REVIEWS OF CONTEMPORARY
LABORATORY METHODS
Arnold M. Weissler, M.D., Editor
A Critical Review of
the Systolic Time Intervals
RICHARD P. LEWIS, M.D., STANLEY E. RITTGERS, M.S.,
WILBUR F. FORESTER, M.S., P.E., AND HARIsIOs BOUDOULAS, M.D.
SYSTOLIC TIME INTERVALS (STI) have become one
of the established "noninvasive" techniques of clinical car-
diology. Indeed the term "noninvasive" was first used in con-
nection with the STI. The STI were one of the first quantita-
tive noninvasive tests of cardiac function and remain one of
the simplest and most reliable to perform.
Most tests of ventricular performance deal with force
and/or distance either alone or as a function of time. The
STI are unique among these tests because time is the only
variable. Hence, the validation of the STI has been a two
stage process; first, to establish that external time measure-
ments were accurate; and second, to correlate the STI with
other parameters of ventricular function, none of which in-
volve time as the primary variable. The first stage has been
Downloaded from http://ahajournals.org by on August 29, 2023
accomplished over the past ten years and work continues on
the second stage.
The STI should not be considered competitive with other
invasive or noninvasive studies. Indeed, no single noninva-
sive test provides all of the clinical information which is
desired. The STI are less useful for differential diagnosis
since the method is primarily a measure of left ventricular
performance. As such, the STI provide a quantitative esti-
mate of the effect (if any) of various cardiovascular dis-
orders upon the left ventricle. Like all noninvasive tech-
niques, the STI have the singular advantage that multiple
observations can be performed. This allows study of the
natural history of disorders and long-term effects of thera-
peutic interventions.
Since several recent reviews have extensively covered the
historical background, pathophysiologic basis of the STI,
and have summarized the changes in the STI which occur in
disease states,' this review will concentrate on three aspects:
a critical evaluation of methodology, a summary of factors
known to influence the STI, and the practical use of the STI
for clinical decision-making.
From the Division of Cardiology, The Ohio State University College of
Medicine, Columbus, Ohio.
Supported in part by a USPHS, NHLI, Cardiovascular Training Grant
No. 5 TOI HL05968-03 and Central Ohio chapter of American Heart
Association.
Address for reprints: Richard P. Lewis, M.D., Director, Division of Car-
diology, Ohio State University Hospitals, 410 West 10th Avenue, Columbus,
Ohio 43210.
146
Methodology
The three basic STI are the pre-ejection period (PEP), the
left ventricular ejection time (LVET), and total electro-
mechanical systole (QS,) (fig. 1).4 ' The measurements are
obtained from simultaneous high speed recordings (100
mm/sec) of an electrocardiographic lead best displaying the
onset of left ventricular depolarization, a carotid pulse trac-
ing, and a phonocardiogram best displaying the initial high
frequency vibrations of the aortic closure sound. The patient
is instructed to breathe quietly and at least 10 cardiac cycles
are averaged to obtain the STI. Because there is a trans-
mission delay (average 18.5 ± 8.2 [1 SD] msec)' in the carotid
pulse tracing, the PEP must be calculated by subtraction of
the LVET from the QS,.
It has long been appreciated that the STI vary inversely
with heart rate.`-" Thus, for deviations of the STI to be
properly interpreted, correction must be made for variation
related to differences in resting heart rate. This was an im-
portant conceptual advance for the clinical application of the
STI. When the heart rate derived from the R-R interval is
used, a simple linear equation best describes the relationship
in the range of the resting heart rate.4 1" The regression equa-
tions of Weissler represent the largest population of nor-
mals and have generally been adopted4 (table 1). It is nota-
ble that the equations for males and females differ slightly.4
Different corrections are required for children until puber-
ty.' 14 "The STI steadily increase in duration from infancy to
puberty (the PEP more than the LVET) and all are slightly
prolonged in the elderly.16' 17
In applying the STI clinically, deviations from the normal
regressions can be expressed in either of two ways: 1) sub-
traction of the actual value from the predicted normal for a
given HR (APEP, ALVET, AQS,), or 2) calculation of each
STI as an "index" value. Clinical experience indicates the
latter is the most useful approach. Calculation of the index
value requires transposition of terms of the regression equa-
tion as shown in table 1. The normal index values, in fact,
represent the expected normal value at zero heart rate. By
calculation of the index value, an immediate estimate of the
deviation from normal (not accountable by HR) is obtained.
For example, a QS, index (QSJI) of 506 msec in a male is
easily recognized as being 40 msec shorter than the expected
normal value of 546 msec.
 SYSTOLIC TIME INTERVALS/Lewis et al. 147

EMIlIq
a
m ==MS
L PHONO
I
l'
1

AORTIC
PRESSURE

LV PRESSURE

_t ECG
I
, PEP z QS2 - LVET
FIGURE 1. On the left is a schematic of the left ventricular events which comprise the STI. On the right is a recording of
aphonocardiogram, external carotid pulse, and electrocardiogram at 100 mm/sec paper speed. The subintervals of the
STI are indicated on both.

Validation of the External Carotid Pulse
It is well known that the systemic pressure pulse under-
goes a change in configuration as it passes down the arterial
tree."8 It attains a higher peak pressure, becomes more
Downloaded from http://ahajournals.org by on August 29, 2023
due to damping effects but rather to phase shifting of the
pulse wave component frequencies causing a reshaping of the
pulse wave.
This phase shift could affect the validity of the carotid

smooth, and the incisural notch becomes less sharply defined pulse since the landmarks could be influenced.20 However,
because of the short distance from the aortic root to the
(fig. 2). carotid artery, this error is negligible. Indeed, several in-
In a recent study to determine the source of this distor-
tion, micromanometer tip catheter measurements of aortic vestigations in which the LVET from the externally recorded
root pressure were compared to simultaneously recorded ex- carotid pulse has been compared to the LVET directly mea-
ternal carotid pulse waves employing an air-coupled Stat- sured from a manometer tip catheter in the aortic root have
ham P23 Db strain gauge."9 Spectral analysis of the pulse shown no significant differences.21-2' This phase shift could,
waves was performed over a range of 0-100 Hz. Comparing
however, produce significant distortion if the LVET is mea-
the average amplitude at each frequency harmonic indicates sured from more distal sites in the arterial tree.'8
that the aortic root and external carotid artery pulses have
the same harmonic content and that frequencies beyond the
fifteenth harmonic contain less than 2% of the fundamental r 1
amplitude (fig. 3). However, phase analysis of the two pulses
showed a definite change in phase angle between the two
pulses the phase shift being most prominent at higher fre-
quencies. The distortion in the carotid pulse is therefore not

TABLE 1. Calculation of STI Index Values from Resting

Regression Equations4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Normal Index Tine, Sec.
Sex Equation (mseo) SD

M QS2I 2.1 HR + QS2

= 546 14 ExternOl
F QSJI = 2.0 HR + QSa 549 14 CaroJid
ACrt'
M LVETI = 1.7 HR + LVET 413 10
F LVETI = 1.6 HR + LVET 418 11
M PEPI = 0.4 HR + PEP 131 10 FIGURE 2. Comparison of a simultaneously recorded manometer
F PEPI = 0.4 HR + PEP 133 10 tip catheter pressure pulse (Millar "Mikro-tip") in the aorta and an
Abbreviations: I = index; SD = standard deviation; M = male; F externally recorded carotid artery pressure pulse employing an air
female; HR - heart rate. coupled Statham P23 Db strain gauge.
 148 CIRCULATION VOL 56, No 2, AUGUST 1977

FIGURE 3. Harmonic amplitudes of blood

pressure pulses at various body sites obtained by
Fourier analysis. Numbers 2 (7 patients) and 3 (2
patients) were obtained from a manometer tip
catheter (Millar "Mikro-tip"), and number 4 (7
patients) by an air coupled P23 Db Statham strain
gauge. Number 1 is the original data of A. T.
Hansen (Pressure measurement in the human
organism; Technisk Forlag, Copenhagen, 1949).

2 3 4 5 6 7 8 9 10 11 12 13 14 15
Harmonic Number

Validation of the QS2
There is no absolute physiologic landmark for the end of
systole which can be derived externally. The QS2I interval,
which ends with the initial high frequency vibrations of the
aortic component of the second heart sound (A2), represents
an estimate in which easily identified physiologic landmarks
are employed. Recent studies have demonstrated that the in-
Downloaded from http://ahajournals.org by on August 29, 2023
cisural notch and the initial high frequency vibrations of A2
nearly coincide and follow aortic valve coaption by less than
5 msec.2627 Aortic valve closure in turn normally follows left
ventricular-aortic pressure crossover by 5-10 msec. It
appears that this gap is relatively constant except in certain
disease states (notably aortic valve disease).
The QS2 could potentially be unreliable if there were a
wide variability in the timing of the end-systolic left ventric-

TC - 1.0

TC - 0.05

TC - 0.25
Tc- O.5 V v

FIGURE 4. The effect of transducer time constant (Tc) on a carotid pulse wave. The data was obtained by electrically
simulating the various time constants. The marked change in configuration with the short T, values represent amplitude
loss and phase shift of the low frequency components. A lthough these pulses resemble a differentiated signal, they are not
since only low frequencies are affected.
 SYSTOLIC TIME INTERVALS/Lewis et al.
ular-aortic pressure crossover. Although this has not been
systematically investigated, studies of patients with atrial
fibrillation and a wide variety of left ventricular systolic
pressures show a remarkable constancy of the aortic and left
ventricular pressure crossover at end-systole relative to peak
aortic pressure.2' 28 Finally, the narrow range of normal for
the QS2I at all heart rates and pressures support the reli-
ability of the QS2I as a useful measure for the duration of
systole.
Validation of the PEP
Recent catheter tip manometer studies in both experi-
mental animals and man have validated the externally mea-
sured PEP.21' 22, 29 The PEP is in fact comprised of the elec-
tromechanical delay (Q to onset of pressure rise in the left
ventricle) and the isovolumic contraction period. The elec-
tromechanical delay in normals is 30-40 msec2' 0, 31 and is
minimally affected in disease unless left bundle branch block
is present.2' 82-7 The electromechanical delay is prolonged by
varying intervals in complete left bundle branch block. It has
been shown that the upstroke of the apexcardiogram offers a
reliable measure of the onset of left ventricular pres-
sure2, 21 28, 89 and as such is useful in defining the electro-
mechanical delay.
The first heart sound occurs 25-50 msec after left ventric-
ular left atrial pressure crossover and as such is not a useful
landmark for subdividing the PEP.26' 40 At present the
clinical usefulness of defining the subintervals of the PEP is
not conclusively established in patients in whom there is no
left ventricular conduction delay.
Downloaded from http://ahajournals.org by on August 29, 2023
Pulse Transducers
When considering a transducer system for pulse record-
ings, the following properties should be known: 1) time con-
stant; 2) frequency response; 3) phase shift. A time constant
which is too short will distort the pulse by poor representa-
tion of lower frequencies in the form of both amplitude loss
and phase shift.41 A poor frequency response (which might
well be caused by the connecting apparatus) causes poor
definition of higher frequencies. It appears that a flat fre-
D 15 -
.o 10
5
4 CA~~~~
0E
5 Transdk
0~
0~ ~ ~ ~ ~
-90
-180
I0 00
Frequency, Hz
149
quency response of from 0.1-30 Hz is required.18-20 42 This
allows faithful reproduction of the first 15 harmonics for
heart rates up to 120 beats/min. Likewise, there should be no
phase shift from 0.1-30 Hz. Phase distortion independent of
the time constant predominately affects the higher frequen-
cies and if present may also be due to the connecting ap-
paratus. Of the above transducer inadequacies, poor ampli-
tude frequency response will not affect the time events of the
recorded pulse but may so distort the pulse configuration as
to distort the landmarks. A phase shift on the other hand can
alter the time events of the pulse.
From direct testing of several transducers in our labora-
tory and a review of the literature, it appears that nearly all
pulse transducers have an adequate frequency response un-
less improper connecting apparatus is attached.42' 43 How-
ever, most, if not all, crystal devices have a short time con-
stant (< 2 seconds). This distorts the pulse configuration and
can efface the time landmarks"' (fig. 4). The ideal transducer
has an infinite time constant.
Figure 5 shows the characteristics of the system employed
in our laboratory. This involves a Statham P23 Db pressure
transducer which is air coupled by attaching a 4" length of
1/8' inner diameter tygon tubing and a brass funnel 1 1/2"
long with maximum and minimum inner diameter of 7/8"
and 1/16", respectively. The transducer output is connected
to a carrier preamplifier. The time constant of this device is
infinite. Its frequency response and phase shift are flat (+ 3
dB to 100 Hz).* The distortion occurring at higher frequen-
cies is due both to our connecting apparatus and the trans-
ducer itself.42 If a connecting apparatus different from the
above is employed, it is imperative to establish that it allows
a flat frequency response and phase shift to at least 30 Hz.
Another transducer with similar excellent characteristics is
the Hewlett-Packard Sanborn APT-16, which has no con-
necting apparatus.", 48
Optimum Paper Speed
The error incurred in reading a given STI can reflect two
factors. One is due to misjudgment of the actual physiologic
*Details of our testing system available upon request.
FIGURE 5. Amplitude and phase response of an
air-coupled Statham P23 Db strain gauge
transducer. Note that both responses areflat (± 3
Db) to 100 Hz.
1000
 150 CIRCULATION
event desired. The second error is due to limitations of the
observer's accuracy. This latter error is related to factors
such as the spacing of the timelines, resolution of the ob-
server's eye, thickness of timelines, etc. It has been sug-
gested that a rapid paper speed minimizes this second error,
although this contention has been disputed by the only direct
study of the problem."
In order to resolve this problem we have completed a
study of seven carotid pulse waves obtained from seven pa-
tients. In each case the pulse wave was at least 4 cm in ampli-
tude and judged to have distinct landmarks. The pulse waves
were recorded on magnetic tape and re-recorded at 25, 50,
and 100 mm/sec paper speed. The timelines were kept at 40
msec intervals for each paper speed. The pulses were elec-
tronically digitized using a Hewlett-Packard 9830A com-
puter and digitizer. The computer-determined data were con-
sidered the most accurate estimate of the LVET against
which to compare the technicians' measurements.
The results are illustrated in figure 6. The mean absolute
deviations from the digitized value for five experienced STI
technicians for each of seven pulses at each paper speed are
shown. It is clear that there is significantly greater vari-
ability in the measurement at slower paper speeds. Of inter-
est, when data from the previous study are analyzed in a
similar manner, essentially identical results are obtained."
In addition to greater measurement variation at slower
paper speeds, this study also revealed that the LVET mea-
16
Downloaded from http://ahajournals.org by on August 29, 2023
14
E
-
12
10
:t
8 to measure the STI.45 4" Theoretically this should reduce sys-
I tematic error. However, improper selection of landmarks by
A puter's selection of landmarks should be verified by a trained
6
4
I The carotid pulse should be as large as possible (4-5 cm)
2 (fig. 7). It is also useful to have thinning of the pulse line with
0
I I I I
0 25 50 75 100
Speed(mm/sec)
FIGURE 6. Mean absolute deviation of left ventricular ejection
times from seven patients at paper speeds of 25, 50, and 100
mm/sec. Each point represents the mean offive observers. Note
that at slower recording speed the variation is greater and the true
value is less well estimated.
VOL 56, No 2, AUGUST 1977
sured at 100 mm/sec paper speed was the most accurate. The
mean LVET for the seven pulses obtained by electronic
digitization was 236 msec. The mean value and average 1 SD
for all seven patients at 25 mm/sec was 243 ± 7.1, at 50
mm/sec it was also 243 ± 5.2 and at 100 mm/sec it was 235
3.6 (P < 0.05).
To determine the interobserver measurement error at 100
mm/sec paper speed, the mean calculated LVET for all
seven pulses for each technician was compared to the mean
electronically digitized LVET. The absolute mean errors
were 3.0, 4.5, 3.6, 3.4, and 5.1 msec (average 3.9 msec).
To determine the intraobserver measurement error, the
same five technicians measured the same pulse at 100
mm/sec paper speed on four separate occasions. The mean
LVET from the four measurements for each observer was
again compared to the electronically digitized value for that
patient. The absolute mean errors were 5.0, 5.4, 3.5, 3.8, and
5.0 msec (average 4.5 msec).
These studies indicate that when the LVET is manually
calculated at 100 mm/sec paper speed by trained tech-
nicians, there is a high degree of accuracy and precision, in
each case to the nearest 5 msec or better. It should be noted
that all of these studies were performed on a single pulse. In
practice, ten pulses are measured and averaged thus reduc-
ing the measurement error even further.
In a similar manner the mean absolute error of the QS2
measurement was derived. The mean absolute error was 1.6
msec. Since the PEP is determined by subtraction of the
LVET from the QS2, the error in the PEP is a function of the
error in both primary variables. The same is true for the
PEP/LVET. In order to estimate error in the measurement of
the PEP/LVET a generalized equation was developed using
the experimentally derived mean absolute error values of the
LVET and QS2.* This equation provides the maximum ex-
pected error (table 2). It is notable that the maximum error
of the PEP/LVET is dependent on both the absolute value of
the PEP/LVET and the HR, with the error being greatest at
rapid heart rates when significant left ventricular dysfunc-
tion is present.
Several investigators have developed computer programs
the computer may induce serious error so that the com-
observer.
Other Technical Considerations
for optimal delineation of the upstroke and incisural notch
sudden changes in direction. In general, only optical
recorders provide these features. It is obvious that the paper
speed must be accurate. At 100 mm/sec an error of 2% pro-
duces a 5 msec error. Thus, the recorder must be frequently
calibrated. In our experience, many commercially available
recording systems do not meet these requirements.
Other common, but important errors include: 1) failure to
average ten consecutive cycles in order to minimize the influ-
ence of respiratory variation; 2) failure to employ the elec-
*Detail of derivation of this equation available upon request.
 SYSTOLIC TIME INTERVALS/Lewis et al. 151

TABLE 2. Maximum Error in Measurement of PEPILVET be generated. As a consequence, the PEP following short cy-
PEP/LVET 0.33 0.50 0.75 cle lengths is relatively prolonged.2 This is most striking at
heart rates above 75 beats/min (R-R < 800 msec). Although
HR 50 an angiographic study has shown a high correlation between
(QS2 = 441 msec) 0.021 -0.026 10.034 the PEP/LVET and left ventricular ejection fraction on a
HR 75 beat-to-beat basis,50 inclusion of many beats with short pre-
(QS2 = 400 msec) i0.023 -0.028 -0.037 ceding R-R intervals can prove misleading. Indeed, if the
HR 100 PEP/LVET obtained from averaging 25 consecutive beats
(QS2 = 336 msec) "0.028 =0.034 =0.045 (which would likely include several beats with R-R intervals
Abbreviations: HR = heart rate. < 800 msec) and the PEP/LVET obtained from averaging
only beats with R-R > 800 msec are compared to the
trocardiographic lead, which best defines the onset of elec- PEP/LVET when sinus rhythm is restored, the former is sig-
trical depolarization; 3) poor quality phonocardiograms nificantly higher and the latter not different.2 Thus STI
which do not identify A2 precisely (sounds of frequencies less calculated by averaging all beats including those following
than 100 Hz must be effectively filtered). There are some short cycle lengths can underestimate potential left ventric-
patients in whom good carotid pulse tracings or phono- ular performance and this error can be minimized by mea-
cardiograms cannot be obtained. In such cases it is obvious suring only those beats with an R-R > 800 msec.
that poor data should not be reported. Few studies of dysrhythmias other than atrial fibrillation
have been reported.51 The STI have minimal clinical value in
Effect of Atrial Fibrillation and Other Dysrhythmias Upon
this setting and are best studied after the tachyarrhythmia is
Measurement of the STI
slowed or reverted.
Premature beats of any type may produce postextrasys-
This common dysrhythmia produces alterations in the tolic potentiation and a decreased PEP/LVET in the follow-
usual relationship of the LVET and PEP to HR. There has ing beat. In atrioventricular conduction disorders, beats with
been some controversy relating to the choice of either the a properly timed atrial systole show a decreased
preceding R-R interval per se or the HR calculated from the PEP/LVET.52' 53 Finally, pulsus alternans may produce
preceding R-R interval to correct the STI. Since HR and marked beat-to-beat changes in the PEP/LVET.54 6 In all
R-R are nonlinearly related, apparently different results are of these situations the beat-to-beat variations need to be con-
obtained from the two methods. In fact, it matters little sidered when reporting the STI.
which method is employed if the difference between the two
Downloaded from http://ahajournals.org by on August 29, 2023

is appreciated. Physiologic Variations of the STI

At slower HRs in atrial fibrillation the LVET does not
lengthen as much as would be expected from the linear Weissler's original HR-STI regressions were based upon
regression lines for normal subjects.2 4749 This is most evi- normal volunteers with no heart disease who were at rest in
dent when the HR calculated from the preceding R-R in- the supine position for one hour.' In clinical use these condi-
terval is used rather than the actual R-R interval itself. This tions are not easily met; the patient is being sent for a "heart
effect of R-R interval on LVET is mitigated when mitral test" which could produce emotional arousal. For practical
stenosis (and impaired diastolic filling) is present. considerations in most busy laboratories the patient cannot
In beats with a short preceding R-R interval there is be at rest more than a few minutes. However, if the 95%
diminished preload and a greater isovolumic pressure must confidence intervals for Weissler's data are considered, nor-
mal patients studied with only 5-10 min rest still fall into his
QS21 540 QS21 540 normal range (fig. 8).2
LVETI 402
It is important to appreciate the day to day variation of
LVETI 385
the STI both in normal persons and in patients with stable
PEP I 155 PEPI 138 chronic heart disease. This has been assessed in both groups
P/ L .47 P/L .39
(table 3).3 As might be expected, the mean one standard
deviations of the STI for individuals are smaller than the one
standard deviation values for the entire normal population
(table 1), both for normals and abnormals. These small stan-
dard deviations indicate the remarkable consistency of the
STI in both groups, although the variability is higher in
patients with chronic heart disease.

Factors Influencing the STI

FIGURE 7. The effect of a low amplitude carotid pulse tracing
upon the time landmarks of the LVET. The same patient was
studied with the carotid tracing on high gain (left) and low gain
(right). The landmarks ofthe low gain tracing are poorly delineated,
producing a 17 msec error in the measurement of the L VET.
Reproduced by permission from Grune & Stratton from "Systolic
time intervals" in Noninvasive Cardiology, 1974.
The factors known to influence the PEPI and LVETI are
listed in table 4. There have been several studies to deter-
mine the mechanism of STI abnormalities by direct correla-
tion with other measurements of left ventricular per-
formance.2' 4, 5, 10, 11, 21, 22, 29, 47, 51-" From these studies the
following facts seem clear. When left ventricular failure oc-
curs, regardless of cause, the PEPI lengthens and the LVETI
 152 CIRCULATION VOL 56, No 2, AUGUST 1977

~~~~~~E
(0 0 LEET - 250

v)

30- 200-

50- 50-

50- 60 70 90 90 n)o 1 10 50 so 70 so 90 K)O

HR HR
FIGURE 8. Individual data on 31 female (left) and 26 male (right)patients who subsequently proved to have no heart dis-
ease. Ninety-five percent confidence intervals of the regression equations for 90 normal female and 121 normal male
volunteers for each of the SrI are superimposed.4 The patients were studied as part of a cardiac evaluation between 8:00
Downloaded from http://ahajournals.org by on August 29, 2023

AM and 3:00 PM with only afive minute rest period. The values obtained under these circumstances were not statistically
different from those of the normal volunteers. Permission for reproduction, see figure 7 legend.

shortens. The prolongation of the PEPI can be attributed The shortening of the LVETI with heart failure is more
almost entirely to a diminished rate of left ventricular complex to explain than the changes in the PEPI. A primary
pressure rise during isovolumic systole (LVdp/dt). Occa- factor in the shortening of LVET must be the relative
sionally this can be masked (i.e., the PEPI is normal in the lengthening of the PEP which induces a delayed onset of
presence of a low rate of pressure rise) if the pressure devel- ejection. During ejection the velocity of fiber shortening is
oped during isovolumic contraction is low.65 This occurs in diminished when left ventricular failure is present, a condi-
patients with a high left ventricular end-diastolic pressure tion which theoretically should produce a prolonged
and low systemic diastolic pressure. LVETI.68 However, the extent of fiber shortening is also
Complete left bundle branch block usually delays activa- reduced and this would tend to shorten the LVETI. It
tion of the left ventricle and therefore prolongs the PEPI (see appears that the diminished extent of fiber shortening in left
earlier section). Lesser degrees of left ventricular conduc-
tion abnormality, however, seem to produce only a minimal TABLE 4. Factors Influencing the Systolic Time Intervals
delay in left ventricular activation.2 In chronic left ventric- Increase ( f ) Decrease ( )
ular disease associated with reduced diastolic compliance,
the PEPI may also be prolonged if preload is inadequate.66 A 1) PEP LV muscle failure aortic valve disease
similar phenomenon occurs with the diminished preload in- left bundle branch I LV isovolumic
block pressure
duced by upright tilt.67 1 preload positive inotropic
negative inotropic agents
TABLE 3. Day-to-Day Variation of the STI2 agents
nse¢ 2) LVET aortic valve disease LV muscle failure
QS2I LVETI PEPI PEP/LVET I preload
positive inotropic
I. Normals agents
(N = 33) 6.2* 5.2 4.3 0.015 negative inotropic
II. Chronic stable agents
LV disease 3) QS2 left bundle branch positive inotropic
(N = 17) 11.8 10.5 7.5 0.036 block agents
* = mean standard deviation. aortic valve disease
LV = left ventricle.
 SYSTOLIC TIME INTERVALS/Lewis et al.
ventricular decompensation usually predominates so that the
net result is a shortened LVETI. Of interest is the fact that
the absolute size of the stroke volume per se does not deter-
mine the LVET.2 It is when stroke volume is small relative to
end-diastolic volume that the LVETI will be shortened. This
can be observed in some patients with primary myocardial
disease in whom a shortened LVETI occurs when the left
ventricular end-diastolic volume is increased, the ejection
fraction reduced, and the stroke volume and cardiac output
are normal.2
Unlike the PEPI, which responds in an opposite manner
to positive and negative inotropic agents, the LVETI is gen-
erally shortened by both. With negative inotropic drugs, the
mechanism is similar to that which occurs with left ventric-
ular failure."7' Positive inotropic agents can increase both
the velocity and extent of fiber shortening. These changes
have opposite effects on the duration of the LVETI. Gen-
erally the increased rate of shortening is the predominant
effect so that the LVETI shortens or is unchanged.72-7' How-
ever, lengthening can occur when a marked hemodynamic
response occurs (i.e., large increase in the stroke volume)."
Finally, diminished preload reduces the stroke volume and
shortens the LVETI." "
The duration of systole (QS2I) is one of the remarkable
constants in the circulatory system. Although most types of
heart disease may produce profound alterations in cardiac
performance, manifest by directionally opposite changes in
the PEPI and LVETI, the QS21 is unchanged from normal
unless drug effects are present.
Since positive inotropic agents generally shorten both the
Downloaded from http://ahajournals.org by on August 29, 2023
LVET and PEPI, the QS2I is the most useful of the STI in
judging the presence of positive inotropic stimulation.
Indeed, studies have shown a strong correlation between
urinary catecholamine excretion and the QS21 in acute myo-
cardial infarction as well as a dose-related shortening of the
QS21 in response to intravenously administered positive ino-
tropic agents.72 78, 76 This is a unique property of the STI. No
other easily derived test of left ventricular function provides
a measure of the inotropic state.
The diagnostic value of the STI for identification of left
ventricular dysfunction can be enhanced by determination of
the ratio of PEP to LVET. In the heart rate range below 110
beats/min this ratio is unrelated to heart rate since the PEP
and LVET shorten proportionally as the heart rate increases.
The PEP/LVET may identify left ventricular dysfunction
when either (or both) the PEPI and LVETI are still within
the normal range. Consequently, the PEP/LVET has
become the single most useful measurement of left ventric-
ular performance from the STI.
Clinical Application
Evaluation of Left Ventricular Performance in Suspected Chronic
Myocardial Disease
This is the area in which the STI were first clinically evalu-
ated. Early studies showed a strong correlation between the
PEP/LVET and cardiac index.4" However, as more patients
were studied, it became clear that while the STI were vir-
tually always abnormal when the cardiac index was reduced,
it was not uncommon to find an abnormal PEP/LVET in the
face of a normal cardiac index.2 This was in accord with the
153
longstanding observation that the cardiac index is one of the
last parameters of left ventricular performance to become
abnormal when disease is present.
Subsequently it was shown that of all direct parameters of
left ventricular performance the PEP/LVET correlates best
with the angiographic ejection fraction (EF) (r = -0.90).6
This is true whether the underlying disease is myocarditis,
cardiomyopathy, coronary artery disease with angina pectoris,
hypertensive heart disease, mitral valve disease, congenital
heart disease involving the left ventricle, or pericardial dis-
ease" 7741 (unpublished observations). In myocardial infarc-
tion (both acute and chronic), pulmonary disease,82' ' and
aortic valve disease, this relationship may be influenced by
other factors but the same basic relationship persists (see
later sections).
With the widespread use of modern potent oral diuretics
the degree of congestive failure may relate poorly to the
severity of underlying left ventricular muscle failure. It is no
longer uncommon for patients with severe left ventricular
dysfunction to be free of excessive fluid retention. In such in-
stances the severity of the underlying muscle disease may be
missed. Measurement of the STI can prevent this error. It is
also not unusual for overdiuresis to occur in these patients."
Worsening of the STI with therapy for congestive heart
failure provides an early clue to this phenomenon.
There are several conditions in which circulatory conges-
tion occurs on a basis other than left ventricular muscle
failure. These include constrictive pericarditis, anemia, renal
failure, cirrhosis with ascites, excessive fluid administration,
cor pulmonale, and certain volume load lesions and shunts.
In these patients the classic bedside signs of cardiomegaly,
gallop rhythm, increased venous pressure, pulmonary rales
and edema may not be reliable for assessing left ventricular
function. The STI are highly useful to establish the role (if
any) of left ventricular muscle dysfunction in these patients
(unpublished observations). In other patients the presence of
myocardial dysfunction may be relatively occult and is dis-
covered only when the STI are measured.71'"
Among patients with chronic left ventricular disease there
is a wide spectrum of severity of left ventricular dysfunction.
In many of these disorders the long-term prognosis is
strongly related to the severity of the performance abnor-
malities. Most standard bedside signs of heart failure are
qualitative and do not consistently distinguish those with
mild from those with severe dysfunction. The STI on the
other hand do provide a quantitative estimate of left ventric-
ular dysfunction.
The STI in Acute Myocardial Infarction (AMI)
The STI have been studied extensively in patients with
acute myocardial infarction."' 65. 76. "- Since left ventric-
ular performance frequently changes during the course of the
illness, serial measurements of the STI assume major impor-
tance. In the early stages of the AMI the STI provide a mea-
sure of the impact of the infarction on the left ventricle, while
later studies are useful in assessing the long-term response to
the healing process.
The most consistent (and unique) phenomenon noted in
AMI is a marked abbreviation of the QS21. This has been
shown to be a direct effect of increased adrenergic tone.76
 154 CIRCULATION
This factor is responsible for considerable confusion in the
interpretation of the STI in AMI. In normal subjects with
the high levels of adrenergic tone characteristic of the early
stages of AMI, the PEPI should be shortened, the LVETI
unchanged, and the PEP/LVET low (0.20-0.25).72 Thus, a
"normal" PEPI or PEP/LVET actually represents left ven-
tricular dysfunction when adrenergic tone is increased.
In spite of the above considerations, the directional
changes in the PEP and LVET with AMI are the same as in
chronic disorders. In almost all instances, patients who are
judged to have more severe left ventricular dysfunction
clinically or by direct measurements with cardiac
catheterization techniques have a more abnormal
PEP/LVET.
The mere presence of an acute MI need not necessarily
produce an abnormal PEP/LVET (although as noted above,
these "normal" PEP/LVET values are probably not nor-
mal). Since the STI are a test of performance and the spec-
trum of AMI size is large, a wide range of values for the
PEP/LVET is found when large groups of AMI patients are
studied. Indeed, the same wide overlap with normal is found
with other measures of ventricular performance, such as car-
diac output, left ventricular dp/dt and echocardiographic
measures.88, 82, 98
If patients with AMI are divided into two groups accord-
ing to the presence or absence of pre-existing disease (prior
AMI, longstanding hypertension, primary myocardial dis-
ease), then a striking (and expected) separation occurs (fig.
9)." Those with no pre-existing disease have a normal
PEP/LVET throughout the acute phase and during the sub-
Downloaded from http://ahajournals.org by on August 29, 2023
sequent nine months of convalescence. Patients with prior
left ventricular disease, on the other hand, have an abnor-
mal PEP/LVET throughout the acute phase and up to six
months are required for final normalization of the
PEP/LVET. These results suggest that the STI may be
useful for guiding the convalescence from an AMI and un-
derscore the fact that patients with AMI are not a
homogeneous population in terms of left ventricular per-
formance.
1 Sol
.30
I I
ADM DISCH
I
3 MOS
I
6 MM 9 MOS
FIGURE 9. Serial changes in the PEP/L VET after acute myocar-
dial infarction. Patients with a previously normal left ventricle (L V)
(N = 12) remain normal throughout while those with a previously
abnormal L V (N = 13) have significant left ventricular dysfunction
for at least three months. Asterisks indicate significant differences
between the groups, bars represent I SEM, dashed line indicates up-
per limit of normal for PEP/L VET. ADM= admission;
DIS = discharge; MO = months. Reproduced by permission from
" Usefulness of systolic time intervals in coronary artery disease, " in
Am J Cardiol 37: 787, 1976.
VOL 56, No 2, AUGUST 1977
The STI are probably of less value when cardiogenic
shock is present. They often are technically difficult to ob-
tain. The low isovolumic pressure and increased adrenergic
tone may result in a normal PEPI in spite of a low left ven-
tricular dp/dt.6" However, serial measurements of the
LVETI may be useful.84 86, Recently, the PEPI has been
combined with arterial and wedge pressure data to provide a
contractility index that may also prove useful.86
Assessment of Left Ventricular Function in Chronic Coronary
Artery Disease (CAD)
The correlation between the PEP/LVET and angio-
graphic left ventricular ejection fraction is lower in CAD
than for other types of chronic disease (fig. 10)." The STI
generally suggest better left ventricular performance than
does the ejection fraction. Several phenomena unique to cor-
onary artery disease appear to be responsible. These include
a low isovolumic pressure (developed pressure < 40 mm Hg)
due to a high LVEDP and low aortic diastolic pressure, un-
derestimation of the ejection fraction by the single plane
RAO view when localized contraction abnormalities are
present, and the occurrence of transient myocardial ischemia
during the left ventriculogram. It is clear that only one of
these factors represents an error in the estimation of left ven-
tricular performance by the STI.
In spite of the above considerations, inspection of figure 10
indicates that a PEP/LVET of greater than 0.5 is associated
with an ejection fraction of less than 40% in 95% of the cases.
It can also be seen that the majority of patients with angina
pectoris without prior AMI have normal STI and ejection
fractions. Patients with prior AMI have either normal or ab-
normal STI depending on the size and number of prior in-
farctions." The STI are frequently abnormal in patients who
have recovered from a previous myocardial infarction when
all clinical and radiologic signs of cardiac disease are
absent."4
The STI have proved useful for assessing the effect of
saphenous vein bypass upon left ventricular per-
100
o ANGINA(n-39)
* PRIOR MI tn-76)
* DEVELOPED PRESSURE
<4OumHg (n-12)
60-
EJECTION4
FRATION
M%)
40-
20 -
me,R \.a
N:o*w
: .k.*
A:
e
-
I
*
x,
r 9. 99-l I.t
r
\
* -0.76
. \
a\ 0
0
0 .20 .40 .60 .80 lOO
PEP/LVET
FIGURE 10. Relationship between the angiographic left ventric-
ular ejection fraction (RAO view) and PEPIL VET in 127 patients
with significant coronary artery disease. The symbols designate
various categories. All patients with an isovolumic systolic pressure
of less than 40 mm Hg had a prior myocardial infarction (MI). The
dashed lines indicate 95% confidence intervals from a previous
study.6" Permission for reproduction, see figure 9 legend.
 SYSTOLIC TIME INTERVALS/Lewis et al.
formance.""96 In many instances ischemic injury at the time
of surgery causes postoperative deterioration of the
PEP/LVET. This requires three months to return to
preoperative baseline. Late deterioration of the STI sug-
gests graft closure. Finally, these studies have demonstrated
the role of excessive adrenergic tone in the early postopera-
tive period which must be considered when comparing pre
and postoperative ventricular performance.
Patients with stable angina pectoris may also have in-
creased adrenergic tone although this is not as marked or as
common as in AMI."4 97 When patients with angina pectoris
on no medication are compared to patients with chest pain
and normal coronary arteries, the QS2I is significantly
shorter in those with angina." Indeed, 16% of patients in our
series had a QS2I less than 515 msec, the lower limit of nor-
mal. Thus, a QS2I of less than 515 msec in a patient with
chest pain strongly suggests the pain reflects ischemic dis-
ease. Another study suggested that patients with shortened
values of the QS2I have the most undeveloped collaterals.9"
Aortic Valve Disease
Recent studies have shown that prolongation of the
LVETI is highly useful for identifying hemodynamically sig-
nificant aortic stenosis.24' 27, 99106 This is enhanced if also
associated with a carotid peak dp/dt of less than 500 mm
Hg/sec.19 Patients with significant aortic stenosis usually
have an LVETI greater than 420 msec. Although the length
of the LVETI does not correlate highly with the severity of
the aortic stenosis, the combination of the LVETI and stroke
volume provides a good estimate of the aortic valve area.108
Downloaded from http://ahajournals.org by on August 29, 2023
I C PHONO-
150- -LV
100- DN
-AO
50-
0-
ECG-
FIGURE 11. Left ventricular (L V) and aortic (Ac) pressures and
intracardiac (IC) phono obtained from two manometer tip catheters
in a patient with significant aortic stenosis. Timelines are 40 msec.
No te the SO msec delay between L V-A o press ure crosso ver and the
dicrotic notch (DN). Permission for reproduction, see figure 7
legend.
155
The mechanism of the prolonged LVETI is not com-
pletely clear. In part, it may reflect relative early opening
and late closing of the aortic valve. There also appears to be
a longer than normal delay of the incisura following left ven-
tricular-aortic pressure crossover (fig. 1 1). This delay may be
related to decreased aortic impedance resulting from post-
stenotic dilatation of the aorta and low aortic pressure.'"
Occasionally this factor alone can prolong the LVETI even
when the aortic stenosis is not critical. Whether actual pro-
longation of left ventricular systole is also a factor (it is in
pulmonary stenosis) has not been established.106 Of interest,
the prolonged LVETI almost always returns to normal after
corrective surgery and often is shorter than normal if under-
lying left ventricular dysfunction is present.100 101, 104
When left ventricular failure develops, the LVETI
shortens as in other types of heart disease. However, it
shortens to the normal range and almost never becomes
shorter than two standard deviations of normal (< 398 msec)
unless excessive adrenergic stimulation is present. The find-
ing of a normal LVETI in a patient with congestive heart
failure and clinical evidence of aortic stenosis strongly sug-
gests the heart failure is associated with significant aortic
stenosis. Conversely, a short LVETI, especially with a nor-
mal carotid dp/dt, strongly suggests aortic stenosis is not the
cause of the heart failure. This is a common clinical prob-
lem in patients over age 50 years.
The PEPI is short in aortic stenosis due to the rapid dp/dt
and low aortic diastolic pressure. However, the aortic clo-
sure sound is often absent or effaced when significant aortic
stenosis is present so that neither the QS2I or PEPI can be
determined.
Significant outflow obstruction also prolongs the LVETI
in muscular subaortic stenosis and may be useful both for
diagnosis and following therapy in this disorder.10'7110
The LVETI is also prolonged by aortic regurgitation."
The mechanisms are probably similar to aortic stenosis (i.e.,
early opening and late closing of the aortic valve, and
delayed incisura) rather than to an increased stroke volume.
The relationship of the duration of the LVETI to the magni-
tude of the regurgitant leak has not been precisely deter-
mined, but it appears that more severe regurgitation pro-
duces a greater prolongation of the LVETI. As is the case
with aortic stenosis, left ventricular decompensation
shortens the LVETI.
Mitral Valve Disease
The studies by Garrard and co-workers demonstrated that
a linear relationship between the PEP/LVET and ejection
fraction is retained in patients with mitral stenosis.61 If
patients with mitral stenosis are compared to those with left
ventricular disease (ischemic heart disease and primary myo-
cardial disease) and similar levels of cardiac output, those
with mitral stenosis have less severe abnormalities in STI
and ejection fraction.2 Mitral stenosis thus appears unique in
that a low cardiac index may be accompanied by a normal
ejection fraction and normal STI. When the PEP/LVET
ratio is significantly increased in patients with mitral
stenosis, the ejection fraction is usually reduced."
In patients with mitral regurgitation, prolongation of the
PEPI and shortening of the LVETI, with a consequent in-
 156 CIRCULATION
crease in PEP/LVET, may result from either the dynamic
effects of mitral regurgitation or the influence of diminished
contractile performance of the left ventricle. Recent studies
employing echocardiographic methods for assessing the con-
tractile performance of the left ventricular chamber in
patients with mitral regurgitation have demonstrated that
prolongation of PEPI and shortening of LVETI is often ob-
served when left ventricular chamber performance is within
normal limits."11 These abnormalities in STI are accentu-
ated when mitral regurgitation is accompanied by
diminished left ventricular performance as reflected in the
determination of the extent of minor axis diameter shorten-
ing. From these studies it appears that mitral regurgitation
per se induces prolongation of PEPI and shortening of
LVETI and that these changes are accentuated in the
presence of left ventricular decompensation. The finding of a
normal PEP/LVET in patients with mitral regurgitation
provides evidence for well compensated contractile per-
formance while an increase in PEP/LVET of 0.5 or greater
favors the coexistence of mitral regurgitation and left ven-
tricular decompensation.
Use of the STI in Clinical Pharmacology
Because of the extreme sensitivity of the test, the ease of
its measurement, and the large body of existing literature,
the STI are ideally suited for studying effects of pharma-
cologic agents upon the heart. Indeed, this may well repre-
sent a most useful future application of the technique. Most
of the studies which have been performed have dealt with
parameters of pharmacologic action such as the onset,
Downloaded from http://ahajournals.org by on August 29, 2023
magnitude, and duration of drug action, establishment of
dose response relationships, and studies of the interplay of
agonists and antagonists. 72, 74, 76, 112 lie Recently the avail-
ability of plasma levels of drug have added a new dimension
to such studies.117
The clinical pharmacology of the various digitalis glyco-
sides has been extensively studied with the STI.2 7 75 112 118
These studies have shown that the QS2I is the most sensitive
of the STI in assessing the presence of positive inotropic
stimulation while the hemodynamic response to the agent is
reflected by the pattern of changes of the PEP and LVET.
Several adrenergic positive inotropic agents have also been
studied.",70, 72, 72, 11S With this class of drugs there are
usually more striking hemodynamic changes than is the case
with digitalis. Thus, caution must be exerted to ensure that
pharmacologic effects on the peripheral circulation do not
cloud the interpretation of the STI.'11121 This is particularly
true when marked changes in arterial pressure occur.
The effect of negative inotropic agents upon the STI have
been less extensively studied.122 In general, it appears that
such agents induce the typical heart failure pattern of pro-
longation of the PEPI and shortening of the LVETI. Many
antihypertensive drugs fall into this category but as yet there
are no major studies of the effects of these agents upon the
STI.
Only recently have studies appeared wherein the STI have
been used as a routine clinical test to improve drug therapy.
The STI appear to have great potential for evaluating
therapy of thyroid disorders,123 following patients receiving
VOL 56, No 2, AUGUST 1977
the cardiotoxic antineoplastic agent, adriamycin,lu2 125 and
for assessing the effect of propranolol therapy for angina
pectoris or acute myocardial infarction.126 127 In the case of
propranolol therapy, the STI not only provide information
concerning left ventricular function but also reflect the
degree (if any) of excessive adrenergic tone which is present.
Summary
The theoretical basis for the use of the systolic time inter-
vals has been largely established. The method has been vali-
dated by direct measures from within the circulatory sys-
tem. Standards for equipment and technique have been
defined. Numerous clinical studies have demonstrated the
value of this quantitative noninvasive technique for assess-
ing left ventricular performance. At present there is need for
further studies of the clinical usefulness of the systolic time
intervals to improve both diagnosis and therapy of various
cardiovascular disorders.
References
1. Weissler AM, Lewis RP, Leighton RF: The systolic time intervals as a
measure of left ventricular performance in man. In Progress in Car-
diology, edited by Yu PN, Goodwin JF. Philadelphia, Lea and Febiger,
1972, pp 155-183
2. Lewis RP, Leighton RF, Forester WF, Weissler AM: Systolic time in-
tervals. In Non-Invasive Cardiology, edited by Weissler AM. New
York, Grune and Stratton, 1974, pp 301-368
3. Lewis RP: Diagnostic value of systolic time intervals in man. In Diag-
nostic Methods in Cardiology, edited by Fowler NO. Philadelphia, FA
Davis, 1975, pp 245-264
4. Weissler AM, Harris WS, Schoenfeld CD: Systolic time intervals in
heart failure in man. Circulation 37: 149, 1968
5. Weissler AM, Harris WS, Schoenfeld CD: Bedside technics for the
evaluation of ventricular function in man. Am J Cardiol 23: 577, 1969
6. Garrod AH: On some points connected with the circulation of the
blood, arrived at from a study of the sphygmograph-trace. Proc R Soc
London 23: 140, 1874-5
7. Bowen WP: Changes in heart rate, blood pressure and duration of sys-
tole resulting from bicycling. Am J Physiol 11: 59, 1904
8. Wiggers CJ: Studies on the consecutive phases of the cardiac cycle. II.
The laws governing the relative durations of ventricular systole and
diastole. Am J Physiol 56: 439, 1921
9. Lombard WP, Cope OM: The duration of the systole of the left ven-
tricle of man. Am J Physiol 77: 263, 1926
10. Braunwald E, Sarnoff SJ, Stanisby WN: Determinants of duration and
mean rate of ventricular ejection. Circ Res 6: 319, 1958
11. Wallace AG, Mitchell JH, Skinner NS, Sarnoff SJ: Duration of the
phases of left ventricular systole. Circ Res 12: 611, 1963
12. Weissler AM, Harris LC, White GD: Left ventricular ejection time in-
dex in man. J Appl Physiol 18: 919, 1963
13. Willems J, Kesteloot H: The left ventricular ejection time. Its relation
to heart rate, mechanical systole and some anthropometric data. Acta
Cardiol 22: 401, 1967
14. Harris LC, Weissler AM, Manske AO, Banford BH, White GD, Ham-
mill WA: Duration of the phases of mechanical systole in infants and
children. Am J Cardiol 14: 448, 1964
15. Golde D, Burstin L: Systolic phases of the cardiac cycle in children.
Circulation 42: 1029, 1970
16. Slodki SJ, Hussain AT, Luisada AA: The Q-II interval. III. Study of
the second heart sound in old age. J Am Geriatr Soc 17: 673, 1969
17. Willems JL, Roelandt J, DeGeest H, Kesteloot H, Joosens JF: Left
ventricular ejection time in elderly subjects. Circulation 42: 37, 1970
18. McDonald DA: Blood Flow in Arteries. Baltimore, Williams and Wil-
kins Co., 1960
19. Rittgers SE: Analysis of the external carotid artery pulse wave and its
measurement as a noninvasive indication of cardiac function in the sys-
tolic time interval. Ohio State University, Mechanical Engineering De-
partment. Master's Thesis, 1974
20. Geddes LA, Baker LE: Principles of Applied Biomedical Instru-
mentation. New York, John Wiley & Sons, Inc., 1968, p 446
21. Bush CA, Lewis RP, Leighton RF, Fontana ME, Weissler AM: Verifi-
cation of systolic time intervals and the true isovolumic contraction
time for the apexcardiogram by micromanometer catheterization of
 SYSTOLIC TIME INTERVALS/Lewis et al.
the left ventricle and aorta. (abstr) Circulation 42 (suppl III): III-121,
1970
22. Martin CE, Shaver JA, Thompson ME, Reddy PS, Leonard J: Direct
correlation of external systolic time intervals with internal indices of
ventricular function in man. Circulation 44: 419, 1971
23. Van de Werf F, Piessens J, Kesteloot H, DeGeest H: A comparison of
systolic time intervals derived from the central aortic pressure and
from the external carotid pulse tracing. Circulation 51: 310, 1975
24. Moene RJ, Mook GA, Kruzinga K, Bergstra A, Bossina KK: Value of
systolic time intervals in assessing severity of congenital aortic stenosis
in children. Br Heart J 37: 1113, 1975
25. Brough RD, Talley RC: Temporal relation of the second heart sound
to aortic flow in various conditions. Am J Cardiol 30: 237, 1972
26. Luisada AA, MacCanon DM, Kumar S, Feigen LP: Changing views
on the mechanism of the first and second heart sounds. Am Heart J 88:
503, 1974
27. Anastassiades PC, Quinones MA, Gaasch WH, Adyanthaya AV,
Waggoner AD, Alexander JK: Aortic valve closure: Echocardio-
graphic, phonocardiographic and hemodynamic assessment. Am Heart
J 91: 228, 1976
28. Sherman JA, Lewis RP: Systolic time intervals in atrial fibrillation.
(abstr) Circulation 45, 46 (suppl II): 11-220, 1972
29. Metzger CC, Chough CB, Kroetz FW, Leonard J: True isovolumic
contraction time: Its correlation with two external indexes of ventric-
ular performance. Am J Cardiol 25: 434, 1970
30. Luisada A, MacCanon DM: The phases of the cardiac cycle. Am
Heart J 83: 705, 1972
31. Oreshkov V: Indirect measurement of isovolumetric contraction time
on the basis of polygraphic tracing (apexcardiogram, carotid tracing
and phonocardiogram). Cardiologia 47: 315, 1965
32. Baragon J, Fernandez-Caamano F, Sozutek Y, Coblence B, Lenegre J:
Chronic complete bundle branch block: Phonocardiographic and
mechanocardiographic study of 30 cases. Br Heart J 30: 196, 1968
33. Silei M: Etude Phonom6canocardiographique des blocs de branche: A
propos de 106 observations. Thesis. Imprimerie Bosc Fr6res, Lyon
1969
34. Adolph RJ, Fowler NO, Tanaka K: Prolongation of isovolumic con-
traction time in left bundle branch block. Am Heart J 78: 585, 1969
35. Santos DE, DeLapaz A, Pietras RJ, Tobin JR, Gunnar RM: The apex-
cardiogram in left bundle branch block. Br Heart J 31: 693, 1969
36. Baragan J, Tsrnandez F, Coblence B, Saad Y, Lenegre J: Left ventric-
Downloaded from http://ahajournals.org by on August 29, 2023
ular dynamics in complete right bundle branch block with left axis
deviation of QRS. Circulation 42: 797, 1070
37. Haft JL, Herman MW, Gorlin R: Left bundle branch block. Etio-
logic, hemodynamic and ventriculographic considerations. Circulation
43: 279, 1971
38. Waagstein F, Hjalmarson AC, Wasir HS: Apexcardiogram and sys-
tolic time intervals in acute myocardial infarction and effects of prac-
talol. Br Heart J 36: 1109, 1974
39. Manolas J, Rutishauser W, Wirz P, Arbenz U: Time relation between
apexcardiogram and left ventricular events using simultaneous high
fidelity tracings in man. Br Heart J 37: 1263, 1975
40. DiBartolo G, Nunez-Day D, Muiesan G, MacCanon DM, Luisada
AA: Hemodynamic correlation of the first heart sound. Am J Physiol
201: 888, 1961
41. Kesteloot H, Willems J, Van Vollenhoven E: On the physical prin-
ciples and methodology of mechanocardiography. Acta Cardiol 24:
147, 1969
42. Geddes LA: Measurement of blood pressure. Chicago, Year Book
Medical Publishers, Inc., 1970
43. Johnson JM, Siegel W, Blomquist G: Characteristics of transducers
used for recording the apexcardiogram. J Appl Physiol 31: 796, 1971
44. Spodick DH, Kumar S, Flessas AP, Sriratanaban GS, Klints RV,
Muench H: Relationship of paper speed to precision of pulse wave
measurements: A multiple observer study of left ventricular ejection
time. Aerospace Med 40: 707, 1969
45. Kyle MC, Freis ED: Computer identification of systolic time intervals.
Comp Biomed Res 3: 637, 1970
46. Zoneraich S, Zoneraich 0, Rodenrys J: Computerized system for non-
invasive techniques. I. Its value for systolic time intervals. Am J Car-
diol 33: 643, 1974
47. Greenfield JC Jr, Harley A, Thompson HK, Wallace AG: Pressure-
flow studies in man during atrial fibrillation. J Clin Invest 47: 2411,
1968
48. Tavel ME, Baugh DO, Feigenbaum H, Nasser WK: Left ventricular
ejection time in atrial fibrillation. Circulation 46: 744, 1972
49. Klingfield P: Systolic time intervals in atrial fibrillation and mitral
stenosis. Br Heart J 36: 798, 1974
50. Flessas AP, Pomposiello JC, Connelly GP, Kowsari M, Ryan TJ:
Determinants of systolic time intervals in the human heart. (abstr) Cir-
culation 43, 44 (suppl II): 11-165, 1971
51. Willems J, Kesteloot H: The left ventricular ejection time. Its relation
157
to heart rate, mechanical systole and some anthropometric data. Acta
Cardiol 22: 401, 1967
52. Kroetz FW, Leonard JJ, Shaver JA, Leon DF, Lancaster JF, Beamer
VL: The effect of atrial contraction on left ventricular performance in
valvular aortic stenosis. Circulation 35: 852, 1967
53. Heindenreich FP, Thompson ME, Shaver JA, Leonard JJ: Left atrial
transport in mitral stenosis. Circulation 40: 545, 1969
54. Harris LC, Nghien Q, Schreiber MH, Wallace JM: Severe pulsus
alternans associated with primary myocardial disease in children. Ob-
servations on clinical features, hemodynamic findings, mechanism, and
prognosis. Circulation 34: 948, 1966
55. Spodick DH, St Pierre JR: Pulsus alternans: Physiologic study by non-
invasive techniques. Am Heart J 80: 766, 1970
56. Weissler AM, Peeler RG, Roehll WH Jr: Relationships between left
ventricular ejection time, stroke volume and heart rate in normal indi-
viduals and patients with cardiovascular disease. Am Heart J 62: 367,
1961
57. Harley A, Starmer CF, Greenfield JC Jr: Pressure-flow studies in man:
Evaluation of the duration of the phases of systole. J Clin Invest 48:
895, 1969
58. Leighton RF, Zaron SJ, Robinson JL, Weissler AM: Effects of atrial
pacing on left ventricular performance in patients with heart disease.
Circulation 40: 615, 1969
59. Talley RC, Meyer JF, McNay JL: Evaluation of the pre-ejection
period as an estimate of myocardial contractility in dogs. Am J Car-
diol 27: 384, 1971
60. Weissler AM, Schoenfeld CD: Effect of digitalis on systolic time inter-
vals in heart failure. Am J Med Sci 259: 4, 1970
61. Garrard CL Jr, Weissler AM, Dodge HT: The relationship of altera-
tions in systolic time intervals to ejection fraction in patients with car-
diac disease. Circulation 42: 455, 1970
62. Ahmed SS, Levinson GE, Schwartz CJ, Ettinger PO: Systolic time in-
tervals as measures of the contractile state of the myocardium in man.
Circulation 46: 559, 1972
63. McConahay DR, Martin MM, Cheitlin MD: Resting and exercise sys-
tolic time intervals. Correlations with ventricular performance in pa-
tients with coronary artery disease. Circulation 45: 529, 1970
64. Lewis RP, Boudoulas H, Welch TG, Forester WF: Usefulness of sys-
tolic time intervals in coronary artery disease. Am J Cardiol 37: 787,
1976
65. Diamond G, Forrester JS, Chatterjee K, Wegner S, Swan HJC: Mean
electromechanical ,P/AT. On indirect index of the peak rate of rise of
left ventricular pressure. Am J Cardiol 30: 338, 1972
66. Bush CA, Lewis RP: Importance of preload in chronic left ventriculiar
disease. Circulation 48 (suppl VI): VI-151, 1973
67. Stafford RW, Harris WS, Weissler AM: The left ventricular systolic
time intervals as indices of postural circulatory stress in man. Circula-
tion 41: 485, 1970
68. Gault JH, Ross J Jr, Braunwald E: Contractile state of the left ventricle
in man. Circ Res 22: (4) 451, 1968
69. Harris WS, Schoenfeld CD, Brooks RH, Weissler AM: Effect of beta
adrenergic blockade on the hemodynamic responses to epinephrine in
man. Am J Cardiol 17: 484, 1966
70. Harris WS, Schoenfeld CD, Weissler AM: Effects of adrenergic recep-
tor activation and blockade on the systolic pre-ejection period, heart
rate and arterial pressure in man. J Clin Invest 46: 1704, 1967
71. Hunt D, Sloman G, Clark RM, Hoffmann G: Effects of beta-adrenergic
blockade on the systolic time intervals. Am J Med Sci 259: 97, 1970
72. Salzman SH, Wolfson S, Jackson B, Schecter E: Epinephrine infusion
in man. Standardization, normal response, and abnormal response in
idiopathic hypertrophic subaortic stenosis. Circulation 43: 137, 1971
73. Whitsett TL, Goldberg LI: Effects of levodopa on systolic pre-ejection
period, blood pressure, and heart rate during acute and chronic treat-
ment of Parkinson's Disease. Circulation 45: 97, 1972
74. Weissler AM, Snyder JR, Schoenfeld CD, Cohen S: Assay of digitalis
glycosides in man. Am J Cardiol 17: 768, 1966
75. Weissler AM, Lewis RP, Leighton RF, Bush CA: Comparative
responses to the digitalis glycosides in man. In Marks BH, Weissler
AM: Basic and Clinical Pharmacology of Digitalis. Fort Lauderdale,
Florida, Thomas, 1972, p 260
76. Lewis RP, Boudoulas H, Forester WF, Weissler AM: Shortening of
electromechanical systole as a manifestation of excessive adrenergic
stimulation in acute myocardial infarction. Circulation 46: 856, 1972
77. Armstrong TG, Lewis BS, Gotsman MS: Systolic time intervals in con-
strictive pericarditis and severe primary myocardial disease. Am Heart
J 85: 6, 1973
78. Khullar S, Lewis RP: Usefulness of systolic time intervals in differen-
tial diagnosis of constrictive pericarditis and restrictive cardio-
myopathy. Br Heart J 38: 43, 1976
79. Ahmed SS, Jaferi GA, Narang RM, Regan TJ: Preclinical abnor-
mality of left ventricular function in diabetes mellitus. Am Heart J 89:
153, 1975
 158 CIRCULATION
80. Spodick DH, Pigott VM, Chirife R: Pre-clinical cardiac malfunction in
chronic alcoholism. Comparison with matched normal controls and
with alcoholic cardiomyopathy. N Engl J Med 287: 677, 1972
81. McDonald I, Hobson ER: A comparison of the relative value of nonin-
vasive techniques - echocardiography, systolic time intervals, and apex
cardiography in the diagnosis of primary myocardial disease. Am Heart
J 88: 454, 1974
82. Alpert JS, Rickman FD, Howe JP, Dexter L, Dalen JE: Alteration of
systolic time intervals in right ventricular failure. Circulation 50: 317,
1974
83. Hooper RG, Whitcomb ME: Systolic time intervals in chronic obstruc-
tive pulmonary disease. Circulation 50: 1205, 1974
84. Toutouzas P, Gupta D, Samson R, Shillingford J: Q-second sound in-
terval in acute myocardial infarction. Br Med J 31: 462, 1969
85. Biamont B, Killip T: Indirect assessment of left ventricular per-
formance in acute myocardial infarction. Circulation 42: 579, 1970
86. Heikkila J, Luomanmake K, Pyorala K: Serial observations on left ven-
tricular dysfunction in acute myocardial infarction. II. Systolic time in-
tervals in power failure. Circulation 44: 343, 1971
87. Jain SR, Lindahl J: Apexcardiogram and systolic time intervals in acute
myocardial infarction. Br Heart J 33: 578, 1971
88. Hamosh P, Cohn JN, Engelman K, Broder MI, Fries ED: Systolic time
intervals and left ventricular function in acute myocardial infarction.
Circulation 45: 375, 1972
89. FabianJ, Epstein N, Coulshed N, McKendrick CS: Duration of phases
of left ventricular systole using indirect methods.II. Acute myocardial
infarction. Br Heart J 34: 882, 1972
90. Austin TW, Ahuja SP, Boughner DR: Atraumatic study of left ventric-
ular events following acute myocardial infarction. AmJ Cardiol 29:
745, 1972
91. Hodges M, Halpern BL, Friesinger GC, Dagenais GR: Left ventricular
preejection period and ejection time in patients with acute myocardial
infarction. Circulation 45: 933, 1972
92. Agress CM, WegnerS, Forrester JS, Chatterjee K, Parmley WM, Swan
HJC: An indirect method for evaluation of left ventricular function in
acute myocardial infarction. Circulation 46: 291, 1972
93. Nieminen M, Heikkila J: Echoventriculography in acute myocardial in-
farction. II. Monitoring of left ventricular performance. Br Heart J 38:
271, 1976
94. Stack RS, Lee CC, Reddy BP, Taylor ML, Weissler AM: Left ventric-
ular performance in coronary artery disease evaluated with systolic time
intervals and echocardiography. AmJ Cardiol 37: 331, 1976
Downloaded from http://ahajournals.org by on August 29, 2023
95. Johnson AD, O'Rourke RA, Karliner JS, Burian C: Effect of myocar-
dial revascularization on systolic time intervals in patients with left ven-
tricular dysfunction. Circulation 45 (supplI): I-1, 1972
96. Boudoulas H, Lewis RP, Vasko JS, Karayannacos PE, Beaver BM: The
role of adrenergichyperactivity in the early postoperative phase of
saphenous vein bypass. Circulation 53: 802, 1976
97. Battock OJ, Alvarez H, Chidsey CA: Effects of propranolol and isosor-
bide dinitrate on exercise performance and adrenergic activity in
patients with angina pectoris. Circulation 39: 157, 1969
98. Buyukozturk K, Kimbris D, Segal BL: Systolic time intervals. Relation
to severity of coronary artery disease, intercoronary collateralization
and left ventricular dyskinesia. Am J Cardiol 28: 183, 1971
99. Moskowitz RL, Wechsler BM: Left ventricular ejection time in aortic
and mitral valve disease. AmJ Cardiol 15: 809, 1965
100. Benchimol A, MatsuoS: Ejection time before and after aortic valve
replacement. Am J Cardiol 27: 244, 1971
101. Parisi AF, Salzman SH, Schecter E: Systolic time intervals in severe
aortic valve disease. Changes with surgery and hemodynamic correla-
tions. Circulation 44: 539, 1971
102. Sherman JA, Lewis RP, Fontana ME, Weissler AM: Use of left ventric-
ular ejection time in assessing the severity aortic
of stenosis. (abstr) Cir-
culation 43, 44: 225, 1971
103. Bache RJ, Wang Y, Greenfield JC
Jr: Left ventricular ejection time in
valvular aortic stenosis. Circulation 47: 527, 1973
VOL 56, No 2, AUGUST 1977
104. Bonner AJ, Sacks HN, Tavel ME: Assessing the severity of aortic
stenosis by phonocardiography and external carotid pulse recordings.
Circulation 48: 247, 1973
105. Saebra-Gomes R, Sulton R, Parker DJ: Left ventricular function after
aortic valve replacement. Br Heart J 38: 491, 1976
106. Curtis EI, Matthews RE, Shaver JE: Mechanism of normal splitting of
the second heart sound. Circulation 45, 46 (suppl II): 11-64, 1972
107. Cohn K, Hancock EW: Effect of amyl nitrite and isoproterenol on the
ejection time in hypertrophic subaortic stenosis. Circ Res 14: 123, 1966
108. Wigle ED, Auger P, Marquis Y: Muscular subaortic stenosis. The direct
relation between the intraventricular pressure difference and the left ven-
tricular ejection time. Circulation 36: 36, 1967
109. Carter WH, Whalen RE, Morris JJ, Orgain ES: Carotid pulse tracings
in hypertrophic subaortic stenosis. Am Heart J 82: 180, 1971
110. White CW, Zimmerman JJ: Prolonged left ventricular ejection time in
the post-premature beat. A sensitive sign of idiopathic hypertrophic sub-
aortic stenosis. Circulation 52: 306, 1975
111. Wanderman KL, Goldberg MJ, Stack RS, Weissler AM: Left ventric-
ular performance in mitral regurgitation assessed with systolic time in-
tervals and echocardiography. Am J Cardiol 38: 831, 1976
112. Forester WF, Lewis RP, Weissler AM: The onset and magnitude of the
contractile response to commonly used digitalis glycosides in normal
subjects. Circulation 49: 517, 1974
113. Levy AM, Leaman DM, Hanson JS: Effects of digoxin on systolic time
intervals of neonates and infants. Circulation 46: 816, 1972
114. Gould L, Gamprect RF, Jaynal F: The effects of phentolamine on the
duration of the phases of ventricular systole in man. Am J Med Sci 259:
260, 1970
115. Byrne MJ, Pigott V, Spodick DH: Cardiovascular responses to
glucagon: Physiologic measurement by external recordings. Am HeartJ
83: 635, 1972
116. Beaver BM, Boudoulas H, Kates RE, Lewis RP: Different time course
of the inotropic and chronotropic effects of propranolol. (abstr) Clin Res
24: 516, 1976
117. Kramer WG, Cobb TC, Reuning RH, Forester WF, Lewis RP: Phar-
macodynamic studies of digoxin employing simultaneous blood levels
and systolic time intervals. Clin Res 21: 814, 1973
118. Bush CA, WebelJ, Leier CV: Treatment of end stage cardiac failure
with dobutamine. (abstr) AmJ Cardiol 37: 125, 1976
119. Shaver JA, Kroetz FW, Leonard JJ: The effect of steady state increases
in systemic arterial pressure on the duration of left ventricular ejection
time.J Clin Invest 47: 217, 1968
120. Leighton RF, Polumbo RA, Zaron SJ, Robinson JL: The use of systolic
time intervals in predicting hemodynamic effects of isoproterenol. Clin
Res 18: 371, 1970
121. Sawayama T, Masakazu 0, MarumatoS, Matsuura T, NikiI:
Influence of amyl nitrite inhalation on the systolic time intervals in nor-
mal subjects and in patients with ischemic heart disease. Circulation 40:
327, 1969
122. Harris WS, Weissler AM, Brooks RH: Vagal influence on left ventric-
ular contraction in man: A cause for prolongation of the pre-ejection
period. (abstr) Clin Res 14: 426, 1966
123. Parisi AF, Hamilton BP, Thomas CN, Mazzaferri EL: The short car-
diac pre-ejection period. An index to thyrotoxicosis. Circulation 46: 900,
1974
124. Rinehart J, Lewis RP, Balcerzak SF: Adriamycin cardiotoxicity in
man. Ann Int Med 81: 475, 1974
125. Bristow MR, Mason JW, Bellingham ME, Daniels JR: Endomyo-
cardial biopsy and systolic time intervals in detection of adriamycin car-
diotoxicity. (abstr) Circulation 54 (suppl II): 11-80, 1976
126. Frishman W, Smithen C, Befler B, Kligfield P, Kiup T: Noninvasive
assessment of clinical response to oral propranolol therapy. Am J Car-
diol 35: 635, 1975
127. Waagstein F, Hjalmarson AC, Wasir HS: Apex cardiogram and
systolic time intervals in acute myocardial infarction and effects of prac-
talol. Br Heart J 36: 1109, 1974