The optimum management of the second stage of labor in women with epidural anesthesia with respect to immediate pushing at complete dilatation versus a period of rest and descend (ie, no pushing, while allowing for spontaneous descent) before the onset of pushing is controversial.1–4 Traditionally, women are asked to begin pushing as soon as the cervix is completely dilated. However, some authors suggest fetal and maternal advantages to waiting before commencing to push in the second stage of labor. These authors propose that a longer second stage, if managed correctly, is not detrimental to the fetus.3,5–9 Other authors' findings indicate possible disadvantages associated with maternal pushing efforts.10
The present study was designed as a randomized, controlled clinical trial to estimate the effects of rest and descend versus pushing in the second stage of labor with epidural anesthesia.
MATERIALS AND METHODS
Patients were enrolled in this study at the time of admission to labor and delivery with the diagnosis of spontaneous or induced labor at term. Care was provided at LDS Hospital, a regional tertiary care teaching center.
Exclusion criteria included refusal of epidural, first epidural dose after complete dilatation, known fetal anomaly, multiple gestation, nonvertex presentation, gestational age less than 37 weeks or over 42 weeks, and pregnancy complicated by pregnancy-induced hypertension, heart disease, or insulin-dependent diabetes. The study was reviewed and approved for protection of human subjects by the Institutional Review Board at the LDS Hospital.
A power analysis was performed to determine the appropriate sample size in relation to the various outcome variables. The computation was based on an α level of .05% and 80% power. The number of subjects needed in each group were: for duration of pushing - 66, for fetal decelerations - 12, and for perineal lacerations and episiotomy - 64. Assumed prevalence rates for this calculation were based upon published values.3 For other variables, umbilical cord pH value spontaneous delivery, maternal fatigue, infant Apgar score, and post-partum endometritis, the required sample size was either too great for practical inclusion, or insufficient prevalence data for similar populations were available.
A sample size of 66 in each group (primigravid experimental group, primigravid control group, multigravid experimental group, multigravid control group) was planned, for a total sample size of 264 women. Patients undergoing vaginal birth after cesarean without prior vaginal delivery were considered to be functionally primigravid.
Initial enrollment consisted of 312 women, and 252 women (81%) completed the study protocol. Eligible women were randomly assigned to either the control group 1 (pushing) or the experimental group 2 (rest and descend before pushing). A computer-generated randomized list was used to assign patients.
A computerized bedside fetal monitoring system was used to record the fetal heart rate (FHR) tracing, as well as all demographic, admission, labor, delivery, and recovery room information. Interventions and care were entered by the registered nurse according to standard labor and delivery protocols, except for frequency of vaginal examination, which was dictated by the study protocol. All patients had continuous electronic fetal and uterine contraction monitoring.
A software program was written that prompted the nurse to enter dependent and confounding variable information into the computer at appropriate times. The software program also prompted the nurse to have the patient document the maternal fatigue indices at the appropriate times. The time of delivery, method, fetal position, cervical dilatation, fetal station, episiotomies, lacerations, instrument deliveries, umbilical arterial cord blood pH, and Apgar scores were recorded.
A single, experienced labor and delivery nurse reviewed each FHR tracing in both the control and experimental groups. The reviewer determined number and type of fetal decelerations during the second stage of labor.
The nurse or physician determined by vaginal examination the beginning of the second stage of labor. Vaginal examinations were conducted frequently, according to the judgment of the nurse, in the late active phase of labor, to determine as closely as possible when the patient was first completely dilated. The nurse was oriented to the importance of determining the onset of the second stage as accurately as possible, as this was the starting point of the study protocol.
Women in the experimental group began a period of rest and descend at the time of complete dilatation and continued until the head was seen at the introitus, or after 120 minutes in primigravidas or 60 minutes in multigravidas. These women were encouraged not to push. Visualizing the head at the introitus was determined by slightly separating the vulva with two fingers during a contraction, with the woman's legs apart. The introitus was examined in this manner every 30 minutes, or sooner if signs of imminent delivery occurred. In the experimental group, the nurse determined station and position at the onset of pushing after the period of rest and descend. Both the primigravidas and multigravidas in the control group were encouraged to begin pushing as soon as they were found to be completely dilated.
In the control group, and in the experimental group after the allowed lapse of time, pushing consisted of coached Valsalva maneuvers directed by the nurse or physician during contractions. Once pushing began, it occurred with each uterine contraction until delivery. Two quick breaths were encouraged between pushes. Patients pushed in lateral tilt or decubitus position.
The nurse recorded the uterine contraction interval every 30 minutes during the second stage of labor for both the control and experimental groups. The uterine contraction pattern was evaluated by averaging to the nearest minute the frequency of uterine contractions over the preceding 30 minutes. In all patients, uterine contractions were clinically assessed by the nurse — the additional use of intrauterine pressure catheters was left to the discretion of the attending physician.
Continuous lumbar epidural anesthesia was given at patient request in the first stage of labor. In over 95% of cases, bupivicaine was the agent. No subdural component was used. The patient's bladder status was assessed by palpation above the symphysis pubis every 30 minutes in the second stage of labor. If indicated, the bladder was emptied.
Fatigue has been conceptualized as a multidimensional concept involving subjective, cognitive, and physical dimensions. In this study, two instruments were used to measure the subjective component of fatigue. These tools were the Modified Fatigue Symptom Checklist and three serially performed fatigue visual analogue scales.11
Baseline maternal fatigue data (Modified Fatigue Symptom Checklist and the first visual analogue scale) were collected as soon as the epidural was in place and the patient was comfortable. The second visual analogue scale was obtained within 15 minutes of complete dilatation of the cervix. The final visual analogue scale was obtained within 60 minutes after the delivery of the placenta.
Statistical analyses consisted of two sample independent Student t test (parametric) or Mann-Whitney U (nonparametric), one sample paired Student t test (parametric) or Wilcoxon signed rank test (nonparametric), and χ2 and Fisher exact test, both one tail and two tail. All data were analyzed for normal distribution using the Kolmogorov-Smirnov test.
Of the 312 women initially randomized, 60 were removed. The reasons for withdrawal were physician request without medical indication (16 patients), physician request for medical complication (eight patients), patient request (one patient), cesarean before the second stage of labor (12 patients), cesarean for abnormal FHR pattern in the second stage of labor (three patients), nursing protocol error (nine patients), and unknown (11 patients).
Our population consisted of 96% white, non-Hispanic women. The women in the multigravid control group (mean 30.2 ± 4.6) were slightly older than the women in the multigravid experimental group (28.2 ± 4.31) (P = .014). No other statistically significant demographic maternal or fetal differences were found between the control and experimental groups.
Outcome and potentially confounding variables for experimental and control groups are detailed in Table 1 and Table 2.
The pelvic station in centimeters above or below the ischial spine was divided by the length of time in minutes, for both the period of rest and descend and for the period of pushing. This gave the mean descent in centimeters per minute. There was no significant difference in the rate of descent between the experimental groups and the control groups from onset of pushing to the delivery of the infant. As expected, the overall rate of descent of the experimental group was significantly slower during their rest and descend phase than when they were pushing (P < .001).
Total pushing time was less for the experimental, compared with the control groups for both primigravid (58 ± 44 minutes compared with 76 ± 41 minutes) (P = .21) and multigravid women (13 ±15 minutes compared with 24 ± 23 minutes) (P < .001).
Fatigue scores are detailed in Table 1. Regardless of group, primigravid patients had greater fatigue scores at all measured time intervals than parous women. Primigravid patients in the rest and descend group had significantly less overall fatigue (visual analogue scale at second-stage onset, visual analogue scale after the third stage of labor) than those in the pushing group (P = .017). No difference was found in calculated fatigue score in parous women.
The number and type of FHR decelerations were divided by the length of the second stage in minutes, reflecting all FHR decelerations for each woman during her second stage of labor. The hypothesis that the total number of fetal decelerations in the second stage of labor would be less with a period of rest before pushing was supported (Table 1). This was found in the experimental group in both the primigravid (P < .001) and multigravid (P = .004) groups when all degrees of variable decelerations were combined. This hypothesis was also supported in both the primigravid and multigravid groups for mild variable decelerations (P = .006 and .028, respectively), for moderate variable decelerations and severe decelerations in primigravidas (P = .056 and .011), and for severe variable decelerations in multigravidas. There were no late decelerations on the monitor strips in this study (Table 1).
There was no statistical difference found in Apgar scores or umbilical arterial cord pH values between groups (Table 1). No difference was found in the incidence of postpartum endometritis.
No significant differences in type of episiotomies or lacerations were found between experimental and control groups. The overall episiotomy rate was 64% in primigravidas and 40% in multigravidas.
There were slightly less instrumental deliveries in the experimental multigravid group, but this was not statistically significant (P = .078). No differences in instrumental delivery were found in the primagravid group (Table 2). Overall, there was a 22% forceps rate and a 4% vacuum extraction rate in primigravidas, and a 5.8% forceps delivery rate and 1.7% vacuum extraction rate in multigravidas. Although 12 patients underwent cesarean delivery before the second stage of labor (and were thus removed from analysis), only three patients who reached the study phase of labor underwent cesarean delivery. All cesarean deliveries in the second stage of labor occurred in nulliparous women in the control group (ie, active pushing). Two of these cesarean deliveries were performed for arrest of descent and one for a nonreassuring FHR tracing. The overall cesarean rate for women initially randomized was 15 of 312 (4.8%).
Potentially confounding variables in the second stage of labor were: bladder status, position changes, uterine contraction pattern, and frequency and duration of pushing once pushing began. We found an increase in the number of position changes in the control group in primigravidas (P = .005), an increase in the average uterine contraction rate in the experimental group in primigravidas (P = .035), and an increase in pushing frequency once pushing began in the experimental group in both primigravidas and multigravidas (P = .001 and .048, respectively).
Our data suggest that although the second stage of labor may be prolonged (up to 4.9 hours) in women allowed to rest and descend, such lengthening of labor is not associated with any discernable adverse maternal or fetal outcomes in women with functioning epidural anesthesia. Among women undergoing a protocol of rest and descend compared with the more traditional second-stage labor management, there was no difference with respect to Apgar scores, umbilical arterial cord pH, perineal injuries, operative delivery, or postpartum endometritis. Potentially beneficial effects of such a policy were documented, including a significant decrease in maternal fatigue index in primigravid patients, fewer decelerations, and decreased pushing time during the second stage in both primigravid and parous patients.
Our inability to document benefits of rest and descend on fatigue scores in multiparous patients is probably related to the overall shorter length of second-stage labors seen in this group. The beneficial effects of this policy on fatigue, as measured in the early postpartum period in primigravid patients were observed despite the natural euphoria of many such women after having given birth to their first child; thus, the actual beneficial effects may be even greater than we were able to assess.
Aldrich et al suggested a reduction in fetal brain oxygenation associated with maternal pushing efforts.10 In our study, women randomized to rest and descend also had higher Apgar scores and umbilical cord pH values than those managed in the traditional manner. However, this difference did not reach statistical significance. This observation may be a reflection of the significant reduction in second-stage decelerations seen in the rest and descend group. Thus, although definitive conclusions cannot be drawn regarding beneficial fetal effects of a policy of rest and descend, it appears that prolongation of the second stage in the carefully controlled manner described in this study is not detrimental to fetal well being.
A statistical benefit was demonstrated with respect to decelerations, decreased pushing time, and fatigue during labor. It must be noted that power calculations do not allow the authors to exclude potential small, undetected differences (β error) in complications such as endometritis, operative delivery, or reduced fetal Apgar score or cord pH in these groups. To adequately address these issues, a larger sample size is necessary. In addition, because our power analyses were based upon the study group as a whole, subgroup analysis of primiparous and multiparous patients fell outside the bounds of the power calculation. Thus, any negative findings, although valid in the study population as a whole, must be taken with caution when applied separately to these parity-based subgroups. Because this investigational protocol required significant alteration in the standard practices of a large number of private practitioners, such numbers were not deemed practical in this study. However, given a complete absence of data to suggest that such complications are, in fact, increased in patients carefully managed according to such rest and descend protocols, our results suggest that a policy of rest and descend remains one appropriate option in properly selected women.
The ACOG recommends that the second stage of labor in patients with epidural anesthesia be limited to 2 hours in parous women, and 3 hours in primigravid patients.12 Our data suggest that in appropriately selected and managed women, extension of these time periods is safe, and may be beneficial for mother and fetus using a policy of rest and descend. However, it is critical to note that a policy of rest and descend in the second stage involves much more than simply ignoring an obstructed labor. All patients underwent careful electronic FHR monitoring, pelvic examinations were limited, and periods of rest were limited by study protocols. In addition, all patients had well-functioning, continuous lumbar epidural anesthesia in place, and were carefully attended to by trained and supportive nursing staff. Where such conditions are not met, it would be inappropriate to depart from traditional recommendations regarding the length of the second stage.12 However, in patients too fatigued to push effectively, or those with multiple variable decelerations, a policy of rest and descend appears to be an appropriate option.