Objective
We sought to evaluate the efficacy of maternal posturing during labor on the prevention of persistent occiput posterior (OP) position.
Study Design
We conducted a randomized trial including 220 patients in labor with a single fetus in documented OP position. Main outcome was the proportion of anterior rotation from OP position.
Results
The rates of anterior rotation were, respectively, 78.2% and 76.4% in the intervention group and the control group without significant difference ( P = .748). Rates of instrumental and cesarean section deliveries were not significantly different between intervention and control groups (18.2% vs 19.1%, P = .89, and 19.1% vs 17.3%, P = .73, respectively). In intervention and control groups, persistent OP position rates were significantly higher among women who had cesarean section (71.4% and 89.5%, respectively) and an instrumental delivery (25% and 33.3%, respectively) than among women who achieved spontaneous vaginal birth (5.8% and 2.8%, respectively). In multivariable analysis, body mass index and parity were found to have significant and independent impact on the probability of fetal head rotation.
Conclusion
Our study failed to demonstrate any maternal or neonatal benefit to a policy of maternal posturing for the management of OP position during labor.
With an estimated prevalence ranging from 2-13% at delivery, persistent occiput posterior (OP) position is the most common malposition in labor and is associated with maternal and neonatal morbidities. Thus it exposes not only to prolonged first and second stages of labor, use of epidural analgesia, postpartum hemorrhage, cesarean and operative vaginal delivery, third- and fourth-degree perineal lacerations, and chorioamnionitis but also to low Apgar score, neonatal trauma, acidemic cord blood gas concentrations, admission to neonatal intensive care unit, and newborn encephalopathy. Several studies based on ultrasound examination during labor have demonstrated that the vast majority of OP deliveries results from the failure of rotation from this position. Thus, the rate of OP position has been estimated to be 30-40% at the early stage of active labor, 20-30% at 10-cm dilatation, and 5-10% at delivery. Various methods have been considered to rotate an OP fetus: use of oxytocin, operative delivery, manual rotation, and maternal posturing before or during labor. But none of them had any proven efficacy in reducing persistent OP position. Maternal posturing is somewhat attractive because it is noninvasive and theoretically harmless for the fetus. In a recent Cochrane review, only a single study about the efficacy of maternal posturing during labor was of sufficient quality to be included. This trial, using a single posture during labor (hands-and-knees), failed to demonstrate either any reduction of persistent OP at birth or improvement of maternal or neonatal outcomes. Two Chinese trials (published in Chinese) have found an increased rate of anterior rotation of OP fetuses using lateral recumbent position during labor. These last 2 studies were excluded from the Cochrane review ; no explanation was given about the reason for such exclusion. However, with only these 3 trials, the presumed benefit of maternal posturing during labor remains understudied and therefore widely debated. Nevertheless, there is in France a growing interest in and use for maternal positioning during labor that is supported more on empirical knowledge than on evidence-based medicine. The aim of this study was to test the efficacy of maternal posture options during labor on prevention of persistent OP position in order to reduce their maternal and neonatal morbidities.
Materials and Methods
This randomized controlled trial took place in a tertiary care center maternity unit (Hôpital Nord, Marseille, France) from Jan. 5, 2009, through Jan. 25, 2011. Ethical approval was obtained from the Committee for the Protection of Persons “Sud Méditérranée I.” Written informed consent was obtained from each participant before randomization. Eligible patients were adult pregnant women (age ≥18 years) in labor at ≥36 weeks of gestation, with cervical dilatation of ≥3 cm and ruptured membranes, and with a single fetus in cephalic OP position confirmed by ultrasound examination. Gestational age calculation was based on the participant’s reported last menstrual period and fetal biometry at first-trimester systematic ultrasound scan. After inclusion, participants were randomly assigned to the intervention or the control group. The randomization was made in permutated blocks of 4.
Intervention
While randomized patients in the control group adopted dorsal recumbent position during labor, those in the intervention group had to adopt postures depending on the station of the fetal head. In both groups, if medically indicated (severe fetal heart rate abnormalities or maternal hypotension), patients were transiently placed in lateral recumbent position for a short delay (<10 minutes) and immediately returned to their designated posture. Because pressures applied to the fetal head change as it progresses into the pelvis, we postulated that maternal position should be adjusted to the different stages of labor. Three different maternal postures were chosen after 3-dimensional computer simulation and analysis of theoretical rotation mechanism by one of the authors (J.-P.R.) using 3-dimensional simulation software (Carrara 7 Pro; DAZ 3D Editor, Inc. Draper, Lake City, UT) ( Figure 1 ). Posture 1 was used in patients with fetal head station between -5 and -3 ( Figure 2 , A) and corresponded to hands-and-knees posture with possible support on a balloon at shoulder level. The expected goal of this posture was to move the fetus away from the maternal spine to facilitate its global rotation, assuming that the physical confrontation of fetal spine with maternal spine convexity and psoas muscle in maternal recumbent position restricts its flexion ( Figure 1 , A). Patients with fetal head station between -2 and 0 were positioned in posture 2 ( Figure 2 , B). Thus, patients were positioned in strict lateral recumbent position on the same side of the fetal spine (eg, left maternal lateral position for left fetal spine) with folded inferior leg and upper leg positioned in the axis of the body. In this position, the maternal pelvic inlet was oblique with the highest iliac crest located on the lower side (ie, the side of the patient that is lying on the bed). The aim of this posture was to induce the flexion of the fetal head by confronting the fetal occiput to the maternal sacro-iliac joint and delay contact of the forehead with the contralateral pubis allowing for complete flexion and thus provoke anterior rotation ( Figure 1 , B). Finally, posture 3 was used in patients with fetal head station >0 ( Figure 2 , C). It consisted of a lateral recumbent position on the same side of the fetal spine, with the inferior leg lying in the axis of the body and the other leg folded at an approximately 90-degree angle with the use of a leg support. In this position, the maternal pelvic inlet was orientated in the exact opposite direction than in posture 2. The aim was to confront the fetal forehead with the levator ani and to therefore induce the flexion of the fetal head, supposing that the rotation of the fetal head would lead to the anterior rotation of fetal shoulders and trunk ( Figure 1 , C).
A simplified and illustrated protocol with a flow chart and detailed descriptions was available in the labor ward to help caregivers. Regardless of the timing of inclusion, the posture was maintained as long as the fetal occiput remained posterior and the patient tolerated it. The maternal posture was subsequently adapted to the evolution of the fetal head station. Because ultrasonography has been demonstrated to be the gold standard for the diagnosis of fetal presentation, an ultrasonographic control of the fetal presentation was performed every hour throughout the labor until anterior rotation of the fetal occiput was achieved. Patients who had achieved anterior rotation were then asked to adopt dorsal recumbent position.
Outcome measures
We considered that the primary outcomes had to be: (1) fetal head position at delivery for spontaneous deliveries; and (2) fetal head position immediately before instrumental delivery or cesarean section for operative deliveries. For all patients who achieved vaginal spontaneous delivery, the midwife or the obstetrician in charge of the delivery systematically recorded the exact type of fetal head presentation observed at the time of vaginal delivery. In cases of operative delivery, the fetal head position was recorded by means of ultrasonography immediately before the instruments (either spatulas or vacuum) were applied. If a cesarean delivery was performed, the fetal head presentation was confirmed by ultrasonography immediately before cesarean section. As secondary outcomes, length of labor, length of pushing, operative deliveries, cesarean sections, perineal lacerations (Anglo-Saxon classification), and neonatal morbidity (5- and 10-minute Apgar scores, umbilical arterial pH) were systematically recorded. Maternal characteristics were collected to verify the comparability of the 2 groups: age, parity, gestational age, height, and initial and final weight. Finally, factors that could have had an impact on head presentation were also recorded for analysis: analgesia, position of placenta, spontaneous or induced labor, and oxytocin use.
Sample size
The sample size calculation was based on a previous nonrandomized pilot study performed in our maternity ward (data not published) and relied on the assumption that 60% of the control group would rotate from OP to occiput anterior vs 85% in the intervention group. To achieve a 90% powered study with set at 0.001, we calculated that 202 patients were required to be included and randomized. To compensate for any secondary exclusion, we arbitrarily increased this number by 10%; inclusion and randomization of 220 patients was therefore required.
Statistical analysis
Statistical analysis was performed using software (PASW Statistics, version 17.0.2; IBM Corp, Armonk, NY). Data are expressed as mean ± SEM or median with interquartile interval. Groups were compared using the χ 2 or Fisher exact test for categorical characteristics, and using the Student t test or Mann-Whitney U test for continuous ones, as appropriate. A multivariate analysis was performed using a logistic regression model to estimate risk factors for persistent OP position. Variables that were found to be significantly associated with outcome or to be marginally significant ( P < .10) in univariate analysis, or that had clinical relevance, were included in the logistic regression model. Calibration was assessed using the Hosmer-Lemeshow goodness-of–fit test to evaluate the discrepancy between observed and expected values. A 2-sided P value < .05 was considered statistically significant. Main outcome measure was analyzed according to the intention-to-treat principle.
Results
Patients
Over 2 years, 220 women were prospectively included and randomized, of whom 110 were assigned to the intervention group and 110 to the control group. Figure 3 shows the flow of participants throughout the trial. No patient was secondarily excluded from the study. Two patients in the control group underwent successful manual rotation immediately before instrumental delivery and–therefore considered as having persistent OP position–they were not excluded from the study because we chose an intention-to-treat protocol. No deviation of protocol occurred in the intervention group. Table 1 summarizes the baseline characteristics of the 2 groups. Except for cervical dilatation at the time of inclusion, the groups were comparable for gestational age, initial and final body mass index (BMI), parity, placental insertion, use of induction of labor and of epidural analgesia, and neonatal weight. In the intervention group, the majority of patients (86/110; 78.2%) were included when the fetal head station was between -2 and 0 cm, and then were recommended to adopt posture 2. Postures 1 and 3 were used in 13.6% (15/110) and 8.2% (9/110), respectively. No patient had to adopt all 3 positions, as all women who adopted posture 1 had their fetuses rotate from OP to occipitoanterior before they had to adopt posture 3. No statistically significant differences were observed between the intervention and control groups regarding the mode of delivery ( Table 2 ). Operative delivery was performed in 41 cases: 26 spatulas and 15 vacuums. Indications for operative delivery were inadequate expulsive efforts in 22 cases, nonreassuring fetal heart tracing in 18 cases, and maternal contraindications to expulsive efforts in 1 case. Forty patients underwent a cesarean section. Indications for cesarean section were: failure to progress before complete dilatation in 24 cases (21 associated with persistent OP), failure to progress at full dilatation in 9 cases (6 associated with persistent OP), nonreassuring fetal heart tracing in 6 cases, and brow presentation in 1 case. Indications for operative delivery or cesarean section were not statistically different between the 2 groups.
| Characteristic | Intervention group (n = 110) | Control group (n = 110) | P value |
|---|---|---|---|
| Maternal age, y | 27.8 ± 0.64 | 28.9 ± 0.51 | .180 |
| BMI | |||
| Initial | 23.7 ± 0.47 | 24.1 ± 0.49 | .482 |
| At time of delivery | 28.6 ± 0.57 | 28.2 ± 0.75 | .680 |
| Parity | |||
| Nullipara | 69 (62.7) | 62 (56.4) | .579 |
| 1 | 23 (20.9) | 29 (26.4) | |
| ≥2 | 18 (16.4) | 19 (17.3) | |
| Gestational age, wk | 39.4 ± 0.10 | 39.4 ± 0.10 | .950 |
| Placental insertion | |||
| Anterior | 49 (44.5) | 61 (55.4) | .367 |
| Posterior | 47 (42.8) | 39 (35.5) | |
| Others | 14 (12.7) | 10 (9.1) | |
| Labor | |||
| Spontaneous | 82 (74.5) | 81 (73.6) | .878 |
| Induced | 28 (25.5) | 29 (26.4) | |
| Median cervical dilatation at time of inclusion, cm | 6.26 | 5.68 | .046 |
| Epidural analgesia | 103 (93.6) | 105 (95.5) | .553 |
| Birthweight, g | 3466.2 ± 41.29 | 3403.13 ± 44.56 | .300 |
| Variable | Intervention group (n = 110) | Control group (n = 110) | P value |
|---|---|---|---|
| Spontaneous deliveries | 69 (62.7) | 70 (63.6) | .889 |
| OA | 65 (94.2) | 68 (97.1) | .441 |
| OP | 4 (5.8) | 2 (2.8) | |
| Operative vaginal deliveries | 20 (18.2) | 21 (19.1) | .965 |
| Fetal head position before operative delivery | |||
| OA | 15 (75) | 14 (66.7) | .558 |
| OP | 5 (25) | 7 (33.3) | |
| Fetal head position at birth | |||
| OA | 18 (90) | 18 (85.7) | 1 |
| OP | 2 (10) | 3 (14.3) | |
| CS deliveries | 21 (19.1) | 19 (17.3) | .727 |
| Fetal head position before CS | |||
| OA | 6 (28.6) | 2 (10.5) | .241 |
| OP | 15 (71.4) | 17 (89.5) |
Main outcome
In the intervention group, 24 (21.8%) patients were diagnosed with persistent OP position compared to 26 (23.6%) in the control group ( P = .75) ( Table 3 ). When considering each mode of delivery, no statistical difference between the 2 groups regarding fetal head position at birth or immediately before operative delivery or cesarean section was observed ( Table 2 ). In the control group, compared to spontaneous deliveries, persistent OP position was significantly more often observed immediately before operative delivery: 2 (2.8%) vs 7 (33.3%), respectively ( P < .001), and was also more often observed immediately before cesarean section than immediately before operative delivery: 17 (89.5%) vs 7 (33.3%), respectively ( P = .003). The same differences were observed in the intervention group: 4 (5.8%) vs 5 (25%), respectively ( P = .004), and 15 (71.4%) vs 5 (25%), respectively ( P < .03).
