Misoprostol for the Induction of Labour at Term
FOR THE
INDUCTION OF LABOUR AT TERM
Jodie Michele Dodd
M.B.B.S., F.R.A.N.Z.C.O.G., C.M.F.M.
Thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy, March 2005 Department of Obstetrics & Gynaecology Faculty of Health Sciences The University of Adelaide
TABLE OF CONTENTS
TABLE OF CONTENTS ...... 2 LIST OF TABLES...... 6 LIST OF FIGURES...... 8 GLOSSARY OF TERMS...... 9 ABSTRACT...... 10 DECLARATION ...... 12 ACKNOWLEDGEMENTS ...... 13 AUTHOR’S CONTRIBUTION...... 15
1. LITERATURE REVIEW ...... 16 1.1 INDUCTION OF LABOUR – AN OVERVIEW...... 16 1.1(1) Introduction...... 16 1.1(2) Historical Perspective...... 16 1.1(3) Prostaglandins ...... 17 1.1(4) Role of Prostaglandins in the Initiation of Labour...... 18 1.1(5) The Role of Prostaglandins in Cervical Ripening...... 21 1.1(6) Therapeutic use of Prostaglandins to Induce Labour...... 22 1.1(7) The Cochrane Systematic Reviews of Induction of Labour ...... 23 1.2 MISOPROSTOL...... 25 1.2(1) Chemistry and Development...... 25 1.2(2) Stability ...... 26 1.2(3) Metabolism and Pharmacokinetics ...... 26 1.2(4) Toxicity and Teratogenicity ...... 27 1.3 MISOPROSTOL FOR THE INDUCTION OF LABOUR...... 28 1.3(1) Introduction...... 28 1.3(2) Research Methodology...... 32 1.3(3) Misoprostol for Induction of Labour – Results from the current Cochrane Systematic Reviews ...... 33 1.3(4) Uterine Hyperstimulation...... 36 1.3(5) Limitations of the Current Clinical Trials of Misoprostol ...... 39 1.3(6) Essential Features of Future Misoprostol Clinical Trials...... 42
2. ORAL MISOPROSTOL VERSUS VAGINAL PROSTAGLANDIN E2 FOR THE INDUCTION OF LABOUR AT TERM – A SYSTEMATIC REVIEW AND META-ANALYSIS...... 44 2.1 INTRODUCTION ...... 44 2.2 METHODS...... 44 2.2(1) Types of studies...... 44 2.2(2) Types of participants...... 44 2.2(3) Types of interventions...... 45 2.2(4) Types of outcome measures...... 45 2.2(5) Search strategies for identification of studies ...... 46 2.2(6) Methods of the review...... 46 2.3 RESULTS ...... 47 2.3(1) Description of Studies...... 47 2.3(2) Meta-analysis ...... 50 2.4 DISCUSSION...... 53
2 2.5 CONCLUSIONS ...... 55
3. MISOPROSTOL FOR CERVICAL RIPENING AND INDUCTION OF LABOUR IN WOMEN WITH INTACT MEMBRANES AT TERM: A RANDOMISED CONTROLLED TRIAL...... 56 3.1 INTRODUCTION ...... 56 3.2 STUDY AIMS & HYPOTHESES...... 57 3.3 METHODS...... 57 3.3(1) Trial Design...... 57 3.3(2) Inclusion Criteria...... 58 3.3(3) Exclusion Criteria...... 58 3.3(4) Randomisation Schedule...... 58 3.3(5) Blinding...... 59 3.3(6) Treatment Allocation ...... 59 3.3(7) Treatment Schedules ...... 59 3.3(8) Study Outcome Measures ...... 61 3.3(9) Data Collection...... 63 3.3(10) Statistical Analysis...... 63 3.3(11) Sample Size...... 63 3.4 RESULTS ...... 64 3.4(1) Baseline Characteristics ...... 64 3.4(2) Primary Outcomes...... 66 3.4(3) Secondary Outcomes – Evidence of Effect...... 67 3.4(4) Secondary Outcomes – Labour and Birth Complications...... 68 3.4(5) Secondary Outcomes - Neonatal Complications...... 69 3.4(6) Secondary Outcomes - Maternal Complications...... 70 3.4(7) Pre-specified Secondary Analyses – Effect of Maternal Parity and Bishop’s Score at Trial Entry...... 71 3.5 DISCUSSION...... 79 3.6 CONCLUSION ...... 83
4. WOMEN’S PREFERENCES FOR CARE ...... 85 4.1 INTRODUCTION ...... 85 4.2 STUDY AIMS AND HYPOTHESES...... 85 4.3 METHODS...... 85 4.4 RESULTS ...... 87 4.4(1) Baseline characteristics ...... 87 4.4(2) Women’s Satisfaction and Preferences for Care at Discharge...... 87 4.4(3) Women’s Likes and Dislikes During Induction of Labour at Discharge...... 88 4.4(4) Midwives' Satisfaction and Preferences for Care...... 90 4.4(5) Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum...... 90 4.4(6) Women’s Likes and Dislikes During Induction of Labour at 6 weeks Postpartum ...... 92 4.5 DISCUSSION...... 93 4.6 CONCLUSION ...... 95
5. TIME OF COMMENCING INDUCTION OF LABOUR – A NESTED RANDOMISED CONTROLLED TRIAL...... 96 5.1 INTRODUCTION ...... 96 5.2 STUDY AIMS AND HYPOTHESES...... 97
3 5.3 METHODS...... 97 5.3(1) Trial Design...... 97 5.3(2) Inclusion and Exclusion Criteria...... 98 5.3(3) Randomisation Schedule...... 98 5.3(4) Blinding...... 98 5.3(5) Treatment Allocation ...... 99 5.3(6) Treatment Schedules ...... 99 5.3(7) Study Outcome Measures ...... 100 5.3(8) Data Collection...... 100 5.3(9) Statistical Analysis...... 100 5.3(10) Sample Size...... 101 5.4 RESULTS ...... 101 5.4(1) Baseline Characteristics ...... 101 5.4(2) Time of Commencing Induction and Primary Outcomes ...... 103 5.4(3) Time of Commencing Induction - Secondary Outcomes: Evidence of Effect ...... 105 5.4(4) Time of Commencing Induction – Secondary Outcomes: Labour and Birth Complications ...... 107 5.4(5) Time of Commencing Induction – Secondary Outcomes: Neonatal Complications ...... 111 5.4(6) Time of Commencing Induction – Secondary Outcomes: Maternal Complications ...... 112 5.4(7) Time of Commencing Induction – Women’s Satisfaction and Preferences for Care at Discharge...... 113 5.4(8) Time of Commencing Induction - Women’s Likes and Dislikes at Discharge ...... 114 5.4(9) Time of Commencing Induction – Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum...... 116 5.4(10) Time of Commencing Induction - Women’s Likes and Dislikes at 6 weeks Postpartum ...... 117 5.5 DISCUSSION...... 118 5.6 CONCLUSION ...... 120
6. COST COMPARISON...... 122 6.1 INTRODUCTION ...... 122 6.2 STUDY HYPOTHESES...... 122 6.3 METHODS...... 123 6.4 RESULTS ...... 124 6.5 DISCUSSION...... 131 6.6 CONCLUSION ...... 132
7. MISOPROSTOL FOR INDUCTION OF LABOUR AT TERM – AN UPDATED META-ANALYSIS...... 133 7.1 INTRODUCTION ...... 133 7.2 STUDY AIMS AND HYPOTHESES...... 133 7.3 METHODS...... 133 7.4 RESULTS ...... 134 7.4(1) Description of Studies...... 134 7.4(2) Meta-analysis ...... 134 7.5 DISCUSSION...... 139
4 7.6 CONCLUSION ...... 140
8. OVERALL CONCLUSIONS ...... 141 8.1 ORAL MISOPROSTOL VERSUS VAGINAL PROSTAGLANDIN E2 GEL FOR INDUCTION OF LABOUR...... 141 8.2 WOMEN’S PREFERENCES FOR CARE ...... 142 8.3 TIME OF ADMISSION FOR INDUCTION OF LABOUR...... 142 8.4 COSTS ASSOCIATED WITH INDUCTION OF LABOUR...... 143 8.5 IMPLICATIONS FOR CLINICAL PRACTICE...... 144 8.6 IMPLICATIONS FOR RESEARCH...... 146
REFERENCES...... 148
5 LIST OF TABLES
Table 2.3(1)Description of Included Studies...... 48 Table 2.3(2)(i) Meta-analysis of primary outcomes ...... 51 Table 2.3(2)(ii) Meta-analysis of secondary outcomes – evidence of effect...... 51 Table 2.3(2)(iii) Meta-analysis of secondary outcomes – maternal complications...... 52 Table 2.3(2)iv Meta-analysis of secondary clinical outcomes – neonatal complications ...... 52
Table 3.3(8)(ii) Secondary outcomes...... 62 Table 3.4(1) Comparison of baseline variables at trial entry...... 65 Table 3.4(2) Primary Outcomes...... 66 Table 3.4(3) Secondary Outcomes – Evidence of Effect...... 67 Table 3.4(4) Secondary Outcomes – Labour & Birth Complications ...... 69 Table 3.4(5) Secondary Outcomes - Neonatal Complications ...... 70 Table 3.4(6) Secondary Outcomes - Maternal Complications ...... 71 Table 3.4(7)(i) Distribution of maternal parity and initial Bishop’s score ...... 72 Table 3.4(7)(ii) Secondary Analysis – Primary Outcomes...... 73 Table 3.4(7)(iii) Secondary Analysis – Evidence of Effect...... 75 Table 3.4(7)(iv) Secondary Analysis – Labour & Birth Complications...... 76 Table 3.4(7)(vi) Secondary Analysis – Maternal Complications ...... 79
Table 4.4(2) Women’s Satisfaction and Preferences for Care at Discharge ...... 88 Table 4.4(3) Women’s Likes and Dislikes During Induction of Labour at Discharge...89 Table 4.4(4) Midwives’ Satisfaction and Preferences for Care...... 90 Table 4.4(5) Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum ..91 Table 4.4(6) Women’s Likes and Dislikes During Induction of Labour at 6 weeks Postpartum ...... 93
Table 5.4(1) Distribution of Time of Commencing Induction ...... 103 Table 5.4(2) Time of Commencing Induction - Primary Outcomes...... 104 Table 5.4(3) Time of Commencing Induction - Secondary Outcomes: Evidence of Effect...... 106 Table 5.4(4) Time of Commencing Induction – Secondary Outcomes: Labour and Birth Complications ...... 108 Table 5.4(5) Time of Commencing Induction – Secondary Outcomes: Neonatal Complications ...... 111 Table 5.4(6) Time of Commencing Induction – Secondary Outcomes: Maternal Complications ...... 113 Table 5.4(7) Time of Commencing Induction of Labour – Women’s Satisfaction and Preferences for Care at Discharge ...... 114 Table 5.4(8) Time of Commencing Induction - Women’s Likes and Dislikes at Discharge ...... 115 Table 5.4(9) Time of Commencing Induction – Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum ...... 116 Table 5.4(10) Time of Commencing Induction - Women’s Likes and Dislikes at 6 weeks Postpartum ...... 118
6 Table 6.3(1) Costs to the Women’s and Children’s Hospital for Induction of Labour 124 Table 6.4(1) Costs for all women ...... 125 Table 6.4(2) Costs nulliparous women with Bishop’s score 0-3...... 126 Table 6.4(3) Costs for nulliparous women with Bishop’s score 4-6 ...... 127 Table 6.4(4) Costs for multiparous women with Bishop’s score 0-3 ...... 128 Table 6.4(5) Costs for multiparous women with Bishop’s score 4-6 ...... 129
Table 7.4(1) Summary characteristics of the current randomised trial...... 134 Table 7.4(2)(i) Meta-analysis of Primary Outcomes...... 135 Table 7.4(2)(ii) Meta-analysis of Secondary Outcomes – Evidence of Effect...... 137 Table 7.4(2)(iii) Meta-analysis of Secondary Outcomes - Maternal Complications....138 Table 7.4(2)(iv) Meta-analysis of Secondary Outcomes – Neonatal Complications ...138
7 LIST OF FIGURES
Figure 1.1(3) Prostaglandin structure and nomenclature...... 18 Figure 1.2(1)(i) Structure of misoprostol and prostaglandin E1 ...... 25 Figure 1.2(1)(ii) Development of misoprostol...... 26 Figure 1.3(5)(i) Absorption profile of oral misoprostol tablets and misoprostol solution ...... 41
Figure 3.4(1) Trial Flow Chart ...... 64
Figure 5.4(1) Trial Flow Chart ...... 102
Figure 6.4(1) Drug Costs per Woman Induced...... 130
Figure 7.4(2)(i) Vaginal birth not achieved in 24 hours...... 136 Figure 7.4(2)(ii) Uterine hyperstimulation with associated fetal heart rate changes....136 Figure 7.4(2)(iii) Caesarean section...... 136 Figure 7.4(2)(iv) Cervix unchanged after 12/24 hours...... 137 Figure 7.4(2)(v) Need for oxytocin infusion ...... 137
8 GLOSSARY OF TERMS
PG prostaglandin HPMC hydroxy propyl methyl cellulose HPLC high pressure liquid chromatography RIA radio immune assay mg milligrams mcg micrograms mL millilitres RCT randomised controlled trial RR relative risk OR odds ratio NNTH number needed to treat to harm 95% CI 95% confidence intervals IOL induction of labour GA gestational age ARM artificial rupture of membranes SROM spontaneous rupture of membranes PROM premature rupture of membranes PE pre-eclampsia HT hypertension IUGR intra-uterine growth restriction APH antepartum haemorrhage CTG cardiotocograph MSL meconium stained liquor EDB epidural block NVD normal vaginal delivery CS caesarean section PPH postpartum haemorrhage NICU neonatal intensive care unit
9 ABSTRACT
Background:
The aims of this randomised, double blind, placebo controlled trial were to compare vaginal PGE2 gel with oral misoprostol in the induction of labour at term.
Methods:
Women randomised to the oral misoprostol group received 20mcg oral misoprostol solution at two hourly intervals and placebo vaginal gel, and those in the vaginal prostaglandin group received vaginal PGE2 gel at six hourly intervals and oral placebo solution.
The primary outcome measures were vaginal birth not achieved in 24 hours, uterine hyperstimulation with associated fetal heart rate changes, and caesarean section.
Women were asked about their preferences for care, and a cost comparison was performed for the two methods of induction of labour. A nested randomised trial compared health outcomes for the woman and her infant related to morning or evening admission for commencing induction of labour.
Results:
A total of 741 women were randomised, 365 to the misoprostol group and 376 to the vaginal PGE2 group.
There were no differences between women in the oral misoprostol group and women in the vaginal PGE2 group, for the outcomes vaginal birth not achieved in 24 hours
(Misoprostol 168/365 (46.0%) versus PGE2 155/376 (41.2%); RR 1.12 95% CI 0.95-
1.32; p=0.134), caesarean section (Misoprostol 83/365 (22.7%) versus PGE2 100/376 (26.6%); RR 0.82 95% CI 0.64-1.06; p=0.127), or uterine hyperstimulation with fetal
10 heart rate changes (Misoprostol 3/365 (0.8%) versus PGE2 6/376 1.6%); RR 0.55 95% CI 0.14-2.21; p=0.401).
Women in the misoprostol group were more likely to indicate that they “liked everything” associated with their labour and birth experience compared with women in the vaginal PGE2 group (Misoprostol 126/362 (34.8%) versus PGE2 103/373 (27.6%); RR 1.26; 95% CI 1.02-1.57; p=0.036).
There were no differences in the primary outcomes when considering morning or evening admission to commence induction.
The use of misoprostol was associated with a saving of $110.83 per woman induced.
Conclusions:
The use of oral misoprostol in induction of labour does not lead to poorer health outcomes for women or their infants, women express greater satisfaction with their labour and birth experience, and with misoprostol induction there is a cost saving to the institution.
11 DECLARATION
This work contains no material which has been accepted for the award of any other degree or diploma in any other university or tertiary institution, and, to the best of my knowledge and belief, contains no material previously published or written by another person, except where due reference has been made in the text.
I give consent to a copy of my thesis, when deposited in the University Library, being available for loan and photocopy.
Jodie Dodd March 2005
12 ACKNOWLEDGEMENTS
I must begin by thanking the 741 women who have selflessly participated in this clinical research project, in the hope of improving health care for women and their infants.
I first encountered my supervisors Professors Jeffrey Robinson and Caroline Crowther, as a 4th year medical student. Their enthusiasm for research and clinical obstetrics inspired me then, and more recently have shaped my career path. I am extremely grateful for the opportunity to have them both as mentors and colleagues, and hope that in the future I might emulate the high standards they set as clinicians, researchers, and teachers.
A project like this doesn’t happen without the assistance and support of many, and I would like to say an enormous thank you to the following people. • Sheree Agett (research midwife) – for her fantastic efforts with recruitment, data collection, follow-up of postnatal questionnaires, and the preparation of treatment packs. • Judy Coffey and the midwifery team on delivery suite – for their incredible support, without which this trial would not have been possible. • The midwives in antenatal clinic, and the obstetric medical staff – for promoting the trial in a positive way to eligible women. • Professor Gus Dekker and the staff at the Lyell McEwin Health Service, and Dr Dirk Ludwig and the staff of the Hervey Bay Hospital, for their support and involvement in the trial. • Kristyn Willson – for her statistical assistance, advice, and patience in dealing with a novice in the realms of statistical programming. • Staff in the Maternal & Perinatal Clinical Trials Unit – for their support and willingness to share the highs and lows that life has to offer. • The Department of Obstetrics & Gynaecology, The University of Adelaide –for the opportunity to study in such a great department (and for the desk with the wonderful view across the park).
13 Any research project requires funding and I would like to thank the following institutions for their financial assistance. • The Department of Obstetrics and Gynaecology, The University of Adelaide for financial assistance and provision of support services. • The Women’s and Babies Division of the Women’s and Children’s Hospital, for meeting the cost of drugs. • The South Australian & Northern Territory branch of the Royal Australian and New Zealand College of Obstetricians and Gynaecologists for their generous donation of $2,500 for consumable items. • Mayne Health for the provision of the Mayne Women’s Health Fellowship, through the Research Foundation of the Royal Australian and New Zealand College of Obstetricians and Gynaecologists. • The Women’s and Children’s Hospital Research Foundation for the provision of a Postgraduate Medical Scholarship, and Project Grant to facilitate research into women’s and infant’s health.
There are a few more people who have been instrumental in shaping the person I am, and whose love and support is immeasurable. • Thank you to my family and friends – your support means everything. • Thank you to Mum, Dad and Nana – you have always supported me in everything I have done and for this I am so grateful. I hope that you are proud of my achievements, and of the person you have helped me to become. • Thank you to Greg – you are my best friend, my lover and my soul mate. I thank you for your unending belief in me and in my ability, for your constant support and encouragement, and for being the person who makes life so worthwhile.
14 AUTHOR’S CONTRIBUTION
I have been responsible for the development of the original protocols, submission of these protocols to Research and Ethics Committees, and obtaining funding for the project. I have devised the information sheets and data sheets, and have coordinated in- service education sessions for midwifery and medical staff. I have been involved in the preparation of treatment packs, recruitment at the Women’s and Children’s Hospital, data collection at that site, checking of all data forms and data entry. I have received statistical advice and software programme assistance from Kristyn Willson, but the interpretation and any errors therein are my responsibility.
15 1. LITERATURE REVIEW
1.1 INDUCTION OF LABOUR – AN OVERVIEW
1.1(1) Introduction
Induction of labour is a common obstetric intervention, performed when the perceived risks to the mother or fetus associated with continuation of the pregnancy are greater than those associated with birth (Riskin-Mashiah 1999). Labour may be induced for medical or obstetric indications (such as hypertensive conditions, impaired glucose tolerance, prolonged pregnancy, intrauterine growth restriction), or for the convenience of mother or obstetrician (so called “social” indications).
In Australia, 26.6% of pregnant women had their labour induced in 2002 (Laws 2004). In South Australia 29.3% of pregnancies were induced for the same year (Laws 2004), with the rate of induction remaining relatively stable over the past 10 years (Chan 1999). It is estimated that prostaglandins are used in approximately 22.5% of all confinements to induce labour (Nassar 2000).
1.1(2) Historical Perspective
An early literature record of induction of labour dates to the 1500’s, with a mixture of juniper berries, cinnamon, and castor oil recommended to expedite birth (Calder 1997). During the 18th century, the concept of induction of labour was to effect early birth for women with a contracted pelvis (O'Dowd 1994), with midwives aware of the use of ergot as an induction agent (Davis 1935). Ergot is a fungal agent that principally attacks rye and other grains, but when eaten prompted gastrointestinal disturbances (Davis 1935). John Stearns reported the use of ergot in the medical literature in 1808, where a preparation was mixed with boiling water and administered to labouring women, thereby hastening the onset of contractions (Stearns 1808). Its use was often associated with marked gastrointestinal symptoms (Stearns 1808), and an increase in the occurrence of stillbirth caused its use to fall into disrepute (O'Dowd 1994). Almost a century later, Henry Dale noted that a preparation from ox pituitary tissue initiated
16 uterine contractions in a pregnant cat (Dale 1906), with Blair Bell subsequently describing the clinical use of pituitary extract in pregnant women for the management of “normal labour”, and uterine atony post-partum (Bell 1909). Davis and colleagues went on to extract the active component from ergot alkaloids, further demonstrating the compound’s uterotonic effects (Davis 1935), with du Vigneaud later identifying both oxytocin and vasopressin in pituitary extracts, allowing purification and the synthetic preparation of Syntocinon in 1953 (du Vigneaud 1953). While initial oxytocin preparations were crude and unreliable in effect, the development of Syntocinon allowed intravenous titration to produce a more predictable and effective method of induction of labour.
The state of the cervix relates directly to the duration of pregnancy, as described by Bishop, who published a quantitative pelvic scoring system based on 500 consecutive vaginal examinations performed on multiparous women who entered labour spontaneously after 36 weeks gestation (Bishop 1964). Subsequent clinical studies have shown correlation between outcome of induction of labour and cervical score, with the use of amniotomy and oxytocin infusion in the presence of an “unfavourable” cervix (as indicated by a low Bishop score), being associated with a longer interval to delivery and a high incidence of caesarean section (Calder 1974; Calder 1975; MacKenzie 1978). This has resulted in the search for more effective agents in the induction of labour designed to ‘ripen’ the cervix, and reduce the risk of caesarean section due to a “failed” induction of labour.
1.1(3) Prostaglandins
The existence of prostaglandins was established through the observations of Kurzrok and Lieb in 1930, who noted the ability of human semen to initiate uterine contractions (Kurzrok 1930). After isolating substances from the male prostate gland shown to initiate contractions of uterine tissue, von Euler named the compounds “prostaglandins”, believing them to be a unique secretion from that gland (von Euler 1936). It was not until later that prostaglandins were recognised as a large family of substances ubiquitous in mammalian biology (Collins 1990; Calder 1997).
17 Prostaglandins consist of 20-carbon fatty acid molecules derived from dietary fatty acids, and consisting of a pentane ring with a fatty acid moiety attached at two adjacent carbons (Elattar 1978; Embrey 1981; Garris 1989; O'Brien 1995) (Figure 1.1(3)). Prostaglandins are classified into six groups (A to F), according to the structure or substitution of the pentane ring. The numbering sequence begins at the carboxylic end and proceeds around the ring to the terminal carbon (Elattar 1978; Garris 1989). The subscript numeral (1,2,3) indicates the number of double bonds in the fatty acid backbone (Elattar 1978; Garris 1989; O'Brien 1995). For example, prostaglandins of the
E series consist of a ketone group at the C9 position, and a hydroxyl group at the C11 position (Elattar 1978; Garris 1989). Prostaglandin E1 has a trans double bond between
C13 and C14, while prostaglandin E2 has an additional cis double bond at the C5-C6 position (Elattar 1978; Collins 1985).
Figure 1.1(3): Prostaglandin structure and nomenclature (Collins 1985)
1.1(4) Role of Prostaglandins in the Initiation of Labour
The initiation of labour in sheep has been well studied, with a role for the fetal pituitary- adrenal system established (Liggins 1967; Liggins 1968; Liggins 1973). These findings
18 have not been replicated in women, with a local membrane mechanism controlling the production of prostaglandins being favoured (Liggins 1977).
The presence of prostaglandins E and F has been demonstrated in amniotic fluid samples obtained from women at term with spontaneous rupture of the membranes and shown to cause muscle obtained from the pregnant uterus to contract (Karim 1966), while amniotic fluid obtained in early pregnancy induces a relaxant effect (Karim
1967). Prostaglandin F2α and its metabolite can be detected in blood prior to the onset of uterine contractions (Karim 1969), and exert a stimulatory effect on myometrial contractility in vivo following intravenous infusion to women in both early and late pregnancy (Embrey 1969).
The concentrations of prostaglandins in amniotic fluid increase with the onset of labour (Keirse 1973; Hibbard 1974; Hillier 1974; Keirse 1974; Dray 1976; Keirse 1977; Mitchell 1977; Mitchell 1978; Whalen 1978), with increasing cervical dilatation (Keirse 1973; Keirse 1974; Keirse 1977), and have been shown to be lower following induction of labour when compared with spontaneous onset of labour (Hillier 1974; Keirse 1974; Willman 1976a; Husslein 1981). Levels are also lower where there is poor progression of cervical dilatation (Keirse 1977; Reddi 1987). Accurate estimates of prostaglandin concentrations in amniotic fluid samples following artificial rupture of the membranes is difficult, as the procedure itself has been documented to generate an increase in prostaglandin production (Mitchell 1976; Keirse 1977; Mitchell 1977). The increasing PGF concentrations seen from 36 weeks gestation in amniotic fluid obtained following ARM, were not demonstrated when samples at similar gestation were obtained from amniocentesis, suggesting a role for the local control of production of these compounds in the pregnant uterus (Mitchell 1976).
Prostaglandin concentrations have been measured in samples of umbilical cord, placental, decidual and myometrial tissue, and amnion and chorion, with a greater concentration of PGE2 compared with PGF2α in all tissues studied (Willman 1976b). At term, the concentration of PGE2 increased most in decidual tissue with the onset of labour, being greater again when labour occurred spontaneously (Willman 1976a). The distribution and concentration of enzymes regulating prostaglandin degradation within
19 the pregnant uterus reflect this, being highest in the fetal membranes and placenta, followed by the myometrium and decidua, with the lowest levels being found in the umbilical cord (Keirse 1975). No significant prostaglandin metabolism has been demonstrated in the amniotic fluid (Keirse 1975). An increase in prostaglandin production in the decidua and amnion has been documented following treatment with oxytocin (Fuchs 1981), as has oxytocin induced release of arachidonic acid (the precursor required for prostaglandin synthesis) from decidual cells obtained from women in labour at term (Wilson 1988).
For prostaglandins to exert a biological effect, there must be interaction with receptors specific to the type of prostaglandin, as determined by the structure of the pentane ring. However, there is no differentiation between molecules that differ in their fatty acid backbone, and both prostaglandin E1 and E2 will interact with the same PGE receptor (O'Brien 1995).
The receptor site specific for prostaglandins of the E series have been designated EP receptors, and those interacting with prostaglandins of the F series, FP receptors (Senior
1993). The EP receptor has been further subdivided into subclasses EP1 to EP3 (Senior 1993). Myometrial samples obtained from caesarean sections performed between 38 and 40 weeks gestation show the greatest concentration of EP and FP receptors at the uterine fundus, with a decrease in receptor concentration towards the cervix, corresponding with a decrease in the content and distribution of uterine smooth muscle
(Senior 1993). Stimulation of the FP receptor by PGF2α has been shown to initiate a purely contractile response of the myometrium, while stimulation of the EP receptor by
PGE2 is stimulatory at low doses and inhibitory at higher doses (Senior 1993).
Stimulation of EP2 receptors produces a relaxant effect on the myometrium when stimulated by high concentrations of PGE2, while the EP3 receptor initiates a contractile response when stimulated by low doses of PGE2 (Senior 1993). Misoprostol an EP1 and
EP3 receptor agonist, initiates a qualitative response similar to PGE2 in that contractions are initiated at lower doses in a dose response fashion, while the effect at higher doses is inhibitory (Senior 1993).
20 Prostaglandin receptors are present in significant numbers in both the pregnant and non- pregnant uterus, rendering it sensitive to prostaglandin stimulation throughout pregnancy (O'Brien 1995). In contrast, oxytocin receptor concentration in the myometrium increases between 100 and 200 fold as gestation advances (Fuchs 1984; O'Brien 1995), reaching a maximum during early labour (Fuchs 1984). This increase in receptor concentration is paralleled by an increase in uterine sensitivity to oxytocin during the second half of gestation (Fuchs 1984). Oxytocin may play a dual role in parturition, activating myometrial receptors resulting in uterine contraction, and also stimulating prostaglandin production in decidual tissue expressing oxytocin receptors (Fuchs 1984).
1.1(5) The Role of Prostaglandins in Cervical Ripening
The historical view of the cervix as a muscular structure with sphincteric properties was challenged by Danforth, whose study of pregnant and non-pregnant surgical specimens indicated the cervix to be composed of fibrous connective tissue, interspersed with small amounts of smooth muscle (Danforth 1947). Cervical specimens obtained from non-pregnant women demonstrate a primarily collagenous connective tissue component, the bundles woven tightly together and held with a “cementing substance”, while those obtained from women in late pregnancy revealed increased tissue oedema, with spaces between the individual collagen fibres, and increased diameter of individual fibres (Danforth 1960). When compared with the non-pregnant cervix, the postpartum cervix has an increased water, glycosaminoglycan, and hyaluronic acid content (Danforth 1974), in addition to a reduction in the concentration of dermatan and chondroitan sulfate, and fewer bridges between collagen fibrils, resulting overall in a loose and disorganised network of fibres (Uldberg 1983). An increase in the local production of prostaglandins from cervical tissue obtained during late pregnancy suggests a role for the cervical softening and dilatation observed during late pregnancy and labour (Ellwood 1980). Changes to the structure of the cervix may be further mediated by an inflammatory process, in which infiltrating macrophages and neutrophils release proteolytic enzymes including collagenase (Liggins 1981; Kelly 1994). The role of prostaglandins in this process may involve initiation of a vascular response and the infiltration of leukocytes.
21 1.1(6) Therapeutic use of Prostaglandins to Induce Labour
Natural prostaglandins are unsuitable for therapeutic use, by virtue of their rapid metabolism (as indicated by a short duration of action when given parenterally and a lack of activity when given orally); numerous side effects (including facial flushing, headaches, hyperthermia, hypotension, shivering, and the more commonly observed gastrointestinal symptoms of nausea, vomiting, and diarrhoea); and chemical instability at room temperature (Collins 1990).
Prostaglandins were first used clinically to induce labour in the late 1960’s (Karim 1968), with subsequent administration of both E and F series prostaglandins to induce labour via intravenous infusion at or near term (Bygdeman 1968; Karin 1968; Karim 1969; Embrey 1970; Karim 1970a), or to induce first trimester abortion (Embrey 1970; Karim 1970b; Wiqvist 1970). While oral prostaglandins have been administered to induce labour, their use was limited by the frequent occurrence of side effects reflecting the high doses required to produce clinical effect (Karim 1971; Craft 1972; Gordon- Wright 1979). In an attempt to reduce side effects, local vaginal application of prostaglandins became the route of administration of choice, initially involving extra- amniotic placement (Calder 1974; Shepherd 1976; Calder 1977), although similar results (in terms of effective induction of labour) were achieved with the less cumbersome approach of intravaginal PGE2 gel or pessaries (MacKenzie 1977; MacKenzie 1978; Gordon-Wright 1979; Shepherd 1979; MacKenzie 1981; Shepherd 1981).
Regardless of the route of administration of prostaglandins, be it oral, intravenous, or intravaginal, the effect in terms of cervical ripening is similar (O'Brien 1995). Local vaginal administration of PGE2 results in systemic absorption, with maximal uterine activity correlating with the peak circulatory level of the compound (O'Brien 1986; O'Brien 1995). In this regard then, the appropriate route of prostaglandin administration should be determined by both cost and ease of administration, in addition to preferences expressed by women.
22 1.1(7) The Cochrane Systematic Reviews of Induction of Labour
A standardised protocol has been developed by the Cochrane Pregnancy and Childbirth Collaborative Review Group to facilitate review of the randomised controlled trial literature concerning methods of induction of labour (Hofmeyr 2005a). The protocol includes the use of standard primary outcome measures, including vaginal birth not achieved within 24 hours (including women who birth vaginally beyond 24 hours and by caesarean section); caesarean section; uterine hyperstimulation with fetal heart rate changes; neonatal morbidity and mortality; and maternal morbidity and mortality. Uterine hyperstimulation has been defined as either uterine tachysystole (the presence of five or more contractions in a ten minute period for more than two consecutive ten minute periods) or uterine hypertonus (the presence of a uterine contraction lasting for more than two minutes) (Curtis 1987). The fetal heart rate changes considered as abnormal include fetal tachycardia (fetal heart rate of greater than 160 beats per minute); fetal bradycardia (fetal heart rate of less than 110 beats per minute); reduced variability of the fetal heart rate (less than five beats per minute); and persistent decelerations of the fetal heart rate (RANZCOG 2001; RCOG 2001).
Secondary outcome measures have also been reported in the protocol in a standardised fashion. These include evidence of clinical effect, complications, and satisfaction. Evidence of clinical effect includes the cervix remaining unchanged or unfavourable after 12 to 24 hours and need for oxytocin administration. Complications relate to both mother and infant, including uterine hyperstimulation without fetal heart rate changes; uterine rupture; use of epidural analgesia; instrumental vaginal birth; meconium stained liquor; Apgar score of less than seven at five minutes; admission to the neonatal intensive care unit; neonatal encephalopathy; perinatal death; childhood disability; maternal side effects (including nausea, vomiting, and diarrhoea); postpartum haemorrhage; serious maternal complications and maternal death. Measures of satisfaction relate to both woman and caregiver. Uterine rupture was defined as all clinically significant ruptures of scarred and unscarred uteri, while asymptomatic dehiscence of the uterine scar identified incidentally at the time of surgery was excluded.
23 Intravenous prostaglandin for the induction of labour is associated with an increase in uterine hyperstimulation, both with (5 trials; 390 participants; RR 6.76; 95% CI 1.23- 37.11) and without (5 trials; 318 participants; RR 4.25; 95% CI 1.48-12.24) fetal heart rate changes (Luckas 2005). As intravenous prostaglandins are no more efficient than oxytocin in effecting vaginal birth within 24 hours (9 trials; 990 participants; RR 0.85; 95% CI 0.61-1.18), but have both increased maternal side effects (8 trials; 940 participants; RR 3.75; 95% CI 2.46-5.70) and cost, its use in the induction of labour has been abandoned (Luckas 2005).
Oral PGE2 has been shown to be no more clinically effective than intravenous oxytocin in achieving vaginal birth within 24 hours (1 trial; 201 women; RR 3.09; 95% CI 0.13- 74.96), but more frequently associated with gastrointestinal side effects of vomiting (3 trials; 305 participants; RR 5.56; 95% CI 2.15-14.38) and diarrhoea (2 trials; 236 participants; RR 8.13; 95% CI 1.03-63.94) (French 2005). The use of oral PGE2 in the induction of labour has been replaced with vaginal prostaglandin preparations, although the availability of misoprostol has regenerated interest in an oral route of administration.
Vaginal prostaglandins are effective in the induction of labour, increasing the likelihood of vaginal birth within 24 hours of the induction process commencing (2 trials; 384 participants; RR 0.19; 95% CI 0.14-0.25), and reducing the need for oxytocin augmentation (11 trials; 1265 participants; RR 0.80; 95% CI 0.69-0.91), without increasing the rate of caesarean section (32 trials; 6243 participants; RR 0.79; 95% CI 0.79-1.00) when compared with placebo (Kelly 2005). Vehicle delivery of the preparation indicates that PGE2 gel is as efficacious as PGE2 vaginal tablets, but relatively more expensive. Vaginal prostaglandin E2 is currently the standard prostaglandin agent in most obstetric hospitals within Australia for cervical ripening and induction of labour.
24 1.2 MISOPROSTOL
1.2(1) Chemistry and Development
Misoprostol is a synthetic prostaglandin structurally related to prostaglandin E1 (PGE1), and is the generic name for (+/-) methyl, 11α, 16-dihydroxy-16-methyl-oxoprost-13E- en-1-oate, a water insoluble, oily, viscous substance (Garris 1989) (Figure 1.2(1)(i)).
Figure 1.2(1)(i): Structure of Misoprostol and Prostaglandin E1 (Collins 1990)
While natural PGE1 is an effective inhibitor of gastric acid secretion when administered intravenously, there are a number of disadvantages to its use in the therapeutic treatment of peptic ulcer disease, including a lack of oral activity, short duration of action, and side effects including rhinorrhoea, trembling, nausea, vomiting, and diarrhoea (Collins
1985). Misoprostol was developed from attempts to manipulate PGE1 to produce an orally active compound while minimising side effects.
The proposed lack of oral activity of PGE1 is thought to relate to the rapid metabolic oxidation of the C15 hydroxyl group (Collins 1985). The placement of a methyl group at the C15 position blocks this oxidative process, the resulting compound having long acting inhibition of gastric acid secretion, as well as potency when administered orally (Collins 1985). A reduction in side effects typical of prostaglandins while maintaining the antisecretory potency of PGE1 was achieved by repositioning the beta side chain from the C15 to C16 position. However, this compound was only weakly active after oral administration, and had a short duration of action. The addition of a methyl group to the
C16 position gave misoprostol, a compound with increased oral potency and duration of action (Figure 1.2(1)(ii)).
25 Figure 1.2(1)(ii): Development of misoprostol (Collins 1990)
1.2(2) Stability
Like all prostaglandins of the E series, misoprostol is chemically unstable at room temperature (Collins 1985; Karim 1987; Collins 1990; Kararli 1991). Under alkaline or acidic conditions in the presence of a small amount of water, a dehydration process occurs, with elimination of the 11-hydroxyl groups to produce an A-type prostaglandin (PGA) (Collins 1990; Chen 2000). Under similar conditions, PGA can further isomerise to a prostaglandin-B form (Collins 1990; Chen 2000). The stability of misoprostol is significantly improved by its dispersion (1:100) in hydroxy propyl methyl-cellulose (HPMC) (Kararli 1991). The presence of the HPMC matrix separates the misoprostol molecules, reducing the mobility of both misoprostol and water. Additionally, any water present preferentially interacts with the hydroxyl groups in the HPMC, sparing interaction with, and resulting in minimal degradation of misoprostol (Kararli 1991).
1.2(3) Metabolism and Pharmacokinetics
After oral administration, misoprostol is rapidly de-esterified to misoprostol acid, which is the major active metabolite in plasma (Garris 1989). This rapid conversion occurs before or during absorption, and as a result, the plasma concentration of misoprostol is virtually undetectable, even shortly after administration (Garris 1989). Misoprostol acid
26 is in turn metabolised via β and σ oxidation to inactive dinor and tetranor metabolites that are excreted renally, with the total urinary excretion of unchanged misoprostol being less than 1% of the orally administered dose (Garris 1989). The dinor metabolite has only weak antisecretory activity, while the tetranor metabolite has no substantial antisecretory activity (Garris 1989).
The in vitro half-life for the de-esterification of misoprostol incubated in human plasma at 37 degrees Celsius is 6.4 minutes, with absorption of total radioactivity occurring within 1.5 hours, and a mean elimination half-life for total plasma radioactivity of 1.7 hours (+/- 0.06) (Schoenhard 1985). In vivo, the mean peak plasma concentration of misoprostol acid is achieved within 12 minutes with an elimination half-life of 20 minutes, and on the basis of radioactive excretion in both urine and faeces, it has been estimated that an average of 88% of the oral dose of misoprostol is absorbed (Karim 1987). Misoprostol acid is 81-89% protein bound, primarily by serum albumin (Nicholson 1990; Foote 1995).
1.2(4) Toxicity and Teratogenicity
The preclinical trials assessing acute toxicity of misoprostol were performed in rats, mice and dogs, with the most common clinical signs observed being diarrhoea, emesis, tremor, and reduced motor activity (Kotsonis 1985). Most of the deaths occurred within 24 hours, while those animals surviving regained “normal” behaviour and appearance within 3 to 4 days. These studies indicated a safety margin of 500 to 1000 fold between the lethal doses observed in animals, and the therapeutic doses administered to humans. Carcinogenic and fertility properties were studied in rats, mice and rabbits, without demonstration of any effects attributable to misoprostol (Kotsonis 1985), the conclusion being that misoprostol showed no evidence of embryotoxicity, fetotoxicity, or teratogenicity (Kotsonis 1985).
27 1.3 MISOPROSTOL FOR THE INDUCTION OF LABOUR
1.3(1) Introduction
Misoprostol is licensed for use in the treatment of gastric ulcer disease, and does not have a product license for use in pregnancy in Australia, or any other country worldwide. Despite this, its use in obstetric and gynaecological practice has increased, being used widely in the management of first and second trimester abortion (Dickinson 1999), and in the third trimester of pregnancy following intrauterine fetal death (Mariani-Neto 1987). While misoprostol is used commonly in this setting, comparison of clinical outcomes with other methods of induction of labour to terminate pregnancy, the optimal route of administration, and the optimal dosing regimen have not been assessed systematically. More recently, misoprostol has been used in the induction of labour at term in the presence of a viable fetus, with both vaginal (Hofmeyr 2005b) and oral (Alfirevic 2005) routes of administration being used. There have also been trials of misoprostol as an oxytocic agent in the management of the third stage of labour (Gulmezoglu 2005).
1.3(1)(i) Safety of Misoprostol 1.3(1)(i)(a) Misoprostol and Uterine Rupture Concern has been raised regarding the use of misoprostol as an induction agent, with several case reports of uterine rupture and its consequences (including maternal or fetal death) appearing in the literature (Phillips 1996; Bennett 1997; Cunha 1999; Plaut 1999; Daisley 2000; Gherman 2000; Akhan 2001; Al-Hussaini 2001; Berghahn 2001; Khabbaz 2001). These reports have involved women with scarred and unscarred uteri, following oral and vaginal administration of misoprostol, and where the induction was performed in the presence of a live fetus and following intrauterine demise. Doses of the drug have varied from those generally considered to be low (25mcg) (Bennett 1997; Plaut 1999; Gherman 2000), up to 600mcg being used in cases of fetal demise (Phillips 1996).
Attempts have been made to address the issue of the safety of misoprostol in the presence of a uterine scar, with both Sciscione (1998) and Wing (1998) commencing randomised controlled trials. Sciscione (1998) compared 50mcg intravaginal
28 misoprostol at four hourly intervals with an intracervical Foley catheter, in which women with a prior caesarean birth were not excluded. While one woman with a scarred uterus suffered uterine rupture following misoprostol, it is not clear from the report the total number of women randomised, the number of women with a scarred uterus who were involved, or the number of women allocated to each treatment arm. Wing and colleagues specifically recruited women with a prior caesarean birth to a randomised controlled trial comparing 25mcg vaginal misoprostol at six hourly intervals with intravenous oxytocin for induction of labour (Wing 1998). Both symptomatic scar rupture and asymptomatic scar dehiscence were combined as a single outcome measure. Seventeen women were allocated to the misoprostol group, and 21 women to the oxytocin group. Two women in the misoprostol group were identified with scar disruption (11.8%), and due to concerns regarding the safety of misoprostol in this clinical setting, the trial was terminated prematurely.
The safety of misoprostol for the induction of labour in women with a previous caesarean birth is unlikely to be addressed in the short term, with only two of the trials (Chuck 1995; Webb 1997) included in the systematic reviews of vaginal misoprostol (Hofmeyr 2005b) and oral misoprostol (Alfirevic 2005) in the induction of labour involving women with a previous caesarean section.
1.3(1)(i)(b) Misoprostol and gastrointestinal stimulation Concern has been raised related to the use of misoprostol and the occurrence of meconium stained amniotic fluid during labour induction (Hofmeyr 2003). There is evidence from in vitro animal studies to indicate the stimulation of ileal smooth muscle by both dinoprostone and misoprostol (Matonhodze 2002), postulated to result in the fetal passage of meconium. However, the systematic reviews of both vaginal (Hofmeyr 2005b) and oral (Alfirevic 2005) misoprostol in the induction of labour provide some reassurance about the safety of misoprostol, particularly in relation to the presence of meconium stained liquor.
1.3(1)(ii) Clinical Toxicity of Misoprostol There have been two reported cases in the literature of overdose with misoprostol in pregnancy (Bond 1994; Austin 1997).
29 Bond and Van Zee (1994) report a case of a woman at 31 weeks gestation who presented having taken 6000mcg misoprostol in combination with trifluoperazine in a suicide attempt. Two hours after ingestion, she complained of shortness of breath and chills, and was noted to be febrile with a tetanic uterus. Initial laboratory investigations were consistent with a metabolic acidosis and hypoxaemia. An obstetric ultrasound performed one hour after initial presentation confirmed fetal death, and the woman gave birth to an appropriately grown stillborn infant four hours after ingestion. Nine hours after ingestion the woman’s temperature returned to normal limits, and she remained asymptomatic prior to discharge three days after presentation. Austin and colleagues (1997) report the case of a woman at 36 weeks gestation who self-administered 6000mcg of misoprostol intravaginally and 600mcg orally to induce labour. Three hours after ingestion at the time of presentation, she complained of chills, cramping pain, nausea and vomiting. Examination confirmed an elevated temperature, a firm uterus suggestive of tonic uterine activity, and fetal demise. Initial laboratory investigations revealed a picture of metabolic acidosis and hypoxaemia, which progressed to a disseminated intravascular coagulopathy (DIC), with evidence of placental abruption at birth. Over the course of admission, her symptoms resolved and she was discharged several days after presentation.
Similar symptoms have been reported in an elderly woman who consumed 3000mcg of misoprostol orally (Graber 1991).
1.3(1)(iii) Cost of Misoprostol Despite these safety concerns, the use of misoprostol as an induction agent has increased rapidly, related in part to its low cost when compared with other prostaglandin preparations commonly employed in obstetric practice. The current cost to the
Women’s and Children’s Hospital for a 1mg dose of prostaglandin E2 gel is (Aus)$40.17, with the cost of a 2mg dose increasing to (Aus)$51.64. The total cost to the hospital for the 1998/1999 financial year for prostaglandin gel alone was (Aus)$48,307, and for the 1999/2000 financial year (Aus)$38,508. In contrast a single 200mcg misoprostol tablet costs 35 cents.
In addition to the obvious cost differential between misoprostol and prostaglandin E2 gel, there are other costs associated with induction of labour. Several authors have
30 attempted to elucidate these differences further (Sanchez-Ramos 1993; Wing 1995a; Sanchez-Ramos 1997). In comparing vaginal misoprostol with intracervical prostaglandin E2 gel, Wing estimated misoprostol induction of labour to be associated with a saving of (US)$163 per woman induced in drug costs (Wing 1995a). Sanchez- Ramos compared the average induction costs of vaginal misoprostol with intravenous oxytocin alone, and with PGE2 gel followed by oxytocin infusion (Sanchez-Ramos 1993). The average induction costs for drugs alone was (US)$49 per woman given misoprostol, increasing to (US)$205 for those women receiving oxytocin alone, and
(US)$315 for those women administered PGE2 gel followed by oxytocin infusion. In a subsequent report, the authors compared misoprostol with dinoprostone inserts, with the average induction costs increasing to (US)$606 per woman with the use of dinoprostone inserts (Sanchez-Ramos 1997). Kramer compared intravenous oxytocin with vaginal misoprostol in a more detailed analysis that considered costs of intrapartum care for both vaginal and caesarean birth, epidural analgesia, post-partum care, and newborn nursery costs (Kramer 1997). The mean hospital charge per woman administered misoprostol was significantly less than for women induced with oxytocin infusion (US$2081 +/- $984 (misoprostol) versus US$2616 +/- $1035 (oxytocin); p<0.005).
The use of misoprostol as an induction agent has significant potential for reduction in both drug and hospital care costs for developed nations, in addition to providing an economical method of induction of labour in under resourced countries. A further advantage in this setting relates to its chemical stability at room temperature, obviating the need for refrigerated storage as is required with other obstetric prostaglandin compounds, in settings where this may be unavailable or unreliable.
While hospital administrators and health economists could make an impressive argument supporting the use of misoprostol as an induction agent, it is unlikely that this will propel manufacturers towards seeking appropriate product licensing for the use of misoprostol in pregnancy. There is considerable economic advantage to health service providers in the use of misoprostol for induction of labour, although its low production cost does not translate into financial incentives for the manufacturers.
31 1.3(2) Research Methodology
The randomised controlled trial (RCT) is accepted widely as the ‘gold standard’ in clinical research methodology, allowing comparison between alternate interventions or forms of care (Chalmers 1989; Greenhalgh 1997c). Any differences observed between alternate forms of care may be due to true differences in the care received, or differences in the pre-treatment characteristics of the individuals involved in the trial (Chalmers 1989). The ability to attribute differences in observed outcome to the actual form of treatment given relies on the extent to which trial participants are comparable in every respect, with the exception of the treatment received (Chalmers 1989). The RCT theoretically reduces systematic bias through the selection of a sample of participants from a particular population and then randomly allocating them to different interventions (Greenhalgh 1997a; Greenhalgh 1997c).
The relationship between randomisation and the estimate of treatment effect has been assessed, with non-randomised clinical trials distorting apparent effects of care in both directions, ranging from an underestimation of effect of 75% to an overestimation of 160% (Kunz 1998). The effect of a randomised control group with historical controls has also been investigated, the use of historical controls being more likely to be associated with significant differences between the study groups in favour of the treatment group, and generally having poorer outcome measures than randomised controls (Sacks 1982). Thus it is not surprising that the use of historical control groups is more likely to yield statistically significant differences between treatments than those trials using randomised controls (Chalmers 1983).
Another source of bias in clinical trials is the effect of knowledge of the treatment allocation by participants, caregivers and those assessing outcome. This form of bias can be reduced by blinding all involved to the treatment allocated (Chalmers 1989). Inadequate allocation concealment has been shown to be associated with a 40% increase in the estimate of treatment effect (Kunz 1998), with a similar exaggeration of effect present in trials where there was inadequate allocation concealment or where there was no formal attempt of blinding (Schulz 1995).
32 1.3(3) Misoprostol for Induction of Labour – Results from the current Cochrane Systematic Reviews
1.3(3)(i) Vaginal Misoprostol for Induction of Labour
The Cochrane Systematic Review assessing vaginal misoprostol for cervical ripening and induction of labour at term included 70 randomised controlled trials in which misoprostol was compared with placebo, intravenous oxytocin, vaginal prostaglandin
E2, and intracervical prostaglandins (Hofmeyr 2005b). Of these 70 trials, only 13 were double blind and placebo controlled (Fletcher 1993; El-Azeem 1997; Farah 1997; Surbek 1997; Gottschall 1998; Ramsey 1998; Srisomboon 1998; Diro 1999; Montealegre 1999; Stitely 2000; Khoury 2001; Ferguson 2002; Meydanli 2003).
Overall, women who were administered misoprostol were less likely not to achieve vaginal birth within 24 hours (7 trials; 1580 participants; RR 0.76; 95% CI 0.68-0.85), but at the expense of increased uterine hyperstimulation with fetal heart rate changes (11 trials; 1607 participants; RR 2.28; 95% CI 1.33-3.93), when compared with vaginal prostaglandins (Hofmeyr 2005b). While doses of more than 25 micrograms at four hourly intervals were “more effective” than conventional methods of induction of labour, it was at the expense of greater occurrence of uterine hyperstimulation. In contrast, lower doses of misoprostol were considered “as effective” as standard prostaglandin preparations, but without an increase in uterine contractile disorders. The authors indicate a need for further trials with appropriate blinding, further assessment of low doses of misoprostol, and of sufficient size to be able to detect differences in rare but clinically important complications (Hofmeyr 2005b).
1.3(3)(ii) Oral Misoprostol for Induction of Labour The current Cochrane Systematic Review of oral misoprostol in the induction of labour included 13 randomised controlled trials (Alfirevic 2005), in which oral misoprostol was compared with placebo, intravenous oxytocin, vaginal prostaglandin E2, intracervical prostaglandin E2, and vaginal misoprostol. Of these 13 trials, only four were double blind and placebo controlled, one comparing oral misoprostol with placebo
(Ngai 1996); one comparing oral misoprostol with vaginal prostaglandin E2 (Tessier 1997); and two comparing oral with vaginal misoprostol (Adair 1998; Bennett 1998).
33 1.3(3)(ii)(a) Oral Misoprostol versus Placebo A single randomised placebo controlled trial was identified, in which a single 200mcg oral dose of misoprostol was compared with placebo in women with pre-labour ruptured membranes at term (Ngai 1996). Women allocated to the misoprostol group were less likely to require oxytocin augmentation (1 trial; 80 participants; RR 0.25, 95% CI 0.10-0.60), had the randomisation to delivery interval reduced by 8.7 hours (1 trial; 80 participants; WMD –522.00 minutes; 95% CI –683.72 to –360.28) when compared with placebo. There was no difference observed in caesarean births (1 trial; 80 participants; RR 1.05, 95% CI 0.23-4.90), or uterine hyperstimulation with associated fetal heart rate changes (1 trial; 80 participants; RR 3.15; 95% CI 0.13-75.09). However, the trial had a small sample size and the confidence intervals around the treatment effect are wide.
1.3(3)(ii)(b) Oral Misoprostol versus Intravenous Oxytocin Two trials have compared oral misoprostol with intravenous oxytocin infusion (Butt 1999; Ngai 2000). Women allocated to the misoprostol groups were administered 50mcg at four hourly intervals (Butt et al. 1999), and 100mcg at four hourly intervals (Ngai 2000).
The meta-analysis of these two trials indicated no differences between the interventions with regards to caesarean birth (2 trials; 188 participants; RR 0.97, 95% CI 0.43-2.22) or uterine hyperstimulation with associated fetal heart rate changes (2 trials; 188 participants; RR 0.96, 95% CI 0.25-3.66).
1.3(3)(ii)(c) Oral Misoprostol versus Vaginal Prostaglandin E2 Two trials have compared oral misoprostol with vaginal prostaglandin E2 involving a total of 962 women. In one study, women were allocated to either the standard vaginal prostaglandin regimen or to 50mcg oral misoprostol, both administered at six hourly intervals, to a maximum of four doses (200mcg misoprostol) in a 24-hour period (Tessier 1997). In the other, women were randomly allocated to either vaginal dinoprostone gel, or to a titrated oral misoprostol solution, with the initial dose being 20mcg, and increased to 40mcg if no clinical effect was observed after two or three doses (Hofmeyr 2001).
The meta-analysis of these two trials indicates no differences between the interventions regarding vaginal birth not achieved in 24 hours (1 trial; 691 participants; RR 1.19; 95%
34 CI 0.94-1.51), caesarean birth (2 trials; 959 participants; RR 0.90; 95% CI 0.70-1.17), or uterine hyperstimulation with associated fetal heart rate changes (2 trials; 929 participants; RR 0.87, 95% CI 0.49-1.56).
1.3(3)(ii)(d) Oral Misoprostol versus Vaginal Misoprostol Seven trials have compared oral misoprostol with vaginal misoprostol, involving differing dosing intervals and regimens (Alfirevic 2005). The dose of oral misoprostol ranged from 50mcg at four hourly intervals (Bennett 1998; Wing 1999b; Dyar 2000; Shetty 2001), up to 100mcg at three hourly intervals (Toppozada 1997) and 200mcg at six hourly intervals (Adair 1998).
The meta-analysis performed (Alfirevic 2005) indicated women receiving oral misoprostol were more likely to remain undelivered vaginally after 24 hours (4 trials; 877 participants; RR 1.27, 95% CI 1.09-1.47). However, oral administration of misoprostol was associated with fewer caesarean births (7 trials; 1276 participants; RR 0.77, 95% CI 0.61-0.97), while there were no differences observed in uterine hyperstimulation and route of administration (6 trials; 1236 participants; RR 1.1, 95% CI 0.78-1.47).
1.3(3)(iii) Implications While both of these reviews have included a total of 83 trials, only 17 of the trials have been appropriately blinded and placebo controlled. In particular, only four of the trials involving an assessment of oral misoprostol have involved appropriate blinding of participants, caregivers and outcome assessors. With women and their caregivers aware of the allocated treatment in 80% of the trials conducted to date, there is potential for the introduction of bias in these results. To improve the quality of evidence available, and the certainty with which differences between treatment groups can be attributed to medication differences, further randomised controlled trials are required, but with the essential inclusion of an appropriate placebo control in order to maintain blinding.
1.3(4) Uterine Hyperstimulation
Uterine hyperstimulation is recognised as a potential complication following prostaglandin induction of labour, with concern that frequent uterine contractions may
35 impair utero-placental perfusion, manifesting as abnormalities of the fetal heart rate. The terminology surrounding uterine hyperstimulation has been one of confusion, including both uterine tachysystole and uterine hypertonus (Curtis 1987). Additionally, uterine hyperstimulation may be associated with fetal heart rate changes, or may occur in isolation. Management of uterine hyperstimulation has involved the use of intravenous or subcutaneous beta-mimetic therapy (Egarter 1990). However, management practices are inconsistent, with some practitioners treating hyperstimulation in the absence of fetal heart rate changes, while others reserve therapy for cases complicated by fetal heart rate abnormalities.
The occurrence of uterine hyperstimulation is well described following misoprostol induction of labour (Hofmeyr 1998; Hofmeyr 2005b). This is particularly the case following vaginal administration of misoprostol, with uterine hyperstimulation (both with and without fetal heart rate changes) occurring more frequently when compared with oxytocin and vaginal prostaglandin E2 (Hofmeyr 2005b). In an attempt to determine any effect of dose on the occurrence of hyperstimulation, lower dosing regimens (defined as 25mcg or less at four hourly or less intervals) have been compared with higher dosing regimens (Hofmeyr 2005b). The use of lower doses of misoprostol was not associated with an increase in the number of women who did not achieve vaginal birth within 24 hours (RR 1.04, 95% CI 0.92-1.18), although more women required oxytocin augmentation (RR 1.23, 95% CI 1.08-1.40). There were no differences observed between the regimens with regards to caesarean birth (RR 1.00, 95% CI 0.79-1.26), meconium stained liquor (RR 0.89, 95% CI 0.66-1.22), or maternal side effects (RR 0.71, 95% CI 0.46-1.11). Women administered lower doses of misoprostol were less likely to experience uterine hyperstimulation, both with (RR 0.50, 95% CI 0.36-0.71) and without (RR 0.61, 95% CI 0.49-0.76) fetal heart rate changes, with trends towards fewer infants with Apgar scores of less than seven at five minutes (RR 0.76, 95% CI 0.40-1.45), and fewer neonatal intensive care unit admissions (RR 0.82, 95% CI 0.64-1.05).
Wing has suggested that there may be a cumulative effect of misoprostol contributing to the occurrence of uterine hyperstimulation (Wing 1995a). A dose of 50mcg vaginal misoprostol administered at 3 hourly intervals resulted in 21 of 25 women experiencing uterine tachysystole (Wing 1995a). In subsequent trials, the dose of misoprostol was
36 reduced (Wing 1995b), and the dosing interval increased to four (Wing 1996) and six hours (Wing 1997). When 25mcg of misoprostol was administered, nine women developed tachysystole after the first dose, with the average interval to onset being 303 minutes (Wing 1995b). An additional 11 women developed tachysystole after the second dose of misoprostol, with the interval to onset reduced to 169 minutes. Increasing the dosing interval from three to six hours (Wing 1996) revealed a statistically non-significant decrease in the occurrence of tachysystole from 14.6% (three hourly intervals) to 11.2% (six hourly intervals). There was no difference in the interval from last dose to onset of tachysystole with either regimen.
These observed effects are in keeping with information obtained from pharmacokinetic studies performed in early pregnancy termination, where misoprostol was administered both orally and vaginally (Zieman 1997; Danielsson 1999). In a study comparing oral and vaginal absorption of misoprostol, two cohorts, each of 10 women were recruited, in which five were pregnant and five not (Zieman 1997). One cohort was administered 400mcg of misoprostol orally, while the other received 400mcg of misoprostol vaginally. Following oral administration, the plasma concentration of misoprostol acid rose quickly, peaking between 12.5 and 60 minutes (34 +/- 17 minutes) after administration, and falling steeply by 120 minutes. In contrast, the plasma concentration of misoprostol acid after vaginal administration rose gradually to reach a maximum level between 60 and 120 minutes (80 +/- 27 minutes), thereafter declining slowly to an average of 61% of the peak level, 240 minutes after administration. The incidence of side effects with misoprostol was similar between the two routes of administration, although numbers were too small for statistical comparison. The bioavailability of vaginal misoprostol was three times greater than that achieved with oral administration, as determined by the area under the concentration curve. Peak plasma levels were slightly lower, obtained more gradually, and persisted for longer duration when misoprostol was administered vaginally rather than orally.
Absorption of 200 and 400mcg doses of misoprostol has been compared following oral and vaginal routes of administration, and correlated with effects on uterine contractility (Danielsson 1999). Thirty women presenting for first trimester termination of pregnancy were recruited between 8 and 11 weeks gestation and received either 200mcg (5 women received oral administration and 6 women received vaginal administration),
37 or 400mcg (10 women with oral administration and 9 women with vaginal administration) of misoprostol. Intrauterine pressure was measured via a Grass polygraph connected to a pressure transducer. At the end of the monitoring period, pregnancy was terminated by suction curettage. Misoprostol was absorbed more quickly after oral administration, with plasma levels peaking after 30 minutes. In contrast, plasma levels after vaginal administration rose more gradually, reaching a peak 1-2 hours after administration, and remained detectable for a longer period. Independent of the route of administration, misoprostol treatment was associated with an increase in uterine tonus. This effect was more rapid following oral administration (mean onset of effect 7.8 minutes +/- 3 minutes), with maximal effect being seen after a mean of 25.5 minutes (+/- 5.0 minutes). In contrast, the onset of uterine contractility following vaginal administration was seen after 20.9 minutes (+/- 5.3 minutes), with maximal effect after 46.3 minutes (+/- 20.7 minutes). Uterine activity was more pronounced 2 hours after vaginal administration when compared with oral administration, the effect being even greater 4 hours after administration. This study reported no difference in effect (in terms of strength of uterine activity) between the two doses of misoprostol when administered via the same route.
The bioavailability of misoprostol administered vaginally is greater than when administered orally (Zieman 1997; Danielsson 1999). However, vaginal absorption of misoprostol is influenced by pH, an effect studied using litmus paper at sterile speculum examination prior to induction of labour (Gunalp 2000). Women with vaginal pH of less than 5 were found to have a shorter induction to birth interval and required fewer doses of misoprostol, when compared with women whose vaginal pH was greater than 5. This difference in birth interval was thought to reflect the improved liquefaction of misoprostol in an acidic medium (Singh 1999). The usual pH of the vagina during pregnancy is less than 4.5, due to the production of lactic acid by normal vaginal bacterial commensals (Gravett 1994). However, the vaginal environment can be changed, particularly in association with vaginal infection and following membrane rupture, resulting in alterations in pH (Gravett 1994).
The studies described above (Zieman 1997; Danielsson 1999) have involved women in early pregnancy following oral doses of misoprostol varying from 200 to 400 mcg. There have been no studies to date assessing the absorption kinetics of smaller doses of
38 oral misoprostol, which would be more relevant to its potential use as an agent to induce labour at term, where doses of 50mcg or less may be more appropriate.
Based on this information, an oral route of administration may be preferable to vaginal administration in the induction of labour. Vaginal administration has been complicated by high rates of uterine tachysystole, with evidence of a cumulative effect over time, as indicated by the shorter interval to onset of hyperstimulation following second and subsequent doses. This is in keeping with the pharmacokinetic profile of vaginally administered misoprostol, with levels 60% of the peak value obtained 240 minutes after vaginal administration. The potential advantage of an oral route of administration, despite reduced bioavailability when compared with vaginal administration, is the rapid peak in plasma levels, with the maximal peak in increased uterine tonus occurring at this point (Danielsson 1999). Additionally, the concentrations of the drug in plasma decreases quickly by 120 minutes after administration, reducing adverse effects secondary to drug accumulation. An oral dosing interval of two hours should allow for maximal plasma concentrations to be achieved (and therefore maximal increase in uterine tonus), reduced drug accumulation (and therefore reduced adverse effects such as uterine hyperstimulation), and allow the lowest possible dose to be administered to each woman while maintaining clinical effect.
1.3(5) Limitations of the Current Clinical Trials of Misoprostol
1.3(5)(i) Dose Administered The majority of clinical trials to date involving misoprostol for induction of labour at term have used doses between 50 and 200mcg. Low doses of misoprostol (less than 25mcg) have specifically been compared with higher doses (greater than 25mcg) in a sub-group analysis (Hofmeyr 2005b). The use of lower doses of misoprostol was not associated with a reduction in clinical efficacy (that is achieving vaginal birth within 24 hours of the induction commencing), indicating that further attention be directed to assessment of both clinical effect and side effects of even lower doses of misoprostol, in an attempt to determine the lowest clinically effective dose.
The investigation of low doses of misoprostol has been hampered to a large extent by the currently available commercial preparation of misoprostol, which is manufactured
39 as 100 and 200mcg scored tablets. Several studies have used 25mcg of misoprostol, obtained by manual quartering of a 100mcg tablet (Varaklis 1995; Wing 1996; Wing 1997; Clark 1998; Wing 1998; Wing 1999a). Dosage variation in attempting to use 25mcg misoprostol has been assessed in which one hundred microgram tablets were scored to produce a 50mcg half tablet that was further halved to yield a 25mcg fragment, and compared using both freehand razor blade and pill cutter techniques (Williams 2001; Williams 2002). These tablet fragments were then assessed for usability, including weight and assay for misoprostol content using liquid chromatography. While the mean fragment weights were not different between the two methods of cutting, there was greater variance among those pills cut with pill cutter (SD+/-5.8) when compared with those cut using the razor blade (SD+/-2.6) (Williams 2002). The cumulative weight recovery of tablet fragments was higher for tablets divided with the razor blade, with 99% of the original tablet recovered compared with 96.6% when using the pill cutter (Williams 2002). Misoprostol content as determined by liquid chromatography indicated that 62% of assayed tablet fragments were within 10% of the expected 25mcg dose, 25% varied between 10 and 20%, and 12% were more than 20% of the expected dose (Williams 2002). While the authors concluded that manual quartering of a standard 100mcg misoprostol tablet produced an accurate low- dose form of misoprostol (Williams 2002), it necessitates weighing of each individual tablet fragment prior to administration.
In an attempt to overcome this difficulty and still administer a low dose (20mcg), misoprostol has been administered as a solution (Hofmeyr 2001). A 200mcg commercially available tablet was dispersed in 200mL water, producing a solution with misoprostol concentration 1mcg/mL. Women were then administered 20mL of solution orally. The bioavailability of oral misoprostol tablets relative to an oral solution has been investigated, where subjects were administered a 400mcg commercial tablet orally, as well as 400mcg of misoprostol / HPMC dispersion in solution (Karim 1987). The absorption profile of the two preparations was indistinguishable (Karim 1987) (Figure 1.3(5)(i)).
40 Figure 1.3(5)(i): Absorption profile of oral misoprostol tablets and misoprostol solution (Karim 1987)
Legend: Circles = administration of two 200mcg misoprostol tablets; squares = administration 400mcg misoprostol in solution.
There is currently no pharmacokinetic data available detailing the long-term stability of misoprostol in solution. Misoprostol solution has been made up immediately prior to administration, and also administered as a 20mL aliquot of the same solution used for a maximum of 12 hours (Hofmeyr 2001). In this trial, measured outcomes were similar across sites, indicating indirectly that misoprostol was not significantly degraded chemically in solution for a period of at least 12 hours. Making a fresh solution prior to each administration is a time consuming process, and additionally discards 80mcg of misoprostol in the remaining solution, which may have economic implications particularly in under-resourced countries.
1.3(5)(ii) Oral Route of Administration The use of oral misoprostol in the induction of labour has been under-investigated to date, with the majority of randomised controlled trials assessing vaginal misoprostol. This is in part due to the increased bioavailability of misoprostol when administered vaginally, and the subsequent likelihood of achieving vaginal birth within 24 hours of the induction process commencing, when compared with oral administration (RR 1.27, 95% CI 1.09-1.47) (Alfirevic 2005). The increased ‘speed’ of the labour and birth process has to be balanced against the reduction in caesarean birth (RR 0.77, 95% CI 0.61-0.97) observed following oral administration (Alfirevic 2005). An oral route of administration has advantages of ease and convenience for both woman and caregiver,
41 with the potential to administer small frequent doses in a controlled fashion to attain clinical effect, while minimising unwanted side effects.
1.3(5)(iii) Comparison of Misoprostol with the Current Standard Preparation Oral misoprostol has been compared with placebo (Ngai 1996), intravenous oxytocin
(Butt 1999; Ngai 2000), vaginal prostaglandin E2 (Tessier 1997; Hofmeyr 2001), intracervical prostaglandin E2 (Bartha et al. 2000), and vaginal misoprostol (Toppozada 1997; Adair 1998; Bennett 1998; Wing 1999a; Dyar 2000; Shetty 2001). Neither oral nor vaginal misoprostol are the current standard methods of cervical ripening and induction of labour in most obstetric units. Misoprostol should therefore be compared with the current standard – vaginal prostaglandin E2 gel for women with intact membranes, and intravenous oxytocin for women with pre-labour ruptured membranes.
1.3(6) Essential Features of Future Misoprostol Clinical Trials
1.3(6)(i) Randomised Trials with Placebo Control As indicated previously, the randomised controlled trial is the “gold standard” in clinical research methodology. While there have been randomised trials assessing misoprostol, the majority of these trials have not included a placebo control.
The commercial preparation of misoprostol limits the suitability of other similar appearing tablets that can safely be administered during pregnancy for use as a placebo. If misoprostol is crushed to a powder for administration as a dispersed solution, it is possible (although time consuming) to develop a similar appearing placebo solution. Similarly, the appearance of the dinoprostone applicator requires transfer of dinoprostone under sterile conditions into a syringe that can be used for the placebo gel also. While both of these procedures are cumbersome, it is possible to develop an appropriate oral and vaginal placebo.
1.3(6)(ii) Investigation of the Lowest Effective Clinical Dose The majority of trials of misoprostol have utilised doses between 50 and 200mcg, due to its commercial preparation as 100 and 200mcg scored tablets. The use of misoprostol dispersed in solution makes it possible to administer low doses in a convenient manner for women and their caregivers.
42 1.3(6)(iii) Broad Inclusion Criteria In order to apply the results of any trial to the wider population concerned, the inclusion criteria must closely reflect the wider population. Induction of labour is performed for indications where there is compromise of either maternal or fetal condition. Trial inclusion criteria should therefore aim to include these women, as it is important to have information regarding the use of misoprostol in these clinical settings prior to its widespread introduction into clinical practice.
1.3(6)(iv) Increased Information Regarding Safety While misoprostol appears to be an effective agent in the induction of labour, its widespread clinical use cannot be recommended until further information becomes available regarding its safety. With events such as uterine rupture, severe neonatal acidosis, and maternal and neonatal death being rare, large numbers of women will need to be induced with misoprostol before reliable clinical information regarding these outcomes is available.
1.3(6)(v) Assessment of Women’s Preferences for Care A perceived advantage of the use of oral misoprostol is its ease and convenience for the woman and her caregivers, in particular reducing the need for repeated vaginal examinations. The use of a vaginal placebo necessitates that all women involved require vaginal examination, a factor that may influence their overall preferences and perception of trial participation. However, information can still be obtained regarding preferences for oral or vaginal administration of medication for any future pregnancy should induction of labour be required.
With these factors as a basis, the current randomised controlled trial comparing oral misoprostol with vaginal prostaglandin E2 was planned.
43 2. ORAL MISOPROSTOL VERSUS VAGINAL
PROSTAGLANDIN E2 FOR THE INDUCTION OF LABOUR AT TERM – A SYSTEMATIC REVIEW AND META-ANALYSIS
2.1 INTRODUCTION
In the current Cochrane review (Alfirevic 2005), oral misoprostol has been compared with placebo (Ngai 1996), intravenous oxytocin (Butt 1999; Ngai 2000), vaginal prostaglandin E2 (Tessier 1997; Hofmeyr 2001), intracervical prostaglandin E2 (Bartha 2000), and vaginal misoprostol (Toppozada 1997; Adair 1998; Bennett 1998; Wing 1999a; Dyar 2000; Shetty 2001). Neither oral nor vaginal misoprostol are the current standard methods of cervical ripening available to clinicians at the Women’s and Children’s Hospital for the induction of labour. Misoprostol should therefore be compared with the current standard induction agent for women with intact membranes – vaginal prostaglandin E2 gel.
The aim of this review was to assess the benefits and harms associated with the use of oral misoprostol for induction of labour in women at term, compared with our current standard method of induction of labour for women with intact membranes, being vaginal prostaglandin E2 gel.
2.2 METHODS
2.2(1) Types of studies Published randomised controlled trials in which oral misoprostol was compared with intravaginal prostaglandin E2.
2.2(2) Types of participants
Women undergoing induction of labour at or near term.
44 2.2(3) Types of interventions
Oral misoprostol for the induction of labour at or near term.
2.2(4) Types of outcome measures
Outcomes include those detailed in the Cochrane Review relating to all methods of induction of labour (Hofmeyr 2005a), and include vaginal birth not achieved within 24 hours, uterine hyperstimulation with associated fetal heart rate changes, and caesarean section (all and those for non-reassuring fetal heart rate tracing). Uterine hyperstimulation was defined as uterine tachysystole (with five or more contractions in a 10 minute period for two consecutive 10 minute periods) or uterine hypertonus (a uterine contraction lasting for more than two minutes) (Curtis 1987). The fetal heart rate changes that were recorded as abnormal included persistent decelerations, fetal tachycardia (fetal heart rate greater than 160 beats per minute), fetal bradycardia (fetal heart rate less than 100 beats per minute), or reduced short term variability (less than five beats per minute) (RANZCOG 2001; RCOG 2001). Secondary outcomes relate to evidence of effect (cervix unfavourable or unchanged after 24 hours (as assessed by a modified Bishop’s score before and at 24 hours if not delivered); need for oxytocin augmentation; and the induction to birth interval); complications (including uterine hyperstimulation without fetal heart rate changes; uterine rupture; need for epidural analgesia; instrumental vaginal birth; presence of meconium stained amniotic fluid; infant Apgar score of less than seven at five minutes; infant admission to the neonatal intensive care unit (and length of stay); neonatal encephalopathy; perinatal death; serious infant morbidity; maternal side effects (including nausea, vomiting, diarrhoea, abdominal pain); postpartum haemorrhage (greater than 600mL and greater than 1000mL; need for blood transfusion); serious maternal complications (admission to intensive care unit, hyperpyrexia, coma); and maternal death); and satisfaction with the induction method (woman not satisfied; and caregiver not satisfied).
Only outcomes with available data appear in the analysis table, and any other outcomes that were not pre-stated are labelled as such.
45 2.2(5) Search strategies for identification of studies This review used the search strategy developed by the Cochrane Pregnancy and Childbirth Group of the Cochrane Collaboration, and involved searching the group’s specialised register of controlled trials (CENTRAL). In addition, a MEDLINE search was conducted using the key words misoprostol, oral; vaginal prostaglandin E2; pregnancy, third trimester; and induction of labo(u)r. The date of last search was February 2005. A manual search was conducted of the reference list of all identified papers.
2.2(6) Methods of the review
Identified trials were evaluated for appropriateness for inclusion and methodological quality without consideration of their results, according to pre-stated eligibility criteria. Eligible trials were assessed using the following criteria for assessment of quality: • Generation of random allocation sequence: adequate, inadequate, unclear; • Allocation concealment: A=adequate, B=unclear, C=inadequate; • Blinding of participants: yes, no, inadequate, no information; • Blinding of caregivers: yes, no, inadequate, no information; • Blinding of outcome assessment: yes, no, inadequate, no information; • Completeness of follow-up data (including any differential loss of participants from each group): A=<3% of participants excluded; B=3-9.9% of participants excluded; C=10-19.9% of participants excluded; D=20% or more excluded; E=unclear; • Analysis of participants in randomised groups.
Data was extracted from the original papers and data from different trials were combined if considered to be sufficiently similar. Meta-analyses were performed using relative risks for binary outcomes, and weighted mean differences for continuous outcomes.
46 2.3 RESULTS
2.3(1) Description of Studies
The search strategy identified seven randomised controlled trials comparing oral misoprostol with vaginal prostaglandin E2, of which five were included, involving a total of 1419 women (Tessier 1997; Gherman 2001; Hofmeyr 2001; Dallenbach 2003; Shetty 2003; Shetty 2004). The dose of misoprostol varied from 20mcg at two hourly intervals (Hofmeyr 2001; Dallenbach 2003), to 50mcg at four hourly (Gherman 2001) or six hourly intervals (Tessier 1997), to 100mcg at four hourly intervals (Shetty 2003; Shetty 2004).
The trial by Matonhodze and colleagues (2003) involved a three-way comparison of Foley catheter and titrated oral misoprostol, titrated oral misoprostol alone, and vaginal dinoprostone. The arms comparing titrated oral misoprostol alone with vaginal dinoprostone formed part of a larger randomised trial and have been included in that report (Hofmeyr 2001).
The trial by Windrim and colleagues (1997) was not included as the control group consisted of “physician chosen” methods of induction, including vaginal prostaglandins, intracervical prostaglandins, artificial rupture of membranes and oxytocin infusion, and it was not possible to determine the numbers of women receiving each intervention.
The characteristics of the included studies are summarised in Table 2.3(1).
47 Table 2.3(1) Description of Included Studies
Trial Identification Trial Characteristics Tessier (Tessier 1997) Participants: women with an “indication” for induction of labour Interventions: 50mcg oral misoprostol at 6 hourly intervals vs vaginal
PGE2 at 6 hourly intervals; maximum 4 doses in 24 hour period Sample Size: 267 women Outcomes: caesarean birth; uterine hyperstimulation; induction to birth interval Randomisation: stated to be “randomly assigned” (A) Allocation Concealment: stated to be “double blind” (A) Blinding: stated to be “double blind” Gherman (Gherman 2001) Methods: Trial conducted at Naval Medical Center, Portsmouth, Virginia between December 1998 and June 1999 Participants: women with singleton gestation in cephalic presentation at greater than 24 weeks gestation, a medical or obstetric indication for induction of labour and Bishop score of less than 6; exclusion criteria included Bishop score greater than 6, prior uterine incision, fetal malpresentation, placenta praevia or unexplained vaginal bleeding, active herpes simplex infection, estimated fetal weight greater than 4500g or less than 2500 g, asthma or cardiovascular disease, hypersensitivity to prostaglandins, contraindication to prostaglandin administration, renal or hepatic dysfunction, or non- reassuring fetal heart rate tracing Interventions: 50mcg oral misoprostol at 4 hourly intervals vs 4mg
vaginal PGE2 at 4 hourly intervals; max. 6 doses in 24 hour period Sample Size: 60 women Outcomes: failure to achieve vaginal birth within 48 hours of induction commencing; caesarean birth; uterine hyperstimulation; induction to birth interval; need for epidural; oxytocin use; Apgar score less than 7 at 5 minutes; meconium stained liquor Randomisation: computer generated random number table with variable blocks (A) Allocation Concealment: sequentially numbered, sealed opaque envelopes (A) Blinding: participants, caregivers and outcome assessors not blinded to treatment allocation
48 Trial Identification Trial Characteristics Hofmeyr (Hofmeyr 2001) Methods: Trial conducted Coronation & Chris Hani Baragwanath Hospitals (South Africa), & The Liverpool Women’s Hospital (UK) Participants: women with a singleton pregnancy in cephalic presentation; exclusion criteria included gestational age less than 34 weeks, prior caesarean birth, uncontrolled diabetes mellitus, antepartum haemorrhage, pre-eclampsia, eclampsia, “fetal distress” Interventions: 20mcg oral misoprostol solution at 2 hourly intervals
(increased to 40mcg if no response after 2 or 3 doses) vs vaginal PGE2 at 6 hourly intervals; maximum 2 vaginal doses in 12 hour period Sample Size: 648 women Outcomes: vaginal birth not achieved in 24 hours; caesarean birth; uterine hyperstimulation (with and without fetal heart rate changes) Randomisation: computer generated random number list; block sizes between two and six (A) Allocation Concealment: sequentially numbered, sealed opaque envelopes (A) Blinding: participants, caregivers and outcome assessors not blinded to treatment allocation Dallenbach (Dallenbach Methods: Trial conducted at Geneva University hospital, Switzerland 2003) between September 1999 and April 2001 Participants: women with a singleton pregnancy, cephalic presentation, gestational age greater than 37 weeks, and Bishop score less than six; women were excluded if non-reassuring fetal heart rate tracing, oligohydramnios, intrauterine growth restriction, fetal malformation, vaginal bleeding or any contraindication to prostaglandin administration Interventions: 20mcg oral misoprostol solution at 2 hourly intervals
(increased to 40mcg) to a maximum of 10 doses vs 2mg vaginal PGE2 at 6 hourly intervals; maximum 4 doses in 24 hour period Sample Size: 200 women Outcomes: vaginal birth not achieved within 24 hours; caesarean birth; uterine hyperstimulation (both with and without fetal heart rate changes); induction to birth interval; need for augmentation of labour; neonatal and maternal morbidity and mortality Randomisation: computer generated random number list (A) Allocation Concealment: sealed opaque envelopes (A) Blinding: participants, caregivers and outcome assessors not blinded to treatment allocation
49
Trial Identification Trial Characteristics Shetty (Shetty 2004) Methods: Trial conducted at the Aberdeen Maternity Hospital (United Kingdom) between October 2001 and February 2002 Participants: women with a singleton gestation in cephalic presentation, gestational age greater than 37 weeks, and medical or obstetric indication for induction of labour; exclusion women with prior caesarean birth, hypersensitivity to prostaglandins, or parity > 5 Interventions: 100mcg oral misoprostol at 4 hourly intervals (to a
maximum of 5 doses) vs 3mg vaginal PGE2 at 6 hourly intervals Sample Size: 200 women Outcomes: vaginal birth not achieved within 24 hours; caesarean birth; uterine hyperstimulation; induction to birth interval; need for oxytocin; need for fetal scalp pH estimation; meconium stained liquor; infant birth weight; admission to NICU Randomisation: computer generated random number list (A) Allocation Concealment: sealed opaque envelopes (A) Blinding: participants, caregivers and outcome assessors not blinded to treatment allocation
2.3(2) Meta-analysis
The pre-specified outcomes were well reported by Hofmeyr (Hofmeyr 2001) and Dallenbach (Dallenbach 2003), although there was no information available about women’s preferences and satisfaction with care.
For the primary clinical outcomes, there were no differences between oral misoprostol and vaginal PGE2, with regards to vaginal birth not achieved within 24 hours (3 trials, 1040 participants, relative risk 1.10, 95% Confidence Intervals 0.94-1.28), uterine hyperstimulation with associated fetal heart rate changes (5 trials, 1387 participants, RR 0.85, 95% CI 0.54-1.33) or caesarean birth (5 trials, 1387 participants, RR 0.94, 95% CI 0.77-1.16) (Table 2.3(3)i). There were no reports of serious neonatal (1 trial, 267 participants, RR not estimable) or maternal morbidity (2 trials, 962 participants, RR not estimable) in those trials reporting these outcomes. Women administered oral misoprostol had greater need for oxytocin augmentation (2 trials, 258 participants, RR 1.13, 95% CI 1.11-1.73), shorter induction to birth interval (4 trials, 1157 participants,
50 WMD –1.38 hours, 95% CI –2.60 to –0.15), and increased nausea reported (2 trials, 258 participants, RR 1.86, 95% CI 1.00-3.46). For the remaining secondary outcomes reported relating to evidence of effect, maternal complications and neonatal complications, there were no differences identified between oral misoprostol and vaginal PGE2 (Tables 2.3(3)ii-2.3(3)iv).
Table 2.3(2)(i) Meta-analysis of primary outcomes
Outcome No. No. Relative Risk 95% CI Trials Participants Vaginal birth >24 hours 3 1040 1.10 0.94-1.28 Hyperstimulation with FHR changes 5 1387 0.85 0.54-1.33 Caesarean section 5 1417 0.94 0.77-1.16 Serious neonatal morbidity 1 267 Not estimable Serious maternal morbidity 2 962 Not estimable
Table 2.3(2)(ii) Meta-analysis of secondary outcomes – evidence of effect
Outcome No. No. Relative Risk 95% CI Trials Participants Cervix unchanged 12-24 hours 1 189 1.13 0.57- 2.24 Need for oxytocin augmentation 2 258 1.38 1.11- 1.73 Induction – birth interval* 4 1157 -1.38 -2.60- -0.15
*Weighted mean difference (interval in hours)
51 Table 2.3(2)(iii) Meta-analysis of secondary outcomes – maternal complications
Outcome No. No. Relative Risk 95% CI Trials Participants Hyperstimulation no FHR changes 2 862 1.04 0.66-1.65 Uterine rupture 2 891 Not estimable Epidural analgesia 2 258 0.98 0.89-1.08 Instrumental vaginal birth 3 1092 0.90 0.68-1.20 Maternal side effects – any 2 891 1.17 0.96-1.42 Maternal side effects – nausea 2 258 1.86 1.00-3.46 Maternal side effects – vomiting 2 891 1.25 0.88-1.77 Maternal side effects – diarrhoea 2 200 0.33 0.01-8.09 Post partum haemorrhage 2 892 1.03 0.78-1.35 Maternal Death 1 692 Not estimable
Table 2.3(2)iv Meta-analysis of secondary clinical outcomes – neonatal complications
Outcome No. No. Relative Risk 95% CI Trials Participants Meconium Stained Liquor 3 458 1.03 0.67- 1.58 Apgar Score <7 at 5 minutes 3 949 0.71 0.34- 1.46 Admission to NICU 3 1091 0.78 0.48- 1.25 Perinatal Death 2 958 1.00 0.06-15.97
52 2.4 DISCUSSION
There are several limitations of the currently published randomised controlled trial literature comparing oral misoprostol with vaginal prostaglandin E2, relating in particular to the need for appropriately blinded studies, further assessment of low doses of misoprostol, broad inclusion criteria representative of all indications for induction of labour, and information about women’s preferences for care.
Of the 5 studies identified and included in this systematic review, only one was described as double blind and placebo controlled (Tessier 1997). While this study indicated that all women received both an oral and vaginal preparation, the success of blinding was not reported. With over 80% of women and their caregivers aware of the allocated treatment, there remains a significant potential for the introduction of bias into these results. Inadequate allocation concealment has been shown to be associated with a 40% increase in the estimate of effect (Kunz 1998), with a similar exaggeration of effect present in trials where there was inadequate allocation concealment or no formal attempt at blinding (Schulz 1995).
Further randomised controlled trials comparing oral misoprostol with vaginal prostaglandin E2 require the essential inclusion of an appropriate placebo control to maintain blinding. This is not without difficulty, however, given appearance of commercially available misoprostol and vaginal PGE2 gel preparations, which limits the use of other similar appearing agents than can be used safely during pregnancy as a placebo.
The availability of misoprostol preparations as 100 and 200mcg scored tablets has similarly hindered the investigation of low doses of oral misoprostol in induction of labour. Dosage variation in attempting to use 25mcg misoprostol through manual quartering of a 100mcg tablet has been assessed (Williams 2001; Williams 2002). While the authors concluded that an accurate low-dose form of misoprostol was achieved (Williams 2002), it does necessitate weighing of individual tablet fragments.
53 Administration of an oral solution in which crushed misoprostol is dispersed in water (Hofmeyr 2001; Dallenbach 2003) has allowed several investigators to assess the efficacy of low doses (20mcg) misoprostol in the induction of labour. The bioavailability of oral misoprostol tablets relative to an oral solution has been investigated previously (Karim 1987), with an identical absorption profile noted to standard tablet preparations.
In order to apply the results of any trial to the wider population concerned, the inclusion criteria must be broad to reflect the variety of indications for which labour is induced. Induction of labour is performed for indications where there is compromise of either maternal or fetal condition. Trial inclusion criteria should therefore aim to include these women, as it is important to have information regarding the use of misoprostol in these clinical settings prior to its widespread introduction into clinical practice. The inclusion criteria for the Tessier study (1997) were women with an “indication” for induction, although details were not provided in the report. The trial by Hofmeyr (2001) excluded women whose induction was performed for antepartum haemorrhage, uncontrolled diabetes mellitus, severe pre-eclampsia, or eclampsia. The trial by Dallenbach (2003) excluded women with a pregnancy suspected to be at risk of “fetal distress” due to non- reassuring fetal heart rate tracing, oligohydramnios or intrauterine growth restriction. These are all valid indications for induction of labour, and for the study results to be as widely applicable as possible, information needs to be obtained on the use of misoprostol in situations with suspected maternal or fetal compromise.
While misoprostol appears to be an effective agent in the induction of labour, further information is required about its safety. With events such as uterine rupture, severe neonatal acidosis, and maternal and neonatal death being rare, large numbers of women will need to be induced with misoprostol before reliable clinical information regarding these outcomes is available. There is a lack of information in the studies to date assessing women’s preferences for care during induction of labour. A perceived advantage of the use of oral misoprostol is its ease and convenience for the woman and her caregivers, in particular reducing the need for repeated vaginal examinations. The use of a vaginal placebo would necessitate that all women involved would require vaginal examinations, a factor that may influence their overall preferences and perception of trial participation. However,
54 information can still be obtained regarding preferences for oral or vaginal administration of medication for future pregnancies should induction of labour be required.
2.5 CONCLUSIONS
While oral misoprostol appears effective and safe, there have been too few trials, with too few participants to allow study of rare but significant clinical events, such as uterine hyperstimulation with fetal heart rate changes, meconium stained liquor, infant Apgar score of less than seven at five minutes, and neonatal intensive care unit admission. These outcomes have been used as indirect markers of rare and more severe outcomes related to neonatal acidaemia. Further information regarding these outcomes is required. Any future randomised controlled trials should address the use of low dose oral misoprostol, with the essential inclusion of appropriate blinding of women, their caregivers and those assessing outcome measures, to improve the reliability of information available.
55 3. MISOPROSTOL FOR CERVICAL RIPENING AND INDUCTION OF LABOUR IN WOMEN WITH INTACT MEMBRANES AT TERM: A RANDOMISED CONTROLLED TRIAL
3.1 INTRODUCTION
The current Cochrane Systematic Review of oral misoprostol in the induction of labour includes 13 randomised controlled trials (Alfirevic 2005), in which oral misoprostol has been compared with placebo, intravenous oxytocin, vaginal prostaglandin E2, intracervical prostaglandin E2, and vaginal misoprostol. Of these 13 trials, only four were double blind and placebo controlled, one comparing oral misoprostol with placebo
(Ngai 1996); one comparing oral misoprostol with vaginal prostaglandin E2 (Tessier 1997); and two comparing oral with vaginal misoprostol (Adair 1998; Bennett 1998).
Five randomised controlled trials have been identified comparing oral misoprostol with vaginal prostaglandin E2 gel (Tessier 1997; Gherman 2001; Hofmeyr 2001; Dallenbach 2003; Shetty 2004). When the results of these trials are combined in a meta-analysis, there are no differences between oral misoprostol and vaginal PGE2 gel, with regards to vaginal birth not achieved in 24 hours (3 trials, 1040 participants, RR 1.10, 95% CI 0.94-1.28), uterine hyperstimulation with associated fetal heart rate changes (5 trials, 1387 participants, RR 0.85, 95% CI 0.54-1.33), or caesarean birth (5 trials, 1387 participants, RR 0.94, 95% CI 0.77-1.16).
Only one of these trials was described as double blind and placebo controlled (Tessier 1997), and with over 80% of women and their caregivers aware of allocated treatment there is considerable potential for the introduction of bias into these results. As highlighted in the systematic review comparing oral misoprostol with vaginal PGE2 gel (Chapter 2), there is a need for further randomised controlled trials to address the use of low dose oral misoprostol, with the essential inclusion of appropriate blinding of
56 women, their caregivers and those assessing outcome measures, to improve the reliability of information available.
3.2 STUDY AIMS & HYPOTHESES
The purpose of this study was to compare the current vaginal prostaglandin E2 gel with oral misoprostol in the induction of labour at term, in a randomised, double blind, placebo controlled trial.
The hypotheses of the trial were that the use of oral misoprostol for the induction of labour at term, when compared with vaginal prostaglandin E2 gel was associated with • No difference in the number of women who remain undelivered 24 hours after the induction process commences; • No difference in the rate of uterine hyperstimulation associated with fetal heart rate changes; and • No difference in the rate of caesarean section (including those performed primarily for fetal distress).
3.3 METHODS
3.3(1) Trial Design
A prospective randomised, double blind, placebo controlled trial comparing oral misoprostol solution with vaginal prostaglandin E2 gel in the induction of labour at term was conducted at the Women’s and Children’s Hospital, the Lyell McEwin Health Service, and The Hervey Bay Hospital, between April 2001 and December 2004. Research and ethics approval was obtained from each of the three collaborating institutions.
Women were recruited from the antenatal clinic, antenatal wards and delivery suite of participating institutions. Potentially eligible women presenting for induction of labour were given the trial information sheet, counselled by a member of the study team, and
57 encouraged to discuss the study with her family. Written informed consent was then obtained.
3.3(2) Inclusion Criteria
Women with a singleton pregnancy at a gestational age of 366 weeks or greater were approached by a member of the study team after the decision had been made by the attending obstetrician to induce labour. Indications for induction of labour included: prolonged pregnancy; pre-eclampsia; hypertensive disorders; antepartum haemorrhage (excluding placenta praevia); oligohydramnios; intra-uterine growth restriction; gestational diabetes; and “other” indications, including social reasons.
3.3(3) Exclusion Criteria
Women with the following conditions were excluded from trial participation: active labour (defined as regular uterine contractions, and dilatation of three centimetres or more and full effacement of the cervix); a “favourable” cervix (defined as a modified Bishop’s score of seven or more); a contraindication to vaginal birth; previous uterine surgery (including prior caesarean section); maternal history of asthma, glaucoma, or heart disease; ruptured membranes; parity of five or more; suspected cephalopelvic disproportion; abnormal fetal lie; breech presentation; fetal distress or a non-reassuring fetal heart rate tracing; multiple pregnancy; intrauterine infection; vasa praevia; placenta praevia; active herpes infection; maternal illness precluding induction of labour; and maternal renal or hepatic failure.
3.3(4) Randomisation Schedule
The randomisation schedule was generated by a non-clinical researcher using a computer generated sequence with variable blocks and stratification for the woman’s parity (0 and 1-4). Treatment packs were prepared by the principal investigator and the research midwife, and labelled by a non-clinical researcher according to the generated randomisation schedule. Opening the treatment pack was the point of randomisation.
58 3.3(5) Blinding
Treatment packs had an identical appearance, each containing six white plastic bottles. Each bottle was labelled (oral medication 1-6) and contained an identical appearing white powder (with either crushed misoprostol or crushed vitamin B6). The solutions generated by dispersing the powder in water were both tasteless. Packs also contained two sealed sterile opaque plastic bags, each with an opaque sterile syringe containing either sterile placebo vaginal gel or sterile vaginal PGE2 gel. Both the placebo and active vaginal gel were transferred into identical appearing syringes under sterile conditions. All packs were sealed and a label placed across the seal to prevent tampering or opening prior to use.
3.3(6) Treatment Allocation
When women who had given consent to trial participation were admitted to delivery suite for induction of labour, trial entry details were confirmed. A study number was allocated to the woman, by taking the next identically appearing, sequentially numbered treatment pack appropriate for her parity. 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 (Prostin) gel. Each pack contained a sheet detailing treatment schedules. The treatment packs were kept at 4O Celsius (although this was not a requirement for misoprostol).
3.3(7) Treatment Schedules
3.3(7)(i) Misoprostol Treatment Packs Each misoprostol pack contained six labelled white plastic bottles, each containing 100mcg of crushed misoprostol. This was obtained by scoring a 200mcg misoprostol tablet and crushing the half tablet to produce a white powder. Also in the pack were two doses of vaginal placebo (tylose) gel in sterile opaque syringes and sealed in sterile opaque packages.
59 3.3(7)(ii) Prostaglandin E2 Treatment Packs Each placebo pack contained six labelled white plastic bottles, each containing 25mg of crushed vitamin B6. Also in the pack were two doses of vaginal prostin gel in sterile opaque syringes and sealed in sterile opaque packages. The prostin was transferred from the applicators provided by the manufacturers and placed into opaque syringes identical to the placebo gel, under sterile conditions. The packs for primigravid women contained two doses of 2mg prostaglandin E2 gel, while packs for multiparous women contained two doses of 1mg prostaglandin E2 gel, consistent with local hospital practices (Government of South Australia 2005).
3.3(7)(iii) Care of all women For all women, a preliminary fetal cardiotocogram (CTG) was recorded for 20 minutes, and had to be within normal limits for the study preparation to be given. A vaginal examination was performed and the initial Bishop’s score recorded. The intravaginal gel was then given (either prostaglandin E2 or placebo gel). The CTG was continued for a period of 20 minutes after the trial medication had been administered.
The oral solution was made up immediately prior to administration of the current required dose by the primary care midwife. The powder (either misoprostol or vitamin
B6) from the plastic bottle was transferred to a brown glass bottle of 100mL capacity. A total of 100mL of water was added to the bottle, the lid replaced, and the contents shaken to mix the powder and water. This produced a solution of 1mcg/mL of misoprostol in the active solution. A 20mL aliquot of solution (to produce a dose of 20mcg misoprostol solution in the misoprostol group) was then taken and administered orally to the woman. The remaining 80mL of solution was discarded. This procedure was repeated at two hourly intervals, with a CTG being obtained for 20 minutes prior and following each oral dose.
The oral solution was administered at two hourly intervals to a maximum of 6 doses over a 12-hour period. The vaginal gel (2mg dose for nulliparous women and 1mg for multiparous women in the prostaglandin E2 group and tylose gel for women in the misoprostol group) was administered at six hourly intervals to a maximum of 2 doses over a 12-hour period. The second dose of vaginal gel if required (2mg dose for nulliparous women and 1mg for multiparous women in the prostaglandin E2 group and
60 tylose gel for women in the misoprostol group), was given concurrently with the fourth oral solution if this was required. All women underwent a vaginal examination six hours after the administration of the first dose of vaginal PGE2 gel, and six hours after the administration of the second dose of vaginal PGE2 gel if this was required. All care provided to women was according to local hospital guidelines, with the exception of the administration of trial medications as described.
All trial medication was ceased once the woman’s cervix was suitable for artificial rupture of the membranes (ARM), following spontaneous rupture of the membranes (SROM), or in the presence of three to four regular uterine contractions in a 10-minute period.
Following ARM, SROM, the onset of labour, or completion of all study medications, women were managed according to the usual delivery suite protocols, including administration of known prostaglandin E2 gel if the cervix remained unfavourable, or the use of oxytocin infusion if required. The further administration of vaginal PGE2 gel was at the discretion of the consultant obstetrician on call for delivery suite if the woman’s cervix remained unfavourable.
3.3(8) Study Outcome Measures
3.3(8)(i) Primary Study Outcomes The primary study outcomes were consistent with those in the Cochrane generic protocol relating to induction of labour (Hofmeyr 2005a), and were • vaginal birth not achieved within 24 hours, • uterine hyperstimulation with associated fetal heart rate changes, • caesarean section (all), and • caesarean section for fetal distress.
Uterine hyperstimulation was defined as uterine tachysystole (with five or more contractions in a 10 minute period for two consecutive 10 minute periods) or uterine hypertonus (a uterine contraction lasting for more than two minutes) (Curtis 1987). The fetal heart rate changes that were recorded as abnormal included persistent decelerations (early, late or variable decelerations), fetal tachycardia (fetal heart rate greater than 160
61 beats per minute), fetal bradycardia (fetal heart rate less than 100 beats per minute), or reduced short term variability (less than five beats per minute) (RANZCOG 2001; RCOG 2001). The presence of uterine hyperstimulation in the presence of fetal heart rate changes was managed with intravenous salbutamol according to delivery suite protocols. Each induction of labour CTG was reviewed by a single investigator to maintain consistency in interpretation while blinded to the treatment allocated.
3.3(8)(ii) Secondary Study Outcomes Secondary study outcomes recorded related to evidence of effect, neonatal complications and maternal complications as described in the Cochrane generic protocol relating to methods of induction of labour (Table 3.3(6)(ii)) (Hofmeyr 2005a).
Secondary outcome Evidence of Effect Cervix unfavourable or unchanged after 24 hours
Further doses of known PGE2 gel Need for oxytocin infusion Induction to birth interval Labour and Birth Complications Uterine hyperstimulation without fetal heart rate changes (defined as uterine tachysystole – five or more contractions in a ten minute period for two or more consecutive ten minute periods; or uterine hypertonus – a uterine contraction lasting for more than two minutes) (Curtis 1987) Uterine rupture (defined as full thickness separation of the uterine muscle discovered at laparotomy; incidental scar dehiscence is not included) Need for analgesia (any analgesia, epidural analgesia) Meconium Stained Liquor Instrumental vaginal birth Postpartum haemorrhage (greater than 600mL, greater than 1000mL, need for blood transfusion) Neonatal Complications Birth weight less than 2.5kg Apgar score of less than 7 at five minutes Cord pH less than 7.18 Admission to neonatal intensive care unit Neonatal death (defined as death of a live born infant within 28 days of birth) Neonatal encephalopathy Maternal Complications Maternal side effects (nausea, vomiting, diarrhoea, flushing) Serious maternal complications (admission to intensive care unit, hyperpyrexia, coma) Maternal death
Table 3.3(8)(ii) Secondary outcomes
62 3.3(9) Data Collection
Data forms were completed by the midwife caring for the woman, with information confirmed and checked by the principal investigator or research midwife prior to the woman being discharged from hospital. All data were then entered into a database created in Access ’97 (Microsoft Corporation 1997) by the principal investigator.
3.3(10) Statistical Analysis
Data were analysed on an intention-to-treat basis and blind to the allocated treatment using the SAS Version 9.1 statistical software programme (SAS Institute Incorporated 2003). Baseline data were examined by treatment group to ensure comparability for prognostic factors, with correction for imbalances using log binomial regression techniques. The primary and secondary outcomes were compared using χ2 tests or Fisher’s Exact Test, with calculation of relative risks (RR) with 95% Confidence Intervals (CI) for dichotomous data. Normally distributed continuous data were compared using the Student’s T-test, and skewed data using non-parametric tests (Wilcoxon Rank Sum). Pre-specified subgroup analyses were planned to assess the effect of maternal parity (nulliparous versus multiparous) and initial Bishop’s score at commencing induction (Bishop’s score 0-3 versus Bishop’s score 4-6) on treatment effect. A p value of less than 0.05 was considered statistically significant.
3.3(11) Sample Size
Using information from the Cochrane review of oral misoprostol for induction of labour at term (Alfirevic 2005), and figures generated by the Women’s and Children’s Hospital Clinical Information Service, a sample size of 738 women was calculated to give 85% power to detect a 50% difference in the number of women undelivered after 24 hours from 20% in the vaginal prostaglandin group to 30% in the misoprostol group (p<0.05); a 32% difference in caesarean section from 28% in the vaginal prostaglandin group to 19% in the misoprostol group; and a 72% difference in the rate of hyperstimulation from 2% in the vaginal prostaglandin group to 6% in the misoprostol group (p<0.05; power 80%).
63 3.4 RESULTS
3.4(1) Baseline Characteristics
A total of 1319 eligible women were approached for trial participation by a clinical researcher at the time of booking the induction of labour, of whom 939 (71.2%) provided written consent to participate (Figure 3.4(1)). Of women who gave consent, 199 (21.2%) entered labour spontaneously prior to the date of booked induction of labour, and 741(78.8%) were admitted for induction of labour and randomised into the trial, 365 women to the oral misoprostol group, and 376 to the vaginal PGE2 group. The treatment allocated at randomisation was received by 740 (99.9%) of women, with one woman randomised to the vaginal prostaglandin E2 group receiving known prostaglandin E2 gel. Outcome data were available on all women to the time of discharge from hospital (Figure 3.4(1)).
Figure 3.4(1) Trial Flow Chart
Eligible Women Approached 1319
Eligible Women Consenting Eligible Women Declined 939 (71.2%) 380 (28.8%)
Spontaneous onset labour prior to scheduled induction 199 (21.2%)
Eligible Women Randomised 741 (78.8%)
Misoprostol Vaginal PGE2 365 women 376 women
Allocated Treatment Received Allocated Treatment Received 365 (100%) 375 (99.7%)
Outcome Data Available Outcome Data Available 365 (100%) 376 (100%)
64 Of the women participating in the trial, 59% were in their first ongoing pregnancy, and 97% were of caucasian background. The most common indication for induction of labour was post dates, accounting for 48% of all inductions, followed by hypertensive conditions (including pre-eclampsia), accounting for 23% of all inductions, and social indications (including maternal discomfort and maternal request), accounting for 14% of all inductions performed.
Baseline characteristics were comparable between the two treatment groups, including maternal age, parity, booking weight, smoking during pregnancy, gestational age, and reason for induction of labour. However, there were differences in the number of women with an initial Bishop’s score of 0-3 between the two treatment groups, and adjustments for this imbalance were made in the analyses using log binomial regression techniques, with the exception of those analyses exploring the effect of Bishop’s score (Table 3.4(1)).
Table 3.4(1) Comparison of baseline variables at trial entry
Misoprostol Group PGE2 Group N=365 % N=376 % Age (years)* 27.9 5.6 28.0 5.6 Nulliparous 213 58.4 221 58.8 Caucasian 355 97.3 362 96.3 Public patient 362 99.2 376 100.0 Indication for induction of labour - post dates 181 49.6 175 46.5 - pre-eclampsia 30 8.2 37 9.8 - hypertension 49 13.4 55 14.6 - intra-uterine growth restriction 24 6.6 32 8.5 - abnormal glucose tolerance 29 8.0 46 12.2 - social 55 15.1 46 12.2 - other 21 5.8 25 6.6 Initial Bishop’s score - Bishop’s score 0-3 246 67.4 214 56.9 - Bishop’s score 4-6 119 32.6 162 43.1 Mean Gestational Age (weeks)* 40.6 2.0 40.4 2.1 Height (m)* 165.1 5.8 165.0 6.6 Weight at booking (kg)# 72.0 48-96 75.0 49-101 Body Mass Index >30 (kg/m2) at booking 94 34.0 98 33.1 Smoking at booking 84 23.0 86 22.9
Figures are numbers and percentages. * = mean and standard deviation # = median and interquartile range
65 3.4(2) Primary Outcomes There were no statistically significant differences identified between oral misoprostol and vaginal PGE2 gel for vaginal birth not achieved in 24 hours (Misoprostol 168/365
(46.0%) versus PGE2 155/376 (41.2%); RR 1.12 95% CI 0.95-1.32; p=0.134), caesarean section (Misoprostol 83/365 (22.7%) versus PGE2 100/376 (26.6%); RR 0.82 95% CI 0.64-1.06; p=0.127), caesarean section performed for fetal distress (Misoprostol 32/365
(8.8%) versus PGE2 35/376 (9.3%); RR 0.91 95% CI 0.57-1.44; p=0.679), or uterine hyperstimulation with associated fetal heart rate changes (Misoprostol 3/365 (0.8%) versus PGE2 6/376 1.6%); RR 0.55 95% CI 0.14-2.21; p=0.401) (Table 3.4(2)).
Table 3.4(2) Primary Outcomes
Misoprostol PGE2 Group Treatment Effect P Group RR 95% CI value All women N=365 % N=376 % - Vaginal Birth not achieved 24 hours 168 46.0 155 41.2 1.12 0.95- 1.32 0.134 - Uterine HSS with FHR changes 3 0.8 6 1.6 0.55 0.14- 2.21 0.401 - Caesarean section - all 83 22.7 100 26.6 0.82 0.64- 1.06 0.127 - Caesarean section - fetal distress 32 8.8 35 9.3 0.91 0.57- 1.44 0.679
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. HSS = hyperstimulation syndrome FHR = fetal heart rate
66 3.4(3) Secondary Outcomes – Evidence of Effect
Women in the oral misoprostol group were significantly more likely to have a Bishop’s score of less than seven 24 hours after commencing the induction (Misoprostol 57/365
(15.6%) versus PGE2 39/376 (10.4%); RR 1.51 95% CI 1.03-2.20; p=0.031), were more likely to require further doses of known vaginal PGE2 gel (Misoprostol 70/365 (19.2%) versus PGE2 47/376 (12.5%); RR 1.51 95% CI 1.01-1.97; p=0.043), and were more likely to require an oxytocin infusion during labour (Misoprostol 203/365 (55.6%) versus PGE2 179/376 (47.6%); RR 1.17 95% CI 1.01-1.34; p=0.034) (Table 3.4(3)). For every 23 women treated with misoprostol, one woman requires further doses of vaginal
PGE2 gel (NNTH 23; 95% CI 12-4021), and for every 13 women treated with misoprostol, one woman requires oxytocin infusion during labour (NNTH 13; 95% CI 7-162). Women administered oral misoprostol had a longer induction to birth interval when compared with women administered vaginal PGE2 gel (Misoprostol median interval 21.2 hours, interquartile range 8.6-33.8 versus PGE2 median interval 18.4 hours, interquartile range 6.3-30.5; p<0.001).
Table 3.4(3) Secondary Outcomes – Evidence of Effect
Misoprostol Group PGE2 Group Treatment P Effect value RR 95% CI All women N=365 % N=376 % - Bishop’s Score <7 after 24 hours 57 15.6 39 10.4 1.51 1.03-2.20 0.031
- Further doses of known PGE2 70 19.2 47 12.5 1.41 1.01-1.97 0.043 - Oxytocin infusion 203 55.6 179 47.6 1.17 1.01-1.34 0.034 - Induction to birth interval* 21.2 8.6-33.8 18.4 6.3-30.5 <0.001
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. *Median and interquartile range (interval in hours); Wilcoxon Rank Sum Test
PGE2 = vaginal prostaglandin E2 gel
67 3.4(4) Secondary Outcomes – Labour and Birth Complications
There were no differences in the mean length of labour for women receiving oral misoprostol compared with women receiving vaginal PGE2 gel (Misoprostol mean interval 7.5 hours, standard deviation 4.1 versus PGE2 mean interval 6.9 hours, standard deviation 4.0; p=0.073). Women randomised to the oral misoprostol group were significantly less likely to have uterine hyperstimulation without associated fetal heart rate changes (Misoprostol 4/365 (1.1%) versus PGE2 17/376 (4.5%); RR 0.23 95% CI 0.08-0.69; p=0.009)(Table 3.4(4)).
Women in the oral misoprostol group were more likely to require analgesia during labour (Misoprostol 351/365 (96.2%) versus PGE2 347/376 92.3%); RR 1.04 95% CI
1.00-1.08; p=0.035) when compared with women in the PGE2 group (Table 3.4(4)). However, there were no statistically significant differences between women in the oral misoprostol group and those in the vaginal PGE2 gel group in the use of epidural analgesia (Misoprostol 243/365 (66.6%) versus PGE2 229/376 (60.9%); RR 1.08 95% CI 0.97-1.21; p=0.149), the presence of meconium stained liquor during labour
(Misoprostol 59/365 (16.2%) versus PGE2 52/376 (13.8%); RR 1.14 95% CI 0.81-1.61; p=0.465), instrumental vaginal birth (Misoprostol 57/365 (15.6%) versus PGE2 77/376 (20.5%); RR 1.06 95% CI 0.77-1.46; p=0.712), blood loss greater than 600mL
(Misoprostol 57/365 (15.6%) versus PGE2 77/376 (20.5%); RR 0.76 95% CI 0.55-1.04; p=0.081), blood loss greater than 1000mL (Misoprostol 17/365 (4.7%) versus PGE2 20/376 (5.3%); RR 0.86 95% CI 0.46-1.63; p=0.646), or need for blood transfusion
(Misoprostol 8/365 (2.2%) versus PGE2 9/376 (2.4%); RR 0.96 95% CI 0.37-2.47; p=0.927).
68 Table 3.4(4) Secondary Outcomes – Labour & Birth Complications
Misoprostol PGE2 Group Treatment Effect P value Group RR 95% CI All women N=365 % N=376 % - Uterine HSS - no FHR changes 4 1.1 17 4.5 0.23 0.08-0.69 0.009 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 351 96.2 347 92.3 1.04 1.00-1.08 0.035 - Need for epidural 243 66.6 229 60.9 1.08 0.97-1.21 0.149 - Meconium stained liquor 59 16.2 52 13.8 1.14 0.81-1.61 0.465 - Length of labour* 7.5 4.1 6.9 4.0 0.073 - Instrumental vaginal birth 65 17.8 63 16.8 1.06 0.77-1.46 0.712 - Blood loss >600mL 57 15.6 77 20.5 0.76 0.55-1.04 0.081 - Blood loss >1000mL 17 4.7 20 5.3 0.86 0.46-1.63 0.646 - Need for blood transfusion 8 2.2 9 2.4 0.96 0.37-2.47 0.927
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. * Mean and standard deviation (interval in hours); Student’s T-Test HSS = hyperstimulation syndrome FHR = fetal heart rate
3.4(5) Secondary Outcomes - Neonatal Complications
Infant birth weight did not differ significantly between the oral misoprostol group and the vaginal PGE2 gel group (Misoprostol 3533g + 534 versus PGE2 3540g + 490; p=0.840). Similarly, there were no significant differences between the oral misoprostol group and the vaginal PGE2 group for the number of infants with birth weight less than
2500g (Misoprostol 15/365 (4.1%) versus PGE2 11/376 (2.9%); RR 1.30 95% CI 0.60- 2.79; p=0.505), Apgar score of less than seven at five minutes of age (Misoprostol
2/365 (0.6%) versus PGE2 5/376 (1.3%); RR 0.42 95% CI 0.08-2.15; p=0.297), cord pH less than 7.18 (Misoprostol 10/365 (7.6%) versus PGE2 19/376 (11.6%); RR 0.62 95%
CI 0.30-1.29; p=0.205), or admission to NICU (Misoprostol 5/365 (1.4%) versus PGE2 2/376 (0.5%); RR 2.66 95% CI 0.52-13.75; p=0.242) (Table 3.4(5)). Of infants admitted to the neonatal intensive care unit, two were for major undiagnosed cardiac defects (one in the misoprostol group and one in the PGE2 group), three required intubation (two in the misoprostol group and one in the PGE2 group), one required blood transfusion
69 following intrapartum feto-maternal haemorrhage (misoprostol group) and one required treatment for pronounced hypoglycaemia (misoprostol group).
Table 3.4(5) Secondary Outcomes - Neonatal Complications
Misoprostol PGE2 Group Treatment Effect P value Group RR 95%CI All infants N=365 % N=376 % - Birth-weight <2.5kg 15 4.1 11 2.9 1.30 0.60- 2.79 0.505 - Apgar <7 @ 5 minutes 2 0.6 5 1.3 0.42 0.08- 2.15 0.297 - Cord pH <7.18 10 7.6 19 11.6 0.62 0.30- 1.29 0.205 - NICU admission 5 1.4 2 0.5 2.66 0.52-13.75 0.242 - Neonatal encephalopathy 0 0.0 0 0.0 Not estimable - Neonatal death* 0 0.0 0 0.0 Not estimable
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. *Neonatal Death = death of a live-born infant within 28 days of birth NICU = neonatal intensive care unit
3.4(6) Secondary Outcomes - Maternal Complications
In considering maternal experience of any side effects, there were no statistically significant differences between the oral misoprostol and vaginal PGE2 gel (Misoprostol
76/365 (20.8%) versus PGE2 99/376 (26.3%); RR 0.78 95% CI 0.60-1.01; p=0.063). Similarly, there were no differences in the number of women experiencing nausea
(Misoprostol 20/365 (5.5%) versus PGE2 30/376 (8.0%); RR 0.68 95% CI 0.39-1.19; p=0.175), vomiting (Misoprostol 4/365 (1.1%) versus PGE2 10/376 (2.7%); RR 0.43
95% CI 0.14-1.37; p=0.154), diarrhoea (Misoprostol 5/365 (1.4%) versus PGE2 9/376 (2.4%); RR 0.53 95% CI 0.18-1.57; p=0.250) or flushing (Misoprostol 6/365 (1.6%) versus PGE2 4/376 (1.1%); RR 1.43 95% CI 0.41-5.05; p=0.578) (Table 3.4(6)). No women in either group were admitted to intensive care, suffered hyperpyrexia or coma. There were no maternal deaths.
70 Table 3.4(6) Secondary Outcomes - Maternal Complications
Misoprostol PGE2 Group Treatment Effect P value Group RR 95% CI All women N=365 % N=376 % - Any side effect 76 20.8 99 26.3 0.78 0.60-1.01 0.063 - Nausea 20 5.5 30 8.0 0.68 0.39-1.19 0.175 - Vomiting 4 1.1 10 2.7 0.43 0.14-1.37 0.154 - Diarrhoea 5 1.4 9 2.4 0.53 0.18-1.57 0.250 - Flushing 6 1.6 4 1.1 1.43 0.41-5.05 0.578 - Intensive care unit admission 0 0.0 0 0.0 Not estimable - Hyperpyrexia 0 0.0 0 0.0 Not estimable - Coma 0 0.0 0 0.0 Not estimable - Maternal death 0 0.0 0 0.0 Not estimable
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect.
3.4(7) Pre-specified Secondary Analyses – Effect of Maternal Parity and Bishop’s Score at Trial Entry
3.4(7)(i) Distribution of maternal parity and initial Bishop’s score Of the 365 women randomised to the misoprostol group, 213 (58.4%) were nulliparous, of whom 143 (67.1%) had an initial Bishop’s score of 0-3, and 70 (32.9%) an initial Bishop’s score of 4-6. There were 152 (41.6%) multiparous women in the misoprostol group, of whom 103 (67.8%) had an initial Bishop’s score of 0-3, and 49 (32.3%) an initial Bishop’s score of 4-6 (Table 3.4(7)(i)).
Of the 376 women randomised to the vaginal PGE2 group, 221 (58.8%) were nulliparous, of whom 126 (57.0%) had an initial Bishop’s score of 0-3, and 95 (43.0%) an initial Bishop’s score of 4-6. There were 155 (41.2%) multiparous women in the vaginal PGE2 group, of whom 88 (56.8%) had an initial Bishop’s score of 0-3, and 67 (43.2%) an initial Bishop’s score of 4-6 (Table 3.4(7)(i)).
71 Table 3.4(7)(i) Distribution of maternal parity and initial Bishop’s score
Misoprostol Group PGE2 Group N=365 % N=376 % Bishop’s Score - Bishop’s score 0-3 246 67.4 214 56.9 - Bishop’s score 4-6 119 32.6 162 43.1 Nulliparous women 213 58.4 376 58.8 - Bishop’s score 0-3 143 67.1 126 57.0 - Bishop’s score 4-6 70 32.9 95 43.0 Multiparous women 152 41.6 155 41.2 - Bishop’s score 0-3 103 67.8 88 56.8 - Bishop’s score 4-6 49 32.3 67 43.2
Figures are numbers and percentages
3.4(7)(ii) Effect of parity and initial Bishop’s score on primary trial outcomes For the outcomes of vaginal birth not achieved within 24 hours, uterine hyperstimulation with associated fetal heart rate changes, caesarean section and caesarean section for fetal distress, there were no statistically significant differences identified on the basis of maternal parity (nulliparous versus multiparous) or the Bishop’s score at trial entry (Bishop’s score 0-3 versus Bishop’s score 4-6) (Table 3.4(7)(ii)). The magnitude of treatment effect was similar across all of the pre-specified subgroups.
72 Table 3.4(7)(ii) Secondary Analysis - Primary Outcomes
Misoprostol Group PGE2 Group Treatment Effect P value RR 95% CI All women N=365 % N=376 % - Vaginal Birth not achieved in 24 hours 168 46.0 155 41.2 1.12 0.95- 1.32 0.134 - Uterine HSS with FHR changes 3 0.8 6 1.6 0.55 0.14- 2.21 0.401 - Caesarean section - all 83 22.7 100 26.6 0.82 0.64- 1.06 0.127 - Caesarean section - fetal distress 32 8.8 35 9.3 0.91 0.57- 1.44 0.679 Bishop’s score 0-3 N=246 % N=214 % - Vaginal Birth not achieved in 24 hours 127 51.6 104 48.6 1.06 0.88-1.28 0.210 - Caesarean section – all 61 24.8 67 31.3 0.79 0.59-1.06 0.120 - Caesarean section – fetal distress 24 9.8 23 10.8 0.91 0.53-1.56 0.726 Bishop’s score 4-6 N=119 % N=162 % - Vaginal Birth not achieved in 24 hours 41 34.5 51 31.5 1.09 0.78-1.53 0.234 - Caesarean section – all 22 18.5 33 20.4 0.91 0.56-1.47 0.694 - Caesarean section – fetal distress 8 6.7 12 7.4 0.91 0.38-2.15 0.825 Nulliparous women N=213 % N=221 % - Vaginal Birth not achieved in 24 hours 133 62.4 119 53.8 1.16 0.99- 1.36 0.157 - Caesarean section - all 67 31.5 81 36.7 0.83 0.64- 1.07 0.155 - Caesarean section – fetal distress 22 10.3 28 12.7 0.79 0.47- 1.33 0.375 Nulliparous women, BS 0-3 N=143 % N=126 % - Vaginal Birth not achieved in 24 hours 101 70.6 80 63.5 1.11 0.94-1.32 0.145 - Caesarean section - all 49 34.3 53 42.1 0.81 0.60- 1.11 0.188 - Caesarean section – fetal distress 17 11.9 18 14.3 0.83 0.45- 1.54 0.600 Nulliparous women, BS 4-6 N= 70 % N= 95 % - Vaginal Birth not achieved in 24 hours 32 45.7 39 41.1 1.11 0.78- 1.58 0.136 - Caesarean section - all 18 25.7 28 29.5 0.87 0.53- 1.45 0.595 - Caesarean section – fetal distress 5 7.1 10 10.5 0.68 0.24- 1.90 0.455 Multiparous women N=152 % N=155 % - Vaginal Birth not achieved in 24 hours 35 23.0 36 23.2 0.99 0.66- 1.49 0.790 - Caesarean section – all 16 10.5 19 12.3 0.82 0.44- 1.54 0.542 - Caesarean section – fetal distress 10 6.6 7 4.5 1.40 0.55- 3.58 0.485 Multiparous women, BS 0-3 N=103 % N= 88 % - Vaginal Birth not achieved in 24 hours 26 25.2 24 27.3 0.93 0.57- 1.49 0.643 - Caesarean section – all 12 11.7 14 15.9 0.73 0.36- 1.50 0.392 - Caesarean section – fetal distress 7 6.8 5 5.7 1.20 0.39- 3.64 0.752 Multiparous women, BS 4-6 N= 49 % N= 67 % - Vaginal Birth not achieved in 24 hours 9 18.4 12 17.9 1.03 0.47- 2.24 0.966 - Caesarean section – all 4 8.2 5 7.5 1.09 0.31- 3.86 0.889 - Caesarean section – fetal distress 3 6.1 2 3.0 2.05 0.36-11.81 0.411
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. BS = Bishop’s Score; HSS = hyperstimulation syndrome; FHR = fetal heart rate
73 3.4(7)(iii) Secondary Analysis – Secondary Outcomes of Evidence of Effect Nulliparous women administered misoprostol, when compared with those administered vaginal PGE2 gel, were significantly more likely to have a Bishop’s score of less than seven 24 hours after the induction commenced (Misoprostol 46/213 (21.6%) versus
PGE2 26/221 (11.8%); RR 1.69 95% CI 1.09-2.64; p=0.024), and to require further doses of vaginal PGE2 gel (Misoprostol 57/213 (26.8%) versus PGE2 27/221 (12.2%); RR 2.15 95% CI 1.35-3.12; p<0.001) (Table 3.4(7)(iii)). These effects were confined to those nulliparous women with a Bishop’s score of 0-3 at the commencement of the induction process (Bishop’s score of less than seven 24 hours after induction commenced (Misoprostol 39/143 (27.3%) versus PGE2 20/126 (15.9%); RR 1.72 95%
CI 1.06-2.78; p=0.022), and need for further doses of vaginal PGE2 gel (Misoprostol
48/143 (33.6%) versus PGE2 19/126 (15.1%); RR 2.23 95% CI 1.39-3.58; p<0.001).
Women administered misoprostol had a longer induction to birth interval when compared with women in the vaginal PGE2 gel group, an effect apparent in women with an initial Bishop’s score of 0-3 (Misoprostol median interval 22.3 hours, interquartile range 13.0 versus PGE2 median interval 20.3 hours, interquartile range 12.3; p=0.039), nulliparous women (Misoprostol median interval 24.5 hours, interquartile range 11.1 versus PGE2 median interval 20.9 hours, interquartile range 12.2; p<0.001), and nulliparous women with an initial Bishop’s score of 0-3 (Misoprostol median interval
25.8 hours, interquartile range 11.3 versus PGE2 median interval 22.8 hours, interquartile range 12.9; p<0.001).
The increased need for oxytocin infusion during labour was confined to multiparous women administered oral misoprostol (Misoprostol 64/152 (42.1%) versus PGE2 48/155 (42.1%); RR 1.36 95% CI 1.00-1.83; p=0.049), regardless of the woman’s cervical score at the start of the induction. However, multiparous women with an initial Bishop’s score of 0-3 in the misoprostol group were less likely to require further doses of vaginal
PGE2 gel (Misoprostol 10/103 (9.7%) versus PGE2 18/88 (20.5%); RR 0.47 95% CI 0.23-0.97; p=0.036).
74 Table 3.4(7)(iii) Secondary Analysis – Evidence of Effect
Misoprostol Group PGE2 Group Treatment Effect P value RR 95%CI All women N=365 % N=376 % - Bishop’s Score <7 after 24 hours 57 15.6 39 10.4 1.51 1.03- 2.20 0.031 - Further doses of known PGE2 70 19.2 47 12.5 1.41 1.01- 1.97 0.043 - Oxytocin infusion 203 55.6 179 47.6 1.17 1.01- 1.34 0.034 - Induction to birth interval* 21.2 8.6-33.7 18.4 6.3-30.5 <0.001 Bishop’s score 0-3 N=246 % N=214 % - Bishop’s score <7 after 24 hours 48 19.5 31 14.5 1.09 0.78-1.52 0.606 - Further doses of known PGE2 58 23.6 37 17.3 1.36 0.94-1.97 0.097 - Oxytocin infusion 137 55.7 104 48.6 1.15 0.96-1.37 0.129 - Induction to birth interval* 22.3 9.3-35.3 20.3 8.0-32.6 0.039 Bishop’s score 4-6 N=119 % N=162 % - Bishop’s score <7 after 24 hours 9 7.6 8 4.9 1.02 0.45-2.29 0.963 - Further doses of known PGE2 12 10.1 10 6.2 1.63 0.73-3.65 0.228 - Oxytocin infusion 66 55.5 75 46.3 1.20 0.95-1.51 0.129 - Induction to birth interval* 19.3 9.0-29.6 17.3 7.4-27.2 0.060 Nulliparous women N=213 % N=221 % - Bishop’s Score <7 after 24 hours 46 21.6 26 11.8 1.69 1.09- 2.64 0.024 - Further doses of known PGE2 57 26.8 27 12.2 2.05 1.35- 3.12 <0.001 - Oxytocin infusion 139 65.3 131 59.3 1.10 0.95- 1.27 0.220 - Induction to birth interval* 24.5 13.4-35.6 20.9 8.7-33.1 <0.001 Nulliparous women, BS 0-3 N=143 % N=126 % - Bishop’s Score <7 after 24 hours 39 27.3 20 15.9 1.72 1.06- 2.78 0.022 - Further doses of known PGE2 48 33.6 19 15.1 2.23 1.39- 3.58 <0.001 - Oxytocin infusion 96 67.1 77 61.1 1.09 0.92- 1.32 0.304 - Induction to birth interval* 25.8 14.5-37.1 22.8 9.9-35.7 <0.001 Nulliparous women, BS 4-6 N= 70 % N= 95 % - Bishop’s Score <7 after 24 hours 7 10.0 6 6.3 1.58 0.56- 4.51 0.9658 - Further doses of known PGE2 9 12.9 8 8.4 1.53 0.62- 3.76 0.3542 - Oxytocin infusion 43 61.4 54 56.8 1.08 0.84- 1.40 0.5542 - Induction to birth interval* 21.7 11.5-31.9 18.6 7.6-29.6 0.0258 Multiparous women N=152 % N=155 % - Bishop’s Score <7 after 24 hours 11 7.2 13 8.4 0.86 0.40- 1.87 0.464 - Further doses of known PGE2 13 8.6 20 12.9 0.59 0.30- 1.14 0.115 - Oxytocin infusion 64 42.1 48 31.0 1.36 1.00- 1.83 0.049 - Induction to birth interval* 15.3 5.0-25.6 16.6 6.2-27.0 0.232 Multiparous women, BS 0-3 N=103 % N= 88 % - Bishop’s Score <7 after 24 hours 9 8.7 11 12.5 0.70 0.30- 1.61 0.234 - Further doses of known PGE2 10 9.7 18 20.5 0.47 0.23- 0.97 0.036 - Oxytocin infusion 41 39.8 27 30.7 1.29 0.88- 1.92 0.189 - Induction to birth interval* 16.4 5.8-27.0 17.8 5.7-29.9 0.466 Multiparous women, BS 4-6 N= 49 % N= 67 % - Bishop’s Score <7 after 24 hours 2 4.1 2 3.0 1.37 0.20- 9.37 0.876 - Further doses of known PGE2 3 6.1 2 3.0 2.05 0.36-11.81 0.411 - Oxytocin infusion 23 46.9 21 31.3 1.50 0.94- 2.38 0.087 - Induction to birth interval* 14.9 5.0-24.8 14.8 6.1-23.5 0.909
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. * Median and interquartile range (interval in hours); Wilcoxon Rank Sum Test BS = Bishop’s Score
PGE2 = vaginal prostaglandin E2 gel
75 3.4(7)(iv) Secondary Analysis –Labour & Birth Complications Nulliparous women randomised to the oral misoprostol group were significantly less likely to have uterine hyperstimulation without associated fetal heart rate changes when compared with women randomised to the PGE2 group (Misoprostol 3/213 (1.4%) versus
PGE2 15/221 (6.8%); RR 0.20 95% CI 0.06-0.69; p=0.011), as were those women with an initial Bishop’s score of 0-3 (Misoprostol 3/246 (1.2%) versus PGE2 11/214 (5.1%); RR 0.24 95% CI 0.07-0.84; p=0.015) (Table 3.4(7)(iv)). This effect was greatest for nulliparous women with a Bishop’s score of 0-3 (Misoprostol 2/143 (1.4%) versus
PGE2 9/126 (7.1%); RR 0.20 95% CI 0.04-0.89; p=0.018). Multiparous women in the misoprostol group were less likely to require an instrumental vaginal birth when compared with women in the vaginal PGE2 group (Misoprostol 7/152 (4.6%) versus
PGE2 17/155 (11.0%); RR 0.42 95% CI 0.18-0.98; p=0.046). The mean length of labour did not differ significantly between women administered oral misoprostol and those administered vaginal PGE2 gel, with no interaction on the basis of maternal parity or initial Bishop’s score.
Table 3.4(7)(iv) Secondary Analysis – Labour & Birth Complications
Misoprostol Group PGE2 Group Treatment Effect P value RR 95% CI All women N=365 % N=376 % - Uterine HSS - no FHR changes 4 1.1 17 4.5 0.23 0.08-0.69 0.009 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 351 96.2 347 92.3 1.04 1.00-1.08 0.035 - Need for epidural 243 66.6 229 60.9 1.08 0.97-1.21 0.149 - Meconium stained liquor 59 16.2 52 13.8 1.14 0.81-1.61 0.465 - Length of labour* 7.5 4.1 6.9 4.0 0.073 - Instrumental vaginal birth 65 17.8 63 16.8 1.06 0.77-1.46 0.712 - Blood loss >600mL 57 15.6 77 20.5 0.76 0.55-1.04 0.081 - Blood loss >1000mL 17 4.7 20 5.3 0.86 0.46-1.63 0.646 - Need for blood transfusion 8 2.2 9 2.4 0.96 0.37-2.47 0.927 Bishop’s Score 0-3 N=246 % N=214 % - Uterine HSS - no FHR changes 3 1.2 11 5.1 0.24 0.07-0.84 0.015 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 236 95.9 200 93.5 1.03 0.98-1.07 0.233 - Need for epidural 163 66.3 137 64.0 1.04 0.90-1.18 0.615 - Meconium stained liquor 43 17.5 33 15.4 1.13 0.75-1.72 0.553 - Length of labour* 7.5 4.1 7.0 3.8 0.266 - Instrumental vaginal birth 14 17.9 36 16.8 1.06 0.71-1.59 0.764 - Blood loss >600mL 37 15.0 47 22.0 0.68 0.46-1.01 0.055 - Blood loss >1000mL 10 4.1 14 6.5 0.62 0.28-1.37 0.233 - Need for blood transfusion 3 1.2 6 2.8 0.44 0.11-1.72 0.221
76 Misoprostol Group PGE2 Group Treatment Effect P value RR 95% CI Bishop’s Score 4-6 N=119 % N=162 % - Uterine HSS - no FHR changes 1 0.8 6 3.7 0.22 0.03-1.86 0.128 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 115 96.6 147 90.7 1.07 1.00-1.13 0.052 - Need for epidural 80 67.2 92 56.8 1.18 0.98-1.42 0.076 - Meconium stained liquor 16 13.5 19 11.7 1.15 0.62-2.13 0.667 - Length of labour* 7.5 3.9 6.7 4.2 0.139 - Instrumental vaginal birth 21 17.7 27 16.7 1.06 0.63-1.78 0.829 - Blood loss >600mL 20 16.8 30 18.5 0.91 0.54-1.52 0.711 - Blood loss >1000mL 7 5.9 6 3.7 1.59 0.55-4.60 0.390 - Need for blood transfusion 5 4.2 3 1.9 2.27 0.55-9.31 0.242 Nulliparous women N=213 % N=221 % - Uterine HSS - no FHR changes 3 1.4 15 6.8 0.20 0.06-0.69 0.011 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 210 98.6 212 95.9 1.03 0.99-1.08 0.113 - Need for epidural 159 74.7 156 70.6 1.05 0.93-1.18 0.411 - Meconium stained liquor 40 18.8 32 13.8 1.27 0.83-1.94 0.279 - Length of labour* 9.2 4.0 8.5 4.2 0.566 - Instrumental vaginal birth 58 27.2 46 20.8 1.31 0.93-1.83 0.121 - Blood loss >600mL 38 17.8 56 25.3 0.70 0.48-1.01 0.056 - Blood loss >1000mL 11 5.2 10 4.5 1.12 0.49-2.60 0.783 - Need for blood transfusion 3 1.4 6 2.7 0.54 0.14-2.14 0.380 Nulliparous women, BS 0-3 N=143 % N=126 % - Uterine HSS - no FHR changes 2 1.4 9 7.1 0.20 0.04-0.89 0.018 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 140 97.9 121 96.0 1.02 0.98-1.06 0.368 - Need for epidural 106 74.1 92 73.0 1.02 0.89-1.17 0.837 - Meconium stained liquor 30 21.0 20 15.9 1.32 0.79-2.21 0.283 - Length of labour* 9.2 3.9 8.7 4.1 0.153 - Instrumental vaginal birth 40 28.0 23 18.3 1.53 0.97-2.41 0.060 - Blood loss >600mL 26 18.2 35 27.8 0.65 0.42-1.02 0.061 - Blood loss >1000mL 8 5.6 7 5.6 1.01 0.38-2.70 0.989 - Need for blood transfusion 3 2.1 4 3.2 0.66 0.15-2.90 0.580 Nulliparous women, BS 4-6 N= 70 % N= 95 % - Uterine HSS - no FHR changes 1 1.4 6 6.3 0.23 0.03-1.84 0.124 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 70 100.0 91 95.8 1.04 1.00-1.09 0.082 - Need for epidural 53 75.7 64 67.4 1.12 0.93-1.36 0.243 - Meconium stained liquor 10 14.3 12 12.6 1.13 0.52-2.47 0.757 - Length of labour* 8.7 4.2 8.3 4.2 0.601 - Instrumental vaginal birth 18 25.7 23 24.2 1.06 0.26-1.81 0.825 - Blood loss >600mL 12 17.1 21 22.1 0.78 0.41-1.47 0.431 - Blood loss >1000mL 3 4.3 3 3.2 1.36 0.28-6.52 0.702 - Need for blood transfusion 0 0.0 2 2.1 0.27 0.01-5.55 0.222 Multiparous women N=152 % N=155 % - Uterine HSS - no FHR changes 1 0.7 2 1.3 0.49 0.05- 5.38 0.561 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 141 92.8 135 87.1 1.07 0.99- 1.15 0.087 - Need for epidural 84 55.3 73 47.1 1.17 0.94- 1.45 0.173 - Meconium stained liquor 19 12.5 20 12.9 0.94 0.52- 1.69 0.834 - Length of labour* 5.6 3.1 5.2 3.0 0.825 - Instrumental vaginal birth 7 4.6 17 11.0 0.42 0.18- 0.98 0.046 - Blood loss >600mL 19 12.5 21 13.6 0.91 0.51- 1.63 0.764 - Blood loss >1000mL 6 4.0 10 6.5 0.60 0.22- 1.62 0.313 - Need for blood transfusion 5 3.3 3 1.9 1.79 0.43- 7.41 0.420 Multiparous women, BS 0-3 N=103 % N= 88 % - Uterine HSS - no FHR changes 1 1.0 2 2.3 0.43 0.04- 4.63 0.471 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 96 93.2 79 89.8 1.04 0.95- 1.13 0.394 - Need for epidural 57 55.3 45 51.1 1.08 0.83- 1.41 0.562 - Meconium stained liquor 13 12.6 13 14.8 0.85 0.42- 1.75 0.666 - Length of labour* 5.3 3.1 5.3 2.6 0.850 - Instrumental vaginal birth 4 3.9 13 14.8 0.26 0.09- 0.78 0.008 - Blood loss >600mL 11 10.7 12 13.6 0.78 0.36- 169 0.531 - Blood loss >1000mL 2 1.9 7 8.0 0.24 0.05- 1.14 0.051 - Need for blood transfusion 0 0.0 2 2.3 0.17 0.01- 3.52 0.124
77 Misoprostol Group PGE2 Group Treatment Effect P value RR 95% CI Multiparous women, BS 4-6 N= 49 % N= 67 % - Uterine HSS - no FHR changes 0 0.0 0 0.0 Not estimable - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 45 91.8 56 83.6 1.10 0.96- 1.23 0.191 - Need for epidural 27 55.1 28 41.8 1.32 0.90- 1.93 0.156 - Meconium stained liquor 6 12.2 7 10.5 1.17 0.42- 3.27 0.762 - Length of labour* 6.2 3.3 5.0 3.5 0.068 - Instrumental vaginal birth 3 6.1 4 6.0 1.03 0.24- 4.38 0.973 - Blood loss >600mL 8 16.3 9 13.4 1.22 0.51- 2.92 0.663 - Blood loss >1000mL 4 8.2 3 4.5 1.82 0.43- 7.78 0.410 - Need for blood transfusion 5 10.2 1 1.5 6.83 0.82-56.68 0.364
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text. * Mean and standard deviation (interval in hours); Student’s T-Test HSS = hyperstimulation syndrome FHR = fetal heart rate
3.4(7)(v) Secondary Analysis – Neonatal Complications There was no correlation in infant outcomes based on maternal parity or initial cervical score.
3.4(7)(vi) Secondary Analysis – Maternal Complications Women with an initial Bishop’s score of 0-3 who were administered misoprostol were less likely to experience nausea when compared with women administered PGE2
(Misoprostol 11/246 (4.5%) versus PGE2 20/214 (9.4%); RR 0.48 95% CI 0.23-0.98; p=0.038). Nulliparous women administered misoprostol were less likely to experience any side effect (Misoprostol 39/213 (18.3%) versus PGE2 63/221 (28.5%); RR 0.63
95% CI 0.45-0.90; p=0.011), or vomiting when compared with PGE2 (Misoprostol
1/213 (1.1%) versus PGE2 9/221 (4.1%); RR 0.12 95% CI 0.02-0.95; p=0.044), but otherwise there was no correlation between experience of side effects and maternal parity.
78 Table 3.4(7)(vi) Secondary Analysis – Maternal Complications
Misoprostol PGE2 Group Treatment Effect P value Group RR 95% CI All women N=365 % N=376 % - Any side effect 76 20.8 99 26.3 0.78 0.60- 1.01 0.063 - Nausea 20 5.5 30 8.0 0.68 0.39- 1.19 0.175 - Vomiting 4 1.1 10 2.7 0.43 0.14- 1.37 0.154 - Diarrhoea 5 1.4 9 2.4 0.53 0.18- 1.57 0.250 - Flushing 6 1.6 4 1.1 1.43 0.41- 5.05 0.578 Bishop’s Score 0-3 N=246 % N=214 % - Any side effect 55 22.4 59 27.6 0.81 0.59- 1.12 0.197 - Nausea 11 4.5 20 9.4 0.48 0.23- 0.98 0.038 - Vomiting 2 0.8 5 2.3 0.35 0.07- 1.78 0.183 - Diarrhoea 4 1.6 7 3.3 0.50 0.15- 1.67 0.294 - Flushing 5 2.0 3 1.4 1.45 0.35- 6.00 0.606 Bishop’s Score 4-6 N=119 % N=162 % - Any side effect 21 17.7 40 24.7 0.71 0.45- 1.15 0.157 - Nausea 9 7.6 10 6.2 1.22 0.51- 2.92 0.647 - Vomiting 2 1.7 5 3.1 0.54 0.11- 2.76 0.455 - Diarrhoea 1 0.8 2 1.2 0.68 0.06- 7.42 0.751 - Flushing 1 0.8 1 0.6 1.36 0.09-21.55 0.826 Nulliparous women N=213 % N=221 % - Any side effect 39 18.3 63 28.5 0.63 0.45- 0.90 0.011 - Nausea 11 5.2 19 8.6 0.60 0.29- 1.23 0.162 - Vomiting 1 0.5 9 4.1 0.12 0.02- 0.94 0.044 - Diarrhoea 4 1.9 6 2.7 0.64 0.18- 2.24 0.487 - Flushing 3 1.4 2 0.9 1.45 0.24- 8.58 0.686 Multiparous women N=152 % N=155 % - Any side effect 37 24.3 36 23.2 1.03 0.69- 1.54 0.880 - Nausea 9 5.9 11 7.1 0.83 0.35- 1.95 0.669 - Vomiting 3 2.0 1 0.7 3.21 0.34-30.58 0.311 - Diarrhoea 1 0.7 3 1.9 0.31 0.03- 2.97 0.311 - Flushing 3 2.0 2 1.3 1.41 0.24- 8.34 0.704
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text.
3.5 DISCUSSION
This randomised trial is the second double blind, placebo controlled study comparing oral misoprostol with vaginal PGE2 gel, and the first involving low dose oral misoprostol solution for the induction of labour at term. The importance of blinding participants and caregivers to treatment allocation in randomised trials has been well described (Chalmers1989; Schulz 1995; Kunz 1998). The blinding of women, caregivers and outcome assessors employed in this trial reduced the potential for bias and increases confidence in the validity of these results. The inclusion criteria represented the spectrum of indications for induction of labour, and with over 70% of women approached agreeing to participate in the trial, these results have external
79 validity and are applicable to the general obstetric population requiring induction of labour at term.
This trial has the largest sample size to date of the randomised trials described comparing oral misoprostol and vaginal PGE2 gel, and was adequately powered to detect clinically important differences in caesarean birth and vaginal birth not achieved within 24 hours. This trial, however, was powered to detect relatively large differences in the less common outcome of uterine hyperstimulation with associated fetal heart rate changes (72% difference from 2% in the vaginal PGE2 group to 6% in the oral misoprostol group). For the rare, but serious maternal and neonatal complications, such as death and morbidity, including maternal uterine rupture and neonatal acidosis, the study was under-powered to detect all but large differences. Given the low frequency of occurrence of these serious adverse outcomes, it would be necessary to recruit into a trial tens of thousands of women and their infants, an extremely unlikely scenario to be successfully achieved in a single study. Nevertheless, the results from this study provide reliable evidence from a randomised controlled trial on the use of oral misoprostol for the induction of labour at term, can be included into a meta-analysis with other randomised controlled trials in future systematic reviews of the literature, and contribute to the available information about the safety of oral misoprostol for rare but serious maternal and infant adverse events.
For the primary study outcomes of vaginal birth not achieved in 24 hours, caesarean birth, and uterine hyperstimulation syndrome with associated fetal heart rate changes, there were no statistically significant differences between oral misoprostol and vaginal
PGE2 gel.
In considering the secondary outcomes relating to evidence of effect, oral misoprostol was less effective than vaginal PGE2 gel, as indicated by more women with a Bishop’s score of less than seven 24 hours after the induction commenced, more women requiring further doses of known prostaglandin E2 gel, more women requiring an oxytocin infusion during labour (including either induction or augmentation of labour), and a longer induction to birth interval. However, there were no differences in the mean length of labour for these women, suggesting that progress of labour, once established, occurred at a similar rate, regardless of the induction agent utilised.
80
The findings from this trial indicate no difference in maternal side effects or serious maternal complications in women administered oral misoprostol when compared with women administered vaginal PGE2 gel. Similarly there was a low incidence of neonatal Apgar score of less than five at seven minutes of age and cord pH of less than 7.18, outcomes that have traditionally been used as surrogate markers of the more severe outcome of neonatal acidemia. Any modification or increase in the dose of oral misoprostol, while improving clinical efficacy, must be balanced against a potential increase in these unwanted adverse outcomes.
The secondary analyses were conducted to explore the effects of maternal parity (nulliparous versus multiparous) and initial Bishop’s score (Bishop’s score 0-3 versus Bishop’s score 4-6) on the primary and secondary study outcomes. There was no apparent effect on the primary study outcomes for women with an initial Bishop’s score of 0-3, nulliparous women, and nulliparous women with an initial Bishop’s score of 0-3. There were more women in the misoprostol group with an initial Bishop’s score of 0-3, and even though corrections were made in the subsequent analyses, there may be potential for the introduction of bias into these results, tending towards more favourable reporting of outcomes in the vaginal PGE2 gel group.
The risk of caesarean section was not associated with maternal parity (nulliparous versus multiparous) or Bishop’s score (Bishop’s score 0-3 versus Bishop’s score 4-6). However, for all women with an extremely unfavourable cervix (Bishop’s score 0-3), induction of labour was associated with an increased risk of caesarean birth (28%), when compared with women with a more favourable cervix (Bishop’s score 4-6) (20%), highlighting the considerable risks associated with induction of labour in the presence of an unfavourable cervix. The implications of caesarean birth relate not only to the increased risk of complications in the index pregnancy when compared with vaginal birth, but also to future childbearing and subsequent mode of birth in the presence of a uterine scar. Caution should therefore be exercised when contemplating induction of labour in women with an extremely unfavourable cervix, to ensure that there is no opportunity to delay the procedure to allow for spontaneous improvement in the cervical score.
81 Nulliparous women administered oral misoprostol, and particularly those nulliparous women with a highly unfavourable cervix (Bishop’s score of 0-3), were more likely to still have an unfavourable cervix (Bishop’s score of less than seven) 24 hours after the induction commenced, and to receive further doses of vaginal PGE2 gel. Nulliparous women, regardless of initial Bishop’s score had a longer induction to birth interval, although there were no differences in the mean length of labour between women in the oral misoprostol group and those in the vaginal PGE2 gel group, again suggesting that the longer induction to birth interval reflects an increase in the duration of time prior to establishing in labour, rather than a prolongation of labour itself.
Fewer nulliparous women administered oral misoprostol experienced hyperstimulation syndrome without associated fetal heart rate changes when compared with those women administered vaginal PGE2 gel, an effect again most pronounced for nulliparous women with an initial Bishop’s score of 0-3, and suggestive of the possible need for an increase in the dose administered to this group of women.
The findings of reduced efficacy associated with misoprostol raise the possibility that the dosing regimen of 20mcg at two hourly intervals used in this trial may have been inadequate. Women administered misoprostol had a significant reduction of 79% in the occurrence of uterine hyperstimulation without associated fetal heart rate changes, and while this may represent a true reduction in the incidence of side effects with misoprostol, in the context of evidence of reduced efficacy it is more consistent with an inappropriately low dose administered. In this regard, an incremental increase in dose to 40mcg after four hours in the absence of uterine activity as described by Hofmeyr (2001) and later Dallenbach (2003) may be more appropriate.
For multiparous women, the dose of 20mcg at 2 hourly intervals is as effective as vaginal PGE2 gel in that similar numbers of women in each group did not achieve vaginal birth in 24 hours, showed no evidence of cervical change, or required additional doses of vaginal PGE2 gel. Multiparous women administered misoprostol were less likely to require an instrumental vaginal birth, and while there was an increased need for oxytocin infusion, there was a reduction in the need for further doses of vaginal
PGE2 gel after 12 hours. Importantly, there was no documented increase in potential
82 adverse effects such as uterine hyperstimulation with associated fetal heart rate changes, or caesarean section.
These findings are in keeping with the clinical observation of increased ease in inducing labour in multiparous women compared with nulliparous women, and accepted differences in the dose of PGE2 vaginal gel used on the basis of a woman’s parity. The vaginal gel dosing regimen adopted in this trial is consistent with current South
Australian Perinatal Guidelines recommending a dose of 1mg vaginal PGE2 be used in multiparous women and 2mg vaginal PGE2 in nulliparous women (Government SA 2005). In this current trial, an identical dose of oral misoprostol of 20mcg at two hourly intervals was used for both multiparous and nulliparous women.
The outcome vaginal birth not achieved in 24 hours as defined in the Cochrane generic protocol (Hofmeyr 2005a) comprises both women who birth vaginally beyond 24 hours and all women who birth by caesarean section. These two components reflect different processes whereby misoprostol may generate adverse health outcomes. If vaginal birth is achieved beyond 24 hours, this may reflect an inappropriately low dose of misoprostol as suggested in this study by other outcomes related to clinical efficacy. The alternative process contributing to adverse health outcomes relates to uterine hyperstimulation with fetal heart rate changes and subsequent birth by caesarean section for non-reassuring fetal heart rate trace. The results of this study indicate a relative reduction in caesarean birth with the use of misoprostol of the order of 18%. For completeness and to ensure clarity of information, future trials should report both components of this outcome.
3.6 CONCLUSION
Oral misoprostol was not associated with differences in the number of women who achieve vaginal birth in 24 hours, uterine hyperstimulation with fetal heart rate changes, or caesarean section, compared with vaginal PGE2 gel. Its use was associated with more women having an unfavourable cervix 24 hours after commencing the induction, an increased need for further doses of vaginal PGE2 gel and oxytocin infusion, but a significant reduction in uterine hyperstimulation without fetal heart rate changes.
83 Importantly, these differences in the process of induction of labour were not associated with an increase in the risk adverse health outcomes for the woman or her infant.
An initial Bishop’s score of 0-3 and nulliparity were both independently associated with greater difficulty in inducing labour, particularly when using misoprostol. However, these differences in the process of initiating labour were not associated with an increase in the risk of adverse health outcomes for the woman or her infant. For nulliparous women, and particularly those with an extremely unfavourable cervix (as indicated by a Bishop’s score of 0-3), a dose of 20mcg of misoprostol at two hourly intervals is less effective when compared with the standard vaginal PGE2 gel. Further information is required about the lowest dose of misoprostol that will attain clinical effect for this group of women who require induction of labour. Any further increase in the oral dose administered, while generating beneficial effects of improved clinical efficacy, must be balanced against the potential increase in unwanted side effects for the woman, and uterine hyperstimulation, both with and without fetal heart rate changes. While there is no apparent difference in the length of labour for this group of women, additional information is required about women’s satisfaction, particularly those women who remain undelivered after 24 hours.
The dose of 20mcg of misoprostol at two hourly intervals utilised in this trial is as effective as the standard vaginal PGE2 gel for multiparous women requiring induction of labour, and should be available for use within institutions.
Misoprostol is not associated with fewer women achieving vaginal birth within 24 hours, or an increased risk of caesarean section, uterine hyperstimulation with fetal heart rate changes, or adverse health outcomes for women or their infants. In view of the relative safety of this low dose of misoprostol in the induction of labour, women’s satisfaction, and particularly for those women not achieving vaginal birth within 24 hours, needs to be incorporated into any future clinical recommendation regarding the widespread or routine use of misoprostol.
84 4. WOMEN’S PREFERENCES FOR CARE
4.1 INTRODUCTION
While induction of labour is common, there is limited information available relating to women’s experiences and preferences for care. There is a current shift in health care policy and practice to incorporate women’s preferences for health related outcomes, with the aim of providing more individualised care. The assessment of women’s preferences is therefore of great importance in the evaluation of any new healthcare intervention. This study was designed to assess both midwives’ and women’s satisfaction with and preferences for care in relation to methods of induction of labour.
4.2 STUDY AIMS AND HYPOTHESES
The use of oral misoprostol is • Women’s preferred method of induction of labour when compared with vaginal
prostaglandin E2 gel; and • Caregiver’s preferred method of induction of labour when compared with
vaginal prostaglandin E2 gel.
4.3 METHODS
This randomised, double blind, placebo controlled trial comparing oral misoprostol with vaginal prostaglandin E2 gel for the induction of labour was performed at the Women’s and Children’s Hospital, the Lyell McEwin Health Service, and Hervey Bay Hospital between April 2001 and December 2004. Women were recruited from the antenatal clinic, antenatal wards and delivery suite of participating institutions, with research and ethics committee approval from each collaborating centre as described previously.
Following birth, and prior to discharge from hospital, women were asked to complete a short questionnaire detailing their experiences of induction of labour. Women were
85 asked to indicate their level of satisfaction with the induction process, their labour, and their birth (categorised into unsatisfied = not at all or mildly satisfied, and satisfied = moderately or extremely satisfied). A list of likes and dislikes relating to the induction process were provided, in addition to a text field in which women could make further comments. Women were asked to mark all applicable statements relating to their likes and dislikes. Additionally, women were asked to nominate the group to which they thought they belonged (oral misoprostol, vaginal prostaglandin E2 gel, or uncertain), as well as their preferences for method of induction in a future pregnancy should it be required. An identical questionnaire was posted to each woman approximately six weeks after birth. Women who did not return the questionnaire by mail were telephoned, and the questionnaire completed over the phone.
A questionnaire was completed by the midwife primarily responsible for the care of the woman during the induction process, detailing satisfaction with the method (categorised into unsatisfied = not at all or mildly satisfied, and satisfied = moderately or extremely satisfied), preferences for induction methods to be used in the future, as well as asking an indication of the group the woman was thought to belong to (oral misoprostol, vaginal prostaglandin E2 gel, or uncertain).
Results were analysed on an intention-to-treat basis and blind to the allocated treatment using the SAS Version 9.1 statistical software programme (SAS Institute Incorporated 2003). Baseline data were examined by treatment group to ensure comparability for prognostic factors as described previously, with adjustment for imbalance in the initial Bishop’s score, using log binomial regression techniques. The outcomes were compared using χ2 tests and Fisher’s Exact Test, with calculation of relative risks (RR) with 95% Confidence Intervals (CI). A planned secondary analysis related to exploration of any differences in satisfaction for those women who remained undelivered 24 hours after the induction process commenced. A p value of less than 0.05 was considered statistically significant.
86 4.4 RESULTS
4.4(1) Baseline characteristics
A total of 1319 eligible women were approached for trial participation by a clinical researcher at the time of booking the induction of labour, of whom 939 provided written consent to participate (71.2%)(Figure 3.4(1)). Of those consenting women, 199 (21.2%) entered labour spontaneously prior to the date of booked induction of labour, and 741(78.8%) were admitted for induction of labour and randomised to the trial. Three hundred and 65 women were randomised to the oral misoprostol group, and 376 to the vaginal PGE2 group. Data on women’s preferences for care at the time of hospital discharge were available for 739 (99.7%) women. Six-week questionnaires were available from 735 (99.2%) women. Baseline characteristics were comparable between the treatment groups, with the exception of initial Bishop’s score as described previously (Table 3.4(1)), and adjustment for this imbalance in the planned analyses was required.
4.4(2) Women’s Satisfaction and Preferences for Care at Discharge
All 739 (100%) women for whom satisfaction questionnaires were available, were satisfied with the care that they received (Table 4.4(2)). The number of women who were not satisfied (defined as either not at all satisfied or only mildly satisfied) with their induction were similar between the oral misoprostol and vaginal PGE2 gel groups
(Misoprostol 68/365 (18.6%) versus PGE2 72/374 (19.3%); RR 0.97 95% CI 0.72-1.31; p=0.840), as were those unsatisfied with their labour (Misoprostol 84/365 (23.0%) versus PGE2 73/374 (19.5%); RR 1.17 95% CI 0.89-1.55; p=0.266) and birth
(Misoprostol 51/365 (14.0%) versus PGE2 53/374; RR 1.01 95% CI 0.71-1.45; p=0.949). Overall, 74.3% of women, given similar circumstances, would participate in the study again, and 83.2% of women would recommend a friend in similar circumstances participate in the study, figures that were not significantly different between the two treatment groups. Women in the misoprostol group were more likely to correctly identify the group to which they had been allocated (Misoprostol 132/365
(36.2%) versus PGE2 91/374 (24.3%); RR 1.49 95% CI 1.19-1.86; p=0.001). Of the
87 women participating in the trial, 57.9% indicated a preference for an oral induction agent, should any future pregnancy require induction of labour, with no differences in preferences between the two treatment groups. Satisfaction with care did not differ between women who had given birth within 24 hours of the induction starting, and those women who remained undelivered after 24 hours.
Table 4.4(2) Women’s Satisfaction and Preferences for Care at Discharge
Misoprostol PGE2 Group Treatment Effect RR P Group 95% CI value All women N=365 % N=374 % - Woman not satisfied with care 0 0.0 0 0.0 Not estimable - Not satisfied with IOL 68 18.6 72 19.3 0.97 0.72-1.31 0.840 - Not satisfied with labour 84 23.0 73 19.5 1.17 0.89-1.55 0.266 - Not satisfied with birth 51 14.0 53 14.2 1.01 0.71-1.45 0.949 - Participate in study again 274 75.1 275 73.1 1.03 0.94-1.12 0.539 - Recommend others participate in study 304 83.3 311 82.3 1.01 0.94-1.07 0.842 - Correctly identify group 132 36.2 91 24.3 1.49 1.19-1.86 0.001 - Preference for oral IOL agent 219 60.0 209 55.9 1.07 0.95-1.21 0.434
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text. IOL = induction of labour
4.4(3) Women’s Likes and Dislikes During Induction of Labour at Discharge
Overall, women’s responses to their participation in the trial were positive, with 29.4% indicating that they “liked being involved in the study”, and 32.2% indicating that they “liked everything” during their induction and labour (Table 4.4(3)). Few women (2.3%) indicated that they “disliked being involved in the study”, or that they “didn’t like anything” during their induction and labour (9.3%), responses that did not differ for women in the oral misoprostol group when compared with those in the vaginal PGE2 group (Table 4.4(3)). Women liked the “reassurance that (their) baby was well” (58.2%), while 18.3% of women indicated that they didn’t like having their baby monitored. Again, these responses were not different between women in the oral misoprostol group and those in the vaginal PGE2 group.
88 While 39.0% of women indicated that they had no side effects from the medication, and 9.5% few medication side effects, 19.8% of women disliked abdominal cramps, 13.1% disliked nausea or vomiting, 5.1% disliked hot flushes, 3.8% disliked headaches, and 1.6% disliked diarrhoea. There were no differences in these responses between women in the oral misoprostol group and those in the vaginal PGE2 group.
The trial protocol required administration of oral medication and fetal heart rate monitoring at two hourly intervals, and 3.1% of women indicated that they did not like the interruptions to sleep that this caused, a response that did not differ between those women in the oral misoprostol group and those in the vaginal PGE2 group.
Likes and dislikes during induction and labour did not differ between women who had given birth within 24 hours of the induction starting, and those women who remained undelivered after 24 hours.
Table 4.4(3) Women’s Likes and Dislikes During Induction of Labour at Discharge
Misoprostol PGE2 Group Treatment Effect RR P value Group 95% CI All women N=365 % N=374 % I Liked… - Didn’t like anything 35 9.6 34 9.0 1.08 0.69-1.71 0.724 - No side effects from medication 142 38.9 146 38.8 1.00 0.83-1.20 0.985 - Few side effects from medication 34 9.3 36 9.6 0.96 0.61-1.51 0.866 - Reassurance baby was well 214 58.6 216 57.5 1.02 0.90-1.15 0.770 - Being involved in this study 104 28.5 113 30.1 0.94 0.75-1.17 0.564 I Disliked… - Like everything 123 33.7 115 30.6 1.11 0.90-1.37 0.319 - Nausea or vomiting 46 12.6 51 13.6 0.94 0.65-1.38 0.784 - Diarrhoea 5 1.4 7 1.9 0.75 0.24-2.34 0.616 - Abdominal cramps 63 17.3 83 22.1 0.78 0.58-1.04 0.092 - Headaches 11 3.0 17 4.5 0.66 0.31-1.39 0.272 - Hot flushes 13 3.6 25 6.7 0.53 0.27-1.02 0.058 - Too many internal examinations 98 26.9 103 27.4 0.95 0.75-1.20 0.642 - Having the baby monitored 65 17.8 70 18.6 0.98 0.72-1.33 0.893 - Being involved in this study 9 2.5 8 2.1 1.21 0.47-3.12 0.692 - Lack of sleep 9 2.5 14 3.7 0.64 0.28-1.46 0.288
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text.
89 4.4(4) Midwives’ Satisfaction and Preferences for Care The primary midwife responsible for care during the induction process completed a staff satisfaction questionnaire, and there were no differences in the number of midwives not satisfied (defined as not at all satisfied or only mildly satisfied) with the induction process (Misoprostol 102/365 (27.9%) versus PGE2 107/374 (28.6%); RR 0.94 95% CI 0.75-1.18; p=0.612)(Table 4.4(4)). Midwives caring for women in the misoprostol group were more likely to correctly identify the group to which the woman was allocated (Misoprostol 108/365 (29.6%) versus PGE2 84/374 (22.5%); RR 1.32 95% CI 1.03-1.68; p=0.049). Overall, 65.1% of midwives indicated a preference for an oral induction of labour agent.
Table 4.4(4) Midwives’ Satisfaction and Preferences for Care
Misoprostol PGE2 Group Treatment Effect P Group RR 95% CI value All women N=365 % N=374 % - Carer not satisfied 102 27.8 107 28.6 0.94 0.75-1.18 0.612 - Carer correctly identify group 108 29.6 84 22.5 1.32 1.03-1.68 0.049 - Carer preference for oral IOL agent 240 65.8 241 64.4 1.02 0.92-1.13 0.619
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text. IOL = induction of labour
4.4(5) Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum
Of the 735 women for whom satisfaction questionnaires were available, 99.7% were satisfied with the care that they received (Table 4.4(5)). The number of women who were not satisfied (defined as either not at all satisfied or only mildly satisfied) with their induction were similar between the oral misoprostol and vaginal PGE2 gel groups
(Misoprostol 77/362 (21.3%) versus PGE2 79/373 (21.2%); RR 1.00 95% CI 0.76-1.33; p=0.975), as were those unsatisfied with their labour (Misoprostol 83/362 (22.9%) versus PGE2 80/373 (21.4%); RR 1.07 95% CI 0.82-1.40; p=0.600) and birth
90 (Misoprostol 48/362 (13.3%) versus PGE2 50/373 (13.4%); RR 0.99 95% CI 0.68-1.43; p=0.990).
Overall, 70.6% of women, given similar circumstances, would participate in the study again, and 80.5% of women would recommend a friend in similar circumstances participate in the study, figures that were not significantly different between the two treatment groups. Women in the misoprostol group were more likely to correctly identify the group to which they had been allocated (Misoprostol 135/362 (37.3%) versus PGE2 89/373 (23.9%); RR 1.49 95% CI 1.19-1.86; p=0.001). Of the women participating in the trial, 58.5% indicated a preference for an oral induction agent, should any future pregnancy require induction of labour, with no differences in preferences between the two treatment groups.
At six weeks post-partum, satisfaction with care did not differ between women who had given birth within 24 hours of the induction starting, and those women who remained undelivered after 24 hours.
Table 4.4(5) Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum
Misoprostol PGE2 Group Treatment Effect P Group RR 95% CI value All women N=362 % N=373 % - Woman not satisfied with care 0 0.0 2 0.5 0.21 0.01-4.28 0.466 - Not satisfied with IOL 77 21.3 79 21.2 1.00 0.76-1.33 0.975 - Not satisfied with labour 83 22.9 80 21.4 1.07 0.82-1.40 0.600 - Not satisfied with birth 48 13.2 50 13.4 0.99 0.68-1.43 0.990 - Participate in study again 263 72.7 256 68.6 1.06 0.96-1.16 0.238 - Recommend others participate in study 296 81.8 296 79.4 1.03 0.96-1.11 0.421 - Correctly identify group 135 37.3 89 23.4 1.49 1.19-1.86 0.001 - Preference for oral IOL agent 215 59.4 215 57.6 1.03 0.91-1.16 0.348
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text. IOL = induction of labour
91 4.4(6) Women’s Likes and Dislikes During Induction of Labour at 6 weeks Postpartum
Overall, women’s responses to their participation in the trial were positive, with 31.7% indicating that they “liked being involved in the study”, and 31.2% indicating that they “liked everything” during their induction and labour (Table 4.4(6)). More women in the misoprostol group indicated that they “liked everything” when compared with women in the vaginal PGE2 group (Misoprostol 126/362 (34.8%) versus PGE2 103/373 (27.6%); RR 1.26 95% CI 1.02-1.57; p=0.036). Few women (5.2%) indicated that they “disliked being involved in the study”, or that they “didn’t like anything” during their induction and labour (9.4%), responses that did not differ for women in the oral misoprostol group when compared with those in the vaginal PGE2 group (Table 4.4(6)). Women liked the “reassurance that (their) baby was well” (62.4%), while 18.5% of women indicated that they didn’t like having their baby monitored. Again, these responses were not different between women in the oral misoprostol group and those in the vaginal PGE2 group.
While 43.5% of women indicated that they had no side effects from the medication, and 8.3% few medication side effects, 16.1% of women disliked abdominal cramps, 12.7% disliked nausea or vomiting, 3.4% disliked hot flushes, and 1.8% disliked diarrhoea. There were no differences in these responses between women in the oral misoprostol group and those in the vaginal PGE2 group. Women in the misoprostol group were less likely to experience headaches when compared with women in the vaginal PGE2 group
(Misoprostol 9/362 (2.5%) versus PGE2 23/373 (6.2%); RR 0.43 95% CI 0.19-0.86; p=0.015).
The trial protocol required administration of oral medication and fetal heart rate monitoring at two hourly intervals, and 2.2% of women indicated that they did not like the interruptions to sleep that this caused, a response that did not differ between those women in the oral misoprostol group and those in the vaginal PGE2 group.
Of women who remained undelivered 24 hours after commencing the induction, more women in the misoprostol group indicated that they “liked everything” during their induction and labour when compared with women in the vaginal PGE2 group
92 (Misoprostol 26/130 (20.0%) versus PGE2 9/98 (9.2%); RR 2.18 95% CI 1.07-4.44; p=0.025). Likes and dislikes during induction and labour otherwise did not differ between women who had given birth within 24 hours of the induction starting, and those women who remained undelivered after 24 hours.
Table 4.4(6) Women’s Likes and Dislikes During Induction of Labour at 6 weeks Postpartum
Misoprostol PGE2 Group Treatment Effect P value Group RR 95% CI All women N=362 % N=373 % I Liked… - Didn’t like anything 34 9.4 35 9.4 1.00 0.64-1.57 0.998 - No side effects from medication 162 44.8 158 42.4 1.06 0.90-1.25 0.516 - Few side effects from medication 24 6.6 37 9.9 0.67 0.41-1.09 0.106 - Reassurance baby was well 230 65.5 229 61.4 1.03 0.93-1.16 0.554 - Being involved in this study 106 29.3 127 34.0 0.86 0.69-1.06 0.165 I Disliked… - Like everything 126 34.8 103 27.6 1.26 1.02-1.57 0.036 - Nausea or vomiting 40 11.0 53 14.2 0.78 0.53-1.14 0.198 - Diarrhoea 5 1.4 8 2.1 0.64 0.21-1.95 0.432 - Abdominal cramps 51 14.1 67 18.0 0.78 0.56-1.10 0.153 - Headaches 9 2.5 23 6.2 0.43 0.19-0.86 0.015 - Hot flushes 15 4.1 10 2.7 1.54 0.70-3.39 0.274 - Too many internal examinations 93 25.7 120 32.2 0.80 0.64-1.00 0.053 - Having the baby monitored 72 19.9 64 17.2 1.16 0.64-1.57 0.342 - Being involved in this study 17 4.7 21 5.6 0.83 0.45-1.55 0.567 - Lack of sleep 6 1.6 10 2.7 0.62 0.23-1.68 0.342
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text.
4.5 DISCUSSION
There is a current shift in health care policy and practice to incorporate women’s preferences for health related outcomes, with the aim of providing more individualised and personalised care. As highlighted by Hofmeyr and colleagues (2001), and in the Cochrane review on oral misoprostol for the induction of labour (Alfirevic 2005), there is a need for information about women’s satisfaction and preferences for care during induction of labour. This is the first randomised controlled trial comparing oral
93 misoprostol and vaginal PGE2 gel for induction of labour at term to specifically obtain women’s views on the care they received, with follow-up to six weeks postpartum available for over 99% of women. However, as all women received both oral and vaginal preparations, it was not possible to specifically assess women’s experiences related to the individual methods of induction of labour. The use of a vaginal placebo necessitated all women undergoing vaginal examination, a factor that may have influenced their overall preferences and perception of trial participation.
Overall, women were highly satisfied with the care received, and with their induction, labour and birth, findings that did not differ between women in the oral misoprostol group and women in the vaginal PGE2 group. Women’s experiences of participation in the study were similarly high, with the majority of women willing to participate again if they found themselves in similar circumstances, and would recommend participation to a friend. These findings suggest that for the majority of women, induction of labour and birth is a positive experience.
Almost 60% of women indicated a preference for the availability of an oral induction of labour agent, indicating a high degree of acceptability among women participating in the trial. This, in conjunction with 27% of women indicating a dislike for the number of internal examinations during induction and labour, highlights the invasive nature of vaginal examinations for a woman, and the negative impact it can have on a woman’s birth experience.
Current guidelines from both the Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG 2001), and the Royal College of Obstetricians and Gynaecologists (RCOG 2001) recommend electronic fetal heart rate monitoring in settings where labour is induced. Almost 60% of women in this trial indicated that they liked the reassurance fetal heart rate monitoring provided, while less than 20% of women involved indicated that they did not like fetal heart rate monitoring. This positive view expressed by women is supportive of the current trial protocol, which required intensive fetal heart rate monitoring for 40-minute periods at two hourly intervals in relation to the administration of oral medication.
94 Findings at 6 weeks postpartum were similar to those obtained prior to discharge from hospital, with women expressing a high degree of satisfaction with their induction (81.1% at discharge versus 78.8% at six weeks postpartum), labour (78.8% at discharge versus 77.8% at six weeks postpartum), and birth (85.9% at discharge versus 86.7% at six weeks postpartum). Women’s willingness to participate in the study again was similar at both time points (74.3% at discharge versus 70.6% at six weeks postpartum), as was their recommendation to a friend (83.2% at discharge versus 80.5% at six weeks postpartum), and their preference for an oral medication (57.9% at discharge versus 58.5% at six weeks postpartum). The likes and dislikes related to induction expressed by women at six weeks post-partum were similar to those expressed at the time of discharge from hospital, however, women in the misoprostol group were more likely to indicate that they “liked everything” in relation to their induction, labour and birth, when compared with women in the vaginal PGE2 group.
Women who remained undelivered 24 hours or more after the induction process commenced were not different in the degree of satisfaction expressed with their induction, labour or birth, when compared with women who birthed in less than 24 hours. Of women who remained undelivered after 24 hours, more women in the misoprostol group indicated that they “liked everything” associated with their induction, labour and birth, when compared with women in the vaginal PGE2 group. While women in the misoprostol group were more likely to remain undelivered after 24 hours, this does not appear to have affected their labour and birth experience in a negative fashion.
4.6 CONCLUSION
Women participating in this randomised trial have indicated a strong preference for the availability of an oral agent to induce labour, and have indicated that the electronic fetal heart rate monitoring required under the trial protocol provided them with reassurance that their baby was “well”. Women administered misoprostol were more likely to “like everything” in relation to their induction, labour and birth experience.
95 5. TIME OF COMMENCING INDUCTION OF LABOUR – A NESTED RANDOMISED CONTROLLED TRIAL
5.1 INTRODUCTION
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 (Runnebaum 1972; Junkermann 1982; Walsh 1984), estrone, estriol and estradiol (Challis 1980; Reck 1988; Bernstein 1989) 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 (Lindow 1996), correlating with a documented nocturnal increase in uterine activity (Hirst 1991; Hirst 1993).
There have been several clinical studies that have retrospectively assessed diurnal variations in onset of labour in women both at term (Fraser 1989; Cagnacci 1998; Mancuso 2004), and before term (Cooperstock 1987; Lindow 2000). These studies consistently demonstrate a peak in the onset of labour in the evening and early hours of the morning. Traditionally, women booked for induction of labour have been admitted to hospital in the evening, with the process of cervical ripening taking place overnight, in the expectation that women will labour 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 labour.
Birth during daylight hours may be beneficial in terms of neonatal outcomes, as 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 (Chalmers 1998; Stewart 1998; Heller 2000; Luo 2001). Using the Swedish Birth Registry, Luo and Karlberg (2001) have reviewed over two million births between 1973 and 1995 and assessed the effect of time of birth on infant outcomes, as well as identifying particular periods of
96 time considered to be “high risk”. For all births occurring at night (defined as between 9pm and 7am), there was a small increased risk of infant mortality (rate 3.65 day versus 3.81 night; absolute risk 0.16; relative risk 1.04; 95%CI 1.00-1.09), this being greatest for infants born at less than 37 weeks gestation (rate 39.47 day versus 42.60 night; absolute risk 3.13; relative risk 1.08; 95%CI 1.01-1.15). In addition, birth between 5pm and 1am, and around 9am were considered to be “high risk” periods for early neonatal mortality (Luo 2001).
There have been no randomised trials that have compared the preferences women hold for timing of admission and induction of labour, in particular morning or evening admission. This nested randomised trial was conducted to assess the preferences women have for timing of induction of labour and the impact that timing of the induction process may have on maternal and infant outcomes.
5.2 STUDY AIMS AND HYPOTHESES
Commencing the process of induction of labour in the morning more closely reflects the physiological timing of onset of labour, and is associated with fewer women who remain undelivered 24 hours after the induction process commences.
5.3 METHODS
5.3(1) Trial Design
This nested prospective randomised controlled trial comparing time of admission to start induction of labour (Morning Admission Group – after 0800 hours versus Evening Admission Group – after 2000 hours) was conducted at the Women’s and Children’s Hospital. Women who consented to participation in the main trial (as described previously) were randomly allocated at the time of booking the induction of labour, to admission in the morning (0800 hours) or in the evening (2000 hours).
97 Women were recruited from the antenatal clinic, antenatal wards and delivery suite of participating institutions as described previously. Potentially eligible women presenting for induction of labour were given the trial information sheet, counselled by a member of the study team, and encouraged to discuss the study with her family. Written informed consent was then obtained.
5.3(2) Inclusion and Exclusion Criteria
Inclusion and exclusion criteria were as described previously in relation to the main trial comparing oral misoprostol with vaginal PGE2 gel (Chapter 3).
5.3(3) Randomisation Schedule
The randomisation schedule was generated by a non-clinical 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 main trial. Treatment packs were prepared by the principal investigator and the research midwife, and labelled by a non-clinical researcher according to the generated randomisation 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. Opening the treatment pack was the point of randomisation.
5.3(4) Blinding
Women, caregivers and outcome assessors were not blinded to the woman’s time of admission for induction of labour to start.
98 5.3(5) Treatment Allocation
When women who had given written consent to participation in the main trial comparing oral misoprostol with vaginal PGE2 gel, they were then randomly allocated to the time of admission for induction of labour to start. 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 main trial, and the induction of labour booked with delivery suite in accordance with the instructions written on the card.
At the time of admission to delivery suite for induction of labour, trial entry details were confirmed. A study number was allocated to the woman, by taking the next identically appearing, sequentially numbered treatment pack appropriate for her parity and for the 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 (Prostin) gel. Each pack contained a sheet detailing treatment schedules. The treatment packs were kept at 4O Celsius (although this was not a requirement for misoprostol). Opening the treatment pack that corresponded to the woman’s parity and time of admission was the point of randomisation to the timing of induction trial.
5.3(6) Treatment Schedules
The treatment schedules have been described previously in the relation to the main trial comparing oral misoprostol with vaginal PGE2 gel (Chapter 3).
Following birth, and prior to discharge from hospital, women were asked to complete a short questionnaire detailing their experiences of induction of labour. Women were asked to indicate their level of satisfaction with the induction process, their labour, and their birth (categorised into unsatisfied = not at all or mildly satisfied, and satisfied = moderately or extremely satisfied). A list of likes and dislikes relating to the induction process were provided, in addition to a text field in which women could make further
99 comments. Women were asked to mark all applicable statements relating to their likes and dislikes. An identical questionnaire was posted to each woman approximately six weeks after birth. Women who did not return the questionnaire by mail were telephoned, and the questionnaire completed over the phone.
5.3(7) Study Outcome Measures
5.3(7)(i) Primary Study Outcomes The primary study outcomes were consistent with the Cochrane generic protocol relating to induction of labour (Hofmeyr 2005a), and have been described in Chapter 3.
The primary outcomes were • vaginal birth not achieved within 24 hours, • uterine hyperstimulation with associated fetal heart rate changes, • caesarean section (all), and • caesarean section for fetal distress.
5.3(7)(ii) Secondary Study Outcomes Secondary study outcomes recorded related to evidence of effect, neonatal complications and maternal complications as described in the Cochrane generic protocol relating to methods of induction of labour (Hofmeyr 2005a) and have been described in Chapter 3 (Table 3.3(6)(ii)).
5.3(8) Data Collection
Data forms were completed by the midwife caring for the woman, with information confirmed and checked by the principal investigator or research midwife prior to the woman being discharged from hospital. All data were then entered into a database created in Access ’97 (Microsoft Corporation 1997) by the principal investigator.
5.3(9) Statistical Analysis
Data were analysed on an intention-to-treat basis and blind to the allocated treatment using the SAS Version 9.1 statistical software programme (SAS Institute Incorporated
100 2003). The primary and secondary outcomes were compared using χ2 tests and Fisher’s Exact Test, with calculation of relative risks (RR) with 95% Confidence Intervals (CI) for dichotomous data. Normally distributed continuous data were compared using the Student’s T-test, and skewed data using non-parametric tests (Wilcoxon Rank Sum). Pre-specified subgroup analyses were planned to assess any differential effect in medication received (oral misoprostol versus vaginal PGE2 gel), the effect of maternal parity (nulliparous versus multiparous) and initial Bishop’s score at commencing induction (Bishop’s score 0-3 versus Bishop’s score 4-6) on the time of admission for induction of labour. A p value of less than 0.05 was considered statistically significant.
5.3(10) Sample Size
Using information from the Cochrane review of oral misoprostol for induction of labour at term (Alfirevic 2005), and figures generated by the Women’s and Children’s Hospital Clinical Information Service, 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 evening admission group to 30% in the morning admission group (p<0.05).
5.4 RESULTS
5.4(1) Baseline Characteristics
A total of 1072 eligible women were approached for trial participation by a clinical researcher at the time of booking the induction of labour, of whom 775 (72.3%) provided written consent to participate in the timing of induction trial (Figure 5.4(1)). Of the 775 women who gave consent, 380 (49.0%) women were scheduled for morning admission and 395 (51.0%) women were scheduled for evening admission. Prior to the date of induction of labour, 155 (20.0%) women entered labour 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 randomised into the timing of induction trial, 280 (45.2%) were randomised to the morning admission group and 340 (54.8%) to the evening admission group.
101 In the morning admission group, 138 women received oral misoprostol and 142 women vaginal PGE2 gel, and in the evening admission group, 167 women received misoprostol and 173 women vaginal PGE2 gel (Table 5.4(1)). There were 164 nulliparous women (58.6%) in the morning admission group and 201 nulliparous women (59.1%) in the evening admission group. In the morning admission group, there were 177 women (63.2%) with an initial Bishop’s score of 0-3, and 203 women (59.7%) in the evening admission group with an initial Bishop’s score of 0-3.
Figure 5.4(1) Trial Flow Chart
Eligible Women Approached 1072
Eligible Women Consenting Eligible Women Declined 775 (72.3%) 297 (27.7%)
Scheduled Morning Admission Scheduled Evening Admission 380 women (49.0%) 395 women (51.0%)
Entered Labour Spontaneously Entered Labour Spontaneously Prior to Booked Induction Prior to Booked Induction 100 women (12.9%) 55 women (7.1%)
Randomised to Morning Admission Group Randomised to Evening Admission Group 280 (45.2%) women 340 (54.8%) women
Misoprostol Vaginal PGE2 Misoprostol Vaginal PGE2 138 women 142 women 167 women 173 women
Outcome Data Available Outcome Data Available Outcome Data Available Outcome Data Available 138 women (100%) 142 women (100%) 167 women (100%) 173 women (100%)
102 Table 5.4(1) Distribution of Time of Commencing Induction
Morning Admission Evening Admission N= 280 % N=340 % Allocated Treatment - Misoprostol 138 49.3 167 49.1 - Vaginal PGE2 gel 142 50.7 173 50.9 Bishop’s Score - Bishop’s score 0-3 177 63.2 203 59.7 - Bishop’s score 4-6 103 36.8 137 40.3 Nulliparous women 164 58.6 201 59.1 - Bishop’s score 0-3 102 62.2 119 59.2 - Bishop’s score 4-6 62 37.8 82 40.8 Multiparous women 116 41.4 139 40.9 - Bishop’s score 0-3 75 64.7 84 60.4 - Bishop’s score 4-6 41 35.3 55 39.6
Figures are numbers and percentages
5.4(2) Time of Commencing Induction and Primary Outcomes
The likelihood of not achieving vaginal birth within 24 hours of starting induction of labour was not different for women in the morning admission group, when compared with women in the evening admission group (Morning admission 121/280 (43.2%) versus Evening admission 151/340 (44.4%); RR 0.97 95% CI 0.81-1.16; p=0.432) (Table 5.4(2)). The occurrence of uterine hyperstimulation syndrome with associated fetal heart rate changes (Morning admission 5/280 (1.8%) versus Evening admission 0/340 (0.0%); RR 7.31 95% CI 0.88-60.58; p=0.998), all caesarean sections (Morning admission 62/280 (22.1%) versus Evening admission 89/340 (26.2%); RR 0.83 95% CI 0.63-1.10; p=0.194) and caesarean section for fetal distress (Morning admission 27/280 (9.6%) versus Evening admission 33/340 (9.7%); RR 0.98 95% CI 0.60-1.59; p=0.934) did not differ for women in the morning admission group compared with women in the evening admission group.
There were no differences identified between morning and evening admission and the induction agent used (misoprostol versus vaginal PGE2 gel), maternal parity (nulliparous versus multiparous), or initial Bishop’s score (Bishop’s score 0-3 versus Bishop’s score 4-6) for primary study outcomes.
103 Table 5.4(2) Time of Commencing Induction - Primary Outcomes
Morning Admission Evening Admission Treatment Effect P value RR 95% CI All women N=280 % N=340 % - Vaginal Birth not achieved in 24 hours 121 43.2 151 44.4 0.97 0.81-1.16 0.432 - Uterine HSS with FHR changes 5 1.8 0 0.0 7.31 0.88-60.58 0.997 - Caesarean section - all 62 22.1 89 26.2 0.83 0.63-1.10 0.194 - Caesarean section - fetal distress 27 9.6 33 9.7 0.98 0.60-1.59 0.934 Misoprostol N=138 % N=167 % - Vaginal Birth not achieved in 24 hours 65 47.1 81 48.5 0.97 0.77-1.23 0.581 - Caesarean section - all 28 20.3 42 25.2 0.78 0.52-1.19 0.255 - Caesarean section – fetal distress 14 10.1 16 9.6 1.04 0.53-2.05 0.913
Vaginal PGE2 N=142 % N=173 % - Vaginal Birth not achieved in 24 hours 56 39.4 70 40.5 0.97 0.74-1.28 0.587 - Caesarean section - all 34 23.9 47 27.2 0.87 0.60-1.27 0.482 - Caesarean section – fetal distress 13 9.2 17 9.8 0.92 0.47-1.83 0.821 Bishop’s score 0-3 N=177 % N=203 % - Vaginal Birth not achieved in 24 hours 84 47.5 107 52.7 0.90 0.74-1.10 0.633 - Caesarean section – all 42 23.7 63 31.0 0.76 0.55-1.07 0.112 - Caesarean section – fetal distress 19 10.7 23 11.3 0.95 0.53-1.68 0.854 Bishop’s score 4-6 N=103 % N=137 % - Vaginal Birth not achieved in 24 hours 37 35.9 44 32.1 1.12 0.78-1.59 0.465 - Caesarean section – all 20 19.4 26 19.0 1.02 0.61-1.73 0.932 - Caesarean section – fetal distress 8 7.8 10 7.3 1.06 0.44-2.60 0.892 Nulliparous women N=164 % N=201 % - Vaginal Birth not achieved in 24 hours 94 57.3 115 57.2 1.00 0.84-1.20 0.284 - Caesarean section - all 49 29.9 71 35.3 0.83 0.61-1.11 0.208 - Caesarean section – fetal distress 22 13.4 23 11.4 1.16 0.67-2.00 0.601 Nulliparous women, BS 0-3 N=102 % N=119 % - Vaginal Birth not achieved in 24 hours 66 64.7 82 68.9 0.94 0.78- 1.13 0.297 - Caesarean section - all 32 31.4 50 42.0 0.75 0.52-1.07 0.103 - Caesarean section – fetal distress 15 14.7 16 13.5 1.09 0.57-2.10 0.788 Nulliparous women, BS 4-6 N= 62 % N= 82 % - Vaginal Birth not achieved in 24 hours 28 45.2 34 41.5 1.09 0.75-1.59 0.763 - Caesarean section - all 17 27.4 21 25.6 1.07 0.62-1.85 0.807 - Caesarean section – fetal distress 7 11.3 7 8.5 1.32 0.49-3.57 0.581 Multiparous women N=116 % N=139 % - Vaginal Birth not achieved in 24 hours 27 23.3 35 25.2 0.92 0.60-1.43 0.903 - Caesarean section – all 13 11.2 18 13.0 0.85 0.44-1.66 0.693 - Caesarean section – fetal distress 5 4.3 10 7.2 0.59 0.21-1.68 0.322 Multiparous women, BS 0-3 N= 75 % N= 84 % - Vaginal Birth not achieved in 24 hours 18 24.0 25 29.8 0.81 0.48-1.36 0.492 - Caesarean section – all 10 13.3 13 15.5 0.86 0.40-1.85 0.701 - Caesarean section – fetal distress 4 5.3 7 8.3 0.64 0.20-2.10 0.457 Multiparous women, BS 4-6 N= 41 % N= 55 % - Vaginal Birth not achieved in 24 hours 9 22.0 10 18.2 1.21 0.54-2.70 0.399 - Caesarean section – all 3 7.3 5 9.2 0.80 0.20-3.18 0.756 - Caesarean section – fetal distress 1 2.4 3 5.5 0.45 0.05-4.15 0.465
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect.
104 5.4(3) Time of Commencing Induction - Secondary Outcomes: Evidence of Effect
The likelihood of a woman having a Bishop’s score of less than seven 24 hours after the induction commenced (Morning admission 35/280 (12.5%) versus Evening admission 41/340 (12.1%); RR 1.04 95% CI 0.68-1.58; p=0.503), requiring further doses of known vaginal PGE2 gel (Morning admission 43/280 (15.4%) versus Evening admission 51/340 (15.0%); RR 0.98 95% CI 0.68-1.42; p=0.915), or the induction to birth interval (Morning admission median interval 18.9 hours (interquartile range 6.2-31.6) versus Evening admission median interval 21.0 hours (interquartile range 9.5-32.5); p=0.145) did not differ between women admitted in the morning when compared with women admitted in the evening (Table 5.4(3)).
Women admitted in the morning were significantly less likely to require oxytocin infusion when compared with women admitted for induction in the evening (Morning admission 126/280 (45.0%) versus Evening admission 184/340 (54.1%); RR 0.83 95% CI 0.70-0.97; p=0.022). This effect was also present for women admitted in the morning who received vaginal PGE2 gel as the induction agent (Morning admission 54/142 (38.0%) versus Evening admission 91/173 (52.6%); RR 0.73 95% CI 0.56-0.93; p=0.012), but not for those women who received misoprostol as the induction agent (Morning admission 72/138 (52.2%) versus Evening admission 93/167 (55.7%); RR 0.93 95% CI 0.75-1.15; p=0.497). Women with an initial Bishop’s score of 0-3 admitted in the morning were less likely to require oxytocin infusion (Morning admission 80/177 (45.2%) versus Evening admission 115/203 (56.7%); RR 0.80 95% CI 0.65-0.98; p=0.026) and had a shorter induction to birth interval (Morning admission median interval 19.3 hours (interquartile range 6.3-32.3) versus Evening admission median interval 22.8 hours (interquartile range 10.9-34.7); p=0.043) when compared with women who were admitted in the evening. Nulliparous women admitted in the morning were less likely to require oxytocin infusion than women admitted in the evening (Morning admission 92/164 (56.1%) versus Evening admission 135/201 (67.2%); RR 0.83 95% CI 0.70-0.98; p=0.029), as were nulliparous women with an initial Bishop’s score of 0-3 who were admitted in the morning (Morning admission 59/102 (57.8%) versus Evening admission 84/119 (70.6%); RR 0.82 95% CI 0.67-1.00; p=0.048), who also had a significantly shorter induction to birth interval (Morning admission median
105 interval 23.3 hours (interquartile range 9.2-37.4) versus Evening admission median interval 25.5 hours (interquartile range 14.9-36.1); p=0.042).
There were no other differences identified related to the time of admission for induction of labour and induction agent used, maternal parity, and initial Bishop’s score.
Table 5.4(3) Time of Commencing Induction - Secondary Outcomes: Evidence of Effect
Morning Admission Evening Admission Treatment Effect P value RR 95%CI All women N=280 % N=340 % - Bishop’s Score <7 after 24 hours 35 12.5 41 12.1 1.04 0.68-1.58 0.503 - Further doses of known PGE2 43 15.4 51 15.0 0.98 0.68-1.42 0.915 - Oxytocin infusion 126 45.0 184 54.1 0.83 0.70-0.97 0.022 - Induction to birth interval* 18.9 6.2-31.6 21.0 9.5-32.5 0.145 Misoprostol N=138 % N=167 % - Bishop’s score <7 after 24 hours 21 15.2 26 15.6 0.98 0.58-1.66 0.938 - Further doses of known PGE2 27 19.6 31 18.6 0.99 0.63-1.57 0.978 - Oxytocin infusion 72 52.2 93 55.7 0.93 0.75-1.15 0.497 - Induction to birth interval* 20.7 7.6-33.8 21.6 9.6-33.6 0.862
Vaginal PGE2 N=142 % N=173 % - Bishop’s score <7 after 24 hours 14 9.9 15 8.9 1.14 0.57-2.28 0.354 - Further doses of known PGE2 16 11.3 20 11.6 0.95 0.52-1.75 0.871 - Oxytocin infusion 54 38.0 91 52.6 0.72 0.56-0.93 0.012 - Induction to birth interval* 16.9 5.4-28.4 20.0 8.9-31.1 0.080 Bishop’s score 0-3 N=177 % N=203 % - Bishop’s score <7 after 24 hours 27 15.3 34 16.8 0.91 0.57-1.45 0.572 - Further doses of known PGE2 32 18.1 43 21.2 0.85 0.57-1.29 0.448 - Oxytocin infusion 80 45.2 115 56.7 0.80 0.65-0.98 0.026 - Induction to birth interval* 19.3 6.3-32.3 22.8 10.9-34.7 0.043 Bishop’s score 4-6 N=103 % N=137 % - Bishop’s score <7 after 24 hours 8 7.8 7 5.1 1.92 0.72-5.14 0.541 - Further doses of known PGE2 11 10.7 8 5.8 1.83 0.76-4.38 0.169 - Oxytocin infusion 46 44.7 69 50.4 0.89 0.68-1.16 0.381 - Induction to birth interval* 18.7 6.4-31.0 17.7 6.3-29.1 0.941 Nulliparous women N=164 % N=201 % - Bishop’s Score <7 after 24 hours 24 21.6 34 16.9 0.87 0.54- 1.40 0.986 - Further doses of known PGE2 30 18.3 38 18.9 0.93 0.61- 1.42 0.734 - Oxytocin infusion 92 56.1 135 67.2 0.83 0.70- 0.98 0.029 - Induction to birth interval* 21.4 8.9-33.9 23.5 12.7-34.3 0.153 Nulliparous women, BS 0-3 N=102 % N=119 % - Bishop’s Score <7 after 24 hours 19 18.6 27 22.7 0.82 0.49-1.39 0.890 - Further doses of known PGE2 22 21.6 31 26.1 0.83 0.51-1.34 0.437 - Oxytocin infusion 59 57.8 84 70.6 0.82 0.67-1.00 0.048 - Induction to birth interval* 23.3 9.2-37.4 25.5 14.9-36.1 0.042 Nulliparous women, BS 4-6 N= 62 % N= 82 % - Bishop’s Score <7 after 24 hours 5 8.1 7 8.5 0.94 0.31-2.84 0.901 - Further doses of known PGE2 8 12.9 7 8.5 1.51 0.58-3.94 0.396 - Oxytocin infusion 33 53.2 51 62.2 0.86 0.64-1.14 0.280 - Induction to birth interval* 20.7 9.7-31.7 21.3 11.7-30.9 0.842
106 Morning Admission Evening Admission Treatment Effect P value RR 95%CI Multiparous women N=116 % N=139 % - Bishop’s Score <7 after 24 hours 11 9.5 7 5.0 1.88 0.75-4.70 0.204 - Further doses of known PGE2 13 11.2 13 9.4 1.14 0.55-2.35 0.718 - Oxytocin infusion 34 29.3 49 35.3 0.83 0.58-1.19 0.300 - Induction to birth interval* 15.3 6.5-24.1 16.6 5.1-28.1 0.955 Multiparous women, BS 0-3 N= 75 % N= 84 % - Bishop’s Score <7 after 24 hours 8 10.7 7 8.3 1.28 0.49-3.36 0.391 - Further doses of known PGE2 10 13.3 12 14.3 0.93 0.43-2.03 0.862 - Oxytocin infusion 21 28.0 31 36.9 0.76 0.48-1.20 0.232 - Induction to birth interval* 16.3 8.6-24.0 18.0 5.7-30.3 0.686 Multiparous women, BS 4-6 N= 41 % N= 55 % - Bishop’s Score <7 after 24 hours 3 7.3 0 0.0 9.33 0.50-175.86 0.091 - Further doses of known PGE2 3 7.3 1 1.8 4.02 0.43- 37.30 0.182 - Oxytocin infusion 13 31.7 18 32.7 0.97 0.54- 1.74 0.916 - Induction to birth interval* 14.3 4.2-24.4 15.4 6.1-24.7 0.833
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. * Median and interquartile range (interval in hours); Wilcoxon Rank Sum Test BS = Bishop’s Score
PGE2 = vaginal prostaglandin E2 gel
5.4(4) Time of Commencing Induction – Secondary Outcomes: Labour and Birth Complications
The occurrence of uterine hyperstimulation syndrome without associated fetal heart rate changes (Morning admission 10/280 (3.6%) versus Evening admission 7/340 (2.1%); RR 1.73 95% CI 0.67-4.48; p=0.260), need for any analgesia (Morning admission 259/280 (92.5%) versus Evening admission 324/340 (95.3%); RR 0.97 95% CI 0.93- 1.01; p=0.140) or need for epidural analgesia (Morning admission 190/280 (67.9%) versus Evening admission 209/340 (61.5%); RR 1.10 95% CI 0.98-1.23; p=0.117) was not significantly different for women admitted in the morning compared with women admitted in the evening (Table 5.4(4)). The presence of meconium stained liquor (Morning admission 42/280 (15.0%) versus Evening admission 51/340 (15.0%); RR 0.99 95% CI 0.68-1.44; p=0.962), instrumental vaginal birth (Morning admission 47/280 (16.8%) versus Evening admission 62/340 (18.2%); RR 0.92 95% CI 0.65-1.30; p=0.636), and length of labour (Morning admission mean interval 7.2 hours (standard deviation 4.1) versus Evening admission mean interval 7.4 hours (standard deviation
107 4.2); p=0.555) was not different between morning and evening admission for induction to start. The number of women with blood loss greater than 600mL (Morning admission 55/280 (19.6%) versus Evening admission 56/340 (16.5%); RR 1.19 95% CI 0.85-1.67; p=0.307), greater than 1000mL (Morning admission 11/280 (3.9%) versus Evening admission 16/340 (4.7%); RR 0.83 95% CI 0.39-1.76; p=0.627), or needing blood transfusion (Morning admission 6/280 (2.1%) versus Evening admission 8/340 (2.4%); RR 0.91 95% CI 0.32-2.59; p=0.856) did not differ between morning and evening admission.
Women with an initial Bishop’s score of 4-6 (Morning admission 12/103 (11.7%) versus Evening admission 31/137 (22.6%); RR 0.51 95% CI 0.28-0.95; p=0.028) and in particular nulliparous women with an initial Bishop’s score of 4-6 (Morning admission 10/62 (16.1%) versus Evening admission 28/82 (34.2%); RR 0.47 95% CI 0.25-0.90; p=0.015) were less likely to require instrumental vaginal birth when admission occurred in the morning compared with the evening. There were no other effects identified between time of starting the induction process and the induction agent used
(misoprostol versus vaginal PGE2 gel), maternal parity (nulliparous versus multiparous), or initial Bishop’s score (Bishop’s score 0-3 versus Bishop’s score 4-6) for labour and birth complications.
Table 5.4(4) Time of Commencing Induction – Secondary Outcomes: Labour and Birth Complications
Morning Admission Evening Admission Treatment Effect P value RR 95% CI All women N=280 % N=340 % - Uterine HSS - no FHR changes 10 3.6 7 2.1 1.73 0.67- 4.48 0.260 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 259 92.5 324 95.3 0.97 0.93- 1.01 0.140 - Need for epidural 190 67.9 209 61.5 1.10 0.98- 1.23 0.117 - Meconium stained liquor 42 15.0 51 15.0 0.99 0.68- 1.44 0.962 - Length of labour* 7.2 4.1 7.4 4.2 0.555 - Instrumental vaginal birth 47 16.8 62 18.2 0.92 0.65- 1.30 0.636 - Blood loss >600mL 55 19.6 56 16.5 1.19 0.85- 1.67 0.307 - Blood loss >1000mL 11 3.9 16 4.7 0.83 0.39- 1.76 0.627 - Need for blood transfusion 6 2.1 8 2.4 0.91 0.32- 2.59 0.856
108 Morning Admission Evening Admission Treatment Effect P value RR 95% CI Misoprostol N=138 % N=167 % - Uterine HSS – no FHR changes 1 0.7 2 1.2 0.60 0.05- 6.53 0.673 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 133 96.4 162 97.0 0.99 0.95- 1.03 0.717 - Need for epidural 101 73.2 108 64.7 1.12 0.97- 1.31 0.132 - Meconium stained liquor 25 18.1 25 15.0 1.20 0.73- 2.00 0.472 - Length of labour* 7.7 4.2 7.7 4.2 0.938 - Instrumental vaginal birth 28 20.3 30 18.0 1.13 0.71- 1.79 0.607 - Blood loss >600mL 22 15.9 25 15.0 1.06 0.63- 1.80 0.826 - Blood loss >1000mL 3 2.2 10 6.0 0.36 0.10- 1.29 0.101 - Need for blood transfusion 0 0.0 7 4.2 0.08 0.00- 1.40 0.151
Vaginal PGE2 N=142 % N=173 % - Uterine HSS – no FHR changes 9 6.3 5 2.9 2.19 0.75- 6.40 0.151 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 126 88.7 162 93.6 0.95 0.88- 1.02 0.131 - Need for epidural 89 62.7 101 58.4 1.07 0.90- 1.28 0.449 - Meconium stained liquor 17 12.0 26 15.0 0.79 0.44- 1.39 0.408 - Length of labour* 6.7 3.9 7.1 4.1 0.423 - Instrumental vaginal birth 19 13.4 32 18.5 0.72 0.43- 1.22 0.224 - Blood loss >600mL 33 23 2 31 17.9 1.30 0.84- 2.01 0.242 - Blood loss >1000mL 8 5.6 6 3.5 1.62 0.58- 4.57 0.353 - Need for blood transfusion 6 4.2 1 0.6 7.31 0.89-60.01 0.289 Bishop’s Score 0-3 N=177 % N=203 % - Uterine HSS - no FHR changes 8 4.5 3 1.5 3.06 0.82-11.35 0.078 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 165 93.2 195 96.1 0.97 0.92- 1.02 0.216 - Need for epidural 122 68.9 131 64.5 1.07 0.93- 1.23 0.365 - Meconium stained liquor 28 15.8 34 16.8 0.94 0.60- 1.49 0.807 - Length of labour* 7.2 4.1 7.6 4.1 0.433 - Instrumental vaginal birth 35 19.8 31 15.3 1.29 0.83- 2.01 0.248 - Blood loss >600mL 37 20.9 32 15.8 1.33 0.86- 2.03 0.195 - Blood loss >1000mL 9 5.1 9 4.4 1.15 0.47- 2.83 0.766 - Need for blood transfusion 5 2.8 4 2.0 1.43 0.39- 5.26 0.585 Bishop’s Score 4-6 N=103 % N=137 % - Uterine HSS - no FHR changes 2 1.9 4 2.9 0.67 0.12- 3.56 0.631 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 94 91.3 129 94.2 0.97 0.90- 1.04 0.386 - Need for epidural 68 66.0 78 56.9 1.16 0.95- 1.42 0.154 - Meconium stained liquor 14 13.6 17 12.4 1.10 0.57- 2.12 0.787 - Length of labour* 7.1 4.2 7.1 4.3 0.941 - Instrumental vaginal birth 12 11.7 31 22.6 0.51 0.28- 0.95 0.028 - Blood loss >600mL 18 17.5 24 17.5 1.00 0.57- 1.74 0.993 - Blood loss >1000mL 2 1.9 7 5.1 0.38 0.08- 1.79 0.201 - Need for blood transfusion 1 1.0 4 2.9 0.33 0.04- 2.93 0.295 Nulliparous women N=164 % N=201 % - Uterine HSS - no FHR changes 8 4.9 6 3.0 1.63 0.58- 4.60 0.356 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 157 95.7 198 98.5 0.97 0.94- 1.01 0.124 - Need for epidural 125 76.2 140 69.7 1.08 0.96- 1.23 0.206 - Meconium stained liquor 28 17.1 32 15.9 1.06 0.67- 1.69 0.795 - Length of labour* 8.9 4.2 9.1 4.3 0.617 - Instrumental vaginal birth 40 24.4 50 24.9 0.98 0.68- 1.41 0.908 - Blood loss >600mL 40 24.4 39 19.4 1.26 0.85- 1.85 0.253 - Blood loss >1000mL 6 3.7 10 5.0 0.73 0.27- 1.97 0.533 - Need for blood transfusion 4 2.4 4 2.0 1.22 0.31- 4.81 0.776 Nulliparous women, BS 0-3 N=102 % N=119 % - Uterine HSS - no FHR changes 6 5.9 2 1.7 3.50 0.72-16.96 0.096 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 97 95.1 117 98.3 0.97 0.92- 1.02 0.173 - Need for epidural 77 75.5 87 73.1 1.03 0.88- 1.21 0.687 - Meconium stained liquor 19 18.6 22 18.5 1.01 0.58- 1.75 0.979 - Length of labour* 9.1 4.2 9.7 4.1 0.393 - Instrumental vaginal birth 30 29.4 22 18.5 1.59 0.98- 1.02 0.057 - Blood loss >600mL 26 25.5 24 20.2 1.26 0.78- 2.06 0.346 - Blood loss >1000mL 4 3.9 7 5.9 0.67 0.20- 2.21 0.504 - Need for blood transfusion 3 2.9 4 3.4 0.88 0.20- 3.82 0.859
109 Morning Admission Evening Admission Treatment Effect P value RR 95% CI Nulliparous women, BS 4-6 N= 62 % N= 82 % - Uterine HSS - no FHR changes 2 3.2 4 4.9 0.66 0.13- 3.50 0.623 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 60 96.8 81 98.8 0.98 0.93- 1.03 0.404 - Need for epidural 48 77.4 53 64.6 1.20 0.97- 1.48 0.097 - Meconium stained liquor 9 14.5 10 12.2 1.19 0.51- 2.75 0.684 - Length of labour* 8.5 4.2 8.5 4.4 0.986 - Instrumental vaginal birth 10 16.1 28 34.2 0.47 0.25- 0.90 0.015 - Blood loss >600mL 14 22.6 15 18.3 1.23 0.64- 2.36 0.525 - Blood loss >1000mL 2 3.2 3 3.7 0.88 0.15- 5.12 0.888 - Need for blood transfusion 1 1.6 0 0.0 3.95 0.16-95.41 0.249 Multiparous women N=116 % N=139 % - Uterine HSS - no FHR changes 2 1.7 1 0.7 2.38 0.22- 25.96 0.476 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 102 87.9 126 90.7 0.97 0.89- 1.06 0.487 - Need for epidural 65 56.0 69 49.6 1.13 0.89- 1.42 0.313 - Meconium stained liquor 14 12.1 19 13.7 0.87 0.46- 1.66 0.681 - Length of labour* 5.3 3.1 5.6 3.2 0.528 - Instrumental vaginal birth 7 6.0 12 8.6 0.70 0.28- 1.72 0.437 - Blood loss >600mL 15 12.9 17 12.2 1.06 0.55- 2.02 0.868 - Blood loss >1000mL 5 4.3 6 4.3 1.00 0.31- 3.18 0.994 - Need for blood transfusion 2 1.7 4 2.9 0.60 0.11- 3.21 0.548 Multiparous women, BS 0-3 N= 75 % N= 84 % - Uterine HSS - no FHR changes 2 2.7 1 1.2 2.24 0.21- 24.20 0.495 - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 68 90.7 78 92.9 0.98 0.89- 1.07 0.615 - Need for epidural 45 60.0 44 52.4 1.15 0.87- 1.51 0.334 - Meconium stained liquor 9 12.0 12 14.3 0.84 0.38- 1.88 0.671 - Length of labour* 5.2 2.8 5.6 3.0 0.445 - Instrumental vaginal birth 5 6.7 9 10.7 0.62 0.22- 1.77 0.369 - Blood loss >600mL 11 14.7 8 9.5 1.54 0.65- 3.62 0.318 - Blood loss >1000mL 5 6.7 2 2.4 2.80 0.56- 14.01 0.189 - Need for blood transfusion 2 2.7 0 0.0 5.59 0.27-114.66 0.132 Multiparous women, BS 4-6 N= 41 % N= 55 % - Uterine HSS - no FHR changes 0 0.0 0 0.0 Not estimable - Uterine rupture 0 0.0 0 0.0 Not estimable - Need for any analgesia 34 82.9 48 87.3 0.95 0.80- 1.13 0.551 - Need for epidural 20 48.8 25 45.5 1.07 0.70- 1.64 0.747 - Meconium stained liquor 5 12.2 7 12.7 0.96 0.33- 2.80 0.938 - Length of labour* 5.4 3.5 5.5 3.5 0.921 - Instrumental vaginal birth 2 4.9 3 5.5 0.89 0.16- 5.11 0.900 - Blood loss >600mL 4 9.8 9 16.4 0.60 0.20- 1.80 0.349 - Blood loss >1000mL 0 0.0 4 7.3 0.15 0.01- 2.68 0.078 - Need for blood transfusion 0 0.0 4 7.3 0.15 0.01- 2.68 0.078
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text. * Mean and standard deviation (interval in hours); Student’s T-Test HSS = hyperstimulation syndrome FHR = fetal heart rate
110 5.4(5) Time of Commencing Induction – Secondary Outcomes: Neonatal Complications
The number of infants with birth weight less than 2.5kg (Morning admission 8/280 (2.9%) versus Evening admission 11/340 (3.2%); RR 0.85 95% CI 0.35-2.09; p=0.731), Apgar score of less than seven at five minutes of age (Morning admission 1/280 (0.4%) versus Evening admission 6/340 (1.8%); RR 0.34 95% CI 0.05-2.17; p=0.101), cord pH of less than 7.18 (Morning admission 13/280 (11.8%) versus Evening admission 13/340 (9.4%); RR 1.20 95% CI 0.58-2.49; p=0.618), or admission to the neonatal intensive care unit (Morning admission 1/280 (0.4%) versus Evening admission 3/340 (0.8%); RR 0.60 95% CI 0.08-4.61; p=0.426) did not differ between morning and evening admission to start induction of labour (Table 5.4(5)). There were no infants with neonatal encephalopathy, and no infant deaths to 28 days of life.
There were no differential effects in outcomes related to the induction agent used
(misoprostol versus vaginal PGE2 gel), maternal parity (nulliparous versus multiparous), or initial Bishop’s score (Bishop’s score 0-3 versus Bishop’s score 4-6), and the time of admission for induction.
Table 5.4(5) Time of Commencing Induction – Secondary Outcomes: Neonatal Complications
Morning Admission Evening Admission Treatment Effect P value RR 95% CI All infants N=280 % N=340 % - Birth-weight <2.5kg 8 2.9 11 3.2 0.85 0.35- 2.09 0.731 - Apgar <7 @ 5 minutes 1 0.4 6 1.8 0.34 0.05- 2.17 0.101 - Cord pH <7.18 13 11.8 13 9.4 1.20 0.58- 2.49 0.618 - NICU admission 1 0.4 3 0.8 0.60 0.08- 4.61 0.426 - Neonatal encephalopathy 0 0.0 0 0.0 Not estimable - Postnatal death* 0 0.0 0 0.0 Not estimable Misoprostol N=138 % N=167 % - Birth-weight <2.5kg 5 3.6 7 4.2 0.82 0.27- 2.51 0.724 - Apgar <7 @ 5 minutes 0 0.0 2 1.2 0.24 0.01- 4.99 0.197 - Cord pH<7.18 4 7.3 6 10.3 0.65 0.19- 2.18 0.487 - NICU admission 0 0.0 2 1.2 0.24 0.01- 4.99 0.197 - Neonatal encephalopathy 0 0.0 0 0.0 Not estimable - Postnatal death* 0 0.0 0 0.0 Not estimable
111 Morning Admission Evening Admission Treatment Effect P value RR 95% CI
Vaginal PGE2 N=142 % N=173 % - Birth-weight <2.5kg 3 2.1 4 2.3 0.91 0.21- 3.97 0.898 - Apgar <7 @ 5 minutes 1 0.7 4 2.3 0.30 0.03- 2.69 0.256 - Cord pH <7.18 9 16.4 7 8.8 1.85 0.74- 4.65 0.190 - NICU admission 1 0.7 1 0.6 1.21 0.08-19.31 0.888 - Neonatal encephalopathy 0 0.0 0 0.0 Not estimable - Postnatal death* 0 0.0 0 0.0 Not estimable
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. *Neonatal Death = death of a live-born infant within 28 days of birth NICU = neonatal intensive care unit
5.4(6) Time of Commencing Induction – Secondary Outcomes: Maternal Complications
The likelihood of any maternal side effect (Morning admission 75/280 (26.8%) versus Evening admission 75/340 (22.3%); RR 1.21 95% CI 0.92-1.60; p=0.182), nausea (Morning admission 20/280 (7.1%) versus Evening admission 26/340 (7.7%); RR 0.93 95% CI 0.53-1.64; p=0.813), vomiting (Morning admission 6/280 (2.1%) versus Evening admission 8/340 (2.4%); RR 0.92 95% CI 0.32-2.62; p=0.877), diarrhoea (Morning admission 7/280 (2.5%) versus Evening admission 7/340 (2.1%); RR 1.18 95% CI 0.42-3.32; p=0.753), or flushing (Morning admission 4/280 (1.4%) versus Evening admission 5/340 (1.5%); RR 0.83 95% CI 0.20-3.50; p=0.936) did not differ for women admitted in the morning and women admitted in the evening (Table 5.4(6)). There were no women admitted to intensive care, no cases of hyperpyrexia or coma, and no maternal deaths. There were no differential effects in outcomes related to the induction agent used (misoprostol versus vaginal PGE2 gel), maternal parity (nulliparous versus multiparous), or initial Bishop’s score (Bishop’s score 0-3 versus Bishop’s score 4-6) and the time of admission for induction.
112 Table 5.4(6) Time of Commencing Induction – Secondary Outcomes:
Maternal Complications
Morning Admission Evening Admission Treatment Effect P value RR 95% CI All women N=280 % N=340 % - Any side effect 75 26.8 75 22.3 1.21 0.92- 1.60 0.182 - Nausea 20 7.1 26 7.7 0.93 0.53- 1.64 0.813 - Vomiting 6 2.1 8 2.4 0.92 0.32- 2.62 0.877 - Diarrhoea 7 2.5 7 2.1 1.18 0.42- 3.32 0.753 - Flushing 4 1.4 5 1.5 0.84 0.20- 3.50 0.936 - Intensive care unit admission 0 0.0 0 0.0 Not estimable - Hyperpyrexia 0 0.0 0 0.0 Not estimable - Coma 0 0.0 0 0.0 Not estimable - Maternal death 0 0.0 0 0.0 Not estimable Misoprostol N=138 % N=167 % - Any side effect 32 23.2 31 18.6 1.23 0.80- 1.91 0.347 - Nausea 8 5.8 9 5.4 1.07 0.43- 2.71 0.880 - Vomiting 1 0.7 3 1.8 0.41 0.04- 3.91 0.439 - Diarrhoea 3 2.2 2 1.2 1.74 0.30-10.25 0.541 - Flushing 2 1.5 3 1.8 0.81 0.14- 4.77 0.812
Vaginal PGE2 N=142 % N=173 % - Any side effect 43 30.3 44 25.4 1.19 0.84- 1.70 0.332 - Nausea 12 8.5 17 9.8 0.86 0.42- 1.74 0.673 - Vomiting 5 3.5 5 2.9 1.22 0.36- 4.12 0.752 - Diarrhoea 4 2.8 5 2.9 0.96 0.26- 3.50 0.950 - Flushing 2 1.4 2 1.2 1.22 0.17- 8.54 0.842
Statistically significant results (p<0.05) in bold text. Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect.
5.4(7) Time of Commencing Induction – Women’s Satisfaction and Preferences for Care at Discharge
All 620 women for whom satisfaction questionnaires were available, were satisfied with the care that they received (Table 5.4(7)). The number of women who were not satisfied (defined as either not at all satisfied or only mildly satisfied) with their induction were similar between the morning admission group and the evening admission group (Morning admission 56/280 (20.0%) versus Evening admission 59/340 (17.4%); RR 1.15 95% CI 0.82-1.59; p=0.418), as were those unsatisfied with their labour (Morning admission 54/280 (19.3%) versus Evening admission 82/340 (24.2%); RR 0.79 95% CI 0.58-1.08; p=0.136) and birth (Morning admission 33/280 (11.8%) versus Evening admission 53/340 (15.6%); RR 0.76 95% CI 0.51-1.14; p=0.192). Overall, 75.2% of women, given similar circumstances, would participate in the study again, and 83.1% of women would recommend a friend in similar circumstances participate in the study,
113 figures that were not significantly different between the morning and evening admission groups.
Table 5.4(7) Time of Commencing Induction of Labour – Women’s Satisfaction and Preferences for Care at Discharge
Morning Evening Treatment Effect P Admission Admission RR 95% CI value All women N=280 % N=340 % - Woman not satisfied with care 0 0.0 0 0.0 Not estimable - Not satisfied with IOL 56 20.0 59 17.4 1.15 0.82-1.59 0.418 - Not satisfied with labour 54 19.3 82 24.2 0.79 0.58-1.08 0.136 - Not satisfied with birth 33 11.8 53 15.6 0.76 0.51-1.14 0.192 - Participate in study again 216 77.1 250 73.5 1.05 0.96-1.15 0.291 - Recommend others participate in study 238 85.0 277 81.5 1.04 0.97-1.11 0.310
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text. IOL = induction of labour
5.4(8) Time of Commencing Induction - Women’s Likes and Dislikes at Discharge
Overall, women’s responses to their participation in the trial were positive, with 28.1% indicating that they “liked being involved in the study”, and 32.3% indicating that they “liked everything” during their induction and labour (Table 5.4(8)). Few women (1.9%) indicated that they “disliked being involved in the study”, or that they “didn’t like anything” during their induction and labour (8.7%), responses that did not differ for women in the morning admission group when compared with those in the evening admission group (Table 5.4(8)). Women liked the “reassurance that (their) baby was well” (58.9%), while 16.6% of women indicated that they didn’t like having their baby monitored. Again, these responses were not different between women in the morning admission group and those in the evening admission group. While 37.4% of women indicated that they had no side effects from the medication, and 10.0% few medication side effects, 19.8% of women disliked abdominal cramps, 13.7% disliked nausea or vomiting, 5.2% disliked hot flushes, 3.9% disliked headaches, and 1.9% disliked
114 diarrhoea. There were no differences in these responses between women in the morning admission group and those in the evening admission group.
Fewer women in the morning admission group indicated a lack of sleep to be a dislike during induction of labour when compared with women in the evening admission group (Morning admission 1/280 (0.4%) versus Evening admission 15/340 (4.4%); RR 0.08 95% CI 0.01-0.61; p=0.015).
Table 5.4(8) Time of Commencing Induction - Women’s Likes and Dislikes at Discharge
Morning Evening Treatment Effect P value Admission Admission RR 95% CI All women N=280 % N=340 % I Liked… - Didn’t like anything 19 9.6 35 10.3 0.66 0.39-1.13 0.127 - No side effects from medication 97 34.6 135 39.7 0.87 0.71-1.07 0.188 - Few side effects from medication 33 11.8 29 8.5 1.38 0.86-2.22 0.180 - Reassurance baby was well 166 59.3 199 58.5 1.01 0.89-1.16 0.845 - Being involved in this study 88 31.4 86 25.3 1.25 0.97-1.61 0.082 I Disliked… - Like everything 99 35.4 101 29.7 1.20 0.96-1.51 0.116 - Nausea or vomiting 36 12.9 49 14.4 0.90 0.60-1.34 0.590 - Diarrhoea 7 2.5 5 1.5 1.71 0.55-5.33 0.356 - Abdominal cramps 47 16.8 76 22.4 0.75 0.54-1.04 0.083 - Headaches 12 4.3 12 3.5 1.21 0.55-2.66 0.629 - Hot flushes 13 4.6 19 5.6 0.83 0.42-1.66 0.601 - Too many internal examinations 73 26.1 99 29.1 0.87 0.67-1.12 0.285 - Having the baby monitored 39 13.9 64 18.8 0.74 0.51-1.07 0.107 - Being involved in this study 6 2.1 9 2.7 0.82 0.30-2.28 0.702 - Lack of sleep 1 0.4 15 4.4 0.08 0.01-0.61 0.015
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text.
115 5.4(9) Time of Commencing Induction – Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum
All 617 (99.5%) women for whom satisfaction questionnaires were available, were satisfied with the care that they received (Table 5.4(9)). The number of women who were not satisfied (defined as either not at all satisfied or only mildly satisfied) with their induction were similar between the morning admission group and the evening admission group (Morning admission 60/278 (21.6%) versus Evening admission 77/339 (22.7%); RR 0.94 95% CI 0.70-1.26; p=0.674), as were those unsatisfied with their labour (Morning admission 60/278 (21.6%) versus Evening admission 80/339 (23.6%); RR 0.90 95% CI 0.67-1.21; p=0.495) and birth (Morning admission 18/278 (6.5%) versus Evening admission 26/339 (7.7%); RR 0.91 95% CI 0.61-1.36; p=0.651). Overall, 71.0% of women, given similar circumstances, would participate in the study again, and 80.2% of women would recommend a friend in similar circumstances participate in the study, figures that were not significantly different between the morning and evening admission groups.
Table 5.4(9) Time of Commencing Induction – Women’s Satisfaction and Preferences for Care at 6 weeks Postpartum
Morning Evening Treatment Effect P Admission Admission RR 95% CI value All women N=278 % N=339 % - Woman not satisfied with care 0 0.0 0 0.0 Not estimable - Not satisfied with IOL 60 21.6 77 22.7 0.94 0.70-1.26 0.674 - Not satisfied with labour 60 21.6 80 23.6 0.90 0.67-1.21 0.495 - Not satisfied with birth 18 6.5 26 7.7 0.91 0.61-1.36 0.651 - Participate in study again 202 72.7 236 69.6 1.04 0.94-1.15 0.465 - Recommend others participate in study 231 83.1 264 77.9 1.07 0.99-1.15 0.105
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text. IOL = induction of labour
116 5.4(10) Time of Commencing Induction - Women’s Likes and Dislikes at 6 weeks Postpartum
Overall, women’s responses to their participation in the trial were positive, with 31.4% indicating that they “liked being involved in the study”, and 30.6% indicating that they “liked everything” during their induction and labour (Table 5.4(10)). Few women (5.5%) indicated that they “disliked being involved in the study”, or that they “didn’t like anything” during their induction and labour (9.9%), responses that did not differ for women in the morning admission group when compared with those in the evening admission group (Table 5.4(10)). Women liked the “reassurance that (their) baby was well” (61.3%), while 16.7% of women indicated that they didn’t like having their baby monitored. Again, these responses were not different between women in the morning admission group and those in the evening admission group. While 44.2% of women indicated that they had no side effects from the medication, and 8.1% few medication side effects, 17.0% of women disliked abdominal cramps, 13.6% disliked nausea or vomiting, 4.5% disliked headaches, 3.9% disliked hot flushes, and 1.8% disliked diarrhoea. There were no differences in these responses between women in the morning admission group and those in the evening admission group.
Fewer women in the morning admission group indicated a lack of sleep to be a dislike during induction of labour when compared with women in the evening admission group (Morning admission 0/278 (0.0%) versus Evening admission 12/339 (3.5%); RR 0.10 95% CI 0.01-0.76; p=0.002).
117 Table 5.4(10) Time of Commencing Induction - Women’s Likes and Dislikes at 6 weeks Postpartum
Morning Evening Treatment Effect P value Admission Admission RR 95% CI All women N=278 % N=339 % I Liked… - Didn’t like anything 26 9.4 35 9.4 0.98 0.60-1.60 0.923 - No side effects from medication 120 43.2 153 45.1 0.95 0.80-1.14 0.599 - Few side effects from medication 24 8.6 26 7.7 1.13 0.66-1.92 0.662 - Reassurance baby was well 168 60.4 210 61.9 0.97 0.86-1.10 0.655 - Being involved in this study 84 30.2 110 32.4 0.93 0.73-1.17 0.531 I Disliked… - Like everything 94 33.8 95 28.0 1.22 0.96-1.55 0.100 - Nausea or vomiting 38 13.7 46 13.6 1.00 0.67-1.50 0.990 - Diarrhoea 7 2.5 4 1.2 2.09 0.62-7.07 0.236 - Abdominal cramps 42 15.1 63 18.6 0.81 0.57-1.16 0.250 - Headaches 9 3.2 19 5.6 0.58 0.27-1.26 0.166 - Hot flushes 11 4.0 13 3.8 1.02 0.46-2.23 0.969 - Too many internal examinations 86 30.9 97 28.6 1.06 0.84-1.36 0.615 - Having the baby monitored 42 15.1 61 18.0 0.84 0.58-1.20 0.326 - Being involved in this study 13 4.7 21 6.2 0.74 0.38-1.46 0.387 - Lack of sleep 0 0.0 12 3.5 0.10 0.01-0.76 0.002
Figures are numbers and percentages with relative risk (RR) and 95% Confidence Intervals (CI) as treatment effect. Statistically significant results (p<0.05) in bold text.
5.5 DISCUSSION
As far as can be ascertained, this is the first randomised controlled trial that has assessed the impact of time of commencing induction of labour on maternal and infant outcomes. Traditionally, women have been admitted to hospital to commence induction of labour 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 well documented diurnal variation in onset of labour, which consistently demonstrate a peak onset of labour in the evening and early hours of the morning (Cooperstock 1987; Fraser 1989; Cagnacci 1998; Lindow 2000; Mancuso 2004).
118 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, caesarean section, or caesarean section for fetal distress, when compared with admission in the evening.
Women with a Bishop’s score of 0-3 on admission, nulliparous women, and in particular nulliparous women with a Bishop’s score of 0-3, required less oxytocin during labour, and had a shorter induction to birth interval when admission occurred in the morning. However, there were no differences in the length of labour documented. Women with a Bishop’s score of 4-6 were less likely to require instrumental vaginal birth when admission occurred in the morning, compared with evening admission, as were nulliparous women with an initial Bishop’s score of 4-6.
It is unclear if these observations reflect the well-documented physiological diurnal rhythm in the onset of labour in the evening and early hours of the morning, or if they are indicative of diurnal variations in clinical practice. With women admitted in the morning for cervical ripening and induction, the onset of labour would be expected to occur in the evening. Staffing levels within maternity units tend to be lower in the evening, and there may be a general reluctance to actively manage labour 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 labour 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 assisted vaginal birth may reflect a degree of impatience and pressure to have women birth at a time more convenient for caregivers.
While there were no significant differences identified in other labour and birth outcomes between morning and evening admission for induction, there was a trend to a reduction in the occurrence of major postpartum haemorrhage (defined as blood loss greater than 1000mL), and need for blood transfusion when admission occurred in the morning. These findings are consistent with the increased use of oxytocin in women admitted in the evening. While the overall number of women with blood loss greater
119 than 600mL is high at almost 18%, women requiring induction of labour are recognised to be at increased risk of obstetric haemorrhage. Furthermore, the rate includes women with caesarean birth, where the average blood loss is greater than that associated with vaginal birth.
For the rare, but serious neonatal complications, such as death and neonatal acidosis, the study was under-powered to detect all but large differences between time of starting induction of labour. While there is a recognised risk of adverse neonatal outcome associated with birth during the evening and night (Chalmers 1998; Stewart 1998; Heller 2000; Luo 2001), it would be necessary to recruit tens of thousands of women and their infants to assess this adequately in prospective randomised trials, a highly unlikely scenario to be achieved.
Overall, women were satisfied with the care they received and their induction, labour and birth. Expression of likes and dislikes during the induction process did not differ between morning and evening admission, although predictably, more women in the evening admission group did not like the interruptions to sleep that were associated with the protocol. However, lack of sleep associated with active labour in the evening was not identified by women in the morning admission group as a dislike.
5.6 CONCLUSION
Admission in the morning to start induction of labour is advantageous, with fewer women needing an oxytocin infusion during labour, fewer women requiring instrumental vaginal birth, and a shorter induction to birth interval when compared with admission in the evening. While there were no differences in degree of maternal satisfaction with their induction, labour and birth, more women in the evening admission group disliked the interruptions overnight and lack of sleep associated with the trial protocol.
120 Women requiring induction of labour, and particularly nulliparous women should be offered admission in the morning, rather than the traditional admission in the evening. This has implications for midwifery and medical staffing levels which currently are greater during daylight hours.
121 6. COST COMPARISON
6.1 INTRODUCTION
There has been a rapid increase in the use of misoprostol in obstetric practice, with some of this enthusiasm relating to its low cost when compared with other prostaglandin preparations. The current cost to the Women’s and Children’s Hospital for a 1mg dose of prostaglandin gel is (Aus)$40.17, with the cost of a 2mg dose increasing to (Aus)$51.64. The total cost to the hospital for the 1998/1999 financial year for prostaglandin gel alone was (Aus)$48,307, and for the 1999/2000 financial year (Aus)$38,508. In contrast a single 200mcg misoprostol tablet costs (Aus)35 cents.
More recent estimates of drug costs from 2001 to the present are not reliable, as the Women’s and Children’s Hospital met the cost of the drugs for this trial, therefore influencing the drug budget during this time.
There is an obvious cost differential between misoprostol and vaginal prostaglandin E2 gel relating to drug costs. However, a cost analysis should also take into consideration other costs to the institution associated with the two methods of induction of labour. Information derived from the current randomised controlled trial, will be used to detail the costs associated with each method of induction of labour, with clinical pathways built into the model.
6.2 STUDY HYPOTHESES
The use of oral misoprostol is associated with • reduced drug costs, and • reduced hospitalisation costs.
122 6.3 METHODS
Women were recruited from the Women’s and Children’s Hospital to a randomised controlled trial comparing oral misoprostol with vaginal PGE2 gel for induction of labour at term as described previously (Chapter 3). The perspective of the health care institution was chosen for this analysis, without consideration of the social and economic costs to the woman and her family.
Costs were specifically compared between oral misoprostol and vaginal prostaglandin
E2 induction of labour, as determined by baseline figures provided by the Women’s and Children’s Hospital business manager, to obtain a cost per woman induced and a cost per 100 women induced. Costs were further determined for both nulliparous and multiparous women with a Bishop’s score at commencement of induction of both zero to three and four to six (as described in Chapter 3). This involved using the 95% confidence intervals about the point estimate for the outcomes caesarean birth and further doses of known vaginal PGE2 gel; the interquartile range about the median for the outcome induction to birth interval; and the standard deviation about the mean for the outcome length of labour. Actual doses of trial medication were collected during the randomised trial.
The costs considered included the drug costs of misoprostol and vaginal PGE2 gel, midwifery care costs during induction and labour (estimated at the salary of a junior registered midwife, level 1.9), and hospital derived costs related to caesarean birth and vaginal birth (Table 6.3(1)). The midwifery care during induction was costed at a rate of one midwife caring for two women, and care during labour and birth was costed at a rate of 1.25 midwives caring for a single woman, according to hospital standards. The hospital derived costs related to vaginal birth and caesarean birth included costs related to oxytocin infusion, analgesia requirements, operating theatre costs and considered an average length of hospital stay on the postnatal ward.
123 Table 6.3(1) Costs to the Women’s and Children’s Hospital for Induction of Labour
Item Cost to WCH 1:2 Midwifery care (per hour) $13.04 per hour 1:1.25 Midwifery care (per hour) $32.60 per hour Misoprostol 35 cents per 200mcg tablet Vaginal PGE2 gel $40.17 per 1mg dose $51.64 per 2mg dose Vaginal birth (including average $3,973.83 length of hospital stay) Caesarean birth (including average $6,349.97 length of hospital stay)
6.4 RESULTS
The cost per woman induced with misoprostol was $4,948.61, compared with $5,059.64 per woman induced with vaginal PGE2 gel, a difference of $110.83 per woman induced in favour of misoprostol (range $15.88 to $121.87) (Table 6.4(1)). For nulliparous women with an initial Bishop’s score of 0-3, the cost per induction with misoprostol was $5,331.42, compared with $5,533.84 per induction with vaginal PGE2 gel, a difference of $202.42 in favour of misoprostol (range $150.12 to $272.73) (Table 6.4(2)). The cost per induction with misoprostol for nulliparous women with an initial Bishop’s score of 4-6 was $5,046.14, compared with $5,156.25 per induction with vaginal PGE2 gel, a difference of $110.11 in favour of misoprostol (range $55.32 to $176.31) (Table 6.4(3)). For multiparous women, the cost per induction with misoprostol was $4,574.67 where the initial Bishop’s score was 0-3, and $4,487.82 where the initial Bishop’s score was 4-6 (Tables 6.4(4) and 6.4(5)). This equates to a difference in cost of $186.98 (range $68.78 in favour of vaginal PGE2 to $300.66 in favour of misoprostol), and $14.42 (range $49.37 in favour of vaginal PGE2 to $31.40 in favour of misoprostol) respectively, when compared with induction with vaginal
PGE2 gel.
124 Table 6.4(1) Costs for all women
Outcome measured Misoprostol Cost PGE2 Group Cost Group Australian $ N=376 Australian $ N=365 Induction to birth interval* - Length of 13.7 65 207.25 11.5 56 384.96 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 7.5 89 242.50 6.9 84 577.44 Doses of trial medication 1 668 583.80 342 x 2mg 17 660.88 239 x 1mg 9 600.63
Further doses known 2mg PGE2 gel 70 3 614.80 47 2 427.08 Vaginal birth 282 1 120 620.00 276 1 096 777.00 Caesarean birth 83 527 047.51 100 634 997.00 Total Cost 1 806 315.80 1 902 424.90 Total Cost per Woman 4 948.81 5 059.64 110.83 Cost per 100 Women induced 494 881.05 505 964.09
Outcome measured – Lower Threshold Misoprostol Cost PGE2 Group Cost Group Australian $ N=376 Australian $ N=365 Induction to birth interval* - Length of 5.2 24 749.92 2.9 35 547.04 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 3.4 40 456.60 3.4 16 670.34 Doses of trial medication 1 668 583.80 342 x 2mg 17 660.88 239 x 1mg 9 600.63
Further doses known 2mg PGE2 gel 42 2 168.88 28 1 445.92 Vaginal birth 297 1 180 227.50 294 1 168 306.00 Caesarean birth 68 431 797.96 82 520 697.54 Total Cost 1 679 984.60 1 736 587.60 Total Cost per Woman 4 602.70 4 618.58 15.88 Cost per 100 Women induced 460 269.76 461 858.42
Outcome measured – Upper Threshold Misoprostol Cost PGE2 Group Cost Group Australian $ N=376 Australian $ N=365 Induction to birth interval* - Length of 22.2 105 663.12 19.6 96 099.58 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 11.6 138 028.40 10.9 133 607.84 Doses of trial medication 1 668 583.80 342 x 2mg 17 660.88 239 x 1mg 9 600.63
Further doses known 2mg PGE2 gel 109 5 628.76 73 3 769.72 Vaginal birth 262 1 041 143.40 252 1 001 405.10 Caesarean birth 103 654 046.91 124 787 396.28 Total Cost 1 945 094.30 2 049 540.00 Total Cost per Woman 5 329.03 5 450.90 121.87 Cost per 100 Women induced 532 902.57 545 090.43
*Median interval in hours ^Mean interval in hours
125 Table 6.4(2) Costs nulliparous women with Bishop’s score 0-3
Outcome measured Misoprostol Cost PGE2 Group Cost Group Australian $ N=126 Australian $ N=143 Induction to birth interval* - Length of 16.2 30 208.46 14.1 23 166.86 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 9.6 44 753.28 8.7 35 736.12 Doses of trial medication 755 264.25 208 10 741.12
Further doses known 2mg PGE2 gel 48 2 478.72 19 981.16 Vaginal birth 94 373 540.02 73 290 089.59 Caesarean birth 49 311 148.53 53 336 548.41 Total Cost 762 393.26 697 263.26 Total Cost per Woman 5 331.42 5 533.84 202.42 Cost per 100 women induced 533 142.13 553 383.53
Outcome measured – Lower Threshold Misoprostol Cost PGE2 Group Cost Group Australian $ N=126 Australian $ N=143 Induction to birth interval* - Length of 9.1 16 968.95 5.3 8 708.11 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 5.4 25 173.72 4.6 18 894.96 Doses of trial medication 755 264.25 208 10 741.12
Further doses known 2mg PGE2 gel 8 413.12 3 154.92 Vaginal birth 104 413 278.32 84 333 801.72 Caesarean birth 39 247 648.83 42 266 698.74 Total Cost 703 747.19 638 999.57 Total Cost per Woman 4 921.31 5 071.43 150.12 Cost per 100 women induced 492 130.90 507 142.51
Outcome measured – Upper Threshold Misoprostol Cost PGE2 Group Cost Group Australian $ N=126 Australian $ N=143 Induction to birth interval* - Length of 23.9 44 566.81 22.9 37 625.62 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 13.2 61 535.76 12.8 52 577.28 Doses of trial medication 755 264.25 208 10 741.12
Further doses known 2mg PGE2 gel 113 5 835.32 45 2 323.80 Vaginal birth 79 313 932.57 57 226 508.31 Caesarean birth 64 406 398.08 69 438 147.93 Total Cost 832 532.79 767 924.06 Total Cost per Woman 5 821.91 6 094.64 272.73 Cost per 100 women induced 582 190.76 609 463.53
*Median interval (in hours) ^Mean interval (in hours)
126 Table 6.4(3) Costs for nulliparous women with Bishop’s score 4-6
Outcome measured Misoprostol Cost PGE2 Cost Group Australian $ Group Australian $ N=70 N=95 Induction to birth interval* - Length of 13.0 11 866.40 10.3 12 759.64 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 8.7 19 853.40 8.3 25 705.10 Doses of trial medication 305 106.75 134 6 919.76
Further doses known 2mg PGE2 gel 9 464.76 8 413.12 Vaginal birth 52 206 639.16 67 266 246.61 Caesarean birth 18 114 299.46 28 177 799.16 Total Cost 353 229.93 489 843.39 Total Cost per Woman 5 046.14 5 156.25 110.11 Cost per 100 women induced 504 614.18 515 624.62
Outcome measured – Lower Threshold Misoprostol Cost PGE2 Cost Group Australian $ Group Australian $ N=70 N=95 Induction to birth interval* - Length of 7.0 6 389.60 3.5 4 335.80 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 4.5 10 269.00 4.1 12 697.70 Doses of trial medication 305 106.75 134 6 919.76
Further doses known 2mg PGE2 gel 1 51.64 1 51.64 Vaginal birth 58 230 482.14 77 305 984.91 Caesarean birth 12 76 199.64 18 114 299.46 Total Cost 323 498.77 444 289.27 Total Cost per Woman 4 621.41 4 676.73 55.32 Cost per 100 women induced 462 141.10 467 672.91
Outcome measured – Upper Threshold Misoprostol Cost PGE2 Cost Group Australian $ Group Australian $ N=70 N=95 Induction to birth interval* - Length of 19.0 17 343.20 17.1 21 183.48 Labour (1:2 midwifery care) Length of labour^ (1:1.25 midwifery care) 12.9 29 437.80 12.5 38 712.50 Doses of trial medication 305 106.75 134 6 919.76
Further doses known 2mg PGE2 gel 29 1 497.56 26 1 342.64 Vaginal birth 42 166 900.86 51 202 665.33 Caesarean birth 28 177 799.16 44 279 398.68 Total Cost 393 085.33 550 222.39 Total Cost per Woman 5 615.50 5 791.81 176.31 Cost per 100 women induced 561 550.47 579 181.46
* Median interval (in hours) ^Mean interval (in hours)
127 Table 6.4(4) Costs for multiparous women with Bishop’s score 0-3
Outcome measured Misoprostol Cost PGE2 Cost Group Australian $ Group Australian $ N=103 N=88 Induction to birth interval* - Length of 11.1 14 908.63 12.5 14 344.00 Labour (1:2 midwifery care) Length of labour^ (1:1 midwifery care) 5.3 17 796.34 5.3 15 204.64 Doses of trial medication 433 151.55 139 5 583.63
Further doses known 2mg PGE2 gel 10 516.40 18 929.52 Vaginal birth 91 361 618.53 74 294 063.42 Caesarean birth 12 76 199.64 14 88 899.58 Total Cost 471 191.09 419 024.79 Total Cost per Woman 4 574.67 4 761.65 186.98 Cost per 100 women induced 457 467.07 476 164.53
Outcome measured – Lower Threshold Misoprostol Cost PGE2 Cost Group Australian $ Group Australian $ N=103 N=88 Induction to birth interval* - Length of 3.6 12 088.08 4.2 4 819.58 Labour (1:2 midwifery care) Length of labour^ (1:1 midwifery care) 2.2 7 387.16 1.5 4 303.20 Doses of trial medication 433 151.55 139 5 583.63
Further doses known 2mg PGE2 gel 8 413.12 14 722.96 Vaginal birth 95 377 513.85 83 329 827.89 Caesarean birth 8 50 799.76 5 31 749.85 Total Cost 448 353.52 377 007.11 Total Cost per Woman 4 352.95 4 284.17 -68.78 Cost per 100 women induced 435 294.67 428 417.17
Outcome measured – Upper Threshold Misoprostol Cost PGE2 Cost Group Australian $ Group Australian $ N=103 N=88 Induction to birth interval* - Length of 18.6 24 982.03 21.4 24 556.93 Labour (1:2 midwifery care) Length of labour^ (1:1 midwifery care) 8.4 28 205.52 8.5 24 384.80 Doses of trial medication 433 151.55 139 5 583.63
Further doses known 2mg PGE2 gel 15 774.60 27 1 394.28 Vaginal birth 82 325 854.06 63 250 351.29 Caesarean birth 21 133 349.37 25 158 749.25 Total Cost 513 317.13 465 020.18 Total Cost per Woman 4 983.66 5 284.32 300.66 Cost per 100 women induced 498 366.14 528 432.02
*Median interval (in hours) ^Mean interval (in hours)
128 Table 6.4(5) Costs for multiparous women with Bishop’s score 4-6
Outcome measured Misoprostol Cost PGE2 Group Cost Group Australian $ N=67 Australian $ N=49 Induction to birth interval* - Length of 8.7 5 558.95 9.8 8 562.06 Labour (1:2 midwifery care) Length of labour^ (1:1 midwifery care) 6.2 9 903.88 5.0 10 921.00 Doses of trial medication 180 63.00 98 3 936.66
Further doses known 2mg PGE2 gel 3 154.92 2 103.28 Vaginal birth 45 178 822.35 62 246 377.46 Caesarean birth 4 25 399.88 5 31 749.85 Total Cost 219 902.98 301 650.31 Total Cost per Woman 4 487.82 4 502.24 14.42 Cost per 100 women induced 448 781.59 450 224.34
Outcome measured – Lower Threshold Misoprostol Cost PGE2 Group Cost Group Australian $ N=67 Australian $ N=49 Induction to birth interval* - Length of 2.1 1 341.82 4.6 4 018.93 Labour (1:2 midwifery care) Length of labour^ (1:1 midwifery care) 2.9 4 632.46 1.5 3 276.30 Doses of trial medication 180 63.00 98 3 936.66
Further doses known 2mg PGE2 gel 0 0.00 0 0.00 Vaginal birth 48 190 743.84 66 262 272.78 Caesarean birth 1 6 349.97 1 6 349.97 Total Cost 203 131.09 279 854.64 Total Cost per Woman 4 145.53 4 176.93 31.40 Cost per 100 women induced 414 553.24 417 693.49
Outcome measured – Upper Threshold Misoprostol Cost PGE2 Group Cost Group Australian $ N=67 Australian $ N=49 Induction to birth interval* - Length of 15.3 9 776.09 15.0 13 105.20 Labour (1:2 midwifery care) Length of labour^ (1:1 midwifery care) 9.5 15 175.30 8.5 18 565.70 Doses of trial medication 180 63.00 98 3 936.66
Further doses known 2mg PGE2 gel 32 1 652.48 22 1 136.08 Vaginal birth 34 135 110.22 48 190 743.84 Caesarean birth 15 95 249.55 19 120 649.43 Total Cost 257 026.64 348 136.91 Total Cost per Woman 5 245.44 5 196.07 -49.37 Cost per 100 women induced 524 544.16 519 607.32
*Median interval (in hours) ^Mean interval (in hours)
129 The drug costs per woman induced with misoprostol (including the need for further known doses of vaginal PGE2 gel) were $11.50 compared with $72.50 per woman induced with vaginal PGE2 gel, a difference of $61.00 per women in favour of misoprostol. The drug costs for women induced with misoprostol ranged from $19.18 per nulliparous woman with an initial Bishop’s score of 0-3, to $4.48 per multiparous woman with an initial Bishop’s score of 4-6. In contrast, the drug costs per woman induced with vaginal PGE2 gel ranged from $93.03 per nulliparous woman with an initial Bishop’s score of 0-3, to $60.30 per multiparous woman with an initial Bishop’s score of 4-6 (Figure 6.4(1)). The difference in drug costs per woman induced ranged from $73.85 per nulliparous woman with an initial Bishop’s score of 0-3, to $55.85 per multiparous woman with an initial Bishop’s score of 4-6, both in favour of misoprostol.
Figure 6.4(1) Drug Costs per Woman Induced
100 90 80 70 60 50 Misoprostol Vaginal PGE2 40
Cost (Aus $) 30 20 10 0 All Para0, Para0, Para1+, Para 1+, women BS0-3 BS4-6 BS0-3 BS4-6
130 6.5 DISCUSSION
Interest in the use of misoprostol as an induction of labour agent has rapidly increased, related in part to its lower cost when compared with other prostaglandin preparations used in obstetric practice. There is an obvious cost differential between oral misoprostol and vaginal PGE2 gel. A change to the use of misoprostol as the primary induction agent for women at term would result in a potential saving of around $11,000.00 for every 100 women induced to the Women’s and Children’s Hospital, even accounting for slightly reduced clinical efficacy, particularly in nulliparous women with an initial Bishop’s score of 0-3. However, even in this group of women, where there may be need for further doses of vaginal PGE2 gel, the potential savings in drug costs alone increase to $7,385.00 per 100 women induced with misoprostol.
In addition to this difference in drug costs, there are other costs associated with induction of labour. Any difference in efficacy between misoprostol and vaginal PGE2 gel needs to be balanced against a potential increase in adverse events for both the woman and infant. Several authors have attempted to elucidate these differences further (Sanchez-Ramos 1993; Wing 1995a; Kramer 1997; Sanchez-Ramos 1997), but none have been readily applicable to an Australian setting.
The results of this cost comparison indicate a modest saving not only in drug related costs, but also hospital related costs of the order of $11,000 for every 100 women induced with misoprostol. This difference is greatest for the induction of nulliparous women with an initial Bishop’s score of 0-3, where the cost differential is in excess of $20,000 for every 100 women induced with misoprostol. The difference in costs reflects not only a reduction in drug costs, but also a 15% reduction in caesarean section in this group of women. From a perspective considering cost alone, this modest saving offsets any reduction in efficacy as indicated by a need for further doses of known vaginal
PGE2 gel, oxytocin infusion during labour, and increased midwifery care costs as indicated by a longer induction to birth interval. However, the cost of midwifery care in this analysis utilised hourly rates for a junior midwife and did not take into account penalty salary rates, or the use of agency staff, who are often “bought-in” during busy periods. In these situations, salary costs would be greater than those considered in this comparison.
131
In an obstetric unit with approximately 1,400 inductions per year, as is the case for the Women’s and Children’s Hospital, changing to the use of misoprostol has the potential to save in excess of $154,000 annually.
While hospital administrators and health economists could make an argument supporting the use of misoprostol as an induction agent, it is unlikely that this will propel manufacturers towards seeking appropriate product licensing for the use of misoprostol in pregnancy. There is economic advantage to health service providers in the use of misoprostol for induction of labour, although its low production cost does not translate into financial incentives for the manufacturers. Efforts should be directed to ensure that a low dose (20mcg) formulation of misoprostol is available, is easy to administer orally for both women and their caregivers, and retains its current low cost to enable its widespread use, particularly in under-resourced countries.
6.6 CONCLUSION
There would be a cost saving to the Women’s and Children’s Hospital associated with a change from vaginal PGE2 gel to oral misoprostol as the primary induction of labour agent of the order of $154,000 annually. The use of misoprostol as an induction agent has potential for reduction in both drug and hospital care costs for developed nations, in addition to providing an economical method of induction of labour in under resourced countries.
132 7. MISOPROSTOL FOR INDUCTION OF LABOUR AT TERM – AN UPDATED META-ANALYSIS
7.1 INTRODUCTION
Systematic reviews collate information from primary RCTs that address the same (or similar) research question (Greenhalgh 1997b; Greenhalgh 1997c), allowing information to be assimilated and increasing the statistical power and available sample size (Sacks 1987; Mulrow 1994; Thacker 1998). It allows an assessment to be made of how applicable the findings are to the general population, in addition to the consistency of results across trials, both in direction and magnitude of effect (Sacks 1987; Mulrow 1994; Thacker 1998).
7.2 STUDY AIMS AND HYPOTHESES
The aims of this study are to incorporate the information generated from the randomised double blind placebo controlled trial comparing oral misoprostol with vaginal PGE2 gel into the currently available published literature, and to update the meta-analysis described in Chapter 2.
7.3 METHODS
The methods of the systematic review have been described previously in Chapter 2. The current randomised controlled trial conforms with the stated inclusion criteria and will be included in an updated meta-analysis.
133 7.4 RESULTS
7.4(1) Description of Studies
The characteristics of the identified randomised controlled trials are described in Chapter 2. The characteristics of the current randomised trial are described in detail in Chapter 3 and summarised below in Table 7.4(1).
Table 7.4(1) Summary characteristics of the current randomised trial
Trial Identification Trial Characteristics Dodd 2005 Methods: Trial conducted at the Women’s and Children’s Hospital, Lyell McEwin Health Service (South Australia) and Hervey Bay Hospital (Queensland) between April 2001 and December 2004 Participants: Women with a singleton pregnancy in cephalic presentation at 366 weeks gestation with an indication for induction of labour; exclusion criteria include active labour, Bishop’s score >7, contraindication to vaginal birth, pervious uterine surgery (including caesarean section), maternal history of asthma, glaucoma, or heart disease, ruptured membranes, parity >5, suspected cephalopelvic disproportion, abnormal fetal lie, breech presentation, non-reassuring fetal heart rate tracing, multiple pregnancy, intrauterine infection, vasa praevia or placenta praevia, active genital herpes, maternal illness precluding induction, maternal hepatic or renal failure Interventions: 20mcg oral misoprostol solution at 2 hourly intervals (maximum 6 doses in 12 hours) and vaginal placebo gel (at 6 hourly intervals to a maximum of 2 doses in 12 hours) vs vaginal PGE2 gel at 6 hourly intervals (maximum 2 doses in 12 hours) and oral placebo (vitamin B6 solution at 2 hourly intervals to a maximum of 6 doses in 12 hours) Sample size: 741 women Primary Outcomes: Vaginal birth not achieved in 24 hours, uterine hyperstimulation with fetal heart rate changes, caesarean section (all and for fetal distress) Randomisation: computer generated sequence using variable blocks with stratification for maternal parity (A) Allocation Concealment: Sequentially numbered identical appearing treatment packs (A) Blinding: All women, caregivers and outcome assessors blinded to allocated treatment
7.4(2) Meta-analysis
There were no differences in the number of women administered oral misoprostol achieving vaginal birth within 24 hours of the induction commencing when compared with women administered vaginal PGE2 gel (4 trials, 1781 participants, RR 1.11, 95% CI 0.99-1.24), with the magnitude and direction of effect similar between trials (Table
134 7.4(2)(i), Figure 7.4(2)(i)). There were no differences in the occurrence of uterine hyperstimulation syndrome with associated fetal heart rate changes (6 trials, 2128 participants, RR 0.80, 95% CI 0.52-1.23) (Figure 7.4(2)(ii)) or caesarean section (6 trials, 2158 participants, RR 0.91, 95% CI 0.77-1.06) (Figure 7.4(2)(iii)) between women administered oral misoprostol and women administered vaginal PGE2. There were no reports of serious neonatal (2 trial, 1008 participants, RR not estimable) or maternal morbidity (3 trials, 1073 participants, RR not estimable) in those trials reporting these outcomes.
Women administered oral misoprostol were more likely to have no evidence of cervical change after 12 or 24 hours (2 trials, 930 participants, RR 1.41, 95% CI 1.01-1.96) (Figure 7.4(2)(iv)), and to require an oxytocin infusion during labour (3 trials, 999 participants, RR 1.22, 95% CI 1.09-1.38) when compared with women administered vaginal PGE2 (Table 7.4(2)(ii)) (Figure 7.4(2)(v)).
For maternal and neonatal complications, there were no differences identified between oral misoprostol and vaginal PGE2 (Table 7.4(2)(iii) and Table 7.4(2)(iv)).
A single trial only reported maternal satisfaction with care, as described in Chapter 4.
Table 7.4(2)(i) Meta-analysis of Primary Outcomes
Outcome No. No. Relative Risk 95% CI Trials Participants Vaginal birth not achieved in 24 hours 4 1781 1.11 0.99-1.24 Hyperstimulation with FHR changes 6 2128 0.80 0.52-1.23 Caesarean section (all) 6 2158 0.91 0.77-1.06 Serious neonatal morbidity 2 1008 Not estimable Serious maternal morbidity 3 1703 Not estimable
135 Figure 7.4(2)(i) Vaginal birth not achieved in 24 hours
Figure 7.4(2)(ii) Uterine hyperstimulation with associated fetal heart rate changes
Figure 7.4(2)(iii) Caesarean section
136 Table 7.4(2)(ii) Meta-analysis of Secondary Outcomes – Evidence of Effect
Outcome No. No. Relative Risk 95% CI Trials Participants Cervix unchanged 12-24 hours 2 930 1.41 1.01- 1.96 Need for oxytocin augmentation 3 999 1.22 1.09- 1.38 Induction – birth interval* 4 1157 -1.38 -2.60- -0.15
Figure 7.4(2)(iv) Cervix unchanged after 12/24 hours
Figure 7.4(2)(v) Need for oxytocin infusion
137 Table 7.4(2)(iii) Meta-analysis of Secondary Outcomes – Maternal Complications
Outcome No. No. Relative Risk 95% CI Trials Participants Hyperstimulation no FHR changes 3 1603 0.77 0.51-1.16 Uterine rupture 3 1632 Not estimable Epidural analgesia 3 999 1.05 0.96-1.14 Instrumental vaginal birth 4 1833 0.89 0.74-1.07 Maternal side effects – any 3 1632 1.00 0.86-1.17 Maternal side effects – nausea 3 999 1.04 0.70-1.56 Maternal side effects – vomiting 3 1632 1.11 0.79-1.54 Maternal side effects – diarrhoea 2 200 0.33 0.01-8.09 Post partum haemorrhage 3 1633 0.90 0.74-1.11 Maternal Death 2 1433 Not estimable
Table 7.4(2)(iv) Meta-analysis of Secondary Outcomes – Neonatal Complications
Outcome No. No. Relative Risk 95% CI Trials Participants Meconium Stained Liquor 4 1199 1.11 0.85- 1.46 Apgar Score <7 at 5 minutes 4 1690 0.64 0.64- 1.24 Admission to NICU 4 1832 0.87 0.56- 1.37 Perinatal Death 3 1699 1.00 0.06-15.97
138 7.5 DISCUSSION
This updated meta-analysis includes 6 randomised controlled trials comparing oral misoprostol with vaginal PGE2 gel, and involves 2,158 women and their infants. While two of the included trials were double blind, 53.3% of the participants and their caregivers were aware of the treatment allocated. However, the findings of the largest randomised controlled trial to date are consistent in direction and magnitude of treatment effect with the other reported studies, including the two randomised trials comparing low dose oral misoprostol solution with vaginal PGE2 gel (Hofmeyr 2001; Dallenbach 2003).
The use of oral misoprostol was not associated with an increase in the chance of a woman not giving birth vaginally within 24 hours, caesarean section, or uterine hyperstimulation with associated fetal heart rate changes. Women who received oral misoprostol were more likely to have no evidence of cervical change after 12 or 24 hours, and to require oxytocin infusion during labour when compared with vaginal
PGE2 gel, a consistent effect across studies.
It was not possible to explore the effect of Bishop’s score at the start of induction (Bishop’s score 0-3 versus Bishop’s score 4-6), or the effect of maternal parity (nulliparous versus multiparous) to determine the existence of any differential treatment effect, as these parameters were reported in one study only (Chapter 3).
There was a consistent trend towards fewer caesarean sections in the oral misoprostol group, particularly in the studies utilising a low dose misoprostol solution, of the order of 10 to 15%. To reliably be able to detect a change of this magnitude would require a combined sample size of the order of 4,000 women, short of the achieved sample in this meta-analysis. The occurrence of uterine hyperstimulation with associated fetal heart rate changes was not different between women receiving oral misoprostol and women receiving vaginal PGE2 gel. However, to be able to detect the 20% difference suggested from the meta-analysis, from 4.1% in the vaginal PGE2 group to 3.3% in the oral misoprostol group, would require a sample size of almost 18,000 women.
139 The extent of rare but potentially serious adverse complications such as uterine rupture, maternal or perinatal death, and neonatal acidaemia remain uncertain, as the present sample size of the meta-analysis is underpowered to detect all but extraordinarily large differences. While it is unrealistic to expect randomised trials to be conducted that are able to recruit tens of thousands of women and their infant’s, regular audit of clinical practice and reporting of such adverse outcomes should be an essential requirement of clinicians and institutions adopting the use of misoprostol in the induction of labour.
The outcome vaginal birth not achieved in 24 hours as defined in the Cochrane generic protocol (Hofmeyr 2005a) comprises both women who birth vaginally beyond 24 hours and all women who birth by caesarean section. These two components reflect different processes whereby misoprostol may generate adverse health outcomes. If vaginal birth is achieved beyond 24 hours, it may reflect an inappropriately low dose of misoprostol. The alternative process contributing to adverse health outcomes relates to uterine hyperstimulation with fetal heart rate changes and subsequent birth by caesarean section for non-reassuring fetal heart rate trace. For completeness and to ensure clarity of information, future trials should report both components of this outcome.
7.6 CONCLUSION
Oral misoprostol is as effective as vaginal PGE2 gel for the induction of labour at or near term, with no increase in the chance of a woman not birthing vaginally within 24 hours, caesarean section, or uterine hyperstimulation with associated fetal heart rate changes. Oral misoprostol was associated with an increased need for further doses of vaginal PGE2 gel and oxytocin infusion, but these differences in the process of induction did not lead to poorer health outcomes for women or infants.
140 8. OVERALL CONCLUSIONS
8.1 ORAL MISOPROSTOL VERSUS VAGINAL
PROSTAGLANDIN E2 GEL FOR INDUCTION OF LABOUR
Oral misoprostol was not associated with differences in the number of women who achieve vaginal birth in 24 hours, uterine hyperstimulation with fetal heart rate changes, or caesarean section, compared with vaginal PGE2 gel.
Oral misoprostol was associated with an increased need for further doses of vaginal
PGE2 gel and oxytocin infusion, but a significant reduction in uterine hyperstimulation without fetal heart rate changes. Differences in the process of induction did not lead to poorer health outcomes for women or infants.
Oral misoprostol for the induction of labour is not associated with an increase in the risk of adverse health outcomes for the woman or her infant.
An initial Bishop’s score of 0-3 at the start of induction and nulliparity are both independently associated with greater difficulty in inducing labour, particularly when using misoprostol. These differences in the process of initiating labour are most pronounced among nulliparous women with an initial Bishop’s score of 0-3. However, these differences are not associated with an increase in the risk of adverse health outcomes for the woman or her infant.
For nulliparous women with a Bishop’s score of 0-3, a dose of 20mcg oral
misoprostol is less effective in achieving vaginal birth within 24 hours when
compared with vaginal PGE2 gel.
141 For multiparous women, 20mcg of oral misoprostol is as effective as vaginal PGE2 gel in the process of initiating labour, with no increased risk of adverse health outcomes for the woman or her infant.
For multiparous women, 20mcg of oral misoprostol is as effective as
vaginal PGE2 gel.
8.2 WOMEN’S PREFERENCES FOR CARE
Overall, women expressed a high degree of satisfaction with their induction, labour, and birth, and a strong preference for an oral induction of labour agent. Women in the oral misoprostol group were more likely to indicate that they “liked everything” associated with their induction and birth experience.
While women in the misoprostol group were more likely to remain undelivered after 24 hours, this did not impact negatively on their birth experience. Of women who were undelivered after 24 hours, more in the misoprostol group indicated that they “liked everything” associated with their induction and birth.
Women have a strong preference for an oral induction of labour agent.
Women administered misoprostol were more likely to be satisfied with all aspects
of their labour and birth experience.
8.3 TIME OF ADMISSION FOR INDUCTION OF LABOUR
Admission in the morning for induction to start was associated with no differences in the chance of a woman remaining undelivered after 24 hours when compared with evening admission. However, admission in the morning was associated with less need for oxytocin infusion, a shorter induction to birth interval, and less need for instrumental vaginal birth.
142
Overall, women expressed a high degree of satisfaction, although women admitted for induction in the evening were more likely to dislike interruptions to their sleep.
Women should be offered admission in the morning to start induction of labour
as it is associated with less need for intervention, a shorter induction to birth
interval, and less disruption to sleep.
8.4 COSTS ASSOCIATED WITH INDUCTION OF LABOUR
A change from the use of vaginal PGE2 gel to oral misoprostol as the primary induction of labour agent would be associated with a saving to the Women’s and Children’s Hospital, where approximately 1,400 women have their labour induced per annum, of $154,000 annually. This modest saving offsets any reduction in efficacy as indicated by a need for further doses of known vaginal PGE2 gel, oxytocin infusion during labour, and increased midwifery care costs as indicated by a longer induction to birth interval.
Use of oral misoprostol as the primary induction of labour agent has the
potential to save institutions such as the Women’s and Children’s Hospital,
where 1,400 women have their labour induced each year, $154,000 annually.
143 8.5 IMPLICATIONS FOR CLINICAL PRACTICE
The use of oral misoprostol for induction of labour is not associated with differences in a woman’s chance of not achieving vaginal birth within 24 hours, caesarean section, uterine hyperstimulation with associated fetal heart rate changes, or adverse health outcomes for the woman or her infant when compared with vaginal PGE2 gel.
The women’s own preferences were clear. Women expressed a strong preference for an oral induction of labour agent, with women administered misoprostol more likely to “like everything” associated with their labour and birth experience. Furthermore, women who remain undelivered after 24 hours and were administered misoprostol were more likely to indicate that that they “like(d) everything” associated with their birth experience. There was a modest cost saving in the use of misoprostol.
Potential Positive Effects in the use of Misoprostol compared with Vaginal PGE2
gel in the Induction of Labour
• No increase in the risk of not achieving vaginal birth within 24 hours
• No increase in the risk of uterine hyperstimulation with associated fetal
heart rate changes
• No increase in the risk of caesarean section
• No difference in the chance of labour and birth complications
• No difference in the chance of neonatal complications
• No difference in maternal side effects or complications
• Women express a strong preference for an oral induction of labour agent • Women “like everything” associated with their labour & birth experience • A cost saving to the institution
144 Use of oral misoprostol for induction of labour is associated with an increased chance of there being no cervical change after 24 hours, of needing further doses of vaginal gel, of needing an oxytocin infusion, of needing any analgesia during labour, and a longer induction to birth interval when compared with vaginal PGE2 gel.
While these relative differences that relate to the birth process are statistically significant, the absolute differences between the two treatment groups are of similar magnitude, and the clinical significance of the negative differences observed is offset by potential benefits.
Potential Negative Effects in the use of Misoprostol compared with Vaginal PGE 2 gel in the Induction of Labour
• Increased chance of no cervical change after 24 hours
(15% misoprostol versus 10% PGE ) 2 • Increased chance of needing further vaginal PGE gel 2 (19% misoprostol versus 13% PGE ) 2 • Increased chance of requiring oxytocin infusion during labour (56% misoprostol versus 48% PGE2) • Increased chance of needing any analgesia during labour (96% misoprostol versus 92% PGE2) • Longer induction to birth interval (21 hours misoprostol versus 18 hours PGE2)
145 Misoprostol should be available for use in the induction of labour of women at term. Misoprostol is not associated with an increase in the chance of remaining undelivered vaginally within 24 hours, caesarean section, uterine hyperstimulation with fetal heart rate changes, or adverse health outcomes for women or their infants. Women express a preference for an oral induction agent that is associated with a more positive birth experience, and its use is associated with modest cost savings to the institution.
Women should be offered admission in the morning to start induction of labour, as it is associated with a reduction in intervention, a shorter induction to birth interval, and less disruption to sleep. This is particularly the case for nulliparous women.
8.6 IMPLICATIONS FOR RESEARCH
Further information is required about the lowest effective dose of misoprostol in nulliparous women, and in particular those women with an initial Bishop’s score of 0-3. However, any increase in dose, while improving clinical efficacy must be balanced against a potential increase in side effects and adverse complications for the woman and her infant. While the use of titrated doses or oral misoprostol has been addressed in the trials by Hofmeyr (2001) and Dallenbach (2003), it would be interesting to explore any variations in the evidence of effect of misoprostol in nulliparous women with an initial Bishop’s score of 0-3 through subgroup analysis of their primary trial data.
The outcome vaginal birth not achieved in 24 hours as defined in the Cochrane generic protocol (Hofmeyr 2005a) comprises both women who birth vaginally beyond 24 hours and all women who birth by caesarean section. For completeness and to ensure clarity of information, future trials should report both components of this outcome.
Further studies are required to address the optimal timing of admission for induction of labour, and the suggestion of diurnal variations in clinical practice identified in the current trial. Any benefits of morning admission for the woman in terms of reduced intervention and a shorter induction to birth interval must be balanced against a
146 recognised risk of adverse neonatal outcome associated with birth during the evening and night. However, to assess the risk of rare adverse neonatal health outcomes adequately, it would be necessary to recruit tens of thousands of women and their infants in prospective randomised trials.
Efforts should be directed towards ensuring that pharmaceutical companies manufacture a low dose (20mcg) formulation of misoprostol that is easy to administer orally for both women and their caregivers, and retains its current low cost to enable its widespread use, particularly in under-resourced countries.
147 REFERENCES
Adair CD, Weeks JW, Barrilleaux PS, Edwards MS, Bulison K, Lewis DF (1998) Oral or vaginal misoprostol administration for induction of labor: a randomized, double blind trial. Obstetrics & Gynaecology 92:810-813. Akhan SE, Iyibozkurt AC, Turfanda A (2001) Unscarred uterine rupture after induction of labor with misoprostol - a case report. Clinical and Experimental Obstetrics and Gynecology 28:118-120. Al-Hussaini TK (2001) Uterine rupture in second trimester abortion in a grand multiparous woman. A complication of misoprostol and oxytocin. European Journal of Obstetrics, Gynecology, and Reproductive Biology 96:218-219. Alfirevic Z (2005) Oral misoprostol for induction of labour (Cochrane Review). In: The Cochrane Library, Issue1. John Wiley & Sons Ltd, Chichester, UK. Austin J, Ford MD, Rouse A, Hanna E (1997) Acute intravaginal misoprostol toxicity with fetal demise. Journal of Emergency Medicine. 15:61-64. Bartha JL, Comino-Delgado R, Garcia-Benasach F, Martinez-del-Fresno P, Moreno-Corral LJ (2000) Oral misoprostol and intracervical dinoprostone for cervical ripening and labor induction: a randomized comparison. Obstetrics & Gynaecology 96:465-469. Bell WB (1909) The pituitary body and the therapeutic value of the infundibular extract in shock, uterine atony, and intestinal paresis. British Medical Journal i:1069- 1613. Bennett BB (1997) Uterine rupture during induction of labor at term with intravaginal misoprostol. Obstetrics & Gynaecology 89:832-833. Bennett KA, Butt K, Crane JM, Hutchens D, Young DC (1998) A masked randomized comparison of oral and vaginal administration of misoprostol for labor induction. Obstetrics & Gynaecology 92:481-486. Berghahn L, Christenses D, Droste S (2001) Uterine rupture during second-trimester abortion associated with misoprostol. Obstetrics & Gynaecology 98:976-977. Bernstein D, Zer J, Zakut H (1989) Diurnal variations in unconjugated plasma and total urinary estriol levels in late normal pregnancy. Gynecol Endocrinol 3:99- 106.
148 Bishop EH (1964) Pelvic scoring for elective induction. Obstetrics & Gynaecology 24:266-268. Bond GR, Van Zee A (1994) Overdosage of misoprostol in pregnancy. American Journal of Obstetrics & Gynecology 171:561-562. Butt K, Bennett KA, Crane JM, Hutchens D, Young DC (1999) Randomized comparison of oral misoprostol and oxytocin for labor induction in term prelabor membrane rupture. Obstetrics & Gynaecology 94:994-999. Bygdeman M, Kwon SU, Mukherjee T, Wiqvist N (1968) Effect of intravenous infusion of prostaglandin E1 and E2 on motility of the pregnant human uterus. American Journal of Obstetrics & Gynecology 102:317-326. Cagnacci A, Soldani R, Melis GB, Volpe A (1998) Diurnal rhythms of labor and delivery in women: modulation by parity and seasons. American Journal of Obstetrics & Gynecology 178:140-145. Calder A (1997) Review of prostaglandin use in labour induction. British Journal of Obstetrics and Gynaecology 104:2-7. Calder AA, Embrey MP (1975) The management of labour. Proceeding of the third study group of RCOG:62-69. Calder AA, Embrey MP, Hillier K (1974) Extra-amniotic prostaglandin E2 for the induction of labour at term. Journal of Obstetrics and Gynaecology of the British Commonwealth 81:39-46. Calder AA, Embrey MP, Tait T (1977) Ripening of the cervix with extra-amniotic prostaglandin E2 in viscous gel before induction of labour. British Journal of Obstetrics and Gynaecology 84:264-268. Challis JR, Partrick JE, Campbell K, Natale R, Richardson B (1980) Diurnal changes in maternal plasma oestrone and oestradiol at 30 to 31, 34 to 35 and 38 to 39 weeks gestational age. British Journal of Obstetrics and Gynaecology 87:983-988. Chalmers I, Hetherington J, Elbourne D, Keirse M, Enkin M (1989) Materials and methods used in synthesizing evidence to evaluate the effects of care during pregnancy and childbirth. In: Chalmers I, Enkin M, Keirse M (eds) Effective care in pregnancy and childbirth. Vol. 1, pp 39-65. Oxford University Press, Oxford.
149 Chalmers JWT, Shanks E, Paterson S, McInneny K, Baird DT, Penney G (1998) Scottish data on intrapartum related deaths are in the same direction as Welsh data. British Medical Journal 317:539. Chalmers TC, Celano P, Sacks HS, Smith H (1983) Bias in treatment assignment in controlled clinical trials. New England Journal of Medicine 309:1358-1361. Chan A, Scott J, McCaul K, Keane R (1999) Pregnancy outcome in South Australia. Pregnancy Outcome Unit, South Australian Health Commission, Department of Health Services, Adelaide. Chen D, Tsay R, Lin H, Chen H, Chao S, Ku H (2000) Stabilisation and sustained release effect of misoprostol with methyacrylate copolymer. International Journal of Pharmaceutics 203:141-148. Chuck FJ, Huffaker BJ (1995) Labor induction with intravaginal misoprostol versus intracervical prostaglandin E2 gel (Prepidil gel): randomized comparison. American Journal of Obstetrics & Gynecology 173:1137-1142. Clark A, Cook V, Hill P, Spinnato J (1998) Cervical ripening and labor induction: misoprostol vs. dinoprostone. American Journal of Obstetrics & Gynecology 178:S30. Collins PW (1990) Misoprostol: discovery, development and clinical applications. Medicinal Research Reviews 10:149-172. Collins PW, Pappo R, Dajani EZ (1985) Chemistry and synthetic development of misoprostol. Digestive Diseases and Sciences 30:114S-117S. Cooperstock M, England JE, Wolfe RA (1987) Circadian incidence of labor onset hour in preterm birth and chorioamnionitis. Obstetrics & Gynaecology 70:852- 855. Craft I (1972) Amniotomy and oral prostaglandin E2 titration for induction of labour. British Medical Journal 2:191-194. Cunha M, Bugalho A, Bique C, Bergstrom S (1999) Induction of labor by vaginal misoprostol in patients with a previous cesarean delivery. Acta Obstet Gynecol Scand 78:653-654. Curtis P, Evans S, Resnick J (1987) Uterine hyperstimulation. The need for standard terminology. Journal of Reproductive Medicine 32:91-95. Daisley H (2000) Maternal mortality following the use of misoprostol. Med. Sci. Law. 40:78-82.
150 Dale HH (1906) On some physiological actions of ergot. Journal of Physiology 34:163- 206. Dallenbach P, Boulvain M, Viardot C, Irion O (2003) Oral misoprostol or vaginal dinoprostone for labor induction: a randomized controlled trial. American Journal of Obstetrics & Gynecology 188:162-167. Danforth DN (1947) The fibrous nature of the human cervix and its relation to the isthmic segment in gravid and nongravid uteri. American Journal of Obstetrics & Gynecology 53:541-560. Danforth DN, Buckingham JC, Roddick JW (1960) Connective tissue changes incident to cervical effacement. American Journal of Obstetrics & Gynecology 80:939-945. Danforth DN, Veis A, Breen M, Weinstein HG, Buckingham JC, Manalo P (1974) The effect of pregnancy and labor on the human cervix: changes in collagen, glycoproteins, and glycosaminoglycans. American Journal of Obstetrics & Gynecology 120:641-651. Danielsson KG, Marions L, Rodriguez A, Spur BW, Wong PY, Bygdeman M (1999) Comparison between oral and vaginal administration of misoprostol on uterine contractility. Obstetrics & Gynaecology 93:275-280. Davis ME, Adair FL, Rogers G, Kharasch MS, Legault RR (1935) A new active principle in ergot and its effects on uterine motility. American Journal of Obstetrics & Gynecology 29:155-167. Dickinson JE, Godfrey M, Evans SF (1999) Efficacy of intravaginal misoprostol in second trimester pregnancy termination: a randomized controlled trial. Journal of Maternal-Fetal Medicine 7:115-119. Diro M, Adra A, Gilles JM, Nassar AH, Rodriguez A, Salamat SM, al e (1999) A double-blind randomized trial of two dose regimens of misoprostol for cervical ripening and labor induction. Journal of Maternal-Fetal Medicine 8:114-118. Dray F, Frydman R (1976) Primary prostaglandins in amniotic fluid in pregnancy and spontaneous labor. American Journal of Obstetrics & Gynecology 126:13-19. du Vigneaud V, Ressler C, Swan JM, Roberts CW, Katsoyannis PG, Gordon S (1953) The synthesis of an octapeptide amide with the hormonal activity of oxytocin. Journal of the American Chemistry Society 75:4879-4880.
151 Dyar TR, Greig P, Cummings R, Nichols K (2000) The efficacy and safety of oral versus vaginal misoprostol for the induction of term labor. American Journal of Obstetrics & Gynecology 182:S135. Egarter CH, Husslein PW, Rayburn WF (1990) Uterine hyperstimulation after low- dose prostaglandin E2 therapy: tocolytic treatment in 181 cases. American Journal of Obstetrics & Gynecology 163:794-796. El-Azeem S, Samuels P, Welch G, Staisch K (1997) Term labor induction with PGE1 misoprostol versus PGE2 dinoprostone. American Journal of Obstetrics & Gynecology 176:S113. Elattar TMA (1978) Prostaglandins - a review of biochemistry, pharmacology and clinical applications. Journal of Oral Pathology 7:175-207. Ellwood DA, Mitchell MD, Anderson ABM, Turnbull AC (1980) The in-vitro production of prostanoids by the human cervix during pregnancy: preliminary observations. British Journal of Obstetrics and Gynaecology 87:210-214. Embrey MP (1969) The effect of prostaglandins on the human pregnant uterus. Journal of Obstetrics and Gynaecology of the British Commonwealth 76:783- 789. Embrey MP (1970) Induction of labour with prostaglandins E1 and E2. British Medical Journal 2:256-258. Embrey MP (1981) Prostaglandins in human reproduction. British Medical Journal 283:1563-1566. Farah LA, Sanchez-Ramos L, Rosa C, Del Valle GO, Gaudier FL, Delke I, Kaunitz AM (1997) Randomized trial of two doses of the prostaglandin E1 analogue misoprostol for labor induction. American Journal of Obstetrics & Gynecology 177:364-369. Ferguson J, Head B, Frank F, Frank M, Singer J, Stefos T, al e (2002) Misoprostol versus low-dose oxytocin for cervical ripening: a prospective, randomized, double-masked trial. American Journal of Obstetrics & Gynecology 187:273- 280. Fletcher HM, Mitchell S, Simeon D, Frederick J, Brown D (1993) Intravaginal misoprostol as a cervical ripening agent. British Journal of Obstetrics and Gynaecology 100:641-644.
152 Foote EF, Lee DR, Karim A, Keane WF, Halstenson CE (1995) Disposition of misoprostol and its active metabolite in patients with normal and impaired renal function. Journal of Clinical Pharmacology 35:384-389. Fraser WD, McLean FH, Usher RH (1989) Diurnal variation in admission to hospital of women in labour. Can J Surg 32:33-35. French L (2005) Oral prostaglandinE2 for induction of labour (Cochrane Review). In: The Cochrane Library, Issue 1. John Wiley & Sons Ltd, Chichester, UK. Fuchs AR, Fuchs F (1984) Endocrinology of parturition: a review. British Journal of Obstetrics and Gynaecology 91:948-967. Fuchs AR, Husslein PW, Fuchs F (1981) Oxytocin and the initiation of human parturition 2: Stimulation of prostaglandin production in human decidua by oxytocin. American Journal of Obstetrics & Gynecology 141:694-697. Garris RE, Kirkwood CF (1989) Misoprostol: a prostaglandin E1 analogue. Clinical Pharmacy 8:627-644. Gherman RB, Browning J, O'Boyle A, Goodwin TM (2001) Oral misoprostol vs intravaginal prostaglandin E2 for preinduction cervical ripening. A randomized trial. J Reprod Med 46:641-646. Gherman RB, McBrayer S, Browning J (2000) Uterine rupture associated with vaginal birth after caesarean section: a complication of intravaginal misoprostol? Gynecol Obstet Invest 50:212-213. Gordon-Wright AP, Dutt TP, Elder MG (1979) The routine use of oral prostaglandin E2 tablets for induction or augmentation of labour. Acta Obstet Gynecol Scand 58:23-26. Gordon-Wright AP, Elder MG (1979) Prostaglandin E2 tablets used intravaginally for the induction of labour. British Journal of Obstetrics and Gynaecology 86:32-36. Gottschall DS, Borgida AF, Feldman DM, Alberti W, Rodis JF (1998) Preinduction cervical ripening comparing 50 and 100mcg of misoprostol. American Journal of Obstetrics & Gynecology 178:S93. Government of South Australia (2005) South Australian Perinatal Guidelines, Vol. 2005 Graber D, Meier K (1991) Acute misoprostol toxicity. Annals of Emergency Medicine 20:549-551.
153 Gravett MG, Sampson JE (1994) Other Infectious Conditions. In: James DK, Steer PJ, Weiner CP, Gonik B (eds) High risk pregnancy management options pp 509- 550. W.B. Saunders Company Limited. Greenhalgh T (1997a) How to read a paper: Assessing the methodological quality of published papers. British Medical Journal 315:305-308. Greenhalgh T (1997b) How to read a paper: Papers that summarise other papers (systematic reviews and meta-analyses). British Medical Journal 315:672-675. Greenhalgh T (1997c) How to read a paper : the basics of evidence based medicine. 1st edition. BMJ Publishing Group, London. Gulmezoglu AM, Forna F, Villar J, Hofmeyr GJ (2005) Prostaglandins for prevention of post-partum haemorrhage (Cochrane Review). In: The Cochrane Library, Issue 1. John Wiley & Sons Ltd, Chichester, UK. Gunalp S, Bildirici I (2000) The effect of vaginal pH on the efficacy of vaginal misoprostol for induction of labor. Acta Obstet Gynecol Scand 79:283-285. Heller G, Misselwitz B, Schmidt S (2000) Early neonatal mortality, asphyxia related deaths and timing of low risk births in Hesse, Germany, 1990-8: an observational study. British Medical Journal 321:274-275. Hibbard BM, Sharma SD, Fitzpatrick RJ, Hamlett JD (1974) Prostaglandin F2alpha concentrations in amniotic fluid in late pregnancy. Journal of Obstetrics and Gynaecology of the British Commonwealth 81. Hillier K, Calder AA, Embrey MP (1974) Concentrations of prostaglandin F2alpha in amniotic fluid and plasma in spontaneous and induced labours. Journal of Obstetrics and Gynaecology of the British Commonwealth 81:257-263. Hirst JJ, Haluska GJ, Cook MJ, Hess DL, Novy MJ (1991) Comparison of plasma oxytocin and catecholamine concentrations with uterine activity in pregnant rhesus monkeys. J Clin Endocrinol Metab 73:804-810. Hirst JJ, Haluska GJ, Cook MJ, Novy MJ (1993) Plasma oxytocin and nocturnal uterine activity: maternal but not fetal concentrations increase progressively during late pregnancy and delivery in rhesus monkeys. American Journal of Obstetrics & Gynecology 169:415-422. Hofmeyr GJ, Alfirevic Z, Kelly T, Kavanagh J, Thomas J, Brocklehurst P, Neilson JP (2005a) Methods for cervical ripening and labour induction in late pregnancy: generic protocol. In: The Cochrane Library, Issue 1. John Wiley & Sons Ltd, Chichester, UK.
154 Hofmeyr GJ, Gulmezoglu AM (2005b) Vaginal misoprostol for cervical ripening and labour induction in late pregnancy (Cochrane Review). In: The Cochrane Library, Issue 1. John Wiley & Sons Ltd, Chichester, UK. Hofmeyr GJ (2003) Induction of labour with an unfavourable cervix. Best Pract Res Clin Obstet Gynaecol 17(5):777-794. Hofmeyr GJ, Alfirevic Z, Matonhodze B, Brocklehurst P, Campbell E., Nikodem VC (2001) Titrated oral misoprostol solution for induction of labour: a multi- centre, randomised trial. British Journal of Obstetrics and Gynaecology 108:952-959. Hofmeyr J (1998) Misoprostol in obstetrics and gynaecology - unregistered, dangerous and essential. South African Medical Journal 88:535-536. Husslein PW, Fuchs AR, Fuchs F (1981) Oxytocin and the initiation of human parturition 1: Prostaglandin release during induction of labor by oxytocin. American Journal of Obstetrics & Gynecology 141:688-693. Junkermann H, Mangold H, Vecsei P, Runnebaum B (1982) Circadian rhythm of serum progesterone levels in human pregnancy and its relation to the rhythm of cortisol. Acta Endocrinol 101:98-104. Kararli TT, Catalano T, Needham TE, Finnegan PM (1991) Mechanism of misoprostol stabilization in hydroxypropyl methylcellulose. Advances in Experimental Medicine and Biology 302:275-289. Karim A (1987) Antiulcer prostaglandin misoprostol: single and multiple dose pharmacokinetic profile. Prostaglandins 33:40-50. Karim SMM (1966) Identification of prostaglandins in human amniotic fluid. Journal of Obstetrics and Gynaecology of the British Commonwealth 73:903-908. Karim SMM, Devlin J (1967) Prostaglandin content of amniotic fluid during pregnancy and labour. Journal of Obstetrics and Gynaecology of the British Commonwealth 74:230-234. Karim SMM (1970a) Use of prostaglandin E2 in the management of missed abortion, missed labour and hydatidiform mole. British Medical Journal 3:196-197. Karim SMM, Filshie GM (1970b) Use of prostaglandin E2 for therapeutic abortion. British Medical Journal 3:198-200. Karim SMM, Sharma SD (1971) Oral administration of prostaglandins for the induction of labour. British Medical Journal 1:260-262.
155 Karim SMM, Trussell RR, Hillier K, Patel RC (1969) Induction of labour with prostaglandin F2alpha. Journal of Obstetrics and Gynaecology of the British Commonwealth 76:769-782. Karim SMM, Trussell RR, Patel RC, Hillier K (1968) Response of pregnant human uterus to prostaglandin F2alpha induction of labour. British Medical Journal 4:621-623. Keirse MJNC, Flint APF, Turnbull AC (1974) F prostaglandins in amniotic fluid during pregnancy and labour. Journal of Obstetrics and Gynaecology of the British Commonwealth 81:131-135. Keirse MJNC, Mitchell MD, Turnbull AC (1977) Changes in prostaglandin F and 13,14-dihydro-15-keto-prostaglandin F concentrations in amniotic fluid at the onset of and during labour. British Journal of Obstetrics and Gynaecology 84:743-746. Keirse MJNC, Turnbull AC (1973) E prostaglandins in amniotic fluid during late pregnancy and labour. Journal of Obstetrics and Gynaecology of the British Commonwealth 80:970-973. Keirse MJNC, Turnbull AC (1975) Metabolism of prostaglandins within the pregnant uterus. British Journal of Obstetrics and Gynaecology 82:887-893. Kelly AJ, Kavanagh J, Thomas J (2005) Vaginal prostaglandin (PGE2 and PGF2 alpha) for induction of labour at term (Cochrane Review). In: The Cochrane Library, Issue 1. John Wiley & Sons Ltd, Chichester, UK. Kelly RW (1994) Pregnancy maintenance and parturition: the role of prostaglandin in manipulating the immune and inflammatory response. Endocrine Reviews 15:684-706. Khabbaz AY, Usta IM, El-Hajj MI, Abu-Musa A, Seoud M, Nassar AH (2001) Rupture of an unscarred uterus with misoprostol induction: case report and review of the literature. Journal of Maternal-Fetal Medicine 10:141-145. Khoury AN, Zhou Q-P, Gorenberg DM, Nies BM, Manley GE, Mechlenburg FE (2001) A comparison of intermittent vaginal administration of two different doses of misoprostol suppositories with continuous dinoprostone for cervical ripening and labor induction. Journal of Maternal-Fetal Medicine 10:186-192. Kotsonis FN, Dodd DC, Regnier B, Kohn FE (1985) Preclinical toxicology profile of misoprostol. Digestive Diseases and Sciences 30:142S-146S.
156 Kramer RL, Gilson GJ, Morrison DS, Martin D, Gonzales JL, Qualls CR (1997) A randomized trial of misoprostol and oxytocin for induction of labor: safety and efficacy. Obstetrics & Gynaecology 89:387-391. Kunz R, Oxman AD (1998) The unpredictability paradox: review of empirical comparisons of randomised and non-randomised clinical trials. British Medical Journal 317:1185-1190. Kurzrok R, Lieb CC (1930) Biochemical studies of human semen: the action of semen on the human uterus. Proceedings of the Society of Experimental Biology and Medicine 26:268-272. Laws PJ, Sullivan EA (2004) Australia's mothers and babies 2002. Australian Institute of Health and Welfare (AIHW), National Perinatal Statistics Unit (NPSU). Sydney. Liggins GC (1968) Premature parturition after infusion of corticotrophin or cortisol into foetal lambs. Journal of Endocrinology 42:323-329. Liggins GC (1981) Cervical ripening as an inflammatory reaction. In: Ellwood DA, Anderson ABM (eds) The cervix in pregnancy and labour, clinical and biochemical investigations., pp 1-9. Churchill Livingstone, Edinburgh. Liggins GC, Fairclough RJ, Grieves SA, Kendall JZ, Knox BS (1973) The mechanism of initiation of parturition in the ewe. Rec. Prog. Horm. Res. 29:111- 159. Liggins GC, Forster CS, Grieves SA, Schwartz AL (1977) Control of parturition in man. Biology of Reproduction 16:39-56. Liggins GC, Kennedy PC, Holm LW (1967) Failure of initiation of parturition after electrocoagulation of the pituitary of the fetal lamb. American Journal of Obstetrics & Gynecology 98:1080-1086. Lindow SW, Jha RR, Thompson JW (2000) 24 hour rhythm to the onset of preterm labour. British Journal of Obstetrics and Gynaecology 107:1145-1148. Lindow SW, Newham A, Hendricks MS, Thompson JW, van der Spuy ZM (1996) The 24-hour rhythm of oxytocin and beta-endorphin secretion in human pregnancy. Clin Endocrinol 45:443-446. Luckas M, Bricker L (2005) Intravenous prostaglandins for induction of labour (Cochrane Review). In: The Cochrane Library, Issue 1. John Wiley & Sons Ltd, Chichester, UK.
157 Luo ZC, Karlberg J (2001) Timing of birth and infant and early neonatal mortality in Sweden 1973-95: longitudinal birth register study. British Medical Journal 323:1327-1330. MacKenzie IZ, Embrey MP (1977) Cervical ripening with intravaginal prostaglandin E2 gel. British Medical Journal ii:1381-1384. MacKenzie IZ, Embrey MP (1978) The influence of pre-induction vaginal prostaglandin E2 gel upon subsequent labour. British Journal of Obstetrics and Gynaecology 85:657-661. MacKenzie IZ, Embrey MP (1981) A simpler approach to labour induction using a lipid based prostaglandin E2 vaginal suppository. American Journal of Obstetrics & Gynecology. Mancuso PJ, Alexander JM, McIntire DD, Davis E, Burke G, Leveno KJ (2004) Timing of birth after spontaneous onset of labour. Obstetrics & Gynaecology 103:653-656. Mariani-Neto C, Leao EJ, Kenj G, De Aquino MM (1987) Use of misoprostol for induction of labour in stillbirths. Rev. Paul. Med. 105:325-328. Matonhodze B, Hofmeyr G, Levin J (2003) Labour induction at term - a randomised trial comparing Foley catheter plus titrated oral misoprostol solution, titrated oral misoprostol solution alone, and dinoprostone. South African Medical Journal 93:375-379. Matonhodze BB, Katsoulis LC, Hofmeyr GJ (2002) Labor induction and meconium: in vitro effects of oxytocin, dinoprostone and misoprostol on rat ileum relative to myometrium. J Perinat Med 30(5):405-410. Meydanli MM, Caliskan E, Burak F, Narin MA, Atmaca R (2003) Labor induction post-term with 25 micrograms vs 50 micrograms of intravaginal misoprostol. International Journal of Gynaecology and Obstetrics 81:249-255. Microsoft Corporation (1997) Microsoft Access 97, 8.0 edition. Mitchell MD, Flint APF, Bibby J, Brunt J, Arnold JM, Anderson ABM, Turnbull AC (1977) Rapid increases in plasma prostaglandin concentrations after vaginal examination and amniotomy. British Medical Journal 2:1183-1185. Mitchell MD, Flint APF, Bibby J, Brunt J, Arnold JM, Turnbull AC (1978) Plasma concentrations of prostaglandins during late human pregnancy: influence of normal and preterm labor. Journal of Clinical Endocrinology and Metabolism 46:947-951.
158 Mitchell MD, Keirse MJNC, Anderson ABM, Turnbull AC (1976) Evidence for a local control of prostaglandins within the pregnant human uterus. British Journal of Obstetrics and Gynaecology 84:35-38. Montealegre JA, Botero LF, Sabogal G (1999) Induccion de trabajo de parto con cervix desfavorable experimento clinico aleatorizado doble ciego de oxiotocina vs microprostol. Revista Colombiana de Obstetricia y Gunecologia 50. Mulrow CD (1994) Rationale for systematic reviews. British Medical Journal 309:597- 599. Nassar N, Sullivan EA, Lancaster P, Day P (2000) Australia's mothers and babies 1998. Australian Institute of Health and Welfare (AIHW) National Perinatal Statistics Unit (NPSU), Sydney. Ngai SW, Chan YM, Lam SW, Lao T (2000) Labour characteristics and uterine activity: misoprostol compared with oxytocin in women at term with prelabour rupture of the membranes. British Journal of Obstetrics and Gynaecology 107:222-227. Ngai SW, To WK, Lao T, Ho PC (1996) Cervical priming with oral misoprostol in pre-labour rupture of membranes at term. Obstetrics & Gynaecology 87:923- 926. Nicholson P, Karim A, Smith M (1990) Pharmacokinetics of misoprostol in the elderly, in patients with renal failure and when coadministered with NSAID or antipyrine, propranolol or diazepam. Journal of Rheumatology 17:33-37. O'Brien WF (1995) The role of prostaglandins in labor and delivery. Clinics in Perinatology 22:973-984. O'Brien WF, Knuppel RA, Cohen GR (1986) Plasma prostaglandin metabolite levels after use of prostaglandin E2 gel for cervical ripening. American Journal of Obstetrics & Gynecology 155:1037-1040. O'Dowd MJ, Philipp EE (1994) The history of obstetrics and gynaecology. The Parthenon Publishing Group, New York. Phillips K, Berry C, Mathers AM (1996) Uterine rupture during second trimester termination of pregnancy using mifepristone and a prostaglandin. European Journal of Obstetrics, Gynecology, and Reproductive Biology 65:175-6. Plaut MM, Schwartz ML, Lubarsky SL (1999) Uterine rupture associated with the use of misoprostol in the gravid patient with a previous cesarean section. American Journal of Obstetrics & Gynecology 180:1535-1542.
159 Ramsey P, Harris D, Ogburn P, Heise R, Magtibay P, Ramin K (1998) Comparative efficacy of prostaglandin analogues dinoprostone and misoprostol as labor preinduction agents. American Journal of Obstetrics & Gynecology 178:S94. Reck G, Renner A, Breckwoldt M (1988) Diurnal variations in unconjugated oestriol levels during early pregnancy and their relation to maternal cortisol levels. Arch Gynecol Obstet 243:169-177. Reddi K, Norman RJ, Deppe WM, Machattie CT, Kemp M, Joubert SM (1987) Amniotic membrane production of prostaglandin F2alpha is reduced in dysfunctional human labor: results of in vivo and in vitro studies. Journal of Clinical Endocrinology and Metabolism 65:1000-1005. Riskin-Mashiah S, Wilkins I (1999) Cervical ripening. Obstetrics & Gynaecology Clinics of North America 26:243-257. Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) (2001) Clinical Guidelines: Intrapartum Fetal Surveillance. RANZCOG, East Melbourne, Victoria. Royal College of Obstetricians and Gynaecologists (RCOG) (2001) Evidence-based clinical guideline number 8: the use of electronic fetal monitoring. RCOG Press, London. Runnebaum B, Rieben W, Bierwirth-von Munsterm, Zander J (1972) Circadian variations in plasma progesterone in the luteal phase of the menstrual cycles and during pregnancy. Acta Endocrinol 69:731-738. Sacks HS, Berrier J, Reitman D, Ancona-Berk VA, Chalmers TC (1987) Meta- analyses of randomized controlled trials. New England Journal of Medicine 316:450-455. Sacks HS, Chalmers TC, Smith H (1982) Randomized versus historical controls for clinical trials. American Journal of Medicine 72:233-240. Sanchez-Ramos L, Chen AH, Kaunitz AM, Gaudier FL, Delke I (1997) Labor induction with intravaginal misoprostol in term premature rupture of membranes: a randomized study. Obstetrics & Gynaecology 89:909-912. Sanchez-Ramos L, Kaunitz AM, Del Valle GO, Delke I, Schroeder PA, Briones DK (1993) Labor induction with the prostaglandin E1 methyl analogue misoprostol versus oxytocin: a randomized trial. Obstetrics & Gynaecology 81:332-336. SAS Institute Incorporated (2003) SAS Version 9.1. SAS Institute, Cary, North Carolina, USA.
160 Schoenhard G, Oppermann J, Kohn FE (1985) Metabolism and pharmacokinetic studies of misoprostol. Digestive Diseases and Sciences 30:126S-128S. Schulz KF, Chalmers I, Hayes RJ, Altman DG (1995) Empirical evidence of bias : dimensions of methodological quality associated with estimates of treatment effects in controlled trials. JAMA 273:408-412. Sciscione AC, Nguyen L, Manley JS, Shlossman PA, Colmorgen GH (1998) Uterine rupture during preinduction cervical ripening with misoprostol in a patient with a previous cesarean delivery. Australian and New Zealand Journal of Obstetrics and Gynaecology 38:96-97. Senior PJ, Marshall K, Sangha R, Clayton JK (1993) In vitro characterization of prostanoid receptors on human myometrium at term pregnancy. British Journal of Pharmacology 108:501-506. Shepherd JH, Bennett MJ, Laurence DT, Moore F, Sims CD (1981) Prostaglandin vaginal suppositories : a simple and safe approach to the induction of labor. Obstetrics & Gynaecology 58:596-600. Shepherd JH, Pearce JMF, Sims CD (1979) Induction of labour using prostaglandin E2 pessaries. British Medical Journal 2:108-110. Shepherd JH, Sims CD, Craft I (1976) Extra-amniotic prostaglandin E2 and the unfavourable cervix. Lancet ii:709-710. Shetty A, Danielian P, Templeton A (2001) A comparison of oral and vaginal misoprostol tablets in induction of labour at term. British Journal of Obstetrics and Gynaecology 108:238-243. Shetty A, Livingstone I, Acharya S, Danielian P, Rice P, Templeton A (2003) A randomised comparison of oral misoprostol and vaginal prostaglandin E2 tablets in labour induction at term. British Journal of Obstetrics and Gynaecology 110:963. Shetty A, Livingstone I, Acharya S, Rice P, Danielian P, Templeton A (2004) A randomised comparison of oral misoprostol and vaginal prostaglandin E2 tablets in labour induction at term. British Journal of Obstetrics and Gynaecology 111:436-440. Singh K, Fong Y, Prasad R, Dong F (1999) Does an acidic medium enhance the efficacy of vaginal misoprostol for pre-abortion cervical priming? Human Reproduction 14:1635-1637.
161 Srisomboon J, Singchai S (1998) A comparison between 25 micrograms and 50 micrograms of intravaginal misoprostol for labor induction. Journal of the Medical Association of Thailand 81:779-782. Stearns J (1808) Account of the pulvis parturiens, a remedy for quickening of childbirth. Medical Repository 11:308-309. Stewart JH, Andrews J, Cartlidge PHT (1998) Numbers of deaths related to intrapartum asphyxia and timing of birth in all Wales perinatal survey. British Medical Journal 316:657-660. Stitely ML, Browning J, Fowler M, Gendron RT, Gherman RB (2000) Outpatient cervical ripening with intravaginal misoprostol. Obstetrics & Gynaecology 96:684-688. Surbek DV, Boesiger H, Pavic N, Holzgreve W (1997) A double blind comparison of the safety and efficacy of intravaginal misoprostol and prostaglandin E2 to induce labor. American Journal of Obstetrics & Gynecology 177:1018-1023. Tessier F, Danserau (1997) A double blind randomised controlled trial comparing oral misoprostol to vaginal prostaglandin E2 for the induction of labour at or near term. American Journal of Obstetrics & Gynecology 176:S111. Thacker SB (1998) Meta-analysis : a quantitative approach to research integration. JAMA 259:1685-1689. Toppozada MK, Anwar MY, Hassan HA, el-Gazaerly WS (1997) Oral or vaginal misoprostol for induction of labor. International Journal of Gynaecology and Obstetrics 56:135-139. Uldberg N, Ekman G, Malmstrom A, Olsson K, Ulmsten U (1983) Ripening of the human uterine cervix related to changes in collagen, glycosaminoglycans, and collagenolytic activity. American Journal of Obstetrics & Gynecology 147:662- 666. Varaklis K, Gumina R, Stubblefield PG (1995) Randomized controlled trial of vaginal misoprostol and intracervical prostaglandin E2 gel for induction of labor at term. Obstetrics & Gynaecology 86:541-544. von Euler US (1936) On the specific vasodilating and plain muscle stimulating substances from accessory genital glands in man and certain animals (prostaglandin and vesiglandin). Journal of Physiology 88:213-234.
162 Walsh SW, Ducsay CA, Novy MJ (1984) Circadian hormonal interactions among the mother, fetus, and amniotic fluid. American Journal of Obstetrics & Gynecology 150:745-753. Webb GM, Raynor BD, Huddleston JF, Randall HW, Choy-Hee L (1997) Induction of labor with an unfavorable cervix: a randomized prospective trial. American Journal of Obstetrics & Gynecology 176:S22. Whalen JB, Clancey CJ, Farley DB, Van Orden DE (1978) Plasma prostaglandins in pregnancy. Obstetrics & Gynaecology 51:52-55. Williams M, Black K, Davis G, Baiano J, O'Brien WF (2001) Dosage variation associated with 25mcg misoprostol administration. American Journal of Obstetrics & Gynecology: Abstract 0265. Williams MC, Tsibris JCM, Davis G, Baiano J, O'Brien WF (2002) Dose variation that is associated with approximated one-quarter tablet doses of misoprostol. American Journal of Obstetrics & Gynecology 187:615-619. Willman EA, Collins PW (1976a) The concentrations of prostaglandin E2 and prostaglandin F2alpha in tissues within the fetoplacental unit after spontaneous or induced labour. British Journal of Obstetrics and Gynaecology 83:786-789. Willman EA, Collins PW (1976b) Distribution of prostaglandins E2 and F2alpha within the foetoplacental unit throughout human pregnancy. Journal of Endocrinology 69:413-419. Wilson T, Liggins GC, Whittaker DJ (1988) Oxytocin stimulates the release of arachidonic acid and prostaglandin F2 alpha from human decidual cells. Prostaglandins 35:771-780. Windrim R, Bennett K, Mundle W, Young DC (1997) Oral administration of misoprostol for labor induction: a randomized controlled trial. Obstetrics & Gynaecology 89:392-397. Wing DA, Ham D, Paul RH (1999a) A comparison of orally administered misoprostol to vaginally administered misoprostol for cervical ripening and labor induction. American Journal of Obstetrics & Gynecology 180:S127. Wing DA, Ham D, Paul RH (1999b) A comparison of orally administered misoprostol with vaginally administered misoprostol for cervical ripening and labor induction. American Journal of Obstetrics & Gynecology 180:1155-1160.
163 Wing DA, Jones MM, Rahall A, Goodwin TM, Paul RH (1995a) A comparison of misoprostol and prostaglandin E2 gel for preinduction cervical ripening and labor induction. American Journal of Obstetrics & Gynecology 172:1804-1810. Wing DA, Lovett K, Paul RH (1998) Disruption of prior uterine incision following misoprostol for labor induction in women with previous cesarean delivery. Obstetrics & Gynaecology 91:828-830. Wing DA, Oritz-Omphroy G, Paul RH (1997) A comparison of intermittent vaginal administration of misoprostol with continuous dinoprostone for cervical ripening and labor induction. American Journal of Obstetrics & Gynecology 177:612- 618. Wing DA, Paul RH (1996) A comparison of differing dosing regimens of vaginally administered misoprostol for preinduction cervical ripening and labor induction. American Journal of Obstetrics & Gynecology 175:158-164. Wing DA, Paul RH (1998) Induction of labor with misoprostol for premature rupture of membranes beyond thirty-six weeks' gestation. American Journal of Obstetrics & Gynecology 179:94-99. Wing DA, Rahall A, Jones MM, Goodwin TM, Paul RH (1995b) Misoprostol : an effective agent for cervical ripening and labor induction. American Journal of Obstetrics & Gynecology 172:1811-1816. Wiqvist N, Bygdeman M (1970) Therapeutic abortion by local administration of prostaglandin. Lancet ii:716-717. Zieman M, Fong SK, Benowitz NL, Banskter D, Darney PD (1997) Absorption kinetics of misoprostol with oral or vaginal administration. Obstetrics & Gynaecology 90:88-92.
164