Clinical review

Preterm labour and premature rupture of membranes

Philip Steer,  Caroline Flint. 

The length of human pregnancy is variable, reflecting the advantages to the fetus, which would benefit by staying in the uterus to grow more, and to the mother, for whom earlier delivery might reduce pelvic damage (see the second article in the series). The result of this interaction is a relatively high incidence of premature deliveries. In the United Kingdom the incidence of preterm delivery (before 37 weeks) is about 7%, and in many developing countries, it is much higher. The baby is delivered before its homoeostatic mechanisms are properly developed and so is prone principally to the respiratory distress syndrome, hypothermia, hypoglycaemia, and jaundice.

View larger version (30K): [in this window] [in a new window]   Distribution of 24 675 deliveries by gestational age (determined by early ultrasound scanning dates) for a British maternity hospital (Queen's Medical Centre, Nottingham), 1988-95

	Causes

Top Causes Diagnosis of preterm labour Management

Socioeconomic factors influence the incidence of preterm labour. Preterm birth is significantly more common in young women, those with low body weight (body mass index <19), those of lower social class, unmarried or unsupported mothers, and smokers. Some medical factors may increase the risk of preterm birthfor example, previous preterm delivery, persistent vaginal bleeding in early pregnancy, and heart disease. Cervical incompetence is a rare cause of preterm labour, sometimes preventable by cervical cerclage (a purse string suture around the cervix as close to the internal os as possible). The Medical Research Council's trial showed a small but significant benefit from this procedure, even in women with equivocal risk factors.

Infection, usually chorioamnionitis, is a significant component in many cases of spontaneous preterm labour. The ascent of any organism through the cervical mucus plug into the uterus stimulates an inflammatory reaction in the placenta, fetal membranes, and maternal decidua. This leads to the release of cytokines such as interleukin 1 and interleukin 6 from endothelial cells and tumour necrosis  from macrophages. These stimulate the cascade of prostaglandin production, which in turn produces cervical ripening and uterine contractions. The commonest groups of organisms are the streptococci, mycoplasmas, and fusiform bacilli. Bacterial vaginosis (Gardnerella vaginalis) associated with a vaginal pH value of 5.4 seems to promote preterm labour, possibly reducing the efficiency of the cervical barrier to infection. Preliminary studies suggest that treatment with metronidazole or clindamycin in women with bacterial vaginosis may reduce the incidence of preterm labour, and prospective studies are under way.

About 30% of preterm births are iatrogenicthat is, induced by obstetricians for maternal indications such as fulminating pre-eclampsia or for fetal indications such as severe intrauterine growth restriction. In recent years, since the increased use of assisted reproduction, multiple pregnancy has become a growing cause of preterm labour. The incidence of fetal abnormality is higher in pregnancies complicated by preterm labour.

	Diagnosis of preterm labour

Top Causes Diagnosis of preterm labour Management

The precise diagnosis of preterm labour is not easy. The only absolute proof is progressive dilatation of the cervix, but once this has happened, it is too late to attempt preventive treatment. The finding of fetal fibronectin, a fetal protein involved in cell to cell adhesion, in vaginal secretions suggests that the cervical mucus plug is becoming ineffective. This has been proposed as a sensitive screen for preterm labour, but its poor specificity, together with a relatively high false positive rate, makes it unsuitable for routine use.

The diagnosis of labour often has to be made on the basis of reported uterine contractions. However, Braxton Hicks contractions are noticed in most pregnancies from about 24 weeks' gestation onwards, and many women find these painful. This means that the diagnosis of preterm labour is often wrong; two thirds of women diagnosed as being in labour will not have delivered within 48 hours, and over one third go to term. Diagnosis with home uterine contraction monitoring has been tested in a number of trials, but no consensus on its value has emerged.

Diagnosis remains essentially clinical, with a careful history and a speculum examination being important components. Abdominal pain of any type, or any vaginal bleeding, requires a careful speculum examination of the cervix. Digital examinations should be avoided if there is any suggestion of ruptured membranes as they increase the risk of ascending infection. Seeing amniotic fluid trickling through the cervix remains the only certain way of diagnosing ruptured membranes. The use of an acidity indicator, such as nitrazine sticks (Amnicator, Corsham), is not reliable, as this indicates only that the vagina is no longer acid, an effect that can be produced by urine or bath water.

When examining the cervix, a vaginal swab should be taken for culture. This will enable appropriate antibiotic treatment if signs of infection develop later.

	Management

Top Causes Diagnosis of preterm labour Management

Tocolysis
Suppression of uterine contractions would seem to be the obvious solution to the problem of preterm labour. However, tocolytic agents do not work effectively for longer than about 48 hours, probably because of tachyphylaxis. Their major use is to postpone deliveryfor example, for in utero transfer to a tertiary centreor to allow the administration of corticosteroids to the mother and so to her fetus to promote surfactant release in the fetal lung and reduce the incidence of the neonatal respiratory distress syndrome by up to 50%. This effect is only significant at gestations up to 34 weeks; after this it is usual to allow preterm labour to progress.

View larger version (22K): [in this window] [in a new window]   Results of meta-analysis of 16 trials of betamimetics in preterm labour. A high proportion of women remained undelivered without any treatment (therefore original diagnosis of preterm labour was erroneous)

View larger version (23K): [in this window] [in a new window]   Possible use of tocolysis in 1445 preterm labours

View larger version (13K): [in this window] [in a new window]   Results of meta-analysis of prophylactic action of betamimetics in pregnancyeffect on selected outcomes in 11 trials

View larger version (23K): [in this window] [in a new window]   Results of meta-analysis of effect of corticosteroids before delivery in selected outcomes in 15 trials

Antenatal steroids
Meta-analyses of the use of tocolytics suggest that they have little effect on perinatal mortality. However, they do allow time for the administration of antenatal steroids, which, if given at least 24-48 hours before birth, can halve the incidence and severity of respiratory distress and mortality in newborn infants. The effect of steroids lasts up to about a week. The benefit of repeated doses of steroids to the fetus has not been shown; their safety and efficacy needs to be tested in a prospective trial.

Preterm delivery
The preferred method of delivery in preterm labour depends on the fetal presentation and the stage of gestation. If the baby is presenting by the head, then vaginal delivery is probably safe in most cases, with caesarean delivery being performed only for the usual obstetric indications. In a breech presentation before 32 weeks it is possibly safer for the baby to be delivered by cesarean section. This, however, considerably raises the risk to the mother as in many cases the operation is not straightforward because the lower segment may not be well formed.

View larger version (16K): [in this window] [in a new window]   Effect of elective v selective caesarean delivery for preterm breech births

The preterm baby
The conditions into which the baby is born have a major influence on its chance of survival. Very preterm babies (those born at less than 28 weeks) do best when delivered in a tertiary referral centre with a neonatal intensive care unit. Deliveries should be conducted by experienced midwives or obstetricians, with an experienced paediatrician present. The delivery room should be warm, and there should be adequate equipment for resuscitation (see final article in this series). If preterm labour starts at home, or in a smaller hospital, transfer of the mother in early labour should be considered, for the mother is the ideal incubator. However, in some cases the mother is also unwellfor example, with high blood pressureand her condition must be taken into account. In one study 17% of such mothers ended their transfer in the adult intensive care unit. Care must be taken to stabilise maternal condition before transfer. Transfer is probably unwise if labour is progressing rapidly or there is substantial vaginal bleeding.

View larger version (104K): [in this window] [in a new window]   Neonatal intensive care unit

View larger version (24K): [in this window] [in a new window]   Infant survival to 1 year in Wales, by birth weight and gestationcumulative data 1993-5

	Acknowledgments

The box listing potential neonatal hazards of antenatal steroids is based on an article by Quinlivan et al (Aust NZ J Obstet Gynaecol 1998;38:1-7[Medline]). The table on fibronectin testing is adapted from Chien P et al (Br J Obstet Gynaecol 1997;104:436-44[Medline]). The graph of distribution of deliveries by gestational age is adapted from Gardosi et al (Br J Obstet Gynaecol 1997;104:792-7[Medline]). The histogram from the cerclage trial is adapted from the MRC/RCOG Working Party (Br J Obstet Gynaecol 1993;100:516-23[Medline]). The chart showing use of tocolysis is adapted from Tucker et al (Obstet Gynecol 1991;77:343-7[Medline]). The graph showing the meta-analysis of prophylactic betamimetics in pregnancy is adapted from Keirse (Cochrane Pregnancy and Childbirth Database; Oxford: Update Software, 1995). The graph showing the meta-analysis of effect of corticosteroids before delivery is adapted from Crowley (Cochrane Pregnancy and Childbirth Database; Oxford: Update Software, 1995). The graph of caesarean delivery is adapted from Grant et al (Br J Obstet Gynaecol 1996;103:1197-200[Medline]). The bar chart showing survival rates is adapted from Emsley et al (Arch Dis Childhood 1998;78:F99-104[Medline]). The photograph of the neonatal unit was taken by Douglas Neil, medical photographer at the Singleton Hospital.

	Footnotes

Philip Steer is professor of obstetrics and consultant obstetrician at the Imperial College School of Medicine, Chelsea and Westminster Hospital, London; Caroline Flint is honorary professor at Thames Valley University and an independent midwife in London.

The ABC of Labour Care is edited by Geoffrey Chamberlain, emeritus professor of obstetrics and gynaecology at the Singleton Hospital, Swansea. It will be published as a book in the summer.