Mortality and Acute Complications in Preterm Infants
Although the mortality rate for preterm infants and the gestational age-specific mortality rate have dramatically improved over the last 3 to 4 decades, infants born preterm remain vulnerable to many complications, including respiratory distress syndrome, chronic lung disease, injury to the intestines, a compromised immune system, cardiovascular disorders, hearing and vision problems, and neurological insult. Infants born at the lower limit of viability have the highest mortality rates and the highest rates of all complications. Few studies have reported mortality and morbidity rates in gestational age-specific categories, which limits the information available for counseling of parents before a preterm delivery and for making important decisions on the timing and the mode of preterm delivery. Although much progress in the treatment of infants born preterm has been made, many of the medications and treatment strategies used in the neonatal intensive care unit have not been adequately evaluated for their efficacies and safety. The high rates of neurological injury in preterm infants highlight the need for better neuroprotective strategies and postnatal interventions that support extrauterine neuromaturation and the neurodevelopment of infants born preterm.
The significance of preterm birth lies in the complications of prematurity sustained by the infant and the impacts of these complications on the infant’s survival and subsequent development. Many clinical research studies of infants born preterm limit their outcomes to neonatal mortality and morbidity. Complications and the disturbance of normal development may result from factors that influence prenatal development and the etiology of preterm birth, but the extent to which this happens is often unknown. Although this chapter is by no means a complete catalog of complications of preterm birth, this chapter discusses how these various complications reflect immaturity; the impact that they have on survival, organ maturation, and health; and the efficacies of a number of intervention strategies designed to prevent and mitigate the effects of these complications. As outlined in Chapter 2, information based on gestational age is preferred over information based on birth weight because of the value of knowledge of gestation age in making decisions regarding preterm delivery and prenatal counseling of the parent.
The complications of preterm birth arise from immature organ systems that are not yet prepared to support life in the extrauterine environment. The risk of acute neonatal illness decreases with gestational age, reflecting the fragility and immaturity of the brain, lungs, immune system, kidneys, skin, eyes, and gastrointestinal system. In general, more immature preterm infants require more life support. There is controversy about how infants at the border of viability should be managed (see also Chapter 2 for discussion of Perinatal Mortality of Infants Born at the Limit of Viability). Neonatologists may vary in terms of how conservative they are with regard to treatment of these infants and some may regard treatment of infants at these very early gestational ages as experimental. The reader is referred to Appendix C for further discussion of ethical aspects of decision-making at the threshold of fetal viability.
The response of the infant’s organ systems to the demands of the extrauterine environment and the life support provided have an important impact on the infant’s short and long-term health and neurodevelopmental outcomes (Chapter 11). These outcomes are also influenced by the etiology of the preterm birth; maternal and family risk factors; and the extrauterine environment, including the neonatal intensive care unit (NICU), home, and community.
Infants born preterm are more likely than infants born full term to die during the neonatal period (first 28 days) and infancy (first year), and mortality rates increase proportionally with decreasing gestational age or birth weight (Alexander et al., 1999; Allen et al., 2000; Lemons et al., 2001;
CDC, 2005i) (see also Figures 9-2 and 9-3 in Chapter 9 and Appendix B). The leading causes of infant mortality in the United States are preterm birth, low birth weight, and birth defects; so preterm birth and low birth weight are major contributors to infant mortality (Alexander et al., 2003; CDC, 2005i; Petrini et al., 2002). Dramatic declines in infant and neonatal mortality and gestational-age specific mortality over the last several decades have been attributed to improvements in obstetric and neonatal intensive care, especially for infants born preterm and small for gestational age (Allen et al., 2000; Alexander and Slay, 2002). However, the United States most recently had an increase in infant mortality from 6.8 to 7.0 per 1,000 live births in 2002 and an increase in the preterm birth rate to 12.3 percent in 2003 (CDC, 2005i) (see also Chapter 1 for discussion of mortality rates and variations in mortality by race and ethnicity).
Intranational and International Comparisons
Large variations in infant mortality rates exist among different geographical regions as well as among racial and ethnic groups (Alexander et al., 1999; Allen et al., 2000; Carmichael and Iyasu, 1998; Joseph et al., 1998) (Chapters 1 and 2 and Appendix B). The United States ranked 28th of 37 industrialized nations in infant mortality in 2001 and has a higher rate of low birth weight. Although increasing preterm birth rates and racial and ethnic disparities in the rates of preterm birth have been implicated, methodological factors are contributors to these differences (Chapter 2 and Appendix B). For example, efforts at resuscitating infants born at the lower limit of viability, thereby classifying them as live births (not still-births), increases the rate of infant mortality because so many infants born before 24 weeks of gestation die (Alexander et al., 2003; CDC, 2005i; MacDorman et al., 2005). The dearth of international comparisons of preterm birth rates is due to similar methodologic concerns about how gestational age, live births, and fetal deaths are recorded and reported (Appendix B).
The rate of infant mortality among African American populations in the United States in 2000 was 14.1 per 1,000 live births, more than twice the national average of 6.9 per 1,000 live births (NCHS, 2002). The rate of access to high-quality neonatal intensive care varies by race and ethnicity (Alexander et al., 2003; Morales et al., 2005; Wise, 2003). Preterm birth rates for African Americans are more than twice those for of Hispanic or white infants (Alexander et al., 2003). Although African American infants born preterm have had a survival advantage over white infants born preterm, this gap is narrowing, and the higher proportion of African American infants born preterm and the higher mortality rate among African American infants born full term play a greater
role in the disparity in African American and white infant mortality rates (Allen et al., 2000) (Appendix B).
Effects of Regionalization and NICUs on Mortality
Access to neonatal intensive care was recognized as an important issue in the 1970s. Schlesinger (1973) was the first to report differences in the rates of neonatal survival among hospitals. The dearth of physicians and nurses skilled in the new techniques and support services needed to care for sick neonates led to the development of regional programs with NICUs with prescribed structures and functions, formalized arrangements for obstetric referrals, and transportation systems for sick neonates (Blackwell et al., 2005) (see Chapter 14). Regionalization initially involved transporting sick newborns from community hospitals to regional medical centers and outreach community education on the stabilization of acutely ill newborns.
The main arguments in favor of regionalized care rested largely on improved neonatal survival after its introduction into a geographically defined region (McCormick and Richardson, 1995). Low birth weight infants born in hospitals without a NICU had higher risk-adjusted mortality rates than those born in hospitals with an intermediate- or high-level NICU, and the mortality rate only marginally improved with subsequent transfer of the infant to an NICU (Cifuentes et al., 2002). The advantage of the earlier identification of high-risk pregnancies and referral to tertiary perinatal centers before delivery is supported by the more favorable outcomes for infants whose mothers were transported to perinatal centers before delivery compared with the outcomes for infants transported after birth (Doyle et al., 2004a; Kollée et al., 1988; Levy et al., 1981).
The concept of regionalized services has evolved to include the prenatal period and a fully integrated system of consultation, referral, and transport (McCormick et al., 1985).
Guidelines for designating levels of perinatal care (Level I, II, or III, depending on the resources available, the delivery volume, and geographic need) have been developed. As the proportion of infants with birth weights of less than 1,000 grams born at Level III perinatal centers has increased, their survival has improved, and the gap in survival between infants born in and out of such centers has increased (Saigal et al., 1989). In addition to gains in safety and expertise, the development of highly integrated vertical networks is inherently cost-effective because of the elimination of fragmented and redundant services.
This level of integration of regionalized perinatal services is difficult to achieve. In Georgia, the rate of delivery of infants with birth weights of less than 1,500 grams at recommended perinatal centers was better for urban mothers than for rural mothers who lived farther away from regional cen-
ters (Samuelson et al., 2002). Other factors associated with a lack of access to subspecialty care include content of prenatal care (e.g., risk assessment, education about signs and symptoms of labor, and communication and transportation plans), delays in assessment of labor, the adequacy of emergency transport for pregnant women, and the willingness to transfer mothers before delivery. Samuelson et al. (2002) speculated that 16 to 23 percent of neonatal deaths among infants with birth weights of less than 1,500 grams could be prevented if 90 percent of infants born outside hospitals with subspecialty care were delivered at the recommended hospitals (assuming that mortality differences were due to the level of care). As advances in health care improve the rates of survival of infants born preterm, access to care in regionalized subspecialty centers becomes increasingly important in determining infant mortality rates.
COMPLICATIONS OF PRETERM BIRTH
Developmental immaturity affects a wide range of organ systems. This section describes the short-term complications of preterm birth in terms of fetal development as well as injury to fragile organ systems during the perinatal and neonatal periods. Many of these complications have lifelong consequences for the health, growth, and development of infants born preterm. As described in Chapter 6, the complex interplay of the mechanisms involved in preterm delivery, including inflammation and cytokine injury, has also been implicated in the pathogenesis of chronic lung disease, necrotizing enterocolitis, retinopathy of prematurity (ROP), and brain white matter injury in the preterm infant. Although some randomized, controlled trials demonstrate the safety and effectiveness of a few treatments for neonates, most standard NICU treatments and interventions have not been adequately investigated. The role that defining and treating the complications resulting from preterm birth plays in the health and neurodevelopmental outcomes of children born preterm argues for more long-term outcome studies and more rigorous studies of new therapies and medications before they are widely adopted.
Lungs and Respiratory System
The primary function of the lung is gas exchange (i.e., they inhale oxygen and exhale carbon dioxide). Fetal breathing movements begin as early as 10 weeks of gestation, and the breathing of amniotic fluid in and out is essential for the stimulation of lung development. Fetal breathing movements tend to be erratic and occur only 30 to 40 percent of the time up to 30 weeks of gestation. The failure of fetal breathing movements or a lack of amniotic fluid that can be breathed in and out results in underdeveloped
lungs (i.e., pulmonary hypoplasia), which can be incompatible with extrauterine life. By approximately 30 to 32 weeks of gestation, the lungs make surfactant, a soaplike substance that helps keep the air sacs (alveoli) open. Infants born before 28 to 30 weeks gestation lack alveoli and breath with their terminal bronchioles and primitive air sacs. After delivery, the breathing pattern generally becomes more regular and continuous, but immature regulatory systems can lead to brief episodes of not breathing (apnea) (see Chapter 6 for discussion of normal lung development and respiratory distress syndrome).
Respiratory Distress Syndrome
About 24,000 infants a year and 80 percent of infants born before 27 weeks of gestation will develop respiratory distress syndrome (RDS). RDS is associated with surfactant deficiency. The incidence of RDS increases with decreasing gestational age and is higher among white infants than African American infants at each week of gestation (Hulsey et al., 1993). Although respiratory distress is less common in infants born at 33 to 36 weeks of gestation and is rare in full-term infants, it can be severe, with a 5 percent mortality rate (Clark et al., 2004; Lewis et al., 1996). Antenatal administration of glucocorticoids to women at risk for preterm delivery reduces the incidence and severity of RDS as well as the rate of mortality (NIH, 1994) (see Chapter 9). Soon after birth, preterm infants with RDS develop rapid breathing, grunting, poor color, and crackling or diminished breath sounds breathing requires increased work. Respiratory failure because of fatigue, apnea, hypoxia, or an air leak (from alveolar injury) results from stiff lungs that need high pressures for ventilation.
RDS is an acute illness treated with respiratory support (oxygen, positive airway pressure, ventilator, or surfactant) as needed and improves in 2 to 4 days and resolves in 7 to 14 days. The optimal methods of providing respiratory support and even the safe and optimal blood levels of oxygen and carbon dioxide in very preterm infants remain quite controversial (Collins et al., 2001; Phelps, 2000; Saugstad, 2005; Thome and Carlo, 2002; Tin, 2002; Tin and Wariyar, 2002; Woodgate and Davies, 2001). The provision of exogenous surfactant through an endotracheal tube improves pulmonary gas exchange and reduces mortality (by 40 percent), air leak (by 30 to 65 percent), and chronic lung disease but does not influence neurodevelopmental or long-term pulmonary outcomes (Courtney et al., 1995; Dunn et al., 1988; Gappa et al., 1999; Ho and Saigal, 2005; Morley, 1991; Soll, 2002a,b,c; Stevens et al., 2002; Ware et al., 1990). A few randomized controlled trials have addressed the effectiveness of high-frequency ventilation or the use of an inhaled gas (nitric oxide) on survival and the severity of lung injury in severely ill preterm infants (Bhutta and
Henderson-Smart, 2002; Henderson-Smart and Osborn, 2002; Mestan et al., 2005; Van Meurs et al., 2005).
Not all acute respiratory illnesses in preterm neonates are RDS. Because congenital pneumonia is difficult to distinguish from RDS, infants with respiratory distress are generally treated with antibiotics. Some infants also have difficulty transitioning from the type of circulation that they have in utero, where gas exchange occurs in the placenta. When they breathe at birth, their circulatory pattern should change to send blood through their lungs. The retention of fetal lung fluid can also cause respiratory distress, but the condition improves as the fluid is reabsorbed.
Bronchopulmonary Dysplasia and Chronic Lung Disease
The chronic lung disease (CLD) that sometimes follows RDS in preterm infants is also called bronchopulmonary dysplasia (BPD). BPD/CLD is a chronic disorder that results from inflammation, injury, and scarring of the airways and the alveoli. It is associated with growth, health, and neurodevelopmental problems during childhood (see Chapter 11). Positive-pressure ventilation, high oxygen concentrations, infection, and other inflammatory triggers all contribute to lung injury; but the primary cause of BPD/CLD is lung immaturity. Especially for infants born at less than 28 to 30 weeks of gestation, the lung tissue is very fragile and the injured lung tissue tends to trap air, collapse, or fill with mucus and other fluids, which further compromise lung growth and development.
Various definitions of BPD/CLD have been used and are based on the respiratory support that an infant requires, but the most commonly used definition is a requirement for oxygen at 36 weeks of postmenstrual age (gestational age plus chronological age). Its incidence varies with gestational age at birth: in a study of infants born in 2002, 28 percent of infants born before 29 weeks of gestation and 5 percent of infants born 29 to 32 weeks gestation required oxygen at 36 weeks of postmenstrual age (Smith et al., 2005). By using this same definition, the incidence of BPD/CLD varies widely among centers:, from 3 to 43 percent among infants with birth weights of less than 1,500 grams (Lee et al., 2000; Lemons et al., 2001).
Infants with BPD/CLD have nutritional and fluid problems because of fluid sensitivity and increased metabolic needs, have difficulties with reactive airways (wheezing), and are quite vulnerable to infections, especially respiratory infections (Vaucher, 2002). Surprisingly few studies of the standard medications used to treat infants with BPD/CLD have been conducted, including diuretics and bronchodilators (Walsh et al., 2006). Modest improvements in survival and BPD/CLD rates have been reported with intramuscular injections of vitamin A (Darlow and Graham, 2002).
The most controversial treatment for preterm infants with BPD/CLD is
systemic postnatal corticosteroids (especially dexamethasone), which arrest alveolar and lung growth but allow the pulmonary system to mature (see Chapter 6). Two studies in the 1980s reported that long courses of relatively high doses of corticosteroids reduced the duration of time that oxygen and mechanical ventilation were needed in preterm infants (Avery et al., 1985; Mammel et al., 1983). More than 40 randomized, controlled trials of postnatal systemic steroids have been published, with most reporting improved gas exchange, fewer days of mechanical ventilation, and a lower incidence of BPD/CLD; but side effects, including glucose problems, high blood pressure, and growth failure were reported (Bhutta and Ohlsson, 1998; Halliday, 1999; Halliday and Ehrenkranz, 2001a,b,c).
Years after systemic steroids were widely adopted for the treatment of BPD/CLD, follow-up studies reported higher rates of cerebral palsy and cognitive impairment in infants randomly assigned to steroids than in those assigned to placebo, and systematic reviews of the available data have expressed similar concerns (Barrington et al., 2001a,b; Bhutta and Ohlsson, 1998; Halliday, 2004; Kamlin and Davis, 2004; O’Shea et al., 1999; Shinwell et al., 2000; Yeh et al., 1998). Two large trials of lower doses of hydrocortisone for the prevention of BPD/CLD were stopped because of adverse side effects (including gastrointestinal perforation) (Stark et al., 2001; Watterberg et al., 1999). One review calculated that for every 100 neonates given steroids within 96 hours of birth, BPD/CLD would be prevented in 9, while 6 would develop gastrointestinal hemorrhage and 6 would develop cerebral palsy.
Inhaled steroids are also frequently used, despite trials that show that they provide no significant benefits (Shah et al., 2004).
Whether corticosteroids should be used to treat the sickest infants with severe BPD/CLD (many of whom may die) remains controversial, especially if lower doses and much shorter courses are used (Doyle et al., 2005; Jones et al., 2005). Whether a drug that provides a short-term gain (and sometimes dramatic results) but increases the likelihood of serious long-term consequences should be used and how one decides between benefit to one organ system (the lungs) but adverse effects on another organ system (the brain) are serious dilemmas.
The likelihood of persistent respiratory problems during infancy is higher in preterm infants with BPD/CLD than in those without BPD/CLD. They may develop significant wheezing with respiratory infections (viral broncholitis) and may need to be rehospitalized, placed back on a ventilator, or even given exogenous surfactant (Kneyber et al., 2005). Preterm infants are especially vulnerable to respiratory syncytial virus (RSV) infection. The American Academy of Pediatrics recommends RSV prophylaxis for 6 months for infants born at 29 to 32 weeks of gestation and for 12
months for infants born at less than 28 weeks of gestation (AAP, 2006). BPD/CLD often results in residual effects on pulmonary function later in life: children who had had BPD/CLD as infants are particularly vulnerable to the effects of secondhand smoke and have higher rates of asthma, persistent growth problems, and neurodevelopmental disabilities (Hack et al., 2000; Jacob et al., 1998; Jones et al., 2005; Thomas et al., 2003; Vohr et al., 2005).
Another complication of preterm birth is apnea, in which infants may stop breathing for 20 seconds or more, sometimes accompanied by a slow heart rate (bradycardia). Immaturity of the control of breathing is the major cause of apnea and bradycardia, although sometimes preterm infants have obstructive apnea (an obstruction to the movement of air in their airways). They require constant monitoring but generally respond quickly to stimulation (or in the case of obstructive apnea, repositioning). They may occasionally need to be given some positive-pressure breaths to get them breathing again. There is no agreement as to what constitutes pathologic apnea or the threshold of apnea that requires treatment (Finer et al., 2006).
A number of strategies have been used to treat preterm apnea. The primary drugs used to treat apnea are the methylxanthines. Both theophylline and caffeine are effective, but caffeine has less toxicity (Henderson-Smart and Steer, 2004). Another drug, doxapram, has been associated with increases in cognitive delay (Henderson-Smart and Steer, 2004; Sreenan et al., 2001). The provision of vestibular stimulation is not as effective as treatment with methylxanthines for the prevention or treatment of apnea (Henderson-Smart and Osborn, 2002). There is no evidence that treatment of gastroesophageal reflux decreases the frequency or severity of apnea (Finer et al., 2006). Frequent apnea unresponsive to medications is treated with nasal positive airway pressure or mechanical ventilation.
Apnea generally resolves as the preterm infant matures. Occasionally, preterm infants continue to have apnea beyond term, and some are discharged on home apnea monitors. The long-term beneficial effects of the treatment of apnea in preterm infants in an NICU have not been demonstrated (Finer et al., 2006). Acute respiratory infections (especially RSV infections) may cause a recurrence of apnea. Although there is relationship between preterm birth and sudden infant death syndrome, the mechanisms are poorly understood and probably do not include apnea of prematurity (Baird, 2004).
The gastrointestinal (GI) tract digests and absorbs food, but it also has immune and endocrine functions and receives a good deal of input from the nervous system. It begins to form as early as the fourth week of gestation, and the stomach and the intestines are fully formed by 20 weeks of gestation (Berseth, 2005). The intestines double in length in the last 15 weeks of gestation (to 275 cm at term). The intestinal absorptive cells form as early as 9 weeks of gestation, and endocrine and immune functions also begin early. Taste buds form at between 7 and 12 weeks of gestation. However, preterm infants have difficulty with digesting nutrients because many specialized cells are not fully functional.
The earliest coordinated reflexes are related to stimulation around the mouth, with mouth opening in response to perioral stimulation occurring at 9.5 weeks of gestation and head turning occurring by 11.5 weeks of gestation (Hooker, 1952; Hooker and Hare, 1954; Humphrey, 1964). The fetus swallows by 10 to 12 weeks of gestation and can suck by 20 weeks of gestation. After birth, the newborn’s GI tract becomes colonized with bacteria, which aids with food digestion. Antibiotics alter this process. The safety and efficacy of giving preterm infants favorable bacteria (i.e., probiotics) for GI tract colonization is being studied (Bin-Nun et al., 2005).
Feeding intolerance is a common complication of preterm birth. The immature GI tract has difficulty digesting food necessary for ongoing growth and development. Very immature and sick infants receive parenteral (intravenous) nutrition with amino acids, glucose, electrolytes, and lipids. Preterm infants below 34 to 35 weeks of postmenstrual age require tube feeding because they cannot coordinate sucking, swallowing, or breathing. Providing the preterm infant with sufficient nutritional requirements for growth and development can complicate the treatment of other conditions.
Necrotizing enterocolitis (NEC) is an acute injury of the small or large intestines that causes inflammation and injury to the bowel lining and that primarily affects preterm infants. NEC occurs in 3 percent of infants born before 33 weeks of gestation and in 7 percent of infants with birth weights less than 1,500 grams (Lee et al., 2000; Lemons et al., 2001; Smith et al., 2005). It typically occurs within 2 weeks of birth and presents as feeding difficulties, abdominal swelling, hypotension, and other signs of sepsis. When NEC is suspected, infants are treated with antibiotics and bowel rest (i.e., no feedings).
The exact cause of NEC is unknown and, like most other complications of prematurity, is multifactorial. The preterm infant’s intestinal lining is fragile, and stresses (infections and insufficient oxygen or blood flow) can injure it. Inflammation is important in terms of both the etiology and the outcomes. Injury to the GI tract lining can progress through the wall of the
intestines, causing perforation and spilling of the intestinal contents into the abdomen, which causes peritonitis and sepsis. The gram-negative bacteria that colonize the GI tract secrete toxins that can cause severe systemic illness and death. Infants with perforated intestines require blood pressure support, surgery for removal of dead or dying bowel, and frequently, an ostomy until the bowel is healed. The damage may affect only a short segment of the intestine, or it may progress quickly to involve a much larger portion. Around the time of surgery, the infant’s nutritional intake is generally severely limited and the infant may require large amounts of blood products, fluids, and drugs (pressors) for the treatment of hypotension.
Survivors can experience significant short-term and long-term morbidities. Patients cannot be fed until the GI tract recovers, so they require parenteral nutrition and fluid. Although intravenous access is difficult in many infants, prolonged parenteral nutrition requires the placement of central venous catheters, which has attendant risks and complications of its own. Prolonged hyperalimentation and the absence of enteral nutrition can also cause liver damage with cholestasis. In addition, patients with significant disease can develop a stricture (narrowing of the bowel as it heals), which may require surgical intervention and which further compromises successful enteral feeding. Infants with extensive involvement of the GI tract are critically ill, and removal of large portions of the bowel results in malabsorption even after they have recovered. Occasionally, the injury is so extensive, that the small amount of intestines left is insufficient for growth and development or incompatible with life. Long-term morbidities can include ileostomy, colostomy, repeated surgical procedures, prolonged parenteral nutrition, liver failure, poor nutrition, malabsorption syndromes, failure to thrive, and multiple hospitalizations.
Because of the devastating nature of NEC, neonates are not fed during an acute illness. Feedings are introduced gradually, with each increase in the volume or the concentration of feedings carefully monitored, and feedings are stopped at the earliest signs of feeding intolerance. This situation creates a tension for the clinician: balancing the complications of the administration of intravenous fluids and parenteral nutrition against the complications of increasing enteral feedings too fast. If difficulty with digestion is anticipated, the infant may initially be given more elemental formulas. The provision of very small amounts of feedings initially stimulates the GI tract to produce the enzymes needed to digest larger volumes and concentrations of subsequent feedings. Attention to feeding regimens may improve feeding tolerance and reduce the incidence of NEC in NICUs (Patole and de Klerk, 2005).
Gastroesophageal reflux (GER) is common in preterm and full-term infants, often presents as regurgitation, and may adversely affect growth and health. It may also be manifested by aspiration pneumonia, wheezing,
or worsening of BPD/CLD because of an inability to protect the airway when refluxing. The presence of a nasogastric feeding tube increases the likelihood of reflux. Severe GER with aspiration of the stomach contents into the lungs is life threatening. GER is often treated with medications, including H2 blockers or protein pump inhibitors, which neutralize gastric acidity (and which may increase vulnerability to infection via the GI tract), and prokinetic compounds, which increase GI motility. The efficacies and safety of these medications have not been established, however. Occasionally, severe cases may require surgery, especially in infants with severe BPD/ CLD. There is no convincing evidence that the medications currently available for the treatment of GER are efficacious in treating or preventing apnea (Walsh et al., 2006).
Skin, which begins to form as early as 6 weeks of gestation, is an important barrier between the fetus or infant and the environment (Cohen and Siegfried, 2005). Skin plays important roles in fluid balance, temperature regulation, and the prevention of infection. The skin of infants born at the lower limit of viability (i.e., 22 to 25 weeks of gestation) is generally gelatinous, is easily injured when touched, allows tremendous loss of fluids, and does not provide an adequate barrier to infection. Fluid and electrolyte (i.e., salt) needs are often difficult to predict and are quite variable during the first several days after birth, until the skin toughens. Frequent procedures and significant infiltrates from intravenous lines lead to multiple scars in preterm infants. At the limit of viability, skin can scar from the removal of chest monitor leads. Covering the skin of preterm infants born before 26 weeks of gestation with a barrier ointment does not prevent but actually increases the risks of infection (Conner et al., 2003).
Infections and the Immune System
The interactions between the fetal and the maternal immune systems during pregnancy are complex (Taeusch et al., 2005). Carefully regulated changes in the fetal immune system are programmed to retain the pregnancy and reduce the likelihood of being attacked by the maternal immune system (i.e., as in an allogeneic graft) yet to prepare the fetus for birth and survival in the extrauterine environment. Many of the mother’s antibodies cross the placenta to protect the growing fetus beginning at 20 weeks of gestation, but most transfer during the third trimester.
Abnormalities of this delicate and complex interplay between the fetal and the maternal immune systems and infections can result in fetal compromise, maternal or fetal death, or preterm birth. Although the mechanism is
not well understood, many data support the association between subclinical infection and preterm birth (see Chapter 9). Infections with the rubella virus, cytomegalovirus, Toxoplasma, the syphilis spirochete, the malaria parasite, and human immunodeficiency virus during pregnancy can have devastating consequences for the fetus and infant (Beckerman, 2005; Pan et al., 2005; Sanchez and Ahmed, 2005). Other maternal infections and subsequent inflammation in the fetus have been implicated as causes of fetal brain injury (including white matter injury, disruption, and programmed neuronal cell death) and, later, neurodevelopmental disabilities (Dammann et al., 2002; Hagberg et al., 2005; Walther et al., 2000).
Preterm infants have immature immune systems that are inefficient at fighting off the bacteria, viruses, and other organisms that can cause infections. The most serious manifestations of infections with these agents commonly seen in preterm infants include pneumonia, sepsis, meningitis, and urinary tract infections. As many as 65 percent of infants with birth weights of less than 1,000 grams have at least one infection during their initial hospitalization (Stoll et al., 2004). Neonates contract these infections at birth from their mothers or after birth through their immature skin, lungs, or GI tract, which lack fully developed immunoprotective functions. They have difficulty confining infections to where they arise and forming abscesses, so sepsis (i.e., a blood-borne infection) frequently develops. Septic infants are generally critically ill, and infection can spread to other parts of the body (resulting in, for example, meningitis, an infection of the membranes that surround the brain). In addition to intravenous antibiotics, septic infants often require support for other organ systems that break down (e.g., respiratory and blood pressure support). Neonates with birth weights of less than 1,000 gram and infections have been found to have poorer head growth, more cognitive impairment, and higher rates of cerebral palsy than those who did not have infections as neonates (Stoll et al., 2004).
Invasive fungal infections occur in 6 to 7 percent of infants in an NICU, and the rates of such infections increase with decreasing gestational age and birth weight (Hofstetter, 2005; Stoll et al., 1996). Candida is the most common fungal species that causes infections in preterm infants and colonizes approximately 20 percent of infants with birth weights of less than 1,000 grams (Kaufman et al., 2001). Disseminated fungal infection, in which the infection is spread throughout the body, has a mortality rate of 30 percent. Prompt treatment with antifungal medication can prevent dissemination and improve survival, but side effects are frequent. Intravenous administration of the antifungal fluconazole as prophylaxis against fungal infections in infants with birth weights of less than 1,000 grams can reduce the rates of colonization and fungal infection (Kaufman et al., 2001).
The immune system has many component parts, and there are significant differences between the immune systems of neonates and adults and
their responses to inflammation and infections with pathogens. Inflammation is implicated in many of the complications of prematurity, including BPD/CLD, NEC, intracranial and especially white matter injury, and ROP. The complex relationships among pathogens, stress, the cytokine system, tissue injury, hormones, and multiple gene-environment interactions in producing or reducing inflammation have important implications for preterm birth, survival, health, brain injury, and neurodevelopmental outcomes (see Chapter 6).
Preterm infants can experience a variety of cardiovascular disorders, ranging from major morphological defects to dysfunctional autoregulation of blood vessels (hypotension). By embryonic day 20, the cells that will form the heart begin to differentiate (Maschoff and Baldwin, 2005; Schultheiss et al., 1995). The primitive heart beats by 4 weeks of gestation and is fully formed at the end of the 6th week. Because gas exchange occurs in the placenta, most of the fetal blood flow bypasses the lungs through the ductus arteriosus.
The ductus arteriosus normally closes after birth, when the lungs expand; air enters the lungs; and blood is redirected from the right side of the heart, through the lungs, back to the left side of the heart, and out to the body. In preterm infants, the duct may not close properly, which results in a patent (open) ductus arteriosus, which can lead to heart failure and reduced blood flow to vital body organs (e.g., the kidney and the GI tract). Heart murmur, active precordium, and bounding pulses are clinical signs; and an echocardiography performed at the bedside can confirm the presence of a patent ductus arteriosus and an otherwise normal anatomy. A patent ductus arteriosus can be asymptomatic and may close spontaneously in the first week of life, or it can complicate a preterm infant’s clinical course and increase the risks of intraventricular hemorrhage (IVH), NEC, BPD/CLD, and death (Shah and Ohlsson, 2006).
Approximately 5 percent of infants with birth weights of less than 1,500 grams are treated for patent ductus arteriosus (Lee et al., 2000). Medication and surgery are equally effective at closing a patent ductus arteriosus, and each has significant side effects and outcomes (Malviya et al., 2006). The most common medication used, indomethacin, has significant side effects because of the decreasing blood flow to the lower body (which results in decreased urine output and gastrointestinal perforation). Ibuprofen is effective and may have fewer side effects, but it has not been as well studied (Shah and Ohlsson, 2006). The value of indomethacin for the prevention of patent ductus arteriosus or the treatment of a patent with asymptomatic ductus arteriosus remains controversial (Cooke et al., 2003; Fowlie, 2005).
Although the focus has been on closing the patent ductus arteriosus, lower rates of mortality or morbidity (BPD/CLD, NEC, or neurodevelopmental disability) have not been demonstrated (Fowlie, 2005).
Hypotension is a frequent concern in preterm infants, but there is no consensus as to what the blood pressure readings should be in preterm infants with gestational ages of less than 26 or 27 weeks. The administration of boluses of normal saline and pressors is used to support blood pressure. Although preterm infants with severe refractory hypotension are often treated with physiological doses of hydrocortisone, its safety or efficacy has not been established.
Apnea and bradycardia are common in premature infants and aremanifestations of immature cardiorespiratory control (Veerappan et al., 2000). However, preterm infants and, indeed, some term infants can have bradycardia during feeding, despite the absence of other cardiorespiratory symptoms and a lack of clinical reflux. The nature of the autonomic nervous system’s contribution to these symptoms is not well understood (bradycardia could be due to increases in reflex parasympathetic autonomic nervous system activity).
Hematopoiesis is the generation of blood cells from stem cell progenitors. It begins in the embryo 7 days after conception (Juul, 2005). Stem cells are active in the aortogonadomesonephron at 10 days and then shift to the liver and, finally, the bone marrow. There are developmental changes in the numbers and functions of hematopoietic stem cells and in the various differentiated blood cells (e.g., red blood cells, white blood cells, and plate-lets). Red blood cells in the fetus contain fetal hemoglobin, which is necessary for intrauterine gas exchange because it has a higher affinity for oxygen. Fetal hemoglobin levels decrease after birth.
Fetal blood loss, fetomaternal hemorrhage, and hemolysis can all result in congenital anemia, but the most common hematologic complication in preterm infants is anemia of prematurity. Anemia of prematurity is an exaggeration of the physiological anemia of infancy because of suppressed hematopoiesis for 6 to 12 weeks after birth and is earlier in onset and symptomatic. Its causes are multifactorial and include blood loss from frequent blood sampling, the shorter survival of red blood cells in preterm infants, a suboptimal response to anemia, and a greater need for red blood cells with growth. Preterm infants often need red blood cell transfusions, and many of the sickest and most immature infants need multiple transfusions. A meta-analysis of a number of randomized controlled trials documented a modest reduction in the number of red blood cell transfusions required after the
administration of recombinant human erythropoietin and iron (Vamvakas and Strauss, 2001).
Auditory System and Hearing
The ear begins to develop at the end of 6 weeks of gestation and is fully developed by 20 weeks of gestation. A response to sound can be demonstrated in fetuses and infants born at 23 and 24 weeks of gestation, and auditory brainstem-evoked responses can be recorded this early in preterm infants (Allen and Capute, 1986; Birnholz and Benacerraf, 1983; Starr et al., 1977). The shape of the waveform changes and the conduction time decreases with increasing gestational or postmenstrual age.
One to two of 1,000 newborns suffer from congenital or perinatally acquired hearing disorders. The prevalence of neonatal hearing disorders has been reported to be increased 10- to 50-fold in infants at risk, which includes preterm infants. In addition to hearing impairment as a result of heredity, which is the cause of the largest percentage of hearing disorders, a number of in utero and neonatal complications (e.g., infections, immaturity, asphyxia, ototoxic medications, and hyperbilirubinemia) have been described to be risk factors for neonatal hearing disorders. Ventilated infants are at increased risk for otitis media. Significant hearing impairment, often requiring hearing aides, occurs in 1 to 5 percent of infants born at gestational ages of less than 25 or 26 weeks (Hintz et al., 2005; Vohr et al., 2005; Wood et al., 2000) (see Chapter 11).
Moderate to severe bilateral hearing impairment can distort the developing child’s perception of speech and may interfere with his or her attempt at speechproduction. If the hearing impairment remains undetectedthrough the critical period of language acquisition, that is, within the first 2 years of life, a profound impairment of receptive and expressive speech and language development can result. Early detection of hearing impairment facilitates early remediation (e.g., hearing aides or cochlear implants) and early intervention for speech and language acquisition (Gabbard and Schryer, 2003; Gravel and O’Gara, 2003; Niparko and Blankenhorn, 2003). The prognosis for functional speech and language skills improves with the early detection and treatment of hearing impairment (Yoshinaga-Itano, 2000).
Most communities are moving toward universal hearing screening for all newborns (White, 2003). The most widely used methods to screen new-borns for their hearing abilities are auditory brainstem responses and otoacoustic emissions (Hayes, 2003). Both methods detect the infant’s response to sounds. The auditory brainstem response provides an electrical recording of the brainstem’s response to sound. The otoacoustic emissions test evaluates the integrity of the cochlea (inner ear) by detecting the low sounds that the cochlea emits in response to sound perception. These tests
are sensitive but have low specificity rates. Neonates who fail a hearing test should have a repeat test and then be referred for confirmatory audiological testing and medical evaluation.
Progressive hearing impairment has been reported in infants with cytomegalovirus infection and persistent pulmonary hypertension of the newborn. These infants and infants who demonstrate a delay in language milestone acquisition should have a follow-up hearing test during the first year of life.
Ophthalmic System and Vision
Preterm infants are more likely than term infants to have significant abnormalities of all parts of the visual system, leading to reduced vision (Repka, 2002). The optic vesicles that will become the eyes form during the fifth and sixth weeks after conception (Back, 2005). The eyeball is well formed by the lower limit of viability (22 to 25 weeks gestation). However, a pupillary membrane covers the anterior vascular capsule of the lens and gradually disappears between 27 and 34 weeks of gestation (Hittner et al., 1977). The retina is a vascular layer in the back of the eye that translates light into electrical messages to the brain. The retina is the one of the last organs to be vascularized in the fetus (Madan and Good, 2005). Blood vessel-forming cells originate near the optic disc (where the optic nerve enters the retina) from spindle cell precursors at 16 weeks of gestation and gradually spread across the surface of the retina, from the center to the periphery. Vessels cover only 70 percent of the retina by 27 weeks of gestation, but in most cases the retina is completely vascularized to the nasal side by 36 weeks of gestation and to the temporal side by 40 weeks of gestation (Palmer et al., 1991).
The visual system functions very early, with the preterm infant blinking in response to bright light by 23 to 25 weeks of gestation and with papillary constriction in response to light by 29 to 30 of weeks gestation (Allen and Capute 1986; Robinson, 1966). By 30 to 32 weeks of postmenstrual age, the preterm infant begins to differentiate visual patterns (Dubowitz, 1979; Dubowitz et al., 1980; Hack et al., 1976, 1981; Morante et al., 1982). Visual acuity progressively improves with increasing postmenstrual age. The full-term neonate sees shapes (approximate visual acuity of 20/150) and colors and has a fixed focal length of 8 in. (anything closer or farther away becomes more blurry).
ROP is the most common eye abnormality in preterm infants. It is a neovascular retinal disorder, and its incidence increases with decreasing gestational age and decreasing birth weight. It is multifactorial in etiology, with the primary determinant being immaturity with an avascular retina (Madan et al., 2005). Environmental factors, including hypoxia, hyperoxia,
variations in blood pressure, sepsis, and acidosis, may injure the endothelia (the cells that line) of the immature retinal blood vessels. The retina then enters a quiescent phase for days to weeks and forms a pathognomonic ridge-like structure of mesenchymal cells between the vascularized and the avascular regions of the retina by 33 to 34 weeks of postmenstrual age. In some infants, this ridge regresses, and the remaining retina is vascularized. In other infants, abnormal blood vessels proliferate from this ridge; and progressive disease can cause exudation, hemorrhage, and fibrosis, with subsequent scarring or retinal detachment (i.e., the retina is pulled off the back of the eye). The presence of plus disease, in which dilated and tortuous blood vessels occur in the posterior pole of the eye, is especially ominous for an adverse visual outcome.
ROP occurs in 16 to 84 percent of infants born with gestational ages of less than 28 weeks, 90 percent of infants with birth weights of less than 500 or 750 grams, and 42 to 47 percent of infants with birth weights of less than 1,000 or 1,500 grams (CRPCG, 1988, 1994; Fledelius and Greisen, 1993; Gibson et al., 1990; Gilbert et al., 1996; Lee et al., 2000; Lefebvre et al., 1996; Lucey et al., 2004; Mikkola et al., 2005; Repka, 2002). Fortunately, severe ROP requiring therapy is less common, occurring in 14 to 40 percent of infants with gestational ages of less than 26 weeks, 10 percent of infants with gestational ages of less than 28 weeks, 16 percent of infants with birth weights of less than 750 grams, and 2 to 11 percent of infants with birth weights of less than 1,000 or 1,500 grams (Coats et al., 2000; Costeloe et al., 2000; Hintz et al., 2005; Ho and Saigal, 2005; Lee et al., 2000; Mikkola et al., 2005; Palmer et al., 1991). ROP resolves without significant visual loss in the majority (80 percent) of infants (CRPCG, 1988; O’Connor et al., 2002). Repka and colleagues (2000) found that involution occurred in 90 percent of infants with ROP by 44 weeks of postmenstrual age.
Treatments have improved the visual outcomes for children with severe ROP (i.e., threshold or plus disease, stages 3 and 4). The ablation of abnormal peripheral vessels with cryotherapy (in earlier studies) and laser therapy (in the last decade) have led to favorable visual outcomes in at least 75 percent of infants with severe ROP (CRPCG, 1988, 1994; Repka, 2002; Shalev et al., 2001; Vander et al., 1997). Continuing improvements in treatments and more timely treatments of severe ROP have served to reduce the proportion of children with severe visual impairment or blindness from 3 to 7 percent down to 1.1 percent in children with birth weights of less than 1,000 or 1,500 grams (Doyle et al., 2005; Hintz et al., 2005; Tudehope et al., 1995; Wilson-Costello et al., 2005) (see Chapter 11). Severe visual impairment or blindness occurs in 0.4 percent of children with gestational ages of 27 to 32 weeks, 1 to 2 percent of children with gestational ages of less than 26 or 27 weeks, 4 percent of children with gestational ages of 24
weeks, and 8 percent of children with gestational ages of less than 24 weeks (Marlow et al., 2005; Vohr et al., 2005).
Timely diagnosis and the prompt treatment of ROP are essential for improving visual outcomes. Screening ophthalmologic examinations require an experienced examiner and careful examination of the retina to the periphery with an indirect ophthalmoscope and lid speculum after dilation of the pupils. Revised guidelines for screening preterm infants for ROP have recently been published, with recommendations for which infants should be screened (infants with birth weights of less than 1,500 grams or gestational ages of less than 32 weeks or selected other preterm infants with an unstable clinical course), the timing and frequency of screening examinations, and the indications for treatment (Section on Ophthalmology, AAP, 2006)
Although visual outcomes have improved, ROP continues to be a major problem, especially in the most immature infants. The primary method of prevention is to prevent preterm births. Prevention of wild swings in blood pressure, blood oxygen and carbon dioxide levels, and acidosis has been recommended (Madan et al., 2005). There is much interest in the role of oxygen in ROP, but the optimal blood oxygen levels and oxygen saturation levels remain controversial (Saugstad, 2005; STOP-ROP, 2000; Tin 2002; Tin and Wariyar, 2002). Although the intravenous administration of high doses of vitamin E appears to reduce the incidence of severe ROP and blindness in infants with birth weights of less than 1,500 grams, it also appears to increase the incidence of sepsis and IVH (Brion et al., 2003).
Other ophthalmologic complications of prematurity include refractive disorders (especially myopia), strabismus (i.e., ocular misalignment), amblyopia (i.e., visual loss associated with reduced development of the visual cortex), optic nerve atrophy, cataracts, and cortical visual impairment (Repka, 2002) (see Chapter 11). Late ophthalmologic problems include angle closure glaucoma (i.e., increased pressure in the eye), retinal detachment, and phthisis (i.e., shrinkage and disorganization of the eye severely affected by ROP); but fortunately, these are rare. Severe threshold ROP is associated with other complications of prematurity, including BPD/CLD and IVH.
Central Nervous System
Neuromaturation is a dynamic process in which the central nervous system (CNS) is formed by a continuous interaction between the programmed genetic processes encoded within the genome and then the intrauterine environment, followed by the extrauterine environment. The successive turning on and then turning off of specific genes propel development forward, whereas surrounding cells, temperature, nutrients, and unknown environmental factors influence cell division, differentiation, function, con-
nections, and migration. At 16 days from conception, the neural plate, which contains the cells that form the brain, is formed. A neural groove then forms and then begins to close to become the neural tube by 3 to 4 weeks from conception. At one end of the neural tube, embryonic brain vesicles form and begin to differentiate into the forebrain, midbrain, and hindbrain (i.e., the prosencephalon, mesencephalon, and rhombencephalon, respectively) (Capone and Accardo, 1996). By the end of the 6th week, the basic subdivisions of the adult brain have formed. Neurons and their glial support cells actively divide during the first trimester, with the peak period of proliferation between the 2nd and 4th months of gestation. Neuronal migration is the mass movement of neurons from where they were formed to an ultimate destination in a specified layer of the brain and occurs between the 3rd and 5th months of gestation.
Fetal movement begins shortly after the brain begins to differentiate and can be detected by prenatal ultrasound as early as 8 to 10 weeks from conception. Fetal and infant activity and sensory input shape the development of the CNS. The fetus moves in response to cutaneous stimulation by 9 to 11 weeks and demonstrates the earliest signs of primitive reflexes (i.e., rooting and grasping) (Hooker, 1952; Hooker and Hare, 1954; Humphrey, 1964). Neurons continue to differentiate, and axons grow out and connect to dendrites to form synapses from the 6 month of gestation to at least 3 years from term. A complex and extensive network of neuronal circuits form; and they are shaped by patterns of electrical activity promoted by sensory input, movement, and responses to the environment. Fetal movements and responses are necessary for the normal development of the limbs and the CNS. Ongoing activity, learning, and sensory input determine which circuits are reinforced, whereas unused circuits are pruned. Myelination covers the neuron with a lipid sheath and reduces conduction times. The myelination process begins as early as 6 months gestation in some regions of the CNS and continues throughout childhood.
Incomplete formation of the CNS makes neonates vulnerable to CNS injury, especially if the infant was born preterm. Injury to the CNS can occur during pregnancy, labor, delivery, the transition to extrauterine life, or a subsequent illness or exposure. Many etiologies of preterm delivery (e.g., infection and maternal illness) contribute to fetal CNS injury. Concerns about the ability of the extremely preterm infants to tolerate the contractions of labor and the trauma of vaginal delivery have raised the question as to whether delivery by cesarean section is neuroprotective (Grant and Glazener, 2001). Trials to evaluate this question have suffered from recruitment problems, and there is not sufficient evidence of improved infant outcomes to balance the increased morbidity for mothers. Infants born preterm also have more difficulties with the transition from placental support to extrauterine life and the many vascular changes that occur.
In preterm infants, the white matter around the ventricles and highly vascular germinal matrix eminence are especially vulnerable to injury (de Vries and Groenendaal, 2002; Gleason and Back, 2005; Madan and Good, 2005). They have difficulties with autoregulation of cerebral blood flow (i.e., maintaining adequate cerebral blood flow, despite changes in blood pressure). Ischemia, hypoxia, and inflammation contribute to CNS injury in the preterm infant, but the relative importance of these factors remains controversial. The most common signs of CNS injury in preterm infants are IVH, intraparenchymal hemorrhage (IPH; bleeding within the substance of the brain), and white matter injury (including periventricular leukomalacia [PVL]). Neuroimaging studies, including ultrasound, computerized tomography, and magnetic resonance imaging (MRI), provide ways to visualize brain injury in infants. Ultrasound has the advantage of being cheaper and easily available (i.e., it can be performed at the bedside), but MRI is increasingly being used for better visualization of the brain parenchyma.
Germinal Matrix Injury, IVH, and IPH
IVH generally begins with bleeding into the germinal matrix just below the lateral ventricles (i.e., a subependymal or germinal matrix hemorrhage). During the late second and early third trimesters, the subependymal germinal matrix supports the development of cortical neuronal and glial cell precursors, which migrate to the cortical layers. The germinal matrix is highly vascularized, with a rich capillary network and a relatively poor supportive matrix. Blood filling the lateral ventricles may dilate the ventricles. The incidence and severity of IVH increase with decreasing gestational age and birth weight. Factors that contribute to IVH include hypotension, hypertension, fluctuating blood pressures, poor autoregulation of cerebral blood flow, disturbances in coagulation, hyperosmolarity, and injury to the vascular endothelium by oxygen free radicals. In 10 to 15 percent of infants a germinal matrix hemorrhage will obstruct venous return and lead to venous infarction of brain tissue (called IPH) (de Vries and Groenendaal, 2002).
Severe IVH can lead to ventricular dilation and posthemorrhagic hydrocephalus if there is an obstruction to the flow of cerebrospinal fluid, with increased intracranial pressure. Intermittent spinal taps or ventricular taps (i.e., drawing off of the cerebrospinal fluid with a needle) can relieve this pressure. This should be done primarily if the infant is symptomatic, as studies have demonstrated no benefits to regular taps of asymptomatic infants (Whitelaw, 2001). Once the blood is mostly cleared from the ventricles, a ventriculoperitoneal (VP) shunt can be surgically placed to drain the cerebrospinal fluid into the abdominal cavity where it is absorbed. De Vries and Groenendaal (2002) found that a third of preterm infants with large IVH required a VP shunt. Neither diuretics nor streptokinase (a clot
buster) reduces the need for a shunt, nor do they improve outcomes (and a borderline increase in motor impairment was detected at 1 year of age after the use of diuretics) (Whitelaw, 2001; Whitelaw et al., 2001).
Infants with subependymal or germinal matrix hemorrhage or IVH without ventricular dilation have a good prognosis; but those with IVH with ventricular dilation, posthemorrhagic hydrocephalus or IPH are at an increased risk of neurodevelopmental disability (de Vries and Groenendaal, 2002). As many as 11 percent of infants with birth weights of less than 1,500 grams have IVH with ventricular dilation or IPH (Lee et al., 2000; Lemons et al., 2001). The prevalence of neurodevelopmental disabilities in preterm infants with severe IVH and ventricular dilation or posthemorrhagic hydrocephalus ranges from 20 to 75 percent (de Vries et al., 2002; Fernell et al., 1994). Although early studies showed a high incidence of neurodevelopmental disabilities with IPH among preterm infants, recent studies have shown that the prevalence of disability varies with the size and the location of the hemorrhage (de Vries and Groenendaal, 2002; de Vries et al., 2002; Guzzetta et al., 1986). A study of infants born between 1979 and 1989 with gestational ages of less than 33 weeks found that the probabilities of a major disability at age 8 years were 5 percent for infants with a normal ultrasound, germinal matrix hemorrhage, or small IVH without ventricular dilation and 41 percent for infants with ventricular dilation, hydrocephalus, or cerebral atrophy (Stewart and Pezzani-Goldsmith, 1994).
Antenatal betamethasone (a corticosteroid) reduces the incidence of IVH in preterm infants, but many other treatments have been less successful (Crowley, 1999; NIH, 1994). There is not enough evidence to support the antenatal use of either phenobarbital or vitamin K to prevent IVH (Crowther and Henderson-Smart, 2003; Shankaran et al., 2002). Postnatal phenobarbital did not significantly improve the incidences of IVH, severe IVH, posthemorrhagic ventricular dilation, severe neurodevelopmental disability, or death; and there was a trend toward a longer duration of ventilation (Whitelaw, 2001). A meta-analysis of five trials of prolonged neuromuscular paralysis with pancuronium treatment in preterm infants with asynchronous breathing concluded that although pancuronium did help decrease the incidences of IVH and pneumothorax, concerns about its safety and long-term pulmonary and neurological effects precluded recommendation of its routine use (Cools and Offringa, 2005). Intramuscular doses of vitamin E may have reduced the incidence of IVH in preterm infants, but they were also associated with an increased incidence of sepsis (and high doses may increase the risk of IVH) (Brion et al., 2003). The prophylactic use of indomethacin in the first hours and days after delivery reduced the incidence and severity of IVH, especially in preterm boys, but the use of indomethacin results in many side effects (e.g., renal complications, NEC, and gut perforation) and it has little sustained effect on neurodevelopmental
outcomes (although it might improve the verbal abilities of boys) (Fowlie and Davis, 2002; McGuire and Fowlie, 2002; Ment et al., 2004; Schmidt et al., 2004). As with other complications of prematurity, the prevention of preterm birth would be the most effective way to prevent IVH and IPH.
White Matter Injury and Periventricular Leukomalacia
Injury to the periventricular white matter is a sign of CNS injury and is a complication of preterm birth. Its pathogenesis is currently the subject of extensive study (Damman et al., 2002; Wu and Colford, 2000) (see Chapter 6). White matter injury includes a spectrum of CNS injuries, from focal cystic necrotic lesions (also called PVL) to ventricular dilation with irregular ventricular edges or cerebral atrophy (as a result of the resorption of injured brain tissue) and extensive and bilateral white matter lesions. A complex interplay of etiologic factors predisposes the preterm infant’s white matter to injury, but the gray matter may be injured as well. Poor blood flow to regions of the brain because of obstruction, low blood pressure or an immature vascular system, poor autoregulation of cerebral blood flow, hypoxia, the vulnerability of preoligodendrocytes (i.e., supporting cells), excitatory neurotransmitters (e.g., glutamate), and harmful inflammatory substances (e.g., cytokines and free radicals) carried by the blood can all contribute brain injury. A meta-analysis found significant relationships between clinical chorioamnionitis, PVL, and cerebral palsy in preterm infants (Wu and Colford, 2000).
Imaging of white matter injury is more difficult than imaging of IVH or IPH (de Vries and Groenendaal, 2002). Ultrasounds should be repeated at 3 to 4 weeks after birth and at 34 to 36 weeks of postmenstrual age to detect signs of white matter injury, which evolve over time. The first sign may be an uneven density of the white matter that resolves (transient echogenicity) or that evolves into cystic lesions. Cystic lesions can collapse, so the timing and the quality of the ultrasound examinations are crucial for the detection of white matter injury. MRI is helpful for the detection of patchy or nonhomogeneous echogenicity.
Children with cystic PVL have a high risk of neurodevelopmental disabilities, and the more extensive that it is, the higher the risk that the children have (e.g., the risk is 100 percent with extensive bilateral cystic PVL) (Holling and Leviton, 1999; Rogers et al., 1994; van den Hout et al., 2000). These children are also at high risk for the development of cerebral palsy, which tends to be more severe with extensive PVL; cognitive impairments; and cortical-visual impairments with visual-perceptual problems. Children with more focal or unilateral cystic PVL also have a high incidence of cerebral palsy (up to 74 percent), but it tends to be a milder motor impairment (Pierrat et al., 2001). Approximately 10 percent of children with
periventricular echodensities develop cerebral palsy, generally in the form of mild spastic diplegia. MRI studies of infants born at term and older children have detected reduced regional cortical volumes (especially in sensorimotor regions and in both white and gray matter) that correlate with cognitive or neuromotor impairments (Inder, 2005; Peterson et al., 2000).
There has been a paucity of studies of strategies for the prevention of white matter injury and the amelioration of the effects of white matter injury. The effects of neuroprotective medications for IVH (e.g., indomethacin) on white matter injury are not clear, especially as the definitions of IVH are changing to include not just PVL but also irregular ventricular dilation and cortical atrophy. Some intriguing studies of insufficient naturally occurring developmentally regulated neuroprotective substances (e.g., hydrocortisone, thyroxine, and erythroietin) have suggested that they are associated with increased rates of mortality, BPD/CLD, and possibly, negative neurodevelopmental outcomes (Kok et al., 2001; Osborn, 2000; O’Shea, 2002; Scott and Watterberg, 1995; Sola et al., 2005; van Wassenaer et al., 2002; Watterberg et al., 1999). Many avenues are available for study and exploration, and research into the causes of preterm brain injury and neuroprotective strategies and the NICU interventions that can be undertaken to improve the neurodevelopmental outcomes of preterm infants is very much needed.
Complications for Near-Term or Late-Preterm Infants
For many years, attention has focused on high-risk obstetric and neonatal intensive care for extremely preterm infants and infants born at the lower limit of viability, although very little attention has been paid to the majority of preterm infants who are born near term (also called late-preterm infants). Although many deliveries of near-term infants are spontaneous or are indicated for maternal or fetal circumstances, it is important to keep in mind that these larger preterm infants born near term are more vulnerable to complications and disabilities than full-term infants. Although complications in near-term infants are not as frequent as they are in more-preterm infants, near-term infants have more perinatal and neonatal complications than full-term infants (Allen et al., 2000; Amiel-Tison et al., 2002; Wang ML et al., 2004). One study found that the incidence of RDS was as high as 15 percent among infants born at 34 weeks of gestation, whereas it was 1 percent among infants born at 35 to 36 weeks of gestation (Lewis et al., 1996). In a New England study of infants born at 35 to 36 weeks of gestation, more preterm infants than full-term control infants had evaluations for sepsis (37 and 13 percent, respectively), problems with temperature stability (10 and 0 percent, respectively), hypoglycemia (16 and 5 percent, respectively), respiratory distress (29 and 4 percent, respectively), and jaun-
dice (54 and 38 percent, respectively) (Wang ML et al., 2004). Poor feeding was more likely to delay the discharge of near-term infants than full-term infants (76 and 29 percent, respectively), and their hospital costs were higher (the mean cost difference between near-preterm and full-term infants was $1,596, with a median increase in cost of $221 per preterm infant). Accurate identification of late-preterm infants, even if they have normal birth weights, allows better anticipation and management of the complications associated with preterm birth.
Accurate estimates of gestational age and better measures of fetal and infant maturity would provide important information for clinical decision making. Health care providers and families need to carefully weigh the advantages of earlier delivery against the health, financial, and economic costs of preterm delivery.
Neurodevelopmental care is an approach to the intensive care of preterm and sick full-term infants in an NICU that supports neuromaturation and that also provides care for acute and chronic illnesses. Just as the intrauterine environment influences fetal development, the NICU environment influences the development of infants born preterm. The elements that make up the provision of neurodevelopmental support include NICU design and lighting, nursing routines and care plans, feeding methods, management of pain, attention to sensory input, activity and signs of stress, and the involvement of the parents in the care of their infants (Aucott et al., 2002). Although a number of studies have been conducted to evaluate the efficacies of various aspects of neurodevelopmental support in improving the outcomes for infants born preterm, few have yielded conclusive results. Conducting good randomized, controlled trials has proven to be quite difficult and expensive. Neurodevelopmental support is an important area that requires further study, both for the efficacy of the interventions that are used and for obtaining a better understanding of how NICU interventions support (or interfere with) the neuromaturation of infants born preterm.
The NICU presents preterm infants with an overwhelming amount of stimuli because of the active hospital environment and the infant’s exposure to multiple medical procedures (Aucott et al., 2002; Gilkerson et al., 1990). To minimize adverse stimuli and to support neuromaturation, NICUs therefore seek to implement strategies that mimic the intrauterine environment and that provide more appropriate stimuli that are geared to the infant’s state of alertness and responses (Aucott et al., 2002; Conde-Agudelo et al., 2005; Phelps and Watts, 2002; Pinelli and Symington, 2006; Stevens et al., 2005; Symington and Pinelli, 2006; Vickers et al., 2004). For example, attention to how infants are positioned and handled can influence the devel-
opment of their posture and muscle tone. Some NICUs have adopted more comprehensive approaches to developmental care, including kangaroo care and the Neonatal Individualized Developmental Care and Assessment Program (NIDCAP) (Conde-Agudelo et al., 2003; Symington and Pinelli, 2006).
It is not unusual for parents of critically ill neonates to feel overwhelmed by the technology that they encounter in the NICU and to have difficulty connecting with their newborn infant underneath all the NICU equipment. Family-centered NICU care is more of a philosophy than a program (Malusky, 2005) and involves providing families with comfortable seating, rocking chairs, privacy, and liberal visiting hours; encouraging them to bring in family photos or tapes of their voices; and saving bathing and feeding for family visits.
Besides providing milk that is more easily digested by vulnerable preterm infants, breast-feeding facilitates attachment by ensuring that the mother has a primary role in her baby’s recovery (Kavanaugh et al., 1997; Meier, 2001). Preterm infants fed breast milk have lower risks of infection and NEC, learn to nipple feed better, have higher cognitive scores, and may have a lower risk of chronic gastrointestinal diseases and allergies (AAP, 2006; Mizuno et al., 2002; Mortensen et al., 2002). Women who breast-feed have less postpartum blood loss, enhanced bone mineralization, and a reduced risk of ovarian and breast cancer (AAP, 2006).
Sensory Input and the NICU Environment
Early attempts to improve an infant’s environment focused on providing sensory stimuli, including rocking, stroking, holding, and moving, as well as auditory stimuli (e.g., the mother’s recorded voice and music) and visual stimuli, either alone or in combination (Aucott et al., 2002; Barnard and Bee, 1983; Mueller, 1996). Most studies of these interventions were flawed by small sample sizes, inadequate controls, or a failure to mask outcome evaluators. Few studies addressed the difficulty of confining an intervention to the study group without carryover to the control group. Finally, most studies failed to take into account background stimulation or the infant’s state of alertness and response to the stimulation.
The ability to control the frequency, duration, and intensity of incoming stimuli is an important aspect of learning. Fetuses and preterm infants respond to sound and light as early as 24 to 26 weeks of gestation (Allen and Capute, 1986; Johansson et al., 1992). Preterm infants can visually fixate and recognize visual patterns as early as 30 to 32 weeks of gestation (Allen and Capute, 1986; Dubowitz et al., 1980; Hack et al., 1976, 1981).
Fragile preterm infants, however, are easily overwhelmed by sensory stimuli and respond by closing their eyes, turning away, or even showing physiological instability (e.g., a drop in oxygen saturation levels). The NICU bombards preterm infants with multiple, invariable stimuli, including bright fluorescent lights, noise, and frequent handling (Aucott et al., 2002; Chang et al., 2001; Robertson et al., 1998). Infants with apnea receive tactile stimulation, and many procedures cause discomfort or pain. The ability to habituate to repeated aversive stimuli can be present as early as 24 to 30 weeks of gestation, but it requires the expenditure of energy, and the ability to respond may be imperfect in preterm infants (Allen and Capute, 1986).
Current NICU efforts focus on modifying the NICU environment, routines, and equipment to reduce noise and bright lights (Ashbaugh et al., 1999; Robertson et al., 1998; Walsh-Sukys et al., 2001). In addition to dimming bright overhead lights, measures can be easily adopted to indirectly shield infants’ eyes. Complete eye shielding or ear coverings are not beneficial, but decreases in light and sound stimuli on a circadian rhythm appear to promote weight gain (Brandon et al., 2002; Phelps and Watts, 2002; Zahr and de Traversay, 1995). Coordinating and clustering nursing and physician care avoids waking the infant unnecessarily, but there are concerns that clustered care may be too stressful for infants born before 30 weeks of gestation (Holsti et al., 2005).
Positive interactions and stimulation may be beneficial, as long as the infant’s responses are carefully monitored (and are therefore contingency based). Lullabies, parents’ voices, and rocking may improve the infant’s weight gain and shorten the length of the hospital stay (Gaebler and Hanzlik, 1996; Gatts et al., 1994; Helders et al., 1989). Rhythmic vestibular stimulation may facilitate quiet sleep but does not significantly influence weight gain, the frequency of apnea, feeding, or neurodevelopmental outcomes (Darrah et al., 1994; Osborn and Henderson-Smart, 2006a,b; Saigal et al., 1986; Thoman et al., 1991). Although kinesthetic stimulation may reduce the frequency of apnea, it does not prevent it and is less effective than medication (Henderson-Smart and Osborn, 2002; Osborn and Henderson-Smart, 2006a,b). Nonnutritive sucking (i.e., providing a pacifier for the infant to suck on during tube feeding) is associated with improved feeding and a shorter length of hospitalization (Pinelli and Symington, 2006). Some data suggest that gentle massage of physiologically stable preterm infants improves weight gain and decreases the length of the hospital stay (Vickers et al., 2004). Many believe that the best auditory, visual, kinesthetic, vestibular, and tactile stimulation interventions are positive interactions with the parents, who can easily be taught how to recognize and monitor their infants for signs of discomfort or sensory input overload.
Pain and Discomfort
The relationships between frequent or chronic pain, the stress response, cortisol levels, and the neurodevelopment of the preterm infant are extremely complex (Grunau, 2002; Grunau et al., 2005). The fetus or preterm infant responds to painful stimuli with increases in cortisol and endorphin levels as early as 23 weeks of gestation, but the neurotransmitters that attenuate pain develop later in postnatal life (Anand, 1998; Fitzgerald et al., 1999; Franck et al., 2000).
Preterm infants have an increased sensitivity to pain, and stimuli (such as handling) may be painful. These frequent painful experiences that preterm infants encounter in an NICU could lead to structural and functional alterations of their nervous system and subsequent altered pain responses through childhood (Anand, 1998; Anand et al., 2001; Grunau et al., 1998, 2001).
Guidelines for the management of pain in newborns have been published (Anand et al., 2001). The most widely used medications for the treatment of severe acute pain are morphine and fentanyl. Soothing measures (nonnutritive sucking of oral sucrose on a nipple) are also provided during minor procedures (Stevens and Ohlsson, 2000; Stevens et al., 2004). Studies have not consistently demonstrated the benefits of the routine treatment of mechanically ventilated newborns with narcotics (preemptive analgesia), however (Bellu et al., 2005; Grunau et al., 2005).
Positioning and Handling
Attention to how preterm infants are positioned and handled in the NICU may influence their posture and motor development after their discharge to home (Aucott et al., 2002). Failure to attend to how infants are positioned in an NICU can have adverse consequences, and small modifications in routine care take no additional time, nor do they incur additional cost.
In utero, fetuses are tightly flexed and contained (i.e., they have firm boundaries) and are bathed in amniotic fluid, which decreases the influence of gravity. Normal neuromaturation can therefore be promoted by positioning the infant in a manner that mimics the infant’s position in the intrauterine environment with extremity flexion and hip adduction, the avoidance of neck and trunk extension, and the promotion of body symmetry.
Because of their physiological instability, critically ill preterm infants receive minimal handling and stimulation but are repositioned on a regular basis according to nursing protocols (Aucott et al., 2002). Attention to how they are positioned can easily be incorporated into their routine care. More comfortable breathing, better oxygenation, and more time in deep sleep
have been noted in preterm and sick infants in the prone position than in infants in the supine position (lying on their side) (Grunau et al., 2004b; Wells et al., 2005).
Infants with narcotic abstinence syndrome have fewer signs of opiate withdrawal and better caloric intake when they are positioned more in the prone position (Maichuk et al., 1999). Although obstructive apnea is often effectively treated by repositioning an infant’s head and neck, placement of the infant in the prone position and kinesthetic and vestibular stimulation are not as effective as methylxanthines for the treatment of apnea of prematurity (Henderson-Smart and Osborn, 2002; Keene et al., 2000, Osborn and Henderson-Smart, 2006a,b).
A study of 21 intubated NICU infants on ventilators and in the supine position documented evidence of obstructed cerebral venous drainage when their heads were turned to the side, with resolution when they were positioned with their heads midline (Pellicer et al., 2002). Another series of studies demonstrated that more-preterm infants who were predominantly in the supine position in the NICU had asymmetries, asymmetric flattening of the skull, an early preference for use of the right hand, and an asymmetric gait (Konishi et al., 1986, 1987, 1997). Other neuromotor abnormalities formerly observed during infancy in a large proportion of preterm infants appear to have been influenced by how they were positioned in the NICU (e.g., shortened tibial bands from being in a frog-legged position and shoulder retraction with neck extensor hypertonia) (Amiel-Tison and Grenier, 1986; de Groot et al., 1995; Georgieff and Bernbaum, 1986). Hyperextension and hip contractures can make hand exploration, rolling, and sitting more difficult to achieve.
Although many small randomized controlled trials of NICU interventions have not definitively demonstrated beneficial effects, interventions that focus on mimicking the intrauterine environment appear to have at least some transient positive effects on motor development (Blauw-Hospers and Hadders-Algra, 2005; Goodman et al., 1985; Piper et al., 1986; Symington and Pinelli, 2006). Several small studies have found that stable preterm infants gain more weight and have improved bone mass when they are provided with some controlled physical activity each day (Moyer-Mileur et al., 1995, 2000). Allowing older infants to play in prone on a firm surface (“tummy time”) improves their ability to control their heads by strengthening antigravity muscles and improves their balance skills and shoulder stability (but has no effect on cognitive outcomes) (Mildred et al., 1995; Ratliff-Schaub et al., 2001).
Parents can use this approach during their visits, which provides an opportunity for them to be involved in their child’s care. Nurses and parents can easily use positioning aids, rolled blankets, or swaddling to position preterm infants symmetrically with their extremities flexed, shoulders
placed forward, and hips adducted to promote normal neuromaturation. Modeling of this approach for families in the NICU also enhances carryover to post-NICU care and promotes the parents’ interest in providing neurodevelopmental support for their infant.
Neonatal Individualized Developmental Care and Assessment Program
Als devised a highly organized comprehensive system for providing neurodevelopmental support in an NICU (Als, 1998). This system, commonly known as NIDCAP, has generated much interest and is often equated with NICU developmental care (Ashbaugh et al., 1999). Its systematic implementation requires development of NICU developmental care teams with dedicated staff trained and certified in NIDCAP, the systematic observation of the behavior of the infants, the coordination of care, and careful monitoring of the infant’s physiological responses. An individual developmental care plan is designed for each infant, with efforts to decrease adverse elements in the NICU environment.
Although some studies, including randomized clinical trials, have demonstrated the beneficial effects of NIDCAP on short-term growth, the duration of ventilation, the need for tube feedings, hospital stays, and cognitive abilities, many of these studies have been criticized for their small sample sizes or because they lacked masked outcome evaluators (Symington and Pinelli, 2006). In addition, for every positive effect reported, other studies have provided conflicting results. Because NIDCAP includes multiple interventions, it has been difficult to determine the efficacy of any single intervention. A recent paper found differences in brain structure and behavior at 8 months of age for infants who were part of NIDCAP, but further research is needed to assess longer-term outcomes (Als et al., 2004).
Cost has been cited as a reason for the lack of full implementation of NIDCAP, but no studies have addressed the economic impact of NIDCAP implementation (Symington and Pinelli, 2006). Only 30 percent of responders to a nursing survey published in 1999 worked in an NICU with a dedicated developmental care team and budget, although most reported the incorporation of aspects of NIDCAP care into their practices. The Cochrane Review concludes, “Before a clear direction for practice can be supported, evidence demonstrating more consistent effects of developmental care interventions on important short- and long-term clinical outcomes is needed. The economic impact of the implementation and maintenance of developmental care practices should be considered by individual institutions” (Symington and Pinelli, 2006).
Kangaroo care provides skin-to-skin care by placing the naked preterm infant in an upright position between the mother’s breasts and allows unlimited breast-feeding. This concept of caring for preterm infants originated in Bogota, Colombia, as a low-cost way to assist preterm infants with temperature regulation, nutrition, and stimulation (Charpak et al., 1996). Kangaroo care is initiated after a routine period of stabilization after birth. A number of studies from developing countries, including a few randomized controlled trials, suggest that kangaroo care improves weight gain (an additional 3.6 grams per day), reduces the incidence of nosocomial (i.e., hospital-acquired) infections, and reduces the incidences of severe illness and respiratory disease up to 6 months of age (Conde-Agudelo et al., 2003). Mothers who provided kangaroo care were more likely to continue to breast-feed and were more satisfied with the care that their infants received in the NICU.
Finding 10-1: Few postnatal intervention strategies that can be used to improve outcomes for children born preterm have been evaluated, and such intervention strategies are needed, especially for more immature preterm infants.
VARIATIONS IN NEONATAL COMPLICATION RATES
The complications of the newborn period noted in this chapter reflect in part the difficulty of establishing extrauterine life with immature organs. However, some of these complications may also reflect the interventions used in the NICU to sustain life. The question about variations in complication rates as a function of differences in management practices in the NICU was initially raised by a report of the substantial variations in the rate of bronchopumonary dysplasia or chronic lung disease among eight NICUs (Avery et al., 1987). The rates varied from a low of 5 percent in one NICU to almost 40 percent in another and could not be explained by the approaches to the management of respiratory distress syndrome reported by the NICUs, with one exception. The site with the lowest rate of chronic lung disease rarely used mechanical ventilation and tolerated blood gas values out of the physiological normal range. The interpretation of these variations was unclear, however. Even for a given gestational age, the severity of the complications may vary among infants, and without some measure of admission severity or case mix, units with higher rates of complications may simply be admitting sicker infants.
Addressing this issue required the development of admission severity measures, which occurred during the 1990s (Richardson et al., 1998). With
the development of two measures that assess the degree to which measures of physiological processes like oxygenation and blood pressure fall outside of the normal range, numerous investigators have documented variations in neonatal outcomes overall (Sankaran et al., 2002) as well as variations in specific complications (Aziz et al., 2005; Darlow et al., 2005; Lee et al., 2000; Olsen et al., 2002; Synnes et al., 2001) that cannot be explained by the severity of the infant’s condition on admission. Likewise, after adjusting for the severity of the infants’ conditions on admission, substantial differences in management have also been noted (Al-Aweel et al., 2001; Kahn et al., 2003; Lee et al., 2000; Richardson et al., 1999a; Ringer et al., 1998). Although most of this work has primarily been done with infants born before 32 weeks of gestation, data that are emerging indicate that such variations are also encountered in the complication rates and management of late-preterm infants (Blackwell et al., 2005; Eichenwald et al., 2001; Lee et al., 2000; Richardson et al., 2003). Although such variations may have less of an impact on survival and morbidity in the late-preterm infants than in earlier preterm infants, even minor variations among hospitals, such as a week’s difference in discharge time between those with the earliest gestational age at discharge compared with those with the latest, may have substantial economic benefits, because these late-preterm infants account for almost half of all NICU stays.
The observation of such variations, some of which appear to be unrelated to the clinical condition of the infant, has prompted efforts to reduce the variation and improve outcomes with existing technologies. As reviewed in a supplement to the journal Pediatrics, several groups are implementing quality improvement strategies to reduce the rates of unnecessary adverse outcomes (Horbar et al., 2003; Ohlinger et al., 2003), with some evidence of success (Chow et al., 2003).
Finding 10-2: Substantial interinstitutional variations in the complication rates for infants born preterm have been documented, and some outcomes, like physical growth, remain suboptimal.
Although the mortality rate for preterm infants and the gestational age-specific mortality rate have dramatically improved over the last 3 to 4 decades, preterm infants remain vulnerable to the many complications of prematurity. Infants born at the lower limit of viability have the highest mortality rates and the highest rates of all complications of prematurity. Few studies have reported mortality and morbidity rates in gestational age-specific categories, which limits the information available for the counseling of the parents before preterm delivery and for decision making on the
timing and the mode of delivery of an infant who will be born preterm. Better methods of evaluating fetal and infant maturity may improve the ability to predict the many complications of prematurity.
Although much progress in the treatment of preterm infants has been made, many of the medications and treatment strategies used in the NICU have not been adequately evaluated for their efficacies and safety. Even though progress in neuroimaging of brain structure of preterm infants is being made, research is needed to provide better indicators of CNS function and means for prediction of long-term neurodevelopmental outcomes. The high rates of neurological injury in preterm infants highlight the need for better neuroprotective strategies and postnatal interventions that support the extrauterine neuromaturation and neurodevelopment of preterm infants. Long-term health and neurodevelopmental outcomes should be the focus of new trials of treatments and intervention strategies for neonates born preterm.