Keywords cervical ripening - cervical ripening balloon - double-balloon catheter - induction of labour - time to delivery
Schlüsselwörter Zervixreifung - Ballonkatheter zur Zervixreifung - Doppelballonkatheter - Geburtseinleitung - Zeitspanne bis zur Entbindung
Introduction
Approximately 22% of births in Germany are induced each year [1 ]. Medication is the main method used, but mechanical,
sequential mechanical-medical, and other methods, such as induction with castor oil, are also used [2 ]
[3 ]
[4 ]
[5 ]
[6 ].
Oxytocin and prostaglandins are available as drugs for induction [7 ]
[8 ]. Prostaglandin E2 (PGE2 – dinoprostone) and prostaglandin E1 (PGE1 – misoprostol) have a cervical ripening and labour-inducing effect, which is
why, unlike oxytocin, they are also used in cases of immature cervical findings. Especially at the beginning of induction, there are often painful uterine contractions without cervical ripening,
which has reduced general acceptance by the pregnant women. An alternative is sequential induction with first mechanical cervical ripening followed by PGE1/2 [7 ]. Such a sequential approach using misoprostol in combination with a double balloon catheter (DBC) was similarly effective in primiparae and
shortened the prostaglandin delivery interval (PDI) [9 ].
Mechanical methods of induction of labour (IOL) include iatrogenic amniotomy, hydroscopic dilators, and cervical single and double balloon catheters. Amniotomy is not suitable as a
stand-alone procedure and should only be used in the presence of mature cervical findings [7 ]. Dilators and balloon
catheters have been used for cervical ripening for decades [10 ]
[11 ]. Dilation of the cervix results in the endogenous release of prostaglandins, leading to cervical ripening and myometrial contractions. IOL with double balloon catheters is
associated with low rates of uterine hyperstimulation [12 ] and low maternal and neonatal morbidity [3 ]. The resulting rate of caesarean section appears to be comparable to that following drug induction with dinoprostone [9 ]. The effectiveness of the balloon catheters is equivalent to that of IOL with dinoprostone [3 ]
[13 ]
[14 ] and misoprostol [3 ]
[15 ],
although a recent meta-analysis by Zhao et al. shows the superiority of vaginal misoprostol with respect to the induction-delivery interval (IDI) [16 ].
The German guideline for induction of labour 2020 [7 ] recommends the use of a DBC for 12 hours, as prolonging the
balloon insertion time from 12 to 24 hours showed no additional effect on IDI [17 ].
More recent studies have found no difference in effectiveness, but a reduction in IDI when the balloon placement time was shortened further [18 ]
[19 ].
This prospective case-control study tests the effectiveness of shortening the DBC insertion time from 12 to 6 hours in everyday clinical practice.
Methodology
Study population
In this prospective case control study, a total of 259 pregnant women were recruited from January to December 2021 at the Department of Obstetrics, University Hospital Jena. The study was
approved by the Ethics Committee of the Friedrich Schiller University of Jena (no. 2021–2365). Women with a term pregnancy (gestational age ≥ 37+0 weeks of gestation) were included if they
were induced with a DBC (Cook Medical, Bloomington, Indiana, USA). Exclusion criteria were age below 18 years, premature rupture of membranes, and suspicious/pathological fetal heart rate
pattern at the time of labour induction. A total of 248 women with IOL were included in the study ([Fig. 1 ]). Incomplete data sets
result in different case numbers, as indicated in the tables.
Fig. 1 Study population. Inclusion criteria: GA ≥ 37 + 0, singleton pregnancy, induction with double balloon catheter; exclusion criteria: age < 18 years, path./susp. CTG at the start of
induction, premature rupture of membranes.
Methods of induction of labour
The DBC was inserted under visual control according to the manufacturer’s instructions [10 ]. In the P12 group, the
balloon was inserted at approximately 8 pm according to the hospital’s internal Standard Operating Procedure (SOP) and usually removed at 8 am the following day. In the P6 group, the balloon
was inserted at around 7 am and removed at 1 pm on the same day, unless the DBC spontaneously fell out. If the balloon insertion failed to induce labour, IOL was continued with misoprostol
(initially 50 µg p. o., then 100 µg p. o. every 4 hours according to vaginal findings, manufactured by the Jena University Hospital pharmacy) or a vaginal insert containing dinoprostone
(Propess, Ferring Arzneimittel, Kiel, Germany). As the clinical situation may required, oxytocin (5 IU oxytocin, Hexal, 5 IU in 500 ml whole electrolyte solution, start with 2 mIU/min,
increase infusion rate by 2 mIU/min every 20–30 min, if necessary) was administered. The selection of IOL method for a participant following a prior caesarean section was made through a
process of shared decision with her, excluding the use of misoprostol [7 ].
Endpoints
The primary endpoint of this study is the induction to delivery interval following induction of labour using the DBC. Secondary outcomes are the need for sequential drug induction of
labour, prostaglandin delivery interval (PDI), mode of delivery, maternal outcomes as blood loss and uterine rupture, and neonatal outcomes, including arterial umbilical pH, base excess,
neonatal acidosis < 7.1, APGAR score 5’ < 7, pathological CTG, and meconium-stained amniotic fluid.
Data management
A paper-based case report form was created for the participants’ patient files. It outlined the study’s workflow and contained a specialized form for study specific data. Relevant
descriptive clinical routine data were extracted from primary clinical documentation. Demographic data were collected on admission and the fetal condition was assessed by cardiotocography
(CTG). Data with clinical relevance for IOL procedures were carefully chosen. This encompassed anamnestic factors that impact IOL, such as participant’s age, co-morbidities, BMI, parity,
history of previous caesarean section, gestational age, and the rationale behind IOL initiation. Moreover, details delineating pregnancy, childbirth, and neonatal outcomes were selected to
comprehensively describe the efficacy and safety of labour induction. This included the sequential progression of IOL, utilization of epidural anesthesia, mode and rationale of delivery,
instances of uterine rupture, APGAR scores, and neonatal intensive care unit (NICU) admissions.
Study related and clinical data was combined into one data sheet. The analysis was conducted according to the intention-to-treat principle.
Descriptive statistics are presented as the median (interquartile range) or number (percentage) for numeric and categorical variables. Baseline statistical analysis was performed using a
Student’s t-test and a Fisher’s chi-squared or exact test. The Wilcoxon signed rank test was used to determine the significance of the IDI and the PDI in the two groups [20 ]. All calculations were performed using R (version 4.2.1) [21 ]
[22 ]
[23 ]
[24 ] and a p-value < 0.05 was considered statistically significant.
Results
During the study period, 109 women with IOL were recruited to the P12 group and 139 to the P6 group. The two groups were comparable in terms of their demographics ([Table 1 ]). The median age of the participants was 31–32 years, they were normal to slightly overweight (P12: BMI 24.7 kg/m2 , P6:
25.9 kg/m2 ), predominantly primiparous (P12: 65.1%, P6: 66.2%), and with a median gestational age of 40 weeks at balloon insertion. Furthermore, the observed anamnestic conditions
entail no significant differences. These include previous caesarean sections (P12: 14.68%, P6: 6.47%, p = 0.055), multiple pregnancies (P12: 2.75%, P6: 0.88%; p = 1.00), hypertension in
pregnancy (P12: 2.8%, P6: 1.5%; p = 0.657), and gestational diabetes (P12: 19.4%, P6: 25.6%; p = 0.329) in pregnancy. The comparability of those factors across both groups obviates the necessity
for statistical adjustments.
Table 1 Demographic description.
P12
P6
p-value
N
median (IQR) or n (%)
N
median (IQR) or n (%)
IOL = induction of labour; IQR = interquartile range; N = number of cases; n = number of cases with complete information; P6/P12 = balloon placement 6 and 12 h
Age [years]
109
31 (29–34)
139
32 (28–35)
0.914
Diabetes
108
21 (19.4%)
137
35 (25.6%)
0.329
Hypertension in pregnancy
108
3 (2.8%)
137
2 (1.5%)
0.657
BMI before pregnancy [kg/m2 ]
109
24.7 (21.9–30.5)
137
25,9 (22–30.1)
0.898
BMI at birth [kg/m2 ]
109
29.6 (27.4–35.4)
136
31,2 (27.2–34.9)
0.94
Gravidity
109
1 (1–2)
139
1 (1–2)
0.373
Parity
109
0 (0–1)
139
0 (0–1)
0.9
71 (65.1%)
92 (66.2%)
38 (34.9%)
47 (33.8%)
Prior caesarean section
109
16 (14.68%)
139
9 (6.47%)
0.055
Multiples
109
3 (2.75%)
139
4 (2.88%)
1.000
Weeks of gestation at IOL
109
40 (38–40)
139
40 (39–40)
0.304
Indication for IOL
109
139
0.403
42 (38.5%)
61 (43.9%)
42 (38.5%)
40 (28.8%)
19 (17.4%)
31 (22.3%)
6 (5.5%)
7 (5.04%)
Primary endpoint: induction to delivery interval
The primary endpoint is the induction to delivery interval. In the P6 group, this is significantly shorter by 558 min (about 9 h) compared to the P12 group (1348 min vs. 1906 min;
p < 0.01) ([Fig. 2 ]
a ). The significance remains when primi- (p < 0.01) and multiparous women
(p < 0.01), and women with (p = 0.02) and without previous caesarean section (p < 0.01) are considered separately. The time to delivery is shortened not only by the six-hour reduction in
balloon placement time, but by an additional 3 hours.
Fig. 2 Proportions of undelivered birthing persons. a After induction of labour with double balloon catheter, b after first prostaglandin
administration.
Secondary endpoints: prostaglandin to delivery interval, pregnancy and neonatal outcome
Furthermore, the PDI between the first prostaglandin administration after balloon removal and delivery in multiparous women was significantly shortened by 260 min (approximately 4.3 h) in
the P6 group compared to the P12 group (590 min vs 850 min, p = 0.001) ([Fig. 2 ]
b ).
The groups did not differ in the secondary outcomes ([Table 2 ]). In both groups, the IOL had to be continued with prostaglandins
in 93.5% of cases. The different types for sequential IOL after removing the DBC – dinoprostone vaginally (P12: 25%, P6: 21.2%) and misoprostol orally (P12: 67.6%, P6: 66.2%) – were used
equally in both groups.
Table 2 Course of birth and neonatal outcome.
P12
P6
N
median (IQR) or n (%)
N
median (IQR) or n (%)
p-value
AGA = appropriate for gestational age; CS = caesarean section; CTG = cardiotocography; DS = dilation stage; ES = expulsion stage; IQR = interquartile range; LGA = large for
gestational age; N = Number of cases; n = Number of cases with information; NICU = neonatal intensive care unit; P6/P12 = balloon placement 6 and 12 h; SGA = small for gestational age;
UA = umbilical artery
Course of birth
Continuation of induction
109
102 (93.6%)
139
130 (93.5%)
1
Dinoproston vaginal insert
108
27 (25%)
137
29 (21.2%)
0.578
Misoprostol oral
108
73 (67.6%)
139
92 (66.2%)
0.923
3 (2–4)
2 (2–4)
0.12
250 (150–350)
175 (150–350)
0.132
Amniotomy
109
6 (5.5%)
139
15 (10.8%)
0.21
Epidural anaesthesia
109
52 (47.7%)
138
53 (38.4%)
0.181
Delivery
Suspicious CTG sub partu
109
36 (33%)
139
46 (33.1%)
1
Position
108
137
0.573
101 (93.5%)
124 (90.5%)
5 (4.6%)
11 (8%)
2 (1.9%)
2 (1.5%)
Oxytocin sub partu
109
62 (56.9%)
139
66 (47.5%)
0.18
Delivery mode
109
138
0.572
66 (60.6%)
88 (63.8%)
14 (12.8%)
12 (8.7%)
29 (26.6%)
38 (27.5%)
Indication for CS
41
44
0.41
6 (14.6%)
2 (4.6%)
22 (53.7%)
26 (59.1%)
8 (19.5%)
7 (15.9%)
5 (12.2%)
8 (18.2%)
Blood loss
108
500 (300–600)
138
400 (300–600)
0.836
Uterine rupture
109
0 (0%)
139
1 (0.7%)
1
Neonatal outcome
Meconium stained amniotic fluid
109
2 (1.8%)
139
0 (0%)
0.192
Birth weight
109
3430 (2995–3850)
139
3480 (3055–3910)
0.126
15 (13.89%)
15 (10.95%)
78 (72.22%)
96 (70.07%)
15 (13.89%)
26 (18.98%)
APGAR 1 min
109
9 (8–9)
139
9 (8–9)
0.349
APGAR 5 min
109
9 (9–10)
139
9 (9–10)
0.877
APGAR 10 min
109
10 (10–10)
139
10 (10–10)
0.332
UA ph
107
7.21 (7.16–7.28)
139
7,23 (7,17–7,27)
0.711
11 (10.3%)
9 (6.5%)
0.397
Base excess
106
−3.5 (−6.5–−1.4)
137
−4 (−6.2–−1.5)
0.695
Transfer NICU
109
11 (10.1%)
139
14 (10.1%)
1
The rates of secondary caesarean section (P6: 27.5%, P12: 26.6%, p = 0.57) and blood loss at birth of 400–500 ml were equally distributed in the two cohorts. In the entire study
population, there was only one case of uterine rupture in the P6 group. This was an attempted vaginal birth after caesarean section with DBC and dinoprostone vaginal insert. This approach was
initiated in response to obstructed labour during the previous delivery. Subsequently, a pathological CTG reading indicated the repeat caesarean section. Intraoperatively, a rupture of the
posterior uterine wall, independent of the previous caesarean scar, was encountered and sewn over without complications. The neonate was well (pH NA 7.22; APGAR 1’/7; 5’/8; 10’/9).
There were no significant differences in neonatal outcome as regards the frequency of pathological CTG sub partu (P12 and P6: 33% both), neonatal acidosis with umbilical arterial pH
< 7.1 (P6: 6%, P12: 10%), amniotic fluid containing meconium (P6: 0%, P12: 1.8%) and APGAR 5’ < 7 (P12 and P6: 3.6% both).
Discussion
We were able to show that reducing the duration of the double balloon catheter placement by six hours led to a more effective induction of labour, as indicated by a reduction in induction to
delivery interval of an additional three hours to the shortened 6-hour double balloon catheter placement duration. We found this effect in both primi- and multiparous women, as well as in
deliveries with and without previous caesarean section. The Prostaglandin-to-Delivery-Interval was also significantly reduced by more than four hours in multiparous women. Shortening of the
induction to delivery interval was not associated with a higher rate of additional necessary sequential drug induction or more surgical deliveries, and there are no differences in the maternal
and neonatal safety parameters studied.
Induction of labour is the most common obstetric intervention in Germany, accounting for more than 20 % of inductions [1 ]. Mechanical induction using double balloon catheter is a routine procedure that has been clinically established for many years. Equivalent effectiveness and safety have
been demonstrated for double balloon placement for 12 and 24 hours. Recent studies have reported comparable effectiveness even when the duration of placement was shortened to 6 hours [18 ]
[19 ].
This effect can be explained by a higher endogenous prostaglandin release and a greater sensitivity to exogenous prostaglandins [25 ]
[26 ].
The comparable studies by Bleicher et al. [18 ] and Lassey et al. [19 ] do not provide information on the timing of onset of the induction of labour. In our clinic, the start of the IOL using the double balloon
catheter was changed from the evening hours at 8 pm for 12 hours to the morning at 8 am for 6 hours, so that the women were spared nighttime induction of labour. This adds valuable new insights
to the previous work of Kehl et al. [27 ]. They found the evening application of the double balloon catheter with planned
administration of misoprostol by maintaining the placement time to be favourable. In our clinical experience, overnight balloon induction often leads to sleep deprivation and early exhaustion
caused by the onset of uterine contractions. A morning induction lasting only 6 hours could prevent this nighttime exhaustion and thus have a positive influence on the duration of induction. As
a result, women and midwives may be more accepting and satisfied with their choice of induction method. While we did not extensively assess this aspect during our study, we received similar
feedback at our centre subsequent to the transition to the 6-hour placement regimen in clinical practices.
Our study’s primary emphasis rested on the evaluation of clinical feasibility. Consequently, we intentionally excluded the prospect of conducting a randomized trial, notwithstanding its
inherent advantages. The strategic decision to implement the modification in double balloon catheter placement duration after a span of 6 months served as a foundational step in ensuring the
comparability of our study cohorts. This assertion was substantiated through statistical comparisons encompassing anamnestic profiles, pregnancy, delivery, and neonatal outcomes. Particularly,
this fostered the hypothesis that the use of supplementary sequential IOL methods (dinoprostone, misoprostol, and amniotomy) are equally safe and efficient for both maternal and neonatal
well-being in both study cohorts. This confirms prior findings of improved efficacy by combining oral misoprostol and mechanical dilation [28 ].
Several considerations come into play when determining the optimal approach for labour induction [7 ]
[27 ]. To include women with multiple pregnancies as well as prior caesarean section into our study cohort, was giving tribute
to the clinical reality and contributes novel insights to existing literature. Our findings confirm the safety of mechanical induction of labour among women with prior caesarean section [29 ] within our study cohort, even when followed by vaginal prostaglandin application.
This study has some limitations such as a monocentric approach and unblinded design of the participants and study personnel. The strength of our study is the real-world design in daily
clinical practice in a delivery ward, together with an intention-to-treat analysis. During the study period, the SOPs were not changed for obstetric practice and in clinical matters other than
the study protocol.
Conclusion
We were able to demonstrate the superiority of shortening the duration of a double balloon catheter placement to six hours, both from the point of view of the women in terms of reducing the
induction to delivery time by around 9 hours and from the point of view of hospital management, by optimizing procedures with comparable patient safety. As a result, the shortened procedure
described above has been introduced into our clinical routine as SOP.
