Original Research ajog.

org

OBSTETRICS
Tension compared to no tension on a Foley transcervical
catheter for cervical ripening: a randomized controlled trial
Gary Fruhman, MD; Jeffrey A. Gavard, PhD; Erol Amon, MD, JD; Kathleen V. G. Flick, MD;
Collin Miller, MSW; Gilad A. Gross, MD

BACKGROUND: Cervical ripening of an unfavorable cervix can be
achieved by placement of a transcervical catheter. Advantages of this
method include both lower cost and lower risk of tachysystole than other
methods. Despite widespread use with varying degrees of applied tension,
an unanswered question is whether there is an advantage to placing the
transcervical catheter to tension compared with placement without
tension.
OBJECTIVE: The purpose of this study was to determine whether
tension placed on a transcervical balloon catheter that is inserted for
cervical ripening results in a faster time to delivery.
STUDY DESIGN: This was a prospective, randomized controlled trial;
140 women who underwent cervical ripening (Bishop score,
6) were
assigned randomly to a balloon catheter with applied tension vs no tension.
Tension was created when the catheter was taped to the patient’s thigh
and tension was reapplied in 30-minute increments. There were 67 pa-
tients in the tension group and 73 patients in the no tension group. Low-
dose oxytocin (maximum, 6 mU/min) was administered after catheter
placement. The primary outcome was time from catheter insertion to
delivery. A secondary outcome was time from insertion to catheter
expulsion. The Kolmogorov-Smirnov test was used to determine whether
the data were distributed normally. Survival curves that used lifetables
were constructed from time of catheter insertion to delivery and from time
of catheter insertion to catheter expulsion and were compared with the use
of the Wilcoxon (Gehan) Breslow statistic. A probability value of <.05 was
set to denote statistical significance.
RESULTS: Baseline characteristics were similar between groups. The
median time from catheter insertion to delivery was not significantly
different between the tension group and the no tension group (16.2 vs
16.9 hours; P¼.814). The median time from catheter insertion to
expulsion, however, was significantly less in the tension group vs the no
tension group (2.6 vs 4.6 hours; P<.001), respectively. Vaginal delivery
within 24 hours was not significantly different between the tension and no
tension groups (41/52 [79%] vs 37/52 [71%]; P¼.365) nor were there
significant differences in cesarean delivery rates between the tension and
no tension groups (17/67 [25%] vs 27/73 [37%]; P¼.139).
CONCLUSION: Application of tension did not result in faster delivery
times but did result in faster times to catheter expulsion.
Key words: cervical ripening, Foley bulb, induction of labor, tension,
transcervical catheter

F rom 1990-2010 the labor induction

rate in the United States increased
from 9.6-23.3%.1 A cervical examination
that is performed before the induction of
labor helps determine whether the cervix
is favorable for a successful vaginal de-
livery. The Bishop score is a measure of
favorability of the cervix for induction.2
A score of
6 suggests that the proba-
bility of vaginal delivery with labor
induction is less than spontaneous
labor, whereas a score of 8 suggests
the likelihood of a vaginal delivery is
similar to spontaneous labor.3 Cervical
ripening helps prepare the cervix when
the Bishop score is
6. Ripening agents
include prostaglandins E2 and E1 and
Cite this article as: Fruhman G, Gavard JA, Amon E,
et al. Tension compared to no tension on a Foley trans-
cervical catheter for cervical ripening: a randomized
controlled trial. Am J Obstet Gynecol 2017;216:67.e1-9.
0002-9378/$36.00
ª 2016 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.ajog.2016.09.082
mechanical agents such as laminaria and
transcervical balloon catheters.
There are advantages and disadvan-
tages to different methods of cervical
ripening. Compared with the trans-
cervical catheter, prostaglandin E2 or E1
agents can be costly and cause more
uterine tachysystole with or without
fetal heart rate changes.4 Furthermore,
prostaglandin E1 is considered contra-
indicated in patients with previous
uterine scars because of an increased
risk of uterine rupture.5,6 One study
suggested that transcervical catheter
ripening can be done safely in the
outpatient setting.7
The use of transcervical catheters for
cervical dilation was first described in the
mid 1800s; however, because of safety
concerns, this method was abandoned.8
The first paper in the modern era uti-
lizing a transcervical catheter for cervical
ripening was published in 1967.9 Most
of the modern literature that has
compared transcervical catheters with
other methods of cervical ripening has
been published in the past 15 years. Past
studies compared the efficacy of trans-
cervical catheters with prostaglandin E2
agents,10-14 E1 agents,15-19 F2a agents,20
or oxytocin.21 Other studies addressed
the utility of adding oxytocin22-24 or
prostaglandins25-28 concurrently with
transcervical catheter placement or
compared transcervical catheters with
or without extraamniotic saline solu-
tion.29-32 Additional trials addressed the
amount of fluid instilled into the cath-
eter balloon,33,34 the efficacy of stylette
use with catheter placment,35 and
infection rates of different methods.36
The ideal method of the use of a trans-
cervical catheter remains controversial.
The fundamental question addressed
in this study is whether tension placed
on the transcervical catheter results in
faster times to delivery. In our institu-
tion, the use of tension with a trans-
cervical catheter for cervical ripening is
not a uniform practice. We hypothesized
that applying continuous tension by
taping the transcervical balloon catheter

JANUARY 2017 American Journal of Obstetrics & Gynecology 67.e1

Original Research OBSTETRICS
to the patient’s thigh would result in
shorter delivery times compared with a
transcervical catheter placed without
any tension. We performed a literature
search using both Google and PubMed
search engines with the use of the terms
cervical ripening, induction of labor,
transcervical catheter, Foley catheter, ten-
sion, traction, and thigh taping and found
our study to be unique.
Materials and Methods
This study is a randomized controlled
trial that took place at a single perinatal
center from February 2015 to February
2016. The Institutional Review Board
of Saint Louis University School of
Medicine approved this protocol. This
trial is registered with clinicaltrials.gov
(NCT02606643).
We included patients with a Bishop
score of
6 with a singleton, cephalic
gestation who had been admitted for
induction of labor. Exclusion criteria
included any medical conditions that
precluded vaginal delivery, cho-
rioamnionitis, vaginal bleeding, intra-
uterine fetal death, a low-lying placenta,
previous cervical surgery, or latex allergy.
Rupture of membranes was not an
exclusion criterion, although it was up to
each provider to decide whether to place
a catheter in this situation.
After the patient was admitted, the
house staff, under attending physician
supervision, selected the initial method
of cervical ripening. The most common
noncatheter method of cervical ripening
was prostaglandin E1. It was at the
discretion of the clinician when to place
the transcervical catheter. A transcervical
catheter was not placed if the Bishop
score was >6. Once the decision was
made to use a transcervical catheter and
the patient was not in labor, she was
approached for study participation and
randomization. Initially, group alloca-
tion occurred by a random pull from a
pile of opaque, sealed envelopes that
contained a group selection. After the
recruitment of the first 23 patients,
randomization was achieved by a
random number generator to assign
group allocation. After informed con-
sent was obtained from each participant,
an 18F 30-mL Foley bulb was placed
67.e2 American Journal of Obstetrics & Gynecology JANUARY 2017
digitally by the house staff, with or
without a stylette, until it was presumed
to be above the internal os. The balloon
was filled with 50 mL of saline solution,
because this is the standard at our insti-
tution. Patients who were assigned
randomly to tension had the catheter
taped with applied tension to the inner
thigh. Tension was replaced approxi-
mately every 30 minutes, as needed.
Patients who were assigned randomly to
no tension did not have any tension
applied to their catheter. After random-
ization, clinicians were not blinded to
the allocated group. If the catheter was
not expelled after 12 hours, the balloon
was deflated, and the catheter was
removed. On catheter insertion, an
oxytocin infusion was started at 1e2
mU/min and increased by 1e2 mU/min
every 20e30 minutes to a maximum of 6
mU/min while the catheter was in place.
After catheter removal, the remainder of
the patient’s labor course was managed
according to provider preference. All
patients had continuous fetal heart rate
and uterine activity monitoring.
Our primary outcome was the time of
catheter placement to the time of de-
livery. Secondary outcomes included the
time of placement to catheter expulsion,
vaginal delivery within 24 hours, cesar-
ean delivery rates, pain scores while the
catheter was in place, the amount of
oxytocin infused while the catheter was
in place, and chorioamnionitis. For this
study, chorioamnionitis was defined
strictly as a combination of maternal
temperature at >100.3 F, maternal pulse
at >100 beats per minute, and fetal
tachycardia (>160 beats per minute),
before or during the time of maternal
fever, with no other identifiable source
for fever or tachycardia. Pain scores were
assessed throughout labor; a pain was
scored based on a scale of 1e10, 10 being
the worst pain the patient ever
experienced.
The sample size calculation was based
on a mean time to delivery of 24.09.6
hours in the no tension group, an hy-
pothesized 20% reduction to 19.29.6
hours in the tension group, an alpha level
of .05, and a power of 80%. These as-
sumptions required a sample size of 63
patients per group, for a total of 126
ajog.org
patients. Mean time to delivery and
standard deviation estimations were
based on previously published
data.14,24,37
Chi-square tests and Fisher’s Exact
tests were used to assess differences in
categoric variables between the 2 groups.
The Kolmogorov-Smirnov goodness of
fit test was used to assess for normality.
Independent Student t tests and the
Kolmogorov-Smirnov statistic (for
nonparametric data)38 were used for
continuous variables, depending on the
normality of the distributions. Survival
curves with the use of lifetables were
constructed from time of catheter
insertion to delivery and from time of
catheter insertion to catheter expulsion
and were compared with the use of the
Wilcoxon (Gehan) Breslow statistic. A
probability value of <.05 initially was set
to denote statistical significance. Data
were analyzed by “intent-to-treat” and
“per protocol.” All analyses were per-
formed with SPSS software for Windows
(version 21.0; SPSS Inc, Chicago, IL). A
planned interim analysis was performed
for safety and efficacy at the midpoint of
patient recruitment. The primary and
secondary outcomes were evaluated at
the interim analysis using a probability
value of <.005, based on the O’Brien-
Fleming group sequential boundaries.39
To account for the interim analysis,
probability values of <.048 were
considered statistically significant in the
final analysis.
Results
One hundred seventy-two patients were
approached to participate in the study.
Thirty-two patients were excluded from
the analysis, including 28 patients who
declined (Figure 1). Randomization
occurred after successful catheter place-
ment. Deviations from study protocol
occurred in both groups. One patient
with a Bishop score of 9 was assigned
randomly to the tension group. In the no
tension group, 1 patient had a previous
loop electrosurgical excision procedure,
and 2 patients had a latex allergy that
prompted silicone balloon placement
(Cook Medical Inc, Bloomington, IN).
Sixty-seven patients were analyzed in the
tension group, and 73 patients were
ajog.org
FIGURE 1
Flow diagram
Assessed for eligibility (n = 172)
Excluded (n = 32)
- Not meeƟng inclusion criteria (n = 4)
- Declined to parƟcipate (n = 28)
Randomized (n = 140)
Allocated to tension (n = 67)
- Received allocated intervenƟon (n = 67)
- Did not receive allocated intervenƟon
(n = 0)
Lost to follow up (n = 0)
DisconƟnued intervenƟon (n = 0)
Analyzed (n = 67)
Excluded from analysis (n = 0)
The diagram shows the flow of the participants through the study.
TOLAC, trial of labor after cesarean delivery.
Fruhman et al. Tension vs no tension on transcervical catheter for cervical ripening. Am J Obstet Gynecol 2017.
analyzed in the no tension group
(Figure 1). All patient data were analyzed
by intent-to-treat.40 A “per protocol”
analysis was also performed that yielded
similar results. The planned interim
analysis was performed on 66 patients.
No patient safety issues were identified.
The survival curve for the primary
outcome of time from catheter insertion
to delivery had a probability value of .35;
the survival curve for the secondary
outcome of time from catheter insertion
to expulsion had a probability value of
<.01. Because neither outcome had a
probability value less than the critical
probability value of .005 for the interim
OBSTETRICS
Allocated to no tension (n = 73)
- Received allocated intervenƟon (n = 72)
- Did not receive allocated intervenƟon
(n = 1). She had rupture of membranes
prior to inserƟon of the catheter.
Lost to follow up (n = 0)
DisconƟnued intervenƟon (n = 2).
One was a TOLAC and desired repeat Cesarean
secƟon during the inducƟon. One stopped her
inducƟon aŌer minimal cervical change was
made and returned the following week.
Analyzed (n = 73)
Excluded from analysis (n = 0)
analysis, the decision was made to
continue the study to determine whether
complete data collection resulted in the
finding of significant differences be-
tween groups for study outcomes.
Baseline maternal characteristics are
described in Table 1. There were no
significant differences between the ten-
sion and no tension groups. Although
not statistically significant, there were
trends toward a greater number of pa-
tients with preeclampsia in a previous
pregnancy and a lower body mass index
in the current pregnancy to be assigned
randomly to the tension group, whereas
there was a trend toward a greater
JANUARY 2017 American Journal of Obstetrics & Gynecology
Original Research
number of patients with suspected fetal
growth restriction who were assigned
randomly to the no tension group.
Because this was a pragmatic trial, we
included patients with a previous cervi-
cal ripening agent. Nonetheless, all pa-
tients at randomization were required to
have a Bishop score of
6 and no labor.
The use of a cervical ripening agent
before the insertion of a transcervical
catheter was similar between groups
(P¼.208). Table 2 shows that there were
no differences as to the type and amount
of agent used. Cervical dilation, efface-
ment, and station were similar between
groups in this subset. Table 3 describes
the primary and secondary outcomes
that were related to the duration of the
transcervical catheter and final mode
of delivery outcomes. The primary
and secondary outcome data were
not distributed normally. Therefore,
medians and interquartile ranges are
reported and compared. Means and
standard deviations for the primary
outcome were included in Table 3 to
facilitate comparisons to other published
research. There was no statistically sig-
nificant difference in the primary
outcome between groups (median time,
16.2 vs 16.9 hours; P¼.814). However,
the median time from catheter insertion
to expulsion was significantly shorter in
the tension group (2.6 vs 4.6 hours;
P<.001). The survival curves are pro-
vided in Figure 2.
The cesarean delivery rates and in-
dications were similar between groups.
In the no tension group, there were 3
“other” indications (Table 3) that
included mid-labor fetal presentation
change, unsuccessful induction, and
patient desire to discontinue induction
and proceed with repeat cesarean de-
livery. Table 4 shows other clinical out-
comes that were related to labor. Table 5
shows that there were no differences in
neonatal outcomes between groups.
Survival curves were generated that
stratified for the use of a cervical
ripening agent before catheter place-
ment (data not shown). In each subset,
times from insertion of the catheter to
delivery and to expulsion in the tension
and no tension groups were similar to
that of the overall study population.
67.e3
Original Research OBSTETRICS ajog.org

Overall the number of complications

TABLE 1
in the study groups was low. One balloon
Baseline demographic, medical/obstetrics history, and current pregnancy
in the no tension group was ruptured at
data for 140 women
the time of removal; 1 of the catheters
Characteristic Tension (n¼67) No tension (n¼73) P value was expelled almost immediately after
Maternal age, ya 26 (23, 30) 26 (22, 30) .39 insertion, and there were 4 cases of
postpartum hemorrhage, all in the no
Race, n (%)
tension group. No differences in cho-
Caucasian 20 (29.9) 17 (23.3) .674 rioamnionitis were noted between
African American 46 (68.7) 55 (75.3) groups, despite a difference in the
Other 1 (1.5) 1 (1.4) duration of catheter placement.
Insurance, n (%) Comment
Government 57 (85.1) 58 (79.5) .386 Our study shows that placement of ten-
Private 10 (14.9) 15 (20.5) sion on a transcervical balloon catheter
Parity, n (%)
for cervical ripening does not shorten the
time to delivery but does shorten time
Nulliparity 38 (56.7) 35 (47.9) .299 to expulsion. The transcervical catheter
Multiparity 29 (43.3) 38 (52.1) is thought to work by mechanically
Gestational age at enrollment, wk a
39.1 (37.6, 40.4) 39.0 (37.6, 40.1) .473 stretching the cervical canal. In addition,
the catheter may cause the release of
Medical/obstetrics history, n (%)
prostaglandins that can result in cervical
Asthma 9 (13.4) 13 (17.8) .477 change.41,42 The release of prostaglandins
Chronic hypertension 9 (13.4) 12 (16.4) .619 may be the reason that the transcervical
Diabetes mellitus catheter ripens the cervix, even when no
tension is applied to the catheter. Low-
Gestational 6 (9.0) 7 (9.6) .897
dose oxytocin was infused while the
Pregestational 0 4 (5.5) .121 catheter was in place; during this time
Preeclampsia 18 (26.9) 10 (13.7) .052 frame, a lower amount of oxytocin was
Miscarriages 8 (11.9) 11 (15.1) .589 used in the tension group (290 vs 782
mU; P¼.004). Most likely, this is due to
Previous cesarean deliveries, n (%)
the shorter time from catheter insertion
0 55 (82.1) 58 (79.5) .444 to expulsion in the tension group.
1 11 (16.4) 11 (15.1) Our observed median time to delivery
2 1 (1.5) 4 (5.5) (approximately 17 hours) was shorter
than our sample size assumption of a
Body mass index at admission, kg/m2a 33.1 (28.9, 41.6) 37.0 (29.0, 42.2) .098
mean time to delivery of 24 hours in the
Reason for induction, n (%) no tension group. Because our data were
Diabetes mellitus 2 (3.0) 5 (6.8) .392 not distributed normally, we believed
Elective 10 (14.9) 10 (13.7) that we were obligated to report and
compare medians, as opposed to means.
Fetal growth restriction 1 (1.5) 7 (9.6)
Our mean time to delivery in the no
Hypertension 25 (37.3) 26 (35.6) tension group was 23.1 hours, which was
Oligohydramnios 5 (7.5) 5 (6.8) similar to other reports.14,27 However,
Other 11 (16.4) 11 (15.1) other investigators did not report results
with the use of median values, so we
Postdates 13 (19.4) 9 (12.3)
cannot compare our median times with
Ripening agents given 45 (67.1) 56 (76.7) .208 their data.
before transcervical
In the previously referenced studies,
catheter insertion, n (%)
heterogeneity exists with regards to the
Bishop score at time of 4 (3, 5) 3.5 (3, 5) .753 use of tension and the methods used to
transcervical catheter
placementa create tension. Methods of tension used
a
in previous studies included attachment
Data are shown as median (interquartile range).
Fruhman et al. Tension vs no tension on transcervical catheter for cervical ripening. Am J Obstet Gynecol 2017.
to a 500-mL bag of fluid placed to
gravity,21,29 and taping the transcervical

67.e4 American Journal of Obstetrics & Gynecology JANUARY 2017

ajog.org OBSTETRICS Original Research

catheter to the patient’s thigh on

TABLE 2
tension,7,10,13-17,19,20,22,25,26,28,30,31,33,35;
Baseline cervical ripening in patients who received an agent before insertion
some studies applied no ten-
of the transcervical catheter
sion.11,12,18,24,27,32 These studies did not
Characteristic Tension (n¼45) No tension (n¼56) P value address whether tension should be
Dilation at time of ripening, cma 1 (0, 1) 0.5 (0, 1) .876 placed on the transcervical catheter.
Lutgendorf et al43 published a ran-
Effacement at time of ripening %a 25 (0, 50) 25 (0, 50) 1
domized study of 45 patients who un-
Station at time of ripening a
e3 (e3, e3) e3 (e3, e3) .999 derwent cervical ripening with a
Ripening agents given before 45 (67.1) 56 (76.7) .208 transcervical catheter. They compared
transcervical catheter insertion, n (%) tension with a 1000-mL bag of fluid
Misoprostol use, n (%) 37 (55.2) 40 (54.8) .959 placed to gravity with taping the cath-
eter with tension to the patient’s inner
Misoprostol amount, mg a
50 (25, 75) 50 (25, 75) 1
thigh. The mean time to expulsion (1.6
Dinoprostone use, n (%) 8 (11.9) 11 (15.1) .589 vs 4.6 hours; P¼.0001) was shorter in
Dinoprostone amount, mga 10 (10, 10) 10 (10, 10) 1 the women whose catheter was attached
Oxytocin use, n (%) 13 (19.4) 18 (24.7) .725 to the bag of fluid. Times to delivery
were not reported.43 In 2013 Gibson
Oxytocin amount, mUa 1133 (515, 5216) 1511 (943, 2796) .723
et al37 published a randomized
Other (laminaria), n (%) 0 1 (1.4) 1 controlled trial of 191 patients that
a
Data are shown as median (interquartile range). compared the effectiveness of attaching
Fruhman et al. Tension vs no tension on transcervical catheter for cervical ripening. Am J Obstet Gynecol 2017.
the transcervical catheter to a 500-mL

TABLE 3
Transcervical catheter duration and delivery outcomes for 140 women
Outcomes Tension (n¼67) No tension (n¼73) P value
Delivery
Median time to delivery, hra 16.2 (12.6, 23.5) 16.9 (12.4, 25.6) .814
Mean time to delivery, hrb 19.1  10.4 23.1  26.2 .814
a
Median time to vaginal delivery, hr 14.8 (12.1, 20.5) 14.6 (11.3, 20.9) .99
Delivery within 24 hrs, n (%) 52 (77.6) 52 (71.2) .388
Vaginal delivery within 24 hr, n/N (%) 41/52 (78.8) 37/52 (71.2) .365
Vaginal deliveries after 24 hr, n/N (%) 9/15 (60.0) 9/21 (42.9) .31
Mode of delivery, n (%)
Vaginal 50 (74.6) 46 (63.0) .139
Cesarean delivery 17 (25.4) 27 (37.0)
Primary indication for cesarean delivery
Arrest of dilation 10 (58.8) 12 (44.4) .309
Nonreassuring fetal heart tracing 7 (41.2) 12 (44.4)
Other 0 3 (11.1)
a
Cervical dilation at cesarean delivery, cm 4.8 (4.1, 6.4) 5.0 (4.3, 7.5) .995
Transcervical catheter duration
Time to catheter expulsion, hra 2.6 (1.3, 4.3) 4.6 (2.3, 7.0) .001
Cervical dilation at time of catheter expulsion, cma 3.5 (3, 4) 3.5 (3, 4) 1
a
Highest pain score while catheter was in place 7 (5, 10) 7 (5, 10) 1
a
Total oxytocin while catheter in place, mU 290 (118, 900) 782 (313, 1584) .004
a
Data are shown as median (interquartile range); b Data are shown as mean  standard deviation.
Fruhman et al. Tension vs no tension on transcervical catheter for cervical ripening. Am J Obstet Gynecol 2017.

JANUARY 2017 American Journal of Obstetrics & Gynecology 67.e5

Original Research
FIGURE 2
Survival curves
Curves A, from the time of catheter insertion to delivery and B, from the time of catheter insertion to
catheter expulsion. Probability values were based on the Wilcoxon (Gehan) Breslow statistic.
Fruhman et al. Tension vs no tension on transcervical catheter for cervical ripening. Am J Obstet Gynecol 2017.
bag of fluid placed over the side of the
bed (the “traction” group) vs taping the
catheter on tension to the patient’s leg
every 30 minutes (the “taping” group).
They found that there was a shorter
length of time from insertion of the
67.e6 American Journal of Obstetrics & Gynecology JANUARY 2017
OBSTETRICS
catheter to expulsion of the catheter in
the former group (median time, 1.5 vs
2.6 hours; P<.001). Our findings were
similar in that delivery times were not
improved with an increase in applied
traction. Furthermore, the median
ajog.org
times to expulsion are identical in their
“taping” group and our tension group,
2.6 hours. In the study by Gibson et al,
there was no difference between their
“taping” and “traction” groups in the
mean time from insertion of the cath-
eter to delivery (18.8  8.0 vs 19.8 
8.5 hours; P¼.39). In our study, the
mean time from insertion to delivery in
our “tension” group was 19.1  10.4
hours, which is very similar to their
“taping” group. Our study can be seen
as an extension of the study of Gibson
et al. Although their intervention group
consisted of a 500-mL bag of fluid
placed to gravity on 1 end of the
spectrum, we placed no tension to the
transcervical catheter at the other end
of the spectrum.
We found that there was a higher rate
of epidural use in the tension group
(98.5% vs 89.0%; P¼.035). This finding
may not be statistically significant
because of the type II error of multiple
comparisons. Alternatively, it potentially
could also indicate that there was a
higher degree of pain in the tension
group. Interestingly, there was no dif-
ference in the median value of the
highest pain score during the time of
catheter placement between the 2 groups
(7/10 vs 7/10; P¼1). This discrepancy of
epidural use in the tension group vs no
difference in the pain scores could be
further evaluated in further research
studies.
Premature rupture of membranes was
not a criterion for exclusion in this study.
In our study, 4 patients (2.9%) had
rupture of membranes before placement
of the transcervical catheter. Although
labor theoretically may be quicker in
these patients, it does not seem that these
small numbers affected our primary or
secondary outcomes. Analysis of our
data, excluding these 4 patients, revealed
a time from insertion of the catheter to
delivery of 16.2 hours in the tension
group and 16.7 hours in the no tension
group (P¼.765). On the other hand, it is
difficult to manage cervical ripening in a
minimally dilated and minimally effaced
cervix that is not in spontaneous labor.
Some practitioners will choose not to
place a transcervical catheter out of
concern for infection. A retrospective
ajog.org
TABLE 4
Labor outcomes for 140 women
Characteristic
Rupture of membranes before delivery, n (%)
Before catheter placement
During catheter placement
After catheter placement before removal
After catheter removal
At delivery
a
Duration of rupture of membranes, hr
Epidural use, n (%)
Meconium
Group B streptococcus colonized, n (%)
Clinical chorioamnionitis, n (%)
Placental disease, n (%)
Acute chorioamnionitis
Funisitis
Unremarkable
Missing
a
Data are shown as median (interquartile range).
Fruhman et al. Tension vs no tension on transcervical catheter for cervical ripening. Am J Obstet Gynecol 2017.
study of 124 term patients with prema-
ture rupture of membranes found a
nonstatistically significant increase in
chorioamnionitis in women who un-
derwent cervical ripening with a trans-
cervical catheter compared with those
using oxytocin alone (28.6% vs 15.9%;
P¼.10).44 A multicenter prospective
randomized trial is addressing the risk of
chorioamnionitis in patients with pre-
mature rupture of membranes and
transcervical catheter use.45
Whether the use of a transcervical
catheter for cervical ripening increases
the risk of chorioamnionitis is contro-
versial.36,46 The overall risk of cho-
rioamnionitis in our study was 10.7%,
and there was no statistically significant
difference between the tension and no
tension groups, although the duration of
catheter placement was longer in the no
tension group. The study of Lutgendorf
et al43 did not evaluate the intrapartum
chorioamnionitis rate. The study of
Gibson et al37 found a chorioamnionitis
rate of 6.8%, without statistical differ-
ence between the “traction” and “taping”
OBSTETRICS Original Research
was no statistically significant difference
in cesarean delivery between groups,
perhaps this represents a type II error
Tension No tension because of small sample size. A post-hoc
(n¼67) (n¼73) P value analysis of our data, which compared a
cesarean delivery rate of 25.4% in the
tension group with 37.0% in the no
3 (4.5) 1 (1.4) .692
tension group, calculated that 250
1 (1.5) 1 (1.4) women per group would be needed to
3 (4.5) 3 (4.1) have this difference be significant at the
58 (86.6) 63 (86.3) P<.05 level.
Approximately 19% of the total pa-
2 (3.0) 5 (6.8)
tients (27/140) had had a previous ce-
7.3 (4.2, 12.6) 6.0 (3.4, 11.5) .236 sarean delivery. This is important
66 (98.5) 65 (89.0) .035 because these patients did not receive
7 (10.4) 8 (11.0) .922 cervical ripening with a previous pros-
taglandin agent. Although we did not
16 (23.9) 23 (31.5) .603
study this specifically, many clinicians
7 (10.4) 8 (11.0) .922 prefer to use a transcervical catheter as a
primary cervical ripening agent in pa-
19 (28.4) 20 (27.4) .899 tients with a previous cesarean delivery.
This subset of patients may warrant
4 (6.0) 6 (8.2) .747
further investigation with respect to the
29 (43.3) 36 (49.3) .475 use of tension vs no tension on the
19 (28.4) 18 (24.7) .620 transcervical catheter.
A major strength of this study was its
design as a randomized controlled trial.
Providers and patients, however, were
not blinded to the type of intervention
groups. Our rates were higher (10.7%), used. Despite 5 minor protocol de-
even though our definition seemed to be viations, a secondary “per protocol”
stricter. analysis was performed, and the results
Most studies that have been published were similar. Patients were included even
on cervical ripening with a transcervical if a previous cervical ripening agent was
catheter used between 30 and 80 mL of used, which is both a strength and a
fluid to fill the balloon. Levy et al33 in limitation of the study, depending on
2004 concluded that the use of 80 mL of one’s perspective. It may limit the ability
fluid, compared with 30 mL, resulted in to compare the outcomes of this study
more advanced dilation at the time of with other studies but may reflect clinical
expulsion, faster labor times, and less practice patterns more accurately and
need for augmentation with oxytocin in therefore be more generalizable. The
primiparous patients. Delaney et al34 randomization reflected only that ten-
compared the use of 30 vs 60 mL of sion or no tension was applied to the
fluid and concluded that 60 mL of fluid transcervical catheter. The randomiza-
did not result in a higher percentage of tion was not stratified by previous cer-
deliveries within 24 hours, although it vical ripening agents, and no conclusion
did result in a higher percentage of de- from this study can be made regarding
liveries by 12 hours. Gibson et al37 used their use in tandem or sequentially with
30 mL of fluid to fill the balloon; 50 mL the transcervical catheter. Recognizing
of fluid were used in our study. The this potential confounder, a Bishop score
mean times to delivery between our of >6 or labor was considered exclu-
studies were similar; the extra 20 mL of sionary before study entry. Furthermore,
fluid in our study most likely had little the patients receiving a previous cervical
effect on time to delivery. ripening agent and the actual type of
Mode of delivery was a secondary agent were similar in both groups
outcome in this study. Although there (Table 2). We demonstrated that there
JANUARY 2017 American Journal of Obstetrics & Gynecology 67.e7
Original Research
TABLE 5
Neonatal outcomes for 140 women
Characteristic
Infant birthweight, gma
Low birthweight <2500 g, n (%)
Macrosomia >4000 g, n (%)
Gender: female, n (%)
a 6. American College of Obstetricians and Gy-
1-Minute Apgar score
5-Minute Apgar scorea
a Obstet Gynecol 2010;116:450-63.
Umbilical cord gasses
Arterial pH
Arterial CO2, mEq/L
Arterial base excess, mEq/L
Venous pH
Venous CO2, mEq/L
Venous base excess, mEq/L
Neonatal disposition, n (%)
Neonatal Intensive Care Unit
Respiratory complications
Intracranial hemorrhage
Proven sepsis
a 11. Jozwiak M, Oude Rengerink K, Benthem M,
Data are shown as median (interquartile range).
Fruhman et al. Tension vs no tension on transcervical catheter for cervical ripening. Am J Obstet Gynecol 2017.
was no difference in the primary
outcome when we controlled for the use
of a previous agent, although this study
was not powered to detect such a dif-
ference. Most of the patients were Afri-
can American; thus, the results of our
study may not be applicable to other
populations. Another limitation is that
we could not standardize the amount of
tension placed to the catheter. Finally, we
recognize that there are inherent differ-
ences in the labor curves comparing
nulliparous and multiparous patients,
although there was no difference in
parity between groups at baseline; these
comparisons are beyond the scope of this
article.
Group allocation from a random pull
from a pile of opaque envelopes is a
known means of randomization.47
However, because random number
generators are a more ideal form of
randomization, we chose to alter our
randomization scheme. Although not
ideal, we believe that this change in
67.e8 American Journal of Obstetrics & Gynecology JANUARY 2017
OBSTETRICS ajog.org
3. American College of Obstetricians and Gy-
necologists. ACOG Practice Bulletin No. 107:
Induction of labor. Obstet Gynecol 2009;114:
386-97.
Tension (n¼67) No tension (n¼73) P value 4. Ramirez MM. Labor induction: a review of
3090 (2805, 3520) 3095 (2695, 3540) .485 current methods. Obstet Gynecol Clin North Am
2011;38:215-25, ix.
9 (13.4) 8 (11.0) .654 5. Wing DA, Lovett K, Paul RH. Disruption of
prior uterine incision following misoprostol for
3 (4.5) 6 (8.2) .367
labor induction in women with previous cesar-
37 (55.2) 37 (50.7) .591 ean delivery. Obstet Gynecol 1998;91:828-30.
8 (6, 8) 8 (7, 9) .999
necologists. ACOG Practice Bulletin No. 115:
9 (9, 9) 9 (9, 9) 1 Vaginal birth after previous cesarean delivery.
7. Sciscione AC, Bedder CL, Hoffman MK,
7.24 (7.20, 7.30) 7.24 (7.16, 7.29) .419 Ruhstaller K, Shlossman PA. The timing of
adverse events with Foley catheter preinduction
55.3 (48.8, 65.4) 55.9 (46.5, 67.6) .54
cervical ripening; implications for outpatient use.
e5 (e6, e3) e5 (e7, e3) .544 Am J Perinatol 2014;31:781-6.
8. Smith JA. Balloon dilators for labor induction:
7.31 (7.27, 7.34) 7.30 (7.25, 7.33) .702
a historical review. J Med Ethics Hist Med
44.0 (39.0, 49.0) 45.0 (40.5, 50.7) .754 2013;6:10.
9. Embrey MP, Mollison BG. The unfavourable
e5 (e6, e3) e5 (e7, e3) .772
cervix and induction of labour using a cervical
balloon. J Obstet Gynaecol Br Commonw
1967;74:44-8.
11 (16.4) 13 (17.8) .827
10. Edwards RK, Szychowski JM, Berger JL,
9 (13.4) 12 (16.4) .619 et al. Foley catheter compared with the
controlled-release dinoprostone insert: a ran-
1 (1.5) 0 (0) .479
domized controlled trial. Obstet Gynecol
1 (1.5) 1 (1.4) 1 2014;123:1280-7.
et al. Foley catheter versus vaginal prostaglandin
E2 gel for induction of labour at term (PROBAAT
trial): an open-label, randomised controlled trial.
Lancet 2011;378:2095-103.
randomization method should have lit- 12. Cromi A, Ghezzi F, Agosti M, et al. Is trans-
cervical Foley catheter actually slower than
tle, if any, effect on the results because prostaglandins in ripening the cervix? A ran-
this occurred early in the study. domized study. Am J Obstet Gynecol 2011;204:
In conclusion, this study shows that 338.e1-7.
taping a transcervical catheter to provide 13. Henry A, Madan A, Reid R, et al. Outpatient
tension for cervical ripening does not Foley catheter versus inpatient prostaglandin E2
gel for induction of labour: a randomised trial.
reduce time to delivery. However, it did BMC Pregnancy Childbirth 2013;13:25.
reduce time to expulsion, which resulted 14. Sciscione AC, McCullough H, Manley JS,
in less oxytocin use while the catheter Shlossman PA, Pollock M, Colmorgen GH.
was in place. The clinical implications of A prospective, randomized comparison of Foley
our study deserve further research. n catheter insertion versus intracervical prosta-
glandin E2 gel for preinduction cervical ripening.
Am J Obstet Gynecol 1999;180:55-60.
Acknowledgment 15. Sciscione AC, Nguyen L, Manley J,
The authors acknowledge the Obstetrics and Pollock M, Maas B, Colmorgen G. A randomized
Gynecology residents of Saint Louis University comparison of transcervical Foley catheter to
School of Medicine and the nursing staff at SSM intravaginal misoprostol for preinduction cervical
Health-St. Mary’s Hospital for their help with pa- ripening. Obstet Gynecol 2001;97:603-7.
16. Kandil M, Emarh M, Sayyed T, Masood A.
tient recruitment and implementation of this study.
Foley catheter versus intra-vaginal misoprostol
for induction of labor in post-term gestations.
References Arch Gynecol Obstet 2012;286:303-7.
1. Osterman M, Martin J. Recent declines in 17. Gelisen O, Caliskan E, Dilbaz S, et al. In-
induction of labor by gestational age. Available duction of labor with three different techniques at
41 weeks of gestation or spontaneous follow-up
at: http://www.cdc.gov/nchs/data/databriefs/
db155.pdf. Accessed January 5, 2016. until 42 weeks in women with definitely unfa-
2. Bishop EH. Pelvic scoring for elective induc- vorable cervical scores. Eur J Obstet Gynecol
tion. Obstet Gynecol 1964;24:266-8. Reprod Biol 2005;120:164-9.
ajog.org
18. Ten Eikelder ML, Oude Rengerink K,
Jozwiak M, et al. Induction of labour at term with
oral misoprostol versus a Foley catheter (PRO-
BAAT-II): a multicentre randomised controlled
non-inferiority trial. Lancet 2016;387:1619-28.
19. Tuuli MG, Keegan MB, Odibo AO, Roehl K,
Macones GA, Cahill AG. Progress of labor in
women induced with misoprostol versus the
Foley catheter. Am J Obstet Gynecol 2013;209:
237.e1-7.
20. Mahomed K. Foley catheter under traction
versus extra-amniotic prostaglandin gel in pre-
treatment of unripe cervix: a randomised
controlled trial. Cent Afr J Med 1988;34:98-102.
21. Kashanian M, Nazemi M, Malakzadegan A.
Comparison of 30-mL and 80-mL Foley catheter
balloons and oxytocin for preinduction cervical
ripening. Int J Gynaecol Obstet 2009;105:174-5.
22. Moraes Filho OB, Albuquerque RM,
Cecatti JG. A randomized controlled trial
comparing vaginal misoprostol versus Foley
catheter plus oxytocin for labor induction. Acta
Obstet Gynecol Scand 2010;89:1045-52.
23. Fitzpatrick CB, Grotegut CA, Bishop TS,
Canzoneri BJ, Heine RP, Swamy GK. Cervical
ripening with foley balloon plus fixed versus in-
cremental low-dose oxytocin: a randomized
controlled trial. J Matern Fetal Neonatal Med
2012;25:1006-10.
24. Pettker CM, Pocock SB, Smok DP, Lee SM,
Devine PC. Transcervical Foley catheter with
and without oxytocin for cervical ripening: a
randomized controlled trial. Obstet Gynecol
2008;111:1320-6.
25. Carbone JF, Tuuli MG, Fogertey PJ,
Roehl KA, Macones GA. Combination of Foley
bulb and vaginal misoprostol compared with
vaginal misoprostol alone for cervical ripening
and labor induction: a randomized controlled
trial. Obstet Gynecol 2013;121:247-52.
26. Lanka S, Surapaneni T, Nirmalan PK. Con-
current use of Foley catheter and misoprostol for
induction of labor: a randomized clinical trial of
efficacy and safety. J Obstet Gynaecol Res
2014;40:1527-33.
27. Perry KG Jr, Larmon JE, May WL,
Robinette LG, Martin RW. Cervical ripening: a
randomized comparison between intravaginal
misoprostol and an intracervical balloon catheter
combined with intravaginal dinoprostone. Am J
Obstet Gynecol 1998;178:1333-40.
28. Ugwu EO, Onah HE, Obi SN, et al. Effect of
the Foley catheter and synchronous low dose
OBSTETRICS
misoprostol administration on cervical ripening:
a randomised controlled trial. J Obstet Gynaecol
2013;33:572-7.
29. Ghanaie MM, Jafarabadi M, Milani F,
Asgary SA, Karkan MF. A randomized controlled
trial of foley catheter, extra-amniotic saline infu-
sion and prostaglandin e2 suppository for labor
induction. J Family Reprod Health 2013;7:
49-55.
30. Maslovitz S, Lessing JB, Many A. Compli-
cations of trans-cervical Foley catheter for labor
induction among 1,083 women. Arch Gynecol
Obstet 2010;281:473-7.
31. Lin MG, Reid KJ, Treaster MR,
Nuthalapaty FS, Ramsey PS, Lu GC. Trans-
cervical Foley catheter with and without extra-
amniotic saline infusion for labor induction: a
randomized controlled trial. Obstet Gynecol
2007;110:558-65.
32. Karjane NW, Brock EL, Walsh SW. Induction
of labor using a foley balloon, with and without
extra-amniotic saline infusion. Obstet Gynecol
2006;107:234-9.
33. Levy R, Kanengiser B, Furman B, Ben Arie A,
Brown D, Hagay ZJ. A randomized trial
comparing a 30-mL and an 80-mL Foley cath-
eter balloon for preinduction cervical ripening.
Am J Obstet Gynecol 2004;191:1632-6.
34. Delaney S, Shaffer BL, Cheng YW, et al.
Labor induction with a Foley balloon inflated to
30 mL compared with 60 mL: a randomized
controlled trial. Obstet Gynecol 2010;115:
1239-45.
35. Forgie MM, Greer DM, Kram JJ,
Bernhard KA, Salvo NP, Siddiqui DS. Foley
catheter placement for induction of labor
with or without stylette: a randomized clinical
trial. Am J Obstet Gynecol 2016;214;397.
e1-397.
36. Heinemann J, Gillen G, Sanchez-Ramos L,
Kaunitz AM. Do mechanical methods of cervical
ripening increase infectious morbidity? A sys-
tematic review. Am J Obstet Gynecol 2008;199:
177-88.
37. Gibson KS, Mercer BM, Louis JM.
Inner thigh taping vs traction for cervical
ripening with a Foley catheter: a randomized
controlled trial. Am J Obstet Gynecol 2013;209:
272.e1-7.
38. Steel RGD, Torrie JH. Nonparametric sta-
tistics. In: Principles and procedures of statistics:
a biomedical approach. New York: McGraw-Hill;
1980:540-2.
JANUARY 2017 American Journal of Obstetrics & Gynecology
Original Research
39. Demets DL, Lan KK. Interim analysis: the
alpha spending function approach. Stat Med
1994;13:1341-52.
40. Schulz KF, Grimes DA. Sample size slip-
pages in randomised trials: exclusions and
the lost and wayward. Lancet 2002;359:
781-5.
41. Sciscione AC. Methods of cervical ripening
and labor induction: mechanical. Clin Obstet
Gynecol 2014;57:369-76.
42. Manabe Y, Manabe A, Takahashi A.
F prostaglandin levels in amniotic fluid
during balloon-induced cervical softening
and labor at term. Prostaglandins 1982;23:
247-56.
43. Lutgendorf MA, Johnson A, Terpstra ER,
Snider TC, Magann EF. Extra-amniotic balloon
for preinduction cervical ripening: a randomized
comparison of weighted traction versus un-
weighted. J Matern Fetal Neonatal Med
2012;25:581-6.
44. Cabrera IB, Quinones JN, Durie D, Rust J,
Smulian JC, Scorza WE. Use of intracervical
balloons and chorioamnionitis in term premature
rupture of membranes. J Matern Fetal Neonatal
Med 2016;29:967-71.
45. FOLCROM Trial: Foley Catheter in Rupture of
Membranes. Available at: https://clinicaltrials.gov/
ct2/show/NCT01973036. Accessed January
24, 2016.
46. Jozwiak M, Bloemenkamp KW, Kelly AJ,
Mol BW, Irion O, Boulvain M. Mechanical
methods for induction of labour. Cochrane
Database Syst Rev 2012;3:CD001233.
47. Schulz KF, Grimes DA. Generation of allo-
cation sequences in randomised trials: chance,
not choice. Lancet 2002;359:515-9.
Author and article information
From the Division of Maternal Fetal Medicine (Drs
Fruhman, Amon, Flick, and Gross) and the Division of
Research (Dr Gavard and Mr Miller), Department of
Obstetrics, Gynecology, and Women’s Health, Saint Louis
University School of Medicine, Saint Louis, MO.
Received June 16, 2016; revised Sept. 5, 2016;
accepted Sept. 7, 2016.
The authors report no conflict of interest.
Current affiliation of Dr Fruhman is Department of
Obstetrics and Gynecology, Staten Island University
Hospital, Staten Island, NY.
Corresponding author: Gary Fruhman, MD. fruhman@
slu.edu
67.e9