1615
Importance of Two-dimensional
Echocardiographic Assessment of Left
Ventricular Systolic Function in Patients
Presenting to the Emergency Room With
Cardiac-Related Symptoms
Peter Sabia, MD; Robert D. Abbott, PhD; Ali Afrookteh, MD; Mark W. Keller, MD;
Dale A. Touchstone, MD; and Sanjiv Kaul, MD
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Background. This prospective study was designed to test the hypothesis that the assessment
of left ventricular systolic function at the time of emergency room (ER) presentation provides
valuable diagnostic and prognostic information in patients with cardiac-related symptoms.
Methods and Results. The study is based on a 2-year follow-up of 171 consecutive patients
evaluated in the ER for such symptoms. In the course of follow-up, one third of the patients (55
of 171) suffered a major cardiac event. For those with left ventricular systolic dysfunction
(LVSD), the age-adjusted rate of early events (occurring within 48 hours of presentation) was
more than eight times higher than for those without LVSD (26.9% versus 3.3%, p<0.01). For
events occurring after 48 hours of ER presentation, LVSD was associated with a nearly fourfold
excess of cardiac events (23.9% versus 6.4%, p<0.01). Other than advanced age, the most
important confounder for early events included an abnormal electrocardiogram diagnostic for
acute myocardial infarction. Confounders for late events included advanced age and a history
of hypertension. LVSD on two-dimensional echocardiography (2DE) was the only finding
associated with early and late events after controlling for other risk factors. In addition, the
prediction of these events derived from the combination of historical, clinical, electrocardiographic, and 2DE findings was significantly improved when accounting for the presence or
absence of LVSD (p<0.01).
Conclusions. We conclude that the 2DE assessment of left ventricular systolic function
provides valuable diagnostic and prognostic information in subjects presenting to the ER with
cardiac-related symptoms. (Circulation 1991;84:1615-1624)
linical and electrocardiographic (ECG) variables are routinely used for evaluating
patients presenting to the emergency room
(ER) with cardiac-related symptoms.1-12 Whereas
such information has both diagnostic and prognostic information, the benefits of evaluating left ventricular systolic function in the ER are unknown.
We hypothesized that the assessment of left ventricular systolic function at the time of ER presen-
tation would provide diagnostic and prognostic
information above and beyond that available from
the historical, clinical, and ECG data. In this study,
left ventricular systolic function was assessed using
two-dimensional echocardiography (2DE), and the
relation between left ventricular systolic function
and cardiac events that occur during or soon after
presentation to the ER versus those that occur later
are described.
From the Divisions of Cardiology, Biostatistics, and Emergency
Medicine, Department of Medicine, University of Virginia School
of Medicine, Charlottesville, Va.
Presented in part at the 62nd Annual Scientific Sessions of the
American Heart Association, New Orleans, La., November 1989,
and in part at the 1st Annual Scientific Session of the American
Society of Echocardiography, Washington, D.C., June 1990.
Supported in part by Biomedical Research Development grant
5S07-RR-05431 from the National Institutes of Health, Bethesda,
Md. S.K. is the recipient of the Clinical Investigator (K08-HL01833) and FIRST (R29-HL-38345) Awards from the National
Institutes of Health.
Address for correspondence: Sanjiv Kaul, MD, Division of
Cardiology, Box 158, University of Virginia School of Medicine,
Charlottesville, VA 22908.
Received December 27, 1990; revision accepted June 4, 1991.
C
1616
Circulation Vol 84, No 4 October 1991
Methods
Patient Sample
One hundred eighty-five consecutive patients who
presented to the University of Virginia Medical
Center ER with cardiac-related symptoms and who
met the inclusion criteria were prospectively enrolled
in this study. Among this group there were 202 ER
visits. In the 14 patients with repeat ER visits, the
first was used as the index visit. Men and women
were considered eligible for the study if they were
older than 30 and 40 years, respectively, and if the
symptoms included chest pain or shortness of breath
that lasted for at least 30 minutes within 4 hours of
ER presentation. Subjects were not eligible for the
study if symptoms could be explained by noncardiac
causes, local trauma, or preexisting abnormalities on
chest x-ray such as skeletal or pulmonary pathology.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Protocol
The protocol was approved by the Human Investigation Committee of the University of Virginia
School of Medicine. All patients entered in the study
gave informed consent. At the time of presentation
to the ER, a member of the study team obtained a
detailed history and performed a complete physical
examination. In addition, a 12-lead electrocardiogram and serum enzyme measurements were obtained, and 2DE was performed.
All but one member of the study team had completed 3 years of internal medicine residency and all
but one had at least 3 months of training in 2DE. A
member of the study team was on call for the study 24
hours a day. The information obtained on 2DE was
not divulged to the ER physician, who made the
patient disposition based entirely on historical, clinical, and ECG data.
Variables Analyzed
The historical information used to evaluate the patients in this report included age, sex, and a history of
hypertension and cigarette smoking. The clinical data
included the presence of jugular venous distension,
third heart sound, and pulmonary rales. Three observers blinded to the historical and clinical data evaluated
the electrocardiograms and 2DE on separate occasions.
The electrocardiogram was classified as either normal or abnormal. An abnormal electrocardiogram
was characterized as having one or more of the
following specific abnormalities: ST elevation or depression more than 1.0 mm (without or with Q
waves), Q waves without ST or T wave changes, T
wave inversion more than 1.0 mm, nonspecific ST or
T wave changes, left ventricular hypertrophy, left or
right bundle branch block, and paced rhythm.13
The 2DE variables that were analyzed included
left and right ventricular and left atrial sizes and left
ventricular systolic function.14 For assessment of the
latter, the left ventricle was divided into 12 segments
(Figure 1) as described by us previously.15 Each
segment was classified in two ways: having either
MITRAL
VALVE
LEVEL
PAPILLARY
MUSCLE
LEVEL
APICAL
LEVEL
FIGURE 1. Method used for assessing total left ventricular
wall motion score. Segments are classified as follows: 1, 6,
1 l=anterior wall; 2, 7=lateral wall; 3, 8, 12=posterior/
inferior wall; 4, 9=inferior interventricular septum; and 5.
10=anterior interventricular septum.
normal or abnormal systolic function and assigning a
wall motion score. Abnormal systolic function was
considered to be present in a segment if it manifested
both abnormal motion and thickening. Based on the
degree of systolic dysfunction, the following scores
were also assigned to each segment: 1 for normal
systolic function; 2 for hypokinesia; 3 for akinesia;
and 4 for dyskinesia.15 The scores for each segment
were then added to determine the total left ventricular wall motion score. In this manner, a left ventricle
showing normal systolic function would be assigned a
wall motion score of 12. The worse the function, the
higher would be the score.
Patient Follow-up
Follow-up was obtained over a 2-year period by
contacting the patient, the patient's family and physician, and by monitoring the hospital admissions.
Study end points included cardiac-related death,
nonfatal myocardial infarction confirmed by new Q
waves or cardiac enzyme changes,'3'16"17 life-threatening arrhythmias requiring medical or surgical intervention, and intractable angina pectoris requiring
revascularization (either coronary artery bypass graft
surgery or percutaneous transluminal coronary angioplasty). Events that occurred within 48 hours of
ER presentation were defined as early events. Late
events included those that occurred after 48 hours of
ER presentation.
Statistical Methods
Since the exact time of onset of early cardiac events
could not be precisely ascertained within the 48-hour
period, logistic regression models were used to evaluate
the association between early outcomes and each ER
finding.'8 Use of the logistic models was further supported by the completeness of patient follow-up during
this brief period. Analysis for late events was based on
proportional hazards regression models,19 since event
times and lengths of follow-up could be easily distinguished among the study participants. All models were
evaluated for fit, and the results suggested that they
Sabia et al Value of Assessing Left Ventricular Systolic Function in Emergency Room Patients
1617
TABLE 1. Percentage of Subjects Presenting to ER With Cardiac-Related Symptoms Who Developed Early and Late
Cardiac Events
Within 48 hours
(n=171)*
(29)
17.0
(0)
0.0
(0)
0.0
1.8
(3)
(32)
18.8
Type of event
Nonfatal myocardial infarction
Cardiac-related death
Coronary revascularization
Serious arrhythmias
Total
Number of events in parentheses.
ER, emergency room.
*Subjects at risk.
tSubjects at risk after excluding those who had early events.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
were useful for conducting tests of significance concerning the data obtained in the ER.
Rates of early and late cardiac events were
adjusted for age and selected ER findings using Cox
models.2021 Unadjusted survival rates were also derived using Kaplan-Meier life-table analysis.22 In
addition, cross-sectional comparisons were made between the various ER findings using the logistic
models, and correlates of left ventricular systolic
dysfunction (LVSD) were examined. Age-adjusted
rates of each ER finding were derived for patients
without and with LVSD.
To further explore the independent effect of
LVSD, the value of the ER findings for determining
late and early events was evaluated in terms of their
additive importance in the order in which they would
normally be observed. The process began by examining the information from age alone, followed by the
added contribution from the historical and clinical
data, followed by the additional value of the electrocardiogram, and finally, by the additive information
provided by 2DE. Tests of significance were based on
increases in the likelihood ratio statistic due to the
sequentially added information.23 For practical purposes, the percent change in the likelihood ratio
statistic relates directly to the percent change in the
overall prognostic power. All differences were considered significant at p<0.05 and all tests of significance were two sided.
Results
Of the 185 patients originally entered into the
study, 12 were excluded from analysis because of
technically inadequate 2DE studies and two were
'excluded because of a complete loss to follow-up.
Among the remaining 171 patients who are described
in this report, there were 88 men and 83 women with
average ages of 59±14 and 62±14 years, respectively.
All Events
Table 1 presents the rates of early and late cardiac
events stratified by the type of event. In the 171
patients, 55 (32%) had adverse cardiac events within
the 2-year follow-up period: 32 had events within 48
hours of ER presentation and 23 had events after 48
Time after ER presentation
After 48 hours
(n=139)t
(11)
6.4
(10)
5.8
(2)
1.2
(0)
0.0
(23)
13.4
Total
(n=171)
(40)
23.4
(10)
5.8
(2)
1.2
(3)
1.8
(55)
32.2
hours of ER presentation. Early events were largely
composed of nonfatal myocardial infarctions (29 of
32). There were no deaths and no revascularization
procedures. Three subjects with early events experienced life-threatening arrhythmias (ventricular tachycardia, complete heart block, and Mobitz type II
block, respectively). Among the 23 late events, the
incidence rates of nonfatal myocardial infarction (11)
and death (10) were similar; two patients underwent
revascularization for intractable angina, and there
were no reports of arrhythmias.
Early Events
Table 2 presents the age-adjusted rates of events
occurring within 48 hours of ER presentation. There
were no significant correlates of early cardiac events
among the historical and clinical findings. An abnormal
electrocardiogram was associated with a more than
threefold excess of early events (p<0.05). Subjects with
ST elevation had a sixfold excess of early events as
compared to those without this finding (p<0.001). The
presence of LVSD was the only 2DE finding that had a
significant positive association with early cardiac events.
There was more than an eightfold excess of early events
among patients who presented in the ER with this
finding (p<0.01). Among the patients without ST
elevation, an age-adjusted sixfold excess of early events
was observed in the presence of LVSD as compared to
its absence (20.5% versus 3.4%,p<0.01).
Late Events
Table 3 provides the age-adjusted rates of events
occurring after 48 hours of ER presentation. Among
the historical data, subjects with a history of hypertension had a ninefold excess of cardiac events as compared to those without such a history (p<0.001). Jugular venous distension on physical examination was
associated with more than a fourfold excess of cardiac
events (p<0.001), whereas a threefold excess was observed in those with a third heart sound (p<0.01).
Patients with an abnormal electrocardiogram experienced late cardiac events at a rate nearly three times
higher than those with a normal electrocardiogram
(p<0.05). Patients with ECG evidence of left ventric-
1618
Circulation Vol 84, No 4 October 1991
TABLE 2. Age-Adjusted Percentage of Subjects Who Developed Cardiac Events Within 48 Hours of ER Presentation
ER finding
Finding
Absent
Present
Historical and clinical findings
History of hypertension
18.8
18.7
(14/74)
(18/97)
History of smoking
21.5
15.9
(19/79)
(13/92)
Jugular venous distension
19.3
15.0
(28/153)
(4/18)
Third heart sound
17.5
29.9
(27/157)
(5/14)
Pulmonary rales
17.2
21.4
(18/117)
(14/54)
Electrocardiographic findings
Normal electrocardiogram
23.6*
6.4
(29/114)
(3/57)
ST segment elevation
14.2
(23/161)
89.Ot
(9/10)
Q waves
18.8
17.8
(29/156)
(3/15)
Nonspecific ST changes
18.7
18.5
(29/153)
(3/18)
LV hypertrophy
21.0
4.1
(31/153)
(1/18)
Left bundle branch block
17.7
27.4
(27/157)
(5/14)
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Two-dimensional echocardiographic findings
LV enlargement
18.6
LV hypertrophy
22.8*
RV enlargement
20.5
Left atrial enlargement
19.6
LV systolic dysfunction
3.3
Number of events/subjects at risk in parentheses.
ER, emergency room; LV, left ventricular; RV, right ventricular.
Significant excess of cardiac events: *p<0.05, tp<0.001, tp<0.01.
ular hypertrophy also experienced a threefold excess of
events (p<O.O1). The presence of abnormal 2DE findings was associated with an excess of late cardiac
events. Although not statistically significant, the rate of
events was more than doubled in the presence of left
atrial and left ventricular enlargement. The presence of
LVSD was associated with a nearly fourfold excess of
late events as compared to its absence (p<O.O1). This
effect remained unaltered after removing subjects with
left ventricular hypertrophy.
Independent Effect of LVSD
In Table 4, age-adjusted rates of various findings
are given for subjects without and with LVSD. Patients with the latter finding were significantly more
likely to have a third heart sound (p<O.O1) and
pulmonary rales (p<0.05). An abnormal electrocardiogram (p<O.OO1), ST elevation (p<O.OO1), and left
bundle branch block (p<O.O1) also occurred with
greater frequency in the presence of LVSD. Patients
with LVSD were also more likely to have enlargement of the left and right ventricles (p<O0.01) and the
left atrium (p<O.001) on 2DE.
Based on the associations among the ER findings
and each cardiac event, the independent effect of
LVSD on early and late outcomes was examined after
adjusting for selected data. For early events, the most
important confounders of LVSD included age and an
abnormal electrocardiogram. After adjustment for
these factors, the association between LVSD and early
cardiac events was diminished slightly, but remained
highly significant (p<O.O1) (Figure 2A).
(24/133)
(29/135)
(28/138)
(23/127)
(2/64)
19.0
6.4
11.7
16.74:
26.9
(8/38)
(3/36)
(4/33)
(9/44)
(30/107)
Figure 2B illustrates the relation between early
events and the degree of LVSD, as exemplified by the
left ventricular wall motion score. Whereas the event
rate increased significantly in patients with a wall
motion score of 13-24 compared to those with a score
of 12, a further increase in events was not noted in
patients with a score of more than 24. The results did
not change appreciably after adjusting for age and
presence of abnormal electrocardiogram.
or events that occurred after 48 hours of ER presentation, the most important confounders affecting the
relation between LVSD and late cardiac outcomes
included age and a history of hypertension. After
adjusting for these factors, 91.5% of those with normal
ventricular systolic function were free of a cardiac event
at the end of 2 years of follow-up as compared to 72.6%
of those who exhibited LVSD (p<O.O1). Unadjusted
and adjusted event-free survival based on the presence
or absence of LVSD is depicted in Figure 3.
Figure 4A illustrates unadjusted late event-free
survival based on the degree of LVSD as exemplified
by the left ventricular wall motion score. The higher
the observed score, the greater seems the likelihood
of suffering an adverse cardiac event. When eventfree survival in these patients with late events was
adjusted for the confounding effects of age and
hypertension, event rates among subjects with different degrees of LVSD became similar (Figure 4B),
although differences between those with and without
LVSD continued to remain highly significant
(p<O.Ol).
Sabia et al Value of Assessing Left Ventricular Systolic Function in Emergency Room Patients
1619
TABLE 3. Age-Adjusted Percentage of Subjects Who Developed Cardiac Events After 48 Hours of ER Presentation
ER finding
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Absent
Finding
Historical and clinical findings
3.0
(2/60)
History of hypertension
(13/60)
17.7
History of smoking
(14/125)
12.0
Jugular venous distension
(18/130)
14.1
Third heart sound
(12/99)
14.0
Pulmonary rales
Electrocardiographic findings
20.6t
Normal electrocardiogram
(20/85)
...
(o/o)
ST segment elevation*
(22/127)
17.8
Q waves
(22/124)
17.5
Nonspecific ST changes
(14/122)
12.2
LV hypertrophy
(20/130)
15.8
Left bundle branch block
Two-dimensional echocardiographic findings
(13/109)
12.2
LV enlargement
(14/106)
14.8
LV hypertrophy
(17/110)
15.3
RV enlargement
(12/104)
12.6
Left atrial enlargement
6.4
(4/62)
LV systolic dysfunction
in
at
risk
parentheses.
Number of events/subjects
ER, emergency room; LV, left ventricular; RV, right ventricular.
*Insufficient data.
Significant excess of cardiac events: tp<0.05, tp<0.001, §p<0.01.
Additive Prognostic Value of LVSD
The increases in the likelihood ratio statistic due
to the sequential addition of the ER findings in the
Present
26.8§
15.5
(21/79)
(10/79)
50.6*:
(9/14)
(5/9)
(11/40)
46.9§
21.0
(3/54)
(1/1)
(1/12)
(1/15)
(9/17)
7.0
...
6.2
8.0
41.4§
25.1
30.6
20.8
20.9
26.1
23.9§
(3/9)
(10/30)
(9/33)
(6/29)
(11/35)
(19/77)
order in which they would normally be observed
(age, historical and clinical, ECG, and 2DE) are
presented in Figure 5. For early events (Figure 5A),
TABLE 4. Age-Adjusted Percentage of Subjects With and Without LVSD at Time of ER Presentation
LVSD
Present
Absent
(n= 107)
(n=64)
Finding
Historical and clinical findings
57.0
56.3
(36)
History of hypertension
55.5
50.8
(34)
History of smoking
8.4
11.6t
(5)
Jugular venous distension
12.7
0.0
(0)
Third heart sound
36.5
22.2
(13)
Pulmonary rales
Electrocardiographic findings
Normal electrocardiogram
23.3
49.7*
(34)
9.6*
0.0
(0)
ST segment elevation
11.9
3.2
(2)
Q waves
7.5
15.5
(10)
Nonspecific ST changes
13.4
5.0
(3)
LV hypertrophy
11.61
1.6
(1)
Left bundle branch block
Two-dimensional echocardiographic findings
0.0
35.2i:*
(0)
LV enlargement
20.7
21.7
(13)
LV hypertrophy
4.7
27.9t
(3)
RV enlargement
6.7
(4)
Left atrial enlargement
35.9**
Number of subjects with the finding in parentheses.
LVSD, left ventricular systolic dysfunction; ER, emergency room; LV, left ventricular; RV, right ventricular.
Significant excess of cardiac events: *p<0.001, tp<0.01, tp<0.05.
(61)
(58)
(13)
(14)
(41)
(23)
(10)
(13)
(8)
(15)
(13)
(38)
(23)
(30)
(40)
Circulation Vol 84, No 4 October 1991
1620
A
a)
I-
Z
W
B
40T
30
H
40T
In
1+
F-
Z
W
30-
H
I.-
*@
0.4
ICL-9
4
0
0
20
'4
20
'Li
I-
'I
I..
Z-1
W
WL
Z 10
10.
ol
PRESENT
ABSENT
O0
LEFT VENTRICULAR
DYSFUNCTION
_
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
E
.I..
12
FIGURE 2. Bar graphs showing event rates
within 48 hours ofpresentation to the emergency
room admission for those with and without left
ventricular systolic dysfunction (panel A), and
for those with different degrees of left ventricular
systolic dysfunction (panel B). Both unadjusted
rates and rates adjusted for age and the presence
of an abnormal electrocardiogram (EKG) are
depicted. *p<0.01 compared to no left ventricular systolic dysfunction and left ventricular wall
motion score of 12.
LEFT VENTRICULAR
WALL MOTION SCORE
UNADJUSTED
ADJUSTED FOR AGE AND ABNORMAL EKG
age made a significant contribution to describing
prognosis (p<O.O1). Older subjects had a higher
risk of adverse cardiac outcomes than those who
were younger. The additional prognostic value of
other historical and clinical information was not
apparent in this subset of patients. The presence of
any rather than a specific abnormality on the electrocardiogram was the most important piece of
information derived from the electrocardiogram
after accounting for age and the clinical findings
(p<0.O1). Presence of LVSD was the most useful
2DE finding, nearly doubling the value of the
_j
J
H
mcc
(0)
W
W
ar
U.
likelihood ratio statistic obtained from age, clinical,
and the EKG data alone (p<O.O1).
For late events (Figure 5B), age had the same effect
on cardiac outcomes as it did for early events
(p<O.OOl). The most useful additional historical or
clinical finding in determining prognosis for late events
included a history of hypertension. After adjusting for
age, there was more than a twofold increase in the
value of the likelihood ratio statistic when hypertension
was included in the model (p<0.001). In contrast,
findings from the electrocardiogram had no marked
prognostic value. Adding LVSD increased the likelihood ratio statistic obtained from age, historical, clinical, and the ECG variables by 30% (p<0.01). Other
2DE findings failed to provide significant additional
information in predicting late events. Separate and
combined adjustment for the other ER findings failed
to alter the preceding results.
-
z
W
W
ILL
0
H
-J
m
m
to
¢
TIME (months)
UNADJUSTED
*- -----.-
ADJUSTED FOR AGE AND HISTORY OF HYPERTENSION
FIGURE 3. Graph showingprobability of surviving a cardiac
event after 48 hours from emergency room presentation for
those without and with left ventricular systolic dysfunction.
Both unadjusted event-free survival and survival data adjusted
for age and a history of hypertension are depicted. Both
unadjusted and adjusted survival rates are significantly different between patients with and without left ventricular (LV)
systolic dysfunction (p<0.01).
Discussion
Our results indicate that one third of all patients
presenting to the ER with symptoms suggestive of
cardiac etiology suffered a major cardiac event within
the ensuing two years. In addition, for those with
normal left ventricular systolic function, less than
10% had cardiac events. In those with LVSD, nearly
half suffered an event. Only one patient with normal
left ventricular systolic function and a normal electrocardiogram had an event in the 2-year follow-up
period. Regardless of the time an event was observed, the direct assessment of left ventricular function in the ER added significantly to the diagnostic
and prognostic value of age and the historical, clinical, and ECG findings. These results imply that, in
addition to routine clinical and ECG evaluation, left
ventricular systolic function should also be assessed
in patients presenting to the ER with cardiac-related
symptoms.
Sabia et al Value of Assessing Left Ventricular Systolic Function in Emergency Room Patients
A
1.0- .*i
B
12
I.0.
1621
12
.....................................
..
0 .9
0.9
13-24
J
13-24
J
z.4
111
U)
cc
>24
>24
L'
IL 0.7
(-
z
W
WU
0.7
z
W
W
0.6+
0.6
o.50
O.oDm_ 01Ew
-
.
12
8
4
16
24
20
|
|
l
12
4
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
..........
UNADJUSTED
-
ADJUSTED FOR AGE AND HISTORY OF HYPERTENSION
Early Events
Of the 32 patients with early events, 29 suffered
an acute myocardial infarction. No historical or
clinical variable was useful in identifying these
patients. Of the ECG abnormalities, ST elevation
(without or with Q waves) was noted in only nine of
these patients. The other 20 had either nonspecific
findings, or evidence of left bundle branch block or
paced rhythm, in the presence of which acute
myocardial infarction could not be diagnosed. In
contrast, only two patients with acute myocardial
infarction failed to have LVSD. Both had non-Q
wave infarction and neither suffered complications.
All the 13 in-hospital complications (life-threatening arrhythmias, postinfarction angina, and cardio-
24
genic shock) occurred in patients with LVSD on
2DE. Only in four such patients was ST elevation
noted on the initial electrocardiogram performed in
the ER.
It should not be surprising that the assessment of
left ventricular systolic function doubled the prognostic value inherent in the combined clinical and ECG
data. Regional LVSD occurs within seconds after
coronary occlusion and may appear several hours
prior to ECG abnormalities.24-26 Because most of the
early events in our study were ischemic, abnormalities on 2DE would be expected to be more sensitive
than would those on electrocardiography.
In terms of therapeutic intervention, it is becoming
increasingly clear that thrombolytic agents should be
B) EVENTS AFTER 4B HOURS
A) EVENTS WITHIN 48 HOURS
50T
507
40.
400
0
X
X
FIGURE 5. Bar graphs showing comparison of additional prognostic value of tests (depicted as increase in
likelihood ratio statistic) performed in succession for
determining early (panel A) and late (panel B) events.
Age alone is compared to age and clinical variables;
age, clinical, and electrocardiographic (EKG) variables; and age, clinical, EKG, and two-dimensional
echocardiographic (2DE) variables. Significant increases in the likelihood ratio statistic correspond to the
following p values: *p<0.05, **p<0.01, and
1.4
o
_I
W
m "- 30i
W L
F
30-
U14
WI-
J
20
z
W
20
TIME (months)
TIME (months)
m
16
FIGURE 4. Graphs showing probability
of surviving a cardiac event after 48 hours
of presentation to the emergency room
based on the degree of left ventricular
systolic dysfunction. Degree of left ventricular systolic dysfunction is based on left
ventricular wall motion scores of 12, 1324, and >24. Unadjusted survival curves
(panel A) and survival data adjusted for
age and a history of hypertension (panel
B) are depicted. The unadjusted rates for
scores of 13-24 and >24 are different at a
p=0. 02, while the adjusted rates for these
scores are not different.
W
m
_
U)49
20'
0
W
10
*
Z
10'
l
_0
ORE _FADD
NOMTO
o0
K**
ORDER OF ADDED INFORMATION
AME
AOE + CLICAL
-s ASE
CLIICAL
+
E
***p<0.001.
1622
Circulation Vol 84, No 4 October 1991
administered as early as possible in the evolution of
acute myocardial infarction.27-31 A means of early
diagnosis of this condition would, therefore, be useful. Our results suggest that 2DE could be of value in
the diagnosis of acute myocardial infarction, particularly in those in whom the ECG changes have not
yet evolved at the time of ER presentation. These
findings are in agreement with previous studies examining the role of 2DE in the ER diagnosis of acute
myocardial infarction.32-34 It is also possible that in
such patients, the extent of LVSD could reflect the
size of the myocardium at risk,35,36 and thus assist the
physician in evaluating the risks and benefits of
thrombolytic therapy.32
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Late Events
Of identical concern are the patients in whom
events occurred after 48 hours of ER presentation.
Among the 139 subjects not experiencing early
events, 23 (17%) had adverse outcomes in the
two-year follow-up. Unlike those experiencing
events within 48 hours, these patients were as likely
to die as they were to suffer a nonfatal infarction. In
addition, the historical and clinical profiles of these
patients were different from those suffering early
events. Victims of late events were more likely to
have a history of hypertension and signs of congestive heart failure, such as jugular venous distension
and a third heart sound. Apart from an increased
incidence of left ventricular hypertrophy, the other
ECG abnormalities in such patients were nonspecific. Despite the presence of congestive heart
failure, which is a known prognosticator of adverse
outcomes, LVSD on 2DE provided significant additional prognostic information in these patients.
The results were similar for the patients who died
and those who suffered an ischemic event (nonfatal
infarction or revascularization).
Unfortunately, the evaluation of patients in the
ER has been largely limited to stratifying them into
high- and low-risk categories for acute events.9-12
The long-term prognosis has not been evaluated in
these patients. This prospectively designed study
suggests that the event rate in such patients is high,
and that identifying LVSD at the time of ER presentation is useful in recognizing those who are more
likely to suffer an event. For instance, this finding
may be a more important marker of death than
clinical and ECG abnormalities in patients with
myocardial disorders such as ischemic or hypertensive cardiomyopathy. Similarly, in the 13 patients
who suffered an ischemic event after 48 hours of ER
presentation (11 with myocardial infarction and two
undergoing revascularization), the presence of
LVSD at the time of the ER visit may have reflected
an acute ischemic episode that preceded the actual
infarction or a need for revascularization.
Importance of the Degree of LVSD
Whereas the presence of LVSD was important in
identifying early and late events, the degree of dysfunc-
tion did not seem to be as important. Most of the early
events (29 of 32) were acute myocardial infarctions.
The presence of regional dyssynergy identifies the
occurrence of an infarction. The number of segments
exhibiting this finding does not necessarily add to the
diagnosis of acute myocardial infarction. Consequently,
one would not expect the degree of LVSD to be
important in determining the occurrence of an infarction in these patients, although other events such as
death may be related to the degree of LVSD.
In contrast to early events, one would have expected some relation between the occurrence of
events and the degree of LVSD in patients experiencing late events who had clinical and other evidence of heart failure. Unadjusted event rates seem
to support this expectation. When data are adjusted
for the confounding effects of age and history of
hypertension, however, the degree of LVSD (among
those with this finding) no longer provides independent predictive power.
The reason for a loss of the prognostic power of
the degree of LVSD in predicting late events once
adjustment for advanced age and a history of hypertension is made is not readily apparent from our data.
Among these patients, those with left ventricular wall
motion score of more than 24 were significantly older
(p<0.05) than those with lower wall motion scores
despite having the same prevalence of hypertension.
As such, controlling for age and hypertension may
have been similar to controlling for duration of
hypertension. As a consequence, subjects with wall
motion scores of more than 24 may have been at
greater risk of late events than those whose scores
were lower (>12 but .24), because the higher scores
may have been more often associated with older ages
and a longer duration of hypertension.
Value of Other ER Findings
Although the presence or absence of LVSD was
clearly related to early and late cardiac events in this
patient sample, the value of routinely measured ER
findings continues to be important. Our results indicate that, whereas historical and clinical data have
questionable value in predicting early events, they
provide important information in terms of late
events. Similarly, while the electrocardiogram is very
useful for predicting early events, it has lesser value
for late events. The presence of LVSD, however,
provides diagnostic and prognostic information for
both early and late events over and above that
already provided by age and the historical, clinical,
and EKG data routinely observed in patients presenting to the ER with cardiac-related symptoms.
Logistics of Performing 2DE in the ER
Although it was not the purpose of this study to
assess the logistics of performing 2DE in the ER, it
may be worth discussing some of the concerns related
to this issue. At the time this study was performed,
there were approximately 42,000 annual visits to our
ER, of whom about two thirds were nonsurgical. Of
Sabia et al Value of Assessing Left Ventricular Systolic Function in Emergency Room Patients
the latter group, approximately 10% constituted patients who complained of chest pain. Of these patients, no more than five or six met our inclusion
criteria over a 24-hour period. Despite being on call
24 hours at a time, therefore, members of the study
team did not have to perform 2DE more than five or
six times a day.
To identify patients with LVSD, physicians trained
in performing and interpreting 2DE studies would be
required to evaluate patients presenting to the ER
with cardiac-related symptoms. This service could be
provided by either a cardiologist or an ER physician
trained in selected aspects of 2DE. Which of these
two options is more practical and cost-effective needs
to be determined. It is likely that the option chosen
will depend as much on local hospital based factors
as on training of the individual.
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Clinical Implications of Our Findings
Our results suggest that a patient presenting to the
ER with cardiac-related symptoms be initially given a
physical examination and an electrocardiogram. If
ECG findings diagnostic for acute myocardial infarction or physical findings of congestive heart failure
are absent, evaluation of left ventricular systolic
function becomes valuable. In our data, patients
without ST elevation or findings of congestive heart
failure had a significantly increased risk of both early
and late events in the presence of LVSD. As a result,
it would seem prudent that subjects with LVSD in
whom an acute event is suspected should be admitted
to the hospital. In patients with acute symptoms, the
absence of LVSD can ease the decision to discharge
a patient from the ER, particularly if the electrocardiogram is also normal. Patients with LVSD not
suspected to have an acute event or an acute exacerbation of a chronic condition necessitating hospital
admission should be aggressively evaluated as outpatients in the very near future. Patients with LVSD on
2DE should not be lost to follow-up.
Acknowledgments
The authors gratefully acknowledge the ER personnel and the medical housestaff at the University of
Virginia Medical Center without whose support this
study would not have been possible. Special thanks are
due to Linda Stephens and Lynn Campbell. We also
acknowledge the technical assistance provided by the
following medical students: John Hoover, David Roberts, Thomas Douglas, Andy Cutler, and Mary Jefferies. The authors are also grateful to George A. Beller,
MD, John P. DiMarco, MD, and Eric R. Powers, MD
for their helpful critique of the manuscript.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
References
20.
1. Short D: The earliest electrocardiographic evidence of myocardial infarction. Br Heart J 1970;32:6-15
2. Goldman L, Cook EF, Brand DA, Lee TH, Rouan GW,
Weisberg MC, Acampora D, Stasiulewicz C, Walshon J,
Terranova G, Gottlieb L, Kobernick M, Goldstein-Wayne B,
Copen D, Daley K, Brandt AA, Jones D, Mellors J,
Jacubowski R: A computer protocol to predict myocardial
21.
22.
23.
1623
infarction in emergency department patients with acute chest
pain. N Engl J Med 1988;318:797-803
Lee TH, Rouan GW, Weisberg MC, Brand DA, Cook EF,
Acampora D, Goldman L: Sensitivity of routine clinical criteria for diagnosing myocardial infarction within 24 hours of
hospitalization. Ann Intern Med 1987;106:181-186
McGuiness JB, Begg TB, Semple T: First electrocardiogram in
recent myocardial infarction. Br Med J 1976;2:449-451
Behar S, Schor S, Kariv I, Barrel V, Modan B: Evaluation of
electrocardiogram in emergency room as a decision-making
tool. Chest 1977;71:486-491
Abbott JA, Scheinman MM: Nondiagnostic electrocardiogram
in patients with acute myocardial infarction. Am J Med 1973;
55:608-613
Lee TH, Rouan GW, Weisberg MC, Brand DA, Acampora D,
Stasiulewicz C, Walshon J, Terranova G, Gottlieb L, Goldstein-Wayne B, Copen D, Daley K, Brandt AA, Mellorf J,
Jacubowski R, Cook F, Goldman L: Clinical characteristics
and natural history of patients with acute myocardial infarction sent home from the emergency room. Am J Cardiol
1987;60:219-224
Goldman L, Weinberg MC, Weisberg M, Bran DA, Acampora
D, Stasiulewicz C, Walshon J, Terranova G, Gottlieb L,
Goldstein-Wayne B, Copen D, Daley K, Brandt AA, Mellors
J, Jacubowski R, Cook EF, Goldman L: A computer derived
protocol to aid in the diagnosis of emergency room patients
with acute chest pain. N Engi J Med 1982;307:588-596
Pozen MW, D'Agostino RB, Selker HP, Sytkowski PA, Hood
WB, Wilson C, Decklebaum L, Fineberg H, Stiratelli R: A
predictive instrument to improve coronary-care-unit admission practices in acute ischemic heart disease. N Engi J Med
1984;310:1273-1278
Brush JE, Brand DA, Acampora D, Chalmer B, Wackers FJ:
Use of the initial electrocardiogram to predict in-hospital
complications of acute myocardial infarction. N Engl J Med
1985;312:1137-1141
Lee TH, Cook EF, Weisberg MG, Sargent RK, Wilson C,
Goldman L: Acute chest pain in the emergency room: Identification of low risk patients. Arch Intern Med 1985;145:65-69
Schroeder JS, Lamb IH, Harrison DC: Patients admitted to
the coronary care unit for chest pain: High-risk subgroup for
subsequent cardiovascular death. Am J Cardiol 1977;39:
829-832
Schamroth L: An Introduction to Electrocardiography, ed 6.
Oxford, England, Blackwell Scientific Publications, 1982.
Weyman AE: Cross-sectional Echocardiography. Philadelphia,
Lea and Febiger, 1982
Touchstone DA, Beller GA, Nygaard TW, Tedesco C, Kaul S:
Effects of successful intravenous reperfusion therapy on
regional myocardial function and geometry in humans: A
tomographic assessment using two-dimensional echocardiography. JAm Coll Cardiol 1989;13:1506-1513
Rolet JWT, Terwelle HF, Van Capella FJL, Lie KI: Infarct
size estimation from serial CK MB determinations: Peak
activity and predictability. Br Heart J 1983;49:373-380
Weidner N: Laboratory diagnosis of acute myocardial infarct:
Usefulness of determination of lactate dehydrogenase
(LDH)-1 level and ratio of LDH-1 to total LDH. Arch Pathol
Lab Med 1982;106:375-377
Walker S, Duncan D: Estimation of the probability of an event
as a function of several independent variables. Biometrika
1967;54:167-179
Cox D: Regression models and life tables (with discussion). J
R Stat Soc, B 1972;34:187-220
Lane PW, Nelder JA: Analysis of covariance and standardization as instances of prediction. Biometrics 1982;38:613-621
Link CL: Confidence intervals for the survival function using
Cox's proportional hazard model with covariates. Technical
Report No. 45, Division of Biostatistics, Stanford University.
Kaplan E, Meier P: Nonparametric estimation from incomplete observations. JAm Stat Assn 1958;53:457-481
Mood AM, Graybill FA: Introduction to the Theory of Statistics.
New York, McGraw-Hill, 1963
1624
Circulation Vol 84, No 4 October 1991
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
24. Tennant R, Wiggers CJ: The effect of coronary artery occlusion on myocardial contraction. Am J Physiol 1935;112:
351-361
25. Hauser A, Gangadharan V, Ramos R, Gordon S, Timmis GC:
Sequence of mechanical, electrocardiographic and clinical
effects of repeated coronary occlusion in human beings: Echocardiographic observations during coronary angioplasty. JAm
Coll Cardiol 1985;5:193-197
26. Lepeschkin E: Modem Electrocardiography. Baltimore, Williams & Wilkins, 1951
27. Rovelli F, De Vita C, Feruglio GA, Lotto A, Selvini A,
Tognoni G: Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction: Gruppo Italiano per lo
Studio della Streptochinasi nell Infarto Miocardico (GISSI).
Lancet 1986;1:397-402
28. ISIS-2 (Second International Study of Infarct Survival) Collaborative Group: Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of
suspected acute myocardial infarction. Lancet 1988;2:349-360
29. Chamberlain DA, deBono DP, Fox KAA, Murray RG: Effect
of intravenous APSAC on mortality after acute myocardial
infarction: Preliminary report of a placebo-controlled clinical
trial: AIMS Trial Study Group. Lancet 1988;1:545-549
30. Wilcox RG, von der Lippe G, Olsson CG, Hampton JR,
Wilcox UK, Jensen G: Trial of tissue plasminogen activator for
mortality in acute myocardial infarction: Anglo-Scandinavian
Study of Early Thrombolysis (ASSET). Lancet 1988;2:525-530
31. GISSI-2: A factorial randomised trial of alteplase versus
streptokinase and heparin versus no heparin among 12,490
patients with acute myocardial infarction. Lancet 1990;336:
65-71
32. Oh JK, Miller FA, Shub C, Reeder GS, Tajik AJ: Evaluation
of acute chest pain syndromes by two-dimensional echocardiography: Its potential application in the selection of patients
for acute reperfusion therapy. Mayo Clinic Proc 1987;62:59-66
33. Loh IK, Charuzi Y, Beeder C, Marshall LA, Ginsburg JH:
Early diagnosis of non-transmural myocardial infarction by
two-dimensional echocardiography. Am Heart J 1982;104:
963-968
34. Peels CH, Visser CA, Funke Kupper AJ, Visser FC, Roos JP:
Usefulness of two-dimensional echocardiography for immediate diagnosis of myocardial infarction in the emergency room.
Am J Cardiol 1990;65:687-691
35. Buda AJ, Zotz RJ, Pace DP, Kranse LC: Comparison of
two-dimensional echocardiographic wall motion and wall
thickening abnormalities in relation to the myocardium at risk.
Am Heart J 1986;111:587-592
36. Kaul S, Pandian NG, Gillam LD, Newell JB, Okada RD,
Weyman AE: Contrast echocardiography in acute myocardial
ischemia: III. An in vivo comparison of the extent of abnormal
wall motion with the area at risk for necrosis. JAm Coll Cardiol
1986;7:383-392
KEY WORDS * echocardiography * systolic function, left
ventricular
Importance of two-dimensional echocardiographic assessment of left ventricular systolic
function in patients presenting to the emergency room with cardiac-related symptoms.
P Sabia, R D Abbott, A Afrookteh, M W Keller, D A Touchstone and S Kaul
Downloaded from http://circ.ahajournals.org/ by guest on June 18, 2017
Circulation. 1991;84:1615-1624
doi: 10.1161/01.CIR.84.4.1615
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1991 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/84/4/1615
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/