Professional Documents
Culture Documents
A Study of The Normal QRS-T
A Study of The Normal QRS-T
161
162 ZIEGLERAND BLOOMFIELD
A~ T v s A~ O R S
FRONTAL PLANE
301 NORMAL ELECTROCARDIOGRAMS
90 o .
80 oJJ7 " "
70
6o
9 /~
50 ........................ ~>
40
~- 30
2O
10
-10
-20
-30
- 3 0i - 2 ~ 0 - l l 0 0I ]
10 210 30] 410 50
I
610 710 810 910 1~)0
A~
Fig. 1. A~ vs. A~ in the frontal plane for 30l normal electrocardiograms in persons over age 35. The
lines A~ --60 ~ = A~ and A~ +35 ~ = A~ encompass approximately 94 per cent of values.
proximately 94 per cent of the distribution do not A~ accurately. Since the slope of the curve
accurately describe the A~ population. Rather it of the mean A~ values becomes steeper past
can be seen that from A~ --30 ~ to 0 ~ the + 3 8 ~ regression lines for the A~ population
normal A~ is always to the right of A~ on each side of + 38 ~ were calculated, along with
(A~ always more than A~ This trend con- 95 7o confidence intervals for these lines. From
tinues from A~ 0 ~ to A~ + 3 0 ~ with a A~ --30 ~ to + 3 5 ~, the equation for the
majority of the A~ values lying to the right of regression line was found to be A~ = 0.059
A~ From A~ + 3 0 ~ to + 6 0 ~ A~ is A~ + 33.71~ with the expression 1.96
distributed evenly on each side of A~ From ~/292.91 + .006 (A~ -- 11.15) 2 13 repre-
A~ + 6 0 ~ to + 1 0 0 ~ nearly all of the A~ senting the 95 7o confidence limits of the normal
values lie to the left of A~ (A~ less than A~ for any A~ from - 3 0 ~ to + 3 5 ~ From
A~ A~ + 4 0 ~ to + 1 0 0 ~ the equation for
Fig. 2 is a plot of the mean and standard the regression line was found to be A~ =
deviations of A~ values for a given A~ A~ 0.317 A~ + 38.18 ~ with the expression 1.96
values have been grouped to make the number of ~]309.71 + .0076 (A~ - 57.14~ 2 represent-
A~ values ten or more for each A~ range. ing the 95 70 confidence limits of the normal A~
The curve of best fit through the mean A~ values for any A~ from + 4 0 ~ to + 1 0 0 ~. Fig. 3
of the sample indicates that two regression lines shows each regression line with its respective 95 70
with corresponding 95 7O confidence limits would limits, and Table 2 gives the 9570 confidence
describe the normal A~ related to a specific limits of the normal A~ at any A~ from
FRONTAL PLANE NORMAL QRS-T ANGLE 163
A~ T vs A~ QRS
DEVIATIONS for A~ T
100 -
90-
80
70 ,s j
60
50-
1~
18 14 1•
~Lfl~-~14
2 4 11
- 4- II
40-
30-
20-
10-
0-
-10
-20
-3 (
I I I I I I I I I I I I I I
-30-20 -10 0 10 20 30 40 50 60 70 80 90 100
A~ Q R S
Fig. 2. Normal A~ vs. A~ values plotted as mean and standard deviations of A~ values.
- 3 0 ~ to 4.100 ~ Both Fig. 3 and Table 2 also regression coefficient for the total sample popula-
compare our data with the earlier studies of tion (b = 0.227) was significantly different
Helm la. (p (.05) than the regression coefficients of the
The regression line for the total sample was also two separate regression lines (b = 0.059, b =
calculated. The equation for this line is A~ = 0.317).
0.227 A~ 4- 32.57 ~ with the expression 3. In comparing the mean A~ values of con-
1.96 ~/255.84 4" .0012 (A~ -- 33.31) 2 secutive 30 ~ ranges (approximately) of A~
representing the 95 % confidence limits of the (Table 3), it can be seen that the mean A~ values
normal A~ for any A~ value from --30 ~ for A~ -30 ~ to 0 ~ and for A~ 0 ~ to
to 4-100 ~ We believe that two separate regression 4-35 ~ are not significantly different. However, for
lines describe the interrelationship of A~ and the next two A~ ranges (A~ -t-40 ~ to
A~ better than one regression line. The fol- 4--60~ and A~ 4-60 ~ to 4,100 ~ the mean
lowing reasons are offered for this view: A~ values rise dramatically and, in each case,
1. The confidence limits found using two sepa- are significantly different from the mean A~
rate regression lines describe our sample popula- values of the A~ range immediately preced-
tion much better than the confidence limits found ing. This table indicates the need to describe our
by using one regression line (compare Figs. 1 sample population with two regression lines; one
and 3). line for the A~ values for A~ --30 ~ to -1-35~
2. By use of the t-test for the confidence limits and another line for the A~ values for A~
of a regression coefficient, it was found that the 4-40 ~ to 4-100 ~.
164 Z I E G L E R AND BLOOMFIELD
A~ T vs A~ ORS
REGRESSION LINES AND 5% FIDUCIAL LIMITS ~','
110 ..~
I00 v\~
90 O~ 0 ~ >
80, 5% F I D U C I A L LIMITS /
70 - r ~ * ~ 1 7 6
0- __ o.~,~.
- I0 oO.**
-20 ~~ ~
-30" ....,1%o""
..o s
..&~o
-40-
-3'0--2'0--1~) (~ 1'0 2'0 3'0 4'0 5'0 6'0 7'0 8'0 9'0 1(30
A ORS
Fig. 3. Regression lines and 5 per cent fiducial limits of A~ vs. A~ for the data in this paper con]pared
to Helm 14.
limits H e l m concluded "that any designated represent the range of the n o r m a l A ~ with
upper limit o f normal (e.g. 50 ~) for the frontal respect to A ~ for his sample. Our data indi-
Q R S - T angle may be too large when A ~ is cate that two regression lines, one with a nearly
located on one side of A ~ and t o o small when zero slope and one with a decidedly steeper slope
A ~ is located on the other side of A ~ '' H e along with the respective confidence intervals of
points out that this is particularly true when each, best represent our sample population. The
A~ is in a nearly horizontal or nearly vertical regression coefficients for both of the regression
position. lines determined for our sample were significantly
Figs. 1, 2 and 3, and Table 2 show that our data different f r o m the regression coefficient of the
generally agrees with that of Helm. Table 2 regression line for H e l m ' s sample (b = .514) at
shows that the lower limits of the normal A ~ the .05 level. Because of differences in the regres-
in our sample vary from - - 2 ~ to + 2 ~ and the sion lines used, the limits of the n o r m a l A ~ of
upper limits vary f r o m + 66 ~ to + 70 ~ as A ~ the two samples show much disagreement at the
goes from - - 3 0 ~ to + 3 5 ~ These limits indicate extremes of the normal A ~ range. Even for
that in nearly electrically horizontal hearts, the A~ values at the extreme leftward limit of the
n o r m a l A ~ always lies to the right of A ~ normal A ~ range, our data suggest that any
and that the Q R S - T angle may be as large as 96 ~ A ~ value less than 0 ~ (T wave inverted in aVF)
For A~ + 4 0 ~ to + 1 0 0 ~ the lower limits of is abnormal. Simonson 15 and Hiss et al. 5 came to
the normal A ~ range from + 1 2 ~ to + 3 0 ~ and a similar conclusion with the former setting the
the upper limits range from + 6 9 ~ to + 8 9 ~, These lower limit of T amplitude in adults above 20
limits show that for nearly vertical hearts, A ~ years of age at --0.2 ram. in a V F and the latter
always lies to the left of A ~ and the Q R S - T giving no negative values for the T wave in aVF
angle may be as large as 70 ~. for adults above 20 years of age. However, H e l m ' s
Fig. 3 and Table 2 which compare the 95 ~o
confidence limits of our sample to those of H e l m
TABLE 2
show a general agreement between the two sam-
The Normal Relationship of A~ to A~ in
pies, especially between A ~ + 15 ~ and + 6 0 ~ Persons Over Age 35
However the two samples differ in several ways.
H e l m ' s data is based on a smaller sample with an 95 % Confidence Limits (in degrees)
A~ of A~
age range of 20 to 71 years of age. Our data is
(Degrees) Ziegler and Bloomfield Helm (14)
based strictly on persons over 35 years of age,
an age group in which nonspecific T wave changes --30 -- 2 t o + 6 6 --36 to +47
and variations in the Q R S - T angle may have in- --25 -- 2 t o + 6 6 --32 to +48
--20 -- lto+66 --29to+50
creased significance. H e l m used a single regression --15 -- 1 to +67 --25 to +51
line with the confidence limits of this line to --10 -- lto+67 --21 to +53
-- 5 0to +67 --18 to +54
0 0 to +67 --14 to +56
5 0 to +68 --11 to +58
TABLE 1 10 + lto+68 -- 8 t o + 6 0
Age Distribution of 301 Normal Electrocardiograms 15 + 1 to +68 -- 4 to + 6 2
in Persons Over Age 35 20 + lto+68 -- l t o + 6 4
25 + 2 t o +69 + 2 t o +66
Age Group N 30 + 2to+69 + 5to+68
35 + 2 to +70 + 7 to +70
35-40 39 40 +12 to +69 + 1 0 to +73
41--45 54 45 +14 to +71 + 1 2 to +76
46-50 45 50 +16 to + 7 2 +15 to +78
51-55 42 55 +17 to +74 +17 to +81
56-60 41 60 +19 to +76 +19 to +84
61-65 25 65 +20 to +77 +22 to +87
66-70 24 70 + 2 2 to +79 +24 to + 9 0
71-75 19 75 +23 to +80 +25 to +93
76-80 8 80 +25 to + 8 2 +27 to +97
Greater than 80 4 85 +26 to +84 +29 to +100
Total: 301 90 +28 to +86 +31 to +104
Mean age of sample = 53.05 95 +29 to +87
Median age of sample = 52.02 100 +30 to +89
166 Z I E G L E R AND BLOOMFIELD
data (Table 2, Fig. 3) indicate that the normal to use. Both Hiss, et al. 5 and Zao, et al. s have
A ~ can rotate as far left as - - 3 6 ~ If this were so, stated that 15 degrees is the limit of accuracy
the normal T wave could be inverted in both II obtainable f r o m the hexaxial reference system.
and aVF. Although one source ~6 lists a m i n i m u m M a r r i o t t 4 has given the system credit for greater
T wave in a V F of --0.6 ram. for adults over 20 accuracy, perhaps even 5 degrees. Burch and
years of age, none of the three sources 5,1~,~6 set Winsor 19 and A s h m a n and Hull 2~ have pointed
the lower limit of the T wave in II below 0 ram. out that true accuracy for A ~ and A ~ re-
Furthermore, clinical expelience is strongly quires measurement of the area encompassed by
against normal inverted T waves in aVF, even the Q R S and T waves. However, Simonson, et
though statistical treatment of the H e l m data al. 21 reported a high positive correlation between
may have led to this conclusion. the areas and the amplitudes of both the Q R S
At the extreme rightward limit of the normal and T waves. We used the transitional zone tech-
A~ range, our data suggests that the normal nique t2 using only the wave amplitudes to de-
A ~ never exceeds 89 ~ and thus that the normal termine the angles because this method is the one
T wave is never inverted in I. Simonson ~5 placed most useful to the day-to-day interpreter of
the lower limit of T wave amplitude for adults electrocardiograms who usually works with the
over 20 years of age in lead I at 0.5 ram., and Hiss, Einthoven triangle. Graettinger, et al. 12 point out
et al.:' show no negative T wave values for adults the fairly accurate results of this method.
over 20. However, Helm's data indicates that the The Q R S - T frontal angle should be determined
normal A ~ can rotate as far tight at 104 ~ This, both in magnitude (the absolute number of de-
too, is contrary to clinical experience. grees of the angle) and in direction (the position
By using one regression line fol his entire popu- of A ~ to the right or left of the A~ Since
lation, Helm did not observe the change in the the hexaxial reference system is useful for every-
A ~ population as A ~ increases which we day electrocardiographic interpretation, the limits
have observed in our data. Since our data is based of normality demonstrated in Fig. 3 and Table 2
on a larger sample and deals with an older popu- for A ~ in relation to A ~ should be of im-
lation, we believe the two separate regression lines mediate practical value.
with corresponding confidence intervals more
accurately describe the relation of A ~ to A ~
in persons over age 35. REFERENCES
We recognize that the method used for the 1. Master, A. M., Lasser, R. P., Rosenfeld, l., and
Donoso, E.: The Electrocardiogram and Chest
determination of A ~ and A ~ is subject to
X-ray. Diseases of the Heart. Philadelphia, Lea
limitations. Zao, et al. s argued that the Einthoven and Febiger, 1963, p. 49.
triangle upon which the hexaxial reference system 2. Burch, G. E., and DePasquale, N. P.: Electro-
cardiography in the Diagnosis of Congenital
is based cannot give completely accurate results.
Heart Disease. Philadelphia, Lea and Febiger,
T h e Burger triangle s,~7,1s is said by some to be 1967, p. 45, p. 61.
more accurate, but it is harder and less practical 3. Garcia-Palmieri, M. R., Rodriguez, R. C., and
Girod, C. E. : The Electrocardiogram and Vector-
cardiogram in Congenital Heart Disease. New
York, Grune and Stratton, 1965, p. 36.
TABLE 3 4. Marriott, H. J. L. : Practical Electrocardiography.
Comparison of A~ Populations for Different A~ Baltimore, Williams and Wilkins, 1968, 4th ed.,
Ranges in Persons Over Age 35 p. 34.
5. Hiss, R. G., Lamb, L. E., and Allen, M. F." Elec-
P Value trocardiographic findings in 67,375 asymptomatic
Between subjects. X. Normal values. Am. J. Cardiol., 6:
A~ Consecutive 200, 1960.
A~ Range N (mean 4- s.d.) Means 6. Friedman, H. H. : Outline of Electrocardiography.
New York, Blakiston Division, McGraw-Hill
--30 ~ to 0 ~ 48 33.4 ~ 4- 19.8 ~ Book Company, 1963, p. 79.
>0.1 7. Grant, R. P.: Clinical Electrocardiography: The
+ 5 ~ to +35 ~ 108 34.7 ~ 4- 16.3 ~ Spatial Vector Approach. New York, Blakiston
<0.001 Division, McGraw-Hill Book Company, 1957, p.
+ 4 0 ~ to +60 ~ 97 43.0 ~ 4- 14.2 ~ 59.
<0.001 8. Zao, Z. Z., Herrman, G. R., and Hejtmancik,
+ 6 5 ~ to +100 ~ 48 54.3 ~ 4- 13.8 ~ M. R. : A study of cardiac vectors in the frontal
plane. Am. Heart J., 56: 65, 1958.
FRONTAL PLANE NORMAL QRS-T ANGLE 167
9. Graybiel, A., White, P. D., Wheeler, L., and 16. Criteria Committee of the New York Heart
Williams, C.: Electrocardiography in Practice. Association: Diseases of the Heart and Blood
Philadelphia, W. B. Saunders Co., 1952. Vessels, Nomenclature and Criteria for Diagnosis.
10. Dimond, E. G. : Electrocardiography and Vector- Boston, New York Heart Association, 1964, 6th
cardiography. Boston, Little, Brown and Com- ed., p. 437.
pany, 1967, 4th ed., p. 59. 17. Burger, H. C., and vanMilaan, J, B. : Heart vector
11. Arbeit, S. R., Rubin, I. L., and Gross, H." Differ- and leads. I. Brit. Heart J., 8: 157, 1946. II. Brit.
ential Diagnosis of the Electrocardiogram. Phila- Heart J., 9: 154, 1947.
delphia, F. A. Davis Company, 1960, p. 15. 18. Zao, Z. Z.: Burger triangle as a method of cor-
12. Graettinger, J. S., Packard, J. M., and Graybiel, recting inaccuracies of the Einthoven Triangle.
A. A new method of equating and presenting bi- Science 122: 375, 1955.
polar and unipolar extremity leads in the electro- 19. Burch, G. E., Winsor, T.: A Primer of Electro-
cardiogram. Am. J. Med., 11: 3, 1953. cardiography. Philadelphia, Lea and Febiger,
13. Snedecor, G. W., Cochran, W. G." Statistical 1955. 5th ed., p. 111.
Methods. Ames, Iowa, The Iowa State University 20. Ashman, R., and Hull, E. : Essentials of Electro-
Press, 1967, 6th ed., p. 135. cardiography. New York, The MacMillan Com-
14. Helm, R. A." The direction of mean QRS and T pany, 1944, 2nd ed., p. 69.
vectors. I. Einthoven frontal plane. Am. Heart J., 21. Simonson, E., Schmitt, O. H., Dahl, J., Fry, D.,
48: 224, 1954. and Bakken, E. E.: The theoretical and experi-
15. Simonson, E.: Differentiation between Normal mental bases of the frontal plane ventricular
and Abnormal in Electrocardiography. St. Louis, gradient and its spatial counterpart. Am. Heart
C. V. Mosby Company, 1961, p. 142. J., 47: 122, 1954.