- Email: [email protected]
The efficacy of early amniotomy in nulliparous labor induction: a randomized controlled trial
Research
www. AJOG.org
OBSTETRICS
The efficacy of early amniotomy in nulliparous labor induction: a randomized controlled trial George A. Macones, MD; Alison Cahill, MD; David M. Stamilio, MD; Anthony O. Odibo, MD OBJECTIVE: The purpose of this study was to assess whether early am-
niotomy reduces the duration of labor or increases the proportion of subjects who are delivered within 24 hours in nulliparous patients who undergo labor induction. STUDY DESIGN: We performed a randomized controlled trial that com-
pared early amniotomy to standard management in nulliparous labor inductions. Inclusion criteria were nulliparity, singleton, term gestation, and a need for labor induction. Subjects were assigned randomly to early amniotomy (artificial rupture of membranes, ⱕ4 cm) or to standard treatment. There were 2 primary outcomes: (1) time from induc-
tion initiation to delivery and (2) the proportion of women who delivered within 24 hours. RESULTS: Early amniotomy shortens the time to delivery by ⬎2 hours
(19.0 vs 21.3 hours) and increases the proportion of induced nulliparous women who deliver within 24 hours (68% vs 56%). These improvements in labor outcomes did not come at the expense of increased complications. CONCLUSION: Early amniotomy is a safe and efficacious adjunct in nul-
liparous labor inductions. Key words: amniotomy, nulliparous labor induction
Cite this article as: Macones GA, Cahill A, Stamilio DM, et al. The efficacy of early amniotomy in nulliparous labor induction: a randomized controlled trial. Am J Obstet Gynecol 2012;207:403.e1-5.
R
ates of labor induction are rising. Recent data from the National Center for Health Statistics demonstrate an induction rate of ⬎22% in 2006, which was more than double what it was in 1990 and impacted ⬎900,000 births in the United States that year.1 Although many recent studies have offered evidence for the improvement in methods for labor induction, 2-7 the induction of labor remains a significant risk factor for cesarean delivery,8,9 which highlights the critical need for additional tools to refine induction practice. Amniotomy, generally thought to be “low-tech,” inexpensive, and safe, has received little research attention, and stud-
From the Department of Obstetrics and Gynecology, Washington University in St. Louis School of Medicine, St. Louis, MO. Received Feb. 27, 2012; revised June 1, 2012; accepted Aug. 21, 2012. The authors report no conflict of interest. Reprints: George A. Macones, MD, Professor and Chair, Department of Obstetrics and Gynecology, Washington University in St Louis, School of Medicine, 4911 Barnes Jewish Hospital Plaza, St. Louis, MO 63110. [email protected]. 0002-9378/$36.00 © 2012 Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2012.08.032
ies to date have neglected to investigate the efficacy of amniotomy in nulliparous women, despite the fact that more inductions are performed in nulliparous women than their multiparous peers.1 Based on clinical trials from those in spontaneous labor,10 early amniotomy timing in labor inductions may shorten the duration of labor. Clinical concern for the rare complications of umbilical cord prolapse and theoretic concerns that rupturing the membranes earlier will lead to a longer overall duration of rupture of membranes with potentially increased rates of chorioamnionitis, neonatal sepsis, and neonatal intensive care unit (NICU) admission have limited the empiric use of the practice of amniotomy in nulliparous labor inductions without level I evidence to support it. The specific aim of this study was to assess whether early amniotomy, defined as artificial rupture of the membranes, at ⱕ4-cm dilation, reduces the duration of labor or increases the proportion of subjects delivered within 24 hours in term nulliparous patients who undergo labor induction. We also sought to assess the safety of early amniotomy, as measured by adverse obstetric outcomes and measures of maternal and neonatal infectious morbidities.
M ETHODS We performed an unblinded, randomized controlled trial at Washington University in St. Louis and the University of Pennsylvania with approval from the Institutional Review Boards at both institutions. Inclusion criteria for this clinical trial were nulliparity, singleton, term gestation (defined as ⬎37 weeks 0 days), and a need for labor induction as determined by the treating physician. Exclusion criteria included HIV infection or cervical dilation of ⬎4 cm at admission examination. Eligible subjects were approached by trained research nurses and were offered enrollment into the clinical trial. Patients who consented were then randomly assigned to early amniotomy, which was defined as artificial rupture of the membranes at ⱕ4 cm or to standard management, which was amniotomy at ⬎4 cm dilation. In the early amniotomy group, amniotomy was performed as early as could be done safely. Decisions about the exact timing of rupture (after random assignment) were made by a team that included residents, fellows, and attendings. There were no specific instructions given regarding the timing of amniotomy in the standard treatment group; this decision was left to the treating physicians. The
NOVEMBER 2012 American Journal of Obstetrics & Gynecology
403.e1
Research
Obstetrics
www.AJOG.org
FIGURE
Flowchart 749 assessed for eligibility
164 excluded 84 not meeting inclusion criteria 80 refused to participate 0 other reasons
585 randomized
292 assigned to early amniotomy 270 received intervention as assigned 22 did not receive assigned intervention (Clinical concern)
293 assigned to standard management 280 received intervention as assigned 13 did not receive assigned intervention (Clinical concern)
0 lost to follow-up
0 lost to follow-up
292 included in analysis 0 excluded from analysis
293 included in analysis 0 excluded from analysis
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
primary method of induction was at the discretion of the treating physician, as were all other intrapartum/postpartum decisions. Random assignment was accomplished centrally. A permuted block randomization procedure was used to formulate assignment lists to assure close to equal numbers of subjects in each treatment group. A uniform block size of 4 was used. There were 2 primary outcomes. The first was time from initiation of induction, defined as time at delivery of the first induction method to delivery. The second was the proportion of women delivered within 24 hours from the initiation of induction. Although it may seem unusual to have 2 primary endpoints, we believed that both were equally clinically relevant and should be treated as primary outcomes. This decision was made a priori. We also assessed a number of 403.e2
secondary endpoints that included cesarean delivery rates and indications for cesarean delivery, chorioamnionitis (oral temperature ⬎38°C during labor), postpartum fever (oral temperature ⬎38°C on 2 separate occasions ⬎6 hours apart, ⬎24 hours from delivery), wound infection (defined as purulent discharge from the incision), endomyometritis (defined as fundal tenderness and fever that require treatment with antibiotics), NICU admission, and suspected neonatal sepsis. Trained research nurses collected all baseline information, information on the course of labor, and information on maternal and neonatal outcomes. Statistical analyses were performed with the intent-to-treat principle. Continuous outcomes were compared with the use of the Student t test or MannWhitney U dependent on their distributions; dichotomous outcomes were as-
American Journal of Obstetrics & Gynecology NOVEMBER 2012
sessed with 2 tests or Fisher exact test where appropriate. Time to delivery was not normally distributed and was compared with the use of the Mann-Whitney U test. Relative risks by group and 95% confidence intervals (CIs) were estimated for the percent of women who delivered within 24 hours and each of the secondary outcomes. Our sample size estimate was based on 1 of our primary outcomes: the proportion of women who delivered within 24 hours. We assumed an alpha error of .05, a beta error of .20 (or 80% power), an incidence of delivery within 24 hours of 50% based on published data, a minimum detectable relative risk of 0.75, and a 1:1 allocation ratio. With these assumptions in mind, we estimated that we would need 290 subjects per group. This strategy for sample size calculation gave us tremendous power for our second primary outcome, time to delivery (a continuous measure). Specifically, we estimated a priori that we had 95% power to detect a 2-hour reduction in time to delivery.
R ESULTS Seven hundred forty-nine women were screened for eligibility; 84 women (11.2%) were deemed ineligible by exclusion criteria. Of the 635 eligible nulliparous women, 585 women (92%) consented and were assigned randomly (Figure); 292 women were assigned to the early amniotomy group, and 293 women were assigned to standard treatment. Those who agreed to participate and those who did not were similar in terms of baseline characteristics (age, gestational age, preexisting medical conditions). The groups were well-balanced with regards to demographics and maternal medical conditions; the mean gestational age at induction was similar between the groups (Table 1). The admission cervical dilation was also similar between the groups. Likewise, the indications for labor induction were similar between the groups. The 2 most common indications for induction were ⬎40-week gestations and gestational hypertension/preeclampsia. The “other” category for indication for induction had a variety of uncommon indications for induction, which included maternal request/social factors (eg, dis-
Obstetrics
www.AJOG.org tance from hospital) and oligohydramnios (Table 2). Methods for induction were similar between the early amniotomy and standard treatment groups. Most women received misoprostol; approximately 30% of the women received a Foley bulb. The induction method categories are not mutually exclusive, because many women received multiple agents. In fact, 73% of women in both groups received ⬎1 agent for induction. As expected, with regards to timing of amniotomy, the early amniotomy group was ruptured earlier than the standard treatment group. Twentytwo women who were assigned randomly to early amniotomy were ruptured after 4 cm of dilation; 13 women who were assigned randomly to standard treatment were ruptured at ⬍4 cm. The primary results of this study are given in Table 3. The average time from the start of induction to delivery was shortened by slightly ⬎2 hours in the early amniotomy group (19.0 vs 21.3 hours, respectively; P ⫽ .04) compared with those with standard treatment. This difference in length of labor occurred mainly in the first stage of labor, which was defined as time from random assignment to complete cervical dilation. A higher proportion of women in the early amniotomy group were delivered within 24 hours of the initiation of induction (68% vs 56%, respectively; P ⫽ .002). Despite these differences in length of labor and delivery within 24 hours, there was no difference in the rate of cesarean deliveries. The rate of chorioamnionitis was increased numerically in the early amniotomy group (11.5% v. 8.5%, respectively; P ⫽ .22), although this difference was not statistically significant. Likewise, there were 2 cord prolapses in the early amniotomy group, and none in the standard treatment group. Selected neonatal outcomes are shown in Table 4. There was no increase in the rate of confirmed or suspected neonatal sepsis or admission to the special care nursery or NICU in women who underwent early amniotomy compared with those who experienced standard care. The infants born to women with cord prolapse both did well, with umbilical arterial pH ⬎ 7.20 and 5-minute Apgar score.
Research
TABLE 1
Baseline characteristics Early amniotomy (n ⴝ 292)
Variable
Standard therapy (n ⴝ 293)
P value
Maternal age, y
22.7 ⫾ 5.8
23.3 ⫾ 6.2
.17
African American race, %
72
68
.30
a
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
Diabetes mellitus, %
3.9
3.5
.81
5.2
.32
..............................................................................................................................................................................................................................................
Chronic hypertension, %
7.2
.............................................................................................................................................................................................................................................. 2a
Body mass index, kg/m
28 ⫾ 4.2
28 ⫾ 3.9
.90
GBS⫹, %
29
30
.66
Pitocin, %
93
93
.87
Misoprostol, %
67
69
.70
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
Cervidil, %
6.8
8.4
.45
..............................................................................................................................................................................................................................................
Foley bulb, %
27
30
.43
More than 1 agent, %
73
73
.80
Epidural anesthesia, %
92
94
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
.88
.............................................................................................................................................................................................................................................. a
Admission dilation, cm
1.1 ⫾ 1.03
1.1 ⫾ 0.97
.54
Dilation at rupture of membranes, cm
3.2
7.4
.001
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. a
Station at rupture of membranes
⫺1 ⫾ 1.2
⫺1 ⫾ 1.5
.50
.............................................................................................................................................................................................................................................. a
6.2 ⫾ 3.0
Cervical examinations, n
5.9 ⫾ 3.4
.67
.............................................................................................................................................................................................................................................. a
3323 ⫾ 516
Birthweight, g
3311 ⫾ 566
.78
.............................................................................................................................................................................................................................................. a
39.7 ⫾ 1.4
Gestational age, wk
39.5 ⫾ 1.4
.16
..............................................................................................................................................................................................................................................
GBS, Group B streptococcus. a
Data are given as mean ⫾ SD.
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
C OMMENT
Although these differences in duration of labor and proportion of women delivered within 24 hours may seem to be intermediate outcomes, we would argue that these are good surrogates for both maternal and neonatal outcomes. For example, it has been well-documented that the length of labor is correlated directly with maternal chorioamnionitis, postpartum fever, and neonatal infec-
The goal of our study was to assess the efficacy and safety of early amniotomy in nulliparous women who undergo labor induction. The results of this clinical trial indicate that early amniotomy shortens labor by approximately 2 hours, increases the proportion of women delivered within 24 hours, but does not impact the rate of cesarean deliveries. TABLE 2
Indications for induction Variable
Early amniotomy, %
Standard, %
P value
⬎40 wk
40
39
.83
Maternal medical indication
13
14
.72
Gestational hypertension/preeclampsia
29
27
.63
6
7
.63
12
13
.63
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
Intrauterine growth restriction
..............................................................................................................................................................................................................................................
Other
..............................................................................................................................................................................................................................................
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
NOVEMBER 2012 American Journal of Obstetrics & Gynecology
403.e3
Research
Obstetrics
www.AJOG.org
TABLE 3
Maternal and labor outcomes
Outcome
Early amniotomy (n ⴝ 292)
Standard (n ⴝ 293)
Randomization at delivery, hra
19.0 ⫾ 9.1
21.3 ⫾ 10.1
Delivery at ⬍24 hr, %
68
56
0.72
0.59–0.89
.002
Cesarean delivery, %
41
40
1.03
0.85–0.25
.75
Amnioinfusion, %
19
19
1.02
0.73–1.42
.87
Chorioamnionitis, %
11.5
1.35
0.83–2.21
.22
Cord prolapsed, %
0.7
0
Abruption, %
0.4
0.6
0.55
0.05–6.03
.62
Postpartum hemorrhage, %
8.2
10.1
0.81
0.48–1.36
.44
Relative risk
95% CI
P value .04
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
8.5
..............................................................................................................................................................................................................................................
.13
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
CI, confidence interval. a
Data are given as mean ⫾ SD.
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
tion.11-13 In addition, a 2-hour difference in labor length has important implications for resource utilization at the hospital level. For example, a 2-hour difference in time to delivery across many inductions could lead to a decrease in staffing of a labor and delivery unit. Last, there is likely enhanced patient satisfaction with shorter labors as well. The shorter duration of labor must be weighed against both maternal and neonatal safety concerns. There were a greater number of cases of maternal chorioamnionitis in the early amniotomy group, although this difference was not statistically significant. For this study, chorioamnionitis was defined purely on the basis of fever in labor. Given that fever is an objective measure, we do not believe that unblinding differentially affected the ascertainment of this outcome. Importantly, this numeric difference in chorioamnionitis did not lead to an increase in the rate of suspected neo-
natal sepsis or NICU admission, and there were no serious maternal consequences as a result of chorioamnionitis. Still, future studies should focus on the occurrence of chorioamnionitis with early amniotomy. There were also 2 cord prolapses in the early amniotomy group and none in the standard treatment group. Interestingly, one of the cord prolapses occurred in a patient in the early amniotomy group who actually was ruptured after 4 cm of dilation. Still, the occurrence of cord prolapse is concerning and warrants further study. Although there has been work on the role of amniotomy in spontaneous labor,14 surprisingly, there has been little previous work on the role of early amniotomy in the context of labor induction. There have been several clinical trials that have compared the combination of amniotomy and oxytocin with other methods of induction,15 but no studies that we are aware of that focus exclu-
TABLE 4
Neonatal outcomes Outcome
Early Relative amniotomy Standard risk 95% CI
P value
5-minute Apgar score
8.6
8.6
Suspected neonatal sepsis, %
9.7
11.1
0.87
0.54–1.41 .58
.72
Special care nursery admission, % 13.6
15.0
0.90
0.61–1.35 .63
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
CI, confidence interval. Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
403.e4
American Journal of Obstetrics & Gynecology NOVEMBER 2012
sively on the timing of amniotomy in labor induction. Amniotomy has been well-studied in the context of those women in spontaneous labor, which included the active management of labor; however, the work on active management of labor may not be generalized necessarily to those women whose labor is induced.16,17 Fraser et al10 performed a randomized clinical trial of amniotomy in nulliparous women in spontaneous labor. In that study, early amniotomy reduced the occurrence of dystocia (defined as 4 hours of cervical dilation of ⬍0.5 cm/hr) and shortened the first stage of labor by 136 minutes. The benefit of amniotomy was greatest in women with ⱖ3-cm initial dilation. A recent Cochrane review summarized the available information on early amniotomy in spontaneous labor.14 This pooled analysis did not support the notion that early amniotomy shortened the first stage of labor or reduced the rate of cesarean deliveries. The authors of this Cochrane review did suggest that additional research was needed. Our study has some notable strengths and limitations. First, our randomization strategy effectively balanced the study groups with respect to potentially confounding effects and, more importantly, maximally balanced them on unmeasured confounders. Second, the study is relatively large in size compared with many other studies of labor induction, which allowed us to reach an adequate sample to test our primary hypothesis. Third, we included a diverse group of patients with various indications for induction and various methods for induction, lending to the generalizability of the results. Last, our “simple” trial design with broad inclusion/exclusion criteria and treating physician decision-making should enhance both generalizability and the translation from efficacy in a study setting to real world effectiveness. There are some potential weaknesses that we believe deserve careful consideration. For practical reasons, this study was unblinded, which could have impacted the study results both with the potential for unequal distribution of cointerventions and the assessment of the secondary outcomes. We
Obstetrics
www.AJOG.org were reassured that the potential impact of cointerventions from practitioners because of group assignment based on the primary outcomes were likely minimal; there were no differences between the groups in rates of any of the induction agents that were used alone or in combination. With regard to the secondary outcomes, it is possible that the knowledge of treatment assignment may have influenced our somewhat subjective newborn infant outcomes. We believe that this potential bias, if present, would have resulted in more neonates in the early amniotomy group being admitted to the special care nursery, thus inflating the relative risk for NICU admission. The fact that we did not observe a difference in admissions is reassuring. Our sample size, although sufficiently large to test our hypothesis and compared with many studies of labor induction, was limited with respect to the confident assessment of differences in rare outcomes, such as cord prolapse. Approximately 10% of subjects in the early amniotomy group were ruptured after 4 cm; likewise, some women were assigned randomly to standard treatment were ruptured earlier. This misclassification is likely random and would therefore likely bias our results towards the null. With these strengths and limitations in mind, this study supports the following conclusions. First, relative to standard
treatment with later amniotomy, early amniotomy shortens the time to delivery by ⬎2 hours and increases the proportion of induced nulliparous women who were delivered within 24 hours. Based on these data, early amniotomy, when deemed safe by the practitioner, may be a useful adjunct in nulliparous labor inductions and may be incorporated into induction algorithms. f REFERENCES 1. Martin JA, Hamilton BE, Sutton PD, et al. Births: final data for 2006. Natl Vital Stat Rep 2009;57:1-102. 2. Gelber S, Sciscione A. Mechanical methods of cervical ripening and induction. Clin Obstet Gynecol 2006;49:642-57. 3. Rayburn WF. Prostaglandin E2 gel for cervical ripening and induction of labor: a critical analysis. Am J Obstet Gynecol 1989;160: 529-34. 4. Hofmeyr GJ, Gulmezoglu AM. Vaginal misoprostol for cervical ripening and induction of labour. Cochrane Database of Systematic Reviews 2003;1:CD000941. 5. Wing DA, Jones MM, Rahall A, Goodwin TM, Paul RH. A comparison of misoprostol and prostaglandin E2 gel for preinduction cervical ripening and labor induction. Am J Obstet Gynecol 1995;172:1804-10. 6. Kelly AJ, Kavanagh J, Thomas J. Vaginal prostaglandin (PGE2 and PGF2a) for induction of labour at term. Cochrane Database of Systematic Reviews 2003;4:CD003101. 7. Zhang J, Branch DW, Ramirez MM, et al. Oxytocin regimen for labor augmentation, labor progression and perinatal outcomes. Obstet Gynecol 2011;118:249-56.
Research
8. Moore LE, Rayburn WF. Elective induction of labor. Clin Obstet Gynecol 2006;49:698-704. 9. Luthy DA, Malmgren JA, Zingheim RW. Cesarean delivery after elective induction in nulliparous women: the physician effect. Am J Obstet Gynecol 2004;191:1511-5. 10. Fraser WD, Marcoux S, Moutquin JM, Christen A, the Canadian Early Amniotomy Study Group. Effect of early amniotomy on the risk of dystocia in nulliparous women. N Engl J Med 1993;328:1145-9. 11. Seaward PG, Hannah ME, Myhr TL, et al. International Multicentre Term Prelabor Rupture of Membranes Study: evaluation of predictors of clinical chorioamnionitis and postpartum fever in patients with prelabor rupture of membranes at term. Am J Obstet Gynecol 1997; 177:1024-9. 12. Tran SH, Cheng YW, Kaimal AJ, Caughey AB. Length of rupture of membranes in the setting of premature rupture of membranes at term and infectious maternal morbidity. Am J Obstet Gynecol 2008;198:700.e1-5. 13. Herbst A, Kallen K. Time between membrane rupture and delivery and septicemia in term neonates. Obstet Gynecol 2007;110: 612-8. 14. Smyth RMD, Alldred SK, Markham C. Amniotomy for shortening spontaneous labour. Cochrane Database of Systematic Reviews 2007;4:CD006167. 15. Howarth G, Botha DJ. Amniotomy plus intravenous oxytocin for induction of labour. Cochrane Database of Systematic Reviews 2001; 3:CD003250. 16. Frigoletto FD, Lieberman E, Lang JM, et al. A clinical trial of active management of labor. N Engl J Med 1995;333:745-50. 17. Lopez-Zeno JA, Peaceman AM, Adashek JA, Socol ML. A controlled trial of a program for the active management of labor. N Engl J Med 1992;326:450-4.
NOVEMBER 2012 American Journal of Obstetrics & Gynecology
403.e5
www. AJOG.org
OBSTETRICS
The efficacy of early amniotomy in nulliparous labor induction: a randomized controlled trial George A. Macones, MD; Alison Cahill, MD; David M. Stamilio, MD; Anthony O. Odibo, MD OBJECTIVE: The purpose of this study was to assess whether early am-
niotomy reduces the duration of labor or increases the proportion of subjects who are delivered within 24 hours in nulliparous patients who undergo labor induction. STUDY DESIGN: We performed a randomized controlled trial that com-
pared early amniotomy to standard management in nulliparous labor inductions. Inclusion criteria were nulliparity, singleton, term gestation, and a need for labor induction. Subjects were assigned randomly to early amniotomy (artificial rupture of membranes, ⱕ4 cm) or to standard treatment. There were 2 primary outcomes: (1) time from induc-
tion initiation to delivery and (2) the proportion of women who delivered within 24 hours. RESULTS: Early amniotomy shortens the time to delivery by ⬎2 hours
(19.0 vs 21.3 hours) and increases the proportion of induced nulliparous women who deliver within 24 hours (68% vs 56%). These improvements in labor outcomes did not come at the expense of increased complications. CONCLUSION: Early amniotomy is a safe and efficacious adjunct in nul-
liparous labor inductions. Key words: amniotomy, nulliparous labor induction
Cite this article as: Macones GA, Cahill A, Stamilio DM, et al. The efficacy of early amniotomy in nulliparous labor induction: a randomized controlled trial. Am J Obstet Gynecol 2012;207:403.e1-5.
R
ates of labor induction are rising. Recent data from the National Center for Health Statistics demonstrate an induction rate of ⬎22% in 2006, which was more than double what it was in 1990 and impacted ⬎900,000 births in the United States that year.1 Although many recent studies have offered evidence for the improvement in methods for labor induction, 2-7 the induction of labor remains a significant risk factor for cesarean delivery,8,9 which highlights the critical need for additional tools to refine induction practice. Amniotomy, generally thought to be “low-tech,” inexpensive, and safe, has received little research attention, and stud-
From the Department of Obstetrics and Gynecology, Washington University in St. Louis School of Medicine, St. Louis, MO. Received Feb. 27, 2012; revised June 1, 2012; accepted Aug. 21, 2012. The authors report no conflict of interest. Reprints: George A. Macones, MD, Professor and Chair, Department of Obstetrics and Gynecology, Washington University in St Louis, School of Medicine, 4911 Barnes Jewish Hospital Plaza, St. Louis, MO 63110. [email protected]. 0002-9378/$36.00 © 2012 Mosby, Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2012.08.032
ies to date have neglected to investigate the efficacy of amniotomy in nulliparous women, despite the fact that more inductions are performed in nulliparous women than their multiparous peers.1 Based on clinical trials from those in spontaneous labor,10 early amniotomy timing in labor inductions may shorten the duration of labor. Clinical concern for the rare complications of umbilical cord prolapse and theoretic concerns that rupturing the membranes earlier will lead to a longer overall duration of rupture of membranes with potentially increased rates of chorioamnionitis, neonatal sepsis, and neonatal intensive care unit (NICU) admission have limited the empiric use of the practice of amniotomy in nulliparous labor inductions without level I evidence to support it. The specific aim of this study was to assess whether early amniotomy, defined as artificial rupture of the membranes, at ⱕ4-cm dilation, reduces the duration of labor or increases the proportion of subjects delivered within 24 hours in term nulliparous patients who undergo labor induction. We also sought to assess the safety of early amniotomy, as measured by adverse obstetric outcomes and measures of maternal and neonatal infectious morbidities.
M ETHODS We performed an unblinded, randomized controlled trial at Washington University in St. Louis and the University of Pennsylvania with approval from the Institutional Review Boards at both institutions. Inclusion criteria for this clinical trial were nulliparity, singleton, term gestation (defined as ⬎37 weeks 0 days), and a need for labor induction as determined by the treating physician. Exclusion criteria included HIV infection or cervical dilation of ⬎4 cm at admission examination. Eligible subjects were approached by trained research nurses and were offered enrollment into the clinical trial. Patients who consented were then randomly assigned to early amniotomy, which was defined as artificial rupture of the membranes at ⱕ4 cm or to standard management, which was amniotomy at ⬎4 cm dilation. In the early amniotomy group, amniotomy was performed as early as could be done safely. Decisions about the exact timing of rupture (after random assignment) were made by a team that included residents, fellows, and attendings. There were no specific instructions given regarding the timing of amniotomy in the standard treatment group; this decision was left to the treating physicians. The
NOVEMBER 2012 American Journal of Obstetrics & Gynecology
403.e1
Research
Obstetrics
www.AJOG.org
FIGURE
Flowchart 749 assessed for eligibility
164 excluded 84 not meeting inclusion criteria 80 refused to participate 0 other reasons
585 randomized
292 assigned to early amniotomy 270 received intervention as assigned 22 did not receive assigned intervention (Clinical concern)
293 assigned to standard management 280 received intervention as assigned 13 did not receive assigned intervention (Clinical concern)
0 lost to follow-up
0 lost to follow-up
292 included in analysis 0 excluded from analysis
293 included in analysis 0 excluded from analysis
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
primary method of induction was at the discretion of the treating physician, as were all other intrapartum/postpartum decisions. Random assignment was accomplished centrally. A permuted block randomization procedure was used to formulate assignment lists to assure close to equal numbers of subjects in each treatment group. A uniform block size of 4 was used. There were 2 primary outcomes. The first was time from initiation of induction, defined as time at delivery of the first induction method to delivery. The second was the proportion of women delivered within 24 hours from the initiation of induction. Although it may seem unusual to have 2 primary endpoints, we believed that both were equally clinically relevant and should be treated as primary outcomes. This decision was made a priori. We also assessed a number of 403.e2
secondary endpoints that included cesarean delivery rates and indications for cesarean delivery, chorioamnionitis (oral temperature ⬎38°C during labor), postpartum fever (oral temperature ⬎38°C on 2 separate occasions ⬎6 hours apart, ⬎24 hours from delivery), wound infection (defined as purulent discharge from the incision), endomyometritis (defined as fundal tenderness and fever that require treatment with antibiotics), NICU admission, and suspected neonatal sepsis. Trained research nurses collected all baseline information, information on the course of labor, and information on maternal and neonatal outcomes. Statistical analyses were performed with the intent-to-treat principle. Continuous outcomes were compared with the use of the Student t test or MannWhitney U dependent on their distributions; dichotomous outcomes were as-
American Journal of Obstetrics & Gynecology NOVEMBER 2012
sessed with 2 tests or Fisher exact test where appropriate. Time to delivery was not normally distributed and was compared with the use of the Mann-Whitney U test. Relative risks by group and 95% confidence intervals (CIs) were estimated for the percent of women who delivered within 24 hours and each of the secondary outcomes. Our sample size estimate was based on 1 of our primary outcomes: the proportion of women who delivered within 24 hours. We assumed an alpha error of .05, a beta error of .20 (or 80% power), an incidence of delivery within 24 hours of 50% based on published data, a minimum detectable relative risk of 0.75, and a 1:1 allocation ratio. With these assumptions in mind, we estimated that we would need 290 subjects per group. This strategy for sample size calculation gave us tremendous power for our second primary outcome, time to delivery (a continuous measure). Specifically, we estimated a priori that we had 95% power to detect a 2-hour reduction in time to delivery.
R ESULTS Seven hundred forty-nine women were screened for eligibility; 84 women (11.2%) were deemed ineligible by exclusion criteria. Of the 635 eligible nulliparous women, 585 women (92%) consented and were assigned randomly (Figure); 292 women were assigned to the early amniotomy group, and 293 women were assigned to standard treatment. Those who agreed to participate and those who did not were similar in terms of baseline characteristics (age, gestational age, preexisting medical conditions). The groups were well-balanced with regards to demographics and maternal medical conditions; the mean gestational age at induction was similar between the groups (Table 1). The admission cervical dilation was also similar between the groups. Likewise, the indications for labor induction were similar between the groups. The 2 most common indications for induction were ⬎40-week gestations and gestational hypertension/preeclampsia. The “other” category for indication for induction had a variety of uncommon indications for induction, which included maternal request/social factors (eg, dis-
Obstetrics
www.AJOG.org tance from hospital) and oligohydramnios (Table 2). Methods for induction were similar between the early amniotomy and standard treatment groups. Most women received misoprostol; approximately 30% of the women received a Foley bulb. The induction method categories are not mutually exclusive, because many women received multiple agents. In fact, 73% of women in both groups received ⬎1 agent for induction. As expected, with regards to timing of amniotomy, the early amniotomy group was ruptured earlier than the standard treatment group. Twentytwo women who were assigned randomly to early amniotomy were ruptured after 4 cm of dilation; 13 women who were assigned randomly to standard treatment were ruptured at ⬍4 cm. The primary results of this study are given in Table 3. The average time from the start of induction to delivery was shortened by slightly ⬎2 hours in the early amniotomy group (19.0 vs 21.3 hours, respectively; P ⫽ .04) compared with those with standard treatment. This difference in length of labor occurred mainly in the first stage of labor, which was defined as time from random assignment to complete cervical dilation. A higher proportion of women in the early amniotomy group were delivered within 24 hours of the initiation of induction (68% vs 56%, respectively; P ⫽ .002). Despite these differences in length of labor and delivery within 24 hours, there was no difference in the rate of cesarean deliveries. The rate of chorioamnionitis was increased numerically in the early amniotomy group (11.5% v. 8.5%, respectively; P ⫽ .22), although this difference was not statistically significant. Likewise, there were 2 cord prolapses in the early amniotomy group, and none in the standard treatment group. Selected neonatal outcomes are shown in Table 4. There was no increase in the rate of confirmed or suspected neonatal sepsis or admission to the special care nursery or NICU in women who underwent early amniotomy compared with those who experienced standard care. The infants born to women with cord prolapse both did well, with umbilical arterial pH ⬎ 7.20 and 5-minute Apgar score.
Research
TABLE 1
Baseline characteristics Early amniotomy (n ⴝ 292)
Variable
Standard therapy (n ⴝ 293)
P value
Maternal age, y
22.7 ⫾ 5.8
23.3 ⫾ 6.2
.17
African American race, %
72
68
.30
a
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
Diabetes mellitus, %
3.9
3.5
.81
5.2
.32
..............................................................................................................................................................................................................................................
Chronic hypertension, %
7.2
.............................................................................................................................................................................................................................................. 2a
Body mass index, kg/m
28 ⫾ 4.2
28 ⫾ 3.9
.90
GBS⫹, %
29
30
.66
Pitocin, %
93
93
.87
Misoprostol, %
67
69
.70
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
Cervidil, %
6.8
8.4
.45
..............................................................................................................................................................................................................................................
Foley bulb, %
27
30
.43
More than 1 agent, %
73
73
.80
Epidural anesthesia, %
92
94
.............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
.88
.............................................................................................................................................................................................................................................. a
Admission dilation, cm
1.1 ⫾ 1.03
1.1 ⫾ 0.97
.54
Dilation at rupture of membranes, cm
3.2
7.4
.001
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. a
Station at rupture of membranes
⫺1 ⫾ 1.2
⫺1 ⫾ 1.5
.50
.............................................................................................................................................................................................................................................. a
6.2 ⫾ 3.0
Cervical examinations, n
5.9 ⫾ 3.4
.67
.............................................................................................................................................................................................................................................. a
3323 ⫾ 516
Birthweight, g
3311 ⫾ 566
.78
.............................................................................................................................................................................................................................................. a
39.7 ⫾ 1.4
Gestational age, wk
39.5 ⫾ 1.4
.16
..............................................................................................................................................................................................................................................
GBS, Group B streptococcus. a
Data are given as mean ⫾ SD.
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
C OMMENT
Although these differences in duration of labor and proportion of women delivered within 24 hours may seem to be intermediate outcomes, we would argue that these are good surrogates for both maternal and neonatal outcomes. For example, it has been well-documented that the length of labor is correlated directly with maternal chorioamnionitis, postpartum fever, and neonatal infec-
The goal of our study was to assess the efficacy and safety of early amniotomy in nulliparous women who undergo labor induction. The results of this clinical trial indicate that early amniotomy shortens labor by approximately 2 hours, increases the proportion of women delivered within 24 hours, but does not impact the rate of cesarean deliveries. TABLE 2
Indications for induction Variable
Early amniotomy, %
Standard, %
P value
⬎40 wk
40
39
.83
Maternal medical indication
13
14
.72
Gestational hypertension/preeclampsia
29
27
.63
6
7
.63
12
13
.63
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
Intrauterine growth restriction
..............................................................................................................................................................................................................................................
Other
..............................................................................................................................................................................................................................................
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
NOVEMBER 2012 American Journal of Obstetrics & Gynecology
403.e3
Research
Obstetrics
www.AJOG.org
TABLE 3
Maternal and labor outcomes
Outcome
Early amniotomy (n ⴝ 292)
Standard (n ⴝ 293)
Randomization at delivery, hra
19.0 ⫾ 9.1
21.3 ⫾ 10.1
Delivery at ⬍24 hr, %
68
56
0.72
0.59–0.89
.002
Cesarean delivery, %
41
40
1.03
0.85–0.25
.75
Amnioinfusion, %
19
19
1.02
0.73–1.42
.87
Chorioamnionitis, %
11.5
1.35
0.83–2.21
.22
Cord prolapsed, %
0.7
0
Abruption, %
0.4
0.6
0.55
0.05–6.03
.62
Postpartum hemorrhage, %
8.2
10.1
0.81
0.48–1.36
.44
Relative risk
95% CI
P value .04
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
8.5
..............................................................................................................................................................................................................................................
.13
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
CI, confidence interval. a
Data are given as mean ⫾ SD.
Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
tion.11-13 In addition, a 2-hour difference in labor length has important implications for resource utilization at the hospital level. For example, a 2-hour difference in time to delivery across many inductions could lead to a decrease in staffing of a labor and delivery unit. Last, there is likely enhanced patient satisfaction with shorter labors as well. The shorter duration of labor must be weighed against both maternal and neonatal safety concerns. There were a greater number of cases of maternal chorioamnionitis in the early amniotomy group, although this difference was not statistically significant. For this study, chorioamnionitis was defined purely on the basis of fever in labor. Given that fever is an objective measure, we do not believe that unblinding differentially affected the ascertainment of this outcome. Importantly, this numeric difference in chorioamnionitis did not lead to an increase in the rate of suspected neo-
natal sepsis or NICU admission, and there were no serious maternal consequences as a result of chorioamnionitis. Still, future studies should focus on the occurrence of chorioamnionitis with early amniotomy. There were also 2 cord prolapses in the early amniotomy group and none in the standard treatment group. Interestingly, one of the cord prolapses occurred in a patient in the early amniotomy group who actually was ruptured after 4 cm of dilation. Still, the occurrence of cord prolapse is concerning and warrants further study. Although there has been work on the role of amniotomy in spontaneous labor,14 surprisingly, there has been little previous work on the role of early amniotomy in the context of labor induction. There have been several clinical trials that have compared the combination of amniotomy and oxytocin with other methods of induction,15 but no studies that we are aware of that focus exclu-
TABLE 4
Neonatal outcomes Outcome
Early Relative amniotomy Standard risk 95% CI
P value
5-minute Apgar score
8.6
8.6
Suspected neonatal sepsis, %
9.7
11.1
0.87
0.54–1.41 .58
.72
Special care nursery admission, % 13.6
15.0
0.90
0.61–1.35 .63
.............................................................................................................................................................................................................................................. .............................................................................................................................................................................................................................................. ..............................................................................................................................................................................................................................................
CI, confidence interval. Macones. Early amniotomy in nulliparous labor induction. Am J Obstet Gynecol 2012.
403.e4
American Journal of Obstetrics & Gynecology NOVEMBER 2012
sively on the timing of amniotomy in labor induction. Amniotomy has been well-studied in the context of those women in spontaneous labor, which included the active management of labor; however, the work on active management of labor may not be generalized necessarily to those women whose labor is induced.16,17 Fraser et al10 performed a randomized clinical trial of amniotomy in nulliparous women in spontaneous labor. In that study, early amniotomy reduced the occurrence of dystocia (defined as 4 hours of cervical dilation of ⬍0.5 cm/hr) and shortened the first stage of labor by 136 minutes. The benefit of amniotomy was greatest in women with ⱖ3-cm initial dilation. A recent Cochrane review summarized the available information on early amniotomy in spontaneous labor.14 This pooled analysis did not support the notion that early amniotomy shortened the first stage of labor or reduced the rate of cesarean deliveries. The authors of this Cochrane review did suggest that additional research was needed. Our study has some notable strengths and limitations. First, our randomization strategy effectively balanced the study groups with respect to potentially confounding effects and, more importantly, maximally balanced them on unmeasured confounders. Second, the study is relatively large in size compared with many other studies of labor induction, which allowed us to reach an adequate sample to test our primary hypothesis. Third, we included a diverse group of patients with various indications for induction and various methods for induction, lending to the generalizability of the results. Last, our “simple” trial design with broad inclusion/exclusion criteria and treating physician decision-making should enhance both generalizability and the translation from efficacy in a study setting to real world effectiveness. There are some potential weaknesses that we believe deserve careful consideration. For practical reasons, this study was unblinded, which could have impacted the study results both with the potential for unequal distribution of cointerventions and the assessment of the secondary outcomes. We
Obstetrics
www.AJOG.org were reassured that the potential impact of cointerventions from practitioners because of group assignment based on the primary outcomes were likely minimal; there were no differences between the groups in rates of any of the induction agents that were used alone or in combination. With regard to the secondary outcomes, it is possible that the knowledge of treatment assignment may have influenced our somewhat subjective newborn infant outcomes. We believe that this potential bias, if present, would have resulted in more neonates in the early amniotomy group being admitted to the special care nursery, thus inflating the relative risk for NICU admission. The fact that we did not observe a difference in admissions is reassuring. Our sample size, although sufficiently large to test our hypothesis and compared with many studies of labor induction, was limited with respect to the confident assessment of differences in rare outcomes, such as cord prolapse. Approximately 10% of subjects in the early amniotomy group were ruptured after 4 cm; likewise, some women were assigned randomly to standard treatment were ruptured earlier. This misclassification is likely random and would therefore likely bias our results towards the null. With these strengths and limitations in mind, this study supports the following conclusions. First, relative to standard
treatment with later amniotomy, early amniotomy shortens the time to delivery by ⬎2 hours and increases the proportion of induced nulliparous women who were delivered within 24 hours. Based on these data, early amniotomy, when deemed safe by the practitioner, may be a useful adjunct in nulliparous labor inductions and may be incorporated into induction algorithms. f REFERENCES 1. Martin JA, Hamilton BE, Sutton PD, et al. Births: final data for 2006. Natl Vital Stat Rep 2009;57:1-102. 2. Gelber S, Sciscione A. Mechanical methods of cervical ripening and induction. Clin Obstet Gynecol 2006;49:642-57. 3. Rayburn WF. Prostaglandin E2 gel for cervical ripening and induction of labor: a critical analysis. Am J Obstet Gynecol 1989;160: 529-34. 4. Hofmeyr GJ, Gulmezoglu AM. Vaginal misoprostol for cervical ripening and induction of labour. Cochrane Database of Systematic Reviews 2003;1:CD000941. 5. Wing DA, Jones MM, Rahall A, Goodwin TM, Paul RH. A comparison of misoprostol and prostaglandin E2 gel for preinduction cervical ripening and labor induction. Am J Obstet Gynecol 1995;172:1804-10. 6. Kelly AJ, Kavanagh J, Thomas J. Vaginal prostaglandin (PGE2 and PGF2a) for induction of labour at term. Cochrane Database of Systematic Reviews 2003;4:CD003101. 7. Zhang J, Branch DW, Ramirez MM, et al. Oxytocin regimen for labor augmentation, labor progression and perinatal outcomes. Obstet Gynecol 2011;118:249-56.
Research
8. Moore LE, Rayburn WF. Elective induction of labor. Clin Obstet Gynecol 2006;49:698-704. 9. Luthy DA, Malmgren JA, Zingheim RW. Cesarean delivery after elective induction in nulliparous women: the physician effect. Am J Obstet Gynecol 2004;191:1511-5. 10. Fraser WD, Marcoux S, Moutquin JM, Christen A, the Canadian Early Amniotomy Study Group. Effect of early amniotomy on the risk of dystocia in nulliparous women. N Engl J Med 1993;328:1145-9. 11. Seaward PG, Hannah ME, Myhr TL, et al. International Multicentre Term Prelabor Rupture of Membranes Study: evaluation of predictors of clinical chorioamnionitis and postpartum fever in patients with prelabor rupture of membranes at term. Am J Obstet Gynecol 1997; 177:1024-9. 12. Tran SH, Cheng YW, Kaimal AJ, Caughey AB. Length of rupture of membranes in the setting of premature rupture of membranes at term and infectious maternal morbidity. Am J Obstet Gynecol 2008;198:700.e1-5. 13. Herbst A, Kallen K. Time between membrane rupture and delivery and septicemia in term neonates. Obstet Gynecol 2007;110: 612-8. 14. Smyth RMD, Alldred SK, Markham C. Amniotomy for shortening spontaneous labour. Cochrane Database of Systematic Reviews 2007;4:CD006167. 15. Howarth G, Botha DJ. Amniotomy plus intravenous oxytocin for induction of labour. Cochrane Database of Systematic Reviews 2001; 3:CD003250. 16. Frigoletto FD, Lieberman E, Lang JM, et al. A clinical trial of active management of labor. N Engl J Med 1995;333:745-50. 17. Lopez-Zeno JA, Peaceman AM, Adashek JA, Socol ML. A controlled trial of a program for the active management of labor. N Engl J Med 1992;326:450-4.
NOVEMBER 2012 American Journal of Obstetrics & Gynecology
403.e5