Today, nearly one in three pregnant women undergoes cesarean delivery in the United States.1 With vaginal birth after cesarean delivery rates nationally at less than 10%,1 delivery by cesarean nearly assures that all subsequent births will likewise be by this route. Minimizing primary cesarean delivery is a priority.2 Two common indications of cesarean during labor are active phase arrest and arrest of descent.3,4 Current labor norms were established by Dr. Emmanuel Friedman in the 1950s5,6 with the original intent of presenting average length of labor.5,6
Recently, Zhang et al suggested that the progression of labor in modern obstetrics may deviate from that established by Friedman.7,8 Changes in obstetric practice over time included a much higher prevalence of epidural analgesia and oxytocin augmentation during labor and a substantially lower rate of forceps and vacuum-assisted vaginal deliveries.9 The American College of Obstetricians and Gynecologists (the College) Practice Bulletin on labor dystocia states the mean durations of second stage of labor in nulliparous and multiparous women are 54 and 19 minutes, respectively.10,11 It defined prolonged second stage of labor as more than 2 hours without epidural or 3 hours with epidural in nulliparous women, and 1 hour without or 2 hours with epidural for multiparous women.10 The basis of these parameters is unclear, and the additional hour allotted for labor with epidural anesthesia appears to be based on the mean effect of epidural.10–12
The median length of labor is frequently reported in describing normal labor progression and 95th percentile thresholds are commonly utilized to define the extreme ends of a distribution. Thus, our research aim was to describe the median and the 95th percentile lengths of second stage of labor based on parity and epidural use. We hypothesized that epidural anesthesia would have been associated with an increase of more than 1 hour at the 95th percentile thresholds during the second stage of labor.
MATERIALS AND METHODS
We designed a retrospective cohort study of all cephalic, live, singleton births to women who delivered at the University of California, San Francisco between 1976 and 2008. The Committee on Human Research at University of California, San Francisco approved the study. The exclusion criteria were no labor, multifetal gestations, noncephalic presentation, placenta previa, intrauterine fetal demise, or known lethal congenital anomalies. We also excluded women who delivered during the first stage of labor by cesarean and therefore would have had a second stage of labor length of 0 minutes. Additionally, we excluded women with missing or incomplete information about the duration of second stage of labor. There were 53,682 singleton births that were delivered during the study period and 48,474 (90.3%) had labor. Of these, 1,225 women (2.5%) had missing information about length of second stage of labor. All deliveries at this academic institution were performed by the attending physicians, clinical nurse midwives, or resident physicians with supervision of either an attending physician or a clinical nurse midwife. The number of deliveries remained stable during the study period, ranging between 1,400 and 1,800 deliveries per year. Of the women who met study inclusion or exclusion criteria who delivered by cesarean, approximately 60% of those cesarean deliveries were performed for arrest of descent, another 12% were performed for fetal intolerance of labor, and 11% were performed for failed forceps or vacuum-assisted vaginal delivery.
The maternal characteristics and labor information were prospectively collected by the managing physicians of the labor and delivery unit. Neonatal information was similarly collected by neonatologists and pediatricians using a neonatal database. The maternal and neonatal databases were then linked using two unique identifiers and cross-checked for complete linkage. Additionally, trained abstractors performed daily chart review and data abstraction to ensure information accuracy and to minimize missing data. The database also underwent monthly review by trained physicians for quality assurance. The database was created in 1976, and detailed perinatal and labor information continued to be prospectively collected and maintained during the entire study period and is ongoing. The length of second stage of labor was defined as time interval between complete cervical dilation to delivery of the fetus.
The primary outcomes were the median and the 95th percentile lengths of second stage of labor. We chose to examine the median and the 95th percentile thresholds because the length of second stage of labor in the study population was not of normal distribution. Length of labor in women with labor epidural was compared with length of labor in women without epidural, stratified by parity. Statistical analysis was performed using Kruskal-Wallis and χ2 tests. Survival analysis was also used to examine the length of second stage of labor. We also performed subgroup analysis. We stratified by mode of delivery, induction and augmentation of labor, and time period of delivery (1976–1989, 1990–1999, and 2000–2008). Additionally, we examined the length of second stage of labor based on neonatal outcomes. We created a composite variable, “neonatal morbidity,” which included those neonates with Apgar score less than 7 at 5 minutes, umbilical artery cord gas pH level less than 7.0, neonatal sepsis, meconium aspiration syndrome, admissions to the neonatal intensive care unit, and birth trauma (cephalohematoma, head laceration, clavicular fracture, skull fracture, facial nerve palsy, and Erb [or brachial plexus] palsy).
The secondary outcomes included maternal and neonatal outcomes. These included mode of delivery, peripartum infection, postpartum hemorrhage, and severe perineal lacerations for maternal outcomes. Neonatal outcomes examined were 5-minute Apgar score less than 7, umbilical cord artery pH level less than 7.0, meconium aspiration syndrome, neonatal sepsis, intensive care unit admission, and birth trauma. The association between length of labor and perinatal outcomes was examined using χ2 test and multivariable logistic regression analysis. A full logistic regression model was built, and subsequent stepwise backward elimination with P<.20 as the cut-off was performed to derive a parsimonious regression model. The only covariate that had a P>.20 was maternal insurance status. However, we opted to keep insurance status in the final regression model because it was associated with epidural use and maternal and neonatal outcomes. Thus, covariates included in the multivariable logistic regressions were maternal age older than 35 years, race or ethnicity, marital status, public insurance, gestational age at delivery, epidural use, induction of labor, and delivery year. Of note, we had intended to include oxytocin augmentation of labor as a covariate in the logistic regression model; however, it was omitted in the regression analyses because of collinearity with induction of labor. We also included mode of delivery and birth weight in the regression models for neonatal outcomes. Statistical analysis was performed using STATA 11.0. Statistical significance was indicated using P<.05 and 95% confidence intervals.
There were 42,268 women who met study criteria with normal neonatal outcomes. Among them, 49.9% (n=21,090) had epidural during labor and 50.1% (n=21,178) did not. In our cohort, epidural use varied by parity, maternal age, and race or ethnicity (P<.001 for all; Table 1). Women who had induction of labor or oxytocin augmentation were also more likely to have epidural analgesia in labor (P<.001 for both; Table 1). The proportion of women who had epidural anesthesia during labor increased during the study period (P<.001; Table 1).
For nulliparous women, the median length of second stage of labor was 47 minutes without epidural and 120 minutes in the presence of epidural use, a prolongation of an additional 73 minutes (P<.001). When we examined the length of labor in nulliparous women with and without epidural at the 95th percentile threshold, it was 197 minutes with epidural and 336 minutes without epidural. The difference of the 95th percentile thresholds between nulliparous women with and without epidural was 139 minutes, or 2 hours and 19 minutes (P<.001; Table 2). This 95th percentile threshold difference in duration of second stage of labor was larger than that of the median values. This relationship between second stage of labor duration and epidural was further explored using Kaplan-Meier survival analysis in which the 95th percentile differences in labor duration of women without and with epidural were longer than that of median differences (Fig. 1). For multiparous women, the median lengths were 14 minutes without epidural and 38 minutes with epidural, a prolongation of 24 minutes (P<.001; Table 2). However, at the 95th percentile thresholds, they were 81 minutes without and 225 minutes with epidural, a difference of 2 hours and 54 minutes.
We further performed subgroup analysis. There were 35,681 women (subcohort) who had vaginal delivery without neonatal morbidity. Among them, nulliparous women without epidural had a median and 95th percentile second stage of labor length of 47 minutes and 190 minutes, respectively. For nulliparous women with epidural, the median and 95th percentile second stage of labor lengths were 112 minutes and 302 minutes, respectively (Table 2). Although the difference in median length between nulliparous women with and without epidural was 65 minutes, the difference in 95th percentile length was 112 minutes. Similarly, for multiparous women who delivered vaginally with normal neonatal outcomes, the median and 95th percentile second stage of labor length were 14 minutes and 79 minutes without epidural, and 38 minutes and 217 minutes with epidural (P<.001; Table 2). The difference in the 95th percentile length between multiparous women with and without epidural was 138 minutes. Similar trends of longer median and 95th percentile thresholds with epidural compared with without epidural were seen for women who had spontaneous vaginal deliver (P<.001; Table 2). Among women who had operative vaginal delivery (n=6,851) by either forceps or vacuum-assisted vaginal delivery, the median and 95th percentile lengths of second stage of labor were longer in those with epidural than those without (P<.001; Table 2).
We also examined length of second stage of labor among women who had induction of labor (n=6,004) and those who did not (n=36,150), and among women who had oxytocin augmentation of labor (n=11,402) and those without augmentation of labor (n=25,270). Again, the median and 95th percentile lengths of second stage of labor were longer with epidural than without for both nulliparous and multiparous women with and without induction (P<.001; Table 3). Although the length of second stage of labor was statistically significantly different by parity and epidural status among women who were induced compared with those who were not induced (P<.001 for all), these thresholds did not appear to be clinically significantly different (Table 3). Similar trends were augmentation of labor (Table 3). However, we did not have information regarding indication of oxytocin augmentation in this dataset to further explore this association.
We explored the relationship between epidural anesthesia and length of second stage of labor stratified by parity and year of delivery. The median length of second stage of labor remained relatively similar across each of the study periods (1976–1988, 1989–1999, 2000–2008; Table 4). Among multiparous women without and with epidural, similar trends were also seen for the median length of second stage of labor. Although statistical significance (P<.001) was reached for these comparisons, the median lengths likely were of little clinical significance (Table 4). In contrast, the 95th percentile thresholds of second stage of labor were progressively longer across the time periods. For nulliparous women without and with epidural, the 95th percentile thresholds were 163 minutes and 270 minutes, respectively, during delivery years 1976 to 1989. The 95th percentile thresholds were 199 minutes and 325 minutes for nulliparous women without and with epidural, respectively, during 1990 to 1999. They were 347 minutes and 432 minutes, respectively, during 2000 to 2008 (P<.001; Table 4). A similar increase in 95th percentile thresholds of second stage of labor was seen for multiparous women without and with epidural and across study periods (P<.001; Table 4).
We then explored the definition for abnormal or prolonged second stage of labor. In this study cohort, using the current definition of prolonged second stage of labor defined by the College,10 16.2% of nulliparous women without epidural and 31.1% of nulliparous women with epidural in labor would be considered as having prolonged second stage of labor; the proportion for multiparous women was 7.7% without and 18.9% with epidural (Table 5). Because the 95th percentile thresholds of second stage of labor for this cohort was longer than the current definition, we examined the proportion of women who would have prolonged second stage of labor diagnosed if 1 additional hour were allotted to current thresholds (ie, 3 hours without epidural or 4 hours with epidural in nulliparous women, and 2 hours without or 3 hours with epidural for multiparous women). Using this criteria, 6.6% of nulliparous women without epidural and 16.2% with epidural would have prolonged second stage of labor diagnosed; similarly, 3.0% of multiparous without and 10.4% with epidural would have prolonged second stage of labor diagnosed (Table 5). Another strategy to define prolonged second stage of labor is to use the 95th percentile thresholds. In doing so, approximately 5% of the population would have prolonged second stage diagnosed, with the thresholds of more than 197 minutes for nulliparous women without epidural, 336 minutes for nulliparous women with epidural, 81 minutes for multiparous women without epidural, and 255 minutes with epidural (Table 5).
We examined whether prolonged second stage of labor would be associated with undesirable neonatal outcomes by using definitions currently established by the College,10 College thresholds plus 1 additional hour, and the 95th percentile thresholds according to the study cohort. Based on the College’s definition,10 18.9% of women would have prolonged second stage of labor diagnosed. Compared with women without prolonged second stage of labor, there were no statistically significant differences in the incidence or odds of neonatal outcomes except for birth trauma (adjusted odds ratio 1.58; 95% confidence interval 1.13–2.22; Table 6). When we defined prolonged second stage of labor using the current College definition plus 1 additional hour, 9.3% (n=3,923) of women would be considered to have a prolonged second stage of labor (Table 6). Using this threshold, there were no differences in neonatal outcome except for birth trauma (adjusted odds ratio, 2.08; 95% confidence interval, 1.38–3.15) among prolonged second stage of labor compared with those without (Table 6). When we defined prolonged second stage of labor using the 95th percentile thresholds based on the study cohort (Tables 1 and 5), 4.7% (n=1,580) would be considered to have a prolonged second stage of labor. Using this definition, again there were no differences in neonatal outcome except for birth trauma (adjusted odds ratio, 2.73; 95% confidence interval, 1.62–4.61) among women with prolonged second stage of labor compared with their counterparts (Table 6). We also examined maternal outcomes associated with prolonged second stage of labor using these three proposed criteria (Table7). Compared with women without prolonged second stage of labor, women whose second stage of labor durations were considered prolonged had eight-times to nine-times the odds of cesarean delivery and two-times to three-times the odds of operative vaginal delivery (Table7). The odds of third-degree or fourth-degree perineal lacerations, postpartum hemorrhage, chorioamnionitis, and endomyometritis were similarly higher in women with a prolonged second stage of labor than those without (Table 3). This was the case for all three proposed criteria for prolonged second stage of labor.
We observed that the length of second stage of labor in women with epidural was longer than for women without epidural analgesia. Although the majority of obstetrician–gynecologists subscribe to the clinical guidelines of giving 1 additional hour to account for epidural use, it appears that the differences from epidural at the 95th percentiles may be approximately double.10 Thus, the current definition of prolonged second stage of labor may be too stringent.
When we examined the length of second stage of labor in women who achieved vaginal delivery without adverse neonatal outcomes, those with an epidural continued to have a second stage of labor duration of more than 1 hour longer compared with women without an epidural. Additionally, the 95th percentile thresholds in women with an epidural who had spontaneous vaginal delivery without neonatal morbidity in this study were similar to those reported by Zhang et al13 in a multicenter study (the Consortium on Safe Labor). Despite differences in study design and population, it seemed reassuring that similar second stage of labor characteristics were independently observed for women who achieved vaginal delivery.
Using the current definitions of prolonged second stage of labor defined by the College,10 approximately 31% of nulliparous and 19% of multiparous women with epidural anesthesia in the study cohort would be identified as having a prolonged second stage of labor. Although the passage of time is not an indication for operative intervention,10 women receiving an epidural during labor remained at higher risk for operative vaginal delivery.14 Although most interventions likely were clinically indicated, it remains possible that some were performed primarily for the diagnosis of abnormal second stage, particularly when 20% to 30% of laboring women who received an epidural would meet such definitions. Thus, we advocate that the definition of prolonged second stage of labor should be reexamined.
Interestingly, in a recent joint workshop on how to reduce the first cesarean delivery by the Eunice Kennedy Shriver National Institute of Child Health and Development, the Society for Maternal-Fetal Medicine, and the College, the group proposed thresholds of 3 hours without epidural and 4 hours with epidural during labor for nulliparous women and 2 hours without and 3 hours with epidural for multiparous women.2 Our data also support longer thresholds for nulliparous women without epidural. Concordant with recent data,7,14,15 we also observed that the length of labor became progressive longer over time. Although the precise reasons for this remained unclear, changing obstetric characteristics, such as higher proportion of induced or augmented labor, fewer forceps and high-station operative deliveries, as well as increased prevalence of obesity and gestational weight gain might play a role.1,7,9 During the study period, immediate pushing during the second stage of labor was usually the practice model and delayed pushing was not widely instituted. Thus, the probability of delayed pushing as a contributor to increased length of second stage of labor over time, although possible, was likely low. Interestingly, when we examined whether birth weight increased during the study period, it remained relatively the same. Thus, we could not consider birth weight as a potential factor for progressively increasing length of second stage of labor during the study period.
We were able to explore length of second stage of labor in relation to epidural use and associated perinatal outcomes in various clinical scenarios. However, there were limitations. First, missing data or inaccurate information could potentially bias our findings. It was reassuring that only 2.5% of potential study participants had missing information regarding length of second stage of labor. Second, the study period was long. This, however, enabled the examination of labor pattern over time. Additionally, we analyzed data from one academic institution, which could potentially limit the generalizability of study findings to the broader population. Although labor management at the study hospital may vary from that of other institutions, our study findings were similar to those reported by a large, multicentered study of labor pattern.13
In summary, the use of epidural anesthesia during labor lengthened the median and the 95th percentile thresholds of second stage of labor. The 95th percentile thresholds of second stage of labor were more than 2 hours longer in women with epidural compared with those without epidural during labor for both nulliparous and multiparous women. To put these findings in context, using current prolonged labor definitions, one would label nearly one-third of nulliparous women with epidural as having abnormal labor. This likely leads to potentially unnecessary interventions. Although labor norms should not be established based on this study alone, our findings, along with those of others, suggest that current definitions of prolonged second stage of labor in the setting of an otherwise reassuring fetal status may be insufficient. There exists a need to establish proper second stage of labor norms to reflect modern obstetrics.
1. Hamilton BE, Martin JA, Ventura SJ. Division of Vital Statistics. Births: preliminary data for 2011. Natl Vital Stat Rep 2012;61:1–29.
2. Spong CY, Berghella V, Wenstrom KD, Mercer BM, Saade GR. Preventing the first cesarean delivery. Summary of a joint Eunice Kennedy Shriver
National Institute of Child Health and Human Development, Society for Maternal-Fetal Medicine, and American College of Obstetricians and Gynecologists Workshop. Obstet Gynecol 2012;120:1181–94.
3. Zhang J, Troendle J, Reddy UM, Laughon SK, Branch DW, Burkman R, et al.. Contemporary cesarean delivery practice in the United States. Am J Obstet Gynecol 2010;203:326.e1–10.
4. Barber EL, Lundsberg LS, Belanger K, Pettker CM, Funai EF, Illuzzi JL. Indications contributing to the increasing cesarean delivery rate. Obstet Gynecol 2011;118:29–38.
5. Friedman EA. Primigravid labor; a graphicostatistical analysis. Obstet Gynecol 1955;6:567–89.
6. Friedman EA. Labor in multiparas; a graphicostatistical analysis. Obstet Gynecol 1956;8:691–703.
7. Zhang J, Troendle JF, Yancey MK. Reassessing the labor curve in nulliparous women. Am J Obstet Gynecol 2002;187:824–8.
8. Zhang J, Troendle J, Mikolajczyk R, Sundaram R, Beaver J, Faser W. The natural history of the normal first stage of labor. Obstet Gynecol 2010;115:705–10.
9. Laughon SK, Branch DW, Beaver J, Zhang J. Changes in labor patterns over 50 years. Am J Obstet Gynecol 2012;206:419.e1–9.
10. Kilpatrick SJ, Laros RK Jr. Characteristics of normal labor. Obstet Gynecol 1989;74:85–7.
11. Albers LL. The duration of labor in healthy women. J Perinatol 1999;19:1114–9.
12. Albers LL, Schiff M, Gorwoda JG. The length of active labor in normal pregnancies. Obstet Gynecol 1996;87:355–9.
13. Zhang J, Landy HJ, Branch DW, Burkman R, Haberman S, Gregory KD, et al.. Contemporary patterns of spontaneous labor with normal neonatal outcomes. Obstet Gynecol 2010;116:1281–7.
14. Anin-Somuah M, Smyth RM, Jones L. Epidural versus non-epidural or no analgesia in labour. The Cochrane Database of Systematic Reviews 2011, Issue 12. Art. No.: CD000331. DOI: 10.1002/14651858.CD000331.pub3.
15. Zhang J, Yancey MK, Klebnoff MA, Schwarz J, Schweitzer D. Does epidural analgesia prolong labor and increase risk of cesarean delivery? A natural experiment. Am J Obstet Gynecol 2001;185:128–34.