94

Induction of labor with misoprostol for premature rupture ofmembranes beyond thirty-six weeks’ gestation

Deborah A. Wing, MD, and Richard H. Paul, MD

Los Angeles, California

OBJECTIVE: Our purpose was to compare vaginally administered misoprostol (Cytotec) with intravenousoxytocin for labor induction in women with premature rupture of membranes beyond 36 weeks’ gestation.STUDY DESIGN: Two hundred subjects with rupture of membranes without labor were randomly assigned toreceive vaginally administered misoprostol or intravenous oxytocin. Twenty-five micrograms of misoprostol(Cytotec) was placed in the posterior vaginal fornix. If cervical ripening (Bishop score of ≥8 or cervical dilata-tion of ≥3 cm) or active labor did not occur, a single repeat dose of misoprostol was given 6 hours later.Oxytocin was administered intravenously by a standardized incremental infusion protocol to a maximumdose of 22 mU per minute.RESULTS: Of the 197 subjects evaluated, 98 received misoprostol and 99 oxytocin. The average intervalfrom start of induction to vaginal delivery was about 1 hour longer in the misoprostol group (811.5 ± 511.4minutes) than in the oxytocin group (747.0 ± 448.0 minutes) (P = .65, log transformed data). Oxytocin admin-istration was necessary in 37 of 98 (37.8%) of misoprostol-treated subjects. Vaginal delivery occurred in 85misoprostol-treated subjects (86.7%) and 82 (85.9%) oxytocin-treated subjects (relative risk 1.17, 95% confi-dence interval 0.78 to 1.78, P = .45) with the remainder undergoing cesarean birth. There was no differencein the incidence of tachysystole (six or more uterine contractions in a 10-minute window for two consecutive10-minute periods) or hypertonus between the two groups. There was no significant difference in frequencyof abnormal fetal heart rate tracings between the two groups (29.6% in the misoprostol group and 28.9% inthe oxytocin group, P = .91). Chorioamnionitis was diagnosed in 28 (28.6%) misoprostol-treated subjects and26 (26.3%) oxytocin-treated subjects (P = .72, relative risk 1.06, 95% confidence interval 0.78 to 1.45). Nosignificant differences were found in the incidence of fetal meconium (8.1% and 9.1%), 1- or 5-minute Apgarscores <7 (11.0% and 10.2% of 1-minute Apgar scores, and 2.0% and 2.0% of 5-minute Apgar scores),neonatal resuscitation (24.5% and 27.6%), or admission to the neonatal intensive care unit (25.5% and32.3%) between the two groups.CONCLUSIONS: Vaginal administration of misoprostol (Cytotec) is an effective alternative to oxytocin infu-sion for labor induction in women with premature rupture of the membranes near term. The incidence of un-toward effects is similar with use of the two agents. (Am J Obstet Gynecol 1998;179:94-9.)

Key words: Induction of labor, misoprostol, premature rupture of membranes

Spontaneous rupture of membranes occurs beyond 36weeks’ gestation and before the onset of labor in approx-imately 10% of pregnancies.1 The management of thiscondition is controversial. Traditionally, induction oflabor is undertaken to prevent chorioamnionitis andneonatal sepsis. The incidence of both these problems isincreased when premature rupture of membranes is pre-

sent >24 hours before delivery than when delivery occurswithin 24 hours.1, 2 Multiple studies have compared theuse of expectant management with active labor induc-tion for treatment of this problem.3-7 Expectant manage-ment results in a lower incidence of cesarean deliveriesperformed for failed labor inductions.4, 5 However, thisbenefit must be weighed against higher rates of maternaland neonatal infection6 and the possibility of cord pro-lapse resulting from delayed time of delivery with expec-tant management.

Prostaglandins have been shown to be effective agentsfor cervical ripening and labor induction.8 Several stud-ies have demonstrated the safety and effectiveness of var-ious preparations of prostaglandin E2 (PGE2) when pre-mature rupture of the membranes is present.9-15 Thesestudies have shown that the use of intravenous oxytocinand the incidence of cesarean delivery are less when

From the Department of Obstetrics and Gynecology, Women’s andChildren’s Hospital, University of Southern California School ofMedicine.Received for publication August 18, 1997; revised December 9, 1997;accepted December 22, 1997.Reprint requests: Deborah A. Wing, MD, Women’s and Children’sHospital, University of Southern California School of Medicine,Department of Obstetrics and Gynecology, 1240 N Mission Road, LosAngeles, CA 90033.Copyright © 1998 by Mosby, Inc.0002-9378/98 $5.00 + 0 6/1/88331

Volume 179, Number 1 Wing and Paul 95Am J Obstet Gynecol

PGE2 is used to induce labor than when it is not used. We have previously shown that vaginal administration

of misoprostol (Cytotec), a synthetic prostaglandin E1analog, is as effective for prelabor induction cervicalripening and the induction of labor as dinoprostone(PGE2) is in different vehicles (Prepidil or Cervidil), theonly Food and Drug Administration (FDA)–approveddrug for this purpose.16-18 Because misoprostol appearsas efficacious as dinoprostone for preinduction cervicalripening and the induction of labor, we believed it couldalso be used safely and effectively in women who havepremature rupture of membranes. Others have reportedthe use of misoprostol for cervical ripening and labor in-duction in women with premature rupture of mem-branes, but to our knowledge there is only one investiga-tion in which premature rupture of membranes was thesole indication for labor induction. Misoprostol was ad-ministered orally in this study.19

Thus the objective of this study was to compare vagi-nally administered misoprostol with oxytocin infusion toinduce labor in women with premature rupture of themembranes beyond 36 weeks’ gestation who had no evi-dence of spontaneous labor.

Material and methods

From May 31, 1995, to Feb. 2, 1997, all women withspontaneous rupture of membranes beyond 36 weeks’gestation at Women’s and Children’s Hospital, LosAngeles County–University of Southern CaliforniaMedical Center were evaluated for study participation.The study was approved by the Los Angeles County–University of Southern California Investigational ReviewBoard. Women meeting study criteria were asked abouttheir willingness to participate in the trial, and if theyelected to do so, an informed consent was obtained.

Rupture of membranes was confirmed by a combina-tion of two or more of the following: (1) visualization ofamniotic fluid in the vaginal vault with the use of a sterilespeculum, (2) pH >6.5 as indicated by colorimetric pHchange by Nitrazine paper (Bristol-Myers Squibb, CherryHill, NJ), or (3) microscopic appearance of an arboriza-tion pattern of the amniotic fluid after the fluid in thevagina was placed on a glass slide and allowed to dry.

Inclusion criteria, in addition to confirmation of rup-tured membranes, were (1) singleton gestation, (2) reac-tive fetal heart rate (FHR) pattern, (3) cephalic presenta-tion, and (4) presumed gestational age of ≥36 weeks asdetermined by dates by last menstrual period, serial pre-natal examinations, or prenatal ultrasonographic exami-nation.

Subjects were excluded if any of the following werepresent: (1) cervical dilatation in >3 cm, (2) uterine con-tractions at a rate of >12 per hour, (3) an estimated fetalweight >4500 gm or other evidence of cephalopelvic dis-

proportion, (4) an estimated fetal weight <1800 g as as-sessed by ultrasonography, (5) placenta previa or unex-plained vaginal bleeding, (6) active herpes simplex, (7)previous cesarean delivery or history of uterine incision,(8) evidence of chorioamnionitis as determined by tem-perature ≥100.4°F and the presence of uterine tender-ness or foul-smelling amniotic fluid, (9) parity >5, (10)any moderate or severe preexisting medical disease, suchas cardiovascular disease or chronic renal failure, and(11) any contraindication for use of prostaglandins, suchas glaucoma or sickle cell disease.

The subjects were assigned by means of a computer-ized random number generator to receive either miso-prostol (Cytotec, GD Searle, Chicago) or oxytocin.Group allocation was predetermined and placed in con-secutively numbered and sealed opaque envelopes. Oncea subject was deemed eligible and gave informed consentfor study participation, she was assigned a sequentialstudy number. The primary investigator, who was respon-sible for maintaining the envelopes and not directly in-volved in patient care, was then contacted and wouldopen the next envelope for the purpose of treatment al-location.

Over the 21-month study period a total of 214 subjectswere asked to participate, and 200 agreed. Of these, 101were randomized to receive misoprostol and 99 to re-ceive oxytocin. Three subjects were excluded from dataanalysis because of deviation from study protocol. Thesesubjects received a total of three doses of misoprostol. Nowomen withdrew from the study protocol.

The initial Bishop score was determined by a seniorobstetrics-gynecology resident, and when possible sub-jects were reexamined by the same individual. RepeatBishop scores were determined before each misoprostoldose and just before oxytocin administration in miso-prostol-treated subjects. A Bishop score of 13 was arbi-trarily assigned for those subjects who had entered theactive phase of labor. Once a patient was in the activephase of labor, routine intrapartum management oc-curred without regard to the treatment allocation.

Induction with either misoprostol or oxytocin wasstarted a minimum of 6 hours after the spontaneous rup-ture of membranes in the absence of uterine contrac-tions. The mean time from rupture of membranes to ini-tiation of induction was 890.6 ± 1071.1 minutes in themisoprostol treatment group and 843.1 ± 981.7 minutesin the oxytocin treatment group (P = .94, log-trans-formed data).

Twenty-five micrograms (one fourth of a 100 µg tablet)of misoprostol was placed in the posterior vaginal fornixby a senior house officer. If the subject did not demon-strate adequate uterine contraction frequency (morethan three contractions in a 10-minute window of obser-vation), the same dose was repeated one time in 6 hours

96 Wing and Paul July 1998Am J Obstet Gynecol

for a maximum dose of 50 µg. If the subject did notdemonstrate adequate uterine activity 6 hours after re-ceiving the second dose of misoprostol, an oxytocin infu-sion was initiated.

Oxytocin was administered by an infusion pump ac-cording to a standard protocol. The infusion was initi-ated with a dose of 1 mU/min with incremental increaseevery 30 minutes to a maximum of 20 mU/min.

Continuous external fetal monitoring and externaltocodynamometry were used. FHR patterns were classi-fied in the manner described by Kubli et al.20 AbnormalFHR patterns were defined as the presence of either fetaltachycardia or bradycardia, late decelerations, or moder-ate-to-severe variable FHR decelerations.

Evaluation of uterine activity monitoring was per-formed to assess the frequency and duration of tachysys-tole, hypertonus, and hyperstimulation syndrome.21

Uterine hypertonus was defined as a single uterine con-traction lasting ≥2 minutes, tachysystole as at least sixuterine contractions in 10 minutes for two consecutive10-minute windows, and hyperstimulation as either hy-pertonus or tachysystole associated with an abnormalFHR pattern. Either intravenous or subcutaneous terbu-taline 0.25 mg or intravenous magnesium sulfate was ad-ministered to treat these contraction abnormalities.

The diagnosis of intraamniotic infection or chorioam-nionitis was made if any of the following were present:maternal temperature ≥100.4°F, foul-smelling amnioticfluid, fundal tenderness, or persistent elevation in FHRbaseline ≥160 beats/min. Chorioamnionitis was treat-ed with antipyretics and appropriate antimicrobialagents.

The primary outcome measure was the occurrence ofvaginal delivery within 24 hours from the start of induc-tion, defined as a successful induction. Induction failurewas defined as failure to achieve cervical dilatation ≥4 cmafter a trial of oxytocin. Other outcome variables wereroute of delivery, and time interval from start of induc-tion to delivery.

Sample size calculations were based on the number ofsubjects achieving a vaginal delivery within 24 hours afterinitiating the induction attempt by use of the method de-scribed by Meinert.22 A 20% difference in the numbers

of subjects achieving a successful induction in the miso-prostol treatment arm versus the oxytocin treatment armwas assumed. It was assumed that 80% of misoprostol-treated patients and 60% of oxytocin-treated patientswould achieve this outcome. On the basis of a type I errorof .05 and a type II error of .2, a sample size of 81 pertreatment arm was necessary. The test was two sided.

Statistical analysis was performed with use of the χ2,Student t, Mann-Whitney U, and Fisher’s exact testswhere appropriate. All tests were 2-sided, with a .05 sig-nificance. Differences in age, estimated gestational age,total doses of the inductive agent required, total oxytocindose, time in tachysystole (log transformed), and time todelivery (log transformed) were analyzed with t tests; dif-ferences in gravidity, parity, route of delivery, presence ofcomplications, and need for oxytocin were analyzed withχ2 tests or Fisher’s exact tests when appropriate. Apgarand Bishop scores were analyzed with use of the Mann-Whitney U test. Relative risk and corresponding 95%confidence interval comparing misoprostol to oxytocintreatments were also reported when necessary.

Results

The subjects were similar with respect to mean age,gravidity, parity, height, weight, ethnicity, and estimatedgestational age at entry. Forty-one (41.8%) of the miso-prostol subjects and 47 (48.0%) of the oxytocin subjectswere nulliparous (P = .35). The mean estimated gesta-tional age for the misoprostol treatment group was 38.7 ±1.5 (SD) weeks and for the oxytocin treatment group38.9 ± 1.6 (SD) weeks (P = .46). The median preinduc-tion Bishop score was 3 in the misoprostol group (range0 to 10) and 4 in the oxytocin group (range 0 to 10) (P =.57, Mann-Whitney U test).

A similar percentage of subjects in both treatmentgroups had a treatment success. Seventy-five (75.8%) ofmisoprostol-treated subjects and 73 (74.5%) of oxytocin-treated women were delivered vaginally within 24 hoursof initiating induction (P = .87). There was no significantdifference in the mean time interval from start of induc-tion to delivery between the two treatment groups. Themean time from start of induction to delivery, regardlessof the route, was 900.8 ± 558.1 minutes for the misopros-

Table I. Time intervals to delivery

Misoprostol (n = 98) Oxytocin (n = 99) Statistical significance

Initiation to delivery (min) 900.8 ± 558.1 833.5 ± 485.3 P = .99*

Initiation to vaginal delivery (min) 811.5 ± 511.4 747.0 ± 448.0 P = .65*

Vaginal delivery in 12 h 44 (44.4%) 49 (50%) P = .30Vaginal delivery in 24 h 75 (75.8%) 73 (74.5%) P = .87

Data presented as mean ± SD or number and percent.*Based on log-transformed data.

Volume 179, Number 1 Wing and Paul 97Am J Obstet Gynecol

tol-treated subjects and 833.5 ± 485.3 minutes for theoxytocin-treated subjects (P = .99, log-transformed data).The mean time intervals from start of induction to vagi-nal delivery were also similar. The mean time from startof induction to vaginal delivery was 811.5 ± 511.4 minutesfor misoprostol-treated women and 747.0 ± 448.0 min-utes for oxytocin-treated women (P = .65, log-trans-formed data) (Table I).

Parity influenced the success of induction, regardlessof the treatment arm. In the misoprostol treatmentgroup 26 nulliparous (63%) and 49 multiparous (86%)women were delivered vaginally within 24 hours after ini-tiating induction, and in the oxytocin treatment group29 nulliparous (60%) and 44 multiparous (86%) womenwere delivered in this time period. Forty-one (41.8%) ofthe misoprostol-treated subjects and 48 (48.4%) of theoxytocin-treated subjects were nulliparous (P = .35). Themean duration from induction to delivery in misopros-tol-treated nulliparous women was 1116.3 ± 546.1 min-utes and in misoprostol-treated parous subjects it was745.8 ± 517.9 minutes, whereas the mean duration frominduction to delivery in oxytocin-treated nulliparouswomen was 1080.1 ± 489.3 minutes and in oxytocin-treated multiparous women it was 601.3 ± 350.90 min-utes. These differences were not statistically different be-tween the two treatment groups (nulliparous subjects P =.87 and multiparous subjects P = .29, log-transformeddata).

The misoprostol-treated women required an average of1.3 ± 0.5 doses. Sixty-eight women received one dose ofmisoprostol, and 30 received two doses. The average timeinterval between doses was 469.7 ± 182.2 minutes (range330 to 1050 minutes). Thirty-seven (37.3%) women inthis group required oxytocin augmentation. The indica-tions for oxytocin augmentation were failure to begin ac-tive labor after two doses of misoprostol (14/37, 37.8%),inadequate uterine activity in the active phase of labor(13/37, 35.1%), and adequate cervical ripening after thefirst dose of misoprostol (10/37, 27.0%).

Uterine tachysystole occurred in six subjects in eachtreatment group. The time interval from the last dose ofmisoprostol to the onset of tachysystole was 243.3 ± 118.8minutes and from initiation of oxytocin to onset oftachysystole 417.0 ± 209.6 minutes (P = .14). Hypertonuswas reported in only one misoprostol-treated subject.There were no instances of uterine hyperstimulation.Abnormal FHR tracings occurred in 29 (29.6%) and 28(28.9%) of the misoprostol-treated and oxytocin-treatedwomen, respectively (P = .91) (Table II).

Evidence of intraamniotic infection occurred in simi-lar frequencies in the two treatment groups. Twenty-eight(28.6%) misoprostol-treated women and 26 (26.3%) oxy-tocin-treated women had intraamniotic infection (P =.71, relative risk 1.06, 95% confidence interval 0.78 to1.45). Clinical suspicion of neonatal sepsis occurred in21 (21.4%) infants born to misoprostol-treated womenand 27 (27.3%) infants of oxytocin-treated women (P =.34, relative risk 0.85, 95% confidence interval 0.60 to1.19). There were 2 cases of confirmed neonatal sepsis,both born to mothers treated with oxytocin.

Eighty-five (85.9%) misoprostol-treated women weredelivered vaginally and 82 (83.7%) oxytocin-treatedwomen were delivered vaginally (relative risk 1.17, 95%confidence interval 0.78 to 1.78). Of the 13 cesareanbirths in women assigned to the misoprostol treatmentarm, 9 were for arrest disorders of labor and 4 were forabnormal FHR tracings. Of the 17 cesarean deliveries inthe women treated with oxytocin, there were 10 for arrestdisorders, 4 for abnormal FHR tracings, and 3 for failedinductions.

Neonatal outcomes were also similar between the twotreatment groups (Table III). The mean birth weightsdid not differ between the two groups (3214 ± 428 g inthe misoprostol-treated patients and 3250 ± 480 g inthe oxytocin-treated patient (P = .59) The percentage ofinfants requiring resuscitation at delivery, demonstratingmeconium passage, assigned Apgar scores <7 at 1 and 5minutes, and requiring admission to the intensive care

Table II. Intrapartum complications

Misoprostol (n = 98) Oxytocin (n = 99) Statistical significance

Abnormal FHR patterns 29 (29.6%) 28 (28.9%) P = .91Repetitive moderate-severe variable decelerations 17 10Late decelerations 3 1Prolonged decelerations 6 9Other 0 1Combinations 3 7

Meconium passage 8 (8.1%) 9 (9.1%) P = .82Thin 5 9Thick 3 0

Chorioamnionitis/intrauterine infection 28 (28.6%) 26 (26.3%) P = .71

Data presented as number and percent.

98 Wing and Paul July 1998Am J Obstet Gynecol

unit were similar between the treatment groups.

Comment

There was no significant difference in the percentageof subjects having success of induction, defined as vagi-nal delivery within 24 hours after initiation of induction,with use of vaginally administered misoprostol or oxy-tocin infusion. There was also no significant difference inthe mean time intervals from start of induction to deliv-ery in women with spontaneous rupture of membranesbeyond 36 weeks of gestation when treated with eithermisoprostol or oxytocin. There was a trend toward ahigher number of failed inductions in the oxytocin treat-ment group, because all three cesarean deliveries per-formed for this indication occurred in oxytocin-treatedsubjects. However, this difference was not statistically sig-nificant.

We were unable to reduce the cesarean delivery rate inwomen with premature rupture of membranes by admin-istering misoprostol rather than oxytocin for labor in-duction. Because the majority of cesarean sections inmisoprostol-treated women were performed for labordystocia, rather than failed induction, misoprostol ap-pears to be effective for causing cervical dilatation and ef-facement and inducing labor in patients with prematurerupture of membranes. We were also unable to reducethe numbers of women receiving antimicrobial therapyfor the diagnosis of chorioamnionitis by administeringmisoprostol rather than oxytocin. Intravaginal misopros-tol administration may have contributed to the relativelyhigh frequency of chorioamnionitis in this treatmentarm.

We did not meet our assumptions for our power calcu-lation, because the numbers of subjects achieving a suc-cessful induction did not differ between the two groups.However, it would appear that misoprostol can be usedwith similar efficacy and safety in patients with prematurerupture of membranes beyond 36 weeks’ gestation.

The dosing regimen of misoprostol used in this investi-gation was chosen on the basis of our past experiencewith this medication for cervical ripening and labor in-duction18 and reflects a degree of conservatism in our ap-

proach to this subset of patients.In our efforts to create the safest dosing regimen possi-

ble for misoprostol, we designed this protocol to reflectthe most conservative dosing regimen we had had expe-rience with to date, that of misoprostol 25 µg every 6hours. It would seem from extrapolation from our mostrecent investigation in which we compared giving miso-prostol every 4 hours to a 24-hour exposure to the dino-prostone vaginal insert Cervidil that a similar dosing fre-quency could also be used safely in patients withruptured membranes.23

REFERENCES

1. Gunn GC, Mishell DR Jr, Morton DG. Premature rupture offetal membranes: a review. Am J Obstet Gynecol 1970;106:469-83.

2. Lanier LR, Scarbrough RW Jr, Fillingim DW, Baker RE Jr.Incidence of maternal and fetal complications associated withrupture of the membranes before onset of labor. Am J ObstetGynecol 1965;93:398-404.

3. Wager MV, Chin VP, Peters CJ, Drexler B, Newman LA. A com-parison of early and delayed induction of labor with sponta-neous rupture of membranes at term. Obstet Gynecol1989;74:93-7.

4. Grant JM, Serle E, Mahmood T, Sarmandal P, Conway DI.Management of prelabour rupture of the membranes in termprimigravidae: report of a randomized prospective trial. Br JObstet Gynaecol 1992;99:557-62.

5. Conway DI, Prendiville WJ, Morris A, Speller DCE, Stirrat GM.Management of spontaneous rupture of the membranes in theabsence of labor in primigravid women at term. Am J ObstetGynecol 1984;150:947-51.

6. Duff P, Huff RW, Gibbs RS. Management of premature ruptureof membranes and unfavorable cervix in term pregnancy.Obstet Gynecol 1984;63:697-702.

7. Hannah ME, Ohlsson A, Farine D, Newson SA, Hodnett ED,Myhr TL, et al. Induction of labor compared with expectantmanagement for prelabor rupture of the membranes at term. NEngl J Med 1996;334:1005-10.

8. Keirse MJNC. Prostaglandins in preinduction cervical ripening:meta-analysis of worldwide clinical experience. J Reprod Med1993;38:89-95.

9. Goeschen K. Premature rupture of membranes near term: in-duction of labor with endocervical prostaglandin E2 gel or in-travenous oxytocin. Am J Perinatol 1989;6:181-4.

10. Ekman-Ordeberg G, Uldbjerg N, Ulmsten U. Comparison of in-travenous oxytocin and vaginal prostaglandin E2 gel in womenwith unripe cervixes and premature rupture of the membranes.Obstet Gynecol 1985;66:307-10.

11. Magos AL, Noble MCB, Wong Ten Yue A. Controlled study com-paring vaginal prostaglandin E2 pessaries with intravenous oxy-

Table III. Neonatal outcomes

Misoprostol (n = 98) Oxytocin (n = 99) Statistical significance

Birth weight (g) 3214 ± 428 3250 ± 480 p = .59Apgar score <7

1 min 11 (11.1%) 10 (10.2%) p = .845 min 2 (2.0%) 2 (2.0%) p = .99*

Neonatal resuscitation 24 (24.5%) 27 (27.6%) p = .63Admission to NICU 25 (25.5%) 32 (32.3%) p = .29

Days in NICU 9.2 ± 6.4 7.7 ± 3.3 p = .29Hyperbilirubinemia 5 (5.1%) 6 (6.1%) p = .99*

Data presented as mean ± SD or number and percent. NICU, Neonatal intensive care unit.*Fisher’s exact test.

Volume 179, Number 1 Wing and Paul 99Am J Obstet Gynecol

tocin for the stimulation of labour after spontaneous rupture ofmembranes. Br J Obstet Gynaecol 1983;90:726-31.

12. Granstrom L, Elkman G, Ulmsten U. Cervical priming andlabor induction with vaginal application of 3 mg PGE2 in sup-positories in term pregnant women with premature rupture ofamniotic membranes and unfavorable cervix. Acta ObstetGynecol Scand 1987;66:429-31.

13. Hauth JC, Cunningham FG, Whalley PJ. Early labor initiationwith oral PGE2 after premature rupture of the membranes atterm. Obstet Gynecol 1977;49:523-6.

14. Lange AP, Secher NJ, Nielsen FH, Pedersen GT. Stimulation oflabor in cases of premature rupture of the membranes at ornear term: a consecutive randomized study of prostaglandin E2-tablets and intravenous oxytocin. Acta Obstet Gynecol Scand1981;60:207-10.

15. Westergaard JG, Lange AP, Pederson GT, Secher NJ. Use of oraloxytocics for stimulation of labor in cases of premature ruptureof the membranes at term: a randomized comparative study ofprostaglandin E2 tablets and demoxytocin resoriblets. ActaObstet Gynecol Scand 1983;62:111-6.

16. Wing DA, Jones MM, Rahall A, Goodwin TM, Paul RH. A com-parison of misoprostol and prostaglandin E2 gel for preinduc-tion cervical ripening and the induction of labor. Am J ObstetGynecol 1995;172:1804-10.

17. Wing DA, Jones MM, Rahall A, Goodwin TM, Paul RH.Misoprostol: an effective agent for cervical ripening and laborinduction. Am J Obstet Gynecol 1995;172:1811-6.

18. Wing DA, Paul RH. A comparison of differing dosing regimensof misoprostol for cervical ripening and labor induction. Am JObstet Gynecol 1996;175:158-64.

19. Ngai SW, To WK, Lao T, Ho PC. Cervical priming with oral miso-prostol in pre-labor rupture of membranes at term. ObstetGynecol 1996;87:923-6.

20. Kubli FW, Hon EH, Khazin AF, Takemura H. Observations onheart rate and pH in human fetus during labor. Am J ObstetGynecol 1969;104:1190-206.

21. Stookey RA, Sokol RJ, Rosen MG. Abnormal contraction pat-terns in patients monitored in labor. Obstet Gynecol1973;42:359-67.

22. Meinert CL. Clinical trials design, conduct and analysis. Oxford(UK): Oxford University Press; 1993.

23. Wing DA, Paul RH. A comparison of intermittent vaginal ad-ministration of misoprostol with continuous dinoprostone forcervical ripening and labor induction. Am J Obstet Gynecol1997;177:612-8.