1954 Friedman EA. The Graphic Analysis of Labor

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

THE GRAPHIC ANALYSIS OF LABOR

EMANUEL A. FRIEDMAN, M.D., NEW YORK, N.Y.


(From the Department of Obstetrics and Gynecology, College of Physicians and SuT[Jeons,
Columbia University, and the Sloane Hospital for Women, Columbia-Presbyterian Medical
Center)

N AN effort to evaluate the effects of various factors upon the course of


I labor, a simple, reproducible, and relatively objective method of recording
and comparing progressive changes was sought. Of the major observable
t'vents that occur during labor, i.e., force, frequency, and duration of uterine
contractility, descent of the presenting fetal part and cervical effacement
and dilatation, only the last-named was selected for detailed study because
it seemed to parallel over-all progress best. A general mathematical ex-
pression was derived based upon the graphic portrayal of changes in dilata-
tion of the cer~ix with time. The curves obtained in all normal cases studied
were near-identical S curves, varying only in slope. The study is presented
because of its unique simplicity and ready adaptability to the study of labor.
In a review of a half century of obstetrical literature, it is noted that
numerous methods of objective study of labor have been devised and put to
brief or prolonged use. Calkins and associate.s 3 - 6 used the clinical evaluation
of cervical resistance (graded according to the effacement, dilatation, and
softness at the onset of labor) and of "motive foree" (intensity and fre-
quency of uterine contraetions) in an effort to predict the expected tot11l
duration of the first stage. A rough rule was derived. It was concluded that
"more aecurate observation of the resistance of the cervix (and the pelvic
floor), a.s well as a more accurate determination of the effectiveness of the
labor pains, will be necessary in order to analyze the causes for the extreme
,-ariations in the lengths of labor so eommonly encountered.'' The frequency
and duration of contrartions have been studied clinically. 11 A labor was
noted to be a function of the duration and number of contractions necessary
for its evolution. Neither method permitted accurate evaluation en passant.
Koller and Abt 14 , 1 " and Abt 1 utilized a cervical dilatation-time function
to demonstrate the effect of rupture of the membranes upon the course of
labors. The graphs presented were not true representations of progressive
change because of the nonlinear ordinate scale, the divisions of which were
unequal. 'The divisions represented 1 to 2 fr., 5 fr., "small palm" (circa
fi em.). ''palm'' (circa 8 em.), full dilatation, delivery of the infant and
of the placenta, respectively, this despite Liepman 's16 plea for standardization
of cervical dilatation nomenclature. The zero of the abscissa time was taken
as the time of rupture of membranes. By superimposing curves aligned at
1568
Volume 68 GRAPHIC ANALYSIS OF LABOR li169
Number A

zero time it could be demonstrated grossly that dispersion of curves before


and after rupture of membranes is different. Linear change was assumed and
analysis was concerned with average slope and with maximum d<~livery time.
The latter allo,~ved but a rough tneasure of control in the eyaluation of a
labor in progress, and then only near its termination.
Apparently, the problem of the effect of amniotomy was foremost, many
studies 2 8 9 ' 12 19 having been concerned with it; and the task of establishing
a "norm" for the course of labors, with which comparisons may be made, was
ignored. Zimmer 21 employed much the same method as has been used in the
present study, but, again, with reference to time of rupture of membranes.
The curves obtained were hyperbolic, their sigmoid characteristics having
been overlooked. A marked change in slope beyond 3 to 4 em. dilatation was
recognized, hut alteration of the curve by dystocia or inertia was not
appreciated.
Other studies involved cumbersome equipment, some mme simple an<l
thereby often less objective. Some, still under swuy, analyze electrical or
mechanical effects of individual contractions. It is not the purpose of this
discussion to evaluate these studies. It is believed however, that they may
border on the frontier of ohstetrical progress. This presentation is put forth
nevertheless to provide a rea listie tool for the study of individual labors. in
progress, by obstetricians outside of uninrsity hospitals.

Methods and Materials


The patients studied were all primigravidas at term who presented them-
selves sufficiently early in their labors to permit adequate study. With few
exceptions they all instituted labor spontaneously and despite a lack of regard
for cephalopelvic relationship in selection nearly all delivered vaginally with
vertex presentation. A series of cases, the first 100 of which are reported
here, was accordingly studied in considerable detail. There were 29 spon-
taneous, 64 prophylactic low forceps and 4 midforceps deliveries. The mid-
forceps procedures were ascribed to relative cephalopelvic disproportion,
secondary inertia, transverse arrest, and fetal distress, respectively. The one
cesarean section was performed because of arrested progress after prolonged
trial of labor including three hours of second stage. One frank breech
presentation is included; the delivery was uncomplicated. There was one
multiple pregnancy in this series, the first twin delivered by prophylactic low
forceps, the second by version and breech extraction because o cord presenta-
tion. All infants were liveborn. One neonatal death occurred on the first
day following a precipitate labor and an uncomplicated spontaneous delivery.
Death was attributed to a large subdural hematoma found on postmortem
examination. There were 3 cases of primary inertia, and 7 of clinically
authenticated secondary inertia. Pitocin infusion was used in attempts to
induce 4 labors: for pre-eclampsia in 3 instances, and ruptured chorioamniotic
membranes for a prolonged period of time at term in one. In addition
Pitocin was invoked to stimulate 5 labors already in progress, but clinically
thought to be slowing. Pitocin was employed, likewise, in the 10 documented
cases of inertia. Episiotomy was performed almost universally except in 2
cases, one sustaining a second degree perineal laceration, the other remaining
intact. Twenty-two of the patients received caudal anesthesia throughout the
greater portion of their labors.
1570 FRIEDMAN Am. ]. (Jbst. & Gyucc.
December, 1954

Cervical dilatation was determined by frequent rectal examinations.


Although the rectal route of determining the status of the cervix has not been
universally accepted as a proper method of following the course of labor,
with minimal training considerable accuracy may be obtained. Where any
aoubt existed, particularly with very soft or very thin cervices, sterile vaginal
examinations were performed. It is to be noted, parenthetically, that deter-
minations are of greatest accuracy at the onset and the termination of dilata-
tion (less than 3 em. and greater than 7 em.). Changes usually occur so
rapidly in the mid-portion of the curve that greater accuracy than that asso-
ciated with the oft-quoted 10 per cent error 17 is not essential. Almost all the
examinations recorded were performed by one individual throughout any one
labor. This was done intentionally in order to eliminate the variability of the
determination of cervical dilatation due to the differences in interpretation.
All readings therefore became relative. For reproducibility, it was felt that
examination should be made at the peak of a contraction, particularly where
there may be significant difference between the contractile and quiescent
states. Dilatation was measured in centimeters for uniformity. The fre-
tluency of examinations required depended upon the progress of labor. Often
half-hourly examinations were insufficient to depict accurately the slope of a
precipitous labor. Usually, however, hourly or bihourly examinations sufficed
amply. No attempt was made to alter the routine labor care as performed
at the Sloane Hospital for Women. Thus enemas, sedation, regional and
inhalation anesthesias, Pitocin infusions, etc. were administered, and amniot-
omy performed as indicated.
The technique, in practice, was quite simple. Square-ruled graph paper
was used with 10 divisions along the ordinate to represent the cervical dilata-
tion in centimeters; each corresponding division along the abscissa denoted
time in hours. The readings as obtained were entered and joined to the
preceding notation by a straight line. The slope of each line was determined
readily in terms of centimeters of dilatation per hour. For example, if the
cervix dilates from 2.0 to 3.5 em. in 2 hours, the slope is (3.5- 2.0) ..;- 2 = 0.75.
Any medication given or procedure performed, and the time of rupture
of membranes was entered. Vaginal examinations were especially marked to
indicate the greater accuracy of these points. The progressive changes of the
rough sigmoid curve obtained were thus available for immediate study.
Major variations were apparent even to the uninitiated observer. Minor
deviations from the "normal" could be detected with a minimum of effort.

Results
Onset of labor is taken according to the classic definition as that time
when regular uterine contractions have been established. This is found in
the cases studied to be from 1.7 to 15 hours before the onset of appreciable
cervical dilatation. This period has been designated as the latent period or
phase one of the first stage of labor. The mean duration of this period is 7.3
hours and a few spurious readings, apparently of normal labors in all other
respects, have been noted with total first phase (well documented) up to 23.2
hours. It is during this interval that the myometrial contraction becomes
oriented and the cervix undergoes changes preparatory to dilatation, such
as softening and effacement. There may be no significant change of the
cervix noted; or, as occurred in most cases, very slow dilatation may occur.
The latter progresses slowly in a linear fashion until 2 to 2.5 em. dilatation
is reached, at which time phase two is entered. Phase two, or the acceleration
period, is marked by a rapid change in the slope of the cervical dilatation
curve. The upswing is apparent on the graphic records. This is the period
Volume 68 GRAPHIC ANALYSIS OF LABOR 1571
Number 6

of a continuously changing and increasing rate of dilatation. Phase three is


entered 'vhen the maximun1 slope is reached. This steady period of rapid
cervical dilatation progresses in a linear fashion for a variable length of time.
It is during this most important phase that the moment-to-moment effects of
''interference,'' such as medication, anesthesia, Pitocin, etc., can be observed.
This linear dilatation period extends from 3 or 3.5 to 8.5 or 9 em. cervical
dilatation. As the first stage of labor terminates, the fourth phase or de-
celeration period is entered. The slope of dilatation once again slows as full
dilatation is reached. This period is analogous to phase two, of acceleration,
in that the slope is changing. Here, however, it is diminishing from the
rapid slope of phase three to the zero slope of the second stage. (Obviously
no further detectable change in cervical dilatation is possible following full
dilatation and retraction of the cervix).
Conclusions
The curves obtained by the simple expediency of plotting cervical
dilatation against time were all quite similar in general shape, resembling
sigmoid curves (Fig. 1). The main variation appeared in phase three, repre-
senting the maximal slope. Where this was steeply inclined (i.e., during rapid
cervical change), the total labor tended to be short; where more flat (slower
change), the labor was prolonged. The mathematical analysis of this type
of curve is quite complex* and has been considerably simplified by examining
the phases separately. In essence, the following general statements have been
derived.
First, the slope of the latent period is flat or nearly flat, slope 0.0 to 0.86
with a mean slope of 0.35 em. per hour, standard deviation 0.20t, and as
previously noted it varies in duration from 1.7 to 15.0 hours, mean duration
7.3 hours, standard deviation 5.5. From this alone it can be seen that a
diagnosis of primary inertia is in grave doubt if made solely on the basis of
lack of cervical dilatation after the onset of clinically good labor within this
period. The duration of this phase is apparently quite sensitive to inter-
ference, prolonged with heavy sedation, and shortened with stimulation. This
latter statement is as yet a clinical impression and has not been proved
statistically. The duration of this prodromal period has no apparent bearing
*The curves resemble those first described by Gompertz in 1825.' 13 represented by
y == abx, and applied as a growth curve in business and population problems. The modified
Gompertz equation, y cr + abx, takes into account the variability of the latent phase. The
complexity of this equation and the tedium of calculating the four constants for each curve
make small its practical use in the present study.
In order to diminish the mathematical complexity somewhat, transformation to linear
functions may be accomplished by using the method of probit analysis developed by Finney10
for the statistical treatment of the sigmoid curve. "The probit proportion P is defined as the
abscissa which corresponds to the probabll!ty P in a normal distribution with a mean 5 and
variance 1."
P - __l_J Y- 5 - Yo u'
- v:r:;;- _ e du.
00
The effect of this transformation from percentage to probit is to convert the normal
sigmoid curve to a straight line on a linear scale of probits. This may be considered as a
stretching of the percentage scale (per cent cervical dilatation) during which process the
sigmoid curve becomes straightened. In essence, the probit is no more than a convenient
mathematical device for solving otherwise Intractable equations. Its application here Is
limited by the asymmetry of the curves obtained.

~
tStandard deviation a = '\In
1572 FRIEDMAN Am. J. Obst. & Gynec.
December, 1954

on the future course of labor, i.e., there is no statistical correlation between


the length or slope of the latent period and the remaining portion of the
labor.* In consideration of the data presented, however, a latent period
greater than fifteen hours is significant, although a presumptive diagnosis of
primary uterine inertia cannot be made.
Second, phase two is short and variable but is very important in <.leter-
mining the ultimate outcome of the labor in question. It leads from the
minimum slope of the latent period to the maximum slope of the third phase.
A slow acceleration generally presages a lower maximum slope and therefore
a prolonged total labor; a rapid change precedes a short labor. From this it
follows that the onset of labor should be measured from the onset of phase
two since the ensuing changes are interrelated and apparently mutually
determinant. The incorporation of the antecedent time elapsed, i.e., the
latent period, gives an erroneous impression of the uterine work involved and
does not permit comparisons among labors. Thus a twenty-four hour labor
may consist of a twenty hour latent period and a short terminal period of
activity; or a latent period of eight hours may he followed by a slow to
average labor of sixteen hours. Equating these two labors is obviously m
error.

l
NORMAL LABOR

I0
~"'
cZ
"i
~ 8' J"
OJ
z
!;?
I
.....
<EO
:lO
""'
.. !
~
OU>

!
6-
~
1
q
J
0
~ ::;
... ::J
~
0
...
0

~ 2

10 12 14 16
TIME, h,.. CAS~ 2

Fig. 1.-Normal labor (No. 704263). Primipara at term. Latent phase 7.0 hours, slope
0.3 em. per hour; active phase 8.5 hours, maximum slope 1.1 em. per hour. The sigmoid
character of a normal labor is apparent.

Third, phase three, of the maximal constant slope, is inversely related


to the total duration of the first stage of labor. Since the major portion of
the cervical change occurs during this period (3.5 to 8.5 em.), it is apparent
that this relationship must exist. The essential linearity of progressive dila-
tion reflected in an unchanging slope allows for quick appraisal of deviations.
It is during this portion of the first stage of labor that short-term effects of
various factors may best be observed. Primary inertia, redefined, is detected
by a low overall slope, cervical dilatation occurring quite slowly, but never-
l:xY
*Coefficient of correlation r ::::
,; z x' Zy 2
Volume 68
Number 6
GRAPHIC ANALYSIS OF LABOR 1573

theless progressing along a normal sigmoid curve (l''ig. 2). Secondary inertia
is reflected in a deceleration of the slope prior to that expected (i.e., before
8.5 em.), the preceding portion of the curve having been normal (Fig. 3).
The flattening of the curve prematurely is readily detected and should alert
lhe obstetrician to seek the cause, whether cephalopelvic disproportion, inertia,
eervical dystocia, excessive medication, exhaustion, or a combination of
[actors. This prompt detection of arrested (or arresting) labor should prove
of ronsiderah]e value.

PRIMARY INERTIA

.
10 w ~w w
~~ cZ z
0~ "i
~.
~
~ ~4
4 o~ 0~

g8 ~
:>
z
g ~~
~0
~:?
~D
:i'D
w
""' "
,; w wu wu
D
"' ""' <I>

~ !
0
~
6
~
0
~

~ ~ H n ~ M 9
TIME, hrs.' C/JS"'57

Fig. 2.-Primary inertia (No. 107138). Primipara at term. with a "primarily inert"
labor. Latent phase ? 21.5 hours. slope 0.1 em. per hour: active phase 17.0 hours, maximum
slope 0. 7 em. per hour. Although the curve is sigmoid in general shape, the abscissa scale is
C'onsiderably expanderl.

SECONDARY INERTIA

Fig. 3.-Secondary inertia (No. 124003). Primipara at term. The active phase was
entered after a somewhat prolonged latent period of 17.4 hours. A maximum slope of 0.6 em.
per hour had been reached when the labor became. inert, flattening the curve prematurely.
PitoC"in infusion subsequently re-established good labor, maximum slope 2.3 em. per hour.

Fourth, the final phase of the first stage of labor reflects best the feto-
pelvic relationship in that cervical retraction about the fetal head is ap-
parently essential in obtaining full dilatation. In simultaneous plottings of
station on these graphs, it is apparent that the major portion of descent of the
fetal presenting part takes place in this fourth phase of the first stage and
1574 FRIEDMAN Am. ]. Obst. & Gynec.
December, 1954

during the second stage of labor. 'l'he descent accompanies full dilatation and
retraction of the cervix, whether cause or effect is conjectural. Prolongation
of the deceleration period has been observed in association with high degrees
of relative cephalopelvic disproportion. Quite often the terminal phase of the
first stage is short or absent (the latter probably merely unobserved). The
deceleration does not closely parallel the acceleration of phase two, asserting
that an independent function is active.
Fifth, the total duration of the a.ctive phase of cervical dilatation (phases
two, three, and four, inclusive) is statistically well correlated with the slope of
phase three. In contrast, there is no apparent correlation between the slope of
phase three and the total duration of the first stage, nor with the total labor.
Deducting the duration of the latent period from the total length of labor gives
far smaller figures for the average duration of primiparous labors in this series
than those usually quoted. 3 18 The length of the active phase was noted to be
from 1.8 to 9.5 hours with a mean of 4.4 hours, standard deviation 1.9. The
figures for the slope of phase three ranged from 0.7 for the relatively inert labor
to a maximum of 6.0 for a precipitous type of labor, with a mean of 3.7 em. per
hour, standard deviation 2.1. The difference between the mean slopes of the
latent (0.35) and of the active phases (3.7) is statistically significant with prob-
ability of less than 3 per thousand.* As mentioned before, however, there is no
statistical correlation between the relative slopes; nor is there a correlation be-
tween the respective durations.
Sixth, study of the second stage of labor by this method is, of course, not
:feasible. Here the total duration was merely noted and, as in previous studies, 4
no statistical relationship could be established between the duration of the total
or any part of the first stage and the length of the second stage. The great
variability of the latter depends to a considerable degree on the cephalopelvic
relationship, the intensity and frequency of contractions, and patient coopera-
tion in voluntary and involuntary expulsive efforts, as well as on the accoucheur's
desires regarding instrumental delivery and episiotomy. The second stage,
therefore, is not a matter for consideration here. Its management is left as a
clinical art.
This has been a preliminary report based upon the analysis of our first one
hundred cases. No attempt has as yet been made to assess the effects of the
various :factors mentioned upon the course o:f labor, but we expect to be able to
accomplish this in the future by utilizing this method. What we have done is to
redefine labor in terms of a new dimension (slope), viewing labor as a dynamic
process, setting time limits solely on the basis of previous activity, and, finally,
demonstrating what may be expected of a normal labor.
Summary
1. The efficacy of plotting cervical dilatation against time as a method
of graphically analyzing labor is demonstrated.
*Standard error of the difference between two means "d"' = ya, + "'' where "' and "'
are standard errors. A difference three times the standard error of difference is significant
(P = 0.008).
Volume 68 GRAPHIC ANALYSIS OF LABOR 1575
Number 6

2. 'fhe sigmoid characteristics of the curves thus obtained in primiparas


at terrn are illustrated.
3. 'l'he first stage of labor is divided into four pha.ses based on the graphic
alterations noted. Their respective limits are defined.
4. Primary and secondary inertia are redefined in terms of deviation from
the ''normal.''
5. The application of this method to the study of the effects of sedation,
stimulation, anesthesia, etc., upon the course of labor is indicated.
6. The value of this method for the study of individual labors, in progress,
is stressed.
I should like to express my sincere appreciation to Dr. D. A. D'Esopo for his interest
and continued encouragement, and to the Resident and Nursing Staffs of the Sloane
Hospital for Women for their cooperation during this study.

References
1. Abt, K.: Gynaecologia 132: 1, 65, 1951.
2. Basset, R.: Ztschr. Geburtsh. u. Gynak. 73: 566, 1913.
3. Calkins, L. A., Irvine, Jed H., and Horsley, Guy W.: AM. J. 0BST. & GYNEC. 19: 294,
1930.
4. Calkins, L.A.: AM. J. OBST. & GYNEC. 22: 604, 1931.
5. Calkins, L. A.: AM. J. OBST. & GYNEC. 27: 349, 1934.
6. Calkins, L. A.: AM. J. OBST. & GYNEC. 42: 802, 1941.
7. Davis, D. S.: Empirical Equations and Nomography, New York, 1943, McGraw-Hill
Book Company, Inc.
8. Eisenhart, H.: Arch. f. Gynak. 35: 386, 1889.
9. Essen-Moiler, M. E.: Acta obst. et gynec. scandinav. 16: 1, 1930.
10. Finney, D. J.: Probit Analysis, Cambridge, England, 1952, Cambridge University
Press.
11. Geisendorf, W.: Gynec. et obst. 35: 355, 1939.
12. Gianella, C.: Ztschr. Geburtsh. u. Gynak. 99: 227, 1931.
13. Kenney, J.: ~/!:athematics of Statistics, New York, 1939, D. Van Nostrand.
14. Koller, T., and Abt, K.: Gynaecologia 130: 419, 1950.
15. Koller, T., and Abt, K.: Rev. fran~. gynec. et obst. 47: 163, 1952.
16. Liepman, W.: Zentralbl. Gynak. 45: 1289, 1921.
17. Mayes, H. W.: Surg., Gynec. & Obst. 55: 771, 1932.
18. Spiller, V.: Lancet 2: 733, 1939.
19. VanderHoeven, P. Q. T.: Ztschr. Geburtsh. u. Gynak. 70: 1, 1912.
20. VanderHoeven, P. C. T.: Ztschr. Geburtsh. u. Gynak. 73: 826, 1913.
21. Zimmer, K.: Arch. Gynak. 179: 495, 1951.

You might also like