Preterm birth is a leading cause of neonatal morbidity and mortality. Over the past several decades, there has been a marked increase in survival of very low–birth-weight infants. This increase in survival has been attributed to increased use of corticosteroids, regionalization of perinatal care, improved methods of mechanical ventilation, availability of exogenous surfactant, and improved nutritional therapy. However, the reduction in mortality has not been accompanied by a reduction in neonatal morbidity or long-term handicaps. It is estimated that 50% of all major neurologic handicaps in children result from premature births.

Despite widespread awareness of the problem and use of therapies believed to be beneficial to prevent preterm births, the rate of preterm delivery has increased in the United States. The majority of spontaneous premature births occur to women who develop preterm labor or preterm premature rupture of the membranes (PPROM). Cervical incompetence may also result in preterm delivery. Historically, researchers and epidemiologists have approached these conditions as being distinct processes that were mutually exclusive of one another. Recent evidence would suggest that many women have overlapping conditions that predispose them to deliver preterm. This concept is depicted in Figure 11.1. For example, a woman who has preterm delivery secondary to PPROM at 27 weeks gestation may have had weeks of “silent” or painless contractions or cervical dilation prior to developing ruptured membranes and delivery. Using this broader conceptual framework, this chapter will review the epidemiology, etiology, prevention, and treatment of women with preterm labor.

Labor occurs when the uterus converts from a state of containment to an environment that attempts to expel the fetus. In humans, the average gestational period is 280 days ±14 days. Therefore, term labor is defined as labor that occurs between 37 and 42 weeks gestation. Preterm labor is defined as labor that occurs between 20 and 37 weeks gestation. In theory, pathologic activation of the normal parturition process results in preterm labor and delivery.

In both term and preterm labor, following an unknown stimulus, the mechanisms that produce labor override those that maintain the pregnancy. Activation of the parturition process results in membrane activation, cervical ripening, and an increase in myometrial responsiveness to endogenous and exogenous signals. Subsequently, labor progresses along a common pathway that results in uterine contractions that are sufficient to cause progressive cervical dilation to allow for expulsion of the fetus. A number of inciting events have been implicated in premature births. These events include decidual hemorrhage (abruption), mechanical factors (overdistension of the uterus, cervical incompetence), hormonal changes (fetal or maternal stress), or subclinical/clinical infection. Infection is associated with as many as one third of preterm deliveries, particularly those occurring at the earliest gestational ages. The role of infection in preterm labor will be reviewed separately in this chapter.

Animal models have helped in understanding labor. Important findings in animal labor models include an increase in oxytocin receptors present in the myometrium, gap junctions developing between myometrial cells, an increased response to agents capable of producing contractions in the uterus, and physical and biochemical changes of the cervix resulting in a softened consistency. Uterine smooth muscle contractility is produced by the actin–myosin interaction, following myosin light chain phosphorylation, which is controlled by myosin light chain kinase. Myosin light chain kinase is activated by calcium as a calmodulin–calcium complex. Cyclic adenosine monophosphate (cAMP) also regulates kinase by inhibiting phosphorylation. Many factors are involved in this control. Some of the proposed theories of labor will be discussed in the following sections.

Preterm birth has been linked with symptomatic nongenital infections such as acute pyelonephritis and pneumonia. A large body of evidence suggests that subclinical infection
may be an important cause of premature labor, especially labors resulting in very early delivery.

The hypothesis linking subclinical infection and premature birth may be summarized as follows. Microbes or their products such as endotoxin enter the uterine cavity during pregnancy, most commonly ascending from the lower genital tract. Blood-borne infection from a nongenital focus occurs less commonly. Microbes or their products then interact, most likely with the decidua or possibly with the membranes, producing prostaglandins or directly leading to uterine muscle contraction. This interaction is most likely mediated through a cytokine cascade. As a result, there is cervical dilation, entry of more microbes into the uterus, and continuation of “the vicious cycle” resulting in premature birth.

The first piece of evidence linking subclinical infection to preterm birth is that the prevalence of histologic chorioamnionitis is increased among preterm births. In membranes from preterm deliveries, there is a consistent and very strong association between positive membrane cultures and the likelihood of membrane infiltration. For example, when the birth weight is greater than 3,000 g, the percentage of placentae showing histologic chorioamnionitis is less than 20%; when the birth weight is below 1,500 g, the percentage is 60% to 70%. Most cases of histologic chorioamnionitis are caused by infection.

Second, clinically recognized infections are increased in mothers and neonates after preterm birth. Sepsis and meningitis are increased 3- to 10-fold in preterm infants. Less widely recognized is the increase in maternal infection after preterm birth. These observations suggest that subclinical infection underlies preterm birth and that the infection became clinically evident during or shortly after birth.

Third, there are associations of preterm birth with various maternal lower genital infections or microbes (Table 11.1). Although Ureaplasma urealyticum in the lower genital tract had been associated with LBW infants in earlier studies, a large National Institutes of Health (NIH) study reported no associations of U. urealyticum in the vagina with any adverse pregnancy outcome (preterm birth, PPROM, LBW, or birth weight <1,500 g). Interestingly, then, U. urealyticum in the lower genital tract is not associated with LBW/preterm pregnancies, even though this organism is one of the most common isolates from the amniotic fluid of women in preterm labor. Lower genital infection with
Chlamydia trachomatis has also not been consistently associated with adverse pregnancy outcome. However, women with active chlamydial infection and with a positive serum antichlamydial immunoglobulin M (IgM) have an increased risk of preterm delivery. Although a consistent association has not been observed between maternal group B streptococci (GBS) colonization and premature birth in several small studies, a large investigation of approximately 13,000 women showed that pregnant women with heavy GBS colonization had a small but significant increase in risk for LBW (odds ratio [OR] 1.2; 95% CI 1.01 to 1.50). There were no significant increases in other adverse outcomes, including preterm birth, among heavily colonized women. Women with light colonization were not at an increased risk for any adverse outcomes. There is increasing evidence of an association between Trichomonas vaginalis and premature birth. Although small, earlier studies had conflicting results, the large Vaginal Infections and Prematurity Study found that the presence of T. vaginalis in the vagina at midpregnancy was significantly associated with preterm LBW (7.1% of women with T. vaginalis vs. 4.5% without T. vaginalis; OR 1.6; 95% CI 1.3 to 1.9). The MFMU Network bacterial vaginosis/T. vaginalis trial showed an association between carriage of T. vaginalis and preterm birth. Considerable data have linked lower genital tract anaerobes with preterm labor. Further, bacterial vaginosis, in which there is a predominance of anaerobes, has been consistently associated with approximately a two- to threefold increase in spontaneous preterm delivery. Among other infections, untreated pyelonephritis has been associated with a risk of preterm delivery of approximately 30%, and asymptomatic bacteriuria is associated with a 60% higher rate of LBW (95% CI 1.4 to 1.9) and a 90% higher rate of preterm delivery (95% CI 1.3 to 2.9).

Fourth, positive cultures of the amniotic fluid/membranes/decidua are found in some patients in premature labor. The range of positive amniotic fluid cultures obtained by amniocentesis from asymptomatic women in premature labor is 3% to 24%. When more sensitive testing for detection of bacteria (polymerase chain reaction) is carried out in amniotic fluid from women in preterm labor, bacteria are detected in 30% to 50%. The most likely route of upper genital tract infection in preterm labor is an ascending path through the vagina and cervix. Similarities in organisms isolated from the amniotic fluid and the lower genital tract support this pathogenic route. It is also possible that bacteria may enter the uterine cavity hematogenously through spread via the placenta, by contamination at the time of instrumentation such as during amniocentesis or chorionic villus sampling, or even by spread from the abdominal cavity via the fallopian tubes. Other sources of organisms for hematogenous spread include bacteremia from periodontal disease or procedures. Among women in spontaneous preterm labor with intact membranes, genital mycoplasmas, anaerobic organisms, and Gardnerella vaginalis (the so-called bacterial vaginosis organisms) are the organisms most commonly found in the amniotic fluid. Sexually transmitted organisms such as Neisseria gonorrhoeae and C. trachomatis are rarely found in the amniotic fluid, and GBS and Escherichia coli are found occasionally. Patients in preterm labor at early gestational ages have the highest likelihood of having a positive culture of the amniotic fluid. It may be speculated that intrauterine infection occurs early in pregnancy (or even has preceded the pregnancy) and may remain without clinical detection for months.

Fifth, biochemical “markers” of infection are often present among women in premature labor. In infection-induced premature labor, the primary site of infection is probably not the amniotic fluid but the decidua or membranes. More sensitive markers of infection potentially include amniotic fluid glucose concentrations, serum white blood cell counts, C-reactive protein, and amniotic or serum cytokines. Unfortunately, relatively few are clinically useful. Among patients in preterm labor, a low amniotic fluid glucose (<14 mg/dL) correlates well with the likelihood of a positive culture. Among the cytokines, an elevated amniotic fluid IL-6 level is probably the most sensitive marker for infection but is not yet widely available for clinical use.

Sixth, bacteria or their products induce preterm birth in animal models. Animal models have provided direct evidence that infection triggers preterm birth in the rabbit, monkey, and mouse.

The evidence linking infection to preterm birth has led to many trials of antibiotic therapy to prevent preterm birth. Antibiotic treatment trials may be classified as one of four designs:

Chapter 12 discusses antibiotics in PPROM. Table 11.2 summarizes current practices for use of antibiotics to prevent preterm birth. The discordant results in antibiotic trials raise the question as to why antibiotics have not consistently prevented preterm birth or neonatal morbidity associated with preterm birth. One explanation is that infection is simply not a significant cause of preterm labor, but this seems unlikely in view of all the other evidence. Another explanation is that studies have had too low a power. However, large meta-analyses, the MFMU Network trials, and the ORACLE trial appear to exclude this possibility. Further, because preterm labor has multiple causes, a true effect of antibiotics may be diluted by those cases of preterm labor not caused by infection. It may also be that only a subset of pregnant women (e.g., perhaps genetically predisposed women) with high cytokine response are at risk for preterm labor after subclinical infection. Another explanation is that the antibiotics studied in most
of the trials were simply the wrong ones (e.g., not including antibiotics with better anaerobic activity), the antibiotics were given too late, or the antibiotic dose or timing were incorrect. Because infection is more likely to cause very early preterm birth (<32 weeks), trials focusing on women at later gestational ages may not show an effect. For example, the ORACLE I trial enrolled women up to 37 weeks gestation, and only 10% delivered at less than 32 weeks. It has also been speculated that bacterial lysis as a result of antibiotic therapy may lead to increased exposure to lipopolysaccharide and thus enhance preterm labor. Finally, it is possible that changes in the vaginal flora during pregnancy are responsible for preterm labor. Early screening and treatment may not identify women who are at risk. Antibiotic therapy may actually increase the risk of preterm birth by changing the vaginal flora. The MFMU bacterial vaginosis/T. vaginalis trial found that women with T. vaginalis who were treated with metronidazole were more likely to deliver preterm than those treated with placebo. The PREMET trial also found that women with a positive fetal fibronectin were more likely to deliver preterm if they were treated with metronidazole than placebo. It is possible that metronidazole therapy changes the vaginal flora in women who do not have bacterial vaginosis in such a way as to increase the risk of preterm birth.

In 2004, 12.5% of women in the United States delivered preterm. The vast majority of preterm deliveries are a result of preterm labor (50%), premature rupture of the membranes (PROM) (33%), or cervical incompetence. The contributions of preterm labor and PROM to preterm deliveries vary depending on a number of factors, including socioeconomic status (Fig. 11.2). In a large study from North Carolina, Meis and colleagues found that PPROM (34%) was the most common reason for delivery of less than 2,500 g infants in women who were receiving public assistance. In contrast, in women who had private insurance, the most common reason for early delivery was preterm labor (52%). Indicated preterm deliveries accounted for 14% and 18% of preterm deliveries, respectively.

Several major and minor risk factors are associated with development of preterm labor and PROM (Tables 11.3, 11.4). One of the most obvious and important risk factors for prematurity is a prior history of preterm delivery. To better quantify this relationship, Mercer and associates performed a subgroup analysis of data collected during a large population-based observational study evaluating risk factors for preterm delivery. In this study, gravid women with any prior spontaneous preterm birth had a 2.5-fold increased risk of spontaneous preterm delivery in the current pregnancy. This risk increased to 10.6-fold if the spontaneous preterm birth occurred prior to 28 weeks gestation. Interestingly, women with a history of loss between 13 and 22 weeks gestation had rates of prematurity that were similar to women who did not have this history (10.1% vs. 8.8%; P =.69).

Another major risk factor for preterm labor and birth is multiple gestation. The rate of multiple gestations has increased dramatically over the past 15 years. The increase in twins and higher-order multiples is largely a reflection of
increased use of ovulation induction and assisted reproductive technologies. Fifty percent of twins deliver prematurely, with a mean gestational age at delivery of approximately 35 weeks. As expected, the percent of preterm deliveries increases in proportion to the number of fetuses. Triplets and quadruplets deliver on average at 32 weeks and 30 weeks, respectively. Until researchers develop techniques to perform artificial reproductive technologies that minimize the risk of having high-order multiples, then these women will continue to be at significant risk for delivering prematurely and suffering the consequences of preterm birth.

Blacks are 1.6 to 2.5 times more likely to deliver prematurely than white women of similar age and socioeconomic status. Although blacks have higher rates of prematurity, the rates of neonatal morbidity are lower in black neonates when compared with whites born at similar gestational ages. This suggests that the gestational period may be shorter in black women. Low socioeconomic status is also strongly associated with prematurity. It is not clear whether this is related to environment, genetic predisposition, infection, or access to medical care.

Table 11.4 also lists a number of “minor” risk factors for preterm labor and delivery. Several of these will be discussed in more detail later in this chapter. In general, the minor risk factors can be broken into two categories: those that are potentially modifiable and those that are not. Many of the minor risk factors are common in pregnancy. Individually, their contribution to prematurity is small; however, the risk is compounded by the addition of other risk factors. The impact of work on preterm birth remains controversial. Prolonged, physically demanding work does appear to independently increase the risk of prematurity and is potentially modifiable.

Over the past 2 decades, many researchers have focused on identification of women who are at risk for preterm delivery. Theoretically, identification of asymptomatic women at risk for preterm delivery would allow obstetricians to effectively intervene to prevent preterm delivery or to decrease neonatal morbidity and mortality in preterm neonates. In an American College of Obstetricians and Gynecologists (ACOG) Practice Bulletin titled Assessment of Risk Factors for Preterm Birth, the authors wrote “the ability to predict whether a women is at risk of preterm delivery has value only if an intervention is available that is likely to improve the outcome.” It is believed that identification of women at risk for preterm birth will be beneficial if it allows women to:

Another potential benefit of screening is to identify women at low risk for preterm delivery and thereby avoid administration of potentially dangerous medications or therapies to these women.

Table 11.5 is a review of “ideal” criteria for a screening test. An ideal screening test should have high sensitivity and positive predictive value as well as high specificity and high negative predictive value. Most screening tests do not meet these requirements and trade off sensitivity for specificity. Depending on the clinical scenario where screening tests are used and the consequences of treatment or no treatment, one must decide which test characteristic to stress. For example, one could argue that given the high morbidity associated with preterm birth, the ideal screening test should have a high sensitivity to allow treatment of the majority of women “at risk” and accept a lower specificity rate. On the other hand, one could argue, as the ACOG did, that avoidance of treatment with potentially hazardous drugs is beneficial in women with symptoms who are at “low risk” for delivery, thus stressing the importance of the test’s specificity and negative predictive value. The following will review available screening tests.