A significant proportion of spontaneous preterm deliveries result from preterm labor and premature rupture of the membranes (PROM). Approximately 70% of preterm births may be associated with clinical evidence of ruptured membranes or underlying maternal, obstetric, or fetal conditions (Table 14-1) (10,11,12,13).

Predisposing factors for preterm labor remain largely unknown. Epidemiologic studies consistently report an association with nonwhite race (African American relative risk [RR] = 3.3), lower socioeconomic status (RR = 1.83-2.65), low prepregnancy weight (odds ratio [OR] = 2.72), and maternal age younger than 17 years or older than 40 years (RR = 1.47-1.95) (14). Lack of prenatal care is strongly associated with higher rates of preterm delivery, low birth weight, and maternal and infant mortality (15). Other particularly important risk factors include maternal history of preterm birth, especially during the second trimester, with or without rupture of membranes, and vaginal bleeding in at least two trimesters (14,16,17,18). Prostaglandins, thought to be
important mediators in the onset of labor at term, may have a role in preterm labor; in the absence of intrauterine infection, however, the data are not convincing (19).

Although PROM occurs in only 3% of pregnancies before term (20), it is associated with 20% to 50% of preterm deliveries (21) and represents the foremost predisposing factor for admissions to neonatal intensive care units (22). The pathophysiology underlying PROM appears to be multifactorial; evidence suggests that inflammatory weakening of the membranes or choriodecidual infection may be involved, especially at early gestational ages (20,23). Some clinical characteristics seen with PROM are also common to cases with preterm labor (e.g., lower socioeconomic status, young maternal age, sexually transmitted diseases, cigarette smoking, and vaginal bleeding) (20). Many studies suggest an association between bacterial vaginosis, a relatively common condition in which the normally predominant lactobacilli are overgrown by anaerobes and other peroxide-producing bacteria, and complications of premature birth, preterm labor, and PROM (14,24,25,26); however, conflicting data exist (27).

In most patients, spontaneous labor begins after the membranes rupture. The latency period, defined as the time between membrane rupture and onset of contractions, is shorter in patients close to term. More than half of preterm patients will be in labor within 24 hours of ruptured membranes, and 50% to 60% will deliver within 1 week despite conservative management (7,20,28,29). PROM remote from viability is associated with significant rates of maternal and fetal morbidity and low rates of neonatal survival (7,8,19,29). Neonatal respiratory distress syndrome (RDS) is the most common complication at all gestational ages (7,8). Fetal demise resulting from intrauterine infection, umbilical cord compression, or placental abruption occurs in approximately 1% of cases of PROM at 30 to 36 weeks of gestation and up to 21.7% of cases of PROM in the mid-trimester (7,8). Prolonged oligohydramnios can also be associated with fetal pulmonary hypoplasia, limb-positioning deformities, and abnormal facies (e.g., low-set ears and epicanthal folds) (19,30).

Pharmacologic tocolysis, employed in an attempt to prevent progression of preterm labor, is implemented most often before 34 weeks of gestation. The various agents used in pharmacologic tocolysis, which are administered orally, subcutaneously, or parenterally, can have potentially serious maternal and/or fetal side effects (Table 14-2) (31). Despite the widespread use and evaluation of tocolysis over the past 20 years, data fail to demonstrate benefits of tocolysis in terms of medical costs, improvements in neonatal survival, or improvements in other long-term outcome parameters (31). The major benefit of tocolysis is its ability to provide short-term pregnancy prolongation in order that corticosteroids can be administered (31).

Corticosteroids, which are given between 24 and 34 weeks of gestation to patients who are at risk for preterm delivery, are used to induce fetal pulmonary maturity. The re- commended protocol includes a single course of betametha- sone (12 mg intramuscularly [i.m.]) given in 2 doses 24 hours apart) or dexamethasone (6 mg i.m. given in 4 doses every 12 hours). Optimal benefit is seen within 24 hours and through 7 days after administration (32,33). A recent meta-analysis evaluating multiple randomized studies confirmed the decreased incidence and severity of neonatal RDS with antenatal administration of corticosteroids (34). Additional neonatal benefits include reductions in rates of intraventricular hemorrhage and death, even with treatment of less than 24 hours’ duration and in gestations earlier than 30 to 32 weeks with ruptured membranes (33). Long-term data of infants exposed in utero to a single course of corticosteroids failed to demonstrate any significant adverse effects on physical development, growth, cognitive or motor skills, or early school performance (31).

The use of antibiotic therapy, particularly aimed at treating bacterial vaginosis, group B streptococcus, and other genital tract organisms, has been studied widely in recent years. Initially shown to decrease rates of preterm delivery, preterm labor, and neonatal sepsis in some studies, the data show mixed results (31,35,36,37). Routine antibiotic administration is not recommended if used only in an attempt at prevention of preterm delivery (31). While conservatively managing preterm PROM remote from term, however, aggressive antibiotic therapy with erythromycin and ampicillin or amoxicillin prolongs pregnancy and
decreases major neonatal morbidity (e.g., RDS, early sepsis, severe intraventricular hemorrhage, severe necrotizing enterocolitis) (20,38,39).

Because a woman’s past obstetric history may be important for subsequent pregnancies, several risk-scoring systems have been developed. The most widely applied system was devised by Papiernik and modified by Creasy (40). Application of these systems to various populations in the United States, however, has been disappointing (41), and applicability to nulliparae has not been accepted. A 1996 study of almost 3,000 gravidas that attempted to develop a risk-assessment system for predicting spontaneous preterm delivery by using clinical information at 23 to 24 weeks of gestation was similarly disappointing (18). A recent study demonstrated the utility of the preterm labor index, a clinical tool to assess the likelihood of overall preterm delivery and delivery within 1 week (42). The preterm labor index, originally proposed in 1973, combines four clinical parameters (uterine contractions, PROM, vaginal bleeding, and cervical dilation) and is comparable to newer biochemical markers for predicting preterm delivery (42,43). Its utility across populations remains to be studied.

Other screening tools have been proposed to help identify the pregnancy at risk for preterm delivery, including home uterine activity monitoring, screening for fetal fibronectin in vaginal secretions, cervical examination, and screening for genital tract colonization and/or infection.

Hypertension complicates 5% to 8% of pregnanciesand constitutes a major cause of maternal and perinatal
morbidity and mortality (80,81,82). Potential fetal and neonatal complications associated with chronic hypertension include prematurity, increased overall perinatal morbidity, IUGR, and fetal death (80,81,82). Several studies document that the risk of perinatal mortality is increased two to four times if the mother is hypertensive compared with the general obstetric population (83). Maternal complications include superimposed preeclampsia, placental abruption, cesarean delivery, and potentially life-threatening complications such as pulmonary edema, hypertensive en-cephalopathy, retinopathy, cerebral hemorrhage, and acute renal failure (80,81,82,83,84,85). Low-risk patients have mild essential hypertension and no organ involvement; those with severe hypertension or superimposed preeclampsia are considered to be at high risk. The risks of significant complications are especially increased in patients with uncontrolled severe hypertension and underlying renal or cardiac disease prior to or early in gestation (82).

Classification and terminology of the different subdivisions of the hypertensive disorders in pregnancy is confusing. A recent recommendation by the National High Blood Pressure Education Program Working Group replaces the term “pregnancy-induced hypertension” with “gestational hypertension” to describe situations in which elevated blood pressure without proteinuria develops after 20 weeks of gestation and blood pressure levels return to normal postpartum (80,85). Up to 25% of women with gestational hypertension will develop proteinuria or preeclampsia; with severe chronic hypertension, this rate can approach 50% (84). Proteinuria is classified by at least 0.3 g of protein in a 24-hour urine specimen, which usually corresponds to 1+ or greater on a urine dipstick evalua-tion (83).

Preeclampsia is defined as hypertension with proteinuria in addition to other possible symptoms of headache, edema, visual disturbances, and epigastric pain; laboratory abnormalities may involve hemolysis, elevated liver enzymes, and low platelet count (known by the acronym HELLP). HELLP syndrome can occur in up to 20% of women with severe preeclampsia and may have a variety of clinical presentations (86). In past years, the diagnosis of preeclampsia included specific blood pressure elevations above the patient’s baseline blood pressure; these clinical parameters have not been found to be a reasonable prognostic indicator of outcome (85). Eclampsia involves the development of seizures and/or coma, which represents central nervous system involvement, in a preeclamptic patient. Severe preeclampsia can be defined by the criteria listed in Table 14-3 (80,85). Thrombocytopenia (platelet count below 100,000/mL) is the most consistent finding in patients with preeclampsia (87). Table 14-4 outlines the risk factors for the development of preeclampsia (80,85).

The precise pathophysiologic factors involved in preeclampsia have been difficult to elucidate. Trophoblastic invasion by the placenta appears to be important because the severity of hypertension appears to be related to the degree of trophoblastic invasion (80,88). Vascular changes, specifically vasospasm, seem responsible for many of the serious clinical manifestations (e.g., hypertension and diminished renal function). Earlier investigations demonstrated support for the involvement of vasospasm in the etiology of preeclampsia. Specifically, the failure of the blunted pressor response to angiotensin II present in normal pregnancy is not seen in preeclampsia (89), and a progressive sensitivity to the pressor effects of infused angiotensin can be demonstrated after 18 weeks in patients destined to become preeclamptic (90). Other factors may involve an imbalance in the production of prostacyclin, a potent vasodilator, relative to levels of thromboxane, a vasoconstrictor, and alterations in the synthesis of nitric oxide and/or endothelin 1 (85).

Some fetal effects of preeclampsia reflect vasospasm with regard to placental perfusion. A decrease in uteroplacental perfusion may result in abruption, IUGR, oligohydramnios, or nonreassuring fetal status demonstrated on antepartum testing. Placental abruption, which occurs in fewer than 2% of patients with chronic hypertension, can be twice as high in those with superimposed preeclampsia (82,83). In 1990, Burrows and Andrews (91) found that neonatal thrombocytopenia occurred in more infants of hypertensive than in those born to normotensive mothers (9.2% compared with 2.2%).

Prematurity significantly contributes to the increased rates of perinatal morbidity and mortality associated with preeclampsia. Conservative management, recommended in a tertiary care setting or in consultation with a maternal- fetal medicine subspecialist, may be attempted for women with severe preeclampsia remote from term (80,82). Delivery remains the only definitive treatment, and ultimately may be required (92,93). In cases managed conservatively, patients are hospitalized; bed rest with frequent clinical and laboratory assessments and corticosteroids are administered to enhance fetal pulmonary maturity. Magnesium sulfate, an agent long recognized for its anticonvulsant effects, is often administered (85). Although phenytoin sulfate may be used, magnesium sulfate is the preferred drug (94,95). Blood pressure control is achieved with the use of antihypertensive agents such as labetalol, hydralazine, or nifedipine (80,81,82). A protocol of high-dose corticosteroid administration stabilizes both clinical and laboratory parameters in patients with HELLP syndrome before term (96,97).

Several different antihypertensive agents are used during pregnancy. α-Methyldopa is frequently used because it is a safe, effective agent that has been studied extensively. Newer antihypertensive drugs that are safe in pregnancy include beta-blockers, labetalol (a combination alpha- and beta-blocker), and calcium channel blockers such as nifedipine. There is concern about using diuretics during pregnancy because of an accompanying plasma volume reduction (83). The National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy suggests that diuretics can be used safely except in situations in which there is already evidence of diminished uteroplacental function (80,81). Angiotensin-converting enzyme (ACE) inhibitors, commonly used in young, nonpregnant adults, are both teratogenic and fetotoxic if administered during gestation. Skull defects (hypoplastic calvaria and encephalocele) and in utero renal failure leading to oligohydramnios, pulmonary hypoplasia, long-standing neonatal anuria, and fetal or neonatal death have been reported (98,99,100). In women using an ACE inhibitor, antihypertensive medication is changed to another agent once pregnancy is confirmed.

Over the years, prevention of preeclampsia has been attempted. Initially, a beneficial effect was demonstrated in high-risk groups that were given low-dose aspirin (60 to 80 mg daily), but subsequent studies have been conflicting and a similar effect was not found in low-risk women (101,102,103,104). Similarly, other studies have obtained disappointing results in evaluating supplementation with calcium, magnesium, zinc, fish oil, or use of different antihypertensive medication (103). More recently, evaluation of the uteroplacental circulation via transvaginal Doppler sonography at 12 to 16 weeks suggests an association with abnormal waveforms and increased risks of developing preeclampsia later in gestation (103,104,105,106,107). Further study is ongoing.

Diabetes mellitus complicates nearly 4% of pregnancies (108,109). The disease is classified based on the requirement for insulin therapy into type 1 (insulin-dependent) or type 2 (non-insulin-dependent) diabetes. The White classification system for diabetes in pregnancy, developed in 1949, is based on age of onset and duration of disease, as well as disease progression with respect to vascular complications (110). With continued improvements in glucose control, assessment of fetal well-being, and neonatal management, the White classification is no longer as helpful as it once was in the management of the pregnant diabetic (111). Instead, the distinction can be made between diabetes that preceded a woman’s pregnancy (pregestational diabetes) and diabetes first recognized during pregnancy (gestational diabetes).

Fetuses of diabetic mothers may have growth disturbances at both ends of the spectrum: IUGR and macrosomia. IUGR, fetal growth less than or equal to the 10th percentile for gestational age, is not an infrequent finding in pregnancies of women with vascular complications of pregestational diabetes. In diabetic pregnancies, IUGR often results from uteroplacental insufficiency, usually associated with maternal hypertension, although fetuses with congenital anomalies also may exhibit signs of IUGR.

Infants of mothers with both pregestational and gestational diabetes are at risk for macrosomia, which refers to fetal growth beyond a specific weight regardless of gestational age, usually above 4,000 g or 4,500 g (112). With the development of a new national reference for fetal growth based on data from over 3.8 million births (113), clinicians can distinguish between macrosomia and large-for-gestational-age (above the 90th percentile for a given gestational age) (112,113). Several large studies support the continued use of 4,500 g as a value above which a fetus should be considered to be macrosomic (112,114,115,116,117). The antenatal sonographic diagnosis of macrosomia remains inaccurate, especially in diabetics (118,119,120). In women with normal glucose tolerance, 2% of infants weighed more than 4,500 g compared with 6% of women with untreated borderline gestational diabetes (112,121). Even in the absence of gestational diabetes, however, higher infant weights are associated with higher maternal glucose levels (112,122,123). In patients with untreated gestational diabetes, up to 20% of infants may be macrosomic (112,124).

Macrosomia is accompanied by the additional risks of prolonged labor, fracture of the clavicle, birth trauma from
shoulder dystocia, and instrumented or cesarean deliveries (125,126,127,128,129). Shoulder dystocia occurs in only 1.4% of all vaginal deliveries (130). Among deliveries of diabetic mothers, however, shoulder dystocia occurs two to six times more frequently than in the nondiabetic population (120,129,131). The complication has been reported to occur in 20% to 50% of infants of diabetic mothers with birth weights above 4,500 g (114,115,116,117,120). Brachial plexus injury is the most serious complication resulting from shoulder dystocia. Fortunately, it is rare, occurring in approximately 0.05% to 0.19% of vaginal deliveries (112,120), but that risk is increased 18- to 21-fold among vaginally delivered infants weighing more than 4,500 g (114,120,132,133,134,135). The majority of cases of brachial plexus injury resolve within 1 year (136,137), yet permanent injury is more commonly seen among infants with birth weights greater than 4,500 g (138). Important information for the practicing clinician to remember is that most cases of shoulder dystocia and brachial plexus injuries occur among infants who are not macrosomic (135). Clavicular fracture, which usually resolves without any permanent effect, can be seen in 0.3% to 0.7% of all deliveries (134,139,140), and its occurrence is increased up to ten times in the macrosomic infant (134).

Polyhydramnios, defined as excessive amniotic fluid, is not an unusual finding in diabetic pregnancies. A four-quadrant sonographic assessment of the amniotic fluid index (AFI) defines polyhydramnios as an AFI greater than the 95th percentile for gestational age (141). In diabetic pregnancies, the etiology is not clear, although fetal malformations or poor glucose control may be related. When polyhydramnios complicates maternal diabetes, higher rates of perinatal morbidity and mortality are reported (142).