The frequency of fetuses in occiput posterior position during labor is approximately 20%. Most of these fetuses turn spontaneously to an occiput anterior position. Nonetheless, approximately 5% remain occiput posterior at delivery.1,2 Several studies show that delivery in a posterior position involves significantly longer labor and higher risks of operative vaginal deliveries, cesarean deliveries, and severe perineal lacerations.3–8
Different techniques have been proposed to promote rotation from a posterior to an anterior position.9 The use of rotational forceps probably declined because of the associated risks of fetal and maternal complications and a reduction of its learning.10,11 Some obstetricians perform rotations with a vacuum extractor; these can be effective but have not been sufficiently evaluated.12 In some hospitals, mothers are placed in specific positions in the delivery room, although data demonstrating the interest of these are sparse.13,14 Others choose expectant management either alone or associated with oxytocin. Finally, several studies have shown that manual rotation is a simple technique that can turn a fetus from a posterior to an anterior position.15–18 One possibility for assessing the interest of manual rotation is to compare outcomes in maternity units with different strategies defined in protocols for the management of posterior positions.
Our hypothesis was that when an occiput posterior or occiput transverse position persists at full dilatation, a strategy of manual rotation would not only increase the number of anterior positions at delivery, but would also reduce the frequency of operative deliveries. The objective of our study was to compare mode of delivery of women with fetuses in posterior position at full dilatation in two level III university hospitals, one with a policy favoring modifications of maternal position and the other with a policy favoring manual rotation using data from the TRACOR (TRAction of the CORd) prospective multicenter randomized trial.19
MATERIAL AND METHODS
This prospective comparative study was conducted in two level III university hospitals in France. Manual rotation is not a recommended practice at maternity unit 1, which delivers 3,000 neonates annually. When a fetus is in occiput posterior or occiput transverse position at full dilatation, the woman's position is adjusted to promote rotation of the fetal head. The focus is on the woman's movement rather than a particular position. Accordingly, different positions are used including asymmetric lateral decubitus, sitting, and on hands and knees. On the other hand, in Hospital 2, which delivered 4,700 neonates a year during the study period, manual rotation is recommended at full dilatation for fetuses in a occiput posterior or occiput transverse position.15,16 Maternal movement (ie, a change in postures) can be used as a complement to manual rotation, especially when rotation fails. Recruitment of patients in these two level III university hospitals was comparable and included women at high risk, women transferred for maternal or fetal disease or disorder, and low-risk women living near the hospital.
The manual rotation technique used in Hospital 2 is that described by Tarnier and Chantreuil.15,16 The patient must be placed in the lithotomy position with the legs in stirrups. Her bladder must be empty. The fetal head is rotated 180° if it is in an occipitosacral position, 135° for posterior positions, and 90° for transverse positions. The hand used by the operator varies according to the position: the right hand holds onto the back of the fetus' right ear for the left positions (left occiput transverse and left occiput posterior) and the left hand is behind the left ear for the right positions (right occiput transverse and right occiput posterior). The operator thus places his or her right hand behind the fetal ear and, during a contraction and a pushing effort, moves the fetal head in a rotating movement forward toward the pubic symphysis to move the occiput relative to the anterior arc of the mother's pelvis. The operator can be a midwife, a senior resident, or a senior obstetrician. Manual rotation is performed either at the diagnosis of full dilatation or after a waiting period of 1 or 2 hours.
The TRACOR trial, designed to assess the effect of controlled cord traction during the third stage of labor in preventing postpartum hemorrhage, included women in labor with a singleton pregnancy, a live fetus in cephalic presentation, delivering at 35 weeks of gestation or more (N=4,355).19 The TRACOR trial was a multicenter randomized controlled trial with two parallel groups and took place in five French university hospitals between January 1, 2010, and January 31, 2011. The trial was approved by the Paris Ile de France III Committee for the Protection of research Subjects (ethics committee) in September 2009. Our study examines the data from two of the five participating centers (N=2,296). In both, the analysis concerns the women who reached full dilatation (n=2,164). The 30 women for whom position at full dilatation was unknown were excluded from the study.
Several secondary objectives were planned for the TRACOR trial and these were taken into consideration in developing the questionnaire; a wide variety of data about labor was therefore collected prospectively. The value of manual rotation for turning posterior positions was one of these secondary objectives. The clinician (midwife or physician) prospectively recorded the fetal position at full dilatation (occiput anterior, occiput transverse, or occiput posterior) determined clinically by digital cervical examination and in cases of clinical doubt by ultrasonography. For this study, we combined occiput posterior and occiput transverse. The performance of manual rotation was also recorded as well as its success or failure. Maternal postures at full dilatation (dorsal or lateral decubitus, seated or semiseated, and on hands and knees) were also collected and classified in two categories: absence of posture change if she remained in dorsal decubitus through the second stage of labor or position change regardless of the position used. Finally, the fetal position at delivery was also recorded at the time of delivery by the delivering clinician.
The principal end point was operative delivery, that is, an instrumental vaginal delivery or a cesarean delivery. The secondary outcomes measured were position at delivery (occiput anterior or occiput posterior and occiput transverse), episiotomy, severe perineal tears (third and fourth degree), the duration of the second stage of labor (period between full dilatation and delivery), and neonatal condition (5-minute Apgar score, arterial pH at birth). These outcome measures occurred and were recorded before the intervention studied in the TRACOR trial (ie, controlled cord traction at the moment of the afterbirth). In this trial, the two arms did not differ for the outcome measures we studied here.
The data about mothers' characteristics (age, ethnicity, body mass index [calculated as weight (kg)/[height (m)]2], parity, uterine scar), obstetric characteristics (induction, epidural, oxytocin administration during labor, duration of labor at 5–10 cm), and neonatal characteristics (gestational age and birth weight) were also recorded.
We calculated the sample size to achieve a power of 90% (α=0.05, hospital ratio 1:2) with the following hypothesis: a policy of manual rotation reduces by two the operative delivery risk (ie, 40% in Hospital 1 and 20% in Hospital 2). Under this assumption, the calculated sample size was 90 women in Hospital 1 and 179 in Hospital 2.
The principal analysis concerned the women with fetuses in posterior or transverse positions at full dilatation (n=111 in Hospital 1 and n=220 in Hospital 2). Nonetheless, to study the comparability of obstetric practices in the two units, analyses were also conducted for the total population of women with fetuses in anterior position at full dilatation included in the TRACOR trial in both hospitals (n=589 in Hospital 1 and n=1,214 in Hospital 2).
We compared the maternal, neonatal, and obstetric characteristics of the women with fetuses in posterior or transverse positions at full dilatation in both hospitals. As part of this comparative study, the principal end point and the secondary criteria were then compared between the two centers in univariable analyses with the χ2 test, Fisher’s exact test (when n=less than 5), and Student’s t test. Then, to study the association between the hospital policy (Hospital 1: no manual rotation; Hospital 2: manual rotation) and operative delivery, we conducted a multivariable analysis with a logistic regression model. Finally, to assess the specific effects of manual rotation, we used univariable and multivariable analysis to examine whether manual rotation was a practice independently associated with a reduction in the risk of operative delivery. We also performed the same analyses among women who delivered vaginally with instrumental delivery as the outcome variable. The logistic regression models included all factors significantly associated with the risk of operative delivery in the univariable analysis (P<.1). The statistical analyses were performed with Stata 12.0 software.
Figure 1A and B describe the different fetal positions at full dilatation in the population in each center. The percentage of women with fetuses in occiput posterior or occiput transverse position at full dilatation was similar: n=111 (15.9%) in Hospital 1 and n=220 (15.3%) in Hospital 2 (P=.75). Table 1 presents the maternal, obstetric, and neonatal characteristics of the study population (women with fetuses in occiput posterior and occiput transverse positions). In Hospital 1, the women were significantly younger, more often French, and had a higher body mass index than those in Hospital 2 (Table 1). Induction of labor and epidural analgesia were both less frequent in Hospital 1 (Table 2). The same differences were found in the total population of women with fetuses in the occiput anterior position at full dilatation (results not reported here).
As expected, the strategies for fetuses in occiput posterior or occiput transverse positions at full dilatation differed significantly between the two maternity units. In Hospital 1, manual rotation was attempted in 0.9% of cases compared with 78.9% in Hospital 2 (P<.01), whereas changes in maternal posture were used, respectively, in 57.6% compared with 21.8% (P<.01). In both units, the maternal postures used were mainly lateral decubitus (Table 2). In Hospital 2, in 172 cases in which rotation was attempted, 155 (90.1%) were effective and successfully rotated the head to an occiput anterior position.
In both maternity units, most of the women with fetuses in occiput posterior or occiput transverse positions at full dilatation delivered the fetus in the occiput anterior position. The rate of posterior position at birth, regardless of mode of delivery, did not differ significantly between the hospitals (13.5% in Hospital 1 compared with 8.2% in Hospital 2, P=.13). The operative delivery rate was higher in Hospital 1 than in Hospital 2 (38.7% compared with 23.2%, P<.01). The cesarean delivery rates at full dilatation in this population were similar (9.9% compared with 8.2%, P=.68), but the rate of instrumental deliveries was significantly higher in Hospital 1 (28.8% compared with 15.0%, P<.01) (Table 3). The rate of episiotomy was lower in Hospital 1; however, this difference was not significant (26.0% compared with 36.1%, P=.07). Neonatal outcomes were similar in both units. In Hospital 1, 1.8% of neonates had 5-minute Apgar scores less than 7 and 5.5% had an arterial pH less than 7.10 at birth compared with 2.3% and 4.8%, respectively, in Hospital 2 (Table 3). Mode of delivery of the women with fetuses in the occiput anterior position at full dilatation was similar in both hospitals (data not reported here).
After adjustment for potential confounding factors in the multivariable analysis, the risk of operative delivery was lower in the hospital with a policy of manual rotation (Hospital 2) (adjusted odds ratio [OR] 0.52, 95% confidence interval [CI] 0.28–0.95) (Table 4). Considering only vaginal deliveries, the risk of instrumental delivery was lower in Hospital 2 (adjusted OR 0.38, 95% CI 0.19–0.78).
Finally, we studied the independent effect of manual rotation regardless of hospital. In the multivariable analysis, the risk of operative delivery was significantly lower when manual rotation was attempted (adjusted OR 0.45, 95% CI 0.25–0.85). On the other hand, changing maternal posture at full dilatation was not associated with a significant modification of the risk of operative delivery (adjusted OR 0.72, 95% CI 0.38–1.36) (Table 5). Considering only vaginal deliveries, the risk of instrumental delivery was lower in women with attempted manual rotation (adjusted OR 0.41, 95% CI 0.21–0.81).
This study shows that a strategy of manual rotation for fetuses in posterior or transverse positions at full dilatation is associated with a reduction in the rate of operative delivery.
There is no published randomized trial comparing the effect of manual rotation and expectant management on mode of delivery in women with fetuses in posterior or transverse position. Only a single study in California tried to answer this question from a retrospective cohort and found results that strongly supported use of manual rotation.17 However, the expectant group in that study was comprised only of women who delivered in a posterior position and not all the women with a posterior position at full dilatation regardless of position at delivery. Women in the expectant group who had a spontaneous rotation (and therefore an anterior delivery) were therefore not considered in the analyses. Because delivery in the posterior position is known to be at higher risk of maternal complications, this selection bias probably increased the risk of cesarean delivery or operative vaginal delivery and of maternal complications in the expectant group. Our two-site study, on the contrary, resembles an intention-to-treat analysis because Hospital 1 corresponds to a control group (manual rotation not recommended) and Hospital 2 to an intervention group (manual rotation recommended); every woman in a posterior or transverse position at full dilatation was included, whether or not she had the management recommended in the hospital she was in, and women who had a spontaneous rotation were included in the analyses at both centers, in particular at center 1 (the control group).
As expected, we observed that most of the posterior positions turned to anterior positions regardless of the strategy used. This may explain the absence of any significant difference between the two centers for fetal position at delivery. This absence of difference may also be the result of a lack of power, in view of the limited numbers observed, because there were half as many deliveries in the posterior position in Hospital 2 as in Hospital 1.
Although the rates of deliveries in the posterior position did not differ significantly between the two hospitals, our results show a significantly lower rate of operative vaginal deliveries in Hospital 2, which performed manual rotations. One hypothesis is that manual rotation facilitates the process of the second stage of labor. That is, when manual rotation is systematically performed at full dilatation, engagement and descent of the fetal head occur in anterior position. In expectant management, however, rotation, when it takes place spontaneously, happens later, often after engagement, even at the perineum. Descent of the head is more laborious in the occiput posterior than in the occiput anterior position. It may be accompanied by a longer duration of the second stage of labor and by more frequent fetal heart rate abnormalities, because the fetus rotates at the very end of labor, at the maternal perineum most often. Obstetricians may well be more likely to perform an instrumental intervention once fetal heart rate abnormalities, even only moderate, occur if the second stage has been long. However, manual rotation can also be associated with fetal heart rate abnormalities, but those were not assessed in our study.
Operative vaginal deliveries have been reported to be associated with a higher risk of severe perineal tears. In our study, the rate of instrumental interventions in Hospital 1 was higher but the rate of severe perineal tears did not differ significantly between the groups. Further interest in reducing the risk of instrumental interventions by using manual rotation lies in the simultaneous reduction of the risk of fetal injuries associated with operative deliveries, although this is low.
A randomized trial testing the effect of manual rotation would require training all the midwives and obstetricians in the participating teams to perform manual rotation before setting up the trial and convincing the teams already trained in it and convinced of its value to not perform manual rotation for the patients included in the expectant group. Moreover, a consistent homogeneous alternative practice would be required in the expectant group. In the absence of this type of trial, “quasi-interventional” observational studies of the type we performed are useful for comparative effectiveness research and provide interesting results. The essential limitation of these studies is the possibility that the groups are not comparable in terms of the women's characteristics and obstetric practices other than those specifically studied. Nonetheless, to study the association between manual rotation and operative delivery, we performed a multivariable analysis so that we could adjust for the potential confounding factors.
We also took patients' changes of posture into account. That is, in Hospital 1, we cannot really talk about “expectant management” but rather of a “maternal postures” strategy, one that remains undefined because the postures used varied. After adjustment, we found no significant reduction in the risk of operative delivery in patients with changes in maternal posture. This absence of effect may be the result of a lack of power. However, if a real effect exists, it would result in an underestimation of the effect of manual rotation. It is thus possible that if the women in Hospital 1 had not changed positions, the effect of manual rotation on the rates of occiput posterior position at delivery and of operative delivery would have been still greater. Nonetheless, there are few studies specifically on the effect of maternal posture and they do not show that these postures have a significant effect in turning fetuses.13,14
Our results could aid midwives and obstetricians in caring for women with fetuses in posterior or transverse positions. In view of the available data, it certainly seems reasonable to attempt a manual rotation at full dilatation to turn posterior or transverse positions into anterior positions. Obstetricians and midwives should therefore be trained to perform this maneuver.
1. Gardberg M, Laakkonen E, Salevaara M. Intrapartum sonography and persistent occiput posterior position: a study of 408 deliveries. Obstet Gynecol 1998;91:746–9.
2. Lieberman E, Davidson K, Lee-Parritz A, Shearer E. Changes in fetal position during labor and their association with epidural analgesia. Obstet Gynecol 2005;105:974–82.
3. Senecal J, Xiong X, Fraser WD. Effect of fetal position on second-stage duration and labor outcome. Obstet Gynecol 2005;105:763–72.
4. Ponkey SE, Cohen AP, Heffner LJ, Lieberman E. Persistent fetal occiput posterior position: obstetric outcomes. Obstet Gynecol 2003;101:915–20.
5. Fitzpatrick M, McQuillan K, O'Herlihy C. Influence of persistent occiput posterior position on delivery outcome. Obstet Gynecol 2001;98:1027–31.
6. Cheng YW, Shaffer BL, Caughey AB. The association between persistent occiput posterior position and neonatal outcomes. Obstet Gynecol 2006;107:837–44.
7. Cheng YW, Shaffer BL, Caughey AB. Associated factors and outcomes of persistent occiput posterior position: a retrospective cohort study from 1976 to 2001. J Matern Fetal Neonatal Med 2006;19:563–8.
8. Carseldine W, Phipps H, Zawada S, Campbell N, Ludlow J, Krishnan S, et al.. Does occiput posterior position in the second stage of labour increase the operative delivery rate? Aust N Z J Obstet Gynaecol 2013:53;265–70.
9. Simkin P. The fetal occiput posterior position: state of the science and a new perspective. Birth 2010;37:61–71.
10. Magnin P, Audra P. Is it legitimate today to perform large rotations with forceps? Revue Française de Gynécologie et d'Obstétrique 1984;79:255–61.
11. Park JS, Robinson JN, Norwitz ER. Rotational forceps: should these procedures be abandoned? Semin Perinatol 2003;27:112–20.
12. McQuivey RW. Vacuum-assisted delivery: a review. J Matern Fetal Neonatal Med 2004;16:171–80.
13. Desbriere R, Blanc J, Le Dû R, Renner JP, Carcopino X, Loundou A, et al.. Is maternal posturing labor efficient in preventing persistent occiput posterior position? A randomized controlled trial. Am J Obstet Gynecol 2013;208:60.e1–8.
14. Stremler R, Hodnett E, Petryshen P, Stevens B, Weston J, Willan AR. Randomized controlled trial of hands-and-knees positioning for occipitoposterior position in labor. Birth 2005;32:243–51.
15. Haddad B, Abirached F, Calvez G, Cabrol D. Manual rotation of vertex presentations in posterior occipital-iliac or transverse position. Technique and value. J Gynecol Obstet Biol Reprod (Paris) 1995;24:181–8.
16. Le Ray C, Serres P, Schmitz T, Cabrol D, Goffinet F. Manual rotation in occiput posterior or transverse positions: risk factors and consequences on the cesarean delivery rate. Obstet Gynecol 2007;110:873–9.
17. Shaffer BL, Cheng YW, Vargas JE, Caughey AB. Manual rotation to reduce caesarean delivery in persistent occiput posterior or transverse position. J Matern Fetal Neonatal Med 2011;24:65–72.
18. Reichman O, Gdansky E, Latinsky B, Labi S, Samueloff A. Digital rotation from occipito-posterior to occipito-anterior decreases the need for cesarean section. Eur J Obstet Gynecol Reprod Biol 2008;136:25–8.
19. Deneux-Tharaux C, Sentilhes L, Maillard F, Closset E, Vardon D, Lepercq J, et al.. Effect of routine controlled cord traction as part of the active management of the third stage of labour on postpartum haemorrhage: multicentre randomised controlled trial (TRACOR). BMJ 2013;346:f1541.