Ellice Lieberman’s 2005 research article on Fetal Position and Epidural Analgesia

Ellice Lieberman and her research colleagues looked at which came first, the posterior positioned baby or the epidural. In previous research studies it was somewhat unclear if women having epidural anesthesia (here called analgesia) were more likely to have a posterior baby or whether women with a posterior baby in labor were more likely to ask for an Epidural. While they tracked that question to an answer they observed multiple changes in fetal position at four times during labor. 

Research article title: Changes in Fetal Position During Labor and Their Association With Epidural Analgesia.

Research Review

Researchers: Ellice Lieberman, MD, DrPH, Karen Davidson, MD, Aviva Lee-Parritz, MD and Elizabeth Shearer, MPH. To learn about Dr. Lieberman copy and paste the following into a new internet window: http://www.hsph.harvard.edu/faculty/ElliceLieberman.html

Journal: Obstetrics and Gynecology, Vol. 105, No. 5., Part 1, pp 974-982. May 2005

Key Research Goal: “The purpose of our study was to evaluate changes in fetal position during nulliparous labor and to determine whether the use of epidural analgesia for pain relief is associated with a higher rate of abnormal fetal position at delivery.”

Method: First-time mothers who were at least 37 weeks pregnant allowed their babies positions to be recorded four times during labor and delivery. Women in the study could not have type 1 diabetes or leiomyomata. Subjects were not randomized and so, decided if they wanted epidurals themselves.

Women also answered questions about the pain they experienced in labor. They reported the degree of pain by giving a “0” for no pain to a1”10” if they perceived their pain as the “worst possible pain.” Also, the researchers asked about the location of where women felt the most pain, whether it was in their abdomen, back or vagina.

The first three times fetal position was determined by ultrasound imaging and the fourth time by the physician attending the birth:

# 1. At enrollment of the study, as soon after admission to L & D as possible,
# 2. At the time of their epidural (immediately before or within one hour after), or if they didn’t have an epidural, or didn’t have one yet, four hours after the first ultrasound,
# 3. Late labor, when the woman was close to full dilation, after 8 cm, and
# 4. The attending physician told the study representative immediately after birth what the position of the baby was at the start of delivery, before forceps or manual rotation was used, if tried at all.

Fetal position was not told to either the mothers or their care providers. The ultrasound pictures were examined by Karen Davidson who specializes in reading ultrasound pictures. Carefully, she looked for bone and brain landmarks to see which way the baby was facing in the pelvis. The anterioposterior diameter of the head is used in a standard way to mark direction of the head as a axis graph is superimposed over the picture. If the baby faced within 45 degrees of the transverse diameter (hip to hip) the baby was labeled transverse. If the baby faced within 45 degrees of the anterioposterior diameter, which would have the symphysis pubis in the front) the baby would be occiput posterior or occiput anterior according to whether the back of the head was posterior or anterior.

The head positions were categorized as occiput anterior, occiput posterior and occiput transverse. Occiput anterior is the group to which the other two were compared.

At the beginning of the study 1,766 first time mothers agreed to be in the study. But 204 of their first ultrasounds were not clear to what the fetal position actually was. And 33 other women had other situations that excluded them from the study. Ultimately, 1,562 women began the study.

The epidural analgesia given to the women was the narcotic mix routine for Boston’s Brigham and Women’s Hospital, the site of the study. Some women received a spinal-epidural combination. Both sets of women had similar numbers of OP babies at enrollment and at delivery. 92% of the study women had epidurals.

Labor started spontaneously for 698 of the women and was induced for 864 women.

What was learned

At enrollment, 92% of women were at 4 cm or less dilation, 8% were more dilated than 4 cm. At the enrollment ultrasound, 48.9% of babies were occiput transverse (OT); 26.9% were occiput anterior (OA); and 24% were occiput posterior (OP).

At the time of delivery, 79.8% were occiput anterior; 8.1% were occiput transverse; and 12.2% were occiput posterior.

The fetal position at or before 4 cm was not a strong predictor of anterior position at the end of labor. Approximately 80% of any position ended up being anterior at delivery. But the occiput posterior position at or before 4 cm was more likely to be posterior or transverse at delivery than OA. OP at enrollment was still OP 15.7% of the time at delivery, was OT 12.3% of deliveries and anterior only 8.8% of the time.

Looking backward from delivery, of the OP babies at birth 31% were OP at the enrollment ultrasound.

Of the OP babies at the next ultrasound, at about the time of the epidural or 4 hours after the enrollment ultrasound nearly 20% delivered in the occiput posterior position. Of the babies that were occiput posterior at 8 cm or more about 20%, again, were OP at delivery. Nearly 80% changed to either OT or OA between 8 cm and delivery.

Anterior babies occasionally turned OP, 5.4% of the time.

Occiput transverse babies remained in OT until delivery 8.3% of the time.

Fetal positions changes were common throughout labor, and over a third of babies (36% of all and 33% of babies who were OA at birth) were OP at sometime in labor.

Thirty-two (32%) percent of women whose babies were OP on all three ultrasounds were OP at delivery.

Epidural analgesia was associated with occiput posterior but not occiput transverse.

Of the women with epidurals 12.9% had OP babies at delivery compared to only 3.3% without epidural analgesia.

The posterior babies tended to be a little bigger than the anterior babies. Women with posterior babies had more longer labors. “However, at any given length of labor, women requesting epidural analgesia were more likely than women not requesting epidural analgesia to have the fetus delivered from an occiput posterior position.”

Women who entered labor spontaneously were asked about pain at enrollment. There was not a significant difference in pain between the women with OP babies than with OA babies at the enrollment interview. Nor did women with posterior babies report more back labor than women in other positions.

Mode of delivery did vary substantially. Women with OA babies were far more likely to give birth vaginally. Cesarean rates were 6.3% for occiput anterior; 64.7% for occiput posterior and 73.8% for occiput transverse head positions at delivery. The article notes the increase of the use of instrumental delivery (forceps or vacuum) and cesarean in the six years since an earlier study at the same hospital. For posterior births instrumental deliveries had increased 29% and cesareans were up 45%. Even anterior births experienced a 16% increase in instrumental deliveries and a 10% increase in cesarean deliveries.

The researchers comment, “The difference in mode of delivery between our earlier cohort [1998] and the current study likely represents a change in practice in our institution and may reflect changes in skills and attitudes in management of occiput posterior deliveries. The higher rate of cesarean we observed is consistent with the trends toward higher cesarean rates nationally.”


Ellice Lieberman points out an intriguing possibility, “Our finding that epidural analgesia may contribute to an increase in the occurrence of occiput posterior position represents a mechanism by which epidural analgesia may decrease spontaneous vaginal deliveries.”

A problem admitted in the study report was that 78% did not have late labor ultrasound data for a couple of reasons. Some women had a cesarean before the late labor ultrasound was due. Some of the films were unable to be read. In fact, each of the three ultrasound categories included a portion of ultrasounds were not able to be interpreted. For instance, only 51% of women had an interpretable late labor ultrasound. Only 77% of the women had an interpretable ultrasound at the epidural/4 hour ultrasound. And many of the 11.6% of excluded women from the initial ultrasound were excluded because the ultrasound was not able to be interpreted.


Discussion by Gail Tully on “Changes in Fetal Position During Labor and Their Association With Epidural Analgesia” for purposes of review and perspective.

Ellice Lieberman and her prestigious colleagues have offered us intriguing data that successfully answers their inquiry on the relationship between epidural analgesia and the incidence of occiput posterior positioning. This work verifies several things other studies have told us about the occiput posterior fetal position. An increase in instrumental and cesarean deliveries is found with posterior, as is a longer labor.

Her work also gives insight to things we have been confused about. For instance, posterior labor does not equal back labor. Lieberman does show that in early labor (up to about 4 cm) posterior position does not cause more back pain than an anterior position. She can be as accurate at this conclusion as the accuracy of the pain scale used during the enrollment interview. This study does not report on back pain patterns in active labor, however. If back pain in active labor is associated with the posterior position we wouldn’t be able to tell from this study where 92% of women had epidurals. All we can know from this study is that back pain was not reported in higher numbers by women with OP babies at about 4 cm. Back ache due to posterior position often does not appear until 4 cm or after, in my observations. And furthermore, that back pain is due more to pelvic shape and ligament tension than fetal position. (Carol Phillips, DC sites the posterior and sacrocervical ligament tension as a cause of back pain.) This means that a good number of women with posterior babies do not report back labor at all.

Data is held back, perhaps inadvertently, hopefully for a later report, that could address, in part, questions about the nature of rotation. The first problem is the lack of differentiation between the right occiput transverse with the left occiput transverse. The second problem, which is a reflection of the first, is that rotation is assumed to follow a uniform pattern. Not all OP babies rotate, and not all rotating OPs rotate the same, and even that a small number of OA babies rotate to the posterior. This we knew from previous studies and by watching laboring women. But because a single pattern isn’t perceived the mistake can be made that there is no pattern at all. Remember that before researchers described the six levels of consciousness in a newborn that newborn behavior was seen as random and meaningless. Now we see a pattern of wakefulness that helps us interpret newborn needs. In time, a pattern of rotation will be recognized that will help us attend to the needs of birthing mothers.

There were many uninterpretible ultrasounds that had to be trashed in the data collection. No more than 51% of the study mothers had a readable ultrasound at all three times, enrollment; 4-hour/epidural, and >8 cm. If these labors were tracked individually to see where the baby was at each ultrasound we could learn whether indeed the fetus rotated aimlessly down the pelvic highway or, as I believe, with a purpose. To track such a pattern we will have to include the left and right occiput transverse delineations.

Neither does this study look at flexion (chin-tucking) or asynclitism. These are also influences in fetal rotation. The role of epidural may have an effect on both of these, especially asynclitism if alternations of the tone in the lower uterine segment are effected by epidural analgesia. Hopefully, these topics will be furthered studied also.

This Lieberman study is, as are her others, so intelligent overall and holds so much information that we who hunger for knowledge feel its impact on the greater birthing community. It packs a whallop. The rippling effects will certainly change practice in the maternity field. I hope the mistake is not made that preventing and correcting occiput posterior fetal position is no longer seen as beneficial.

On the positive side, the point may be made that we can’t predict labor outcome by OP position in and of itself. Approximately 80% of posterior babies will rotate at any given stage of labor.

Sadly, this report allows a candid glimpse of a prestigious institution dunked by rising cesarean rates and the loss of obstetrical skills for normal birth. The researchers admit the rising intervention rate for all births in their institution since 1998. This rise may act like as a bizarre form of permission in the odd way the medical field has of using defeat to promote the technological model of childbirth.

Still unanswered are the active childbirth movement’s questions about the role of maternal positioning and movement over time in labor.

Would we be able to help more OPs be born without surgery? Would we be able to address some cases of transverse arrest that weren’t specifically singled out but surely existed in this study. And could we assist, without trauma, the 32% of persistent posteriors?

In other words, research is lacking, since active birthing techniques are seldom studied, and when they are they may not be used in the ways active birth supporters, like trained doulas and midwives might actually use them. Research has not invalidated common labor progress techniques suggested in Penny Simkin and Ruth Ancheta’s Labor Progress Handbook, Jean Sutton and Pauline Scott’s Optimal Fetal Positioning or my own Spinning Babies Workshops or Web Site.

The reason I have so much to say is that Ellice Lieberman and her colleagues said so much. My comments are not criticism, but discussion reflecting a wish for even more detailed research. You owe it to yourself and to the mothers you help to read this article in full. It is available for free downloading from


Keep your copy close by because you will refer back to it from time to time in your studies on fetal positioning.  

Flag Counter

Pageviews since January 1, 2014
Spinning Babies - on the web since 2002


Spinning Babies on Sense and Sensibility


Hi Gail, 

I want to say thank you! A first time mother's posterior birth went well [with] a short labor. Membranes ruptured about 8am. Her labor started about 6pm: 3cm dilated; LOP station -1; cervix soft and effaced. 

We did rebozo sifting, sidelying release, abdominal lift and tuck, standing sacral release, Dip the hip, circling on the birth ball, lunge and she birthed at 9.58pm in sidelying position with hands around the husband's neck.

Baby's head showed more molding on the right side. What you taught in the workshop and website made a difference!

Chiew Gin, Doula, Singapore