In recent years, there have been considerable improvements in the care and outcome for newborn infants, particularly those with complex health care needs. In addition to physicians and nurses, an increasing number of health care professionals, including respiratory therapists, physiotherapists, pharmacists, dietitians, occupational therapists and others, have become important members of the neonatal health care team.

That evolution of enhanced roles in neonatal care began with the development of increased interests, knowledge and skills among all health care professionals. Just as pediatricians sub-specialized in neonatology, neonatal nurses, respiratory therapists, dietitians, social workers and pharmacists developed a range of knowledge and skills specific to the performance of clinical service in the neonatal unit, and in neonatal research and education.

Subsequently, practitioners developed skills that previously had been limited to other groups of health care professionals, most often physicians. Nurses and respiratory therapists performed such delegated acts as arterial puncture or intubation following a process of certification. Dietitians and clinical pharmacists ordered parenteral nutrition; a counter signature by a physician was often required.

The third stage in the evolution of enhanced performance was the development of staff with a wider range of specialized knowledge and technical and nontechnical skills designed to meet specific needs in a limited practice environment. This is exemplified by the development of roles, such as a transport nurse or transport respiratory therapist, through a process of hospital-based certification that may follow university or college courses and/or educational programs developed in individual hospitals. While enhanced roles have developed somewhat differently among various nonmedical health professional groups, they share several common factors:

• the increasing complexity and range of technology provided in neonatal care;
• the availability of a different set of qualifications and skills provided by other professionals;
• increasing survival, particularly of very low birth weight infants, resulting in increasing patient numbers requiring specialized care;
• the need to supplement declining numbers and availability of residents working in the neonatal intensive care unit (NICU); and
• the need for all health care professionals to develop or establish a broader scope of practice and increased academic responsibility.

Because of their size and stage of development, the health needs of neonates are quite different from those of adults.

In this continuing education unit, you will explore some of those needs, review developmental stages and the various diseases that afflict newborn infants (including numerous congenital defects). You will also learn about the various scoring and diagnostic tests utilized in neonatal care.

Upon successful completion of this course you will be able to:

• Describe the stages of development in fetal circulation.
• Identify and discuss the stages of fetal lung development.
• Explain the role surfactant plays in the development of neonates.
• Identify and discuss the processes of labor and delivery, the complications that can occur, and preventive steps that can be taken.
• Explain how to interpret neonatal radiographs
• List the markers for a high-risk pregnancy, and high-risk infants.
• Identify the clinical signs and appropriate treatments for common neonatal pathology.

We will begin with a review of pre and postnatal anatomical and physiological development. As you will see below, Table 1 summarizes the various stages of development experienced by the embryo.

Fetal Circulation

In utero, the prenatal circulation depends heavily on the mother's circulatory system for survival and development. During development, fetal nutrition, oxygenated blood, excretion, respiration, and protection are provided by the placenta. Development begins when the blastocyst attaches itself somewhere near the upper portion of the uterine cavity. Diffusion of substances between maternal and fetal blood occurs in the intervillous space created there.

While the ductus venosus and foramen ovale rarely cause problems at birth, if the ductus arteriosus does remain patent or reopens in response to hypoxia, this can lead to problems. If fetal circulation is maintained it can lead to a massive shunt causing hypoxia and pulmonary hypertension. The normal newborn shunt of 20 to 25% is much higher in the presence of a patent ductus arteriosus.

Pulmonary Circulation and Lung Development

The difference between fetal circulatory and nonfetal circulatory systems is that, since the fetus does not use the lungs for gas exchange, very little blood actually perfuses the pulmonary circulation. In the fetus, the body has mechanisms to bypass the lungs.

By 16 to 20 weeks of gestation, the process of pulmonary arterial branching has been nearly completed. The central pulmonary trunk in the fetus has elastic laminae and its walls have become thick. Prior to birth, blood bypasses the lungs in utero, with only about 10% of the cardiac output carried by the pulmonary circulation. With the majority of cardiac output being shunted past the lungs via the ductus arteriosus, fetal pulmonary vascular resistance (PVR) is very high, making flow through the ductus the path of least resistance.

Branching away from the main pulmonary artery, arterial elastic laminae decrease. Arteries with diameters of 2 mm down to approximately 200 microns undergo a transition to a more muscular type of vessel where there are changes in PVR. These periacinar are located adjacent to the terminal bronchioles. As the vessels get smaller, the amount of muscle gradually decreases and eventually disappears entirely in vessels adjacent to the alveoli.

Stages of Lung Development

Fetal lung development occurs in five phases:

1. During the embryonic phase, which begins at about 4-6 weeks of gestational age, the lung begins as a bud from the foregut. That bud branches into right and left primary bronchi. The branching continues, forming the proximal airways. Common malformations originating in this phase are the laryngeal cleft and tracheoesophageal fistula. There are few structural pulmonary abnormalities that occur during the embryonic phase because an embryo damaged during this period does not usually survive.
2. During the pseudoglandular phase, which lasts from about week seven to week sixteen, the predominant feature involves the formation of conducting airways. At about week 8, the diaphragm is also formed. During the pseudoglandular phase, the mucous glands, cilia, goblet cells, and cartilage also begin to appear in the conducting airways. Respiratory epithelium begins to differentiate during this phase, so injuries then can result in abnormal bronchial positions, connections, or number of bronchi. If the diaphragm does not form sufficiently to separate the thoracic and abdominal cavities, diaphragmatic hernia may result.
3. During the canalicular phase, weeks seventeen to twenty-eight, the gas exchanging area of the lung develops. Multivesicular and lamellar bodies associated with surfactant production begin to appear at about 20 weeks, and differentiation of Type I and Type II pneumocytes begins during this phase. Pulmonary capillaries are near alveoli at this point, but not near enough for effective gas exchange. Alveolar wall thickness is approximately 45 microns at 20 weeks gestation, and it decreases to about 20 microns at 32 weeks, then eventually to fifteen microns at term. In comparison, adult thickness is about 1 micron.

Surfactant produced during the canalicular phase is immature and easily destroyed. Its chemical composition (thus its functional capability) changes dramatically in the latter stages of gestation. The alveolar surface area at the end of this phase is approximately 1 square meter, and it increases to about 4 square meters at birth. Injuries to the fetus during this time can cause damage to the gas-exchanging area of the lung, causing a deficiency in alveoli, which may be severe enough to produce pulmonary hypoplasia. The gas-exchanging portion of the lung matures during the final two phases:

4. During the saccular phase, which lasts from weeks twenty-nine to thirty-five, interstitial tissue space decreases and airspace walls narrow. They become more compact, and lateral projections extend from the walls to divide the airspaces into smaller units. At 32 to 36 weeks alveoli are present.
5. The alveolar phase, from week thirty-six on, is devoted to final development and maturation of the alveoli. The number of alveoli at birth has been estimated to be anywhere from 10 to 150 million, and increases after birth becoming complete by the time the infant is 2 to 3 years old. At this point, the structure of the lung is usually sufficient to survive injuries, but injuries in this phase may interfere with alveolarization and lung function

Fetal Breathing and Lung Fluid

The prenatal lungs do not function as gas exchange organs, but they do serve important purposes:

1. The lung is a primary source of amniotic fluids.
2. Lungs act as reservoirs of carbohydrates needed for fetal energy.
3. They produce surfactant beginning at about 24 weeks

Fetal breathing movements, which appear to be essential for normal development, begin at around 12 weeks gestation. Factors that can affect fetal breathing movements include:

• maternal cigarette smoke, which stops fetal breathing for many hours.
• maternal consumption of alcohol or drugs
• decreases in 02, C02, or glucose
• stress
• prostaglandins

Arousal of the fetus appears to be more related to fetal breathing movements than central nervous system or chemoreceptor stimulation. When fetal breathing movements are absent, the cause is related to chromosomal and other abnormalities, or death. The fetus has breathing movements approximately 30% of the time during the last 10 weeks of gestation.

In utero, the lungs are filled with fluid, and fetal breathing movements exchange this fluid with amniotic fluid. Diagnostic testing of lung maturation via amniocentesis is made possible as a result of this fluid exchange. When Type II pneumocytes mature, they secrete surfactant into the lung and amniotic fluid.

Normal fetal development requires the presence of adequate amounts of lung and amniotic fluid. A diminished amount of either can result in a hypoplastic lung. Hypoplasia is a decrease in either lung weight or volume at birth. A decrease in alveolar number or an increase in alveolar size also can occur. An absence of fetal breathing movements or a lack of adequate space for lung growth may also cause hypoplasia.

The presence and chemical composition of surfactant can be tested in the amniotic fluid samples. Caution needs to be taken while obtaining these samples via amniocentesis because at 14-17 weeks, amniocentesis may reduce birth weight and lung volume; while at 22-25 weeks, amniocentesis may cause a decrease in the size and number of alveoli

Lung fluid, which contains glucose and other carbohydrates, acts as a storage reservoir that can be utilized after delivery. When the lung fluid is absorbed at birth, the materials are made available for use by the newborn's entire body. After delivery, the flow of fluid from lungs into the interstitium and lymphatics is facilitated by increased alveolar pore size.

During lung development, collagen is the dominant connective tissue in airways, blood vessels, and nonrespiratory components of the lung. While collagen fibers appear disorderly in actively branching airways, they are orderly in formed airways. Elastin, not collagen, is the dominant connective tissue in lung parenchyma.

Elastin first appears at 20 to 25 weeks gestation. Neither collagen nor elastin is particularly prominent at birth; however, the amount of elastin increases rapidly within 6 months after birth, helping to explain the easy rupture and decreased elastic recoil of the newborn's lungs. As the amount of elastin increases, the lung becomes more elastic.

A hypoplastic lung, one that has defective or incomplete development, is generally smaller than normal. Lung size is assessed via lung weight, volume, or DNA; however, lung weight is considered a poor indicator of development because most problems increase lung weight. When a lung weight-body ratio is used for assessment, a ratio <0.015 in a fetus less than 28 weeks old or a ratio <0.012 in an older fetus, is considered hypoplastic.

A better indicator of lung size is a measurement of inflated lung volume, because lung volume is unaffected by most fetal lung diseases. The correlation between crown-rump length and lung volume over the last half of gestation is excellent. Lung volume can therefore be predicted from crown-rump length, with lungs less than 69% of predicted volume being considered hypoplastic. Using lung DNA to reveal the number of cells present is another technique for assessing lung size. If lung DNA is <100 mg/kg body weight, the lungs are considered hypoplastic.

Surfactant

Surfactant is the active agent in the alveoli that cuts surface tension and reduces the need for high pressures to open the alveoli on inspiration. Surfactant is also important for changing capillary and interstitial pressures, facilitating removal of fluids from the lungs, and lowering pulmonary vascular resistance at birth.

The production of surfactant, a mixture of phospholipids (70-80%) and protein in relatively consistent proportions, sharply decreases after 34 to 35 weeks gestation. Evaluation of the content of the surfactant provides valuable information regarding the surfactant-producing system's maturity. Just prior to term, the lungs' volume stabilizes, and the lipid composition of aspirates changes. After about 20 weeks gestation, phosphatidylcholine in surfactant is produced and saved.

By contrasting amount of lecithin with that of sphingomyelin in the amniotic sample, a ratio of lecithin/sphingomyelin (L/S) can be calculated. A ratio of greater than 2 indicates lung maturity, while ratios less than 1 can be suggestive of pulmonary immaturity and the potential for respiratory distress syndrome (RDS). However, the RCP should be aware of the potential for deceptively high L/S ratios in infants whose mothers are diabetic, or if there is blood or meconium in the amniotic fluid.

The so-called shake-test can provide an estimate of the presence of surfactant. Mix the amniotic fluid with ethanol and shake the mixture approximately 15 seconds. Because surfactant produces stable bubbles, a closed ring of bubbles seen at the container's edge after 15 minutes indicates the presence of adequate surfactant. When no bubbles are seen after 15 minutes, an L/S ratio test should be conducted.

Surfactant production can be accelerated in premature (<34 weeks gestation) neonates by administering corticosteroids, which can also reduce the incidence and severity of RDS in some infants. Since steroids can mask the presence of infections in infants, they should be used with caution.

Fetal lung maturation can also be accelerated by the release of catecholamines during birth. The secretion of surfactant can be accelerated by: beta-adrenergic drugs, methylxanthines, a decrease in PaC02, alveolar stretch, and cAMP (Cyclic Adenosine Monophosphate). Inhibition of surfactant secretion can be caused by: decreased pulmonary blood flow, cholinergic stimulation, hypoxia, hyperoxia, and decreased pH levels.

The High-Risk Infant

Infants at high-risk are those who are expected to need special medical procedures at delivery. Both maternal and infant conditions can put a newborn at risk, and some of the factors include but are not limited to those on the following list. At the end of the outline, you will find a more in-depth discussion of some risk factors.

Maternal issues that can put an infant at risk for a problem delivery include:

• Health: Obesity or overweight condition, diabetic, emotionally stressed, viral infection early in pregnancy, exposure to radiation
• Lifestyle issues: tobacco, drug or alcohol use
• Obstetrical issues and complications:
• Previous delivery problems (still borns, preemies)
• First pregnancy late in life
• Multiple birth (twins, etc.)
• Post or prematurity
• Breech positioning
• Cesarean section
• Toxemia of pregnancy
• Abnormal or insufficient placenta
• Prolapsed cord
• Premature rupture of amniotic sac
• Meconium in amniotic fluid

At the time of delivery, the infant can present with conditions that signal high-risk condition, including:

• Less than 36 weeks of gestation
• Acute Respiratory Distress Syndrome (ARDS)
• Infection, blood diseases, or other anomalies
• The need for medications or special surgeries at delivery

It is crucial to assess fetal risk factors prior to birth. Approximately 29% of all pregnancies are deemed to be at risk for at least one of the complications listed above. In addition, approximately 5-10% of those at-risk pregnancies require the administration of CPR.

Reviewing of maternal history is an obvious way to identify potentially high-risk neonates. Complications may occur if you find evidence of a history of heart or lung disease, use of controlled substances, cigarette smoking, or infections. Low socioeconomic status or lack of education can also be indicators of potential problems because they are often related to inadequate prenatal care.

Expectant mothers >17 years of age or <38 should be seen as potentially being at risk, and the history of both current and all previous pregnancies should be reviewed for risk factors no matter the age of the soon-to-be mother. When reviewing clinical records, remember that the term gravida refers to pregnancy. Para refers to a pregnancy that terminated with the delivery of a viable neonate. Primipara refers to the mother's first delivery. Multiparous refers to a woman who has had two or more pregnancies which resulted in viable fetuses.

A mother's prior pregnancies involving problems should be considered carefully because history has a strong tendency to repeat itself in matters relating to childbirth. A history of fetal asphyxia, prematurity, RDS, maternal toxemia, ruptured membranes, infections, or bleeding during the current pregnancy are all reasons to consider the current fetus as being at risk for complications.

Multiple gestations can lead to problems, including: a breech birth, placental and cord problems, intrauterine growth retardation, and increased chance for premature death. Mortality increases with twins, particularly identical twins. Twin transfusion syndrome, where the circulations are connected, is also possible. This causes one baby to be polycythemic and the other to be anemic. The polycythernic baby manifests congestive heart failure and increased bilirubin levels. The anemic baby manifests hypotensive symptoms.

The presence of maternal diabetes mellitus (DM) is another cause for concern because problems commonly associated with DM include:

• prematurity
• congenital anomalies
• a predisposition to toxemia
• birth injury due to large baby
• still birth

Less severe DM is associated with delayed maturation of the lung, while severe DM causes chronic intrauterine stress and can accelerate lung maturation. Infants of diabetic mothers are more susceptible to infection, and more likely to be hypoglycemic, hypocalcemic, or have hyperbilirubinemia.

Toxemia involves the spread of bacterial toxins by the bloodstream and is a condition resulting from metabolic disturbances, such as those that occur during pregnancy. The resultant maternal hypertension can have serious consequences and lead to eclampsia (convulsions and/or coma). The term pre-eclampsia, which is often used interchangeably with toxemia, means a toxemia of late pregnancy which is characterized by hypertension, edema and proteinuria.

Toxemia during pregnancy causes a decrease in placental blood flow leading to uteroplacental insufficiency (UPI). UPI may occur in post-maturity infants, cyanotic maternal heart disease, or chronic hypoxia from maternal pulmonary disease.

UPI is more likely in the older primigravida, and can result in: intrauterine growth retardation, fetal death, chronic asphyxia, or the passage of meconium. When UPI is suspected, it can be assessed by measuring maternal urinary estriol levels. Urinary estriol normally increases throughout pregnancy, but measurements showing low or falling levels are indicative of UPI.

The placenta normally implants in the upper wall of the uterine cavity, but when an implantation takes place in the lower portion of the uterus, it is called placenta previa, or if the placenta separates prematurely from the uteral wall (abruptio placentae), the fetus is placed at risk. There are three types of placenta previa:

• A low implantation occupies the lower portion of the uterus, but does not cover the cervical opening.
• A partial placenta previa,covers a portion of the cervical opening, but does not cover it completely.
• In total placenta previa, the placenta is implanted low and completely covers the cervical opening.

All types of placenta previa can be readily diagnosed by ultrasound, and all cause varying degrees of obstruction to fetal passage and increase the chance of premature labor, early separation of the placenta, and hemorrhage. The most serious of these complications involves the early separation of the placenta from the uterus, referred to as abruptio placentae.

Separation of the placenta frequently causes premature labor, complete with its attendant risks, to begin. Fetal mortality approaches 50% due to the acuteness of blood loss, and maternal mortality ranges from 2 to 10% in severe cases ending in fetal death. The most common cause of abruption is maternal hypertension of any origin, including preeclampsia. Treatment of abruptio placentae includes strict management of blood volume, maintaining a hematocrit of 30-vol%. This is accomplished by IV administration of blood or crystalloid solutions.

Premature rupture of membranes (ruptures occurring 24 or more hours prior to delivery) put premature neonates at risk because of the increased potential for infection. Infants are considered postmature after the 42nd week of gestation, at which time the placenta begins to deteriorate. These babies often appear small for their gestational age and show signs of dwindling away. Postmaturity also predisposes to increased morbidity, including intrauterine asphyxia, meconium passage, difficult labor, and even premature death.

The delivery circumstances can also be predictive of potential problems. Vaginal delivery literally squeezes much of the fluid out of the neonate's lungs, easing the transition to life outside the mother's womb. On the other hand, cesarean-section deliveries don't allow for the squeezing action, increasing the chances that neonate will need special treatment to clear fluid out of the lungs and airway.

The neonate's amniotic fluid should be examined at birth for odor, color, consistency, and the presence of meconium. Normal amniotic fluid is thin, pale, and watery. Thick or foul smelling fluid may be indicative of infection. Yellow fluid can be related to infection or hypoxia. If part of the placenta is not perfused, the amniotic fluid may have a red wine color.

Meconium, a dark greenish stool, passes into the amniotic fluid in about 3-5% of preterm births, 10% in term babies, and about 40% of the time in post-term fetuses of more than 42 weeks gestation. The presence of thick particulate meconium in the amniotic fluid requires that as soon the head is delivered, the mouth, oro- and laryngopharynx be thoroughly suctioned to remove any meconium present. Upon delivery of the distressed neonate, the trachea should immediately intubated, suction applied to the end of the endotracheal tube, and the tube withdrawn.

Suction pressure should be set at 100 mm Hg, and suction applied for no more than 3-5 seconds. If meconium is suctioned out of the trachea, the neonate should be re-intubated with a new endotracheal tube and the procedure repeated until no meconium is suctioned. Blowby oxygen can then be delivered to help alleviate hypoxia with positive pressure ventilation beginning after completion of suctioning.

Each hospital needs to have a protocol covering the problem created when a severely depressed newborn has also aspirated meconium. Which risk is greater: that of blowing meconium further into the lungs with PPV before the trachea is clear, or risking asphyxia by not providing PPV until the trachea is clear? The general consensus on the issue seems to be that in severely depressed newborns, it may not be possible to clear the trachea of all meconium before initiating PPV. Be sure you know your hospital's policy on this clinical dilemma.

The average fetal heart rate (FHR) in early gestation is 140 beats per minute, dropping to an average of 120/min near term. FHR should be monitored continuously during labor, with normal ranging from 120 to 160/min. Fetal cardiac status can be measured either by simple auscultation with a stethoscope, or with relatively easy-to-use sophisticated electronics. Either way, routine monitoring of fetal heart rates has so significantly diminished adverse results of delivery that nearly every labor room now has a fetal heart monitor.

There is a normal beat-to-beat variation in the FHR, and an intact neural system responds to stimuli by increasing or decreasing FHR. For example, in a normal fetus, a loud noise causes a transient increase in FHR. If there is no FHR response to stimuli, higher brain-functions may not exist, and a total lack of variation may indicate brain stem activity only.

FHR monitoring can identify fetal distress (see Table 2) that is difficult to diagnose otherwise, and because the FHR monitor shows heart responses to asphyxia, it is an excellent way to identify infants who are being asphyxiated in utero. Since FHR also should correlate with contractions (normally, there is an increase of 15-20 beats in FHR with each contraction), FHR is usually monitored along with uterine contractions so the correlation between the two can be observed.

If fetal distress is suspected following FHR monitoring, the assessment of fetal scalp pH is used as a secondary tool to determine fetal well-being. The acid-base balance of the fetus is determined by the viability of the placenta, and its ability to exchange oxygen and carbon dioxide between maternal and fetal blood. If that exchange is disrupted, either at the placenta or in the cord, the resultant drop in pH can be measured.

There are two reasons for the drop in pH:

1. as blood gas exchange decreases, fetal PaCO2 increases, decreasing the pH;
2. facing hypoxia, the fetus begins to metabolize glycogen without oxygen, resulting in a dramatic increase in lactic acid. This metabolic acid, combined with increased PaCO2, causes the pH to drop.

The fetal scalp blood sample is obtained through the cervix between contractions. Normal fetal blood pH is considered to be above 7.25. A pH of 7.2 to 7.24 shows slight asphyxia, and a pH of less than 7.2 signifies severe asphyxia. Since maternal pH can influence fetal pH, it may also be necessary to determine the acid-base status of the mother concurrently. Fetal scalp pH is useful only in the presence of abnormal FHR tracings, since normal tracing indicates a healthy infant in most instances.

Equipment Preparation for the High-Risk Infant

In the presence of any of the high-risk situations just discussed, practitioners should anticipate respiratory problems and be prepared with whatever equipment will be necessary to:

• create a patent airway
• deliver warm and moist oxygen
• ventilate the patient
• deliver emergency medications
• maintain infant warmth
• provide for any or all of the above

Thermoregulation

The neonate is at highest risk of heat loss shortly after delivery. Thermoregulation is of utmost importance in the care of a newborn. The goal in the delivery room is to maintain an environmental temperature such that the neonate's core temperature remains in the normal range of 36.5 to 37.5°C. For this goal to be achieved, all avenues of heat loss must be minimized or eliminated. The mechanisms of heat loss in the newborn include:

• Evaporation can occur shortly after birth, when the infant is wet. As the liquid dries and evaporates it takes valuable heat with it. Immediately after delivery, every newborn should immediately be completely dried with a warmed towel or blanket. The head and face are particularly important. This drying helps reduce the evaporative heat loss. Thereafter, the neonate should be kept dry.
• Radiant loss involves the loss of heat from merely being placed near a cold surface, causing the transfer of heat from the warm baby to the nearby cooler object. After being dried, the neonate should be wrapped in a warm blanket and immediately placed beneath a radiant heater for resuscitation or examinations. Since a tremendous amount of heat loss can occur from the infant's head, a cap or other covering should be used.
• Both convective and conductive losses pose threats to the neonate's thermoregulation. Conduction takes place when the infant is placed on a cold surface, which pulls heat away from the baby. Convective heat loss is caused by air turbulence in the room that cools the infant. The neonate should be placed on a warming mattress and kept covered as much as possible to avoid convective heat loss. As quickly as possible, the neonate should be placed in a pre-warmed incubator. The longer the baby is left out in the open, the greater the chance of hypothermia.

The goal regarding neonates' thermoregulation is to achieve and maintain a neutral thermal environment (NTE), an environment that allows the infant to maintain his/her internal temperature without increasing oxygen consumption. Too much or too little heat are both adverse situations that lead to increased 02 consumption and apnea. To maintain that proper temperature:

• a thermistor is placed on the neonate's skin to monitor skin temperature
• the thermistor is connected to a servo control on the heat source
• the servo control then adjusts the heat to maintain the NTE. used as a general guideline:

WEIGHT
(grams)

Having reviewed development of fetal anatomy and physiology, risk factors, monitoring, anticipation, and preparation for problems, it is time to discuss the actual delivery and the first few minutes of extrauterine life.

The labor and delivery processes go through three distinct stages, including:

• Stage 1: The mother's contractions begin to dilate cervix and continue until dilation measures approximately 10 cm.
• Stage 2: The fetus is forced through the mother's cervical canal as a result of the contractions and abdominal wall pushing forces. The newborn's head normally presents first, and then the fetus rotates about 90 so the shoulders can present, permitting it to pass through the canal. As the infant's delivery is completed, the umbilical cord is clamped.
• Stage 3: When the placenta is expelled, the delivery is complete.

The Infant

At birth, expansion of the lungs causes PVR to fall dramatically. Ventilation helps to overcome fetal pulmonary vasoconstriction through several mechanisms. The first is a rise in Pa02 associated with filling the alveoli with air. This has a direct vasodilator effect on the capillaries. As more blood passes through the lungs and is exposed to air, vasoconstricting prostaglandins are inactivated. In the fetus, these prostaglandins are necessary to help maintain a high PVR and keep blood flowing through the ductus arteriosus.

A rising Pa02 also increases circulating bradykinin levels. This helps constrict the ductus and force more blood through the pulmonary circulation. Mechanical expansion of the lungs stretches and straightens the capillaries which, in turn, decrease the resistance to flow. In utero, the capillaries are very kinked and the path is tortuous. This keeps the PVR high. Stretching the alveoli and surrounding tissue corrects this.

As the lungs expand with air, P02 rises, the pathway is straightened, and prostaglandins are metabolized. More blood is now able to flow through the pulmonary capillaries and less through the ductus. Catecholamine release, triggered by the delivery process and the cutting of the umbilical cord, causes active constriction of the ductus to begin. Flow through the ductus is furthered impeded by rising aortic pressure.

Following delivery, neonatal pulmonary artery pressure declines incrementally from an average mean pressure of 39mm Hg at ten hours, to about 29 mm Hg at 15 hours. During the same time frame, neonatal aortic pressure rises to between 80-100 mm Hg at 20 hours post delivery.

The normal newborn shunt (20-25%) is considerably different from that of normal adults (5-10%), possibly due to continued leakage through the neonate's ductus and foramen ovale.

Neonate's average blood pressure varies according to gestational age and weight. The average neonatal systemic blood pressure at one hour is:

Average blood pressure at 12 hours is:

Infant Scoring Systems

As soon as the delivery is complete, there are numerous assessments to be made in order to determine the infant's health status. These include checking the respiratory and cardiac status, and weight. The Apgar scoring system (see Figure 1), named after Dr. Virginia Apgar, was developed as an objective way to evaluate the general status of the newborn at one minute and five minutes after birth. APGAR is also an acronym for what the practitioner will assess: The practitioner evaluates newborn Appearance, Pulse, Grimace, Activity, Respiratory rate and effort.

The five areas examined are respiratory effort, heart rate, muscle tone, reflex irritability, and color (see Figure 1). Each area is given a score of 0, 1, or 2 depending on the response noted. A score of "0" indicates maximum distress/dysfunction for that parameter. A score of "2" means the opposite. The first score is assessed at 1 minute after delivery, with a second evaluation performed at 5 minutes. Since the Apgar is an objective assessment of the infant's status, a 5-minute score that is higher than the 1-minute score indicates the effectiveness of the resuscitation.

After assigning numerical scores for the categories, scores are totaled, with normal infants scoring 7 to 10, moderately depressed infants scoring 4 to 6, and severely depressed infants scoring less than 4. Realistically, in the clinical setting the latter infants are not scored immediately because they are obviously in severe distress, and resuscitation measures are instituted before there is time to total scores.

The Apgar evaluations can be done every 5 minutes as needed, up to 20 minutes or when the resuscitation ends. The 5-minute Apgar score is predictive of future impairment, with a low score being associated with a likelihood of long-term damage. For example, an Apgar score of two or less at one minute is associated with a high mortality rate. An Apgar score of 8-10 is considered normal.

Figure 1. The Apgar scoring system.

As you can see, the Apgar is an excellent method for assessing the effectiveness of resuscitation; however it should not be used as the sole basis for making resuscitative decisions. One limitation of the Apgar system is that it was designed to assess normal full term infants, not preemies, so it is less valuable in their assessment. For evaluating premature neonates, umbilical cord pH or the Silverman-Anderson scoring system may be more valuable than Apgar.

In order to assess the degree of respiratory distress in neonates, practitioners often use the Silverman-Anderson scoring system. Like the Apgar system it evaluates five parameters and assigns a numerical score for each parameter. However, unlike the Apgar score, the lower the total score the better the baby's condition in the Silverman-Anderson system. The best score possible in each category is a "0" the worst is a "2". Parameters assessed are: retractions of the upper chest, lower chest, and xiphoid, nasal flaring, and expiratory grunt.

Table 3. Silverman-Anderson Scoring System

As you can see from Table 3, neonates with no retractions, flaring or grunting with synchronized respiratory movements are scored with "0s". Infants with visible retractions of the lower chest and xiphoid, with the upper chest lagging compared to the lower on inspiration, receive a "1". Minimal nasal flaring and an expiratory grunt heard only with a stethoscope also receive a "1". Marked retractions with a "see-saw" movement of the upper and lower chests scores a "2". Marked nasal flaring and audible expiratory grunting also receive a "2". Normal babies have a cumulative score close to "0". Severely depressed babies score close to "10".

RCPs need to be alert to the clinical signs of cardiac and pulmonary diseases in infants, including:

• Tachypnea (respiratory rate above 60)
• Wheezing (can indicate edema in small airways)
• Retractions (can indicate increased work of breathing)
• Rales (indicate fluid in small airways)
• Grunting (can increase the functional residual capacity)
• Nasal flaring (sign of respiratory distress)

Meconium Aspiration

Meconium is present in the amniotic fluid of nearly 10% of all infants at birth, and of those, between 20-25% go on to suffer some form of significant pulmonary disorder. Pneumothorax is frequently a complication of meconium aspiration.  Clinical indications include: tachypnea, rasping or faint respirations, patchy infiltrates on x-rays, hyperinflation, and severe cyanosis.

Aggressive suctioning is called for to eliminate airway obstructions, and placement of a nasogastric tube is needed to evacuate swallowed meconium and stomach contents. Treatment for metabolic acidosis is required, and supplemental oxygen with mechanical ventilation may be needed in order to maintain the infant's ABGs. Oxygen consumption and carbon dioxide production can be kept to a minimum through thermoregulation.

Pneumothorax

Tension pneumothorax can present as a complication of meconium aspiration, ventilation with positive pressure, pneumonia, hyaline membrane disease, and diaphragmatic hernia. Any newborn in respiratory distress should be reviewed for the presence of pneumothorax. Since pneumothorax can be seen in nearly 1% of all normal deliveries, even asymptomatic infants require observation of vital signs.

Clinical signs include onset of respiratory agitation or distress, tachypnea, nasal flaring and grunting, cyanosis, and movement of the apical pulse from the site of the pneumothorax. The most effective differential diagnosis can be made from radiographs. Severe distress may require insertion of a closed system chest tube with continuous suction. The rate of absorption can be enhanced with pure oxygen, but retrolental fibroplasia is a risk.

Pneumonia

Enteric organisms such as E. coli and group B streptococcus are the most frequent causative organisms of perinatal infections. Postnatal pneumonia is most often caused by contamination of the neonate's airway by infected humidifier reservoirs, poor hand washing, and other contaminated equipment. Nonbacterial organisms that can cause pneumonia are acquired by contact with an infected birth canal or nosocomial infection.

The diagnosis of neonatal pneumonia is clearly an inexact science. It is generally based on the history, physical examination, chest x-ray results, and lab data. Symptoms of pneumonia in a neonate, which often present within 48 hours of delivery, include tachycardia, signs of respiratory distress, flaccidity, pale skin, cyanosis, and foul smelling amniotic fluid indicating the presence of infection.

Other clinical signs may include an inconsistent WBC (either depressed below 5,000 or elevated above 15,000), elevated temperature, and/or x-rays showing unilateral or bilateral streaky densities in the perihilar region. Signs of a pneumonia acquired postdelivery can include an increasing tachycardia, poor feeding, lethargy, and aspiration of feedings.

Treatment of neonate pneumonia includes aggressive pulmonary suctioning, thermoregulation, fluid and electrolyte control, supplemental oxygen therapy, identification of the pathogen, and treatment with broad-spectrum antibiotics. Clinical symptoms need to be treated as they appear, with blood gas values closely monitored and treated.

Diaphragmatic Hernia

Diaphragmatic hernia, which occurs in about 1 in 2,200 births, is an extreme emergency and must be treated and corrected immediately upon diagnosis. Herniation of abdominal contents into the thorax is caused by an incomplete embryologic formation of the diaphragm. Ninety percent of the time it occurs on the left side, slightly lateral and posterior, through the foramen of Bochdalek.

When the herniation occurs on the left side, the stomach and intestines may enter the thorax and compress the lung, pushing the mediastinum to the right. The degree of distress noted in the neonate depends on the severity of the herniation. As the neonate begins breathing, the presence of the abdominal contents compresses the lungs, making it very difficult to complete inspiration. As air further distends the intestines and stomach, compressing the lungs even more, the neonate's respiratory distress worsens.

Symptoms of diaphragmatic hernia include cyanosis, respiratory distress, a flattened abdomen, excess amniotic fluid, and bowel sounds in the chest. Chest x-rays showing the loops of bowel in the thorax serve to confirm the diagnosis. In left-sided hernias, heart sounds can be heard in the right chest: X-rays in right-sided hernias show a large density created by the liver in the right thorax.

The treatment includes immediate insertion of a nasal gastric tube attached to suction and  evacuate abdominal gas. Ventilation, if needed, should be done through an endotracheal tube using rates near or above 100/min, and low PIP and PEEP pressures in order to avoid barotrauma. Surgical repair of the defect should be done through the abdomen or chest, and an umbilical artery catheter should be used to monitor blood gases and pressure.

Postoperative therapies should last for at least 24 hours, and usually includes a chest tube, mechanical ventilation, and therapies necessary for maintaining ABG's and preventing atelectasis. Improvement in patient status can usually be seen by the third postoperative day, allowing for medications to be decreased and the infant to be slowly weaned from the ventilator

Bronchopulmonary Dysplasia (BPD)

Ironically, the increasingly sophisticated protocols and equipment for treating prematurely born infants have not had an impact on bronchopulmonary dysplasia--in fact, its incidence has actually increased in the last two decades. Despite advances in the study of BPD, its exact etiology remains unknown; however, most cases of BPD occur subsequent to the treatment of RDS. Ironically, the treatment for RDS is considered to be the primary cause of BPD, which involves high pressures and high FIO2s.

The pathophysiology of BPD appears to linked to the following four factors:

• oxygen toxicity
• barotrauma
• presence of a PDA
• fluid overload

Prolonged exposure to high concentrations of oxygen leads to edema and thickening of the alveolar membrane, and ultimately to hemorrhage of the alveolar tissues, which eventually become necrotic. As the lung attempts to heal itself, the new cells are damaged by the same factors, and the disease is perpetuated.

The diagnosis of BPD can be made from a chronic need for oxygen therapy and ventilator support, and confirmed by chest x-rays and laboratory studies. Lab studies include arterial blood gas analysis, which shows evidence of chronic lung disease (ie., hypoxia, hypercarbia and increased bicarbonate levels). As the patient progresses through the disease, the ECG will show a right axis deviation of the heart and possible hypertrophy of the right ventricle.

Pulmonary function studies will show an increased respiratory rate, decreased tidal volumes, and normal minute ventilation. Airway resistance, especially in the lower airways, is increased and the lung compliance is typically decreased as a result of airway and lung parenchymal damage.

Chest x-rays on neonates with a history of high FI02 and positive pressure for several days may show density in all areas with a streaking appearance to the density. The chest x-ray (CXR) characteristics in BPD are generally seen as falling into four stages:

• Stage I: In the first 3 days of life, the CXR is typical of RDS, with bilateral frosted or ground glass appearance.
• Stage II: In days 4-10 of life, the lungs become opaque with granular infiltrates that obscure the cardiac markings.
• Stage III: This occurs during the first 10-20 days of life, and begins showing multiple small cyst formations within the lung fields with a visible cardiac silhouette. There may also be some areas of lung hyperexpansion.
• Stage IV: This occurs following day 28 of life with CXRs showing an increased lung density and the formation of larger, irregular cysts.

Treatment of BPD

There are a variety of approaches to treating BPD, but the most important goal in treatment is to avoid or reduce those factors leading to its development and perpetuation. During mechanical ventilation of the neonate, the goal is to use the lowest possible airway pressures to achieve sufficient gas exchange. If possible, it is recommended to use pressures, rates, and FIO2s that maintain the PaO2 at 45 to 55 mm Hg. Transcutaneous monitors and pulse oximeters are used to maintain these parameters, and to avoid the need for numerous arterial blood gases. Besides preventive measures, treatment includes:

Mechanical ventilation: Use an endotracheal tube small enough to allow a small leak in order to prevent subglottic stenosis in long-term cases.  If treatment is planned for more than 1-2 months, a tracheostomy may be preferable. Adequate humidification of inspired gases is important for avoiding mucus plugging from thickened secretions. Extubate as quickly as tolerated.

Respiratory therapy procedures: Patients generally need chest physical therapy, suctioning, and aerosolized bronchodilators.

Fluid therapy: should be aimed at maintaining adequate hydration and urination. Diuretics such as furosemide are often needed.  If patients lose excess water rapidly, they may be subjected to pneumothoraces if ventilator pressures and rates are not decreased. Long-term use of diuretics calls for maintaining calcium and phosphorus levels.

Right-heart failure: Symptomatic right-sided heart failure may be treated with digoxin in addition to diuretics. BPD patients need frequent blood work, which depletes volumes, so transfusions may be needed to maintain a hematocrit above 40%.

Nutrition: BPD patients may require 120 to 150 cal/kg/day to achieve growth and meet needs of lung repair.

Vitamin E: deficiency tends to increase incidence of oxygen toxicity; administration of vitamin E supplements decreases lung injury caused by administration of oxygen.

Respiratory Distress Syndrome (RDS)

This syndrome, also known as hyaline membrane disease, is one of the most predominant lung problems experienced by neonates. It mainly strikes infants under 35 weeks old, affecting the younger newborns more than older infants. Diagnostic improvements and treatment advances including CPAP, PEEP have significantly cut the RDS mortality rates, but it remains a serious problem.

The etiology of RDS is well understood: a significant deficiency in pulmonary surfactant production. This deficiency decreases lung compliance, increases the infant's work of breathing (WOB), tires an already weakened system and causes atelectasis, decreased alveolar ventilation, hypoperfusion, and even asphyxia. Problems during pregnancy, including maternal diabetes and bleeding prior to labor, can be factors contributing to the incidence of RDS.

Although many factors contribute to the deficiency of surfactant, the main contributor is prematurity of the neonatal pulmonary system. Although surfactant is produced near gestational week 22, it can easily be disrupted by hypoxemia, hypothermia, and acidosis, all of which plague the premature neonate. It is not until the mature surfactant is produced near week 35 that these stressors do not disrupt the production, and the fetal lungs are considered mature.

The symptoms of RDS usually worsen gradually for the first 48-72 hours, followed by stabilization, and a slow recovery period. Stabilization of the disease is often associated with diuresis. The highest incidence of mortality from RDS occurs within the first 72 hours. If death occurs following 72 hours, RDS is usually secondary to complications such as barotraumatic air leaks, intracranial hemorrhages, or infections rather than being due to the lung disease.

The ideal treatment for RDS would obviously be to prevent it from occurring. The administration of glucocorticoids to the mother at least two days prior to delivery has been shown to promote fetal lung and surfactant development. The difficulty in treating RDS is in maintaining adequate alveolar ventilation without inflicting damage on the lungs.  The goal of treatment is to support the patient's respiratory system adequately while minimizing complications--something that is easy to envision, but difficult to accomplish.

Treatment of RDS involves a variety of issues, including:

• maintenance of a patent airway and respiratory acid-base balance
• remaining alert to other systems being affected by decreased ventilation
• providing crucial support until the infant matures

Treatment of RDS also requires adequate hydration, including electrolyte balance. Diuretics, such as furosemide, are used widely in the management of fluid balance in the neonate. Maintenance of thermoregulation is also of vital importance in treating RDS. The use of a pulse oximeter and transcutaneous monitor, along with supportive blood gases, allows for the titration of ventilatory support to meet the neonate's needs, and should be considered mandatory equipment for treating RDS.

Successful management of neonatal RDS patients requires anticipation of potential complications. That anticipation can prevent some complications and allow for rapid treatment of others. Potential complications include:

• Intracranial hemorrhage occurs in 40% of infants weighing less than 1500 g, and the risk increases as positive pressure is initiated.
• Barotraumatic injury leading to pulmonary air leaks, particularly as higher ventilator pressures are needed to maintain adequate ventilation and oxygenation.
• Disseminated intravascular coagulation (DIC), which leads to profuse bleeding throughout the body, is caused by a disruption of coagulation factors.  Neonates with RDS have an increased incidence of DIC.
• Infection is common because of the presence of an endotracheal tube.  Sterile techniques when intubating and suctioning can reduce chances of pulmonary infection.
• Patent ductus arteriosus(PDA) is another common complication of RDS.

Congenital Anomalies

Neonates experience a variety of respiratory and cardiac anomalies. The anomalies that can inflict the fetal respiratory tract during development include:

• an atresia of the upper esophagus with an accompanying fistula between the lower esophagus and trachea
• esophageal atresia without a fistula
• a normal esophagus and trachea with a fistula connecting the two ("H" type)
• lower esophageal atresia with the upper esophagus attaching to the trachea
• both upper and lower esophageal attachments to the trachea
• choanal atresia (tissue blockage at the posterior nasal chamber)
• herniation of the diaphragm
• Pierre-Robin(micrognathia) which causes respiratory distress because of airway occlusion by the tongue

Cardiac defects occur in about 1 percent of all newborn deliveries, and the anomalies include:

• The ductus arteriosus sometimes fails to close following delivery (PDA), causing the shunting of blood away from the lungs and making it difficult for the newborn to maintain oxygenation.
• Defects in the atrial septum also cause blood to shunt from the left atrium to the right.
• Ventricular septal defects allow blood to shunt from the left ventricle to the right.
• The Tetralogy of Fallot is a well-known defect that includes ventricular septal defects, an overriding aorta, hypertrophy of the right ventricle, and pulmonary valve obstruction.
• Transposition of the great vessel occurs when the aorta arises from the right ventricle, and the pulmonary artery arises from the left ventricle.
• Coarctation of the aorta involves a constriction of the aorta, severely impeding the flow of blood.
• With tricuspid atresia, blood flow between the right atrium and ventricle is interrupted and shunting through the foramen ovale occurs.
• Anomalous venous return involves the return of pulmonary blood flow to the right atrium instead of the left.
• In truncus arteriosus, one large vessel acts both as the aorta and pulmonary artery.
• Hypoplastic left-heart syndrome involves outflow from the left ventricle being impeded by coarctation of the aorta and stenosis of the aortic valve.

The respiratory care provided neonates with these anomalies depends on whether the defects increase or decrease the flow of blood to the lungs. Defects that reduce pulmonary flow include tricuspid atresia and the Tetralogy of Fallot.  Increased blood flow is caused by VSD, coarctation of the aorta, subaortic stenosis, PDA, and anomalous venous return.

Lung compliance is generally increased in neonates with decreased pulmonary blood flow. Using high ventilatory pressures can further compromise blood flow and worsen V/Q ratios.  Changing the frequency of ventilation instead of inspiratory pressures can help maintain low mean airway pressure while still meeting the neonate's ventilatory needs. Judicial use of oxygen is frequently required since high PaO2 levels will increase the chance of closure of the PDA, which may be the only source of pulmonary blood flow.

Neonates who are experiencing increased pulmonary blood flow have decreased lung compliance and require higher ventilatory pressures and PEEP in order to maintain adequate V/Q ratios. Higher pulmonary blood pressures are less affected by increases in ventilatory pressures for these neonates.  In all cases, RCPs caring for these patients need to be prepared to adjust ventilator settings to compensate for any changes in compliance.

Neonates can be affected by a wide variety of chest diseases, many of which can be diagnosed through radiography. We will discuss and illustrate some of these in great detail. The rest are simply shown on this Index so you can be aware of their existence. We begin by showing some normal neonatal radiographs, and then move on to illustrate and discuss some of the diseases.

• Normal Neonatal Chest, Inspiratory
• Normal Neonatal Chest, Expiratory
• Normal Neonatal Chest, Prominent Thymus
• Normal Neonatal Chest, Prominent Skin Folds
• Normal Neonatal Chest, Lordotic
• Normal Neonatal Chest, Rotated
• Neonatal Chest with Normally Positioned Tubes and Lines
• Neonatal Chest with Normally Positioned Extracorporeal Membrane Oxygenation (ECMO) Catheters
• Bronchopulmonary Dysplasia (BPD)
• Chylothorax
• Congenital Lobar Emphysema (CLE)
• Cystic Adenomatoid Malformation (CAM)
• Diaphragmatic Hernia (Congenital Diaphragmatic Hernia) (CDH)
• Erythroblastosis Fetalis (Immune Hydrops Fetalis) (Hemolytic Disease of the Newborn)
• Hyaline Membrane Disease (Respiratory Distress Syndrome) (HMD) (RDS)
• Meconium Aspiration Syndrome
• Neuromuscular Paralysis
• Patent Ductus Arteriosus (PDA)
• Persistent Fetal Circulation (PFC)
• Phrenic Nerve Paralysis
• Pneumomediastinum (PMS)
• Pneumonia, Aspiration
• Pneumonia, Chlamydia
• Pneumonia, Neonatal (Group B Streptococcus)
• Pneumopericardium (PPC)
• Pneumothorax (PTX)
• Pneumothorax, Anteromedial (PTX)
• Pneumothorax, Tension (PTX)
• Pulmonary Hypoplasia Due to Fetal Anuria Syndrome
• Pulmonary Hypoplasia Due to Skeletal Dysplasias
• Pulmonary Interstitial Emphysema (PIE)
• Pulmonary Lymphangiectasia
• Wet Lung Disease (Transient Tachypnea of the Newborn) (TTN) (Retained Fetal Lung Liquid)

Normal Neonatal Chest, Inspiratory
Clinical Presentation:
Not applicable.

Etiology/Pathophysiology:
Not applicable.

Pathology:
Not applicable

Imaging Findings:
When interpreting a chest x-ray in the neonate, the entire film should be examined, and not just the chest. Use of the "ABC" approach ensures that all areas of the film are systematically examined.

A - Abdomen - check for: bowel gas pattern suggesting ileus or obstruction, free intraperitoneal air, abnormal calcifications, abdominal situs, and diaphragm position.

B - Bone - check for: fractures, lytic or blastic lesions, and metabolic bone diseases.

C - Chest - check for: midline trachea and mediastinum, abnormal mediastinal and cardiac contours, position of the aortic arch, pleural effusions, pulmonary vascularity, pneumomediastinum, pneumothorax, pneumopericardium, infiltrates, and atelectasis. In older infants and children, a good inspiratory chest film is one in which the relationship of the 6th anterior rib ends intersect the domes of the diaphragm. This may be difficult to evaluate in the neonate where proper positioning is difficult.

Supine inspiratory chest radiograph of a neonate.

DDX:
Not applicable

References:
Haller JO, Slovis TL: Introduction to radiology in clinical pediatrics. Yearbook Medical Publishers (Chicago) 1984.

Normal Neonatal Chest, Expiratory
Clinical Presentation:
Not applicable

Etiology/Pathophysiology:
Unexpanded alveoli cause decreased thoracic volume.

Pathology:
Collapsed alveoli.

Imaging Findings:
In older infants and children, a good inspiratory chest film is one in which the relationship of the 6th anterior rib ends intersect the domes of the diaphragm. This may be difficult to evaluate in the neonate where proper positioning is difficult because the neonate is connected to a number of life support systems. Because the volume of the thorax is decreased in an expiratory film, the following are seen: increased pulmonary opacity, confluent and prominent pulmonary vasculature shadows, and an increase in the size and prominence of the heart and mediastinal contents

DDX:

• Pneumonia
• Cardiomegaly
• Mediastinal mass
• Vascular congestion
• Congestive heart failure
• Pulmonary edema.

Supine inspiratory chest radiograph of a neonate

Supine expiratory chest radiograph (left) and inspiratory chest radiograph (right) in the same neonate.

Normal Neonatal Chest, Prominent Thymus
Clinical Presentation:
Not applicable

Etiology/Pathophysiology:
Not applicable

Pathology:
Not applicable

Imaging Findings:
The thymus is a thin, bilobed organ located in the superior mediastinum that has a variable size and shape. The thymus lies anteriorly in relationship to the heart and great vessels. The relative size of the thymus increases with expiration and decreases with inspiration. The thymus decreases in size during periods of stress, such as during sepsis. Occasionally the thymus may extend inferiorly to the level of the diaphragm. Thymic contour is variable. Because the thymus is a soft organ, overlying ribs may indent it, causing a "wave" sign. The right lobe of the thymus can insinuate into the minor fissure, causing a "sail" sign.

DDX:

• Abnormal cardiac contour
• Mediastinal mass

AP and lateral chest radiographs show a prominent thymus in a neonate.

AP chest radiograph demonstrates a thymic wave sign with the left border of the thymus being indented by overlying ribs.

AP and lateral chest radiographs reveal a thymic sail sign with the right lobe of the thymus insinuating into the minor fissure.

Normal Neonatal Chest, Prominent Skin Folds
Clinical Presentation:
Not applicable

Etiology/Pathophysiology:
Not applicable

Pathology:
Not applicable

Imaging Findings:
Skin folds can be seen as curvilinear densities projecting over the lung bases laterally. They can mimic a pneumothorax, but can be differentiated from pneumothorax because the skin fold margins extend beyond the confines of the lung and pleura.

DDX:

• Pneumothorax

Supine chest radiograph demonstrates a skin fold projecting over the lateral aspect of the left lung base. Notice how it extends beyond the confines of the lung and pleura.

Normal Neonatal Chest, Lordotic
Clinical Presentation:
Not applicable.

Etiology/Pathophysiology:
Not applicable

Pathology:
Not applicable

Imaging Findings:
The anterior arc of a rib on a normally aligned film should be directed downward, below the normally horizontal posterior rib. If the x-ray tube is angled cephalad or if the infant is not lying flat, a lordotic film is obtained. This results in the anterior arc of the rib projecting cephalad above the posterior rib. In severe lordotic distortion the ribs can appear dysplastic, the lung volumes decreased, the cardiac silhouette may have an elevated apex and appear enlarged, and the central portions of the diaphragm may appear elevated, simulating a diaphragmatic hernia. A normal appearing lateral view taken at the same time of the frontal view will confirm the lordotic nature of the frontal film.

DDX:

• Dysplastic ribs
• Cardiomegaly
• Diaphragmatic hernia

Supine lordotic chest radiograph demonstrates falsely appearing dysplastic ribs, low lung volumes, elevated cardiac apex and central diaphragm elevation.

Supine lordotic chest radiograph demonstrates falsely appearing dysplastic ribs, low lung volumes, elevated cardiac apex and central diaphragm elevation.

Normal Neonatal Chest, Rotated
Clinical Presentation:
Not applicable

Etiology/Pathophysiology:
Not applicable

Pathology:
Not applicable

Imaging Findings:
In a properly aligned frontal chest radiograph the distance from the spine to the anterior end of the ribs should be equal, bilaterally, at each level. A rotated film can simulate abnormal mediastinal shift. DDX:

• Abnormal mediastinal shift

Supine rotated chest radiograph simulates mediastinal shift to the right.

Neonatal Chest with Normally Positioned Tubes and Lines
Clinical Presentation:
Not Applicable

Etiology/Pathophysiology:
Not Applicable

Pathology:
Not Applicable

Imaging Findings:
The position of the tubes and lines on a neonatal chest x-ray should be as follows:

Endotracheal tube (ETT) tip: beneath the thoracic inlet and above the carina

Nasogastric tube (NGT) tip: within the stomach

Feeding tube (FT) tip: within the third portion of the duodenum

Central venous line tip placed from subclavian/jugular/antecubital approaches should be within the superior vena cava (SVC)

Central venous line tips placed from a femoral approach should be low in the inferior vena cava (IVC) [below L3] or at the junction of the inferior vena cava and right atrium (RA)

Umbilical artery catheter (UAC) tip: can be either high [between T7 and T11] or low [below L3]. On the lateral film the UAC dips into the pelvis from the umbilicus through one of the paired umbilical arteries and then courses through the internal iliac artery and then into the common iliac artery and aorta. The UAC generally projects over the left side of the spine on the AP film.

Umbilical venous catheter (UVC) tip: at the junction of the right atrium (RA) and the superior vena cava (SVC). On the lateral film the UVC extends cephalad from the umbilicus through the umbilical vein and then courses into the portal vein, across the ductus venosus, and into the inferior vena cava. The UVC generally projects over the right side of the spine on the supine x-ray.

DDX:
Not applicable

Supine chest radiograph showing the endotracheal tube tip projecting between the clavicles and the carina, the nasogastric tube tip projecting over the stomach, the umbilical arterial catheter tip projecting at the level of the T8 vertebral body and the umbilical venous catheter tip projecting at the level of the inferior vena cava / right atrium junction.

Neonatal Chest with Normally Positioned Extracorporeal Membrane Oxygenation
(ECMO) Catheters
Clinical Presentation:
Patient in respiratory failure.

Etiology/Pathophysiology:
ECMO is a technique for pulmonary bypass, used to support patients with severe respiratory and or cardiac failure who are not responsive to conventional therapy. The idea is to allow the lungs time to heal with mechanical ventilation being reduced to minimum levels. Through large bore cannulas unoxygenated blood is removed from the body, passed through the ECMO circuit, which oxygenates the blood, and then reintroduced into the body through a large bore cannula. The most common indications for ECMO are meconium aspiration, congenital diaphragmatic hernia and neonatal pneumonia, which are severe enough to result in pulmonary hypertension and right-to-left shunting.

Pathology:
Not Applicable

Imaging Findings:
The Endotracheal tube (ETT), Nasogastric tube (NGT), Feeding tube (FT), Central venous line, Umbilical arterial catheter (UAC), and Umbilical venous catheter (UVC) tips should be in their normal positions. The tips of the ECMO arterial and venous catheters are often non-opaque, and their exact positions are often difficult to ascertain.

In arterial-venous (AV) ECMO the tip of the arterial catheter should be within the aortic arch and the tip of the venous catheter should be within the right atrium.

In venous-venous (V-V) ECMO the tip of the sole venous catheter should be within the right atrium pointing toward the tricuspid valve

Body wall edema is present because the patient is paralyzed while on the ECMO circuit.

The lungs are opaque due to a combination of fluid in the alveoli, atelectasis, and effusion.

DDX:
Not applicable

Supine chest radiograph demonstrates the tip of the arterial cannula projecting over the aortic arch, the tip of the venous cannula projecting over the right atrium, the tip of the endotracheal tube projecting between the clavicles and the carina and the tip of the nasogastric tube projecting over the stomach.

Supine chest radiograph demonstrates the tip of the venous cannula projecting over the right atrium, the tip of the endotracheal tube projecting at the carina and the tip of the nasogastric tube projecting over the stomach.

Bronchopulmonary Dysplasia (BPD)
Clinical Presentation:
Premature infant who had severe lung disease (usually hyaline membrane disease) and was treated with ventilatory and oxygen therapy.

Etiology/Pathophysiology:
BPD is an end-stage lung disease due primarily to oxygen toxicity from chronic ventilatory support. Other contributing factors include the effects of intermittent positive pressure ventilation, patent ductus arteriosus, and problems with pulmonary toilet. It is most commonly seen as a sequela to hyaline membrane disease, but can also be seen as a sequela to meconium aspiration, persistent fetal circulation, or congenital heart disease.

Pathology:
Initially, generalized capillary leakage and mucosal necrosis is seen. At 1-2 weeks exudative alveolar and airway necrosis occurs along with hyaline membrane formation, mucosal squamous metaplasia and interstitial edema. At 2-3 weeks overdistended alveoli and scarred lung are observed. At several months, large lung cysts and progressive interstitial and alveolar septal fibrosis is seen.

Imaging Findings:
As ventilation techniques change, the classic radiographic stages of BPD are rarely seen. Classically, over time, the imaging findings progress. Initially the typical "ground glass" pattern of hyaline membrane disease is seen. At 1-2 weeks complete opacification of the lungs ("white out") is observed. At 2-3 weeks multiple small cystic lucencies of relatively uniform size and distribution are seen giving the lung a bubbly appearance. By several months of age, lung volume is increased, and the small cystic lucencies have coalesced into larger ones surrounded by fibrotic stranding. In most survivors, clinical and radiologic signs of BPD clear within 2-3 years.

DDX:
Not applicable

Patient "A", supine chest radiograph obtained at 1 week of age reveals a ground glass appearance to the lungs.

Supine chest radiograph obtained in the same patient at 1 month of age shows the development of small cystic lucencies in the lungs.

Supine chest radiograph obtained in the same patient at 2 months of age shows continued development of small cystic lucencies in the lungs.

Patient "B", supine chest radiograph obtained at 1 day of age reveals a ground glass appearance to the lungs.

Supine chest radiograph obtained in the same patient at 10 days of age reveals complete opacification of the lungs.

Supine and lateral chest radiographs obtained in the same patient at 20 days of age show the development of small cystic lucencies in the lungs and increased lung volume

Supine chest radiograph obtained in the same patient at 5 months of age shows the small cystic lucencies to have coalesced into larger lucencies with interspersed fibrotic stranding.

Chylothorax
Clinical Presentation:
Respiratory distress. Fifty percent present in first 24 hours of life, and 70% present within one week of birth.

Etiology/Pathophysiology:
Chylothorax is the accumulation of lymphatic fluid in the pleural space. Abruptly elevated venous pressure during delivery can lead to thoracic duct rupture, which leads to intrapleural accumulation of lymph fluid. Initially the fluid is serous, but turns chylous after milk feedings. This is the most frequent cause of a large pleural effusion in newborn. It is rarely bilateral, and is rarely associated with generalized lymphangiomatosis. It is diagnosed and usually managed successfully via thoracenteses.

Pathology:
Not applicable

Imaging Findings:
Usually unilateral and usually on right side (60%). It is difficult to find the exact site of lymph extravasation with contrast studies.

DDX:

• Erythroblastosis fetalis
• Congestive heart failure
• Cystic adenomatoid malformation
• Urine ascites
• Hemothorax from trauma
• Pneumonia

Supine chest radiograph demonstrates a large right-sided pleural fluid collection

Congenital Lobar Emphysema (CLE)
Clinical Presentation:

Etiology/Pathophysiology:
Congenital overdistension of the lobe can be due to an intrinsic airway obstruction or alveolar overgrowth. Usually only one lobe is involved. A definite etiology is seen in only 50% of patients.

Pathology:
Intrinsic cartilage anomaly or compression by an extrinsic vascular structure.

Imaging Findings:
Progressive overdistension of a lobe that compresses adjacent lobes and causes mediastinal shift to the contralateral side is seen. The overdistended lobe appears oligemic. The most frequently affected lobes are left upper lobe (43%), right middle lobe (32%), and right upper lobe (20%). It is rarely seen in the lower lobes. CLE can begin by having fluid in it, causing it to look like an opaque lung mass, but then the fluid clears via resorption and is replaced by air and it takes on its normal cystic appearance.

DDX:

• Cystic Adenomatoid Malformation - Usually has multiple cysts, unlike congenital lobar emphysema which has only one cyst.
• Diaphragmatic Hernia - Loops of bowel in thorax usually easily identified, can be confirmed via upper GI.
• Bronchial obstruction due to foreign body or mucous plug - Seen in older children
• Extrinsic obstruction from vascular structures
• Congenital lung cyst

Photomicrograph demonstrates lung affected by congenital lobar emphysema

Supine chest radiograph showing a large cystic lucency in the left upper lobe

Supine chest radiograph obtained in the same patient at 7 months of age shows the small cystic lucencies to have coalesced into larger lucencies with interspersed fibrotic stranding. Increased lung volumes are also present.

Supine chest radiograph showing a large cystic lucency in the right middle lobe

Supine chest radiograph showing a large cystic lucency in the right upper lobe.

Series of three chest radiographs showing a congenital lobar emphysema presenting initially as a fluid-filled mass in the right upper lobe (left) that slowly clears of fluid (middle) and finally is overdistended with air (right).

AP and lateral chest radiographs show a cystic mass in the left upper lobe.

Gross photograph of the resected distended left upper lobe in the same patient.

Gross photograph of the sectioned left upper lobe in the same patient shows several large cysts.

Supine chest radiograph at one hour of life shows a solid mass in the left hemithorax causing mediastinal shift to the right, evidenced by the position of the endotracheal tube.

Supine chest radiograph at three hours of life in the same patient shows a multiseptated cystic mass in the left hemithorax causing mediastinal shift to the right.

Cystic Adenomatoid Malformation (CAM)
Clinical Presentation:
The most common presentation is acute respiratory distress in the newborn in the first few hours of life. Alternatively, it can present at several months or several years of age as recurrent pneumonias.

Etiology/Pathophysiology:
Congenital hamartomatous lesion of the lung.

Pathology:
There are 3 subtypes, all of which lack normal bronchial communications:
Type I - multiple large air or fluid filled cysts, usually greater than 2.0 cm in diameter
Type II - variably sized less bulky lesion with smaller cysts
Type III - bulky mass composed of multiple tiny, microscopic cysts resembling bronchi that involves the entire lobe.  Adenomatous hyperplasia with an increase in terminal bronchiolar structures, as well as a polypoid arrangement of mucosal epithelium is seen.

Imaging Findings:
Seen with equal frequency in any lobe. Can rarely be bilateral.
Type I - multiple large air or fluid filled cysts that produce a mediastinal shift and compression of adjacent lung. The most common type is Type I. It may have air fluid levels in its cysts. If infected, it may appear homogeneously opacified.
Type II - variable sized less bulky lesion with smaller cysts with less mediastinal shift and respiratory distress.
Type III - bulky radiographically solid mass composed of multiple tiny cysts that involve the entire lobe.

DDX:

• Diaphragmatic hernia - the abdominal organs are in the chest cavity. 
• CAM - abdominal organs are in a normal position
• Congenital lobar emphysema
• Sequestration
• Post infectious pneumatocele
• Foreign body, trachea - in an older child.

Supine chest radiograph showing a complex cystic mass in the left upper lobe.

Gross photograph of the resected distended left upper lobe in the same patient.

Gross photograph of the sectioned left upper lobe in the same patient shows multiple large cysts.

AP and lateral chest radiographs show a cystic mass in the left upper lobe.

Gross photograph of the resected distended left upper lobe in the same patient.

Gross photograph of the sectioned left upper lobe in the same patient shows several large cysts.

AP and lateral chest radiographs show a cystic mass in the right upper lobe.

Gross photograph of the sectioned right lung in the same patient shows several large cysts

Autopsy gross photograph showing the abnormally enlarged left lung containing multiple large cysts.

AP chest radiograph shows cystic masses in the left and right lungs.

Supine chest radiograph showing a large cystic lucency in the left upper lobe.

Patent Ductus Arteriosus (PDA)
Clinical Presentation:
Neonate who suddenly develops increased oxygen and ventilatory requirements.

Etiology/Pathophysiology:
The ductus arteriosus (DA) usually closes within 24 hours of birth and is obliterated anatomically in 1 - 8 weeks. The DA can reopen in a neonate with hypoxemia or severe pulmonary disease. Initially, postnatal pulmonary hypertension may prevent a left to right shunt from developing, but by a week after birth the physiologic decrease in the patient's pulmonary hypertension may allow a left to right shunt to develop through the PDA.

Pathology:
Not applicable

Imaging Findings:
Initially the CXR shows the underlying pulmonary disease. As the PDA opens, the CXR shows a slight increase in heart size and prominence of central pulmonary vessels which can progress. Pulmonary interstitial edema leading to some obscuring of vascular sharpness is usually seen secondary to left heart failure. An aortogram shows opacification of pulmonary arteries, veins and right atrium as well as the aorta. The definitive diagnosis is via echocardiogram.

DDX:
Not applicable

Supine radiograph on day one of life of a newborn with Hyaline Membrane Disease.

Supine radiograph on day seven of life in the same patient. The lungs are clearing of parenchymal disease.

Supine radiograph on day eight of life in the same patient shows the heart enlarging and the pulmonary vasculature becoming very prominent as the ductus opens.

Supine radiograph on day eleven of life in the same patient with the ductus still open showing pulmonary edema.

Supine radiograph on day one of life of a newborn with Hyaline Membrane Disease.

Supine radiograph on day eight of life in the same patient shows the heart enlarging and the pulmonary vasculature becoming very prominent as the ductus opens.

Supine radiograph on day eleven of life in the same patient with the ductus still open showing pulmonary edema.

Supine radiograph on day one of life of a newborn with Hyaline Membrane Disease.

Supine radiograph on day seven of life in the same patient. The lungs are clearing of parenchymal disease.

Supine radiograph on day eight of life in the same patient shows the heart enlarging and the pulmonary vasculature becoming very prominent as the ductus opens.

Supine radiograph on day eleven of life in the same patient with the ductus still open showing pulmonary edema.

Since every baby born is potentially a case for resuscitation, the skills and knowledge required for proper resuscitation are likely the most important ones possessed by caregivers who work with newborns. As a result, it is strongly recommended that each practitioner complete the Neonatal Resuscitation program offered jointly by the American Academy of Pediatrics and the American Heart Association. The information on resuscitation offered in this CEU is intended as a review, not a comprehensive course on resuscitation.

The purpose of the neonatal resuscitation is to reverse asphyxia before irreparable damage occurs. A successful resuscitation can be divided into three steps, also known as the ABCs of Resuscitation:

The resuscitation procedure begins immediately at the time of birth. The procedure should follow the AHA's most current standards, and as soon as the fetus's head is presented, When the head of the fetus is presented, suctioning of the nose and mouth should begin. Apgar scores are then recorded at both one and five-minute intervals (see Table 3).

Table 3. Resuscitation and Apgar scores.

Medications that are used for resuscitation and that should be readily available in the delivery room include: sodium bicarbonate, isoproterenol, dextrose, epinephrine, calcium, Narcan, dopamine, atropine, and volume expanders. If endotracheal intubation is necessary, it should be performed gently but gently.

It is necessary to have endotracheal tubes available ranging in size from 2.5 to 4.0 mm. The following table provides guidelines for choosing the proper size uncuffed tube:

Effective resuscitation of newborns greatly depends on how well the delivery room and its trained personnel are prepared for handling emergencies.

Croup (Laryngotracheal bronchitis--LTB)
Croup is the name given to a group of inflammatory diseases that primarily affect infants and children. The most common manifestation of croup is laryngotracheitis, which is the result of a viral organism, with approximately 75% of all cases involving parainfluenza virus. The remaining 25% are caused by RSV, influenzae, and mycoplasma pneumonia.

The onset of laryngotracheitis is much like that of a common cold, complete with rhinorrhea, fever, cough, and upper airway congestion. The symptoms available for diagnosis of croup present much more slowly (usually 3-4 days) than is seen in epiglotitis. The patient awakens in the night with a tight, barky cough, and has upper airway stridor on inspiration and expiration. There is also a degree of distress relative to the amount of airway obstruction. A-P x-rays show the narrowing of the airway at the laryngeal level, and the trachea on the x-rays of these patients has been described as being shaped like an hourglass, a pencil, and a steeple.

Laryngotracheobronchitis is the name given to a bacterial superinfection of laryngotracheitis, and it involves the upper airway structures and then progresses to the bronchial airways and structures. Spasmodic croup is an apparent allergic response that results in the sudden onset of a barky cough, shortness of breath, and stridor. The patient with spasmodic croup is typically healthy, with no signs of upper respiratory infection. The distinguishing feature can be seen in some familial history of spasmodic croup.

The treatment of croup varies according to its severity, with mild cases being successfully monitored and treated at home with room humidifiers, hydration, and close observation.  More serious signs such as retractions, increased respiratory rate or nasal flaring are indicators of the need for medical intervention. Since croup has a viral origin, nebulized racemic epinephrine (0.2 to 0.5 ml mixed with 2.5 ml saline) is used because it causes local vasoconstriction of the swollen subglottic tissues and reduces the edema.

If this treatment fails to eliminate the stridor, the patient needs to be hospitalized, with medication nebulizer treatments continued as needed every 1-2 hours to relieve the airway occlusion. Weaning treatments would be applied every 4-6 hours after symptoms subside. While croup is not usually considered life-threatening, any disease that causes swelling of the airway requires close monitoring for signs of worsening condition. Croup patients, therefore, should be monitored for breath sounds, respiratory rate, and the presence of retractions at four-hour intervals until airway swelling subsides.

Epiglottitis

This is an acute inflammatory disease of infants and children that affects not only the epiglottis, but also the surrounding aryepiglottic folds and arytenoid cartilages. While there are several viruses that can cause epiglottitis, it is generally considered to be bacterial in nature with most cases being caused by type b Haemophilus influenzae. Unlike LTB, epiglottitis is a life-threatening disease requiring prompt diagnosis and treatment.

There is a rapid onset with moderate to severe respiratory distress, sore throat, possibly high-pitched stridor with drooling, and difficulty in swallowing. Mortality from epiglottitis is due to blockage of the trachea by edematous, inflamed tissues leading to asphyxiation. The rapid onset of tracheal blockage makes intubation extremely difficult, and the probability of anoxic brain damage becomes more pronounced.

Because of its fulminant nature, the treatment of epiglottitis needs to be handled as an emergency. First priority should be given to the establishment of an airway, either by intubation or tracheostomy, until the swelling has subsided.

Following the establishment of an airway, treatment of the infection involves administration of antibiotics such as chloramphenicol and ampicillin. The epiglottic inflammation usually can be eliminated within 24-36 hours, at which time the patient may be extubated. If paralyzation is necessary, mechanical ventilation may be employed since the otherwise healthy lung requires very low pressures, rates, and FIO2s to maintain ventilation.

Bronchiolitis

Acute bronchiolitis is a viral infection that leads to swelling, inflammation, and constriction in the bronchioles. While its consequences are gravest among infants less than six months, it presents potentially serious problems in children up to three years of age. The respiratory syncytial virus (RSV) is the primary causative agent, causing about 75% of all cases. RSV is highly contagious, and requires extreme care in hand washing and other precautions in order to prevent nosocomial outbreaks.

Clinically, the child with bronchiolitis begins with a typical upper airway infection complete with rhinorrhea, cough, and fever that lasts 2-3 days. With the onset of bronchiolar involvement, the cough worsens, and patients under 3 years of age begin showing signs of small airway obstruction and congestion. These signs include: expiratory wheeze, tachypnea, cyanosis in severe cases, low-grade fever, possible intercostal retractions, bilateral crepitant rales, and hyperinflation. Infants older than 3 years are rarely affected to this degree.

The treatment depends on the age of the infant and severity of symptoms. Infants less than 4 months of age may require hospitalization to control intake of fluids and potential apnea. The level of distress and blood gas values determine the next step of treatment for those infants hospitalized. Older infants or those with less severe distress, apnea, and worsening blood gas values are treated with Ribavirin because of its specific antiviral activity against RSV.

While Ribavirin is not needed by all bronchiolitis patients, infants considered at high-risk benefit greatly from its administration. The medication is delivered via a special small-particle aerosol generator (SPAG), with 12-18 hours a day of treatment for at least 3 and no more than 7 days. RCPs should take extreme care to avoid unintentionally inhaling any of the ribavirin being administered. Other treatments include aerosolized bronchodilators, Alpha-2A-interferon, immunoglobulin A, and RSV immune globulin.

Asthma

Asthma is briefly mentioned here because it is an airway disorder in which a patient's hyperactive airways spasm and constrict, swell, and pour secretions into the lumen in response to a variety of stimuli. It is the most common pediatric lung disease, ant the most frequent cause of hospitalization in the United States. About 1 in every 12 school-age children in the U.S. have asthma, and between 3-5% of the adult population has asthma, with 50% of them acquiring it prior to age 10.

While asthma's exact etiology is not known for sure, there are several factors that precipitate acute attacks, including:

• Allergens (molds, pollens)
• Indoor irritants (dusts,second hand smoke, animal danders)
• Outdoor irritants (air pollution, smoke)
• Viral infections
• Foods
• Aspirin and related medications
• Exercise
• Emotional reactions

Asthma can be classified on the basis of the types of agents which are causative:

• Extrinsic asthma is caused by external agents such as allergies to pollens, dust and various medications.
• Intrinsic asthma is caused by internal allergens such as emotional reactions, exercise, and respiratory tract infections.

Asthmatic episodes are usually characterized by the presence of coughs, wheezing, and dyspnea. They also are associated with decreased alveolar ventilation with severe bronchospasm, and prolonged expiration along with sibilant or sonorous rales.

Asthma attacks generally vary widely in severity from person to person, and episode to episode. Status asthmaticus involves prolonged periods of bronchospasm which are generally not responsive to treatment, and can even become life-threatening.

Asthma treatments require discovering what precipitates the patient's attacks, finding ways for avoiding those factors, and treating the effects of the attacks when they occur. Oxygen is nearly always indicated to treat for hypoxemia, keeping PaO2 above 55 mm Hg. Bronchospasm requires administration of aerosolized medications, usually with bronchodilators. Corticosteroids are effective prophylactics for hyperreactivity. Mechanical ventilation and psychological counseling are often also necessary.

Despite advances in treatment protocols, asthma continues to be a significant cause of respiratory distress among pediatric populations. Drug treatment is mainly focused on preventing the release of inflammatory mediators, reversing bronchospasms, and reducing inflammation in the airways. There continue to be promising research findings focusing on new longer acting medications which are designed to prevent or slow the onset of asthma.

There are a variety of other pulmonary-related problems affecting the neonatal/pediatric populations, and requiring the attention of health care professionals, including:

Aspiration Syndromes: Since infants are seemingly obsessed with putting objects of all sorts into their little mouths, they are at high-risk of having their narrow airways accidentally obstructed. Aspiration of gastric contents is commonly the result of reflux in neonates, and the pathologic result of aspiration is airtrapping and tissue damage secondary to hydrocarbon aspiration. The trapping of air behind the object placed in the infant's mouth causes hyperexpansion of the lung, reduced ventilation, and possible pneumothoraces. The aspiration of hydrocarbons can cause toxic damage to the epithelial lining of the lung.

Treatment of an aspirated foreign object primarily involves actions aimed at removing the object. Chest physiotherapy often helps dislodge the object, while more severe cases may require the use of a bronchoscope to retrieve it.

Smoke Inhalation: Household fires give off many acids and aldehydes in their smoke, and inhalation of these noxious fumes can cause serious toxicities in the blood and tissues. Neonates, whose pulmonary systems are delicate even under normal conditions, are at high-risk for serious injury. Treatment of patients for smoke inhalation and carbon monoxide poisoning involves the immediate application of 100% oxygen, preferably under pressure.

The victims of smoke inhalation are also at risk for development of RDS, making it necessary for the caregivers to closely monitor their status. Antibiotics are given to combat lung infections, nebulized bronchodilators are helpful in maintaining the patency of the airways, and mechanical ventilation is required for comatose patients in order to establish an airway and administer oxygen.

Sudden Infant Death Syndrome (SIDS): While SIDS accounts for the highest number of deaths among infants under a year old, little is known regarding its etiology. The common event in all SIDS deaths is a quiet cessation of breathing during sleep. In fact, the diagnosis of SIDS is not made until an autopsy has been performed. In 1992, the American Academy of Pediatrics recommended that infants not be placed in the prone position to sleep because of association with SIDS.

While no exact relationship between the prone position and SIDS has been established, the reduction in the number of infants sleeping prone has recently been cited as a factor in the decline in SIDS-related deaths. Other factors that have been identified as elevating the risk for SIDS among infants include:

• the use of natural fiber mattresses
• swaddling
• recent illness
• the use of heating in the bedroom
• low birth weight
• prematurity
• a five minute Apgar score under 7
• low maternal age and education level
• multiple births
• maternal smoking ·
• male gender
• black race

Even with a knowledge of the various risk factors, it remains impossible to identify those infants who will die of SIDS, and the incidence remains at about 2 out of every 1,000 births.

Cystic Fibrosis (Mucoviscidosis)

CF is a chronic pulmonary disease that is characterized by tenacious mucus production which causes obstruction, leading to hyperinflation, an increased chest diameter (barrel chest), atelectasis, and infection.

A majority of patients with CF are diagnosed in childhood, but a few are not diagnosed until their mid to late teens. The most reliable diagnostic indicator of CF is the determination of abnormally high sweat chloride levels. Other symptoms associated with CF include: a chronic paroxysmal cough, cyanosis, digital clubbing, atelectasis, hemoptysis, hypoxemia, hypercapnia, and pneumothorax.

Until a cure is found for CF, treatment is primarily aimed at improving long-term survival and improving the quality of life. Protocols call for treating the pulmonary, dietary and psychological aspects of the disease. Therapies available for effectively treating CF patients include administration of mists, aerosols, positive-pressure, physiotherapy, antibiotics, and oxygen.

One of the issues associated with neonatal care is that of access. Not all neonates have equal access to the specialized care units we have discussed. Therefore, we offering for your review:

Taken from: The Future of Children Vol. 5 No. 1 Spring 1995

Abstract

The birth of a high-risk infant is still a relatively rare, not totally predictable event; and the management of high-risk newborns requires highly skilled personnel and sophisticated technology. In the early days of neonatal intensive care, scarce resources led to regionalized systems of neonatal and, later, perinatal services, generally based on voluntary agreements but sometimes reinforced by planning legislation. At present, a vastly increased pool of skilled professionals and technical resources is available in the context of a rapidly changing medical care system characterized by intense competition, coalescence of services under large managed care plans, and substantial cost pressures. The evidence suggests that, in many areas, these forces have led to the dismantling of regional networks; however, the full potential for these changes to hinder or facilitate access to neonatal intensive care remains to be assessed.

The issue of access to special neonatal hospital care emerged only in the twentieth century. Prior to the turn of the century, most births occurred in the home setting, and the limited care repertoire for ill newborns could likewise be provided in that setting. Indeed, it was only in the late 1800s that care of infants less than two years of age was wrested from obstetricians and placed in the domain of pediatricians.1

In an era when at least 10% of all infants died before their first birthday, the fate of an infant born smaller than average was considered particularly grim and beyond the scope of most medical care. Attempts to improve the outcomes of small babies by providing nutrition and warmth were slow in evolving. Not all such efforts were restricted to medical settings, and popularization of the special care for small newborns was accomplished through exhibits at world's fairs and other such currently unlikely settings.2

In the first half of this century, the place of delivery shifted from home to hospital with the growing realization that hospital deliveries could reduce the toll of the complications of birth for women. However, birth in a hospital still offered little survival advantage to the low birth weight infant because not much could be done to manage the respiratory distress or hyaline membrane disease that reflected pulmonary immaturity. Despite this lack of success, transport services were organized in some cities, notably Chicago, to bring sick neonates born at home into the hospital for care.2

After World War II, many of the environmental causes of infant mortality past the first month of life had been controlled or reduced by sanitation, immunizations, and antibiotics. Deaths in the neonatal period began to dominate the rates of infant mortality with low birth weight infants accounting for the majority of these neonatal deaths.3 Thus, the problems of the low birth weight newborn received renewed attention. Until the late 1960s, however, the track record of infant special care units in improving infant outcomes was undistinguished.

Uncritical implementation of interventions poorly grounded in science led to practices that are now seen as useless and silly at best and, in some instances, extremely harmful.4 Not only were individual interventions of limited utility, the whole enterprise of special infant care was questionable as neonatal mortality rates remained unchanged for more than 15 years.3

In the late 1960s, however, basic scientific inquiry into the problems of the newborn began to bear fruit in empirically grounded and efficacious interventions. In particular, critical observations about hyaline membrane disease, the major cause of death among premature infants, led to more effective strategies for managing respiratory problems.5

While the capability to provide effective, exogenous surfactant, the missing chemical, would come much later (see the article by Horbar and Lucey in this journal issue), basic scientific discoveries led to more efficacious mechanical ventilation techniques.6 To underscore the recentness of this success, it should be realized that President Kennedy's son died of hyaline membrane disease in 1963, and the first report of this technique was not published until 1971.6

With the advent of more effective management of neonatal problems in the 1970s, the issue of assuring prompt access to infant care units emerged. At this time, access was severely limited by the small number of centers offering specialized newborn services, the few trained specialists, and the related complexity of care. The solution appeared to be region wide organization of referrals to the small number of available centers.

Regionalization

The concept of the regionalization of health services is not new. The first articulation of this concept in English was presented to the British Parliament in 1920 in the Report of the Consultative Council on Medical and Allied Services, generally referred to as the Dawson Report. Regionalization was considered to include an organized and integrated hierarchical array of medical services, both preventive and curative. At the base was primary care, consisting of those services most frequently used for common and/or simple health problems. Uncommon or more complex problems were referred to secondary or consultative care, and ultimately to tertiary, usually university-based services.7 Explicit in this model was a balance between ready access to care and the efficient deployment of resources at a population level.7

During the World War II era in the United States, several states and other jurisdictions experimented with regional plans.7 The first attempt to incorporate regional planning nationally occurred in the Hill-Burton Act (Public Law 79-725) in 1946 for hospital construction, and the Heart Disease, Cancer and Stroke Amendments of 1965 (Public Law 89-239) which authorized grants to establish Regional Medical Programs (RMPs) to facilitate access to diagnostic and therapeutic advances in the specified conditions. The voluntary nature of the arrangements, the lack of a clear mission, and the categorical nature of the planning effort limited their success.8-10

Subsequent planning legislation--the Comprehensive Health Planning and Public Health Service Amendments of 1966 and the National Health Planning and Resources Development Act of 1974--did not emphasize regionalization as the model of organization.8, 10 However, they did encourage assessment of health services needs at the state and regional levels and the development of plans for increasing access to care. Further, some of the regulatory authority for approving major capital expenditures could be used to curtail unneeded expansion of expensive forms of health care such as duplicative intensive care facilities. In some areas, health professionals forged effective alliances with planning efforts to foster regionalized neonatal care plans.11 However, the general flaws built into planning legislation precluded universal effectiveness.12 In the wake of the failure of national planning efforts, alternative approaches emerged to the organization of health services generally.13

Regionalization and NICU Care

In the absence of strong national planning for all health services, increasing access to neonatal care relied on voluntary efforts led by health professionals, who noted the experience of a few major centers, and on a more general perception that neonatal care was becoming beneficial. In 1973, Schlesinger cited the improved survival documented in certain centers and recommended the models in Wisconsin, the Province of Quebec, and Arizona for moving sick neonates to special care.14 The specific rationale for such organizations was the dearth of physicians and nurses skilled in the new techniques and other support services as delineated carefully by Dwyer15 and endorsed by the American Medical Association House of Delegates.11

The major elements of such regional programs included the structure and function of neonatal intensive care units (NICUs), the formalization of arrangements with referring obstetric units, and transportation systems for sick neonates.11, 14,15 In this early view, regionalization essentially involved interhospital transfer of infants from community hospitals to the medical center and some reverse outreach educational services to the community hospitals on stabilization of the acutely ill newborn. In some instances, health professionals were able to use federal planning legislation to activate these regionalization efforts;11 in other instances, more informal arrangements among groups of hospitals occurred.

The regionalization of neonatal care represented only a small segment of regionalized services as envisioned by the Dawson model and, moreover, provided little threat to existing care arrangements. The mother remained in the care of her obstetrician. The newborn usually had no primary pediatrician. The limited number of highly trained NICU physicians and nurses precluded rapid increase in NICUs at the community level. Even among insured couples, coverage for expensive care for sick newborns might not be adequate, and thus, the transfer of such care provided little financial threat to the referring hospitals.

Evidence of the effectiveness of such regional models of neonatal intensive care is sparse. Arguments for its importance rest largely on decreases in neonatal mortality rates after the introduction of NICU care in geographically defined regions where it was not previously available14,16,17 and improvements in survival for infants with hyaline membrane disease among those managed through a formal referral system as compared with those managed through other arrangements.18

Regionalization and Perinatal Care

Evidence also accumulated that outcomes could be improved further by earlier identification of high-risk pregnancies and referral to tertiary perinatal centers before delivery. For example, very low birth weight infants born in hospitals with NICUs did better than those not born in such centers,19-21 even accounting for the availability of neonatal transfer.22 Additional evidence was seen in the more favorable outcomes of infants whose mothers were transported to perinatal centers before delivery compared with infants transported after birth.23-25

In response, the notion of regionalized care was expanded to include the prenatal period. Such regional systems would address the needs of all pregnant women in a population through systematic risk assessment and referral to the appropriate consultants when problems emerged. These systems would also assure community providers ready access to consultation, special laboratory facilities, and ongoing education. The intent was to improve perinatal care at all levels for a defined region based on the experience of several programs.26-28 The evolving concept of perinatal regionalization thereby broadened to include a large array of services (see Table 1) with a fully integrated system of consultation, referral, and transport, as advocated in several commentaries and editorials,11,27,29 and codified by a committee of the major pediatric, obstetric, and nursing organizations in a March of Dimes-sponsored publication entitled Toward Improving the Outcome of Pregnancy.30 This manifesto specified in detail the services required for level 1, 2, or 3 perinatal care as well as guidelines for delivery volume and/or geographic necessity for each level of care.

In contrast to the relatively sparse data on neonatal systems, more substantial evidence on the effectiveness of these comprehensive perinatal systems is available. To assess the effect of regionalization on previously less-well-organized areas, the Robert Wood Johnson Foundation conducted an eight-site, five-year demonstration program. The evaluation of that program revealed that regionalization occurred and that the rapidity of regionalization was closely correlated with the rate of decline in neonatal mortality, as measured by the proportion of low and very low birth weight infants born in tertiary centers. The sharp decreases in neonatal mortality were not offset by increases in infant morbidity. Unfortunately, the national movement toward regionalization had become so widespread that a specific effect of the demonstration program could not be detected when compared with the progress in similar control regions not funded by the foundation.31 Another Robert Wood Johnson Foundation program, aimed at improving care in 10 rural sites by fostering improved linkages with tertiary centers, also resulted in improved neonatal outcomes.32 In addition, reports accumulated on the success of individual programs and on specific aspects of regionalized programs.33-39

While clearly improving infant outcomes, the regional coordination of services and transfer of patients was not without adverse effects. These included:

• initial separation of the newborn from its family and the resulting anxiety and, in some instances, grief experienced by one or both parents,40,41
• prolonged neonatal hospital stays and the financial and emotional costs associated with frequent visiting,42
• disruption of established patient-physician relationships,43
• loss of local medical services,44
• disruption of continuity ofcare and parent-provider relationships through retrotransport to community hospitals to alleviate NICU crowding45 even though this is a cost-effective strategy for convalescent care.46,47

Despite these disadvantages, regionalization occurred, reinforced by the general impetus to central planning and scarcity of resources. Other reinforcements included more reliable funding through both public and private payers48 and, perhaps, the malpractice crises which made practitioners more cautious in the management of high-risk obstetric patients.49

The Current Era

In the early 1980s, changes in the availability of resources and cost containment strategies began to place stresses on regional programs. The current status is summarized in the following section.

Although state planning agencies and health departments could provide reinforcement to regionalization through funding and certification of need authorization, regional perinatal care could best be described as voluntary associations among hospitals and providers. In the 1970s these voluntary associations were reinforced by the impetus toward centralized planning noted above, the scarcity of skilled perinatal and neonatal personnel, the complexity and expense of well-equipped units, more reliable funding through public and private payers, and perhaps also the malpractice crisis. Beginning in the early 1980s, even as the evidence of effectiveness began to accumulate, other forces placed stress on regional programs.

Expansion of Perinatal and Neonatal Intensive Care

During the 1970s and 1980s, both the number of specialized physicians, or neonatologists, and the number of NICUs and NICU beds increased dramatically. This change in availability of NICU care has received little empirical examination, in part because of the difficulty of tracking the information. This dearth of information has been remedied with a directory of neonatologists prepared under the sponsorship of the Section on Perinatal Pediatrics of the American Academy of Pediatrics.

While data are still being accumulated, preliminary information reveals that, overall, about 3,000 neonatologists are active in the United States, or about 7.4 per 10,000 live births. In 1994, about 500 hospitals reported having a NICU for a total of about 12,000 NICU beds, or 3 per 1,000 live births.50 The number of neonatologists per 10,000 live births is at least twice that of several other industrialized countries51 and exceeds the upper bound estimate calculated by the American Academy of Pediatrics in 1985 of 5 per 10,000 live births.52 However, the rate of very low weight births has been increasing, and the birth weight at which survival is now more routine is decreasing. These trends indicate that the need for neonatologists may exceed 1985 estimates. To what extent the current status represents an excess needs to be examined further.

The availability of NICU beds is also difficult to assess. The reported number of about 3 per 1,000 live births exceeds the estimated need of one NICU bed per 1,000 live births. However, it is within the overall estimates of 5 to 6 total beds per 1,000 live births, which includes intermediate or convalescent beds.53-55 While these preliminary data are subject to some error and the shift in care patterns may indicate some increased need for resources over prior estimates, NICU care is no longer a scarce resource nationally, although access to care may not be uniform for all groups because of geographic factors and some of the financing issues discussed below.

Financing Neonatal Care

The growth of NICUs has occurred during an era when the financing of obstetric and newborn services was itself in rapid flux. Financing has become a dominant factor in determining both availability and access because of the exceedingly high cost of neonatal intensive care and the impact of inadequate reimbursement on regional neonatal services

Private Insurance

Early studies of NICU care56 indicated incomplete private insurance coverage with a large pool of uncompensated care. More recent data reveal that, even for privately insured patients, most plans do not begin to cover the full costs of maternity care, and many women are uninsured. For those eligible, services may be covered by Medicaid, but reimbursement is often at considerably lower rates than are typical of private insurance.48 Thus, a substantial portion of NICU care remains uncompensated, and the costs of this care are shifted to other payers.48,57 Strong evidence indicates that uninsured newborns receive less care than those privately insured or on Medicaid, even when the numbers are adjusted for illness severity.58

Absent national health care reform, the current changes in private insurance coverage suggest that the problems of financing neonatal intensive care are likely to grow worse. Employed young couples are most likely to lack employer-based insurance because they have entry level jobs or work in small businesses. Even if the employed parent is covered, dependency coverage is decreasing. The result is that private insurance coverage for children has declined substantially.59 What coverage does exist may be exhausted by the catastrophic costs of NICU admission, as such care has been demonstrated to account for a substantial number of all episodes of catastrophic illness.60,61

Public Financing

Some portion of this financial burden may be offset by recent major expansions of public financing of perinatal care. First, the high cost of neonatal intensive care can result in rapid spend-down to medically needy status for Medicaid coverage, even for the insured population. Second, many states have expanded medical assistance coverage of pregnant women as a means of ensuring adequate prenatal care, and this coverage is automatically extended to their newborns.62 Likewise, recent changes in criteria for supplemental Social Security may provide support--both income and access to Medicaid--for children deemed disabled.63

Publicly financed health care is a fragile and vulnerable source of support, however. Eligibility requirements are based on income and vary by state. With the recent downturn in the economy, many states face severe economic constraints in providing for Medicaid patients. While the majority of Medicaid dollars go to support nursing home care for the elderly, few politicians are willing to face such a powerful lobby. Thus, attempts to reduce Medicaid costs focus on the smaller fraction going to poor women and children. Because neonatal intensive care represents a significant portion of hospital care for children, it has become a target for reducing Medicaid costs, as indicated by the recent Oregon waiver proposal that set birth weight limits for initiating intensive care.

Prospective Payment Systems

The shift to case-based reimbursement under Medicare does not directly affect newborns. However, the adoption of diagnosis-related groups (DRGs) by Medical Assistance (Medicaid) in several states has produced serious problems because the original neonatal DRGs were seriously flawed.64-67 Specifically, the original DRGs reflected only the experience of community-based hospitals and the limited data available in the small sample of hospitals used to derive the DRGs.

There are several problems related to the use of DRGs as the basis for reimbursement of NICU care. First, these DRGs performed poorly in explaining resource use because they lacked birth weight, a key predictor of cost. Second, for highly regionalized care such as NICUs, all outliers were concentrated in a few centers. A third serious flaw was the disincentive for back-transfer. While DRGs were adopted in only a few states, the threat of their use by other states and by private insurers sparked great concern among regional centers. This concern prompted the National Association of Children's Hospitals and Related Institutions (NACHRI) to develop pediatric modified DRGs in which the original seven DRGs were expanded to 46 categories, determined by birth weight and need for surgery and/or mechanical ventilation.68 Although the modified DRGs are an enormous improvement, they have not been broadly adopted by payers. Thus, the impact of case-based reimbursement in its more refined form remains to be determined. Other pressures on the arrangements of perinatal services come from the massive changes in organization that are being seen generally in health care. In particular, these changes involve rapid emergence of managed care and intense competition among providers to secure plan contracts.

Organization of Services

Managed Care

Managed care plans have expanded from the more traditional health maintenance organization (HMO) to include a broader array of service and financial arrangements. Significant elements of managed care plans include arrangements with selected care providers, utilization review, and strong financial incentives for members to use selected providers and to follow designated procedures.69 The major strategy to reduce costs is a reduction in hospital admissions, as well as more prudent purchasing of lower cost or discounted services. The extent to which such plans do reduce costs, the implications of more active management of clinical care, and the relative advantages of various models remain open to question.70

Neonatal intensive care admissions are virtually all emergent so it may be difficult to reduce them. Where managed care may influence access to NICU services is its ability to channel obstetric and newborn patients to specific facilities and providers and, thereby, to have a major effect on reinforcing or weakening regionalized care (see Table 2). A major concern is that managed care organizations will direct all obstetric care into lower level hospitals because of their significantly lower operating cost, despite the strong evidence of better outcomes in higher level facilities. The increasing market penetration by managed care in many regions indicates that this will be an important factor in coming decades.

Competition

In view of the changes described above, both insurance plans and providers are now competing heavily for patients. Competition is particularly keen for obstetric patients because hospitals recognize that women make most of the medical care decisions and that a positive birth experience may ensure family loyalty to a particular institution. While competition may drive individual hospitals to upgrade their services, it has also spurred the rapid proliferation of NICUs, particularly for suburban areas where competition for well-insured patients is strongest. These NICUs are established to provide a sense of security in the availability of intensive care if needed but are often small and, therefore, inherently inefficient, although they may operate profitably by attracting a profitable case mix of insured patients. The loss of these patients has further eroded the financial viability of regional centers57 and has begun to redefine them as level 4, or "quaternary," centers, receiving a declining number of only the smallest infants or those with complex multispecialty needs. This reduced and very expensive population base cannot support the overhead costs of sustaining regional organization including outreach education, consultation, transport, and infant follow-up. The competition among hospitals has also curtailed the cooperative arrangements underlying regional perinatal organizations. Furthermore, the dispersion of patients into many small NICUs will significantly hinder research, medical education, and outcomes evaluation (see Table 3). While competition has also tended to sustain inefficient small obstetric services in suburban markets to draw the loyalty of women clients, in both inner cities and rural areas, competition may lead to closure of community obstetric services, thereby reducing access to prenatal care.

Deregionalization

The aggregate effect of these changes in manpower, facilities, technology diffusion, financing, competition, and health care organization has been a cessation or even reversal of the general trend toward regionalization. This deregionalization has been noted in several studies. The National Perinatal Information Center studied six regions using in-depth interviews with hospital executives, neonatologists, and obstetricians. While there was great variation from region to region, most agreed there was a general deterioration in perinatal regionalization, that competition had replaced cooperation, and that traditional levels of care were blurring as all facilities escalated the level of care provided.

These trends also occurred in many community hospitals where the volume of patients was inadequate to maintain professional skills or provide a cost-effective revenue base.71 another study of the Hartford region identified similar concerns, centered on the balance between competition and cooperation. It noted the potential for dispersing the NICU population into smaller competing NICUs versus a single unresponsive monopoly on regionalized services.72 Similar concerns have been voiced by others who have called for negotiated cooperation agreements in place of traditional regionalization schemes.73 Interestingly, deregionalization following National Health Service reforms in the United Kingdom has also produced adverse effects on perinatal care.74 These issues provoked the Committee on Perinatal Health to reconvene to formulate an agenda for regionalization in the 1990s and beyond.75 They recommended improvements in health education, prenatal care, system organization, access to inpatient and specialty services, documentation and evaluation, and adequate financing of perinatal care.

The reality of perinatal regionalization is that market forces are forcing hospital closures, consolidations, and mergers. Patients are increasingly channeled by payer-provider negotiations rather than historical regional designations. However, the long-term outcome may not be bad. Consolidated obstetric services are inherently safer and more efficient. Regionalized perinatal care is also inherently cost effective, utilizing graded levels of care according to need. The development of highly integrated vertical networks that eliminate redundant services may actually strengthen the regionalization of perinatal care. This level of integration, however, may not be achieved in many regions, and even where it is, the transition may cause serious dislocations for perinatal care.

Conclusion and Recommendations

The rapid proliferation of trained professionals and the diffusion of technology in a context of major organizational change and new financial pressures would appear to signal the unraveling of early regionalized models of delivering care. Little empirical evidence exists, however, to estimate the effect of these changes. Moreover, changes in access to NICU care in response to the various pressures noted above are unlikely to be uniform and will depend on demographics, geography, malpractice experience, and state regulatory environments. Several courses of action are recommended to assure access to appropriate levels of neonatal care.

Access and Equity

The first of these recommendations is for ongoing surveillance of access to NICU care through vital statistics and other record systems. Abundant evidence indicates that higher levels of care are associated with better outcomes.21,31,32,75 Shifts in births to lower level hospitals for cost reasons may have a significant impact on overall outcomes. Competition will close not only inefficient services but also poorly reimbursed ones. Hospitals caring for poor and underinsured populations will be stressed regardless of quality or efficiency. To the extent efficiency is also achieved through consolidation of existing facilities, the burden of seeking alternative services may be carried disproportionately by disadvantaged and/or rural communities. Thus, relevant public health and professional organizations should be routinely assessing births and birth-weight-adjusted mortality by hospital of delivery for the newborn population as a whole and for high- risk subgroups.

Quality of Care

The second recommendation is for rapid expansion in rigorous examination of the quality of NICU care. Even among designated tertiary hospitals, there have been instances where quality of NICU care has been questioned,76,77 and such quality differences in the past have been more likely to occur in areas serving largely minority populations.77 However, variations also occur in units where the level of practice is assumed to be high,78,79 suggesting that some portion of the variation may reflect practice patterns, as well as differences in the population that is served. The proliferation of small NICUs has raised additional questions about the extent to which such small units will have sufficient numbers of patients to maintain the skills of providers.77-80

To preserve quality in this rapidly changing health care environment, it is necessary first to measure it. This seemingly simple task has not yet been achieved for perinatal care. It is vital to develop comprehensive case mix adjustment tools77,81 and to pursue comparative outcomes research.78,79 Incorporated into these quality assessments must be explicit recognition of costs. The third-party payers will demand quality at the lowest price.

Adequate Reimbursement

Perinatal and neonatal services remain underfinanced even among populations with insurance. In the current competitive atmosphere, hospitals will be unable to subsidize these services. Besides the support of direct medical services, however, adequate support of other services formerly provided in regionalized networks is necessary. Particularly vulnerable are consultation services, outreach education, assessment of outcomes among discharged children, development and maintenance of information systems, and surveillance and planning activities. Adequate reimbursement must also reflect the increased costs of caring for many high-risk newborns, especially among the disadvantaged population.

Training

Finally, some equilibrium between the needs of training programs and the need for skilled neonatal professionals must be achieved. This equilibrium will require alterations in the size, number, and staffing of teaching hospitals. In addition, it may also require the provision of new training experiences to assure the ongoing assessment of neonatal practices as noted above.

In the absence of significant reductions in prematurity rates, NICU care will continue to be needed. Moreover, even with reduced rates of prematurity, such units are also critical to the well-being of children with malformations and acute complications of the newborn period. The current changes in neonatal care provide both exciting new opportunities to maximize its utility and grave challenges in assuring equity.

From The Future of Children Vol. 5 No. 1 Spring 1995

Abstract

The development and dissemination of neonatal intensive care technology has been associated with improved survival for critically ill newborn infants, particularly those with birth weights of less than 1,500 grams (3 pounds, 5 ounces). Despite these advances, there are concerns about the long-term health status of surviving infants and the costs of their initial and subsequent care. In this article, the authors review current evidence for the effectiveness of neonatal intensive care and discuss several approaches to evaluating neonatal intensive care technology. They discuss a four-step process originally proposed by Roper for assessing and improving neonatal intensive care practices which includes
(1) monitoring of practices, outcomes, and costs;
(2) analysis of variation in practices, outcomes, and costs;
(3) assessment of the efficacy of individual interventions, and
(4) feedback and education to alter clinical behavior. The authors conclude that organized networks of neonatal intensive care units can play a crucial role in this process.

The development and dissemination of neonatal intensive care technology has been associated with increasing survival for low birth weight infants, particularly those weighing less than 1,500 grams (3 pounds, 5 ounces) at birth. Other factors have also contributed to this trend, including the regionalization of perinatal care, the emergence of subspecialists in maternal-fetal medicine, neonatal medicine, and pediatric surgery, and diagnostic and therapeutic advances in high-risk obstetrics. Despite improvements in survival, however, there are continuing concerns about the high costs of neonatal intensive care and the quality of life of survivors. In this section discusses the evaluation of neonatal intensive care technologies and emphasize the role that organized networks of neonatal intensive care units can play in evaluating and continuously improving the effectiveness and efficiency of neonatal intensive care.

Background

Neonatal intensive care for critically ill newborns was introduced in the late 1960s when methods for providing assisted ventilation to small infants were first developed. Because neonatal intensive care units required specialized personnel and facilities beyond those available at most hospitals, national efforts were made to regionalize their location.1 This strategy was successful in increasing the availability of neonatal intensive care and resulted in improved survival rates for low birth weight infants.2 By 1976 there were more than 125 neonatal intensive care units in North America, primarily at major university medical centers.3 There are now reported to be more than 700 such units, including both level 2 and level 3 nurseries.4 Clearly, there has been a dramatic increase in the number of neonatal intensive care units. As a result, the majority of newborns receiving intensive care are now treated in smaller non-university hospitals. This trend, called "deregionalization," is discussed in detail in the article by McCormick and Richardson in this journal issue.

Neonatal-perinatal medicine became a board-certified subspecialty of pediatrics in 1975. The training period consists of three years of general pediatric residency plus three years of neonatology fellowship in one of 106 approved training programs. All neonatologists who entered training programs after 1989 are required to take a recertification examination in general pediatrics and neonatology every seven years. A survey of perinatal centers performed in 1983 estimated that there were 1,509 physicians practicing neonatology in the United States.5 In 1993, the Accreditation Council for Graduate Medication Education listed 528 physicians in neonatal training programs and 2,498 board-certified neonatologists. There is no official estimate of the total number of pediatricians currently practicing neonatology, some of whom are not board certified, but a database maintained by Abbott Laboratories lists 3,740 individuals interested in neonatology. A survey is being carried out by the American Academy of Pediatrics to gather more accurate data on the number of neonatologists who are actually in practice. These data are important because questions have been raised as to whether there are currently too many neonatologists in the United States as compared with the number in other countries.6

The growth in the numbers of neonatologists and neonatal intensive care units has been accompanied by rapid expansion in the range and complexity of medical, surgical, and diagnostic services provided to critically ill newborns. In 1981, Sinclair and colleagues enumerated some of the specific preventive and therapeutic maneuvers included in neonatal intensive care programs.7 They concluded that, although the efficacy of some specific intensive care maneuvers had been validated in randomized controlled trials, the overall effectiveness of neonatal intensive care programs required further evaluation with rigorous scientific methods. Since that article was written, a large number of new intensive care techniques have been introduced. Also, neonatal intensive care has been provided to progressively smaller and less mature infants so that infants at 23 weeks gestational age and 400 grams (14 ounces) birth weight are now receiving intensive care at many institutions.

A representative list of the technologies and procedures that are included in current neonatal intensive care programs is shown in Box 1. This list is not exhaustive but does give an indication of the range of diagnostic and therapeutic measures which are used in modern neonatal intensive care. Most of the specific measures have still not been subjected to rigorous scientific testing, and there is considerable variation among neonatologists in how these techniques are used.

BOX 1

Neonatal Intensive Care Technologies and Procedures

1. Environmental Control incubators
   radiant warmers
   servo-controlled thermoregulation
   heat shields
   plastic wrap
   humidification
   control of noise and light

2. Vascular Access umbilical artery/vein catheters
   peripheral artery/vein catheters
   central venous catheters
   infusion and syringe pumps

3. Physiologic Monitoring temperature
   cardiorespiratory
   electrocardiogram
   thoracic impedance
   apnea/bradycardia alarms
   trend monitors
   systemic blood pressure
   oscillometric method
   indwelling artery catheter
   central venous pressure
   oxygenation/ventilation
   arterial blood sampling
   capillary blood sampling
   pulse oximetry
   transcutaneous PO2,CO2
   pulmonary function testing

4. Laboratory Testing micro sampling methods
   bedside glucose testing
   routine chemistry, hematology, serology
   microbiology
   pulmonary maturity
   genetic analysis
   metabolic screening

5. Diagnostic Imaging radiography
   ultrasonography
   Doppler echocardiography
   CT-scanning
   MRI scanning
   nuclear medicine scanning

6. Nutritional Support parenteral nutrition
   enteral feeding techniques
   special formula
   breast milk supplements
   vitamins
   minerals
   trace elements

7. Blood Products red blood cells
   white blood cells
   platelets
   plasma
   cryoprecipitate
   coagulation factors
   irradiation of blood products

8. Respiratory Support supplemental oxygen
   continuous positive airway pressure
   chest physiotherapy
   conventional ventilation
   high-frequency ventilation
   surfactant
   extracorporeal membrane oxygenation

9. Delivery Room Resuscitation

10. Neonatal Pharmacotherapeutics diuretics
    xanthines
    steroids
    indomethacin
    antimicrobials
    heparin
    vasopressors
    sedatives/analgesics

11. Phototherapy

12. Neonatal Surgery/Anesthesia

13. Psychosocial Interventions unlimited parental visiting
    parental involvement in care
    skin-to-skin contact
    infant stimulation
    reducing noxious stimuli
    nonnutritive sucking

14. Neonatal Transport skilled transport teams
    air/ground transport

15. Other Diagnostic Testing electroencephalogram
    evoked response audiometry

The Impact of Neonatal Intensive Care

Mortality

Several lines of evidence support the hypothesis that neonatal intensive care has resulted in decreased mortality. First, there have been marked declines in neonatal and infant mortality rates coincident with the introduction and refinement of neonatal intensive care.8-10 This decline has been almost entirely due to improvements in the survival of very low birth weight infants rather than a decrease in the number of very low birth weight infants.10 Between 1965 and 1975, neonatal mortality rates decreased by 35%. Because other factors known to affect survival at a given birth weight did not change during this period, Lee and colleagues concluded that improvements in perinatal medical care were responsible.8 A more recent drop in infant mortality has been associated with a specific neonatal intensive care practice.11- 13 Between 1989 and 1990, the infant mortality rate dropped by 6% (from 9.7 to 9.1/1,000 births). Kleinman estimated that as much as one-half of this change may have been due to the introduction of surfactant therapy, which reduces serious lung disease, into neonatal intensive care.11

The second line of evidence regarding the effect of neonatal intensive care on infant mortality is the observation showing that low birth weight infants born in hospitals with tertiary level neonatal intensive care units have lower mortality rates than infants born in hospitals without such units.14

Paneth and colleagues found that preterm low birth weight infants born in New York City were at a 24% lower risk of death if they were born at a hospital providing tertiary level neonatal intensive care.15 Similar findings have been reported for total population cohorts of infants from New York City,16 several states in the United States,17,18 and the Netherlands.19

Finally, increased access to neonatal intensive care--either because a new neonatal intensive care unit is opened20 or because regionalized neonatal services are instituted2--has been associated with decreases in neonatal mortality. In Hamilton-Wentworth County, Ontario, mortality for very low birth weight infants dropped by nearly 17% after a regional neonatal intensive care unit was opened in 1973.20

All of these findings suggest that neonatal intensive care has played a role in improving survival for low birth weight infants. Other factors, including regionalized transport systems for high-risk women and infants, advances in obstetrical care and neonatal surgery, and improved training in maternal-fetal and neonatal medicine, have also had an effect.

Long-Term Outcomes

The impact of neonatal intensive care on long-term outcomes for low birth weight infants is more complex. (See the article by Hack, Klein, and Taylor in this journal issue for a detailed discussion of long-term developmental outcomes.) In a case study published in 1987, the Office of Technology Assessment concluded that "neonatal intensive care has contributed to improved long-term developmental outcomes for premature infants. The great decline in mortality among all subgroups of very low birth weight infants over the past 10 years, however, means there are now larger absolute numbers of both seriously handicapped and normal survivors."21 A recent review of population- based registries of cerebral palsy found that one unavoidable side effect of the increasing success of neonatal intensive care is a moderate rise in the childhood prevalence of cerebral palsy.22 In addition to neurobehavioral disabilities and poor school performance, children born at very low birth weight who survive are also at higher risk for a broad range of other adverse health outcomes at school age.23-27 As a result, survivors of neonatal intensive care have significantly increased medical care costs throughout childhood, which may pose a substantial financial burden to their families.28,29 These long-term consequences must be carefully balanced against the gains in survival in any evaluation of neonatal intensive care.

Extreme Immaturity

There is considerable uncertainty about the effectiveness of neonatal intensive care for extremely immature infants. Difficult questions arise as attempts are made to identify lower limits of birth weight and gestational age below which neonatal intensive care is either futile or does more harm than good for the individual infant, its family, and society. (For a discussion of these ethical issues, see the article by Tyson in this journal issue.) Based on infants born between 1982 and 1988, Hack and Fanaroff concluded that, with few exceptions, survival is limited to infants with birth weights of 600 grams (1 pound, 5 ounces) or more or to those whose gestational age is at least 24 weeks.30 They found no improvement in the outcomes for such infants between 1982 and 1988. Allen and colleagues, in a study of infants of 22 to 25 weeks gestation born from 1988 to 1991, found that no infants born at 22 weeks gestation survived as compared with 15% of infants born at 23 weeks, 56% born at 24 weeks, and 79% born at 25 weeks gestation.31 Only 2% of infants born at 23 weeks gestation survived without severe brain injury as compared with 21% of those born at 24 weeks and 69% of those born at 25 weeks of gestation. Allen and colleagues concluded that aggressive resuscitation and intensive care treatment are indicated for infants born at 25 weeks gestation, but not for those born at 22 weeks gestation. They recommend that, for infants born at 23 or 24 weeks gestation, discussions involving parents, health care providers, and society at large are required. In contrast, in 1991 the law in Japan was amended to lower the limit of viability from 24 to 22 weeks because of increasing numbers of survivors with gestational ages below 24 weeks.32,33 Questions about which infants should receive intensive care will become even more pressing as scarce health care resources are carefully scrutinized and reallocated.

Clinical Evaluation Science

Clinical evaluation science provides a natural framework for evaluating neonatal intensive care. This emerging field, otherwise known as outcome research or medical effectiveness research, uses a variety of analytic techniques to understand the relationships between the structure, process, outcomes, and costs of medical care.34 Clinical evaluation science attempts to identify and explain the variations in the practice and outcomes which have been observed to occur among regions, among hospitals, and among physicians. Ultimately, the goal is to change inappropriate practice patterns by disseminating information to physicians and patients. The Patient Outcome Research Teams sponsored by the Agency for Health Care Policy and Research are a major example of this strategy in action.35

Roper and colleagues have described a four-step process for assessing the effectiveness of medical care and improving clinical practice which illustrates many of the techniques of clinical evaluation science.36 This process includes (1) monitoring of practice, outcomes, and costs, (2) analysis of variations in practice, outcomes, and costs (3) assessment of the efficacy of individual interventions, and (4) feedback and education to alter behavior. We will discuss each of these steps as they apply to neonatal intensive care and describe the role that neonatal research networks can play in this process.

Monitoring

Monitoring the medical interventions, outcomes, and costs for large populations requires accurate and timely data collected using uniform definitions. A major barrier to monitoring neonatal intensive care on a large scale is the lack of adequate data sources. In the Medicare population, for which considerable effectiveness research has been done, monitoring is facilitated by the availability of a universal database of Medicare claims.

No similar national data sources are available for monitoring outcomes, interventions, and costs for perinatal patients. Studies in perinatal patients have, therefore, relied on vital statistics, linked birth and death certificates, state Medicaid files, medical record reviews, and hospital discharge abstracts.

Unfortunately, these sources often lack the clinical detail necessary to properly monitor neonatal intensive care. A pioneering example of how existing data sources can be used for monitoring neonatal care is the system developed by Williams, which uses linked birth and death certificates, for the state of California.37

The March of Dimes Birth Defects Foundation, in its publication, Toward Improving the Outcome of Pregnancy: The 90s and Beyond, recognizes that improved data sources will be crucial to the evaluation of neonatal intensive care.38 Comprehensive perinatal data systems at the state and regional level, which include vital statistics and information about clinical practice, will be needed. As integrated delivery systems are developed under health reform, there will be opportunities to create modern information systems which can be used to assess the effectiveness and costs of neonatal intensive care.

Neonatal Network Data Bases

Currently, databases maintained by neonatal research networks provide an important source of data for monitoring the practice and outcomes of neonatal intensive care. There are now at least eight such networks collecting information on infants receiving neonatal intensive care (see Box 2).

BOX 2

Neonatal Research Networks

1. BAPM Perinatal Clinical Trials Group
   British Association of Perinatal Medicine
   c/o Secretary BAPM Perinatal Trials Service
   NEPU, Radcliffe Infirmary
   Oxford, OX2 6HE UK

2. Canadian Perinatal Clinical Trials Network *
W. Fraser, MDBR
L'Hopital St. Francois d'Assise
10 rue de L'Espinay Quebec, G1L 3L5 CANADA

3. International Neonatal Network* W. O. Tarnow-Mordi, MD
   Department of Child Health
   Ninewells Hospital and Medical School
   Dundee, DD1 9SY UK

4. National Perinatal Information Center* David E. Gagnon, MPH
   Executive Director
   One State Street, Suite 102
   Providence, RI 02908

5. NICHD Neonatal Research Network Linda Wright, MD
   NICHD/NIH
   Room 4B03, Bldg. 6100
   9000 Rockville Park
   Bethesda, MD 20892

6. Perinatal Trials Service National Perinatal Epidemiology Unit *
   Diana Elbourne, PhD
   NEPU, Radcliffe Infirmary
   Oxford, OX2 6HE UK

7. Study Group for Complications of Perinatal Care* T. Macpherson, MD
   Department of Pathology
   Magee Women's Hospital
   Pittsburgh, PA 15213

8. Tokyo Metropolitan Maternal & Child Health Service Center Data Base Project
   M. Hirayama, MD
   Tokyo MCH Center
   Tokyo, JAPAN

9. Vermont-Oxford Trials Network* Lynn Stillman
   Neonatal Research & Technology Assessment, Inc.
   52 Overlake Park
   Burlington, VT 05401.

Two examples of networks with ongoing databases are the Vermont-Oxford Trials Network 39 and the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network.40 These networks perform randomized trials41-43 and maintain databases for observational studies. The Vermont-Oxford Trials Network consists of a broad range of university and non-university neonatal units. Membership is voluntary, most units are small, and participating investigators do not receive salary support for participation. The NICHD Neonatal Research Network consists of large university research centers with network-funded research personnel.
Both of these networks maintain databases for infants with birth weights below 1,500 grams (3 pounds, 5 ounces). In 1992, the Vermont- Oxford Trials Network collected information about more than 5,000 very low birth weight infants at 68 neonatal units. In 1993, it collected data on more than 6,600 infants at 84 neonatal units.
The NICHD Neonatal Research Network collected data on nearly 3,000 infants at 12 neonatal units in 1992. Combined, these networks represent approximately 20% of all very low birth weight infants born in the United States. Network databases such as these are now a major resource for monitoring the process and outcomes of neonatal intensive care.

Analysis of Variations

Since 1973, when Wennberg and Gittlesohn demonstrated that there were marked variations in the utilization of surgical procedures among hospital service areas within the state of Vermont, a large body of research has confirmed that both patterns of care and patient outcomes vary among geographic areas and hospitals in ways that cannot be explained by differences in the patient populations which are served.44 In addition to large differences in utilization rates for surgical procedures, diagnostic services, and hospital admissions, there is also wide variation in hospital mortality for a number of different medical conditions.45

In the following sections, we will review the data concerning variations in interventions and outcomes for neonatal intensive care and discuss approaches to risk adjustment for neonatal patients, which may be useful in determining the causes of the observed variations.

Variations in the Use of Interventions

Large variations in the use of prenatal corticosteroids exist despite their proven effectiveness in reducing morbidity and mortality among preterm infants. Corticosteroid treatment of women at risk for preterm delivery induces lung maturation in the fetus and improves neonatal outcomes.46 (For further discussion of the effectiveness of corticosteroids, see the article by Ricciotti in this journal issue.)

There is clear and convincing evidence from numerous randomized controlled trials that antenatal corticosteroid treatment not only reduces the risk of respiratory distress syndrome in preterm infants of treated women but also reduces the risk of death and intraventricular hemorrhage.47 Despite this evidence, many obstetricians prescribe antenatal steroids infrequently for women at risk for preterm delivery, and some obstetricians do not prescribe them at all. At 73 centers participating in the Vermont-Oxford Trials Network in either 1991 or 1992, 26% of the 8,749 infants weighing 501 to 1,500 grams (from 1 pound, 2 ounces to 3 pounds, 5 ounces), were born to women who had received antenatal steroids.48 Twenty-five percent of the centers in the network had treatment frequencies of 11% or less; 25% of centers had frequencies of 36% or more; only 10% of centers had frequencies of 60% or more. Data from the NICHD Neonatal Research Network also indicate wide variation in the use of antenatal steroid therapy for very low birth weight infants.40 Overall, 16% of infants in the NICHD Network were delivered to women who had received steroids with a range of 1% to 33%.

Reports from these two neonatal networks also document substantial variation among neonatal intensive care units (NICUs) in the use of a number of other postnatal interventions and procedures. Table 1 shows the overall frequencies for selected interventions and their interquartile ranges at 68 centers which participated in the Vermont- Oxford Trials Network in 1992.49

The variation persisted even within 250-gram birth weight categories. Data for variation in postnatal interventions are also provided by the NICHD Neonatal Research Network. Methods of delivery room management, use of phototherapy, exchange transfusions, indwelling vascular catheters, and parenteral nutrition all exhibited considerable variation among the NICHD Network Centers.40 The persistence of variation within relatively narrow birth weight categories suggests that the variation is due in large part to differences in practice styles among the units.

Variations in Outcomes After Neonatal Intensive Care

As previously discussed, neonatal intensive care has resulted in increased birth-weight-specific survival rates and a decrease in the overall infant mortality rate. Infants born at hospitals with level 3 neonatal intensive care units have lower neonatal mortality than infants born at hospitals without such units.15-19 Even among level 3 neonatal intensive care units, however, there are substantial variations in both mortality and morbidity among the survivors.

Avery and colleagues found that the incidence of chronic lung disease in infants weighing 700 to 1,500 grams (from 1 pound, 9 ounces to 3 pounds, 5 ounces) varied significantly among the eight institutions studied even after adjusting for birth weight, race, and gender.50 The investigators suggested that the observed variation was due to differences in respiratory care practices among the centers. Horbar, in a study of 11 neonatal intensive care units, found differences among centers both in the frequency of chronic lung disease and in neonatal mortality.51 Again, the differences persisted after adjustment for birth weight, race, and gender. Kraybill and colleagues, in a survey of 10 neonatal units in North Carolina, found significant differences among centers in the frequency of chronic lung disease.52 They also suggested that differences in respiratory care practices might explain the findings. Hack and colleagues, reporting for the NICHD Neonatal Research Network, indicate that there are large intercenter differences in morbidity, particularly with respect to chronic lung disease, necrotizing enterocolitis, intraventricular hemorrhage, and jaundice.40 Wide variation in most morbidities have also been documented for centers in the Vermont-Oxford Trials Network (see Table 2).49 ?

These data suggest that there are differences among neonatal intensive care units with respect to short-term morbidity and mortality. While some of these differences may be due to differences in the way specific outcomes are diagnosed at the different centers, the extent to which they are due to differences in the quality of medical care is unknown. Data regarding variation in long-term neurodevelopmental outcomes and other morbidities among centers are not available.

Risk Adjustment

Variations in the outcomes of hospitalized patients have been used as indicators of the effectiveness of medical care. However, before inferences can be drawn from observed differences in mortality or other outcomes among hospitals, it is necessary to account for differences in case mix. Variation in hospital mortality has three major sources: the underlying risk of a hospital's patient population, the effectiveness and appropriateness of care provided at the hospital, and sampling variations (the likelihood that the mortality observed in the study group truly represents the experience in the total population).53 Statistical models for predicting mortality risk based on patient characteristics have been developed for use in a number of different clinical situations, including adult medical and intensive care, pediatric intensive care, and neonatal intensive care.54-58 These risk adjustment models can be used to compare the observed outcomes at a particular hospital with the outcomes that would be expected based on the demographic characteristics of the hospital's patients as well as the severity of their illnesses measured by physiologic and laboratory values. After differences in patient risk and sampling variations have been accounted for, residual variation in outcome is assumed to reflect differences in the effectiveness and/or appropriateness of medical care.

One of the earliest examples of risk adjustment for the evaluation of perinatal care was reported by Williams.37 He applied a model for predicting neonatal death based on birth weight, race, sex, and multiple birth to more than 3 million infants born at 504 hospitals in California during the years 1960 to 1973. After the model had been used to account for newborn risk and the effect of chance, there was still a twofold variation in mortality at these hospitals. This residual unexplained variation was presumably the result of differences in the effectiveness of perinatal care.

More recently, risk adjustment models have been developed specifically for neonatal intensive care. Richardson and colleagues have developed the Score for Neonatal Acute Physiology (SNAP), which is patterned after the Acute Physiology and Chronic Health Evaluation (APACHE) score used in adult intensive care and the PSI used in pediatric intensive care.59 The SNAP can be applied to all NICU admissions regardless of birth weight. The SNAP is predictive of neonatal mortality even within narrow birth weight strata and is correlated with other indicators of severity, including nursing workload, therapeutic intensity, and physician estimates of mortality risk. Furthermore, the SNAP increases the accuracy of neonatal mortality risk prediction when used along with birth weight, five-minute Apgar score, and size for gestational age.60 In the future, use of scoring systems such as the SNAP will help to refine the risk adjustment analyses and provide us with a clearer picture of variations in neonatal mortality across hospitals.

The International Neonatal Network has developed the Clinical Risk Index for Babies (CRIB), a scoring system for predicting mortality risk for infants weighing 1,500 grams (3 pounds, 5 ounces) or less.61 The CRIB score is based on birth weight, gestational age, maximum and minimum fraction of inspired oxygen, maximum base excess, and presence of congenital malformations. The score uses values obtained within 12 hours of admission. The CRIB score is more accurate than birth weight alone in predicting mortality risk, and higher scores are associated with an increased risk for major cerebral abnormality.

Because postadmission data may reflect the results of treatments provided in the neonatal intensive care unit rather than the infants' underlying risk, mortality prediction models based only on admission data are preferred if the goal of risk adjustment is to identify differences in the effectiveness of care. Both the SNAP and the CRIB score use information collected during the first 12 to 24 hours after admission to the neonatal intensive care unit for predicting mortality risk.

Figure 1 shows the standardized neonatal mortality ratios (SNMRs) at 68 centers participating in the Vermont-Oxford Trials Network in 1992, and illustrates the existing variation in mortality rates in these centers. In this model, which is based on factors present at the time of admission, the observed variations cannot be attributed to the infant's birth weight, race, gender, health at birth, receipt of prenatal care, and location of the birth because the effects of these factors have been statistically controlled. The SNMR is the ratio of the number of observed deaths at a center to the number of deaths predicted based on the patient characteristics in the model. An SNMR of 1 means that a hospital has exactly the number of deaths which would be expected; values greater than 1 indicate that more deaths occurred than were expected; values less than 1 indicate that fewer deaths occurred than were expected. Although some hospitals have SNMRs that are less than 1 and others have SNMRs that are greater than 1, in most instances the 95% confidence limit includes the values of 1, which means that these hospitals do not appear to have too many or too few deaths. Improved predictor models which include major birth defects among the predictor variables are currently being developed.

Neonatal mortality prediction could serve several purposes. One purpose is the prediction of individual patient risk. It is unlikely that any model will be accurate enough to aid in patient care decisions such as when to withhold or withdraw life support. However, prediction of individual risk may be useful for identifying infants who died despite having a low predicted probability of death. These cases could then be chosen for audit as part of local quality improvement efforts.

A second purpose for neonatal risk prediction is the identification of outlier hospitals where the quality or effectiveness of care is low. Given the relatively small number of very low birth weight infants treated at individual neonatal intensive care units, the confidence intervals for estimates of measures like the SNMR will be large.62 This will severely limit the power of even very accurate statistical models to identify outlier hospitals. Aggregating cases over multiple years increases the ability to detect outliers. It remains to be proven, however, that targeting hospitals in this way accurately identifies units providing less effective care,63 as methods to adjust for the underlying risks and differences in the units remain imperfect.

A third purpose for neonatal risk prediction models is their use in studies of hospital characteristics associated with outcome. The power to detect differences in risk-adjusted mortality rates among groups of hospitals within large neonatal networks will be greater than the power to detect individual outliers. Several studies have already shown that hospital characteristics are associated with outcomes for adult and neonatal patients.37,64 Williams, in the study discussed above, found that, after adjusting for patient risk, hospitals with larger delivery services, urban hospitals, hospitals performing above-average numbers of cesarean sections, those recording Apgar scores, and hospitals with higher specialist-to-generalist ratios had lower mortality rates.37 Conversely hospitals with more Spanish-surnamed mothers and private proprietary hospitals had higher mortality rates. Paneth and colleagues have shown that risk-adjusted neonatal mortality rates at level 3 hospitals are lower than at either level 1 or level 2 hospitals in New York City.15,16 The International Neonatal Network has also shown that mortality rates adjusted for risk using the CRIB score are lower in tertiary as opposed to nontertiary neonatal care units in the United Kingdom.57

We are currently using data from the Vermont-Oxford Trials Network to investigate whether hospital characteristics, services, and staffing patterns are associated with differences in mortality for very low birth weight infants. It is not known whether patient volume, teaching status, hospital ownership, and use of ancillary personnel such as neonatal nurse practitioners influence the costs and outcomes of neonatal intensive care. Because of trends toward deregionalization of care and changing patterns of referrals due to managed care, it will be increasingly important to understand how these factors affect both costs and outcome. Neonatal networks will be valuable laboratories for answering health services questions about the delivery of neonatal intensive care.

ssessment of Interventions

Randomized Controlled Trials

The third step in the process for evaluating and improving the effectiveness of neonatal intensive care is assessing the efficacy of specific interventions. The gold standard for evaluating the efficacy of an intervention is the randomized controlled trial. The National Perinatal Epidemiology Unit in the United Kingdom has demonstrated how large, simple randomized controlled trials (RCTs) can be applied to answer important questions in perinatal medicine.65 Most medical innovations lead to only small or moderate gains over standard treatment, but large trials are required to demonstrate such gains. For example, to have a reasonable chance of demonstrating that a new treatment reduces mortality from 25% to 20% requires a study that enrolls nearly 3,000 infants, many more infants than are generally admitted to any one NICU in a year. In the past, too many of the trials performed in neonatology were single-center studies enrolling relatively few patients. It is now clear that multicenter networks will be needed to perform the large trials required to demonstrate small, but clinically important, differences between treatments. The recent multicenter trials of surfactant therapy,66 high-frequency ventilation,67 and cryotherapy for retinopathy of prematurity68 are examples of how this approach can be used successfully to test new therapies for neonatal intensive care.

A number of new therapies and technologies that are now in the early stages of clinical development will require multicenter randomized clinical trials to demonstrate their safety and efficacy for treatment of low birth weight infants (see Table 3).69-82 Several of these treatments have already been tested in initial small trials or are currently being evaluated in multicenter trials.

It is instructive to compare the situation in neonatology with that in pediatric oncology. Eighty to ninety percent of children in the United States with cancer are cared for at institutions participating in one of two national collaborative networks (Children's Cancer Study Group or Pediatric Oncology Group), using strictly defined treatment protocols.83 These groups continually refine their treatment protocols based on the results of ongoing randomized clinical trials. As a result, the outcomes for children with cancer have improved steadily.84 We should strive for a similar organizational structure in neonatal intensive care. Despite a central role in evaluating neonatal technologies, randomized trials will not be capable of answering all of the questions that arise about interventions in neonatal intensive care.

Feinstein has pointed out that randomized controlled trials are not feasible for studying multiple therapeutic candidates, minor changes in therapy, instabilities caused by rapid technological improvements in available treatment, long-term adverse side effects, studies of etiologic or other suspected "noxious" agents, and the diverse clinical roles of diagnostic technology.85 Additional methods will be required to address these issues. Observational studies based on large neonatal databases from multicenter networks will have a role in addressing those questions that cannot readily be answered using randomized trials.

Meta-Analysis: Combining Data from Many Randomized Trials

The evaluation of specific interventions requires a synthesis of all the available evidence. Traditionally, such a synthesis took the form of a qualitative literature review. Recently, statistical techniques for combining the findings of individual studies have come into widespread use. The formal overview, or meta-analysis, arrives at a summary measure of effect size by combining the results of individual trials.86 Meta- analysis provides a powerful tool for evaluating the evidence from randomized controlled trials. However, investigators must approach meta- analysis with the same methodologic rigor as any other research. Overviews should include all relevant trials which meet predefined inclusion criteria. The results of the trials are extracted from published reports and, in some instances, supplemented with unpublished data obtained from the investigators. These results are then tabulated, and summary measures of effect size are calculated. Several different statistical methods for determining the summary measures are available.87

The method of meta-analysis has been extensively applied to synthesizing the results of randomized trials in perinatal medicine. Formal meta- analyses, or overviews, have been included in the Oxford Database of Perinatal Trials, which contains a bibliography of all randomized trials in prenatal medicine published since 1940.88,89 Several obstetrical overviews90 and neonatal overviews have also been published.91

These reference works will be updated and published in electronic format, providing extensive statistical summaries of the available evidence for perinatal interventions. The highly respected Cochrane Collaboration has been established to coordinate and disseminate overviews of randomized trials which can be used to inform evidence- based decision-making. An example of a recent meta-analysis is the overview by Mercier and Soll of the effects of natural surfactant therapy for the treatment of respiratory distress syndrome (see Figure 2).92 This overview is based on 12 separate randomized trials. Taken together, these trials indicate that surfactant therapy reduces the risks for serious lung diseases and mortality.

The individual trials did not each demonstrate these results. Only by pooling the evidence from all of the trials do such clear conclusions emerge. Formal overviews such as this one are powerful tools for evaluating the efficacy and safety of specific interventions.

A novel form of meta-analysis has recently been described for combining the results of randomized trials and medical practice databases. This method--called cross-design synthesis--identifies studies conducted using similar research designs, assesses the potential biases associated with each design, makes secondary adjustments of study results to correct for known biases, and develops models for synthesizing results which minimize hidden biases.93 This untested method may prove useful in the future because it combines the strengths of randomized trials for producing unbiased comparisons with the ability of patient data bases to include a broad range of patients treated in real-world conditions.

Economic Evaluation

Faced with limited reserves and the need to choose among alternative programs, policymakers must consider not only whether neonatal intensive care is effective but also whether it is worth the cost. Increasingly, analysts have come to rely on cost-effectiveness analyses to evaluate medical care and interventions.94 In cost-effectiveness analyses, an estimate is made of the incremental costs of an improvement in health status (such as the cost of an additional year of life), attributable to a particular intervention and compared with the incremental costs of other interventions.

Cost-effectiveness analysis was one technique used in the most comprehensive economic evaluation of neonatal intensive care which was based on patients receiving intensive care in Hamilton-Wentworth County, Ontario, between 1973 and 1977.95 In addition to measuring all of the costs flowing from neonatal intensive care, such as health care and special services received by survivors after discharge, the study attempted to measure the condition of survivors based on physical function, role function, social and emotional function, and health problems. These outcome data were used to adjust life years gained by NICU survivors by a factor that takes into account the quality of life for those who survived with disabilities. The resulting units are called quality-adjusted life-years (QALYs). For infants weighing from 1,000 to 1,499 grams (between 2 pounds, 3 ounces and 3 pounds, 5 ounces), the cost (in 1978 Canadian dollars) was $59,500 per additional survivor, $2,900 per life-year gained, and $3,200 per QALY gained. For infants weighing from 500 to 999 grams (between 1 pound, 2 ounces and 2 pounds, 3 ounces), the costs were $102,500 per additional survivor, $9,300 per life-year gained, and $22,400 per QALY gained.

These cost-effectiveness ratios have little intrinsic meaning by themselves but can be compared with ratios for other major health interventions. Figure 3 shows the cost per QALY, adjusted to 1986 U.S. dollars, for neonatal intensive care in the two birth weight groups and for other selected health care interventions.96 Several interventions-- including coronary bypass surgery for single-vessel disease, school tuberculin testing, continuous ambulatory peritoneal dialysis, hospital hemodialysis, and liver transplantation--all have higher estimated costs per QALY gained than neonatal intensive care for infants weighing from 500 to 999 grams (between 1 pound, 2 ounces and 2 pounds, 3 ounces). For example, a liver transplant costs $250,000 for each QALY gained while NICU care for a moderately low birth weight infant costs less than $10,000 for each QALY.

The Office of Technology Assessment, in its 1987 evaluation of neonatal intensive care, concluded: "Neonatal intensive care results in both increased survival and increased costs. Moreover, neonatal intensive care becomes more expensive as it is employed in increasingly marginal cases. The worth of a life saved, however, is ultimately a value judgment involving ethical and social considerations. The results from cost-effectiveness studies alone cannot guide decisions regarding who should receive care."21

These conclusions are still valid today. In addition, advances in neonatal intensive care have influenced both the costs and the outcomes of this care. Since the Canadian study was done, survival rates for infants weighing less than 1,000 grams (2 pounds, 3 ounces) have increased. A recent report from the province of Alberta, Canada, suggests that the condition of NICU survivors may also be improving.97 The introduction of surfactant therapy has changed both outcomes and costs.12,13 It is estimated that use of surfactant among very low birth weight infants has resulted in a 30% decrease in mortality and a similar decrease in use of hospital resources.13 Several other investigators have concluded that surfactant therapy is cost-effective, but definitive analyses are still required.13,98-103 An economic evaluation of neonatal intensive care based on patients treated in the 1990s is urgently needed.

Ultimately policymakers and society will have to decide whether resources should be allocated to neonatal intensive care as opposed to other purposes. Joyce and colleagues put this choice into perspective by comparing the cost-effectiveness of different strategies to reduce infant mortality.104 They estimated that, if there was a direct causal relationship between prenatal care and infant mortality, it would cost between $30 and $40 (1984 U.S. dollars) to save an additional life by expanding the number of women who receive prenatal care in the first trimester, as opposed to $2,000 or $3,000 to save an additional life by expanding the number of low birth weight infants who receive neonatal intensive care.

Unfortunately, there is growing evidence that providing access to quality prenatal care alone is insufficient to reduce low birth weight.

Changing Physician Behavior

The final step in the process of evaluating and improving the effectiveness and efficiency of neonatal intensive care involves changing physicians' practices. Greco and Eisenberg have recently reviewed the general methods that can be used to alter the practice behavior of physicians.105 These include
(1) educational processes such as continuing medical education and practice guidelines,
(2) feedback of information comparing individual physician practices and patient outcomes with benchmark standards,
(3) continuous quality improvement programs and other efforts to engage physicians in change,
(4) administrative interventions, and
(5) financial incentives or penalties. In general, whether an intervention is successful depends on the particular circumstances in which it is used; combinations of methods appear to be the most effective.

Neonatal networks have the potential to influence practice patterns of neonatologists using several of these approaches. Membership in a database that produces benchmarking reports, participation in meetings, communication with other network members, and the use of standardized research protocols supplies many elements needed to change neonatologists' behavior.

Participation in multicenter randomized clinical trials may itself be an educational process capable of causing changes in practice. Surfactant therapy for neonatal respiratory distress syndrome was rapidly adopted in nearly all neonatal intensive care units in North America soon after surfactants were commercially released in 1991. The strength of the experimental evidence and advertising by commercial sponsors were partly responsible for the rapid acceptance of surfactant therapy. The fact that hundreds of neonatal units actually participated in the randomized controlled trials conducted in the preceding decade also played a role. Information from neonatal network databases has the potential to influence physician behavior and neonatal intensive care practices.106,107 The Vermont-Oxford Trials Network currently provides members with reports containing feedback on how local practices and outcomes compare with those of the total network. These reports are intended for use in continuous quality improvement programs. The reports include data on risk-adjusted mortality, length of hospital stay, and adverse outcomes such as intraventricular hemorrhage (bleeding in the brain), chronic lung disease, necrotizing enterocolitis (severe damage to the intestines), nosocomial infection (infections acquired in the hospital), and retinopathy of prematurity (impaired vision or blindness). It will be important to evaluate how these reports are actually used by the members of the network and to determine whether such feedback leads to changes in physician behavior and improvements in the quality of neonatal intensive care.

The Role of Parents in Evaluating Neonatal Intensive Care

Parents and families must play a critical role in evaluating neonatal intensive care. Assessments of the benefits and costs of neonatal intensive care should incorporate the views and experiences of a broad range of families whose infants received intensive care. A recent conference attended by neonatologists and parents who had personal experience with neonatal intensive care developed a set of "Principles for Family-Centered Neonatal Care," which will help parents and professionals work together.108 Two of these principles focus directly on issues relating to the evaluation of neonatal intensive care. One principle urges that new treatments be introduced only in the context of properly controlled trials and states, "Experienced parents should have a voice in determining the research agenda, in establishing outcomes of interest, and in educating other parents about the need for ethically and scientifically sound research in neonatology." The other principle states, "Parents and professionals must work together to promote meaningful long-term follow-up for all high risk NICU survivors."

Only a small percentage of infants receiving neonatal intensive care are currently enrolled in formal follow-up studies, and there are many unanswered questions about the quality of life, particularly for extremely low birth weight survivors.109,110 Those follow-up data that are available concentrate primarily on defining the incidence of neurologic deficits. A more complete evaluation of societal and familial costs and benefits is needed which takes into account the experience and opinions of parents and families and examines a broad range of health outcomes.111 An evaluation of this kind will require collaboration between professionals and the parents and families whose infants have been cared for in neonatal intensive care units. There is great interest in developing simple report cards by which patients can compare or judge results of various therapies. The New York State Department of Health is a leader in this field. Since 1991, it has supplied data to the public ranking cardiac surgical services and individual surgeons.112 Other states have begun publishing similar data. This movement is bound to spread to neonatology. Comparative data will be of little value, however, unless they are carefully gathered and reported by a neutral source in a format that parents can understand. Neonatal networks will be capable of compiling the necessary data. Creative new ideas for presenting information, such as the use of interactive videodisks, may help to explain the information to parents.

Conclusion

Nearly 15 years ago, Sinclair and colleagues concluded that, although the efficacy of specific neonatal intensive care interventions had been demonstrated, the overall effectiveness and efficiency of neonatal intensive care programs required validation in randomized controlled trials.7 Such trials have not yet been done. It is unlikely that they will be for several reasons. First, there is widespread agreement that neonatal intensive care saves lives. Second, there is considerable disagreement about what to include in a standard package of neonatal intensive care given the multitude of elements that make up modern intensive care and the marked variations in practice which currently exist. Third, diagnostic and therapeutic technologies are changing so rapidly that any package of services would quickly become obsolete as new tests and treatments were introduced. Although the randomized controlled trial will remain the gold standard by which all new therapies must be judged, a broader range of methods drawn from clinical evaluation science will be needed if the effectiveness of neonatal intensive care as a whole is to be monitored and continuously improved. The proper question is no longer "Does neonatal intensive care work?" but rather "How can neonatal intensive care be made more effective and efficient?" Neonatal research networks will play an important role in finding the answers and applying them.

1. National Committee on Perinatal Health. Toward improving the outcome of pregnancy. White Plains, NY: National Foundation-March of Dimes, 1976.
2. McCormick, M.C., Shapiro, S., and Starfield, B.H. The regionalization of perinatal services: Summary of the evaluation of a national demonstration program. Journal of the American Medical Association (1985) 253:799-804.
3. Butterfield, L.J. Organization of regional perinatal programs. Seminars in Perinatology (1977) 1:217-33.
4. American Hospital Association. Hospital statistics. Chicago: AHA, 1989.
5. Merenstein, G.B., Rhodes, P.G., and Little, G.A. Personnel in neonatal pediatrics: Assessment of numbers and distribution. Pediatrics (1985) 76:454-56.
6. Silverman, W.A. Is neonatal medicine in the United States out of step? Pediatrics (1993) 92:612-13.
7. Sinclair, J.C., Torrance, G.W., Boyle, M.H., et al. Evaluation of neonatal-intensive care programs. New England Journal of Medicine (1981) 305:489-94.
8. Lee, K.S., Paneth, N., Gartner, L.M., et al. Neonatal mortality: An analysis of the recent improvement in the United States. American Journal of Public Health (1980) 70:15-21.
9. Williams, R.L., and Chen, P.M. Identifying the sources of the recent decline in perinatal mortality rates in California. New England Journal of Medicine (1982) 306:207-14.
10. Buehler, J.W., Kleinman, J.C., Hogue, C.J.R., et al. Birth weight- specific infant mortality, United States, 1960 and 1980. Public Health Reports (1987) 102:151-61.
11. Hilts, P.J. U.S. reports drop in infant deaths. New York Times. April 6, 1991, at A1.
12. Horbar, J.D., Wright, E.C., Onstad, L., et al. Decreasing mortality associated with the introduction of surfactant therapy: An observational study of neonates weighing 601 to 1300 grams at birth. Pediatrics (1993) 92:191-96.
13. Schwartz, R.M., Luby, A.M., Scanlon, J.W., and Kellogg, R.J. Effect of surfactant on morbidity, mortality, and resource use in newborn infants weighing 500 to 1500 g. New England Journal of Medicine (1994) 30:1476-80.
14. Paneth, N. Technology at birth. Editorial. American Journal of Public Health (1990) 80:791-92.
15. Paneth, N., Kiely, J., Wallenstein, S., and Susser, M. The choice of place of delivery: Effect of hospital level on mortality in all singleton births in New York City. American Journal of Diseases of Childhood (1987) 141:60-64.
16. Paneth, N., Kiely, J.L., Wallenstein, S., et al. Newborn intensive care and neonatal mortality in low-birth-weight infants: A population study. New England Journal of Medicine (1982) 307:148-55.
17. Gortmaker, S., Sobol, A., Clark, C., et al. The survival of very low-birth weight infants by level of hospital at birth: A population study of perinatal systems in four states. American Journal of Obstetrics and Gynecology (1985) 152:517-24.
18. Mayfield, J.A., Rosenblatt, R.A., Baldwin, L.M., et al. The relation of obstetrical volume and nursery level to perinatal mortality. American Journal of Public Health (1990) 80:819-23.
19. Verloove-Vanhorick, S.P., Verwey, R.A., Ebeling, M.C.A., et al. Mortality in very preterm and very low birthweight infants according to place of birth and level of care: Results of a national collaborative survey of preterm and very low birthweight infants in the Netherlands. Pediatrics (1988) 81:404-11.
20. Horwood, S.P., Boyle, M.H., Torrance, G.W., and Sinclair, J.C. Mortality and morbidity of 500- to 1,499-gram birth weight infants live-born to residents of a defined geographic region before and after neonatal intensive care. Pediatrics (1982) 69,5:613-20.
21. U.S. Congress, Office of Technology Assessment. Neonatal intensive care for low birthweight infants: Costs and effectiveness. Health Technology Case Study 38. OTA-HCS-38. Washington, DC: OTA, 1987.
22. Bhushan, V., Paneth, N., and Kiely, J.L. Impact of improved survival of very low birth weight infants on recent secular trends in the prevalence of cerebral palsy. Pediatrics (1993) 91:1094-1100.
23. McCormick, M.C. The contribution of low birth weight to infant mortality and childhood morbidity. New England Journal of Medicine (1985) 312:82-90.
24. McCormick, M.C. Long-term follow-up of infants discharged from neonatal intensive care units. Journal of the American Medical Association (1989) 261:1767-72.
25. McCormick, M.C., Brooks-Gunn, J., Workman-Daniels, K., et al. The health and developmental status of very low-birth-weight children at school age. Journal of the American Medical Association (1992) 267:2204- 8.
26. Hack, M., Taylor, G., Klein, N., et al. School-age outcomes in children with birth weights under 750g. New England Journal of Medicine (1994) 331:753-59.
27. Saigal, S., Rosenbaum, P., and Stoskopf, B. Comprehensive assessment of the health status of extremely low birth weight children at eight years of age: Comparison with a reference group. Journal of Pediatrics (1994) 125:411-17.
28. Shankaran, S., Cohen, S.N., Linver, M., and Zonia, S. Medical care costs of high-risk infants after neonatal intensive care: A controlled study. Pediatrics (1988) 81:372-78.
29. McCormick, M.C., Bernbaum, J.C., Eisenberg, J.M., et al. Costs incurred by parents of very low birth weight infants after the initial neonatal hospitalization. Pediatrics (1991) 88:533-41.
30. Hack, M., and Fanaroff, A. Outcomes of extremely-low-birth-weight infants between 1982 and 1988. New England Journal of Medicine (1989) 321:1642-47.
31. Allen, M.C., Donohue, P.K., and Dusman, A.E. The limit of viability--Neonatal outcome of infants born at 22 to 25 weeks' gestation. New England Journal of Medicine (1993) 329:1597-601.
32. Nishida, H., and Ishizuka, M.E. Survival rate of extremely low birthweight infants and its effect on the amendment of the Eugenic Protection Act in Japan. Acta Pediatrica Japonica (1992) 34:612-16.
33. Nishida, H. The viability limit of gestation for the fetus and premature neonates. Asian Medical Journal (1992) 35:487-94.
34. Aday, L.A., Begley, C.E., Lavison, D.R., and Slater, C.H. Evaluating the medical care system: Effectiveness, efficiency and equity. Ann Arbor, MI: Health Administration Press, 1993.
35. Raskin, I.E., and Maklau, C.W. Medical treatment effectiveness research: A view from inside the Agency for Health Care Policy and Research. Evaluation and the Health Professions (1991) 14:161-86.
36. Roper, W.L., Winkenwerder, W., Hackbarth, G.M., and Krakauer, H. Effectiveness in health care: An initiative to evaluate and improve medical practice. New England Journal of Medicine (1988) 319:1197-1202.
37. Williams, R.L. Measuring the effectiveness of perinatal medical care. Medical Care (1979) 17:95-110.
38. National Committee on Perinatal Health. Toward improving the outcome of pregnancy: The 90s and beyond. White Plains, NY: March of Dimes Birth Defects Foundation, 1993.
39. Investigators of the Vermont-Oxford Trials Network Database Project. The Vermont-Oxford Trials Network: Very low birth weight outcomes for 1990. Pediatrics (1993) 91:540-45.
40. Hack, M., Horbar, J.D., Malloy, M.H., et al. Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Network. Pediatrics (1991) 87:587-97.
41. The Vermont-Oxford Trials Network. A multicenter randomized trial comparing synthetic surfactant to modified bovine surfactant extract in the treatment of neonatal respiratory distress syndrome. Pediatric Research (1994) 35:259A.
42. Fanaroff, A., Wright, E., Korones, S., et al. A controlled trial of prophylactic intravenous immunoglobulin (IVIG) to reduce nosocomial infections in VLBW infants. New England Journal of Medicine (1994) 330:1107-13.
43. Horbar, J.D., Wright, L.L., and Soll, R.F. A multicenter randomized trial comparing two surfactants for the treatment of neonatal respiratory distress syndrome. Journal of Pediatrics (1993) 123:757-66.
44. Wennberg, J., and Gittlesohn, A. Small area variations in health care delivery. Science (1973) 182:1102-8.
45. Health Care Financing Administration. Medicare hospital information 1988, 1989, 1990. Vol. 55. Technical supplement. Washington, DC: U.S. Department of Health and Human Services, 1992.
46. Ballard, P.L. Glucocorticoid effects in vivo. In Monographs in endocrinology. Vol. 28: Hormones and lung maturation. P.L. Ballard, ed. Berlin: Springer-Verlag, 1986, pp. 24-71.
47. Crowley, P. Promoting pulmonary maturity. In Effective care in pregnancy and childbirth. I. Chalmers, M. Enkin, M. Keirse, eds. Oxford: Oxford University Press, 1989, pp. 746-64.
48. Horbar, J.D. Antenatal corticosteroid administration and neonatal outcomes for infants 501 to 1500 grams: Vermont-Oxford Trials Network 1991-1992. Program and abstracts of the NIH Consensus Development Conference on the Effect of Corticosteroids for Fetal Maturation on Perinatal Outcomes. Bethesda, MD: National Institute of Health, 1994, pp. 47-52.
49. The Vermont-Oxford Trials Network 1992 Annual Report. Burlington, VT: Neonatal Research and Technology Assessment, 1993.
50. Avery, M.E., Tooley, W.H., and Keller, J.B. Is chronic lung disease in low birth weight infants preventable? A survey of eight centers. Pediatrics (1987) 79:26-30.
51. Horbar, J.D., McAuliffe, T.L., and Adler, S.M. Variability in 28- day outcomes for very low birth weight infants: An analysis of 11 neonatal intensive care units. Pediatrics (1988) 82:554-59.
52. Kraybill, E.N., Bose, C.L., and D'Ercole, A.J. Chronic lung disease in infants with very-low-birthweight: A population based study. American Journal of Diseases of Children (1987) 141:784-88.
53. Jencks, S.F., Daley, J., Draper, D., et al. Interpreting hospital mortality data: The role of clinical risk adjustment. Journal of the American Medical Association (1988) 260:3611-16.
54. Hadorn, D.C., Keeler, E.B., Rogers, W.H., and Brook, R.H. Assessing the performance of mortality prediction models. Santa Monica, CA: RAND, 1993.
55. Pollack, M.M., Ruttiman, U.E., Getson, P.R., et al. Accurate prediction of the outcome of pediatric intensive care: A new quantitative method. New England Journal of Medicine (1987) 316:134-39.
56. Richardson, D.K., Gray, J.E., McCormick, M.C., et al. Score for neonatal acute physiology: A physiologic severity index for neonatal intensive care. Pediatrics (1993) 91:617-23.
57. International Neonatal Network. The CRIB (clinical risk index for babies) score: A tool for assessing initial neonatal risk and comparing performance of neonatal intensive care units. Lancet (1993) 342:193-98.
58. Horbar, J.D., Onstad, L., Wright, E., et al. Predicting mortality risk for infants weighing 501 to 1500 grams at birth: A National Institutes of Health Neonatal Research Network report. Critical Care Medicine (1993) 21:12-18.
59. Richardson, D.K., Gray, J.E., McCormick, M.C., et al. Score for neonatal acute physiology: A physiologic severity index for neonatal intensive care. Pediatrics (1993) 91:617-23.
60. Richardson, D.K., Phibbs, C.S., Gray, J.E., et al. Birth weight and illness severity: Independent predictors of neonatal mortality. Pediatrics (1993) 91:969-75.
61. See note no. 57, International Neonatal Network, for more information about the CRIB score.
62. Horbar, J.D. Birthweight-adjusted mortality rates for assessing the effectiveness of neonatal intensive care. Medical Decision Making (1992) 12:259-64.
63. Park, R.E., Brook, R.H., and Kosecoff, J. Explaining variations in hospital death rates. Journal of the American Medical Association (1990) 264:484-90.
64. Hartz, A.J., Krakauer, H., Kuhn, E.M., et al. Hospital characteristics and mortality rates. New England Journal of Medicine (1989) 321:1720-25.
65. National Perinatal Epidemiology Unit 1991 Report. Oxford: National Perinatal Epidemiology Unit, 1992.
66. Soll, R.F., and McQueen, M.C. Respiratory distress syndrome. In Effective care of the newborn infant. J.C. Sinclair and M.B. Bracken, eds. Oxford: Oxford University Press, 1992, pp. 325-55.
67. HIFI Study Group. High-frequency oscillatory ventilation compared with conventional mechanical ventilation in the treatment of respiratory failure in preterm infants. New England Journal of Medicine (January 12, 1989) 320:88-93.
68. Cryotherapy for Retinopathy of Prematurity Cooperative Group. Multicenter trial of cryotherapy for retinopathy of prematurity: Preliminary results. Pediatrics (1988) 81:697-706.
69. Shannon, K. Recombinant erythropoietin in anemia of prematurity: Five years later. Pediatrics (1993) 92:614-16.
70. Shenai, J.P., Kennedy, K.A., Chytil, F., and Stahlman, M.T. Clinical trial of vitamin A supplementation in infants susceptible to bronchopulmonary dysplasia. Journal of Pediatrics (1987) 111:269-77.

    Top

71. Hallman, M., Bry, K., Hoppu, K., et al. Inositol supplementation in premature infants with respiratory distress syndrome. New England Journal of Medicine (1992) 326:1233-39.
72. Yeh, T.F., Torre, J.A., Rastogi, A., et al. Early postnatal dexamethasone therapy in premature infants with severe respiratory distress syndrome: A double-blind, controlled study. Journal of Pediatrics (1990) 117:273-82.
73. La Force, W.R., and Brudno, S. Controlled trials of beclomethasone dipropionate by nebulization in oxygen and ventilator-dependent infants. Journal of Pediatrics (1993) 122:285-88.
74. Traystman, R.J., Kirsch, J.R., and Koehler, R.C. Oxygen radical mechanisms of brain injury following ischemia and reperfusion. Journal of Applied Physiology (1991) 71:1185-95.
75. Valaes, T., Petmezaki, S., Henschke, C., et al. Control of jaundice in preterm newborns by an inhibitor of bilirubin production: Studies with tin-mesoporphyrin. Pediatrics (1994) 93:1-11.
76. Eibi, M.M., Wolf, H.M., Furnkrantz, H., and Rosenbranz, A. Prevention of necrotizing enterocolitis in low-birth-weight infants by IgA-IgG feeding. New England Journal of Medicine (1988) 319:1-7.
77. Lester, B.M., Trunick, E.Z. Introduction: Guidelines for stimulation with preterm infants. Clinics in Perinatology (1990) 17:xv- xvii.
78. Abman, S.H., Kinsella, J.P., Shaffer, M.S., and Wilkening, R.B. Inhaled nitric oxide in the management of a premature newborn with severe respiratory distress and pulmonary hypertension. Pediatrics (1993) 92:606-9.
79. Revenis, M.E., Glass, P., and Short, B.L. Mortality and morbidity rates among lower birth weight infants (2000 to 2500 grams) treated with extracorporeal membrane oxygenation. Journal of Pediatrics (1992) 121:452-58.
80. Greenspan, J.S., Wolfson, M.R., Rubenstein, S.D., and Shaffer, T.H. Liquid ventilation of human preterm neonates. Journal of Pediatrics (1990) 117:106-11.
81. Bancalari, E. Mechanical ventilation--New strategies. Program/Syllabus Hot Topics '93 in Neonatology. Columbus, OH: Ross Laboratories, 1993, pp. 119-32.
82. Phelps, D.L. Retinopathy of prematurity. Current Problems in Pediatrics (September 1992) 22,8:349-71.
83. Bleyer, W.A. The impact of childhood cancer in the United States and the world. CA:A Cancer Journal for Clinicians (1990) 40:355-67.
84. Kramer, S., Meadows, A.T., Pastore, G., et al. Influence of place of treatment on diagnosis, treatment, and survival in three pediatric solid tumors. Journal of Clinical Oncology (1984) 2:917-23.
85. Feinstein, A.R. An additional basic science for clinical medicine: IV. The development of clinimetrics. Annals of Internal Medicine (1983) 99:843-48.
86. Sacks, H.S., Berrier, J., Reitman, D., et al. Meta-analyses of randomized controlled trials. New England Journal of Medicine (1987) 316:450-55.
87. Bracken, M.B. Statistical methods for analysis of effects of treatment on overviews of randomized trials. In Effective care of the newborn infant. J.C. Sinclair and M.B. Bracken, eds. Oxford: Oxford University Press, 1992, pp. 13-18.
88. A classified bibliography of controlled trials in perinatal medicine, 1940-1984. Oxford: Oxford University Press, 1985.
89. Chalmers, I. Oxford database of perinatal trials. Version 1.2, Spring 1992 disk issue 7. Oxford: Oxford University Press, 1992.
90. Chalmers, I., Enkin, M., and Keirse, M., eds. Effective care in pregnancy and childbirth. Vols. 1 and 2. Oxford: Oxford University Press, 1989.
91. Sinclair, J.C., and Bracken, M.B., eds. Effective care of the newborn infant. Oxford: Oxford University Press, 1992.
92. Mercier, C.E., and Soll, R.F. Clinical trials of natural surfactant extract in respiratory distress syndrome. Clinics in Perinatology (1993) 20:711-35.
93. Korper, S., Mosteller, F., and Weissman, N. Cross-design synthesis: A new form of meta-analysis for combining results from randomized clinical trials and medical-practice databases. International Journal of Technology Assessment in Health Care (1993) 9:440-49.
94. Racine, A.D., Joyce, T.J., and Grossman, M. Effectiveness of health services for pregnant women and infants. The Future of Children (Winter 1992) 2,2:40-57.