Dodd, Jodie M. PhD, FRANZCOG1; Crowther, Caroline A. MD, FRANZCOG1; Robinson, Jeffrey S. FRANZCOG1
Several retrospective clinical studies have assessed diurnal variations in onset of labor and time of birth in women both at term1–3 and before term.4,5 These studies consistently demonstrate a peak in spontaneous birth in the evening and early hours of the morning.
There is a well-documented diurnal variation in the concentrations of steroid hormones in blood, plasma, and urine in pregnant women, with a nadir in progesterone,6–8 estrone, estriol, and estradiol (E2)9–11 concentrations documented in the early hours of the morning. In contrast, oxytocin concentrations demonstrate a reciprocal relationship, with greater concentrations measured in the evening and night,12 correlating with a documented nocturnal increase in uterine activity.13,14 Contributing to this enhanced myometrial contractility is an increase in oxytocin receptor concentrations and increased sensitivity to oxytocin.15–17
In our institution, women scheduled for induction of labor are admitted to hospital in the evening, in the expectation that women will labor and give birth during daylight hours. Physiologically, evening may not be the most appropriate time to commence the induction process, but may represent more closely the optimal time to be in active labor.
This nested randomized trial was conducted to assess the preferences women have for timing of induction of labor and the effect that timing of induction may have on maternal and infant outcomes. We hypothesized that beginning induction of labor in the morning would more closely reflect the physiologic timing of onset of labor and be associated with fewer women who remain undelivered 24 hours after the induction commenced.
MATERIALS AND METHODS
This nested prospective randomized clinical trial comparing time of admission to start induction of labor (Morning Admission Group—after 0800 hours—compared with Evening Admission Group—after 2000 hours) was conducted at the Women's and Children's Hospital, Adelaide, Australia between April 2001 and December 2004. Research and Ethics committee approval was received from Women's and Children's Hospital, Adelaide, Australia. Women who consented to participation in our randomized trial comparing oral misoprostol solution with vaginal prostaglandin E2 gel for induction of labor (the MILO trial)18 were randomly allocated at the time their admission for induction was scheduled with delivery suite, to admission in the morning (0800 hours) or in the evening (2000 hours) (the timing of induction trial).
Briefly, women with a singleton pregnancy in cephalic presentation, at gestational age of more than 36+6 weeks, with any indication for prostaglandin induction of labor and who had consented to participate in the MILO trial, were eligible.
Details of the MILO trial comparing oral misoprostol with vaginal PGE2 gel have been published previously.18 Eligible women were recruited from the antenatal clinic and delivery suite, with gestational age greater than 36+6 weeks, a singleton pregnancy in cephalic presentation, Bishop's score less than 7, and an indication for prostaglandin induction of labor. Women assigned to the oral misoprostol group received 20 mcg oral misoprostol solution at 2-hour intervals (maximum of six doses) and placebo vaginal gel, and those in the vaginal prostaglandin group received vaginal PGE2 gel at 6-hour intervals (maximum of two doses) and oral placebo solution. Treatment packs appeared identical and were sealed to prevent tampering. All women, caregivers, and research investigators remained blind to the induction agent received.
The randomization schedule was generated by a nonclinical researcher using a computer-generated sequence with variable blocks and stratification for the woman's parity (0 and 1–4). Timing of admission (Morning Admission or Evening Admission) was written on a card, folded, and placed inside sequentially numbered, sealed opaque envelopes. The envelope was opened after a woman had consented to participation in the MILO trial, and the time for induction was scheduled in accordance with the written instructions. Treatment packs were prepared by the principal investigator and the research midwife and labeled by a nonclinical researcher according to the generated randomization schedule accounting for the woman's parity and time of admission to delivery suite. The four possible combinations were therefore nulliparous woman and morning admission, multiparous woman and morning admission, nulliparous woman and evening admission, and multiparous woman and evening admission. Women, caregivers and outcome assessors were not blinded to the woman's time of admission for induction of labor to start. At the time of admission to delivery suite for induction of labor, trial entry details were confirmed. A study number was allocated by the midwife taking the next identically appearing, sequentially numbered treatment pack appropriate for the woman's parity and time of admission to delivery suite. The study number allocated was recorded on the trial entry form. The study treatment pack was opened, and contained either oral misoprostol and vaginal placebo (tylose gel), or oral placebo (vitamin B6) and vaginal prostaglandin E2 gel (Prostin E2, Pharmacia & Upjohn Company, Kalamazoo, MI). Each pack contained a sheet detailing treatment schedules. The treatment packs were kept at 4°C (although this was not a requirement for misoprostol). After birth and before discharge from hospital, women were asked to complete a short questionnaire detailing their experiences of induction of labor (Dodd JM. Misoprostol for the induction of labour at term [dissertation]. Adelaide, Australia: The University of Adelaide; 2005), and at 6 weeks postpartum. Women who did not return the questionnaire by mail were telephoned, and where possible, the questionnaire completed over the phone. The primary study outcomes were consistent with the Cochrane generic protocol relating to induction of labor,19 and were vaginal birth not achieved within 24 hours (this included women who birthed vaginally beyond 24 hours and all women having a cesarean delivery); uterine hyperstimulation with associated fetal heart rate changes; cesarean delivery (all); and cesarean delivery for fetal distress. Secondary study outcomes related to evidence of effect, neonatal complications, and maternal complications as described in the Cochrane generic protocol relating to methods of induction of labor.19 Data forms were completed by the midwife caring for the woman, with information confirmed and checked by the principal investigator or research midwife before the woman was discharged from hospital. All data were then entered into a database created in Access '97 (Microsoft Corp., Edmond, WA) by the principal investigator. Data were analyzed on an intention-to-treat basis and blind to the allocated time of admission for induction using SAS 9.1 (SAS Institute Inc., Cary, NC) statistical software. The primary and secondary outcomes were compared using χ2 tests and Fisher exact test, with calculation of relative risk (RR) with 95% confidence interval (CI) for dichotomous data. Normally distributed continuous data were compared using the Student t test, and skewed data using nonparametric tests (Wilcoxon rank sum). Prespecified subgroup analyses were planned to assess the effect of maternal parity (nulliparous compared with multiparous) and initial Bishop's score at commencing induction (Bishop's score 0–3 compared with Bishop's score 4–6) on the time of admission for induction of labor. A P value of less than .05 was considered statistically significant. A sample size of 620 women was calculated to give 80% power to detect a 50% difference in the number of women undelivered after 24 hours from 20% in the morning admission group to 30% in the evening admission group (P<.05).
A total of 1,072 eligible women were approached for trial participation by a clinical researcher at the time of booking the induction of labor, of whom 775 (72.3%) provided written consent to participate (Fig. 1). Of the 775 women who gave consent, 380 (49.0%) women were scheduled for morning admission and 395 (51.0%) women for evening admission. Before the date of scheduled induction, 155 (20.0%) women entered labor spontaneously (100 [64.5%] women scheduled for morning admission and 55 [35.5%] women scheduled for evening admission). Of the 620 (70.0%) women who were admitted for induction and randomly assigned into the timing of induction trial, 280 (45.2%) were assigned to the morning admission group, and 340 (54.8%) were assigned to the evening admission group.
Baseline characteristics between the two groups were comparable (Table 1). More women allocated to the morning admission group entered labor spontaneously before the scheduled date of induction. This was unexpected, and thought to reflect the increased ease in scheduling admission for induction of labor in the evening, consistent with the unit's usual policy. As there were no other differences in baseline characteristics, no adjustments were made in subsequent analyses.
Admission and commencing induction of labor in the morning, compared with the evening was not associated with statistically significant differences in the number of women who did not achieve vaginal birth within 24 hours (morning admission 121 of 280 [43.2%] compared with evening admission 151 of 340 [44.4%]; RR 0.97, 95% CI 0.81–1.16; P=.432), uterine hyperstimulation with associated fetal heart rate changes (morning admission 5 of 280 [1.8%] compared with evening admission 0 of 340 [0.0%]; RR 7.31, 95% CI 0.88–60.58; P=.998), cesarean delivery (morning admission 62 of 280 [22.1%] compared with evening admission 89 of 340 [26.2%]; RR 0.83, 95% CI 0.63–1.10; P=.194), or cesarean for fetal distress (morning admission 27 of 280 [9.6%] compared with evening admission 33 of 340 [9.7%]; RR 0.98, 95% CI 0.60–1.59; P=.934)(Table 2). There were no statistically significant differences identified between morning and evening admission and maternal parity (nulliparous compared with multiparous), or initial Bishop's score (Bishop's score 0–3 compared with Bishop's score 4–6) for primary trial outcomes (Table 2).
The use of oxytocin was significantly less in women who were admitted in the morning compared with women admitted for induction in the evening (morning admission 126 of 280 [45.0%] compared with evening admission 184 of 340 [54.1%]; RR 0.83, 95% CI 0.70–0.97; P=.022; number needed to treat to benefit 11; 95% CI 6–68)(Table 3). Women with a Bishop's score of 0–3 on admission were less likely to require oxytocin infusion when admission occurred in the morning compared with the evening (morning admission 80 of 177 [45.2%] compared with evening admission 115 of 203 [56.7%]; RR 0.80, 95% CI 0.65–0.98; P=.026; number needed to treat to benefit 9; 95% CI 5–70). Nulliparous women were less likely to require oxytocin infusion when admission occurred in the morning compared with the evening (morning admission 92 of 164 [56.1%] compared with evening admission 135 of 201 [67.2%]; RR 0.83, 95% CI 0.70–0.98; P=.029; number needed to treat to benefit 9; 95% CI 5–72). Nulliparous women with a Bishop's score of 0–3 on admission were less likely to require oxytocin infusion when admission occurred in the morning compared with the evening (morning admission 59 of 102 [57.8%] compared with evening admission 84 of 119 (70.6%); RR 0.82, 95% CI 0.67–1.00; P=.048; number needed to treat to benefit 8; 95% CI 4–715).
Women with a Bishop's score of 0–3 on admission had a shorter induction to birth interval when admission occurred in the morning compared with the evening (Morning admission median 19.3hours; interquartile range 6.3–32.3 hours compared with Evening admission median 22.8 hours; interquartile range 10.9–34.7 hours; P=.043). Nulliparous women with a Bishop's score of 0–3 on admission had a shorter induction to birth interval when admission occurred in the morning compared with the evening (Morning admission median 23.3 hours; interquartile range 9.2–37.4 hours compared with Evening admission median 25.5 hours; interquartile range 14.9–36.1 hours; P=.042).
There were no other statistically significant differences identified for secondary outcomes related to evidence of effect (Table 3).
Instrumental vaginal birth was significantly less likely for women with a Bishop's score of 4–6 when admission occurred in the morning compared with the evening (Morning admission 12/103 (11.7%) compared with Evening admission 31/137 (22.6%); RR 0.51 95% CI 0.28–0.95; P=.028; number needed to treat to benefit 10; 95% CI 5–62) (Table 4). Instrumental vaginal birth was significantly less likely for nulliparous women with a Bishop's score of 4–6 when admission occurred in the morning compared with the evening (Morning admission 10/62 (16.1%) compared with Evening admission 28/82 (34.2%); RR 0.47 95% CI 0.25–0.90; P=.015; number needed to treat to benefit 6; 95% CI 4–24) (Table 4). There were no other statistically significant differences identified for secondary outcomes related to labor and birth complications (Table 4).
There were no statistically significant differences between the time of admission for induction of labor and the occurrence of either maternal or infant complications (Table 5).
Overall, women were satisfied with the care they received, although women admitted in the evening disliked the lack of sleep associated with the induction of labor protocol (morning admission 1 of 280 [0.4%] compared with evening admission 15 of 340 [4.4%]; RR 0.08, 95% CI 0.01–0.61; P=.015).
In our institution, women have been admitted to hospital to commence induction of labor in the evening, in the expectation that birth will occur during daylight hours. This time of birth may be more convenient for caregivers, and has implications for staffing levels within maternity units. However, this is in contrast to the documented diurnal variation in onset of labor, which consistently demonstrate a peak in spontaneous birth in the evening and early hours of the morning.1–5
The results of our trial indicate that admission and commencing the induction process in the morning was not associated with any differences in the number of women who did not achieve vaginal birth within 24 hours, uterine hyperstimulation with associated fetal heart rate changes, cesarean delivery, or cesarean delivery for fetal distress, when compared with admission in the evening.
Women admitted in the morning for induction of labor required less oxytocin, had a shorter induction to birth interval, and were less likely to require instrumental vaginal birth compared with evening admission. It is unclear whether these differences reflect physiologic variations in diurnal rhythm in the onset of labor and timing of spontaneous birth or they are indicative of diurnal variations in clinical practice. With women admitted in the morning for cervical ripening and induction, the onset of labor would be expected to occur in the afternoon or evening. Staffing levels within maternity units tend to be lower in the evening, and there may be a general reluctance to manage labor actively at this time, indicated by fewer women receiving an oxytocin infusion when admitted in the morning. Conversely, women admitted for induction in the evening would be expected to labor and give birth during daylight hours, a time when midwifery staffing levels are greater. During this time there is greater availability of both junior and consultant medical staff, and in this setting, the increased use of oxytocin and operative vaginal birth may reflect a degree of impatience and pressure to have women give birth at a time more convenient for caregivers.
Our finding of an increase in operative vaginal birth among women whose induction commenced in the evening is consistent with the findings of another, smaller trial in which time of admission for induction of labor with vaginal prostaglandin E2 gel was randomly determined.20 This trial was conducted in the Netherlands, and recruited 126 women, with the study outcomes relating primarily to time of birth and women's preferences for time of admission. Women who were admitted in the evening for induction of labor were more likely to require instrumental vaginal birth (RR 4.2, 95% CI 1.4–13.0) when compared with women admitted for induction in the morning.
Our trial identified no significant differences in other labor and birth outcomes between morning and evening admission for induction. There was a reduction observed in the occurrence of major postpartum hemorrhage (defined as blood loss greater than 1,000 mL) and need for blood transfusion when admission occurred in the morning, although these differences did not reach conventional levels of statistical significance. These findings are consistent with the increased use of oxytocin in women admitted in the evening.
Birth during daylight hours may be beneficial in terms of neonatal outcomes, because there have been several reports documenting an increase in the risk of early neonatal mortality, particularly related to asphyxial causes, for infants who are born during the night when compared with those infants born during the day.21–25 Using the Swedish Birth Registry, Luo and Karlberg24 reviewed over two million births between 1973 and 1995 and assessed the effect of time of birth on infant outcomes. In a similar study, Gould and colleagues25 reviewed in excess of three million infants using linked birth–death certificate data in California. Both studies identified an increase in neonatal mortality with birth in the evening when compared with birth during the day. For the rare, but serious neonatal complications, such as death and neonatal acidosis, our study was underpowered to detect all but large differences between times of starting induction of labor. Although there is a recognized risk of adverse neonatal outcome associated with birth during the evening and night,21–25 it would be necessary to recruit tens of thousands of women and their infants to assess this adequately in randomized trials, a highly unlikely scenario to be achieved.
Overall, women were satisfied with the care they received and their induction, labor, and birth. Expression of likes and dislikes during the induction process did not differ between morning and evening admission, except predictably, more women in the evening admission group did not like the interruptions to sleep that were associated with the induction protocol, involving the administration of oral misoprostol every 2 hours. However, women admitted in the morning did not identify lack of sleep associated with active labor in the evening as a dislike. Although women involved in the study by Oei and colleagues20 indicated a preference for admission to occur in the morning, the authors concluded that overall, there was no benefit in starting the induction process in the morning compared with the evening.
Admission in the morning to start induction of labor has advantages for women, with fewer women needing an oxytocin infusion during labor, fewer women requiring instrumental vaginal birth, and a shorter induction-to-birth interval when compared with admission in the evening. Although there were no differences in degree of maternal satisfaction with their induction, labor, and birth, more women in the evening admission group disliked the interruptions overnight and lack of sleep associated with the trial protocol. For women who require induction of labor, consideration should be given to admission in the morning. This has implications for midwifery and medical staffing levels.
1. Fraser WD, McLean FH, Usher RH. Diurnal variation in admission to hospital of women in labour. Can J Surg 1989;32:33–5.
2. Cagnacci A, Soldani R, Melis GB, Volpe A. Diurnal rhythms of labor and delivery in women: modulation by parity and seasons. Am J Obstet Gynecol 1998;178:140–5.
3. Mancuso PJ, Alexander JM, McIntire DD, Davis E, Burke G, Leveno KJ. Timing of birth after spontaneous onset of labor. Obstet Gynecol 2004;103:653–6.
4. Cooperstock M, England JE, Wolfe RA. Circadian incidence of labor onset hour in preterm birth and chorioamnionitis. Obstet Gynecol 1987;70:852–5.
5. Lindow SW, Jha RR, Thompson JW. 24 hour rhythm to the onset of preterm labour. BJOG 2000;107:1145–8.
6. Runnebaum B, Rieben W, Bierwirth-von Zander J. Circadian variations in plasma progesterone in the luteal phase of the menstrual cycles and during pregnancy. Acta Endocrinol (Copenh) 1972;69:731–8.
7. Junkermann H, Mangold H, Vecsei P, Runnebaum B. Circadian rhythm of serum progesterone levels in human pregnancy and its relation to the rhythm of cortisol. Acta Endocrinol (Copenh) 1982;101:98–104.
8. Walsh SW, Ducsay CA, Novy MJ. Circadian hormonal interactions among the mother, fetus, and amniotic fluid. Am J Obstet Gynecol 1984;150:745–53.
9. Challis JR, Patrick JE, Campbell K, Natale R, Richardson B. Diurnal changes in maternal plasma oestrone and oestradiol at 30 to 31, 34 to 35 and 38 to 39 weeks gestational age. Br J Obstet Gynaecol 1980;87:983–8.
10. Reck G, Renner A, Breckwoldt M. Diurnal variations in unconjugated oestriol levels during early pregnancy and their relation to maternal cortisol levels. Arch Gynecol Obstet 1988;243:169–77.
11. Bernstein D, Zer J, Zakut H. Diurnal variations in unconjugated plasma and total urinary estriol levels in late normal pregnancy. Gynecol Endocrinol 1989;3:99–106.
12. Lindow SW, Newham A, Hendricks MS, Thompson JW, van der Spuy ZM. The 24-hour rhythm of oxytocin and beta-endorphin secretion in human pregnancy. Clin Endocrinol (Oxf) 1996;45:443–6.
13. Hirst JJ, Haluska GJ, Cook MJ, Hess DL, Novy MJ. Comparison of plasma oxytocin and catecholamine concentrations with uterine activity in pregnant rhesus monkeys. J Clin Endocrinol Metab 1991;73:804–10.
14. Hirst JJ, Haluska GJ, Cook MJ, Novy MJ. Plasma oxytocin and nocturnal uterine activity: maternal but not fetal concentrations increase progressively during late pregnancy and delivery in rhesus monkeys. Am J Obstet Gynecol 1993;169:415–22.
15. Fuchs AR, Fuchs F, Husslein P, Soloff MS. Oxytocin receptors in the human uterus during pregnancy and parturition. Am J Obstet Gynecol 1984;150:734–41.
16. Fuchs AR, Behrens O, Liu HC. Correlation of nocturnal increase in plasma oxytocin with a decrease in plasma estradiol/progesterone ratio in late pregnancy. Am J Obstet Gynecol 167:1559–63.
17. Honnebier MB, Myers T, Figueroa JP, Nathanielsz PW. Variation in myometrial response to intravenous oxytocin administration at different times of the day in the pregnant rhesus monkey. Endocrinology 1989;125:1498–503.
18. Dodd JM, Crowther CA, Robinson JS. Oral misoprostol for the induction of labour at term: a randomised controlled trial. BMJ 2006;332:509–13.
19. Hofmeyr GJ, Alfirevic Z, Kelly T, Kavanagh J, Thomas J, Brocklehurst P. Methods for cervical ripening and labour induction in late pregnancy generic protocol. In: The Cochrane Library, Issue 1. Chichester (UK): John Wiley & Sons; 2005.
20. Oei SG, Jongmans L, Mol BW. Randomized trial of administration of prostaglandin E2 gel for induction of labor in the morning or the evening. J Perinat Med 2000;28:20–5.
21. Chalmers J, Shanks E, Paterson S, McInneny K, Baird D, Penney G. Scottish data on intrapartum related deaths are in same direction as Welsh data. BMJ 1998;317:539–40.
22. Stewart JH, Andrews J, Cartlidge PH. Numbers of deaths related to intrapartum asphyxia and timing of birth in all Wales perinatal survey, 1993–5. BMJ 1998;316:657–60.
23. Heller G, Misselwitz B, Schmidt S. Early neonatal mortality, asphyxia related deaths, and timing of low risk births in Hesse, Germany, 1990-8: observational study. BMJ 2000;321:274–5.
24. Luo ZC, Karlberg J. Timing of birth and infant and early neonatal mortality in Sweden 1973-95: longitudinal birth register study. BMJ 2001;323:1327–30.
25. Gould JB, Qin C, Chavez G. Time of birth and the risk of neonatal death. Obstet Gynecol 2005;106:352–8.
Figure. No caption available.