Nulliparous active labor, epidural analgesia, and cesarean delivery for dystocia

Obstetrics Nulliparous active labor, epidural analgesia, and cesarean delivery for dystocia James A. Bofill, M D , a Robert D. Vincent, MD, b Elaine L. Ross, MD, c Rick W. Martin, MD, c Patricia F. Norman, MD, a Carol F. Werhan, MD, a and John C. Morrison, MD ~

Dayton, Ohio, Birmingham, Alabama, and Jackson, Mississippi OBJECTIVE: Our purpose was to examine the effect of epidural analgesia on dystocia-related cesarean delivery in actively laboring nulliparous women. STUDY DESIGN: Active labor was confirmed in nulliparous women by uterine contractions, cervical dilatation of 4 cm, effacement of 80%, and fetopelvic engagement. Patients were randomized to one of two groups: epidural analgesia or narcotics. A strict protocol for labor management was in place. Patients recorded the level of pain at randomization and at hourly intervals on a visual analog scale. Elective outlet operative vaginal delivery was permitted. RESULTS: One hundred women were randomized. No difference in the rate of cesarean delivery for dystocia was noted between the groups (epidural 8%, narcotic 6%; p = 0.71). No significant differences were noted in the lengths of the first (p = 0.54) or second (p = 0.55) stages of labor or in any other time variable. Women with epidural analgesia underwent operative vaginal delivery more frequently (p = 0,004). Pain scores were equivalent at randomization, but large differences existed at each hour t~qereafter. The number of patients randomized did not achieve prestudy estimates. A planned interim analysis of the results demonstrated that we were unlikely to find a statistically significant difference in cesarean delivery rates in a trial of reasonable duration. CONCLUSIONS: With strict criteria for the diagnosis of labor and with use of a rigid protocol for labor management, there was no increase in dystocia-related cesarean delivery with epidural analgesia, (Am J Obstet Gynecol 1997;177:1465-70.)

Key words: Epidural analgesia, cesarean delivery, dystocia

There is controversy regarding the effect of epidural analgesia o n the rate of cesarean birth for the indication of dystocia. R a n d o m i z e d trials 1' 2 have a d d e d credence to the impression of retrospective studies by T h o r p et a l ) ' 4 that there is an increased incidence of cesarean delivery for dystocia in w o m e n receiving epidural analgesia. A metaanalysis of studies c o m p a r i n g "epidural" with "no epiduraI" groups reached the conclusion that a 10% increase in the rate of cesarean delivery for the indication of dystocia is to be expected when epidural analgesia is provided. 5 O n the other hand, the consensus a m o n g others appears to be that epidural analgesia has littIe, if From the Department of Obstetrics and Gynecology,, Wrigtzt State University School of Medicine,,~ the Department of Anesthesiology, University of Alabama at Birmingham,b and the Departmenls of Obstetrics and GynecologyCand Anesthesiolog%~ University of Missis~ sippi Medical Center. Supported by the VicksburgHospital Medical Foundation. Received for publication January 16, 1997; revised May 6, 1997; acceptedffuly 1, 1997. Reprints not availablefrom the authors. Copyright © 1997 by Mosby-Year Book, Inc. 0002~9378/97 $5.00 + 0 6/1/84466

any, impact on dystocia-related cesarean birth rates. 6 Indeed, an investigation by Chestnut et al. 7 n o t e d n o difference in the rates of cesarean delivery, for the indication of dystocia regardless of whether epidural analgesia was initiated early or late in labor. Similarly, Gribble a n d Meier, 8 in a retrospective study, reported no significant increase in the rate of cesarean delivery o n their obstetric service in spite of widespread availability of epidural analgesia compared with a time when such c o n d u c t i o n analgesia was n o t available. Finally, Philipsen a n d J e n s e n 9 were u n a b l e to d o c u m e n t a significant increase in the rate of cesarean delivery for those w o m e n who were r a n d o m i z e d to receive epidural analgesia compared xdth w o m e n who were r a n d o m i z e d to receive parenteral pethidine. Retrospective studies, although they may use innovative statistics, c a n n o t address the causal relationship of c o n d u c t i o n analgesia a n d cesarean delivery for dystocia.S, 4, 8, 10 O t h e r investigations have n o t controlled crucial variables that could be related to the issue of interest. Such variables include the confirmation of active labor 1465

1466

Boflllet al.

before epidural placement, strict protocols for active labor management, consistent diagnosis of dystocia, and guidelines for the depth of epidural analgesia so that motor blockade is minimized. Differences in study design have added to the confusion regarding the effect of epidural analgesia on nulliparous active labor. Some studies have also been criticized because they did not provide a control group that did not receive epidural analgesia.7. ,1.12 Our own unpublished institutional experience was that epidural analgesia, when administered to women in active labor, appeared to have a negligible impact on cesarean delivery rates. The purpose of this study was to prospectively examine the effect of epidural analgesia on the rate of cesarean delivery for the indication of dystocia in actively laboring nulliparous women.

Material and methods Healthy nulliparous women at 36 to 42 weeks' gestation and in spontaneous active labor were candidates for this study. We chose to study only nulliparous patients because, as a group, they are more prone to dystocia and are more likely to require cesarean delivery- than are women who have had a previous vaginal birth. Parturients with serious medical problems (e.g., insulin-dependent diabetes mellitus, chronic hypertension requiring medication, pregnancy-induced hypertension, etc.) or those with twin gestations were excluded from this trial. Women who had received cervical ripening or oxytocin inductions were not allowed entry. If all inclusion criteria were met, candidates were offered participation when they requested medication for pain. M1 subjects in this trial signed an informed consent document that was approved by the Institutional Review Board of the University of Mississippi Medical Center. Randomization was accomplished by selection of the next in a series of opaque envelopes indicating epidural or narcotic analgesia. The randomization envelopes were prepared by an uninvolved third party by use of a computer-generated list of random numbers. Strict guidelines regarding labor management were in force throughout the duration of the study. Active labor was defined as regular uterine contractions with cervical dilatation of >--4 cm (but 200 Montevideo units per 10 minutes for 2 to 3 hours). A second opinion, obtained from a perinatologist blinded to the patient's analgesia group assignment, was sought for any patient being considered for a dystocia-related cesarean delivery. Delivery by elective outlet forceps or vacuum was permitted for the patients in this trial. Women receiving narcotic analgesia were treated with 1 to 2 mg intravenous doses of butorphanol every 1 to 2 hours, as needed. Promethazine 25 mg was given intravenously every 3 to 6 hours as required for nausea. The Institutional Review Board stipulated that epidural "rescue" be allowed on this study. The women who were randomized to receive butorphanol were allowed epidural analgesia 4 hours after randomization if the pain was not adequately controlled by parenteral medication in the j u d g m e n t of the patient. Subjects randomized to the epidural arm of the study received regional analgesia according to the guidelines established by the Department of Anesthesiology. With the patient sitting or in the left lateral position, the skin over the L2-L3, L3-L4, or L4-L5 interspace was scrubbed with povidone-iodine solution and infiltrated with 1% lidocaine. After intravenous hydration with 500 to 1000 ml of lactated Ringer's solution, the epidural space was identified with use of a loss-of-resistance technique with an 18-gauge Hustead needle. After 5 to 10 ml of normal saline solution was injected through the needle, a 20-gauge multiorifice catheter was advanced 3 to 5 cm into the epidural space and secured. A test dose of 3 ml of 1.5% lidocaine with 1:200,000 epinephrine was then administered at the discretion of the anesthesiologist. Three to five ml boluses of 0.25% bupivacaine, with or without 50 to 100 I*g of fentanyl, were injected through the epidural catheter until a T10 sensory level of analgesia was obtained. A continuous infusion was then initiated with 0..125% bupivacaine with 1.5 p~g/ml fentanyl titrated to maintain this level of analgesia. The epidural infusions were continued throughout the first and second stages of labor. If necessary, 5 to 10 ml of 3% 2-chlorprocaine was administered to provide perineal analgesia for delivery. A standard visual analog pain scoring system was used during this study. A simple 10-cm nongraduated line was presented to the patient with !'no pain" at one end of the line and "worst pain imaginable" at the other end. The patient was asked to make a mark on this line to represent her level of pain after the nurse asked the question: "Can you show me how much pain you experienced with your last contraction?" Pain scores were obtained at randomization and at hourly intervals. At the conclusion of the study these pain scores were measured from the "no pain" end to the point where the patient

Bofii/et al.

Volume 177, Number 6 Am J Obstet Gynecol

indicated h e r self-perceived level of pain and a n u m e r i c value was assigned to the nearest 0.1 cm. Data collection was a c c o m p l i s h e d prospectively. Fetaln e o n a t a l assessment i n c l u d e d A p g a r scores at 1 and 5 minutes after birth as assigned by the obstetric or pediatric nurses. C o r d arterial b l o o d gas specimens were collected and analyzed. N a l o x o n e usage, if r e q u i r e d for p r e s u m e d opioid-induced neonatal respiratory depression, was recorded. T i m e was an i m p o r t a n t variable in this study. Because all the w o m e n were in labor at the time of admission, we considered the first stage of labor to be f r o m the time of admission until the time of c o m p l e t e cervical dilatation and the second stage as the time f r o m c o m p l e t e cervical dilatation until delivery. R a n d o m i z a t i o n time was considered to e x t e n d f r o m admission until r a n d o m i z a t i o n was accomplished. T h e length of time between r a n d o m i z a t i o n and c o m p l e t e cervical dilatation and f r o m r a n d o m i z a t i o n to delivery was calculated. T h e length of time the m e m b r a n e s were r u p t u r e d was m e a s u r e d by subtracting the time of amn i o t o m y or of spontaneous r u p t u r e f r o m the time of delivery. T h e highest dose of oxytocin the patient received was r e c o r d e d , as was the length of time the patient received this a g e n t (initiation of oxytocin a u g m e n t a t i o n to time of delivery). A sample size analysis was p e r f o r m e d before the initiation o f the study, but this was problematic. T h e expected rates of dystocia-related cesarean delivery were 15% and 5% in the epidural and narcotic groups, respectively. With significance at the p < 0.05 level and a power of 80%, we calculated that 159 patients were r e q u i r e d in each group to detect a 10% difference in the rate of cesarean section for dystocia. An interim analysis of the results was p l a n n e d at 75% of e x p e c t e d r a n d o m izations. After the p r o t o c o l had b e e n a p p r o v e d and initiated, it was considered that the cesarean delivery rates used for sample size analysis should have reflected the results of the only published prospective study with a similar design. 1 With use o f the dystocia-related cesarean delivery rates of 16% in the epidural g r o u p and 2% in the narcotics g r o u p ) a sample size of 130 patients (65 patients in each of two groups) would be r e q u i r e d to show a statistically significant difference (with p < 0.05 and a p o w e r of 80%). Statistics used were Fisher's exact test or Pearson X2 analysis with Yates' correction for categoric variables and ANOVA with multiple measures for continuous variables. A p g a r scores were c o m p a r e d with use o f the n o n p a r a m e t r i c Mann-Xghitney U test. Significance was c o n s i d e r e d at the f) < 0.05 level. T h e statistical p r o g r a m used was SAS 6.1 (SAS Institute, Cary, N.C.).

Results D u r i n g a p e r i o d of 15 m o n t h s (August 1995 t h r o u g h O c t o b e r 1996) 100 patients were r a n d o m i z e d in this trial:

1467

Table I. D e m o g r a p h i c data

Epidural (n = 49)

Narcotics (n = 51) Significance

Age (yr) 19.1 (3.9) 18.9 (2.9) Height (in) 64.8 (2.7) 64.6 (3.0) Weight (lb) 180.1 (38.7) 174.3 (36.8) Gestational age (wk) 39.3 (1.1) 39.4 (1.4) Race Black 46 42 White 1 9 Other 2 0 Cervix (era) At admission 4.2 (1.0) 4.2 (0.9) At randomization 4.7 (1.0) 4.7 (0.9) At oxytocin initiation 5.4 (1.4) 5.0 (1.0)

p p p p

= = = =

0.77 0.69 0.44 0.94

p = 0.008

p = 0.95 p = 0.56 p = 0.21

Data are presented as mean and SD or as number: Racial data were analyzed by Fisher's exact test; all other data, by analysis of variance.

Table 1I. Labor parameters m

Epidural } Narcoti& (n= 49) ] (n= 51) Significance Admission to randomization Admission to first analgesia Rupture of membranes to delivery Admission to complete dilatation (first stage of labor) Complete dilatatior/to delivery (second stage of labor) Randomization to complete dilatation Randomization to delivery Initiation ofoxytocin to delivery

94 (60)

76 (55)

p = 0.11

129 (57) 116 (62) 417 (171) 357 (186)

p = 0.26 p = 0.10

375 (I43) 357 (153)

p = 0.54

63 (53)

57 (49)

p = 0.55

287 (133). 283 (148)

p = 0.89

365 (159) 343 (169) 288 (157) 271 (183)

p = 0.51 p = 0.68

All data are recorded in minutes. Data are presented as mean and SD. Data were analyzed with analysis of variance. Oxytocin augmentation was required in 34 women in epidural group and 42 women in narcotic group.

49 in the epidural group and 5l in the narcotic group. Two w o m e n in the epidural group were delivered before obtaining regional analgesia and 12 w o m e n in the parenteral analgesia arm eventually received regional analgesia ("epidural rescue"). T h e "intent to treat" study design specified that these patients r e m a i n in their r a n d o m i z e d group for all statistical considerations. Table I demonstrates that the only significant difference in patient d e m o g r a p h i c characteristics was that there were m o r e white w o m e n in the narcotic group (p = 0.008). Thirty-four w o m e n in the epidural g r o u p and 42 in the narcotic group r e q u i r e d oxytocin a u g m e n t a t i o n of labor (p = 0.13). M a x i m u m oxytocin doses were similar in those w o m e n who r e q u i r e d a u g m e n t a t i o n (epidural 14.5 + 8.8 m U / m i n , narcotic 13.7 + 7.9 m U / m i n , p = 0.65). Likewise, Table I shows that cervical dilatation at

1468

Bofill~'t a[.

admission, at the time of randomization, and at the initiation of oxytocin augmentation were not significantly different between these two groups. Thirty-nine women in the epidural arm and 46 in the narcotic arm of the stud}, underwent surgical amniotomy (p = 0.14). Table II demonstrates that the length of time the membranes were ruptured was not significantly different when the two groups were compared. There was no difference between the groups in the length of time between admission and randomization (p = 0.11). Likewise, the length of the first and second stages of labor were not dissimilar. Also, there were no significant differences between the epidural and narcotic groups in the time from admission to first analgesia, from randomization to complete cervical dilatation, or from randomization to delivery. For those patients who required oxytocin, the time interval between initiation of oxytocin and delivery was similar between the groups. No significant differences in neonatal outcome were noted. Median Apgar scores at 1 (epidural 9 [range 2 to 9], narcotics 9 [range 2 to 9]; p = 1.00) and 5 (epidural 9 [range 7 to 10], narcotics 9 [range 7 to 10]; p = 0.53) minutes and mean cord arterial pH measurements (epidural 7.27 +- 0.06, narcotics 7.27 +- 0.08; p = 0,62) were equivalent. The mean weight of the infants on the epidural arm (3151 +- 426 gm) was somewhat smaller than that of their counterparts on the narcotic arm (3326 +_ 382 gin, p = 0.03), but this 175 gm (mean) difference did not appear to be clinically significant. There were a total of eight cesarean births in this trial; five of these were in women in the epidural group. One of the abdominal births in the epidural group was performed for the indication of nonreassuring fetal status (repetitive severe variable fetal heart rate decelerations with poor fetal heart rate variability). Four cesarean deliveries in the epidural group were performed for the indication of dystocia, whereas all three of the cesarean deliveries in the narcotic group were for this variable (p = 0.71). One of the women in the narcotics group who was delivered by cesarean was the recipient of a "rescue" epidnral. This patient was randomized at 4 cm, required oxytocin augmentation of labor for a protraction disorder of labor at 5 cm, received "rescue" epidural analgesia 4 hours after randomization, and had an eventual cesarean delivery after arrest of dilatation at a cervical dilatation of 7 cm. There were fewer spontaneous deliveries (n = 5) and more operative vaginal deliveries (n = 39) in women with conduction analgesia compared with those receiving narcotics (n = 20, n = 28, respectively; p = 0.004). Thirty-three of the 39 operative vaginal deliveries on the epidnral arm and 17 of the 28 operative vaginal deliveries accomplished on the narcotic arm of the study were undertaken for the purposes of resident training. Eleven patients in the epidural group and nine in the narcotics group had either an occiput

December 1997 _,MnJ Obstet Gynecol

posterior or occiput transverse position of the fetal head at the time of delivery (p = 0.58). None of the operative vaginal deliveries were initiated from the midpeMs and only two (one from each group) required >45 degrees of instrumental rotation. Fig. 1 demonstrates that, although the pain scores (mean ± SD) at the time of randomization were equi> alent between the two groups, these scores were significantly different at each point of measurement thereafter. Twelve women randomized to the narcotics arm received "rescue" epidural analgesia. For the women in the narcotics arm, 12 patients received one dose, 16 two doses, 17 three doses, five four doses, and one patient required six doses of butorphanol. Only one infant in this study (narcotics arm) received naloxone in the delivery room for suspected opioid-induced respiratory depression. Six women in the epidural group and one woman in the narcotics group became hypotensive after recei~'ing analgesia (p = 0.057). Emesis occurred in seven women in the epidural arm and four women in the narcotics arm (p = 0.35). Nausea without emesis was noted by five women in the epidural group and six in the narcotics group (p = 0.80). Comment There was no increase in dystocia-related cesarean delivery with epidural analgesia in this trial of spontaneously laboring nulliparous women. We attribute this to several factors. This study had rigid entry requirements that eliminated many of the pitfalls of earlier trials. Without a doubt the patients in this trial were in active labor. Criteria for entry stipulated that the cmwix be dilated at least 4 cm, effaced at least 80%, the vertex engaged, and the patient having frequent uterine contractions. None of the patients were being treated with oxytocin at the time of randomization or at the time they received the first dose of pain medication. This stands in sharp contrast to the study of Thorp et al., ~ where 12 patients (narcotics 3, epidural 9) were receiving oxytocin before receiving analgesia. Indeed, 32% (30/93) of the patients on the study o f T h o r p et a l l would not have met the entry criterion for cervical dilatation specified in our study. Although none of the patients in this trial were receiving oxytocin at tim time of entry into the study, 76% eventually underwent labor augmentation. This compares with a total of 43% (40/93) of patients in the study of Thorp et al. 1 who received oxytocin augmentation of labor. We surmise that this is the second reason for our inability to find a deleterious etiect of epidural analgesia on dystocia-related cesarean delivery. Our study strictly adhered to a protocol for the active management of labor that included early diagnosis of desktory labor and aggressive uterine stimulation. The resident physicians were allowed to perform elective outlet forceps or vacuum deliveries and this fact

~V%lume 177, Nmnber 6

Bofill et af.

1469

.~n J Obstet Gynecol

,,,

Epidural [ Narcotics

10

J

8 o o

2

6

f

0

1

2

3

4

5

Time (hours)

Fig. 1. Pain scores are recorded on y axis whereas dme (in hours) is recorded on x axis. Pain scores for epidural group are shown as solid t~4angles. Pain scores for narcotics group are shown as solid squares. E)Yor bars denote I SD. At randomization there is no significant difference in pain scores (p = 0.26). Difference in pain scores at hours 1 to 4 are significant (p < 0.001). Difference in pain scores at 5 hours is also significant (p = 0.01). Data were analyzed with analysis of variance.

separates the current study from all of the other reports and is in keeping with our institutional tradition regarding resident instruction in operative vaginal delivery. It serves two purposes: our house officers have ample opportunity to learn the techniques of operative vaginal delwery (both forceps and vacuum) and it prevents the needless prolongation of the second stage of labor, which can be remedied by a simple "lift-out" procedure. Likewise, our house officers and staff are comfortable performing rotational forceps and vacuum maneuvers and this may have prevented at least one cesarean section on each arm of the study. Again, those providers unwilling to attempt such operative vaginal deliveries would have no recourse except cesarean delivery. It is clear that the women on the epidural arm of o u r study have similarities and differences compared with the women on the epidural arm of the study o f T h o r p et al. ~ For example, in both reports the patients are similar in height, weight, and gestational age. However, the infants delivered to the women of both groups of the current study were smaller than those reported by Thorp et al) This is probably a reflection of population differences because our patients are typically black, young, and economically deprived. The women in the current study also had significantly greater cervical dilatation at study entry, shorter first and second stages of labor, and higher rates of operative vaginal deliver}'. Again, this probably reflects the fact that the women in our study were definitely in active labor compared with the women in other reports. ~'~ The subjects in our population frequently have anthropoid pelvic architecture, which pre-

disposes to malpositioning of the occiput and operative vaginal delivery,, thus making the finding of fewer cesarean deliveries even more gratifying. The current trial has certain limitations. First, the study did not meet the original sample size requirements. For this calculation we relied on the metaanalysis provided by Morton et al., 5 which noted that a 10% higher rate of cesarean delivery for dystocia should be expected in the epidural arm. This metaanalysis5 used retrospective and prospective data and also included data from multiparous patients ° and therefore may not have been an appropriate model for our study. At the initiation of the current trial there was only one published randomized trial that included only nulliparous patients. ~- These data noted dystocia-related cesarean delivery rates of 16% and 2% in the epidural and narcotic groups, respectively. With the power and significance levels as noted above, sample size analysis projects that 65 patients would be required in each arm of the study. Although there were differences in entry criteria between the study o f T h o r p et aI) and our investigation, it is apparent that overall this model for sample size analysis was the best available for our trial. We planned an interim analysis of results after 75% of randomizations because we were mindful of the criticism of repeated "peeking" at the data in an attempt to look for significance. Widl this second sample size analysis an interim review of the data was due after 98 patients had been randomized. The study protocol was discontinued for several reasons. First, these data were analyzed at this point because the only previous prospective study with a comparable design ~ had noted statistical significance after randomizing a similar n u m b e r of patients. Second, the interim analysis of the pain scores also argued persuasively in favor of termination of the trial because a clearly deleterious effect (increase in abdominal delivery rate) was not noted on the epidural arm of the study. Third, we underestimated the reluctance of patients with regard to randomization. We estimate that fewer than half of the women who were approached regarding entry into this trial actually accepted randomization. Finally, interim analysis demonstrated dystocia-related cesarean section rates of 8% in the epidural and 6% in the narcotic groups; we calculated that our study would require >2500 patients in each group to find a significant difference. It is important to note that these data may not be extrapolated to other clinical situations. Our anesthesiologists provided the best level of pain relief possible without significant impairment to the patient's expulsive efforts. Clearly, a regional technique with significant motor involvement below the T-10 level will predispose to inadequate expulsive efforts. The pain scores shown in Fig. 1 demonstrate that we greatly ameliorated, but did

1470

Bofill eta[.

not entirely abolish, the pain of labor with this analgesic technique. Also, this trial does n o t answer questions regarding the effect of epidnral analgesia on w o m e n whose labors are induced. Likewise, these data may n o t be reproducible unless there is a c o m m i t m e n t for the early detection and aggressive t r e a t m e n t o f protraction disorders of labor. This study does n o t address the multiparous patient n o r the w o m a n attempting a vaginal birth after cesarean delive W. Most importantly, these data do n o t solve the thorny d i l e m m a posed by T h o r p et al. I regarding the h i g h e r rates of cesarean birth for nulliparous w o m e n who receive epidural analgesia at cervical dilatations of < 4 cm. In spite of these problems, these data confirm that providers may offer c o n d u c t i o n analgesia to nulliparous w o m e n in active labor without increasing the risk of a b d o m i n a l delivery. REFERENCES

1. Thorp JA, Hu DH, Albin RM, McNitt J, Meyer BA, Coben GR, et al. The effect of intrapartum epidural analgesia on nulliparous labor: a randomized, controlled, prospective trial. Am J Obstet Gynecol 1993;169:851-8. 2. Ramin SM, Gambling DR, Lucas MJ, Sharma SK, SidawiJE, Leveno KJ. Randomized trial of epidural versus intravenous analgesia during labor. Obstet Gynecol 1995;86:783-9. 3. ThorpJA, Eekert LO, Ang MS,Johnston DA, Peaceman AM, Parisi VM. Epidural analgesia and cesarean section for dystocia: risk factors in nulliparas. Am J Perinatol 1991;8: 402-10.

December 1997 .4anJ Obstet Gynecol

4. ThorpJA, Parisi VM, Boylan PC,Johnston DA. The effect of continuous epidural analgesia on cesarean section for dystocia in nulliparous women. Am J Obstet Gynecol 1989;161: 670-5. 5. Morton SC, Williams MS, Keeler EB, GamboneJC, Kahn KL. Effect of epidural analgesia for labor on the cesarean deliveDT rate. Obstet Gynecol 1994;83:1045-52. 6. Dewan DM, Cohen SE. Epidural analgesia and the incidence of cesarean section: time for a closer look. Anesthesiology 1994;80:1189-92. 7. Chestnut DH, McGrathJM, Vincent RD, Penning DH, Choi WW, Bates JN, et al. Does early administration of epidural analgesia affect obstetric outcome in nulliparous women who are in spontaneous labor? Anesthesiology 1994;80: 1201-8. 8. Gribble RK, Meier PR. Effect of epidural analgesia on the primary cesarean section rate. Obstet Gynecol 1991;78: 231-4. 9. Philipsen T, Jensen NH. Epidural block or parenteral pethidine as analgesic in labor: a randomized study concerning progress in labor and instrmnental deliveries. EurJ Obstet G?mecol Reprod Biol 1989;30:27-33. 10. Lieberman E, LangJM, Cohen A, D'Agostino R, Datta S, Frigoletto FD. Association of epidural analgesia with cesarean delivery in nulliparas. Obstet Gynecol 1996;88: 993-1000. 11. Chesmut DH, Vandewalker GE, Owen CL, Bates JN, Choi WW. The influence of continuous bupivacaine analgesia on the second stage of labor and method of delivery in nulliparous women. Anesthesiology, 1987;66:774-80. 12. Chestnut DH, Vincent RD, McGrathJM, Choi WW, Bates JN. Does early administration of epidural analgesia affect obstetric outcome in nulliparous women who are receiving intravenous oxytocin? Anesthesiology 1994;80:11932OO.