The two components of low birth weight — preterm delivery and fetal growth retardation — lead singly or in combination to a heavy burden of mortality, short- and long-term morbidity, and high medical and social costs. Although different types of long-term morbidity are still being identified, the primary focus of concern has been neurodevelopmental disability resulting from the cerebral palsies, hydrocephalus, and visual and auditory deficits among survivors of very low birth weight. Additional special risks in such infants include chronic lung disease, acute lower respiratory infections, and rehospitalization in infancy and early childhood.

It is well known that signs of respiratory distress or disease (RD), such as high breathing frequency, chest wall retractions, grunting, and cyanosis, are common in newborn infants. The signs may be very transient or develop into a potentially life-threatening condition. When they are first observed, they often have a very ambiguous prognostic significance.

Surfactant deficiency has come to be understood as equivalent clinically to lung immaturity, except in situations of extreme prematurity or lung hypoplasia. Thus, surfactant deficiency and the diagnosis respiratory distress syndrome (RDS) have been applied quite uniformly to preterm infants with respiratory distress who have no other identifiable problems. While the diagnosis can be made more specific by assay of surfactant phospholipid composition (12), such measurements are seldom made in clinical practice. Perinatal discussions of the premature with respiratory distress also have been focused toward surfactant by the availability of multiple tests of “lung maturation” based on the assay of surfactant components in amniotic fluid (19). While surfactant is necessary for normal lung function, adequate surfactant is not sufficient to assure normal gas exchange in the preterm. There are no simple ways to separate surfactant deficiency states clearly from other aspects of lung development, such as airway development, alveolarization, and the development of the pulmonary vasculature in the preterm infant. The anatomic data of Hislop et al. (13) and Langston et al. (22) indicate a considerable variability in the number of alveoli in the preterm infant at each gestational age, resulting in large differences in potential gas exchange surfaces in different infants. Airway development also differs between infants of comparable gestational ages, as is evident clinically by susceptibility to the development of pulmonary interstitial emphysema. Assuming that surfactant treatments for RDS fully correct the surfactant deficiency, a realistic expectation based on animal studies (16), surfactant treatments should be useful in the near future to eliminate the surfactant deficiency component of lung immaturity syndromes. Any residual lung disease can then be better characterized.

For the past 25 years, attempts have been made to treat respiratory distress syndrome (RDS) by administration of surfactant into the airways. This practice is not yet an accepted form of therapy. Lack of generally available surfactant during the present era of prospective studies will hopefully decrease the therapeutic misuse of this novel medicine.

Continuous positive airway pressure was rediscovered and described by Gregory et al. (13) and has since been accepted as an aid in the treatment of newborn infants with pulmonary disease. A variety of technical solutions with beneficial effects for the newborn infant with respiratory distress has been described. Early treatment with continuous positive airway pressure is reported to reduce the length of time in which the demand for oxygen is increased and to make it possible, during a subsequent period on mechanical ventilation, to ventilate the infants at lower airway pressures than when continuous positive airway pressure is started late and the pulmonary disease is allowed to progress more rapidly (8;19).

Respiratory insufficiency has previously been a frequent cause of neonatal death, especially in preterm infants. As late as in 1967, Silverman and associates (66) found that in infants with hyaline membrane disease (HMD), mechanical ventilation with a body-enclosing negative pressure respirator did not improve survival. Before 1970, the mortality among infants who required respiratory therapy was high (20;46;70). At the end of the 1960s and the beginning of the 1970s, several new methods were introduced to improve ventilation of newborn infants. Kirby and coworkers (41) introduced intermittent mandatory ventilation as a way of weaning from mechanical ventilation. In a series of studies, Reynolds and coworkers evaluated the effects of different peak airway pressures, respiratory frequencies, and inspiratory:expiratory ratios on arterial blood gases and right to left shunts (32;58;59;60).

The improved survival rate of premature infants with respiratory failure is attributable to advances in mechanical ventilation, although an adverse consequence has been an increased incidence of bronchopulmonary dysplasia (BPD) (1;32). Positive pressure ventilation with its attendant “barotrauma” is suspected in the causation of BPD. While many attempts to alter respirator variables, such as pressure and time components, have produced optimal patterns for gas exchange, evidence is lacking to support any one pattern that minimizes the incidence of chronic lung injury. The high incidence of BPD has promoted a search for alternative methods of ventilation that might reduce lung injury through a reduction in peak pressure applied to the lung. Additional motivation has come from the need for oxygenation when mechanical ventilation has failed or pulmonary interstitial emphysema has developed. Less compelling reasons have come from the desire to avoid high swings in thoracic pressure that might adversely affect cardiac output, venous return, and cerebral blood flow.

Respiratory movements are controlled by two pairs of neuronal groups in the brain stem: nucl. tractus solitarius and nucl. paraambigualis. These neuronal pools receive drive inputs from the forebrain and hypothalamus, and from central and peripheral chemoreceptors (9). Whether there is endogenous spontaneous respiratory rhythmic activity has been a matter of controversy, but this has recently been discovered in brainstem preparations of the neonatal rat in studies by Onimaru and Homma (27).

Extracorporeal membrane oxygenation (ECMO) is a means of diverting a fraction or all of the circulation through a device that permits gas exchange across a permeable membrane. The site of removal and the site of return of blood are dictated primarily by practical considerations based on the volume of flow of blood to be diverted and whether a particular organ is to be bypassed. The prototype of extracorporeal oxygenation is heart-lung bypass, used for various types of cardiac surgery, in which the entire venous return is diverted through an oxygenator (bubble or membrane type) and returned to the aorta. Since the earliest reports of the use of ECMO in neonates, venoarterial bypass has been the preferred route, with blood drained from the right atrium and returned either to the carotid artery or to the femoral artery, although the former seems to have been used most often (2;4;11;12;14;18). In some reports, veno-venous bypass has also been accomplished with removal of blood from the right atrium and return into the umbilical or femoral vein (13;19). Gas exchange across the lung may also continue, although the ventilator rate, tidal volume, or peak inspiratory pressure are usually reduced markedly during the period of extracorporeal oxygenation. Anticoagulation is accomplished with few problems by infusion of heparin sulfate throughout the duration of the procedure. There now have been enough reports in the literature (individual cases or series of patients) to demonstrate that the procedure can be carried out with a minimum of technical difficulties in newborn infants (2;3;4;11;12;18), although the selection of “appropriate” patients reduces the morbidity and technical challenges of the procedure.

Respiratory distress in newborn and young infants often develops as a result of acute lung injury, in which disruption of the normal barrier function of the pulmonary endothelium and epithelium causes protein-rich interstitial and alveolar edema. Several conditions may initiate acute lung injury, including aspiration of meconium or gastric contents, bacterial or viral infection, overzealous resuscitation, and birth associated with incomplete lung development that requires ventilatory support with positivepressure mechanical ventilation and high concentrations of inspired oxygen. The latter condition usually occurs after premature birth, but it also may occur as a consequence of impaired fetal lung growth secondary to diaphragmatic hernia or chest compression from lack of liquid in the amniotic cavity. Acute lung injury sometimes progresses to a chronic form of lung disease, which is characterized by edema, fibrosis, airway distortion, and nonuniform inflation of the lungs.

Bronchopulmonary dysplasia (BPD) is a progressive disorder of the lungs of newborn infants initially involving distal airways, but followed in weeks by the development of abnormalities in all parts of the lung architecture. BPD develops primarily but not exclusively in premature infants. The incidence in the general population has never been assessed accurately in a prospective manner. One retrospectively calculated incidence figure is 1.2 infants per 1,000 total live births per year in the United States (17). With approximately 4 million births annually in North America, there are about 5,000 babies each year who develop BPD. The additional weeks, months, and, in some cases, years of hospitalization, the frequent rehospitalizations, and the mortality of 20–50% beyond 1 month of age (17) underscore the disproportionate financial and emotional toll of BPD not only on patients and families but caregivers as well. The glacial rates of improvement in the illness, the frequent setbacks, and the profound disruption to family life all support the contention that BPD is a major public health problem.

Persistent pulmonary hypertension of the newborn or persistent fetal circulation is a clinical syndrome that is usually apparent within the first 2 days after birth because of the presence of hypoxemia (2;12;19). The syndrome was first described in an abstract by Gersony, Due, and Sinclair (6) in 1969. Two infants were reported who had “RV decompensation, cyanosis and clear lung fields… in the absence of recognizable cardiac, pulmonary, hematologic or CNS disease.” The syndrome has been associated with aspiration of meconium, diaphragmatic hernia, asphyxia, hemorrhage, shock, and maternal infection (4;18). In other cases, there is no clear antecedent event. Despite considerable interest in the problem and a wealth of research related to pulmonary vasoregulation and vascular development in the fetus and newborn, the etiology of the syndrome remains obscure 20 years since its recognition.

Two observations about the ductus arteriosus are established: most normally developed infants will have a patent ductus arteriosus (PDA) at the time of delivery, and most fullterm infants will have anatomic obliteration of the ductus lumen by 3–6 months of age. Studies designed to determine how a PDA contributes to an infant's morbidity or when a PDA becomes a persistently patent ductus arteriosus have been hampered by the lack of consistent diagnostic criteria for defining the condition. Ellison et al. (22) attempted to evaluate several commonly used criteria for diagnosing a large leftto-right shunt through the ductus arteriosus by noting the occurrence of each sign both before and 36–48 hours after surgical ligation. No single criterion sufficed as an indicator of PDA. Certain signs, such as continuous murmur or hyperactive left ventricular impulse, were specific for a PDA, but lacked sensitivity; conversely, M-mode echocardiography and ventilatory support criteria were very sensitive but lacked specificity.

Systemic blood pressure (BP) is the product of cardiac output and total peripheral resistance. Cardiac output is controlled by the heart rate, myocardial contractility, preload, and afterload. Vascular resistance (vascular hindrance × viscosity) is under local autoregulation and general neurohumoral control through sympathetic adrenergic innervation and circulating catecholamines. Sympathetic innovation predominates in organs receivingflowin excess of their metabolic demands (skin, splanchnic organs, kidney), while innervation is poor and autoregulation predominates in the brain and heart. The distribution of blood flow depends on the relative resistances of the organ circulations. During stress (hypoxia, low cardiac output), a raise in adrenergic tone and in circulating catecholamines leads to preferential vasoconstriction in highly innervated organs, so that blood flow is directed to the brain and heart. Catecholamines also control the levels of the vasoconstrictors renin, angiotensin II, and vasopressin. These general principles also apply to the neonate.

As those involved in neonatal intensive care are well aware, water and heat balance play a very important role in the well-being of the newborn infant. While some aspects of water and heat balance have been studied extensively, others have been less elucidated. Recent studies have demonstrated that both insensible and sensible water loss depend on the maturity of the infant and that both environmental factors and the activity of the infant can have dramatic effects on both these modes of water loss. As heat loss is affected by the insensible and sensible water loss from the infant, the same factors also have implications for heat loss, but heat loss is also influenced by other factors. The aim of this article is to review some of the knowledge relevant to the fields of water and heat balance in the neonatal period.

The kidney is structurally and biochemically immature at birth. As a consequence, renal function is low (3;10;ll;18). Glomerular filtration rate (related to body surface area or to body weight) is approximately 25% of that in adults. The capacity of several different tubular transport systems is lower in the infant than in the adult (2;5;13;21;28). A low transporting capacity of the neonatal kidney will sometimes result in undesired losses of electrolytes, amino acids, and peptides. The capacity to concentrate urine is low (7;29), and disturbances of serum tonicity, therefore, are common. The low concentrating capacity can be attributed to renal immaturity. The capacity of the newborn fullterm as well as preterm infant to release antidiuretic hormone is normal 7;31).

In a review of the literature concerning the nutritional needs of newborn infants in intensive care, one surprisingly finds that only a few studies on this subject have been performed. This is in contrast to the extensive literature regarding the nutritional needs of healthy, growing newborn infants. It is astonishing, further, to find that there are many reports on the metabolic effects of illness in adults but hardly any on those in children and newborn infants. In fact, in a recent textbook of pediatric intensive care, in the chapter on nutrition and metabolism in the critically ill child, the authors state, “the metabolic derangements that occur in children have not been well delineated during critical illness, so that data and findings discussed have been taken mostly from the adult literature.”

The introduction of noninvasive imaging techniques allowed diagnosis of intracranial pathology in the newborn infant during life, which previously was only possible at postmortem. This has stimulated new interest in correlating pathology with the pathophysiological changes that lead to these events, the neurological signs associated with them, and their ultimate effect on the neurological state of the survivors.

In spite of major developments in prenatal supervision, perinatal asphyxia remains an important reason for the development of brain damage (18). Epidemiological investigations suggest that perinatal asphyxia actually represents a factor of increasing frequency as a cause of severe cerebral injury (9).