Evaluation of a QRS Scoring System
for Estimating Myocardial Infarct Size
I.
Specificity and Observer Agreement
GALEN S. WAGNER, M.D., CHRISTOPHER J. FREYE, B.S., SEBASTIAN T. PALMERI, M.D.,
STEVEN F. ROARK, M.D., NANCY C. STACK, B.A., RAYMOND E. IDEKER, M.D., PH.D.,
FRANK E. HARRELL, JR., PH.D., AND RONALD H. SELVESTER, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
SUMMARY We evaluated a simplified version of a previously developed QRS scoring system for estimating infarct size using observations of Q- and R-wave durations and R/Q and R/S amplitude ratios in the standard 12-lead ECG. Groups of subjects with a minimal likelihood of having myocardial infarcts and minimal
likelihood of having common noninfarction sources of QRS modification were studied to establish the specificity of each of the 37 criteria. Only two criteria required modification to achieve 95% specificity. These 37 criteria form the basis of a 29-point QRS scoring system. A 98% specificity was achieved when a score of more
than 2 points was required to identify a myocardial infarct.
Fifty patients were studied to determine the intra- and interobserver agreement with this scoring system.
Each criterion achieved at least 91% intra- and interobserver agreement.
These impressive levels of specificity and observer agreement must be matched by high sensitivity of the scoring system and a good correlation between the point score and infarct size in patients with proven infarcts if the
point score is to be useful for detecting and sizing infarcts. Sensitivity and correlation between point score and
infarct size are evaluated in later studies in this series. The standard ECG is inexpensive and can be obtained
repetitively and noninvasively; its QRS complex may be an important means of estimating the size, presence
and location of myocardial infarcts.
bility of those criteria that involve certain quantifiable
aspects of the QRS complex.
MANY CLINICAL METHODS have been evaluated to determine their ability to estimate the size of
myocardial infarcts. For determining the presence and
location of infarcts, however, the 12-lead ECG remains
the standard. Because it is universally available, noninvasive, inexpensive and easily repeatable, it is important to know the extent to which the ECG could be
used for estimating infarct size.
Studies of the sequence of activation of both canine
and human hearts' 2 have provided the basis for computer simulations that suggest an orderly and predictable sequence of changes in the QRS complex associated with infarcts of various locations and sizes.37
Through pilot studies in patients with localized wall
motion abnormalities seen by ventriculography,
Selvester et al.8 refined the results of these computer
simulations to produce both qualitative and quantitative criteria for determining infarct size. This study
was designed to evaluate the specificity and reproduci-
Methods
The QRS Scoring System
The size of a myocardial infarct is estimated
electrocardiographically using this QRS scoring
system to accumulate a point score. Ten ECG leads (1,
II, aVL, aVF, V1,) are weighted according to their
ability to reflect infarct size as established by Selvester
et al.3 8 Certain criteria in the original scoring
system7' 8 are not evaluated. Qualitative criteria, such
as slurs or notches, are not evaluated because of the
difficulty of establishing definitions that can be applied
uniformly to clinically performed ECGs from a large
number of subjects. Criteria based on the absolute
amplitude of a deflection are not evaluated because of
the extent of variation of these amplitudes in the normal population." This leaves a simplified scoring
system with 37 criteria that can be applied to the 10
leads; these criteria are based upon the durations and
the ratios of the amplitudes of the QRS complex. The
criteria that apply to each lead are translated into
points. A maximum of 29 points can be accumulated
(table 1).
The primary observation in each lead is the duration of the initial Q or R wave, which is determined by
reference to the 40-msec time lines. The amplitude of
each deflection in the QRS is measured with handheld calipers.
When multiple criteria for either the duration or the
amplitude ratio are present in a single lead, only the
criterion that yields the most points is counted. For
example, if the Q wave in lead aVF is 50 msec, only the
3 points for the Q wave of 50 msec or greater are
From the Departments of Pathology, Medicine, and Community
and Family Medicine, Duke University Medical Center, Durham,
North Carolina, and Rancho Los Amigos Hospital, University of
Southern California, Downey, California.
Supported in part by contracts PH-43-NHLI-67-1440 and HRA230-76-0300, Health Resources Administration, DHHS; by
research grants HL-17670, HL-17667, HL-17682, HL-1 1310, HL14688, HL-17532 and MO-RROOO-13, NHLBI; by training grant
LM-07003 and grants LM 03373 and LM 00042, National Library
of Medicine; and by grants from the Prudential Insurance Company
of America and the Kaiser Family Foundation.
Dr. Ideker is recipient of Research Career Development Award
HL-00546, NHLBI.
Address for correspondence: Galen S. Wagner, M.D., P.O. Box
31211, Duke University Medical Center, Durham, North Carolina
27710.
Received October 8, 1980; revision accepted May 5, 1981.
Circulation 65, No. 2, 1982.
342
QRS SCORING SYSTEM/Wagner et al.
TABLE 1. Original QRS Scoring System
R duraQ duration
Lead (msec)
I
.30
tion
(msec)
Ratios
R/Q<1
II
aVL
.40
>30
> 30
R/Q.1
aVF
> 50
>40
.30
R/Q.1
R/QC2
V1
Any
.50
>40
R/S.1
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
V2
Any or < 20
>60
>50
V3
V4
Any or < 30
R/S>1.5
>20
R/QorR/S<0.5
R/QorR/S<1.5
V5
>30
R/QorR/S<1
R/QorR/S<3
V6
.30
R/Q or R/S < 1
R/Q or R/S < 3
Points
each Lead
crit. max.
1
1
2
2
1
2
1
1
2
3
2
1
2
1
1
2
1
1
1
2
1
1
1
1
2
1
1
2
1
1
2
1
5
4
4
1
3
3
3
counted; points for shorter Q waves in that lead are
disregarded.
An amplitude ratio criterion is not applied if the
ratio includes a Q or R wave that has not met a duration criterion in the same lead. For example, if the Q
wave in lead I is less than 30 msec, the criterion of Rto-Q amplitude in lead I is not considered. However,
an amplitude ratio that does not include a Q wave or
an initial R wave, such as the R-to-S amplitude ratio
in leads V,, V5 or V6, is considered regardless of
whether the duration criterion is satisfied in that lead.
Specificity
The three groups studied were selected because of
minimal likelihood of their having had myocardial infarcts. Using Hewlett-Packard three-channel recorders, standard 12-lead ECGs were obtained from 243
volunteers, as well as from 106 patients who underwent diagnostic cardiac catheterization. The volunteers included 146 males and females, ages 20-30
years, from Duke University Medical Center and 97
males, ages 45-55 years, from the University of
Southern California. They had no history suggesting
ischemic heart disease, diabetes mellitus or hypertension, and all had normal cardiovascular physical
examinations. The catheterized group was studied at
Duke University Medical Center for evaluation of
chest pain. Their mean age was 44 years (range 17-63
years). They had no evidence of coronary arterial
343
occlusive disease by selective angiography and had entirely normal left ventriculograms within 48- hours
after the ECG. Patients with evidence of left or right
ventricular hypertrophy, left or right bundle branch
block, or left anterior (frontal-plane axis > 450) or
posterior (frontal-plane axis . 1000) fascicular block
were not included."0
A QRS score was determined for each of the 349
subjects by adding the points for each criterion that
was satisfied. A specificity of at least 95% was arbitrarily established as the standard for the 37 criteria,
and any criterion with less than 95% specificity was
modified until it met this standard.
Observer Agreement
A standard 12-lead ECG was obtained from 50 patients 3 weeks after an acute myocardial infarct documented by elevated levels of isocreatine kinase. A
Hewlett-Packard three-channel recorder was used.
Each ECG was independently scored by three
observers - one trained technician and two cardiologists - who had experience applying the QRS scoring system. After an interval of at least 1 week, the
ECGs were given to each observer to be scored
another time. This format was followed initially using
photocopies of the ECGs and subsequently using the
original recordings.
Intra- and interobserver agreement of the overall
QRS scoring system was assessed by variance components analysis. Each of the 37 criteria was also examined independently. The percentage of cases in
which an observer agreed with his previous observation from the same ECG was determined for each
criterion. The results for each observer were combined
to determine overall intraobserver agreement.
To assess interobserver agreement, each reading by
one observer was compared with all other readings of
the same ECG by the other two observers. Thus, 12
pairs of readings were available for each criterion. The
percentage of cases in which one observer agreed with
another was determined for each criterion.
Results
Specificity
A total point score was determined for each of the
349 control subjects (fig. 1). The distribution of QRS
scores was similar in the volunteer and the
catheterized groups. Three hundred twenty-five of the
349 subjects (93%) scored 0 or 1 points. Seventeen
(5%) scored 2 points. Only seven of the 349 (2%)
scored more than 2 points. Only one subject scored 4
points.
The specificity of each of the criteria for Q- and Rwave duration and R/Q or R/S ratios was determined (table 2). Specificities of 29 of the 37 individual
criteria were either 99% or 100%. All of the criteria
using Q- and R-wave duration achieved the required
95% specificity level. All but two of the criteria for
amplitude ratios also achieved this standard. The
criterion of R/S . 1.5 in lead V4 achieved 92%
specificity and that of R/S < 3 in lead V5 achieved
VOL 65, No 2, FEBRUARY 1982
CIRCULATION
344
TABLE 2. Specificity of Original QRS Scoring System
100
Original
90
80
70
z 60
LU
50
40
0-1
30
20
10
II
C')
aVL
aVF
VI
>3
3
2
1
SCORE
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
100
100
100
99
97
99
100
100
97
100
100
99
100
97
100
100
95
100
99
98
99
95
99
<
0
FIGURE 1. The QRS scores ofall patients in the three control groups. The bars indicate percentages of the 243 subjects in the young noncatheterized group, the 97 subjects in
the middle-aged noncatheterized group, and the 106 subjects in the catheterized group who attained the indicated
QRS score.
V2
>
V3
V4
92%. More restrictive criteria for these amplitude
ratios were therefore required to satisfy the 95%
specificity requirements. Use of R/S < 1.0 in lead V4
and R/S < 2.0 in lead V5 yielded specificities of 98%
and 97%.
Table 3 shows the modified 29-point QRS scoring
system. A specificity of at least 95% was achieved for
each of the 37 criteria. The single-point R/Q criteria
in leads V4 and V5 were modified, in addition to the
single-point R/S criteria, to simplify the scoring
100
100
100
<
V5
92
99
100
100
<
V6
system.
100
92
99
100
100
<
Observer Agreement
The range of scores awarded to the 50 ECGs in the
observer agreement test set was 0-17. The performances of the observers did not differ significantly when
evaluating the photocopied and the original sets of
ECGs. Therefore, only the performances on the
Is
Specificity
(%)
criteria
Q>30
R/Q < 1
Q>40
Q >30
Q >30
R/Q < 1
Q >50
Q >40
Q >30
R/Q < 1
R/Q 2
AnyQ
R 50
R >40
R/S >1
Any Q
R<20
R>60
R > 50
R/S 1.5
Any Q
R<30
Q >20
R/Q < 0.5
R/S < 0.5
R/Q 1.5
R/S < 1.5
Q >30
R/Q 1
R/S < 1
R/Q < 3
R/S < 3
Q>30
R/Q ' 1
R/S 1
R/Q < 3
R/S < 3
Lead
I
- OCSERVER
18
#1
-
OSBERVER
100
99
original ECGs are presented. Variance components
analysis yielded the following estimates of variability
of overall scores: SD of scores within observers
0.594; and SD of scores between observers 0.624.
=
=
18 -
# 2
OBUR
${ 3
16
z_ 14
Lc
CY 12
:
:
~~~~~
~
.
16
16
14
14
12
12
10
10
,. .
_
<
a
.
is
;
.
6
6
LU
4
4
O
4
C-)
C',
2
0
2
2
_
i
0
I
0
2
4
6
8
1214
i63i
0
, ,
-s
2
I
4
6
8
.
. .
10
I
12
. .
0
16
-3
0 2 4 6 8
t2
163
SCORE OF SECOND READI NG
FIGURE 2. The agreement in QRS score between first and second readings of each of the three observers. Each observer
evaluated the same 50 ECGs during both readings. Each dot indicates one pair of readings. The numbers 11 for observer I, Sfor
2, and 13 for observer 3 are used in place of individual dots because of the large number of times O points were scored on each of
the readings.
QRS SCORING SYSTEM/ Wagner et al.
345
I1
18
18
16
16
16
14
# 14
CM 14
12
LL
12
L
10
Le'
10
co
u
a
#
C.)
cn
Cd,
co
Co
6
C>
co
.9-
8
4
9.II
2 0
0
al
2
b
* :i
4
i
0
12
CR
OBERE
O*
.
1
6
*
12
CAO10
Co
12
8
cw
6
_..
.
.
10
1
a
2
4
6
a
I
8
10
I
1
I
I
I
14
0
I1
16
18
SCORE OF OBSERVER # 3
_9....
0
.
..
.
2
-13
0
.
.
C.,
V), 4
.
..
4
2
La-
...
I
2
I
4
a
I
6
I
*
S
a
.
10
12
14
I
I
16
A
18
SCORE OF OBSERVER # 3
FIGURE 3. The QRS scores assigned to each of the 50 ECGs during the first reading by the three observers. Each dot indithe scoring of one of the 50 ECGs by the two observers. The numbers 9, 13 and 9 are used in place of individual dots because of the large number of time 0 points were awarded by each of the observers.
cates
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
The agreements within and between observers in estimating overall scores are shown in figures 2 and 3.
The mean intraobserver agreements for the 37 criteria by the three scorers were 99%, 97% and 97%.
The intra- and interobserver agreements. using each
of the 37 criteria are presented in table 4. Each of the
criteria achieved greater than 90% intraobserver
agreement and greater than 91% interobserver agreement.
Discussion
The ECG has become an essential element in the
diagnosis and management of patients with ischemic
heart disease." Early validation of its value in predicting the presence of coronary artery disease and for
diagnosing, localizing and sizing myocardial infarcts
depended on experimental animal studies and postTABLE 3. Modified QRS Scoring System
Duration
Amplitude
ratios
Lead
(msec)
>
I
Q 30
(1)
R/Q < 1
II
Q>40
(2)
Q>30
(1)
Q>30
aVL
(1)
R/Q c 1
Q>50
(3)
R/Q. 1
Q>40
aVF
(2)
Q>30
(1)
R/Q<2
Any Q
(1)
R>50
(2)
V,
R>40
R/S > 1
(1)
AnyQorR'20 (1)
R ' 60
V2
(2)
R>50
R/S > 1.5
(1)
V3
AnyQ orR<30 (1)
Q>20
V4
(1) R/QorR/S<0.5
Max
points
(1) 2
(1)
(2)
2
2
(1)
5
(1)
4
(1)
(2)
(1)
(2)
(1)
(2)
(1)
4
1
3
R/QorR/S<1
(1) R/QorR/S<1
3
R/Q or R/S < 2
Q>30
V6
(1) R/QorR/S<1
3
R/QorR/S<3
The number of points awarded for meeting each criterion
is in parentheses.
V5
Q>30
mortem examinations.'2 Because selective coronary
artery cineangiography is widely available for de-
termining the
coronary
presence or
absence of obstructive
artery disease, recent reviews have tested the
sensitivity and specificity of the ECG against catheterization findings.
In a review of 1000 patients who underwent coronary arteriography at the Cleveland Clinic Foundation, Proudfit et al.'5 demonstrated that almost all patients with QRS changes diagnostic of a myocardial
infarction also had significant coronary artery
obstruction (174 of 176; 99%).15 This association was
less striking in patients with nondiagnostic QRS
changes (34 of 50; 68%) and in patients with only STsegment and T-wave changes (six of 19; 32%). Many
other investigators have tested the sensitivity and
specificity of the routine ECG.16' 17 Although certain
limitations have been demonstrated, it is remarkably
useful.
Although the relationship among the ECG,
myocardial infarction, coronary anatomy, and myocardial function is well established, the strength of this
relationship requires further definition. In 1977, Awan
et al."8 studied 48 patients within 3 months of acute
myocardial infarction. All patients had Q-wave mapping with a 35-lead precordial ECG blanket performed within 24 hours of cardiac catheterization. The
number of pathologic Q waves (2 0.04 second) detected by the 35-lead precordial blanket was designated the Q index. The Q index correlated well with
hemodynamic variables predictive of myocardial infarct size.
Askenazi et al.'9 attempted to correlate the number
of Q waves in the 12-lead ECG with indexes of left
ventricular function derived from catheterization.
They studied 73 patients with coronary artery disease,
60 of whom had abnormal Q waves. Although the
number of Q waves alone did not reliably predict left
ventricular function, the sum of the R-wave amplitudes calculated from the six precordial leads and two
augmented leads correlated with left ventricular ejection fraction. Ideker et al.20 showed that the ejection
fraction correlates inversely with the size of old,
346
CIRCULATION
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
TABLE 4. Intra- and Interobserver Agreement Using Modified QRS Scoring System
InterIntraobserver
observer
agreement
agreement
Modified
criteria
Lead
(%)
%)
I
97
Q.30
99
99
99
R/Q ' 1
99
Q ' 40
99
II
92
93
Q.30
95
96
Q>30
aVL
96
97
R/Q 1
96
98
Q ' 50
aVF
94
97
Q.40
92
94
Q .30
98
98
R/Q ' 1
97
97
R/Q'2
92
91
AnyQ
V,
99
R .50
99
95
97
R'40
99
99
R/S .1
98
97
Any Q
V2
92
94
R'20
99
99
R.60
95
97
R.50
99
99
R/S ' 1.5
97
97
Any Q
V3
95
R<30
97
98
99
Q'20
V4
98
99
R/Q'0.5
97
97
R/S ' 0.5
100
100
R/Q ' 1
99
98
R/S'1
97
97
Q'30
V5
99
99
R/Q 1
96
96
R/S ' 1
99
99
R/Q .2
98
97
R/S'2
98
98
Q'30
V6
100
100
R/Q .1
99
99
R/S'1
99
99
R/Q .3
94
97
R/S '3
fibrotic myocardial infarcts. Thus, R waves may contain information about infarct size.
Using another approach to estimate infarct size
from the ECG, Selvester and co-workers' developed a
QRS scoring system from knowledge of the normal
sequence of ventricular depolarization in dogs. They
first used this information to develop a computer
simulation of normal human ventricular activation
using 20 dipoles to represent various segments of the
heart (seven for the septum, nine for the left ventricle,
and four for the right ventricle.)3 Four electrodes were
placed around this "heart" to act as a vectorcardiographic (VCG) recording system. When normal activation was simulated, the model produced tracings
resembling normal VCGs; by increasing the strengths
of dipoles representing the left or right ventricular segments, the model could produce VCGs resembling left
or right ventricular hypertrophy. Selvester and associates4 5 found that VCGs resembling those seen after
infarction could be produced by eliminating or
decreasing the strengths of one or more of the 20
VOL 65, No 2, FEBRUARY 1982
dipoles to simulate the loss of electrically active muscle in an old myocardial infarct. When as little as 0.1%
of the left ventricle was eliminated from the production of the signal, notches or bites were produced in
the VCG. This model was subsequently expanded to
generate the standard 12-lead ECG and the body surface map in a simulated male human torso.6 Criteria
for the quantitation of myocardial infarcts from the
ECG were developed from this model.7
Durrer et al.2 reported the normal activation sequences of seven isolated, perfused hearts obtained
from patients who died from cerebral destructive
processes. The sequences were similar to those in the
dog. Selvester and co-workers made use of this activation sequence to develop the model from which the
QRS scoring criteria evolved. Further refinement of
the criteria was based on the location, degree and extent of akinesis in biplane ventriculograms from 100
patients with documented infarcts and proximal coronary occlusion.
The present study shows that the simplified
Selvester QRS scoring system achieves an acceptable
level of specificity when confounding variables such
as left or right ventricular hypertrophy, left or right
bundle branch block, or left anterior or posterior
fascicular block are not evident on the ECG. Each
criterion exhibited at least 95% specificity and the
total 29-point scoring system achieved 98% specificity
when a score of more than 2 points was required for
identification of infarction. The confounding variables noted above will probably diminish the specificity, thus limiting the value of this scoring system for
identifying and sizing infarcts in some patients.
However, the specificity must be established. Certain
of the criteria may retain their importance in the
presence of other factors that can alter the QRS complex.
The QRS scoring system also permits acceptable
levels of both intra- and interobserver agreement.
Each of the 37 criteria achieved 91 % or greater intraobserver agreement and 92% or greater interobserver
agreement. The criteria that were associated with the
lowest level of agreement were those that required the
identification of any Q wave (V,) or of 20- or 30-msec
Q or R waves. Appearance of a minimal initial positive deflection in lead V1 might be considered baseline
noise by one observer and a true R wave by another.
Accurate identification of 20- or 30-msec deflections
may be somewhat more difficult than identification of
40-msec Q or R waves because of the 40-msec time
lines on standard ECG paper.
Machines that record three leads simultaneously
facilitate measurements by allowing reference among
leads for identifying the onset of Q and R waves. Scoring can be reproducibly performed even when a photocopy of the original tracing is used. However,
reference to the calibration signal must indicate that
there has not been excessive filtration of the signal obscuring the wave forms. Accurate measurement would
be facilitated if paper speed were increased to 50
mm/sec rather than 25 mm/sec. However, the level of
agreement achieved in the present study suggests that
QRS SCORING SYSTEM/Wagner et al.
standard electrocardiographic techniques provide
satisfactory results.
The value of the QRS scoring system depends on its
degree of sensitivity and on how well it can estimate
the size of myocardial infarcts. These questions are
being investigated. Studies should also be performed
to evaluate sequential changes in the QRS complex
occurring on serial ECGs during the early phases of
acute infarction and later during the period of gradual resolution of these changes.
In conclusion, the QRS scoring system is specific.
Even Q waves as small as 20 msec in some leads and
30 msec in other leads of a standard ECG are rarely
present in normal controls. Both intra- and interobserver agreements are acceptable.
Acknowledgment
The authors express appreciation to Gail McKinnis for preparation of the manuscript.
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
References
1.
2.
3.
4.
5.
6.
7.
Scher AM, Young AC: The pathway of ventricular depolarization in the dog. Circ Res 4: 461, 1956
Durrer D, van Dam RT, Freud GE, Janse MJ, Meijler FL, Arzbaecher RC: Total excitation of the isolated human heart. Circulation 41: 899, 1970
Selvester RH, Collier CR, Pearson RB: Analog computer
model of the vectorcardiogram. Circulation 31: 45, 1965
Selvester RH, Kalaba R, Collier CR, Bellman R, Kagiwada H:
A digital computer model of the vectorcardiogram with distance and boundary effects: simulated myocardial infarction.
Am Heart J 74: 792, 1967
Selvester RH, Kalaba R, Bellman R, Kagiwada H, Collier CR:
Simulated myocardial infarction with a mathematical model of
the heart containing distance and boundary effects. Proc Long
Island Jewish Hosp Symposium on Vectorcardiography, 1965,
p 403
Selvester RH, Solomon JC, Gillespie TL: Digital computer
model of a total body electrocardiographic surface map. Circulation 38: 684, 1968
Selvester RH, Wagner JO, Rubin HB: Quantitation of myocardial infarct size and location by electrocardiogram and vectorcardiogram. In Boerhave Course in Quantitation in Cardiology, edited by Snellin HA. The Netherlands, Leyden
University Press, 1972, p 31
347
8. Selvester RH, Solomon J, Sapoznikov D: Computer simulation of the electrocardiogram. In Computer Techniques in Cardiology, edited by Cady LD. New York, Marcel Dekker, 1979,
p 417
9. Simonson E: Differentiation Between Normal and Abnormal in
Electrocardiography. St Louis, CV Mosby, 1961
10. Bonner RE, Crevasse JL, Ferrer MI, Greenfield JC: A new
complete program for comparative analysis of serial scalar electrocardiograms and performance of the 1976 IBM program.
Comput Biomed Res 11: 103, 1978
11. Castellanos A, Myerburg RJ: The resting electrocardiogram.
In The Heart, 4th ed, edited by Hurst JW. New York,
McGraw-Hill, 1978, p 298
12. Hillis LS, Askenazi J, Braunwald E, Radvany P, Muller JE,
Fishbein MC, Maroko PR: Use of changes in the epicardial
QRS complex to assess interventions which modify the extent
of myocardial necrosis following coronary artery occlusion.
Circulation 54: 591, 1976
13. Gunnar RM, Pietras RJ, Blackaller J, Dadmun SE, Szanto PB,
Tobin JR: Correlation of vectorcardiographic criteria for myocardial infarction with autopsy findings. Circulation 35: 158,
1967
14. Savage RM, Wagner GS, Ideker RE, Podolsky SA, Hackel
DB: Correlation of postmortem anatomic findings with electrocardiographic changes in patients with myocardial infarction: retrospective study of patients with typical anterior and
posterior infarcts. Circulation 55: 279, 1977
15. Proudfit WL, Shirey EK, Sones FM: Selective cine coronary
arteriography: correlation with clinical findings in 1,000 patients. Circulation 33: 901, 1966
16. Bodenheimer MM, Banka VS, Trout RG, Pasdar H, Helfant
RH: Correlation of pathologic Q waves on the standard electrocardiogram and the epicardial electrogram of the human heart.
Circulation 54: 213, 1976
17. Horan LG, Flowers NC: Diagnostic power of the Q wave:
critical assay of its significance in both detection and localization of myocardial deficit. In Advances in Electrocardiography, edited by Schlant RC, Hurst JW. New York, Grune &
Stratton, 1972, p 321
18. Awan NA, Miller RR, Vera Z, Janzen DA, Amsterdam EA,
Mason DT: Noninvasive assessment of cardiac function and
ventricular dyssynergy by precordial Q wave mapping in anterior myocardial infarction. Circulation 55: 833, 1977
19. Askenazi J, Parisi AF, Cohn PF, Freedman WB, Braunwald E:
Value of the QRS complex in assessing left ventricular ejection
fraction. Am J Cardiol 41: 494, 1978
20. Ideker RE, Behar VS, Wagner GS, Starr JW, Starmer CF, Lee
KL, Hackel DB: Evaluation of asynergy as an indicator of myocardial fibrosis. Circulation 57: 715, 1978
Evaluation of a QRS scoring system for estimating myocardial infarct size. I. Specificity
and observer agreement.
G S Wagner, C J Freye, S T Palmeri, S F Roark, N C Stack, R E Ideker, F E Harrell, Jr and R
H Selvester
Downloaded from http://circ.ahajournals.org/ by guest on July 28, 2017
Circulation. 1982;65:342-347
doi: 10.1161/01.CIR.65.2.342
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1982 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/65/2/342
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation is online at:
http://circ.ahajournals.org//subscriptions/