericardial effusion failed to demonstrate low voltage on their ECGs. Furthermore, no significant correlation existed between the volume of the effusion and the electrocardiographic low voltage (r +0.05; Fig 1 and 2). The 36 patients in group 2 most commonly had an average limb-lead voltage below 5 mm (35/36 or {1'/ percent). Seventeen patients had an average precordial voltage below 10 mm, while 16 patients (44 percent) had an average voltage in both limb and precordial leads below our standards. The condi-

=

1400.0

tions most commonly associated with low voltage and the number of patients with each condition were as follows: pleural effusion, 19 (53 percent); pericardial effusion, 13 (36 percent) ; obesity, six (17 percent); pneumonia, four (11 percent); congestive cardiomyopathy, three (8 percent); emphysema, three (8 percent); hypothyroidism, one (3 percent); and amyloidosis, one (3 percent). Many patients had two or more factors contibuting to their voltage. Although pericardial effusion was found in 13 of these patients, only flve (14 percent) of these had a pericardial effusion without a coexisting pleural effusion. Those who had low voltage on both precordial and limb leads were more likely to have a pericardial effusion (7 /16; 44 percent) than those who had either low limb or precordial voltage alone (6/20,30 percent); however, this difference was not statistically significant (Fisher's exact test of twotailed probability of randomness, P = 0.61). Pericardia] drainage performed at the time of

•

11200.0

~

•

1000.0 1

;1 ~ ~

100 0 .

~!

L

~

•

800.0

400.0

200.0

.

....,-- , ... -. •

....

• :..

•. I

• •• • .......... I •••• • •••• • ••

• •

2. Plot of average voltages in precordial leads against volume of pericardial effusion demonstrates no correlation (r +0.05).

=

O+-----r---oy----......---,......--.----....-------. o 5.0 10D 15.0 20.0 18.0 30.0 35.0

158 UNYERFERTH, WILLIAMS, FULKERSON

CHEST, 75: 2, FEBRUARY, 1979

FiGURE

AVERAGE PRECORDIAL VOLTAGE R+S WNE IN MLUMETERS

0 - 6 - - - - -.....- - - - - - , . - - - - - - ,

o

eee

1000

laoo

or the myocardial edema seen in hypothyroidism 7.11.1"16 also beget low voltage. Low voltage in limb leads may also be seen in cases where the mean QRS axis is predominantly directed in a horizontal or sagittal plane. While this might cause a reduced frontal-plane voltage, the precordial voltage could remain normal. The electrical conductivity of body tissue varies with the physical properties of that tissue. In order of increasing resistivity, these tissues are blood, liver, cardiac muscle, skeletal muscle, lung, and fat. 7 An increase in the distance between the heart and the recording electrode is frequently the cause of low voltage, as seen in obesity or chronic pulmonary disease; however, in these cases the interposed adipose tissue or air in the pulmonary field may also act to increase the electrical isolation between the heart and the recording electrode. The poor correlation in group 1 between the voltage of the pericardial effusion and the reduction of electrocardiographic voltage is due to our inability to correct for the multiplicity of factors which affect voltage. The patients with low voltage in group 2 had one or a combination of causes for their low voltage, but only 36 percent had pericardial effusion. The contributing factors in this group's low voltage

MILLILITERS OF PERICARDIAL EFFUSION

40

REMOVED BY PERICARDIOCENTESIS

FIGURE

3. Electrocardiographic voltage .(vertical axis) rises

in limb leads with aspiration of volume of fluid (horizontal axis.}.

inferior pericardiotomy produced 150 to 1,400 ml (mean, 580 ml) of fluid in the patients in group 3. For each 100 ml of fluid aspirated, the electrocardiographic voltage rose an average of 0.48 mm (range, o to 4.8 mm) in each limb lead and 0.83 mm (range, o to 9.0 mm) in each precordial lead. There was an increase in either precordial or limb lead voltage or both (Fig 3 and 4) in every case.

20

2 ~

~

~

31 (I)

DISCUSSION

+ a:

The amplitude of the electrocardiographic voltages recorded from the surface of any patient is subject to a number of modifying factors. The surface ECG may record a reduced QRS amplitude if there are abnormalities of either generation or transmission of voltage. Generally, low voltage results from one of the following conditions: (1) a reduced voltage arising from the cardiac muscle; (2) an unusual mean QRS axis; (3) increased resistivity of interposed tissues; and (4) increased distance between the heart and the recording electrode. The voltage generated by the depolarization of cardiac muscle may be reduced if the ventricles have thin walls or show multiple scars, as from previous infarctions. 7.11 Infiltrative cardiomyopathies such as amyloidosist-"

1&1

CHEST, 75: 2, FEBRUARY, 1979

~

10

~

0

C

1&1 ....J

ID

2

:;

0+-------....-----......- - - -... o eoe 1000 11500 MILLILITERS OF PERICARDIAL FLUID REMOVED BY PERICARDIOCENTESIS FIGURE 4. Electrocardiographic voltage (oertical axis) rises in precordial leads with aspiration of volume of Huid (hori-

zontal axis).

ECG VOLTAGE IN PERICARDIAL EFFUSION 159

in order of decreasing frequency were pleural effusion, pericardial effusion, obesity, pneumonia, cardiomyopathy, emphysema, hypothyroidism, and amyloidosis. The application of electrocardiographic low voltage to the evaluation of pericardial effusion is suggested in group 3. The aspiration of fluid from these 22 patients resulted in a predictable rise of the voltage in all patients. Voltage in each limb lead rose an average of 0.48 mm, and each precordial lead increased in amplitude an average of 0.83 mm for each 100 ml of fluid removed. Because there was an average volume of 580 ml of pericardial fluid removed, the mean electrocardiographic change after pericardiocentesis was a remarkable rise of 2.8 mm in each limb lead and 4.8 mm in each precordial lead. This information might be applied to a patient whose previous ECGs are available for comparison. Any patient who develops or increases a pericardial effusion might be expected to reduce his electrocardiographic voltage. Therefore, a reduction in the voltage of a previously normal ECG, especially in the absence of pleural effusion, suggests development of a pericardial effusion. Since these electrocardiographic changes would be serial observations in the same patients, many of the variables discussed previously are controlled. Only three of our patients in group 3 had an ECG from before the effusion available for comparison. Serial tracings did, indeed, show a reduction in the voltage during the period of pericardial effusion in these three patients. Thus, a decrease in the voltage may alert the clinician to the development of an effusion. CONCLUSION

Electrocardiographic low voltage is neither specific nor sensitive in the diagnosis of pericardia! effusion and is not diagnostic as a single observation in the individual patient. Low voltage may be helpful if a previous ECG is available for comparison. A

160 UNYERFERTH, WILLIAMS, FULKERSON

reduction in voltage suggests the accumulation of a pericardial effusion and should lead to more definitive studies. REFERENCES

1 Sprague H, White PD: Significance of electrocardiograms of low voltage. J Clin Invest 3: 109-121, 1926 2 W illius FA, Killins W A: The occurrence and significance of electrocardiograms of low voltage. Arch Intern Med 40:332-339, 1927 3 Oppenheimer BS, Mann H: An electrocardiographic sign in pericardial effusion. Proc Soc Exp Biol Med 20:431432, 1923 4 Master AM, Lasser RP, Rosenfeld I, et al: The Electrocardiogram and Chest X-ray in Disease of the Heart. Philadelphia, Lea and Febiger, 1963, p 50 5 Friedman HH: Outline of Electrocardiography. New York, McCra\v-Hill Book Co, Inc, 1963, pp 63-64 6 Littman D: Textbook of Electrocardiography. New York, Harper and Bow, 1972, pp 55-56 7 Marriott HJL: Practical Electrocardiography. Baltimore, Williams and Wilkins Co, 1972, pp 21-22 8 Schaffer AI: Cardiography in General Practice. Baltimore, Williams and Wilkins Co, 1952, p 70 9 Simonson E: Differentiation between Normal and Abnormal in Electrocardiography. St. Louis, CV Mosby Co, 1961, p 266 10 Goldman ~IJ: Principles of Clinical Electrocardiography. Los Altos, Calif. Lange Medical Publications, 1967, p 267 11 Constant J: Learning Electrocardiograhy: A Complete Course. Boston, Little, Brown and Co, 1973, pp 321322 12 Horowitz xrs, Schultz CS, Stinson EB, et al: Sensitivity and specificity of echocardiographic diagnosis of pericardial effusion. Circulation 50:239-247, 1974 13 Buja LM, Khai NB, Roberts WC: Clinically significant cardiac amyloidosis. Am J Cardio126:394-405, 1970 14 Graybiel A, White PD, Wheeler L, et al: Electrocardiography in Practice. Philadelphia, WB Saunders Co, 1952, p 242 15 Lepeschkin E: Modern Electrocardiography. Baltimore, Williams and Wilkins Co, 1951, pp 278-281 16 Thacher C, White PD: The electrocardiogram in myxedema. Am J Med Sci 171:61-66,1926 17 Rush S, Abildskov JA, McFee R: Resistivity of body tissues at low frequencies. Circ Res 12: 40-50, 1963

CHEST, 75: 2, FEBRUARY, 1979