Research
ajog.org
OBSTETRICS
Prostaglandin vaginal gel induction of labor comparing amniotomy with repeat prostaglandin gel Michael Beckmann, MBBS, FRANZCOG; Sailesh Kumar, MBBS, FRANZCOG, CMFM, FRCOG, FRCS, PhD; Vicki Flenady, RN, RM, MMedSci, PhD; Ellen Harker, MBBS OBJECTIVE: The purpose of this study was to compare 2 inductions of labor protocols. STUDY DESIGN: Women with live singleton pregnancies at 37 þ 0
weeks gestation who were booked for prostaglandins 2 (PGE2) vaginal gel induction with a modified Bishop’s score of 5 hours shorter in the amniotomy group (24.8 vs 30.0 hours; mean difference, 5.2 h; 95% confidence interval, e2.5 to e7.8). Fewer women in the amniotomy group remained undelivered after 24 hours (47.1% vs 67.7%; P < .01). However, the likelihood of an in-hours birth and the length of hospital stay were no different between the groups. There was no difference in the mode of birth or any of the secondary outcomes. CONCLUSION: After an initial dose of PGE2 vaginal gel, an amniotomy
(once technically possible) is associated with a shorter induction of laboreto-birth time compared with the use of repeat doses of PGE2. Administering more PGE2 with the aim of starting contractions or making the cervix “more favorable,” appears to have no clinical advantage.
RESULTS: Two hundred forty-five women were assigned randomly
into
either
the
amniotomy
(n
¼
121)
or repeat-PGE2
Key words: cervical ripening, induced labor
Cite this article as: Beckmann M, Kumar S, Flenady V, et al. Prostaglandin vaginal gel induction of labor comparing amniotomy with repeat prostaglandin gel. Am J Obstet Gynecol 2015;213:859.e1-9.
M
ore than 25% of women now undergo induction of labor (IOL).1-4 One of the most common methods of commencing an IOL is to use vaginal prostaglandin (PGE2 vaginal gel) From the Department of Obstetrics and Gynecology, Mater Health Services (all authors), and Mater Research Institute and the University of Queensland School of Medicine (Drs Beckmann and Kumar), Brisbane, QLD, Australia. Received April 27, 2015; revised June 12, 2015; accepted July 29, 2015. Supported by a $30,000 seeding grant from the JP Kelly Foundation. The authors report no conflict of interest. Corresponding author: Michael Beckmann, FRANZCOG. [email protected] 0002-9378/$36.00 ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2015.07.043
then perform an artificial rupture of membranes (ARM) followed by a Syntocinon infusion.5 It is well-established that, in commencing an IOL, cervical priming of an unripe cervix is of value; however, if the woman is not in labor 6 hours after administration of PGE2 vaginal gel, there is much variation in practice. Some clinicians give more PGE2 vaginal gel and then perform an ARM once the modified Bishop’s score is 7; others give no more PGE2 vaginal gel and attempt an ARM, regardless of the modified Bishop’s score. Proponents of giving more PGE2 vaginal gel argue that the likelihood of failed IOL is lower, maternal satisfaction is higher, need for Syntocinon is reduced, and discomfort is less if the cervix is more favorable before ARM. Supporters of an “early ARM” approach believe that clinical outcomes are unchanged and that the IOL-to-birth
time is delayed unnecessarily by giving more PGE2 vaginal gel. In the United States,3 Australia,6-9 and the United Kingdom,2 clinical practice guidelines provide no evidence-based recommendations regarding subsequent dosage and frequency of PGE2 vaginal gel for cervical priming. There is very little objective data of the value of repeated dosing of PGE2 vaginal gel; the single randomized controlled trial that addressed this question reported no difference in the duration of induced labor or any clinical outcomes.10 This study investigated women who underwent IOL who had already received an initial dose of PGE2 vaginal gel. The aim of this study was to determine whether there is advantage or disadvantage in continuing to administer more PGE2 rather than perform an ARM, if technically possible.
DECEMBER 2015 American Journal of Obstetrics & Gynecology
859.e1
Research M ATERIALS
AND
Obstetrics M ETHODS
A randomized controlled trial was undertaken at Mater Health Services Brisbane between March 2010 and August 2013. All women with live singleton pregnancies at or beyond 37 þ 0 weeks gestation who were booked for IOL with the use of PGE2 vaginal gel and with a modified Bishop’s score 4 hours delay” was chosen as the discriminator as to whether a protocol violation had occurred (Table 1).
R ESULTS Between March 2010 and August 2013, there were 2057 women booked for PGE2 vaginal gel IOL at Mater Health Services, Brisbane, Australia who were eligible for inclusion. Two hundred sixty-seven women were approached by the medical officer on the evening of their IOL, of whom 22 declined to participate. Sealed opaque envelopes were prepared, and 250 of the envelopes were opened. However, data were available for only 245 eligible women; it is not known the reason that these 5 additional envelopes were opened, which was not recognized until recruitment was ceased. Of the 245 women who were assigned randomly, 124 were allocated to the repeat-PGE2 group and 121 to the amniotomy group. Outcome data were available for all 245 participants. The flow of participants is presented in the
Obstetrics
ajog.org
Research
FIGURE 1
Trial protocol
PGE2, prostaglandins 2; ARM, artificial rupture of membranes. Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
Consolidated Standards of Reporting Trials diagram (Figure 2). The baseline characteristics of women who were assigned randomly to the repeat-PGE2 group and amniotomy
group did not differ with respect to age, body mass index, gestation at IOL, ethnicity, indication for IOL, or modified Bishop’s score at the commencement of IOL, either overall or in the subgroup of
those treated as-per-protocol. Most women were white with a mean booking body mass index in the overweight range. PGE2 IOL was performed most commonly because the woman was
TABLE 1
Protocol violations Repeat PGE2 group (n [ 61 protocol violations)a
Amniotomy group (n [ 35 protocol violations)
After initial PGE2 dose, >4 hour delay to undertake first review (n ¼ 29)
After initial PGE2 dose, >4 hour delay to undertake first review (n ¼ 25)
After subsequent doses of PGE2 (if indicated), >4 hour delay to undertake next review (n ¼ 11)
After subsequent doses of PGE2 (if indicated), >4 hour delay to undertake next reviews (n ¼ 4)
After ARM, >4 hour delay to commence Syntocinon (n ¼ 10)
After ARM, >4 hour delay to commence Syntocinon (n ¼ 2)
Performing ARM at 2nd or 3rd review when the modified Bishop’s score < 7 (n ¼ 32)
Administering more PGE2 vaginal gel at 2nd or 3rd review when the modified Bishop’s score 5 (n ¼ 4)
Failure to perform ARM at any review when the modified Bishop’s score 7 (n ¼ 0)
—
ARM, artificial rupture of membranes; PGE2, prostaglandins 2. a
Some participants had >1 protocol violation.
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
DECEMBER 2015 American Journal of Obstetrics & Gynecology
859.e3
Research
Obstetrics
ajog.org
FIGURE 2
Consolidated Standards of Reporting Trials flow diagram
Enrolment
Eligible (n=2057)
Excluded (n=1834) • Declined to participate (n=22) • Not approached (n=1812)
Randomized (n=245)
Allocated to repeat-PGE2 group (n=124) • Received allocated intervention (n=120) • Did not receive allocated intervention (n=4) (spontaneous labour after randomisation)
Allocation
Allocated to amniotomy group (n=121) • Received allocated intervention (n=120) • Did not receive allocated intervention (n=1) (spontaneous labour after randomisation)
Follow-Up Discontinued intervention (n=0)
Discontinued intervention (n=0)
Analysis Analysed (n=124) Excluded from analysis (n=0)
Analysed (n=121) Excluded from analysis (n=0)
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
postterm, and >80% of IOLs were performed for an identifiable clinical indication. In addition, the baseline characteristics of the 2147 eligible women who were not approached were analyzed and confirmed to be similar to the study participants (Table 2). The morning after an initial dose of PGE2 vaginal gel, 10.0% of those women who had been assigned randomly to the amniotomy group and 12.8% of those who were assigned randomly to the repeat-PGE2 group had a favorable cervix (modified Bishop’s score, 7). Of the women who were assigned
randomly to the amniotomy group, most (80.8%) received just a single dose of PGE2 vaginal gel; 15.0% received 2 doses; 4.2% received 3 doses, and 89% were administered a Syntocinon infusion. Of the women assigned randomly to the repeat-PGE2 group, only 36.7% received a single dose of PGE2 vaginal gel; 47.5% received 2 doses; 15.8% received 3 doses, and a similar proportion (87%) were administered a Syntocinon infusion. The primary outcome is presented in Table 3. Overall, the time for IOL-tobirth was >5 hours shorter in women
859.e4 American Journal of Obstetrics & Gynecology DECEMBER 2015
randomized to the amniotomy group (24.8 vs 30.0 hours; mean difference, 5.2 hours [95% confidence interval, e2.5 to e7.8]). This result was also significant in the subgroup of those treated as-per-protocol. In an examination of the time intervals that comprise the overall IOL-to-birth time, there was a trend toward a 1-hour shorter interval from ARM until birth in women who received repeat doses of PGE2. However, a 1-hour shorter interval from first dose PGE2 to next review and a 5- to 6-hour shorter interval between the first dose of PGE2 and rupture of membranes
ajog.org
TABLE 2
Baseline characteristics Intention-to-treat, n (%)
As-per-protocol, n (%)
Amniotomy group (n [ 121)
Repeat- PGE2 group (n [ 124)
P value
Amniotomy group (n [ 86)
Age, y
29.3 5.2
29.8 4.8
.39
28.8 5.0
29.2 4.4
.67
29.5 4.9
29.7 5.80
.48
Body mass index, kg/m2a
26.6 5.9
26.9 6.2
.89b
26.8 6.5
26.3 5.9
.71b
26.7 6.3
26.0 7.0
.18b
Nulliparity, n (%)
92 (76.0)
93 (75.0)
.15
61 (70.9)
47 (74.6)
.25
40.1 1.5
40.3 1.4
.29
40.1 1.5
40.5 1.4
.06
3.6 1.3
3.6 1.4
.98
3.5 1.4
3.7 1.4
.24
Baseline characteristic a
Gestation, wk
a
Modified Bishop’s scorea Ethnicity, n (%)
.85
P value
Recruited (n [ 245)
185 (75.5) b
Not recruited (n [ 1812)
1268 (70.0)
.31b
3.6 1.5
3.5 1.5
.56
.31
.22
86 (69.4)
69 (80.2)
43 (68.3)
179 (73.1)
1168 (64.5)
Asian
10 (8.3)
14 (11.3)
6 (7.0)
9 (14.3)
24 (9.8)
344 (19.0)
Aboriginal and/or Torres Strait Islander
2 (1.7)
3 (2.4)
2 (2.3)
1 (1.6)
5 (2.0)
41 (2.3)
Pacific Islander/Maori
4 (3.3)
4 (3.2)
1 (1.2)
0
8 (3.3)
48 (2.7)
12 (9.9)
17 (13.7)
8 (9.3)
29 (11.9)
211 (11.6)
Indication for induction of labor, n (%)
.76
.55
.12
64 (52.9)
68 (54.8)
43 (50.0)
39 (61.9)
132 (53.9)
853 (47.1)
Diabetes mellitus
20 (17.0)
20 (16.1)
16 (18.6)
9 (14.3)
40 (16.3)
215 (11.9)
Hypertension
6 (5.0)
10 (8.1)
4 (4.7)
4 (6.3)
16 (6.5)
198 (10.9)
Isoimmunization
0
1 (0.8)
0
1 (1.6)
1 (0.4)
7 (0.4)
Suspected small for gestational age/fetal growth restriction
4 (3.3)
1 (0.8)
5 (5.8)
1 (1.6)
5 (2.0)
67 (3.7)
Cholestasis
5 (4.1)
5 (4.0)
4 (4.7)
3 (4.8)
10 (4.1)
63 (3.4)
Antepartum hemorrhage
0
0
0
0
0
11 (0.6)
Social
3 (2.5)
2 (1.6)
3 (3.5)
2 (3.2)
5 (2.0)
167 (9.2)
Other
19 (15.7)
17 (13.7)
12 (14.0)
4 (6.4)
36 (14.7)
232 (12.8)
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
859.e5
Research
Data are given as mean standard deviation; b Mann-Whitney U test.
Obstetrics
DECEMBER 2015 American Journal of Obstetrics & Gynecology
Postterm
PGE2, prostaglandins E2.
.09
39.9 1.4
93 (76.8)
10 (15.8)
P value
40.1 1.4
White
Other
a
b
Repeat- PGE2 group (n [ 63)
Eligible for recruitment, n (%)
Research
Obstetrics
ajog.org
TABLE 3
IOLetoebirth time Intention-to-treat, mean (SD)
Time intervals
Amniotomy Repeat-PGE2 Mean difference (95% confidence group group (n [ 121) (n [ 124) interval)
As-per-protocol, mean (SD) Amniotomy Repeat-PGE2 Mean difference group (95% confidence group P value (n [ 86) (n [ 63) interval)
P value
Between 1st dose PGE2 and birth, hr
24.8 (8.3)
30.0 (12.0)
5.2 (2.5e7.8)
< .01
Between 1st dose PGE2 and next review
12.4 (3.6)
13.7 (5.0)
1.4 (0.3e2.5)
Between 1st dose PGE2 and ARM
14.2 (5.3)
20.6 (9.6)
Between review after 1st 12.3 (7.2) dose and birth Between ARM and birth
9.9 (4.9)
22.9 (8.3)
26.8 (10.8)
3.9 (0.7e7.1)
.02a
—
11.0 (2.0)
12.0 (3.1)
0.9 (0.1e1.7)
—
6.4 (4.4e8.4)
—
12.4 (4.5)
17.8 (8.2)
5.3 (3.2e7.5)
—
16.4 (10.1)
4.1 (1.9e6.3)
—
11.6 (7.3)
17.0 (11.5)
5.4 (2.3e8.4)
—
8.7 (5.2)
e1.3 (e2.6 to 0.1)
—
9.7 (4.8)
8.0 (4.8)
Between 1st dose PGE2 100.6 (32.4) 103.2 (32.6) e2.6 (e10.9 to e5.6) and discharge home, hr Delivery not achieved within 24 hr
57 (47.1)
84 (67.7)
0.70 (0.56e0.87)b
In-hours birth (8:00 AM-5:00 PM)
51 (42.2)
46 (37.1)
1.13 (0.83e1.55)b
.53
a
95.7 (33.4) 94.6 (30.1)
< .01a 29 (33.7) .42
41 (47.7)
e1.6 (e3.4 to 0.2) — 1.1 (e9.7 to 11.7) .85
33 (53.3)
0.64 (0.44e0.94) .02a
30 (47.6)
1.00 (0.71e1.41) .99
ARM, artificial rupture of membranes; PGE2, Prostaglandin E2. a
P < .05; b Presented as relative risk (95% confidence interval) and not mean difference (95% CI).
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
contributed to the overall finding of a shorter IOL-to-birth time in women who were assigned randomly to the amniotomy group. Despite a shorter duration of IOL and fewer women remaining undelivered after 24 hours in the amniotomy group, the likelihood of an in-hours birth was no different between the groups. In addition, the shorter IOL-to-birth time in the amniotomy group did not eventuate in a shorter overall length of hospital stay compared with women in the repeat-PGE2 group. The secondary outcomes are presented in Table 4. No differences were shown in the mode of birth. Despite trends toward fewer failed IOLs in the amniotomy group, there were no differences overall in the indications for cesarean delivery. In both the overall group and in the subgroup of those treated as-per-protocol, no differences were shown in the incidence of uterine hyperstimulation that required the removal of PGE2 vaginal gel and/or administration of acute tocolysis, use of epidural analgesia, need for
broad-spectrum antibiotics in labor, postpartum hemorrhage or third-stage blood loss, or admission of the baby to the nursery.
C OMMENT After an initial dose of PGE2 vaginal gel, most women did not have a favorable cervix. The administration of more PGE2 was associated with an IOL-tobirth time that was >5 hours longer than performing an amniotomy, if at all possible. Attempting to rupture the membranes when the modified Bishop’s score was 5 hours would be demonstrated. For example, women who were induced and birthed out-ofhours had a similar IOL-to-birth time compared with those who birthed inhours (28.8 vs 25.2 hours); however, their length of hospital stay was almost 11 hours longer (106.2 vs 95.4 hours). In a post-hoc analysis of a comparison of 90 women who were induced and who birthed in-hours with 90 women who birthed after-hours (matched for parity, mode of birth, IOL-to-birth time), an inhours birth was associated with a hospital length-of-stay that was 20 hours shorter (89.9 vs 110.6 hours). In addition, we also observed that women in the amniotomy group who were reviewed before 6:00 AM (as specified in the trial protocol) had a much higher likelihood of an in-hours birth (63.0% vs 38.4%) and a correspondingly shorter length of hospital stay (91.0 vs 103.6 hours). This underscores the pivotal role of the first
Obstetrics
ajog.org PGE2 review in achieving a timely and efficient induction process. It is hypothesized that an in-hours birth aligns with a hospitals’ morning discharge practices, maximizes the chances of a shorter hospital length-of-stay, and is more likely to be achieved if review occurs before 6:00 AM. After an initial dose of PGE2 vaginal gel, performing an amniotomy (once technically possible) is associated with a significantly shorter IOL-to-birth time compared with the use of repeat doses of PGE2 vaginal gel. The administration of more PGE2 vaginal gel with the aim of starting contractions or making the cervix “more favorable,” is not associated with clinical benefits over an early amniotomy. Future studies should explore the health care costs that are associated with IOL protocols and women’s experiences and preferences regarding duration of induced labor, preferred time of day/night to labor and give birth, and the experience of discomfort with ARM and repeat PGE2 vaginal gel dosing. ACKNOWLEDGMENTS The authors thank the women and their families who participated in the study, the research midwives (Jackie Chaplin and Anne Tremellen), and the midwives and doctors of the Mater Mothers’ Hospitals who supported this study.
REFERENCES 1. Laws P, Sullivan E. Australia’s mothers and babies. Canberra, Australia: Australian Institute of Health and Welfare; 2010. 2. National Institute of Clinical Excellence. Induction of labour clinical guideline. London: NICE; 2008.
3. American College of Obstetricians and Gynecologists. Induction of labor. ACOG Practice Bulletin no. 107. Obstet Gynecol 2009;114: 386-97. 4. World Health Organization. WHO recommendations for induction of labour. Geneva: Department of Reproductive Health and Research; 2011. 5. Thomas J, Fairclough A, Kavanagh J, Kelly AJ. Vaginal prostaglandin (PGE2 and PGF2a) for induction of labour at term. Cochrane Database Syst Rev 2014;6:CD003101. 6. Induction of labour Clinical Guideline. Queensland Government, 2011. Available at: http://www.health.qld.gov.au/qcg/documents/ g_iol5-1.pdf. Accessed August 27, 2015. 7. Induction of Labour Techniques - Clinical Guideline. Government of South Australia, 2013. Available at: http://www.sahealth.sa.gov.au/ wps/wcm/connect/publicþcontent/saþhealth þinternet/resources/policies/inductionþofþ labourþtechniquesþ-þsaþperinatalþpractice þguidelines. Accessed August 27, 2015. 8. Induction of Labour with Prostaglandin E2 Vaginal Gel (Prostin) Clinical Practice Guideline Victorian Government, 2014. Available at: http:// docs.health.vic.gov.au/docs/doc/79AB736038 E8281ACA2579880007FE4F/$FILE/induction_ of_labour_with_prostaglandin_e2_pge2_vaginal_ gel_prostin.pdf. Accessed August 27, 2015. 9. Use of prostaglandins for cervical ripening prior to the induction of labour. RANZCOG, 2012. Available at: http://www.ranzcog.edu.au/ doc/use-of-prostaglandins-for-cervical-ripeningprior-induction-of-labour.html. Accessed August 27, 2015. 10. MacKenzie IZ, Burns E. Randomised trial of one versus two doses of prostaglandin E2 for induction of labour: 1. clinical outcome. BJOG 1997;104:1062-7. 11. Gijsen R, Hukkelhoven CW, Schipper CM, Ogbu UC, de Bruin-Kooistra M, Westert GP. Effects of hospital delivery during off-hours on perinatal outcome in several subgroups: a retrospective cohort study. BMC Pregnancy Childbirth 2012;12:92. 12. De Graaf JP, Ravelli AC, Visser GH, et al. Increased adverse perinatal outcome of hospital delivery at night. BJOG 2010;117:1098-107.
Research
13. Armijo-Olivo S, Warren S, Magee D. Intention to treat analysis, compliance, drop-outs and how to deal with missing data in clinical research: a review. Phys Ther Rev 2009;14: 36-49. 14. Marley J. Editorial: efficacy, effectiveness, efficiency. Aust Prescr 2000;23:114-5. 15. Chua S, Arulkumaran S, Yap C, Selamat N, Ratnam SS. Premature rupture of membranes in nulliparas at term with unfavorable cervices: a double-blind randomized trial of prostaglandin and placebo. Obstet Gynecol 1995;86:550-4. 16. Doany W, McCarty J. Outpatient management of the uncomplicated postdate pregnancy with intravaginal prostaglandin E2 gel and membrane stripping. J Matern Fetal Med 1997;6:71-8. 17. Rayburn W, Gosen R, Ramadei C, Woods R, Scott J Jr. Outpatient cervical ripening with prostaglandin E2 gel in uncomplicated postdate pregnancies. Am J Obstet Gynecol 1988;158:1417-23. 18. Egarter C, Kofler E, Fitz R, Husslein P. Is induction of labor indicated in prolonged pregnancy? Results of a prospective randomised trial. Gynecol Obstet Invest 1989;27:6-9. 19. Mahmood TA, Dick MJ, Smith NC, Templeton AA. Role of prostaglandin in the management of prelabour rupture of the membranes at term. BJOG 1992;99:112-7. 20. Ohel G, Rahav D, Rothbart H, Ruach M. Randomised trial of outpatient induction of labor with vaginal PGE2 at 40-41 weeks of gestation versus expectant management. Arch Gynecol Obstet 1996;258:109-12. 21. Kravet SJ, Levine RB, Rubin HR, Wright SM. Discharging patients earlier in the day: a concept worth evaluating. Health Care Manag (Frederick) 2007;26:142-6. 22. Powell ES, Khare RK, Venkatesh AK, Van Roo BD, Adams JG, Reinhardt G. The relationship between inpatient discharge timing and emergency department boarding. J Emerg Med 2012;42:186-96. 23. Shepperd S, Lannin NA, Clemson LM, McCluskey A, Cameron ID, Barras SL. Discharge planning from hospital to home. Cochrane Database Syst Rev 2013;1:CD000313.
DECEMBER 2015 American Journal of Obstetrics & Gynecology
859.e9
ajog.org
OBSTETRICS
Prostaglandin vaginal gel induction of labor comparing amniotomy with repeat prostaglandin gel Michael Beckmann, MBBS, FRANZCOG; Sailesh Kumar, MBBS, FRANZCOG, CMFM, FRCOG, FRCS, PhD; Vicki Flenady, RN, RM, MMedSci, PhD; Ellen Harker, MBBS OBJECTIVE: The purpose of this study was to compare 2 inductions of labor protocols. STUDY DESIGN: Women with live singleton pregnancies at 37 þ 0
weeks gestation who were booked for prostaglandins 2 (PGE2) vaginal gel induction with a modified Bishop’s score of 5 hours shorter in the amniotomy group (24.8 vs 30.0 hours; mean difference, 5.2 h; 95% confidence interval, e2.5 to e7.8). Fewer women in the amniotomy group remained undelivered after 24 hours (47.1% vs 67.7%; P < .01). However, the likelihood of an in-hours birth and the length of hospital stay were no different between the groups. There was no difference in the mode of birth or any of the secondary outcomes. CONCLUSION: After an initial dose of PGE2 vaginal gel, an amniotomy
(once technically possible) is associated with a shorter induction of laboreto-birth time compared with the use of repeat doses of PGE2. Administering more PGE2 with the aim of starting contractions or making the cervix “more favorable,” appears to have no clinical advantage.
RESULTS: Two hundred forty-five women were assigned randomly
into
either
the
amniotomy
(n
¼
121)
or repeat-PGE2
Key words: cervical ripening, induced labor
Cite this article as: Beckmann M, Kumar S, Flenady V, et al. Prostaglandin vaginal gel induction of labor comparing amniotomy with repeat prostaglandin gel. Am J Obstet Gynecol 2015;213:859.e1-9.
M
ore than 25% of women now undergo induction of labor (IOL).1-4 One of the most common methods of commencing an IOL is to use vaginal prostaglandin (PGE2 vaginal gel) From the Department of Obstetrics and Gynecology, Mater Health Services (all authors), and Mater Research Institute and the University of Queensland School of Medicine (Drs Beckmann and Kumar), Brisbane, QLD, Australia. Received April 27, 2015; revised June 12, 2015; accepted July 29, 2015. Supported by a $30,000 seeding grant from the JP Kelly Foundation. The authors report no conflict of interest. Corresponding author: Michael Beckmann, FRANZCOG. [email protected] 0002-9378/$36.00 ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajog.2015.07.043
then perform an artificial rupture of membranes (ARM) followed by a Syntocinon infusion.5 It is well-established that, in commencing an IOL, cervical priming of an unripe cervix is of value; however, if the woman is not in labor 6 hours after administration of PGE2 vaginal gel, there is much variation in practice. Some clinicians give more PGE2 vaginal gel and then perform an ARM once the modified Bishop’s score is 7; others give no more PGE2 vaginal gel and attempt an ARM, regardless of the modified Bishop’s score. Proponents of giving more PGE2 vaginal gel argue that the likelihood of failed IOL is lower, maternal satisfaction is higher, need for Syntocinon is reduced, and discomfort is less if the cervix is more favorable before ARM. Supporters of an “early ARM” approach believe that clinical outcomes are unchanged and that the IOL-to-birth
time is delayed unnecessarily by giving more PGE2 vaginal gel. In the United States,3 Australia,6-9 and the United Kingdom,2 clinical practice guidelines provide no evidence-based recommendations regarding subsequent dosage and frequency of PGE2 vaginal gel for cervical priming. There is very little objective data of the value of repeated dosing of PGE2 vaginal gel; the single randomized controlled trial that addressed this question reported no difference in the duration of induced labor or any clinical outcomes.10 This study investigated women who underwent IOL who had already received an initial dose of PGE2 vaginal gel. The aim of this study was to determine whether there is advantage or disadvantage in continuing to administer more PGE2 rather than perform an ARM, if technically possible.
DECEMBER 2015 American Journal of Obstetrics & Gynecology
859.e1
Research M ATERIALS
AND
Obstetrics M ETHODS
A randomized controlled trial was undertaken at Mater Health Services Brisbane between March 2010 and August 2013. All women with live singleton pregnancies at or beyond 37 þ 0 weeks gestation who were booked for IOL with the use of PGE2 vaginal gel and with a modified Bishop’s score 4 hours delay” was chosen as the discriminator as to whether a protocol violation had occurred (Table 1).
R ESULTS Between March 2010 and August 2013, there were 2057 women booked for PGE2 vaginal gel IOL at Mater Health Services, Brisbane, Australia who were eligible for inclusion. Two hundred sixty-seven women were approached by the medical officer on the evening of their IOL, of whom 22 declined to participate. Sealed opaque envelopes were prepared, and 250 of the envelopes were opened. However, data were available for only 245 eligible women; it is not known the reason that these 5 additional envelopes were opened, which was not recognized until recruitment was ceased. Of the 245 women who were assigned randomly, 124 were allocated to the repeat-PGE2 group and 121 to the amniotomy group. Outcome data were available for all 245 participants. The flow of participants is presented in the
Obstetrics
ajog.org
Research
FIGURE 1
Trial protocol
PGE2, prostaglandins 2; ARM, artificial rupture of membranes. Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
Consolidated Standards of Reporting Trials diagram (Figure 2). The baseline characteristics of women who were assigned randomly to the repeat-PGE2 group and amniotomy
group did not differ with respect to age, body mass index, gestation at IOL, ethnicity, indication for IOL, or modified Bishop’s score at the commencement of IOL, either overall or in the subgroup of
those treated as-per-protocol. Most women were white with a mean booking body mass index in the overweight range. PGE2 IOL was performed most commonly because the woman was
TABLE 1
Protocol violations Repeat PGE2 group (n [ 61 protocol violations)a
Amniotomy group (n [ 35 protocol violations)
After initial PGE2 dose, >4 hour delay to undertake first review (n ¼ 29)
After initial PGE2 dose, >4 hour delay to undertake first review (n ¼ 25)
After subsequent doses of PGE2 (if indicated), >4 hour delay to undertake next review (n ¼ 11)
After subsequent doses of PGE2 (if indicated), >4 hour delay to undertake next reviews (n ¼ 4)
After ARM, >4 hour delay to commence Syntocinon (n ¼ 10)
After ARM, >4 hour delay to commence Syntocinon (n ¼ 2)
Performing ARM at 2nd or 3rd review when the modified Bishop’s score < 7 (n ¼ 32)
Administering more PGE2 vaginal gel at 2nd or 3rd review when the modified Bishop’s score 5 (n ¼ 4)
Failure to perform ARM at any review when the modified Bishop’s score 7 (n ¼ 0)
—
ARM, artificial rupture of membranes; PGE2, prostaglandins 2. a
Some participants had >1 protocol violation.
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
DECEMBER 2015 American Journal of Obstetrics & Gynecology
859.e3
Research
Obstetrics
ajog.org
FIGURE 2
Consolidated Standards of Reporting Trials flow diagram
Enrolment
Eligible (n=2057)
Excluded (n=1834) • Declined to participate (n=22) • Not approached (n=1812)
Randomized (n=245)
Allocated to repeat-PGE2 group (n=124) • Received allocated intervention (n=120) • Did not receive allocated intervention (n=4) (spontaneous labour after randomisation)
Allocation
Allocated to amniotomy group (n=121) • Received allocated intervention (n=120) • Did not receive allocated intervention (n=1) (spontaneous labour after randomisation)
Follow-Up Discontinued intervention (n=0)
Discontinued intervention (n=0)
Analysis Analysed (n=124) Excluded from analysis (n=0)
Analysed (n=121) Excluded from analysis (n=0)
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
postterm, and >80% of IOLs were performed for an identifiable clinical indication. In addition, the baseline characteristics of the 2147 eligible women who were not approached were analyzed and confirmed to be similar to the study participants (Table 2). The morning after an initial dose of PGE2 vaginal gel, 10.0% of those women who had been assigned randomly to the amniotomy group and 12.8% of those who were assigned randomly to the repeat-PGE2 group had a favorable cervix (modified Bishop’s score, 7). Of the women who were assigned
randomly to the amniotomy group, most (80.8%) received just a single dose of PGE2 vaginal gel; 15.0% received 2 doses; 4.2% received 3 doses, and 89% were administered a Syntocinon infusion. Of the women assigned randomly to the repeat-PGE2 group, only 36.7% received a single dose of PGE2 vaginal gel; 47.5% received 2 doses; 15.8% received 3 doses, and a similar proportion (87%) were administered a Syntocinon infusion. The primary outcome is presented in Table 3. Overall, the time for IOL-tobirth was >5 hours shorter in women
859.e4 American Journal of Obstetrics & Gynecology DECEMBER 2015
randomized to the amniotomy group (24.8 vs 30.0 hours; mean difference, 5.2 hours [95% confidence interval, e2.5 to e7.8]). This result was also significant in the subgroup of those treated as-per-protocol. In an examination of the time intervals that comprise the overall IOL-to-birth time, there was a trend toward a 1-hour shorter interval from ARM until birth in women who received repeat doses of PGE2. However, a 1-hour shorter interval from first dose PGE2 to next review and a 5- to 6-hour shorter interval between the first dose of PGE2 and rupture of membranes
ajog.org
TABLE 2
Baseline characteristics Intention-to-treat, n (%)
As-per-protocol, n (%)
Amniotomy group (n [ 121)
Repeat- PGE2 group (n [ 124)
P value
Amniotomy group (n [ 86)
Age, y
29.3 5.2
29.8 4.8
.39
28.8 5.0
29.2 4.4
.67
29.5 4.9
29.7 5.80
.48
Body mass index, kg/m2a
26.6 5.9
26.9 6.2
.89b
26.8 6.5
26.3 5.9
.71b
26.7 6.3
26.0 7.0
.18b
Nulliparity, n (%)
92 (76.0)
93 (75.0)
.15
61 (70.9)
47 (74.6)
.25
40.1 1.5
40.3 1.4
.29
40.1 1.5
40.5 1.4
.06
3.6 1.3
3.6 1.4
.98
3.5 1.4
3.7 1.4
.24
Baseline characteristic a
Gestation, wk
a
Modified Bishop’s scorea Ethnicity, n (%)
.85
P value
Recruited (n [ 245)
185 (75.5) b
Not recruited (n [ 1812)
1268 (70.0)
.31b
3.6 1.5
3.5 1.5
.56
.31
.22
86 (69.4)
69 (80.2)
43 (68.3)
179 (73.1)
1168 (64.5)
Asian
10 (8.3)
14 (11.3)
6 (7.0)
9 (14.3)
24 (9.8)
344 (19.0)
Aboriginal and/or Torres Strait Islander
2 (1.7)
3 (2.4)
2 (2.3)
1 (1.6)
5 (2.0)
41 (2.3)
Pacific Islander/Maori
4 (3.3)
4 (3.2)
1 (1.2)
0
8 (3.3)
48 (2.7)
12 (9.9)
17 (13.7)
8 (9.3)
29 (11.9)
211 (11.6)
Indication for induction of labor, n (%)
.76
.55
.12
64 (52.9)
68 (54.8)
43 (50.0)
39 (61.9)
132 (53.9)
853 (47.1)
Diabetes mellitus
20 (17.0)
20 (16.1)
16 (18.6)
9 (14.3)
40 (16.3)
215 (11.9)
Hypertension
6 (5.0)
10 (8.1)
4 (4.7)
4 (6.3)
16 (6.5)
198 (10.9)
Isoimmunization
0
1 (0.8)
0
1 (1.6)
1 (0.4)
7 (0.4)
Suspected small for gestational age/fetal growth restriction
4 (3.3)
1 (0.8)
5 (5.8)
1 (1.6)
5 (2.0)
67 (3.7)
Cholestasis
5 (4.1)
5 (4.0)
4 (4.7)
3 (4.8)
10 (4.1)
63 (3.4)
Antepartum hemorrhage
0
0
0
0
0
11 (0.6)
Social
3 (2.5)
2 (1.6)
3 (3.5)
2 (3.2)
5 (2.0)
167 (9.2)
Other
19 (15.7)
17 (13.7)
12 (14.0)
4 (6.4)
36 (14.7)
232 (12.8)
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
859.e5
Research
Data are given as mean standard deviation; b Mann-Whitney U test.
Obstetrics
DECEMBER 2015 American Journal of Obstetrics & Gynecology
Postterm
PGE2, prostaglandins E2.
.09
39.9 1.4
93 (76.8)
10 (15.8)
P value
40.1 1.4
White
Other
a
b
Repeat- PGE2 group (n [ 63)
Eligible for recruitment, n (%)
Research
Obstetrics
ajog.org
TABLE 3
IOLetoebirth time Intention-to-treat, mean (SD)
Time intervals
Amniotomy Repeat-PGE2 Mean difference (95% confidence group group (n [ 121) (n [ 124) interval)
As-per-protocol, mean (SD) Amniotomy Repeat-PGE2 Mean difference group (95% confidence group P value (n [ 86) (n [ 63) interval)
P value
Between 1st dose PGE2 and birth, hr
24.8 (8.3)
30.0 (12.0)
5.2 (2.5e7.8)
< .01
Between 1st dose PGE2 and next review
12.4 (3.6)
13.7 (5.0)
1.4 (0.3e2.5)
Between 1st dose PGE2 and ARM
14.2 (5.3)
20.6 (9.6)
Between review after 1st 12.3 (7.2) dose and birth Between ARM and birth
9.9 (4.9)
22.9 (8.3)
26.8 (10.8)
3.9 (0.7e7.1)
.02a
—
11.0 (2.0)
12.0 (3.1)
0.9 (0.1e1.7)
—
6.4 (4.4e8.4)
—
12.4 (4.5)
17.8 (8.2)
5.3 (3.2e7.5)
—
16.4 (10.1)
4.1 (1.9e6.3)
—
11.6 (7.3)
17.0 (11.5)
5.4 (2.3e8.4)
—
8.7 (5.2)
e1.3 (e2.6 to 0.1)
—
9.7 (4.8)
8.0 (4.8)
Between 1st dose PGE2 100.6 (32.4) 103.2 (32.6) e2.6 (e10.9 to e5.6) and discharge home, hr Delivery not achieved within 24 hr
57 (47.1)
84 (67.7)
0.70 (0.56e0.87)b
In-hours birth (8:00 AM-5:00 PM)
51 (42.2)
46 (37.1)
1.13 (0.83e1.55)b
.53
a
95.7 (33.4) 94.6 (30.1)
< .01a 29 (33.7) .42
41 (47.7)
e1.6 (e3.4 to 0.2) — 1.1 (e9.7 to 11.7) .85
33 (53.3)
0.64 (0.44e0.94) .02a
30 (47.6)
1.00 (0.71e1.41) .99
ARM, artificial rupture of membranes; PGE2, Prostaglandin E2. a
P < .05; b Presented as relative risk (95% confidence interval) and not mean difference (95% CI).
Beckmann. PGE2 vaginal gel induction: amniotomy vs repeat PGE2. Am J Obstet Gynecol 2015.
contributed to the overall finding of a shorter IOL-to-birth time in women who were assigned randomly to the amniotomy group. Despite a shorter duration of IOL and fewer women remaining undelivered after 24 hours in the amniotomy group, the likelihood of an in-hours birth was no different between the groups. In addition, the shorter IOL-to-birth time in the amniotomy group did not eventuate in a shorter overall length of hospital stay compared with women in the repeat-PGE2 group. The secondary outcomes are presented in Table 4. No differences were shown in the mode of birth. Despite trends toward fewer failed IOLs in the amniotomy group, there were no differences overall in the indications for cesarean delivery. In both the overall group and in the subgroup of those treated as-per-protocol, no differences were shown in the incidence of uterine hyperstimulation that required the removal of PGE2 vaginal gel and/or administration of acute tocolysis, use of epidural analgesia, need for
broad-spectrum antibiotics in labor, postpartum hemorrhage or third-stage blood loss, or admission of the baby to the nursery.
C OMMENT After an initial dose of PGE2 vaginal gel, most women did not have a favorable cervix. The administration of more PGE2 was associated with an IOL-tobirth time that was >5 hours longer than performing an amniotomy, if at all possible. Attempting to rupture the membranes when the modified Bishop’s score was 5 hours would be demonstrated. For example, women who were induced and birthed out-ofhours had a similar IOL-to-birth time compared with those who birthed inhours (28.8 vs 25.2 hours); however, their length of hospital stay was almost 11 hours longer (106.2 vs 95.4 hours). In a post-hoc analysis of a comparison of 90 women who were induced and who birthed in-hours with 90 women who birthed after-hours (matched for parity, mode of birth, IOL-to-birth time), an inhours birth was associated with a hospital length-of-stay that was 20 hours shorter (89.9 vs 110.6 hours). In addition, we also observed that women in the amniotomy group who were reviewed before 6:00 AM (as specified in the trial protocol) had a much higher likelihood of an in-hours birth (63.0% vs 38.4%) and a correspondingly shorter length of hospital stay (91.0 vs 103.6 hours). This underscores the pivotal role of the first
Obstetrics
ajog.org PGE2 review in achieving a timely and efficient induction process. It is hypothesized that an in-hours birth aligns with a hospitals’ morning discharge practices, maximizes the chances of a shorter hospital length-of-stay, and is more likely to be achieved if review occurs before 6:00 AM. After an initial dose of PGE2 vaginal gel, performing an amniotomy (once technically possible) is associated with a significantly shorter IOL-to-birth time compared with the use of repeat doses of PGE2 vaginal gel. The administration of more PGE2 vaginal gel with the aim of starting contractions or making the cervix “more favorable,” is not associated with clinical benefits over an early amniotomy. Future studies should explore the health care costs that are associated with IOL protocols and women’s experiences and preferences regarding duration of induced labor, preferred time of day/night to labor and give birth, and the experience of discomfort with ARM and repeat PGE2 vaginal gel dosing. ACKNOWLEDGMENTS The authors thank the women and their families who participated in the study, the research midwives (Jackie Chaplin and Anne Tremellen), and the midwives and doctors of the Mater Mothers’ Hospitals who supported this study.
REFERENCES 1. Laws P, Sullivan E. Australia’s mothers and babies. Canberra, Australia: Australian Institute of Health and Welfare; 2010. 2. National Institute of Clinical Excellence. Induction of labour clinical guideline. London: NICE; 2008.
3. American College of Obstetricians and Gynecologists. Induction of labor. ACOG Practice Bulletin no. 107. Obstet Gynecol 2009;114: 386-97. 4. World Health Organization. WHO recommendations for induction of labour. Geneva: Department of Reproductive Health and Research; 2011. 5. Thomas J, Fairclough A, Kavanagh J, Kelly AJ. Vaginal prostaglandin (PGE2 and PGF2a) for induction of labour at term. Cochrane Database Syst Rev 2014;6:CD003101. 6. Induction of labour Clinical Guideline. Queensland Government, 2011. Available at: http://www.health.qld.gov.au/qcg/documents/ g_iol5-1.pdf. Accessed August 27, 2015. 7. Induction of Labour Techniques - Clinical Guideline. Government of South Australia, 2013. Available at: http://www.sahealth.sa.gov.au/ wps/wcm/connect/publicþcontent/saþhealth þinternet/resources/policies/inductionþofþ labourþtechniquesþ-þsaþperinatalþpractice þguidelines. Accessed August 27, 2015. 8. Induction of Labour with Prostaglandin E2 Vaginal Gel (Prostin) Clinical Practice Guideline Victorian Government, 2014. Available at: http:// docs.health.vic.gov.au/docs/doc/79AB736038 E8281ACA2579880007FE4F/$FILE/induction_ of_labour_with_prostaglandin_e2_pge2_vaginal_ gel_prostin.pdf. Accessed August 27, 2015. 9. Use of prostaglandins for cervical ripening prior to the induction of labour. RANZCOG, 2012. Available at: http://www.ranzcog.edu.au/ doc/use-of-prostaglandins-for-cervical-ripeningprior-induction-of-labour.html. Accessed August 27, 2015. 10. MacKenzie IZ, Burns E. Randomised trial of one versus two doses of prostaglandin E2 for induction of labour: 1. clinical outcome. BJOG 1997;104:1062-7. 11. Gijsen R, Hukkelhoven CW, Schipper CM, Ogbu UC, de Bruin-Kooistra M, Westert GP. Effects of hospital delivery during off-hours on perinatal outcome in several subgroups: a retrospective cohort study. BMC Pregnancy Childbirth 2012;12:92. 12. De Graaf JP, Ravelli AC, Visser GH, et al. Increased adverse perinatal outcome of hospital delivery at night. BJOG 2010;117:1098-107.
Research
13. Armijo-Olivo S, Warren S, Magee D. Intention to treat analysis, compliance, drop-outs and how to deal with missing data in clinical research: a review. Phys Ther Rev 2009;14: 36-49. 14. Marley J. Editorial: efficacy, effectiveness, efficiency. Aust Prescr 2000;23:114-5. 15. Chua S, Arulkumaran S, Yap C, Selamat N, Ratnam SS. Premature rupture of membranes in nulliparas at term with unfavorable cervices: a double-blind randomized trial of prostaglandin and placebo. Obstet Gynecol 1995;86:550-4. 16. Doany W, McCarty J. Outpatient management of the uncomplicated postdate pregnancy with intravaginal prostaglandin E2 gel and membrane stripping. J Matern Fetal Med 1997;6:71-8. 17. Rayburn W, Gosen R, Ramadei C, Woods R, Scott J Jr. Outpatient cervical ripening with prostaglandin E2 gel in uncomplicated postdate pregnancies. Am J Obstet Gynecol 1988;158:1417-23. 18. Egarter C, Kofler E, Fitz R, Husslein P. Is induction of labor indicated in prolonged pregnancy? Results of a prospective randomised trial. Gynecol Obstet Invest 1989;27:6-9. 19. Mahmood TA, Dick MJ, Smith NC, Templeton AA. Role of prostaglandin in the management of prelabour rupture of the membranes at term. BJOG 1992;99:112-7. 20. Ohel G, Rahav D, Rothbart H, Ruach M. Randomised trial of outpatient induction of labor with vaginal PGE2 at 40-41 weeks of gestation versus expectant management. Arch Gynecol Obstet 1996;258:109-12. 21. Kravet SJ, Levine RB, Rubin HR, Wright SM. Discharging patients earlier in the day: a concept worth evaluating. Health Care Manag (Frederick) 2007;26:142-6. 22. Powell ES, Khare RK, Venkatesh AK, Van Roo BD, Adams JG, Reinhardt G. The relationship between inpatient discharge timing and emergency department boarding. J Emerg Med 2012;42:186-96. 23. Shepperd S, Lannin NA, Clemson LM, McCluskey A, Cameron ID, Barras SL. Discharge planning from hospital to home. Cochrane Database Syst Rev 2013;1:CD000313.
DECEMBER 2015 American Journal of Obstetrics & Gynecology
859.e9
