Secondary Logo

Journal Logo

Original Research

When Has an Induction Failed?

Simon, Charla E. MD; Grobman, William A. MD, MBA

Author Information
doi: 10.1097/01.AOG.0000157437.10998.e7
  • Free

The rate of induction of labor in the United States has increased to at least 18%.1 This increase underscores the need to better understand the typical labor course during labor inductions, because the labor curve is different from that of spontaneous labors. For example, Friedman2 and Friedman and Sachtleben3 examined patients undergoing induction separately from spontaneously laboring patients and found a foreshortening of all phases of labor. Others have also reported that nulliparous patients requiring cervical ripening before induction have a longer active phase than spontaneously laboring patients.4

One potential consequence of protracted labors is a greater cesarean rate. Indeed, several authors have reported an increased risk of cesarean delivery in nulliparous women at term undergoing induction of labor regardless of the indication.5–7 One reason that a cesarean may occur that is particular to inductions is a “failed induction,” when women do not progress into the active phase of labor. The exact definition of a “failed induction” has not been established, but one group of investigators has suggested an outcomes-based definition.8 In their framework, a failed induction of labor may be diagnosed in women whose continued lack of progression into the active phase makes it unlikely that they would safely proceed to a vaginal delivery. This group suggested that nulliparous women could safely remain in the latent phase for up to 12 hours, whereas women who remained in the latent phase for longer periods had a quite low chance (13%) of subsequent vaginal delivery. This framework and outcome has not been further explored or validated.

The present study was performed to further determine the most clinically relevant definition of a failed induction of labor. Because the study was performed at an institution where there is not routine limitation of the latent phase to 12 hours and many clinicians allow latent phases of up to 24 hours before considering an induction to have failed, the study population allowed the 12-hour guideline to be explored and longer latent periods to be more fully evaluated.


Nulliparous patients undergoing induction of labor at Northwestern Memorial Hospital at 36 weeks or more of gestation with a singleton gestation in vertex presentation, between August 2002 and February 2003, were identified through the use of the labor and delivery log book. Patients with spontaneous painful contractions greater than 4 per hour or spontaneous rupture of membranes at the time of presentation were excluded. Other exclusion criteria included intrauterine fetal demise, prior vaginal delivery of any fetus beyond 16 weeks gestation, known major fetal anomaly, and prior hysterotomy. Data were abstracted from charts and delivery records of all women who met inclusion criteria. These data included demographic information, obstetric course, and maternal and neonatal morbidity.

The latent phase of labor was defined as starting at the time after both oxytocin had been initiated and amniotomy had been performed and concluding when a patient achieved a cervical examination of at least 4 cm dilation and 80% effacement or 5 cm dilation if less effaced.8 Patients requiring cervical ripening underwent placement of an extra-amniotic saline infusion catheter for 6 hours; this time was not included in the length of the latent phase. The induction protocol at our institution specifies starting oxytocin at 2 mU/min and moving upwards by 2 mU/min every 15 minutes until contractions occur every 2.5 to 3 minutes. Intrauterine pressure catheters and fetal scalp electrodes are not routinely used. Individual physicians were not required to make decisions during the labor course according to any preestablished protocol.

Maternal morbidity that was recorded included chorioamnionitis, postpartum hemorrhage, transfusion, and postpartum hysterectomy. Chorioamnionitis was defined as temperature in labor of 38°C or more and at least 1 of the following: uterine tenderness, fetal tachycardia (greater than 160 beats per minute), or purulent amniotic fluid. Postpartum hemorrhage was diagnosed by a clinical estimation of blood loss more than 500 mL for a vaginal delivery or more than 1,000 mL for a cesarean delivery. Neonatal outcomes included Apgar scores at 1 and 5 minutes and umbilical cord arterial and venous blood gases. Neonatal morbidity that was examined included brachial plexus injury, need for mechanical ventilation, necrotizing enterocolitis, intraventricular hemorrhage, seizure activity, neonatal sepsis, and death.

Before data collection, the study was approved by the institutional review board. Statistical analysis was performed using Minitab 13 (Minitab, Inc., State College, PA). Proportional data were compared using χ2 statistics. Normality of continuous data were assessed by the Kolmogorov-Smirnov test. Normally distributed data were compared using 1-way analysis of variance, whereas continuous data that were not normally distributed were compared with the Kruskal-Wallis test. All tests were 2-tailed, and P < .05 was considered statistically significant.


During the study period, 400 nulliparous women arrived at labor and delivery for induction of labor. Three patients underwent an emergency cesarean for fetal indications within minutes of being placed on an external monitor, and these women were excluded from analysis, leaving 397 records available for analysis. Characteristics of the 397 women included in this study are listed in Table 1.

Table 1
Table 1:
Characteristics of the Study Population

During the induction of labor, the mean maximum dosage of oxytocin used was 21.1 mU/min ± 9.5 mU/min. The median time from initiation of oxytocin until artificial rupture of membranes was 60 minutes (interquartile range [IQR] 0–340). The median interval from amniotomy until active labor was 384 minutes (IQR 240–609). The median interval from amniotomy to delivery was 655 minutes (IQR 457–926). In 236 women (59%) the delivery was spontaneous and vaginal, in 54 (15%) it was operative and vaginal, and in 103 (26%) it was by cesarean. The primary indications for cesarean delivery were as follows: 62 cases (60%) of arrest of dilation, 22 cases (21%) of arrest of descent, and 19 (19%) cases of nonreassuring fetal status.

The risk of cesarean delivery stratified by the length of the latent phase of labor is presented in Table 2. These cesarean rates increase as the length of the latent phase of labor extends and are significantly different from one another (P < .01). The risk of cesarean delivery for all patients who remained in the latent phase at a given interval after rupture of membranes is presented in Figure 1. The cohort of women continued to achieve a vaginal delivery in the majority of cases until 18 hours had passed without entering the active phase of labor. Of the 62 patients delivered by cesarean for arrest of dilation, only 8 patients (2% of the total population undergoing induction) never achieved the active phase of labor. In only 2 instances was this cesarean undertaken before a patient had remained in the latent phase for at least 12 hours.

Table 2
Table 2:
Risk of Cesarean Stratified by Length of Latent Phase of Labor
Fig. 1.
Fig. 1.:
Risk of cesarean for all patients who remained in the latent phase of labor at the designated interval after rupture of membranes.Simon. Failed Induction. Obstet Gynecol 2005.

Maternal and neonatal morbidity were examined as a function of the length of time (stratified by 6-hour intervals) that patients remained in the latent phase of labor. Chorioamnionitis and postpartum hemorrhage were related to the length of the latent phase of labor (Table 3). However, only 2 transfusions (0.5%) occurred, both in women who entered the active phase in less than 12 hours and eventually underwent a cesarean. There were no cesarean hysterectomies.

Table 3
Table 3:
Maternal Morbidity Stratified by Length of Latent Labor

The mean birth weight of the delivered neonates was 3,470 g, with a standard deviation of ± 486 g. Neonatal morbidity data are presented in Table 4. There were no significant differences in the frequency of meconium, special care nursery admission, or mean umbilical cord blood gases. There was also no significant association with arterial cord gas pH less than 7.00 or 5-minute Apgar score less than 4, although these events happened uncommonly (in 2 and 5 cases, respectively). Similarly, there was no statistical association between length of latent labor and any other serious neonatal morbidity, although only 3 infants had these morbid events (3 cases of mechanical ventilation and 2 cases of seizure). There were no neonatal deaths.

Table 4
Table 4:
Neonatal Morbidity Stratified by Length of Latent Phase of Labor


Although multiple risk factors for cesarean have been identified in women undergoing induction of labor, little information exists regarding the length of the latent phase and its value in predicting eventual route of delivery, as well as maternal and neonatal morbidity.9–11 Although some studies have examined the association between the length of the latent phase and maternal and neonatal morbidity, these studies have used a predefined definition of “prolonged latent phase,” and have not determined whether this definition was optimal. It remains unclear whether the criteria that individual physicians choose increase the cesarean rate or contribute to excess maternal and neonatal morbidity

One group of investigators has attempted to define a failed induction using an outcomes-based approach.8 Specifically, they examined the association of the length of the latent phase during labor induction with the frequency of cesarean and maternal and neonatal morbidity, and determined that “benefit to continued induction” accrues for as long as 12 hours in the latent phase; after this time, the frequency of vaginal delivery dropped precipitously to 13%. During their study, these investigators instituted a protocol to ensure that physicians waited through 12 hours of latent labor. It is possible that this benchmark time influenced physician behavior, and the low vaginal delivery rate after the 12-hour interval was dependent upon physician's internalization of the benchmark and less reflective of the optimal outcome that could be achieved. Also, the study's outcome analysis grouped patients with and without spontaneous rupture of membranes. Indeed, in the study by Rouse et al,8 the patients whose indication for induction was spontaneous rupture of membranes had a median length of the latent phase that was 5 hours shorter than those without spontaneous rupture.

In the present study, we chose to use a similar outcome-based methodology as Rouse et al,8 but chose to study only nulliparous women receiving oxytocin induction for reasons other than spontaneous rupture of membranes in an effort to analyze a group likely to be homogenous with regard to labor course. Moreover, we performed the study in an setting where it is common to wait at least 12 hours after initiation of induction, and typically longer, to perform a cesarean in the latent phase for failure to progress. This setting enhanced the ability to explore the course of induction for longer than 12 hours. Indeed, only 8 patients (2%) underwent cesarean in latent labor for an arrest disorder, and only 2 of these had a latent phase of less than 12 hours.

In some respects, the present results agree with those of Rouse et al.8 As those investigators demonstrated and the present study confirms, only a minority of women undergoing labor induction will have latent phases that persist beyond 12 hours. Moreover, the results of both studies illustrate that women who remain in the latent phase for up to 12 hours do not seem to demonstrate significantly increased maternal or neonatal morbidity.

However, the present study also suggests that continuation of the latent phase of labor for longer than 12 hours may provide patients with a reasonable chance of a vaginal delivery without the frequent occurrence of untoward health outcomes. In contrast to the patients in the study by Rouse et al8 who had a 13% chance of vaginal delivery if they remained in latent labor at 12 hours, the patients in the present study had a 59% chance of vaginal delivery. Even after 18 hours of latent labor, a vaginal delivery was achieved in 32% of the cases. It does not seem, moreover, that the benefits of a vaginal delivery were offset by unacceptable rates of maternal or neonatal complications. Although both postpartum hemorrhage and chorioamnionitis increased with the length of the latent phase, this did not translate into more frequent maternal transfusion, cesarean hysterectomy, or neonatal sepsis. Additionally, there were no differences in measures of neonatal morbidity.

Conclusions of the present results should not be overstated. This was an observational study at a single institution, and as such, definitive statements regarding the appropriate length of the latent phase during a labor induction and the definition of a “failed induction” cannot be made with certainty. Also, although no significant differences were found with regard to significant maternal and neonatal morbidity, events such as transfusion proved to be so uncommon that this study would have lacked the power to detect differences even if they existed. Additionally, as in the study by Rouse et al,8 egress from latent labor was defined not by an accelerated rate of cervical change but by cervical examination. Nonetheless, our results suggest that it is not unreasonable to allow up to18 hours of latent labor before recommending a cesarean for a “failed induction.”


1. American College of Obstetricians and Gynecologists. Induction of Labor. ACOG Practice Bulletin No. 10. Washington, DC: ACOG; 1999.
2. Friedman EA. Primigravid labor: a graphicostatistical analysis. Am J Obstet Gynecol 1955;6:567–89.
3. Friedman EA, Sachtleben MR. Determinant role of initial cervical dilatation on the course of labor. Am J Obstet Gynecol 1962;84:930–5.
4. Rinehart BK, Terrone DA, Hudson C, Isler CM, Larmon JE, Perry KG Jr. Lack of utility of standard labor curves in the prediction of progression during labor induction. Obstet Gynecol 2000;182:1520–6.
5. Cammu H, Martens G, Ruyssink G, Jean-Jacques A. Outcome after elective labor induction in nulliparous women: A matched cohort study. Am J Obstet Gynecol 2002;186:240–4.
6. Dublin S, Lydon-Rochelle M, Kaplan RC, Watts DH, Crichlow CW. Maternal and neonatal outcomes after induction of labor without an identified indication. Obstet Gynecol 2000;183:986–94.
7. Seyb S, Berka R, Socol M, Dooley S. Risk of cesarean delivery with elective induction of labor at term in nulliparous women. Obstet Gynecol 1999;94:600–7.
8. Rouse DJ, Owen J, Hauth JC. Criteria for failed labor induction: prospective evaluation of a standardized protocol. Obstet Gynecol 2000;96:671–7.
9. Reis F, Gervasi M, Florio P, Bracalente G, Fadalti M, Severi F, et al. Prediction of successful induction of labor at term: role of clinical history, digital examination, ultrasound assessment of the cervix, and fetal fibronectin assay. Am J Obstet Gynecol 2003;189:1361–7.
10. Maghoma J, Buchmann EJ. Maternal and fetal risks associated with prolonged latent phase of labour. J Obstet Gynaecol 2002;22:16–9.
11. Chelmow D, Kilpatrick SJ, Laros RK. Maternal and neonatal outcomes after prolonged latent phase. Obstet Gynecol 1993;81:486–91.

Cited By

This article has been cited 2 time(s).

Obstetrics & Gynecology
Factors Associated With Cesarean Delivery in Nulliparous Women With Type 1 Diabetes
Lepercq, J; Le Meaux, J; Agman, A; Timsit, J
Obstetrics & Gynecology, 115(5): 1014-1020.
PDF (213) | CrossRef
Obstetrics & Gynecology
Induction of Labor in the Nulliparous Gravida With an Unfavorable Cervix
Mercer, BM
Obstetrics & Gynecology, 105(4): 688-689.
PDF (114) | CrossRef
© 2005 The American College of Obstetricians and Gynecologists