Elective Induction at 40 Weeks? "Decision-Based Evidence Making" Strikes Again

Today on Science & Sensibility, contributor Henci Goer takes a look at a systematic review released in spring that examined the impact of elective inductions on the cesarean rate.  Sound analysis or a house of cards?  Looking closer at the studies reviewed provides insight into how the conclusions reached by the investigators might need to be examined more closely.  Henci does that in this review.  Have you read this new systematic review?  Did you come to the same conclusions?  I invite you to share your thoughts in our comments section below. – Sharon Muza, Community Manager, Science & Sensibility.

flickr photo by catharticflux http://flickr.com/photos/catharticflux/2710057340  CC licensed.

flickr photo by catharticflux
http://flickr.com/photos/catharticflux/2710057340
CC licensed.

Yet another systematic review has surfaced “Induction of labor at full term in uncomplicated singleton gestations: a systematic review and metaanalysis of randomized controlled trials”  in which reviewers claim that electively inducing healthy women, this time at 40, not 41 weeks, offers benefits and doesn’t increase the cesarean surgery rate (Saccone 2015).

Let’s take a closer look.

Reviewers included five trials: three of them conducted in the 1970s (Cole 1975; Martin 1978; Tylleskar 1979), the fourth published in 2005 (Nielsen 2005), and the fifth in 2014 (Miller 2014). Already we have a problem. Induction management in the 1970s is sufficiently different from management today that results are unlikely to apply to contemporary care, but let’s get down to specifics. Two of the 1970s trials were deemed inadequate for inclusion in the Cochrane review of elective induction (Gulmezoglu 2012), and Miller 2014 is published only as an abstract. Quality systematic reviews exclude abstracts because they don’t provide enough information to evaluate the study. For these reasons, these three trials should be taken off the table..

That leaves us with the other two. Nielsen 2005 states in the title “Comparison of elective induction of labor with favorable Bishop scores versus expectant management: a randomized clinical trial” that it is confined to women with favorable Bishop scores. Anyone familiar with elective induction research should know that inducing when the cervix is ready to go won’t increase the cesarean rate compared with spontaneous onset, but inducing with an unripe cervix is a different story even when using cervical ripening agents (Dunne 2009; Jonsson 2013; Le Ray 2007; Macer 1992; Prysak 1998; Thorsell 2011; Vahratian 2005). As you move the induction date earlier and earlier, more and more women will have an unfavorable cervix, so including a trial limited to women with a ripe one will tilt the playing field in favor of induction. Furthermore, half the participants were multiparous women (113/226). Women with prior vaginal births will go on having vaginal births pretty much no matter what you do to them, which raises another point: inducing earlier means a higher percentage of the inductees will be first-time mothers because first time mothers tend to run longer pregnancies (Mittendorf 1990). Nulliparous women are much more vulnerable to anything that pushes them in the direction of a cesarean. That’s not all: The authors tell us that their hospital has a 7% cesarean rate for dystocia in women at term. If a hospital has a cesarean rate much higher than that—and many do—then results can’t be generalized to it, although, frankly, if the doctors are performing cesareans left and right, induction or spontaneous onset may not make much difference. In short, Nielsen (2005) doesn’t make a compelling argument for 40-week elective induction.

flickr photo by Selbe <3 http://flickr.com/photos/stacylynn/11944718954 shared under a Creative Commons (BY-NC-ND) license

flickr photo by Selbe
http://flickr.com/photos/stacylynn/11944718954
shared under a Creative Commons (BY-NC-ND) license

This brings us to the last trial, Cole (1975). Investigators allocated healthy women either to induction at 40 weeks (111 women) or 41 weeks (117 women). As with Nielsen, half the women had prior vaginal births. Despite being healthy, 22 women were induced for “obstetric complications” (undefined) in the 41-week induction group before reaching 41 weeks. If their doctors induced labor because they had concerns, then this would likely put the women at heightened risk for cesarean. Another 32 women were induced for exceeding 41 weeks. This means that overall, nearly half (46%) of the comparison group didn’t begin labor spontaneously, which would mask any association between induction and cesarean. Leaving the induction vs. spontaneous onset issue aside, the U.S. cesarean rate in the early 1970s was around 5%, which means it was a rare woman who would have one regardless of circumstances. Again, not exactly a strong case for inducing at 40 weeks.

What about the benefits? The best reviewers can come up with are a clinically meaningless reduction in mean blood loss (-58 ml); a lower rate of meconium-stained amniotic fluid (4% vs. 14%), not, mind you, a reduction in meconium aspiration, and therefore clinically meaningless as well; and an equally meaningless reduction in mean birth weight of -136 g (5 oz). If they had found something more impressive, surely they would have reported it.

Really? This merited a pre-publication media blast? Because it amounts to a textbook example of “garbage in, garbage out.” I can see only three possibilities to explain it: either 1) the authors and peer reviewers at the American Journal of Obstetrics and Gynecology (AJOG) don’t know as much as they should about what constitutes a quality systematic review, 2) they are so steeped in medical model thinking—“How early can we get the baby out of that treacherous maternal environment?”—that their judgment is compromised, or 3) we have a “pay no attention to what’s behind the curtain” effort to promote elective induction. I don’t know which is the more troubling, but if it’s the last one, the sad thing is that because it’s got the magic words “systematic review,” “meta-analysis,” and “randomized controlled trials” in the title, it’s likely to succeed.

References

Cole, R. A., Howie, P. W., & Macnaughton, M. C. (1975). Elective induction of labour. A randomised prospective trial. Lancet, 1(7910), 767-770.

Dunne, C., Da Silva, O., Schmidt, G., & Natale, R. (2009). Outcomes of elective labour induction and elective caesarean section in low-risk pregnancies between 37 and 41 weeks’ gestation. J Obstet Gynaecol Can, 31(12), 1124-1130.

Gulmezoglu, A. M., Crowther, C. A., Middleton, P., & Heatley, E. (2012). Induction of labour for improving birth outcomes for women at or beyond term. Cochrane Database Syst Rev, 6, CD004945.

Jonsson, M., Cnattingius, S., & Wikstrom, A. K. (2013). Elective induction of labor and the risk of cesarean section in low-risk parous women: a cohort study. Acta Obstet Gynecol Scand, 92(2), 198-203. doi: 10.1111/aogs.12043

Le Ray, C., Carayol, M., Breart, G., & Goffinet, F. (2007). Elective induction of labor: failure to follow guidelines and risk of cesarean delivery. Acta Obstet Gynecol Scand, 86(6), 657-665.

Macer, J. A., Macer, C. L., & Chan, L. S. (1992). Elective induction versus spontaneous labor: a retrospective study of complications and outcome. Am J Obstet Gynecol, 166(6 Pt 1), 1690-1696; discussion 1696-1697.

Martin, D. H., Thompson, W., Pinkerton, J. H., & Watson, J. D. (1978). A randomized controlled trial of selective planned delivery. Br J Obstet Gynaecol, 85(2), 109-113.

Miller, N., Cypher, R., Pates, J., & Nielsen, P. E. (2014). Elective induction of nulliparous labor at 39 weeks of gestation: a randomized clinical trial. Obstet Gynecol,132(Suppl 1):72S.

Mittendorf, R., Williams, M. A., Berkey, C. S., & Cotter, P. F. (1990). The length of uncomplicated human gestation. Obstet Gynecol, 75(6), 929-932.

Nielsen, P. E., Howard, B. C., Hill, C. C., Larson, P. L., Holland, R. H., & Smith, P. N. (2005). Comparison of elective induction of labor with favorable Bishop scores versus expectant management: a randomized clinical trial. J Matern Fetal Neontal Med, 18:59-64.

Prysak, M., & Castronova, F. C. (1998). Elective induction versus spontaneous labor: a case-control analysis of safety and efficacy. Obstet Gynecol, 92(1), 47-52.

Saccone, G., & Berghella, V. (2015). Induction of labor at full term in uncomplicated singleton gestations: a systematic review and metaanalysis of randomized controlled trials. American journal of obstetrics and gynecology.

Thorsell, M., Lyrenas, S., Andolf, E., & Kaijser, M. (2011). Induction of labor and the risk for emergency cesarean section in nulliparous and multiparous women. Acta Obstet Gynecol Scand, 90(10), 1094-1099. doi: 10.1111/j.1600-0412.2011.01213.x

Tylleskar, J., Finnstrom, O., Leijon, I, et al. (1979). Spontaneous labor and elective induction – a prospective randomized study. Effects on mother and fetus. Acta Obstet Gynaecol Scand, 58:513-518.

Vahratian, A., Zhang, J., Troendle, J. F., Sciscione, A. C., & Hoffman, M. K. (2005). Labor progression and risk of cesarean delivery in electively induced nulliparas. Obstet Gynecol, 105(4), 698-704.out

About Henci Goer

Henci Goer

Henci Goer, award-winning medical writer and internationally known speaker, is the author of The Thinking Woman’s Guide to a Better Birth and Optimal Care in Childbirth: The Case for a Physiologic Approach She is the winner of the American College of Nurse-Midwives “Best Book of the Year” award. An independent scholar, she is an acknowledged expert on evidence-based maternity care.

3 Comments

Henci, thanks for your review

July 15, 2015 07:00 AM by Michele L. Ondeck, MEd, RN, LCCE, FACCE
Henci, thanks for your review and encouraging our thinking!

Hi, thank you for your feedbac

September 4, 2015 07:00 AM by Gabriele Saccone
Hi, thank you for your feedback. I have just some comments: 1) You wrote "and Miller 2014 is published only as an abstract. Quality systematic reviews exclude abstracts because they don??t provide enough information to evaluate the study. For these reasons, these three trials should be taken off the table" Reply: This is not true. We emailed the authors (Miller et al) in order to get the informations we needed and to get unpublished and additional data. We clarified this in the manuscript. Not reported this in your blog is misunderstandable 2)Nielsen 2005 states in the title ??Comparison of elective induction of labor with favorable Bishop scores versus expectant management: a randomized clinical trial? that it is confined to women with favorable Bishop scores. We reported this info in Table 1. Moreover, we called for future call in nulliparous and in women without a favorable Bishop score. 3) The best reviewers can come up with are a clinically meaningless reduction in mean blood loss (-58 ml); a lower rate of meconium-stained amniotic fluid (4% vs. 14%), not, mind you, a reduction in meconium aspiration, and therefore clinically meaningless as well; and an equally meaningless reduction in mean birth weight of -136 g (5 oz). If they had found something more impressive, surely they would have reported it. We reported MSAF and not meconium-aspiration since that the original trials reported only data RE MSAF. However, it is known that the higher risk for MSAF, the higher risk for meconium-aspiration! FINALLY I can see only three possibilities to explain it: either 1) the authors and peer reviewers at the American Journal of Obstetrics and Gynecology (AJOG) don??t know as much as they should about what constitutes a quality systematic review, 2) they are so steeped in medical model thinking????How early can we get the baby out of that treacherous maternal environment????that their judgment is compromised, or 3) we have a ??pay no attention to what??s behind the curtain? effort to promote elective induction. I don??t know which is the more troubling, but if it??s the last one, the sad thing is that because it??s got the magic words ??systematic review,? ??meta-analysis,? and ??randomized controlled trials? in the title, it??s likely to succeed. REPLY - 1) I think we are quite expert in systematic review (please see our manuscript on PUBMED) 2) RCT are level 1 data and meta-analyses of RCTs are level 1A data. So the best data possibile with the highest-quality possible. So more data comments or opinion. The medicine should be based on these data. 3) this is not clear for me all the best Gabriele Saccone, MD

I appreciate your taking the t

September 5, 2015 07:00 AM by Henci Goer, BA
I appreciate your taking the time to respond to my criticism of your review. I have read your manuscript. I do not dispute that you and your co-author accurately represent and characterize the data from the trials you include, but this in no way negates their caveats and weaknesses; therefore I stand by what I wrote. To reiterate: ? Two of the five trials you include were deemed inadequate for inclusion in the Cochrane review of trials of elective induction of labor. While reviewers examining the same topic may disagree on which trials to exclude, nevertheless, the Cochrane is highly respected, and a Cochrane review rejection raises a red flag for me. ? One of the three remaining trials is available only as an abstract. In your first rebuttal point, you say you contacted the authors for additional data, but this doesn??t replace the editing, review, and acceptance process of a peer-reviewed journal. Indeed, in its section on selecting studies, the Cochrane manual for conducting reviews enjoins prospective reviewers to ??Retrieve? and ??examine? the ??full-text reports? of studies under consideration http://handbook.cochrane.org/chapter_7/7_2_3_a_typical_process_for_selecting_studies.htm, which by definition, eliminates abstracts from consideration. ? Both of the two remaining trials include women who have had prior vaginal births. Induction has little effect on cesarean rate in such women, but the same cannot be said of first-time mothers. Combining the two groups masks induction??s potential effect in this more vulnerable population. ? Both of the two remaining trials took place in environments where the cesarean rate was much lower than is commonly found in low-risk women in the vast majority of hospitals today, which means their results cannot be generalized to typical contemporary practice. ? One of the two remaining trials is confined to women with a favorable cervix. Women with a favorable cervix are at low risk for cesarean with labor induction in contrast to women with an unfavorable cervix even with treatment to ripen the cervix. ? In the other remaining trial, half the women assigned to expectant management in a population of healthy, low-risk women were induced, thereby obscuring any effect of elective induction on increasing the cesarean rate. Your review??s conclusion reads: ??Induction of labor at full-term in uncomplicated singleton gestations is not associated with increased risk of cesarean deliveries and has similar outcomes compared to expectant management.? Based on the flaws I list, I don??t see how this conclusion is justified. While rebuttal point #2 notes that you acknowledge that one of the trials is in women with a favorable cervix and further that you call for research into nulliparous women, the reality is that few people read past the abstract, and among those who do, fewer still will think through the implications of the limitations of the included trials or actually turn to the source material. The point of a systematic review, after all, is to spare practitioners the tedious process of sifting through the evidence on topics of interest in order to determine care. This means that the review??s conclusion that elective induction at 40 weeks ?? favorable cervix or not, first-time mother or not ?? is harmless will be accepted as a sound guide to practice and will be cited in support of intervening in healthy pregnancies. Last but not least, in point #3, you state that RCTs and systematic reviews are Level 1 data and as such are the ??best? and ??highest quality? data possible, which is true, at least for some, although not all, research questions, but it does not follow that studies with these designs should be uncritically accepted for this reason, as you seem to suggest. As with any other study type, they may be well or poorly conducted and as with any other study type, they are subject to the biases of the investigators. Your review is an example of this. From my perspective, having read your manuscript and the included trials, I think your conclusion could equally well read: ??Because of the limitations of the trials available to us, we were unable to determine whether elective induction of labor at 40 weeks gestation increased the cesarean rate in populations of interest such as nulliparous women or women with an unfavorable cervix; however, we were able to establish that doing so had no clinically significant benefits.?

To leave a comment, click on the Comment icon on the left side of the screen.  You must login to submit a comment.  

Recent Stories
What's to LOVE about LamazeLIVE 2019 - Start Planning Now!

Research Review: Evidence of Racial Disparities in Access to Human Milk in the NICU

Series: Brilliant Activities for Birth Educators - Introducing the Late Pregnancy Cast of Characters