Materials and Methods
We performed a retrospective cohort analysis of all women who underwent induction of labor at BTH from October 2011 through October 2012. During this time, a standardized induction-of-labor protocol was implemented at our institution. The protocol allows evidence-based options for cervical ripening (Foley catheter alone, double balloon [Cook balloon] alone, or misoprostol) followed by a protocol for intrapartum management of latent labor that is a modification of the protocol described by Rouse et al. Once oxytocin is initiated, the protocol mandates amniotomy within 24 hours, followed by placement of an IUPC with titration of oxytocin to MVUs of 200-300. After 12 hours of oxytocin administration and ruptured membranes, a digital cervical examination is performed to assess the cervical dilation and effacement. If the examination is not consistent with active labor (defined as a cervix that is not at least 4 cm dilated and 90% effaced or 5 cm dilated, regardless of effacement), cesarean delivery for failed induction of labor is performed after a minimum of 12 hours and a maximum of 18 hours of oxytocin administration with ruptured membranes.
Women for this study were identified through the use of the labor and delivery log books. Inclusion criteria were gestational age of at least 24 weeks, live singleton gestation in cephalic presentation, unfavorable cervix (dilated no >2 cm), and initiation of oxytocin while still in latent labor. In selecting the time period for our retrospective review, we recognized that not all inductions were being managed per protocol after its implementation, so we selected a few months before protocol implementation and more after implementation in an effort to capture a relatively equal distribution of women in both groups for comparison.
The medical records were reviewed and categorized into 2 study groups: protocol adherent vs nonadherent. Protocol-adherent inductions were defined as those in which (1) amniotomy was performed within 24 hours of the start of induction of labor with oxytocin, (2) IUPC was placed at the time of amniotomy or within the following 6 hours and while the woman was still in latent labor, (3) oxytocin was titrated to achieve MVUs of 200-300 or progression of cervical dilation was deemed adequate at fewer MVUs, and (4) oxytocin was administered for at least 12 hours, and up to 18 hours, after membrane rupture before cesarean delivery for the indication of failed induction of labor was performed. A failed induction of labor was defined as delivery by cesarean section during latent labor as a result of failure to enter active labor. Additionally, if the induction of labor management was demonstrating adherence to the protocol and then an event occurred during latent labor that required cesarean delivery (eg, abnormal fetal heart rate tracing unresponsive to intrauterine resuscitation), the woman was classified as protocol adherent because the treatment had not demonstrated an unwillingness to follow the protocol, but rather the protocol was interrupted by an indicated cesarean delivery.
The primary outcome that was evaluated was the rate of failed induction of labor. Women were identified as having failed induction of labor if the operative report indication for their cesarean delivery was listed as such and if the medical record documentation supported the diagnosis. Secondary maternal outcomes included cesarean delivery rate, time from onset of induction to delivery, number of intrapartum cervical examinations, chorioamnionitis, endometritis, postpartum hemorrhage (defined as ≥500 mL for vaginal delivery and ≥1000 mL for cesarean delivery), and need for blood transfusion. Secondary neonatal outcomes included 1- and 5-minute Apgar scores, admission to and length of stay in the neonatal intensive care unit, shoulder dystocia, and neonatal death.
The 2 study groups were compared by t -test for continuous outcomes and χ 2 /Fisher exact test for grouped outcomes. Nonnormal or ordinal data were compared by Wilcoxon rank test. Logistic regression and generalized linear regression were used for multivariable analysis and for estimation of adjusted probability values. A probability value of < .05 was considered significant. All analysis was performed in SAS software (version 9.3; SAS Institute Inc, Cary, NC). This study was approved by the Baylor College of Medicine Institutional Review Board for Human Subject Research.
Results
During the 1-year study period, there were 3992 deliveries, of which 928 inductions were identified. Of the 928 inductions, 599 met the inclusion criteria. Of the 599 women included in the study for analysis, 369 were identified as having protocol-adherent treatment, and 230 were identified as having protocol-nonadherent treatment of their labor induction ( Figure ). Characteristics of the 599 women who were included in this study are listed in Table 1 . With the exception of higher parity in the protocol-adherent group, maternal demographic characteristics were similar in both groups. Use of cervical ripening and indications for induction differed between the groups ( Table 1 ).
Variable | Protocol adherent (n = 369) | Protocol nonadherent (n = 230) | P value |
---|---|---|---|
Characteristic | |||
Age, y a | 29.1 ± 6.5 | 28.3 ± 6.6 | .13 |
Ethnicity, n (%) | |||
Hispanic | 310 (84) | 187 (81) | .81 |
White | 7 (1.9) | 5 (2.2) | .81 |
African American | 39 (10.6) | 28 (12.2) | .81 |
Other ethnicity | 12 (3.3) | 10 (4.4) | .81 |
Parity, n (%) | |||
Nulliparous | 132 (36) | 112 (49) | .002 |
Multiparous | 237 (64) | 118 (51) | .002 |
Average gestational age, wk a | 38.1 ± 2.4 | 38.2 ± 2.2 | .81 |
Initial cervical dilation, cm | 1.16 ± 0.7 | 1.1 ± 0.76 | .1 |
Gestational age, n (%) | |||
<36 wk | 35 (9) | 22 (10) | .98 |
≥36 wk | 334 (90) | 208 (90) | .98 |
Cervical ripening, n (%) | .04 | ||
None | 235 (63.4) | 142 (61.7) | .63 |
Misoprostol | 109 (29.5) | 58 (25.2) | .25 |
Cook balloon | 20 (5.4) | 28 (12.2) | .003 |
Foley catheter | 4 (1) | 2 (1) | .99 |
Dinoprostone | 1 (<1) | 0 | .99 |
Previous cesarean delivery, n (%) | 39 (11) | 18 (7.8) | .26 |
Body mass index, kg/m 2 a | 33.6 ± 6.8 | 33.4 ± 6.5 | .74 |
Indication for induction, n (%) | |||
Diabetes mellitus | 57 (15) | 35 (15) | .94 |
Hypertensive disorders | 114 (31) | 71 (31) | .99 |
Abnormal fetal growth | 15 (4.07) | 8 (3.5) | .72 |
Cholestasis | 35 (9.5) | 8 (3.5) | .006 |
Preterm premature rupture of membranes | 4 (1.1) | 9 (3.9) | .04 |
Term rupture of membranes | 11 (3) | 25 (10.9) | < .001 |
Abnormal fetal testing | 45 (12.2) | 26 (11.3) | .74 |
Abnormal amniotic fluid | 29 (7.9) | 18 (7.8) | .99 |
All other indications | 59 (16) | 30 (13) | .32 |
Table 2 shows the outcomes of all patients who underwent induction of labor. On univariate analysis, the primary outcome of failed induction of labor among all patients was lower in the protocol-adherent than the protocol-nonadherent group (1.4% vs 7.8%, respectively; P < .0001). In addition, the median total time (in hours) from induction to delivery, defined as time from initiation of oxytocin to delivery, among all patients was also lower in the protocol-adherent than the protocol-nonadherent group (12 vs 15 hours, respectively; P < .001). A difference of 3 hours was observed. We also looked at the remainder of our secondary outcomes by univariate analysis. The mode of delivery was the only outcome found to be statistically different between the 2 groups. In the protocol-adherent group, the rate of spontaneous vaginal delivery was 75% vs 63% in the nonadherent group ( P = .004). In the protocol-adherent group, the overall rate of cesarean delivery was 22% vs 33% in the nonadherent group ( P = .002). There was no difference in the rate of operative vaginal delivery. The 2 groups had similar rates of cesarean delivery for the indication of arrest of active phase and arrest of descent. Cesarean delivery for a category II fetal heart rate tracing occurred more frequently in the protocol-adherent group ( P = .0002), and cesarean delivery for failed induction of labor occurred more frequently in the protocol-nonadherent group ( P = .003). At the time of cesarean delivery for the indication of failed induction of labor, the median cervical dilation was the same in both the protocol-adherent group (4 cm, with a range of 3–4 cm) and in the protocol-nonadherent group (4 cm, with a range of 0.5–5 cm; P = .94).
Outcome | Protocol adherent (n = 369) | Protocol nonadherent based (n = 230) | P value |
---|---|---|---|
Failed induction of labor, n (%) | 5 (1) | 18 (8) | < .0001 |
Mode of delivery, n (%) | |||
Spontaneous vaginal delivery | 274 (75) | 145 (63) | .004 |
Operative vaginal delivery | 14 (4) | 9 (4) | .94 |
Cesarean delivery | 80 (22) | 76 (33) | .002 |
Overall | .0005 | ||
Failed induction of labor | 5 (6) | 18 (24) | .003 |
Arrest of active phase | 20 (24) | 24 (32) | .37 |
Arrest of descent | 2 (2.5) | 7 (9) | .09 |
Category II fetal heart rate | 51 (62) | 25 (33) | .0002 |
Other | 3 (3.7) | 1 (1.3) | .62 |
Median time from induction of labor to delivery, h | 12 | 15 | < .001 a |
Apgar <7 at 1 min, n (%) | 25 (6.8) | 25 (10.9) | .08 |
Apgar <7 at 5 min, n (%) | 7 (1.9) | 6 (2.6) | .56 |
Median neonatal intensive care unit length of stay, d | 3 | 4 | .32 a |
Postpartum hemorrhage, n (%) | 87 (24) | 62 (27) | .37 |
Estimated blood loss of postpartum hemorrhage, mL b | 920 ± 482 c | 1108 ± 483 d | .001 |
Endometritis, n | 2 (0.5%) | 6 (2.6%) | .06 |
Chorioamnionitis, n (%) | 34 (9%) | 30 (13%) | .14 |
Digital cervical examinations, n b | 5.1 ± 2.2 | 6.1 ± 2.5 | < .0001 |
Shoulder dystocia, n (%) | 8 (3) | 6 (4) | .52 |
Blood transfusion, n (%) | 13 (3.5) | 14 (6) | .14 |
Control for possible confounders was made by stratification (on parity) and by multivariate analysis for which we adjusted for ruptured membranes and cervical ripening by including those variables on the regression model ( Table 3 ). The rate of failed induction remained significantly lower in the protocol-adherent group than the protocol-nonadherent group, both among nulliparous (3.8% vs 9.8%, respectively; P = .043) and multiparous women (0% vs 6%, respectively; P = .0004). The adjusted odds ratio for nulliparous women was 3.1 (95% confidence interval, 1.08–10.18). The median total time in hours from induction to delivery was also lower in the protocol-adherent than the protocol-nonadherent group within the nulliparous women (16.0 vs 19.5, respectively; P = .0002) and the multiparas (10.75 vs 12.25, respectively; P < .0001). On average, nulliparous women in the protocol-adherent group spent 3.5 fewer hours in labor than did the women in the nonadherent group. Multiparous women in the protocol-adherent group spent 1.5 fewer hours in labor than did the women in the nonadherent group. We looked at the remainder of our secondary outcomes in the multivariate analysis stratified by parity. The mode of delivery among nulliparous women was the only outcome found to be statistically different between the 2 groups. The protocol-adherent group had fewer cesarean deliveries (34% vs 66%) than the protocol-nonadherent group ( P = .04). Cesarean delivery for a category II fetal heart rate tracing occurred more frequently in the protocol-adherent group ( P = .02).
Outcome | Nulliparous women (n = 244) | Multiparous women (n = 355) | ||||
---|---|---|---|---|---|---|
Protocol adherent (n = 132) | Protocol nonadherent (n = 112) | P value a | Protocol adherent (n = 237) | Protocol nonadherent (n = 118) | P value a | |
Failed induction of labor, n (%) | 5 (3.8) | 11 (9.8) | .043 | 0 | 7 (6) | .0004 |
Cesarean delivery, n (%) | 44 (33.6) | 52 (66) | .04 | 36 (15) | 24 (20) | .25 |
Median time from induction of labor to delivery, h | 16.0 | 19.5 | .0002 | 10.75 | 12.25 | < .0001 |
Apgar score <7, n (%) | ||||||
1-minute | 10 (7.6) | 15 (13.4) | .11 | 15 (6.4) | 10 (8.5) | .64 |
5-minute | 2 (1.5) | 5 (4.5) | .22 | 5 (2.1) | 1 (0.9) | .32 |
Median neonatal intensive care unit length of stay, d | 3.0 | 4.0 | .35 | 4.0 | 4.0 | .99 |
Postpartum hemorrhage, n (%) | 38 (29) | 35 (31) | .52 | 49 (21) | 27 (23) | .37 |
Endometritis, n (%) | 0 | 5 (4.5) | .93 | 2 (0.9) | 1 (0.9) | .91 |
Chorioamnionitis, n (%) | 23 (17.6) | 21 (18.8) | .79 | 11 (4.7) | 9 (7.6) | .38 |
Digital cervical examinations, n b | 6.4 ± 2.3 | 7.03 ± 2.8 | .009 | 4.4 ± 1.9 | 5.3 ± 2.0 | < .0001 |
Shoulder dystocia, n (%) | 1 (1.2) | 2 (3.5) | .32 | 7 (3.5) | 4 (4.2) | .57 |
Blood transfusion, n (%) | 5 (3.8) | 7 (6.3) | .39 | 8 (3.4) | 7 (6) | .28 |