Vaginal delivery among women who underwent labor induction with vaginal dinoprostone (PGE2) insert: a retrospective study of 1656 women in China
Introduction
Induction of labor is the nonspontaneous initiation of uterine contractions to accomplish delivery prior to the onset of spontaneous labor [1]. It has become one of the most common obstetrical procedures. According to the World Health Organization and the Centers for Disease Control, the rate of labor induction has been increasing since 1990 and accounts for 25% of all pregnant women in developed countries [2]. In the past decade, the rates of labor induction at term have doubled from approximately 9% to 23% in the United States [3] and from 10% to 20.4% in China [4]. The goal of labor induction is to achieve successful vaginal delivery by ripening the cervix.
Induction of labor has a major impact on the mater- nal and neonatal outcomes, which in turn affects the satisfaction with the birth experience and increases the pressure of nurses in labor wards. The unripe cervix is a major obstacle to the labor induction and vaginal delivery [5]. Multiple pharmacologic and mechanical agents, such as oxytocin, misoprostol and foley cath- eter, are available for cervical ripening and labor induc- tion. However, each has its own inherent risks of potentially harmful outcomes, such as tachysystole, uterine rupture, and infection [6]. Therefore, it is still required to explore an optimal induction method.
Prostaglandin E2 (PGE2), also known as dinoprostone, is the only prostaglandin approved by the US Food and Drug Administration for cervical ripening in labor induction [7]. Currently, dinoprostone has been proven to be a safe and effective ripening agent for labor induction and thus has been widely used in clin- ics [8]. It allows slow-release of exogenous PGE2 locally to promote relaxation of cervical smooth muscle, which in turn facilitates the synthesis and releases of endogenous PGE2. The unique design of retrieval tape makes it possible to withdraw in the initiation of labor or occurrence of any adverse reactions. The actions of prostaglandin engage both cervical ripening and uter- ine contractions, and therefore improve the success rate of vaginal delivery. Vaginal dinoprostone insert has also been demonstrated to be effective for induc- tion of labor in full-term pregnancy with an unfavor- able cervix [9,10]. Administration of PGE2 insert improved cervical ripening and reduced oxytocin aug- mentation and meconium-stained amniotic fluid when compared with the placebo/no treatment group [8]. Moreover, the delivery rate within 24 hours in the dinoprostone group (20—30%) was significantly higher than that in the placebo group (0–1%) [11,12]. Compared with those receiving intravenous oxytocin, patients receiving vaginal dinoprostone showed a higher rate of vaginal deliveries [13]. The dinoprostone vaginal insert has been used in China for many years. However, the sample size of this kind of studies was relatively smaller. Predictive factors of vaginal delivery in dinoprostone-induced labor varied among the stud- ies. Moreover, high heterogeneity of the studies due to the difference in population and obstetric manage- ment underlines the significance of contextualization in the data regarding pregnancy outcomes. Therefore, a retrospective study was performed using a large cohort of women at a tertiary maternity hospital, one of the top five maternal and child health institutions with the comprehensive service ability in China. We aimed to investigate the clinical use of vaginal dino- prostone insert and potentially relevant factors for vaginal delivery in dinoprostone-induced labor.
Materials and methods
This was a retrospective cohort study of dinoprostone- induced labor among women presented to Hubei Maternal and Child Health Hospital (Wuhan, China). A total of 14 954 puerperae were identified between January and August 2016, with an estimated annual delivery rate of around 22 430. Among them, 1730 pregnancies were delivered by labor induction with dinoprostone vaginal inserts. Finally, 1656 pregnancies that met the inclusion criteria were included in this study (Figure 1).
Figure 1. Flow diagram.
Inclusion criteria were (1) singleton gestation; (2) vertex presentation; (3) gestation age ≥38 weeks; (4) normal fetal heart rate (FHR); and (5) a Bishop Score ≤6. Patients met the following criteria were excluded (1) multiple gestations; (2) fetal anomaly, breech or transverse lie; (3) planned cesarean deliveries (for elective, medical, or obstetric reasons); (4) any ante- natal complication; (5) previous cesarean section or uterine surgery; and (6) any contraindication to dino- prostone (e.g. prostaglandin allergy) or vaginal deliv- ery. The choice of labor induction agent was made at the discretion of the obstetrician. This study was approved by the Ethical Committee of the Hubei Maternal and Child Health Hospital (Wuhan, China).
Dinoprostone vaginal insert (10 mg, Propess, Ferring, Saint-Prex, Switzerland) was placed into the posterior vaginal fornix for labor induction. The dose was repeated if the cervix was still unfavorable (Bishop Score ≤6). The suppository was removed when there was tachysystole or abnormal FHR tracing. Intravenous oxytocin augmentation was initiated in women with inadequate uterine contractions or failure to progress 30-min after removal of the insert. The failure to pro- gress was defined as failure of progressive cervical dilatation and fetal descent, and/or inefficient uterine activity. Continuous electronic fetal monitoring was performed during active labor. Obstetric complications included premature rupture of membranes (PROM), oligohydramnios, and gestational diabetes or hyper- tension at the time of labor induction.
Demographic data and maternal and fetal outcomes were obtained from the medical record of each partici- pant. Continuous data are presented as means ± SD. Categorical data were presented as frequencies and percentages. Univariate and multivariate logistic regres- sion analysis were carried out to identify the potential factors that may associate with the vaginal delivery.
Results
Clinical data
A total of 1656 women with dinoprostone-induced labor between January and August 2016 were finally included in this study. Among them, 396 (23.91%) women gave birth by cesarean section, while the remaining 1260 (76.09%) delivered vaginally. Baseline characteristics of the pregnancies between the vaginal delivery and cesarean section groups are presented in Table 1. Maternal age in the vaginal delivery group was significantly higher than that in the cesarean sec- tion group (28.15 ± 3.07 versus 28.02 ± 3.04; p ¼ .003). Baseline FHR and mean birth weight in the vaginal delivery group was significantly lower than those in the cesarean section group (144.11 ± 6.50 versus 145.08 ± 6.51; p ¼ .01; 3.35 ± 0.40 versus 3.50 ± 0.42; p < .001; respectively). In addition, women who gave birth by cesarean section were more likely to be nul- liparous (p < .001). Meanwhile, the medication time was significantly longer in women of cesarean section group (p < .001). There were no statistically significant differences between the two groups with respect to body mass index (BMI), biparietal diameter (BPD), baseline Bishop score, abortion, and use of oxytocin (p > .05, Table 1). However, women delivered vaginally had significantly higher Bishop Score at dinoprostone removal when compared with those in the cesarean section group (p < .001). The indications for labor induction were delayed pregnancy (42.14 versus 48.99%), PROM (32.70 versus 28.54%), oligohydramnios (10.08 versus 4.55%), gesta- tional diabetes or hypertension (14.44 versus 17.68%), and others (0.63 versus 0.25%) in vaginal and cesarean delivery groups. The data were statistically significantly different (p ¼ .001). No uterine hyperstimulation was observed in this study.Parturition time was further analyzed within the dif- ferent time points (12, 24, and 36 h). As shown in Obstetric complications include premature rupture of fetal membranes, oligohydramnios gestational hypertension and diabetes, and so on. ωindicates statistical significance at p < .05; ωωindicates statistical significance at p < .01; ωωωindicates statistical significance at p < .001. Discussion The present study was designed to investigate the relevant factors for vaginal delivery among Chinese women with dinoprostone-induced labor. A large sam- ple of 1656 women was analyzed in this study to explore the potential factors affecting the vaginal delivery in dinoprostone-induced labor. Vaginal deliv- ery can be influenced by various factors. It is still con- troversial whether or not induction of labor is associated with a decreased vaginal delivery rate [6]. Previous studies have demonstrated that the cesarean delivery rate was almost twofold [14] or higher [15] in women with labor induction when compared with spontaneous labor. Labor induction is correlated with a lower vaginal delivery rate [16,17]. However, other studies reported no change or even increased vaginal delivery rate [8,18]. Vaginal prostaglandins have been demonstrated to be highly effective in achieving cer- vical ripening and vaginal delivery [8]. A vaginal delivery rate of 55.3% was reported in vaginal prosta- glandins group which was significantly higher than that of intravenous oxytocin (34.0%) [19]. Moreover, a high vaginal delivery rate of 78.68% was observed in a Contemporary obstetric Cohort, and the vaginal deliv- ery rates were 68% and 90% in nulliparas and multipa- ras of labor induction [20]. In China, the vaginal delivery rate in dinoprostone-induced labor has also varied across studies. A vaginal delivery rate of 68.3% (54/79) was reported in a prospective study of preg- nant women undergoing labor induction at the Obstetrics Department of Sun Yat-sen Memorial Hospital of Sun Yat-Sen University [21]. However, a comparatively higher vaginal delivery rate of 81.9% (163/199) was observed in a multicenter randomized controlled trial [22]. Meanwhile, a recent multicenter study of 20 hospitals showed a vaginal delivery rate of 78.06% (1085/1390) in the Pearl River Delta of Guangdong Province [23]. Consistent with the above results, our study showed a vaginal delivery rate of 76.09% (1260/1656) for all participants, and 74.37% and 93.3% in nulliparas and multiparas. All these results are in favor of labor induction for vaginal deliv- ery, despite the difference in population, agents used for labor induction and parity (nulliparas: multiparas ¼ 1:1.07 versus 1:10.04). The vaginal delivery rate within 24 h has also been extensively studied in prostaglan- din-induced labor. A multicenter, prospective clinical study revealed a total vaginal delivery rate of 64.29% (180/280) within 24 h [24]. A cross-section study among 685 patients showed a similar 24-h vaginal delivery rate of 62.5% [25]. However, other studies revealed a 24-h vaginal delivery rate of 55.3% (26/47) [19,21], consistent with the results obtained in our study (55.61%, 921/1656). In this study, the vaginal delivery rates within 12 h, 24 h, and 36 h of labor induction were 26.75%, 55.68%, and 69.01%, respect- ively. This variation seems to be caused by the differ- ent population selected, agents used for labor induction, and parity. There was comparatively large proportion of primiparous women in our study due to the Family planning (one-child) policy in China, with a ratio of primiparas to multiparas of 10:1. With the implementation of two-child policy in China, the rate of vaginal delivery is expected to be increasing. Meanwhile, it also requires us to be better prepared for maternal and neonatal care. Nevertheless, our study did provide direct insights into the clinical use of dinoprostone vaginal insert, which may be helpful to better management of pregnancies in dinoprostone labor induction according to these valuable data. Success in labor induction or vaginal delivery may be influenced by several factors. Characteristics of pregnant women, such as race, parity, gestational age, BMI, fetal weight, and obstetrics management may affect labor induction [26]. In this study, multivariate regression analysis revealed that maternal age, parity, oxytocin use, and birth weight were important varia- bles for predicting the success of labor induction. Cervical status is an important factor in predicting the likelihood of successful induction of labor. As the best tool for assessing cervical status, the Bishop score has been demonstrated to be closely correlated with vagi- nal delivery [27]. Elevated BMI was associated with the increased rate of cesarean delivery, which turned out to be an independent risk factor for cesarean delivery [28]. However, our study showed that BMI and Bishop Score may not be the main predictor for vaginal deliv- ery in our population. The rate of successful labor induction was in consistent with the low Bishop score. In our hospital, dinoprostone is used with caution and only the doctors with more than 10 years of work experience were qualified to evaluate the Bishop score. Bishop score of patients recruited in this study ranged from 4 to 5. However, our results showed no effect of Bishop score on the vaginal delivery, consist- ent with the previous study [29]. Therefore, whether Bishop Score and BMI are the potential factors for vaginal delivery in dinoprostone-induced labor remain to be further studied with a specific population. Advanced maternal age has proven to be a risk fac- tor for failure of cervical ripening with PGE2 [30]. Fetal weight increased risk of cesarean delivery [31] and maternal complications [32]. A gestational age ≤39 was associated with a decreased risk of cesarean deliv- ery [31,33]. BPD was also defined as the risk factors for cesarean delivery in nulliparous women with an unfavorable cervix at or beyond 41 weeks of gestation [34]. Fetal weight was also identified to be an accurate parameter in prediction of cesarean delivery [35]. These results were partially consistent with the find- ings of our study, which indicated that maternal age, parity, oxytocin augmentation, and birth weight were important factors for vaginal delivery in dinoprostone- induced labor. However, the retrospective design of this study may prevent us from appropriate data col- lection and interpretation. Despite this, data of our study did show the efficiency of dinoprostone vaginal insert and potential predictors for vaginal delivery in our setting, which allows us to provide better service clinically.
Our study of 1656 Chinese pregnancies showed a vaginal delivery rate of 76.09% in dinoprostone- induced labor. Maternity age, parity, baseline fetal heart rate, and birth weight were significant factors that associated with vaginal delivery when vaginal dinoprostone insert was administrated for induction of labor. Our study provides a reliable clinical evidence for understanding the efficiency of dinoprostone and predictors of vaginal delivery with a larger sample size of Chinese patients. It may be helpful to guide the clinical use of dinoprostone and provide better service according to the clinical condition of pregnant women.