Introduction

Although preterm deliveries constitute a small proportion of all births, their contribution to serious complications, especially those leading to perinatal death and morbidity, is hugely disproportionate. In 2010 it is estimated that 14.9 million babies worldwide (around 11.1% of all births) were premature. Globally, preterm birth is the single biggest cause of neonatal death. Babies born at ‘term’ (conventionally considered to be 37−42 weeks of gestation) have consistently better outcomes than those born ‘preterm’, with the risk of neonatal mortality and morbidity rising exponentially as the gestation of delivery decreases. Preterm labour is the single biggest cause of preterm birth, so that effective ‘treatment’ of preterm labour could have a major impact on global perinatal health. Such treatments include those aimed at preventing or halting preterm labour and also those that improve outcomes for babies of women in preterm labour. After decades in which there were few effective therapies, some promising strategies are emerging, which improve outcomes in a subset of women and babies. Despite this, the global toll of the adverse effects of preterm birth continues to rise, with preterm labour remaining the single biggest cause of neonatal mortality and morbidity in resource rich countries.

Definition

The definition of preterm birth is not without controversy. The ICD10 (International Statistical Classification of Diseases and Related Health Problems 10th Revision) definition of preterm labour is the onset (spontaneous) of labour before 37 weeks of gestation ( http://apps.who.int.easyaccess1.lib.cuhk.edu.hk/classifications/icd10/browse/2010/en#/O60 ), thus preterm birth under this definition is considered to be birth before 37 completed weeks of gestation. The lower gestational limit is not defined under this system, although the WHO recommends that all babies born with any signs of life should be considered live births (and hence would be included). The lack of a consensus about the lower limit of preterm birth causes problems in comparing data among countries, with many countries (including Scotland, the USA and Brazil) not defining their lower gestational limit, some (such as Switzerland and Denmark) using a lower limit of 22 weeks and others (including Australia and Canada) using 20 weeks as the lower gestational limit of preterm birth. Thus a woman who delivers a baby at 21 weeks with no signs of life would be likely to be considered to have had a miscarriage in Switzerland and Denmark, and probably in the majority of countries with no defined lower limit, but would be considered to have had a stillbirth in Australia and Canada. The birth would be defined as a preterm birth in the latter two countries but not the former two. Comparisons are further complicated by the use in some countries of low birth weight as a surrogate for preterm birth: this is inappropriate because not all small babies are preterm, and not all preterm babies are small. Lastly, due to the phenomenon of delayed ovulation, where ultrasound is used to estimate gestational age (as is common in many resource rich countries), the calculated mean duration of pregnancy is consistently shorter, and the rate of prematurity is around 20% higher, than when gestation is calculated from the date of the last menstrual period.

A recent report by the Global Alliance to Prevent Prematurity and Stillbirth has highlighted that similar aetiologies (albeit in different proportions) are involved in a pregnancy loss in the second trimester and in the mid third trimester, and that the risk of adverse outcome for the neonate decreases progressively as gestation advances, even beyond 37 weeks’ gestation. They propose a new definition and classification system whereby preterm birth would be ‘any birth (which includes stillbirths and pregnancy terminations) that occurs after 16 weeks’ gestation and before term (i.e., 39 weeks’ gestation). The complete population of preterm deliveries within the gestational range as described earlier includes live births, stillbirths, multiple pregnancies, pregnancy terminations, and newborn infants with congenital malformations’. The recommendation from this group is that ‘gestational age estimation should, whenever possible, be corroborated by an early, high quality ultrasound and the best obstetric estimate be used for all gestational age determinations’.

Preterm Labour Versus Preterm Birth

The focus of this chapter is preterm labour, although this is not the only pathway to preterm birth. A categorization of (spontaneous) preterm labour, preterm premature rupture of membranes and elective (induced) preterm birth has been widely used, with Scottish data suggesting that the proportions of each (amongst all singletons delivering preterm) are 62%, 15% and 23% respectively ( Fig 9-1 ). Villar proposes that preterm birth is defined by pathway to delivery (spontaneous or care giver initiated) AND signs of initiation of parturition (evidence of initiation of parturition (including preterm premature rupture of membranes) or no evidence of initiation of parturition) AND the presence of significant fetal, maternal or placental pathological conditions. Under this classification, both preterm labour and preterm premature rupture of membranes would be considered to have evidence of initiation of parturition, whereas elective (induced) preterm birth would not. The pathway to delivery would be spontaneous in women presenting in preterm labour and those with preterm premature membrane rupture (because oxytocin augmentation of contractions is also considered in the spontaneous category) but would be care giver initiated in women undergoing elective (induced) preterm birth.

Singleton preterm births in Scotland, expressed as a percentage of all singleton births, live and still, 1978 to 2010.

Incidence of Preterm Birth

The incidence of preterm birth continues to rise globally, although both Scotland and the USA have seen a modest downturn in rates over the last few years. In the USA, this has been attributed to one or more of the widespread use of progesterone prophylaxis for prevention of preterm birth, a reduction in elective late preterm births and a change in maternal risk factor profiles. In Scotland, the reduction in preterm births has been attributed to the ban on smoking in public places which came into effect in 2006.

Aetiology and Mechanisms

The ‘cause’ of preterm labour is incompletely understood. Preterm labour is often accom­panied by one or more of the following pathologies: intrauterine infection, intrauterine inflammation, utero-placental ischaemia, utero-placental haemorrhage, uterine stretch and maternal stress. It is not possible to determine whether these events ‘cause’ preterm labour, although there is strong circumstantial evidence of the role of intrauterine infection and inflammation. This is firstly because, even using relatively insensitive culture techniques, around 25–40% of women in preterm labour have demonstrable intrauterine infection. The proportion rises progressively as gestational age of labour onset declines. Secondly, intrauterine infection/inflammation stimulates an inflammatory response, including production of pros­taglandins, implicated in increasing cervical ripening and myometrial contractility. Lastly, in animal models, intrauterine injection of micro-organisms or pro-inflammatory agents (such as lipopolysaccharide) is effective in stimulating preterm labour.

Risk Factors

The risk factors associated with preterm labour are listed in Table 9-1 .

Outcomes

There is a clear dose−response relationship between gestation of preterm birth and risk of perinatal death, with outcomes being worst in babies born at earlier gestational ages, and the nadir not being reached until 40 weeks’ gestation. For example, UK data show the combined rate of early neonatal, late neonatal and postneonatal death is 580 per 1000 babies born at 24 weeks, 98 per 1000 babies born at 28 weeks, 12 per 1000 for babies born at 34 weeks and 1 per 1000 for babies born at 40 weeks’ gestation. Babies who survive preterm birth also have a greater incidence of morbidity, and again this is inversely proportional to gestational age at delivery. For example, in the EPICure study (a prospective cohort of around 300 children who were born before 25 weeks’ gestation in 1995, and who survived to reach the neonatal unit), 49% had neuromotor or sensory (sight or hearing) disability (with 23% of the total having severe disability) when assessed at 30 months of age, and the remainder had no disability according to the study criteria. Subsequent studies have confirmed a ‘dose dependent’ effect of prematurity on long-term adverse health outcomes, which is inversely proportional to gestational age at delivery. As with death, the nadir of ‘prematurity’ on educational attainment at school is not reached until birth at 40 weeks of gestation, suggesting that even those who apparently have ‘no disability’ suffer long-term adverse consequences of prematurity.

Prediction

Although clinical risk factors for preterm labour have been identified, and predictive tests proposed, no strategy is sufficiently effective to be in widespread use in clinical practice. One of the most widely used clinical indicators, history of spontaneous preterm birth in a previous pregnancy, is associated with likelihood ratios of 4.62 (95% CI 3.28–6.52) and 2.26 (95% CI 1.86–2.74) respectively for birth before 34 and 37 weeks’ gestation in a subsequent pregnancy. The most widely used and the most effective tests for preterm labour prediction in asymptomatic women are detection of fetal fibronectin (fFN) in vaginal fluid and cervical length measurement. The predictive ability of these strategies varies with the gestation of testing, the definition of a positive test (e.g. the length of the cervix or the quantitation of the fetal fibronectin) and the gestation of delivery being predicted. Typical summary likelihood ratios from meta-analyses of a range of studies are shown in Table 9-2 . There is some evidence that fFN testing reduces the risk of preterm birth, with odds ratios of 0.54 (95% CI 0.34–0.87), although no benefit in terms of reduction in adverse outcomes was seen. Newer tests are continually being proposed and evaluated – each of cervico­vaginal fluid prolactin and proteome profile and matrix metalloproteinase-8 in amniotic fluid shows promise, but they require further studies to define their efficacy.

Diagnosis

Preterm labour is a diagnosis that can confidently be made only in established labour. Many women present with symptoms suggestive of preterm labour (e.g. uterine contractions) but are found to have a closed cervix on examination. A proportion of such women will labour and deliver within a short space of time, but it is often very difficult for both women and their care givers to identify those who are, and those who are not, in the early stages of preterm labour. Again, cervicovaginal fluid fibronectin and cervical length measurements are amongst the best tests. In this scenario, the negative likelihood ratio (i.e. the effect of a negative test on the confidence with which preterm labour can be excluded as a diagnosis) is often the most helpful. For birth within 7−10 days of testing, fetal fibronectin has a negative likelihood ratio (i.e. a negative test reduces the risk of preterm birth) of 0.36 (95% CI 0.28–0.47), and cervical length measurement of < 15mm has a negative likelihood ratio of 0.026 (95% CI 0.0038–0.182) in singleton pregnancies.

Management

Treatment with the object of reducing the incidence, risks and complications of preterm labour falls into three principal categories:

• •
  

  Measures aimed at preventing preterm labour including early recognition and treatment of infection, cervical cerclage, progesterone prophylaxis and risk modification such as cessation of smoking and drug abuse.

  
  
• •
  

  Tocolysis to try to abolish or arrest preterm labour.

  
  
• •
  

  Obstetric interventions aimed at minimizing the complications of preterm delivery.