DOI: 10.1055/s-0036-1597629 - volume 38 - Dezembro 2016
Ugo Indraccolo, Gennaro Scutiero, Pantaleo Greco
Cervical length assessment prior to induction is a component of Bishop's score that is already used to predict the success of the labor induction. A meta-analysis supports the ability of Bishop's score in predicting the success of labor induction.1 The rationale of the clinical cervical length assessment has been translated in the practice to assess the cervical length by sonography. Several studies support that sonographic cervical assessments are predictors of the success of the induction,2,3but there is no consensus regarding transvagi-nal ultrasonography as a predictor of the success of the induction. Moreover, to the best of our knowledge, only the study conducted by Kang et al4 has investigated the value of changes in the length of the cervix in predicting the success of the induction.
The aim of this short report is to demonstrate that the cervical shortening, as assessed by sonography, is a predictor for the success of the induction of labor. This additional evidence may be used in meta-analyses.
A sample of women undergoing labor induction for many indications by using dinoprostone (0.5 mg/3 ml - Pfizer Italia srl) intracervically were enrolled in the Obstetric and Gynecological Unit of University of Foggia (Ospedali Riunti of Foggia) between October 2006 and April 2008. The patients provided their consent to undergo trans-vaginal ultrasonography before and after labor induction for cervical assessment. Before the induction, the cervix was measured by taking a sagittal scan from the inner os to the vaginal edge. Three hours after the induction, the same assessment was repeated. The patients were enrolled during the shifts of two Authors (UI and GS), who were able to repeat the measurements in the succeeding hours after labor induction. The measurements were collected with the transvaginal probe of the sonograph (Aloka 5500, ALOKA CO. LTD, Tokyo, Japan) set at 160° and 27 Hz, with the patient in the dorsal lythotomic position.
The length measured after the induction was subtracted from the value of the length measured before the induction in order to obtain the cervical shortening. This cervical shortening was introduced in the logistic regression analyses to calculate the odds ratios for vaginal delivery within 6, 12, 24 and 48 hours after the labor induction. Cervicometry before induction, multiparity, oxytocin infusion, and more administrations of prostaglandin agonists were also inde-pendent variables.
Receiver operating characteristic (ROC) curves were also calculated in order to assess the success of the prediction for the cervical shortening and the likelihood of a vaginal deliv-ery within 6, 12, 24 and 48 hours.
The Statistical Package for the Social Sciences (SPSS) software version 16.0 was used for the calculations. Alpha was set as < 0.05. The study was approved by the local Ethics Committee.
A sample of 63 women was enrolled (with a mean age of 28.3 years, 95% confidence interval [CI]: 26.9-29.6 years). 16 patients (25.4%) were multiparas (10 patients with only 1 previous vaginal birth; 5 patients with 2 previous vaginal births; and 1 patient with 3 previous vaginal births). The mean gestational age was 283.8 days (95% CI: 282.4-285.2 days). Inductions were indicated for patients: with over 41 weeks of gestation (20 patients - 31.7%); in the 41th week of gestation (15 patients - 23.8%); with premature rupture of membranes (14 patients - 22.2%); with hypertensive disor-ders of the pregnancy (2 patients - 3.2%); with sonographic diagnosis of intrauterine growth restriction or large for gestational age baby (5 patients - 7.9%); with oligohydram-nios, or abnormal cardiotocographic patterns, or reduction of fetal movements (10 patients - 15.9%). Sometimes, the indications for the inductions overlapped.
A total of 27 patients (42.9%) underwent cesarean sections, while vacuum delivery was needed in 1 case (0.02%). The indications for cesarean section were: obstructed labor (15 cases - 55.5%); abnormal cardiotocographic patterns (9 cases - 33.3%); and induction failure (4 cases - 14.8%). In one case, both dystocia and abnormal fetal heart rate were found. The indication for the vacuum delivery was failure to progress.
►Table 1 reports the adjusted odds ratios of the logistic regression analyses along with the descriptive statistics. The cervical shortening after the first administration of intra-cervical dinoprostone is an independent predictor of deliv-ery within 6, 24 and 48 hours, while multiple administration of intracervical dinoprostone increases the odds ratios of vaginal delivery within 12 and 24 hours. Each centimeter in the cervical shortening increases in 24.4% the odds of vaginal delivery within 6 hours; in 16.1% the odds of vaginal delivery within 24 hours; and in 10.5% the odds of vaginal delivery within 48 hours.
The areas under the curves (AUCs) for the cervical short-ening are: delivery within 6 hours - 0.784 (standard error: 0.160, p 1/4 0.036, ►Fig. 1); delivery within 12 hours - 0.621 (standard error: 0.100, p 1/4 0.160, ►Fig. 2); delivery within 24 hours - 0.690 (standard error: 0.081, p 1/4 0.014, ►Fig. 3); and delivery within 48 hours - 0.630 (standard error: 0.071, p 1/4 0.076, ►Fig. 4).
Many studies have assessed the ability of transvaginal ultra-sound examination of the cervix in predicting the successful outcome of induced labors. A wide review of these studies was recently done by Papillon-Smith and Abenhaim.5 Some au-thors have assessed the cervical length parameter alone or in combination with other maternal characteristics for predict-ing the outcome of induction, while others have compared the ultrasonographic cervical length versus the digital examina-tion as a predictor of the success of the induction. The level of heterogeneity among the studies makes it difficult to conclude that ultrasonography cervical length is suitable for wide clinical use,5 and a recent meta-analysis of two randomized trials6 does not prove the superiority of the ultrasonography versus Bishop's score. Even the elastosonography of the cervix has received some criticism regarding the prediction of the outcome of the induction.7 Moreover, if the induc-tion is needed, unfavorable ultrasonographic findings and unfavorable Bishop's scores do not change the clinical management of the patients. Practicing obstetricians need a tool to check the effectiveness of the first dose of prostaglandin agonists administered for labor induction, and ultrasonograpy cervical shortening seems the easier and more precise tool.
Table 1 Descriptive
statistics and results of the logistic regression analyses (adjusted odds ratio with 95% confidence intervals)
| Vaginal delivery within 6 hours | Vaginal delivery within 12 hours | Vaginal delivery within 24 hours | Vaginal delivery within 48 hours | ||
|---|---|---|---|---|---|
| means ± SD | OR95% CI of ORp | ||||
| Cervical shortening | 6.88 ± 6.49 mm | 1.2441.030–1.5030.024* | 1.0600.947–1.1870.308 | 1.1611.029–1.3100.016* | 1.1051.010–1.2090.030* |
| Cervicometry before induction | 26.21 ± 10.44 mm | 0.8230.673–1.0070.058 | 0.9400.862–1.0240.156 | 0.9330.859–1.0140.103 | 0.9530.888–1.0220.180 |
| rate (%) | OR95% CI of ORp | ||||
| Multiparity | 25.4% | 0.4900.008–30.9620.736 | 3.3200.527–20.9190.201 | 2.7910.556–14.0020.212 | 2.2980.624–8.4650.211 |
| Oxytocin infusion | 65.1% | 0.5890.006–55.7520.819 | 1.2490.219–7.1200.802 | 1.2110.284–5.1690.796 | 2.5050.815–7.7010.109 |
| > 1 dose of dinoprostone | 44.5% – 2 doses11.1% – 3 doses | Unreliable | 0.0290.004–0.2460.001* | 0.1970.053–0.7270.015* | 1.0650.326–3.4760.917 |
Abbreviations: CI, confidence interval; OR, odds ratio; SD, standard deviation. _significant results.
Fig. 1
Receiver operating characteristic (ROC) curve for delivery within 6 hours.
Fig. 2
Receiver operating characteristic (ROC) curve for delivery within 12 hours.
Fig. 3
Receiver operating characteristic (ROC) curve for delivery within 24 hours.
Fig. 4
Receiver operating characteristic (ROC) curve for delivery within 48 hours.
This short report confirms what was already reported by Kang et al4 in a wider sample of women with induced labors. Therefore, despite the biases related to the low sample size and the heterogeneity of the patients with induced labor in our study, we must reiterate the conclusion reached by Kang et al4: the presence of cervical shortening after the first induction is a predictor of a successful delivery if the induc-tion was repeated in the second day.
By assessing the AUCs, we found that the best predictions for vaginal delivery are achieved for births within 6 and 24 hours, while the cervical shortening poorly predicts vaginal delivery within 48 hours. Whether patients with poor cervi-cal shortening (less than 1 cm) 3 hours after induction should postpone the induction rather than receive more doses of dinoprostone should be assessed. A randomized trial should clarify this concern, in order to avoid the harms that come from the unnecessary use of prostaglandin agonists in pa-tients who are not responsive to them.
In conclusion, the greater the sonographic cervical short-ening 3 hours after labor induction, the higher the likelihood of vaginal delivery within 6, 24 and 48 hours.
This study was performed at the Institute of Obstetrics and Gynecology, Department of Surgical Sciences, University of Foggia (via Pinto - 71100, Foggia, Italy) during the shifts of the Authors. The Authors do not travel currently to the University of Foggia.