I. GENERAL PRINCIPLES. Mechanical ventilation is an invasive life support procedure with many effects on the cardiopulmonary system. The goal is to optimize both gas exchange and clinical status at minimum fractional concentration of inspired oxygen (FiO₂) and ventilator pressures/tidal volume (V_T). The ventilator strategy employed to accomplish this goal depends, in part, on the infant’s disease process. In addition, recent advances in technology have brought more options for ventilatory therapy of newborns.

1. Continuous positive airway pressure (CPAP) is usually administered by means of a ventilator, stand-alone CPAP delivery system, or “bubble” CPAP systems. Any system used to deliver CPAP should allow continuous monitoring of the delivered pressure and be equipped with safety alarms to indicate when the pressure is above or below the desired level. Alternatively, CPAP may be delivered by a simplified system providing blended oxygen flowing past the infant’s airway, with the end of the tubing submerged in 0.25% acetic acid in sterile water solution to the desired depth to generate pressure (“bubble CPAP”). Stand-alone variable flow CPAP devices, in which expiratory resistance is decreased via a “fluidic flip” of flow at the nosepiece during expiration, are also available.

2. General characteristics. A continuous flow of heated, humidified gas is circulated past the infant’s airway, typically at a set pressure of 3 to 8 cm H₂O, maintaining an elevated end-expiratory lung volume while the infant breathes spontaneously. The air-oxygen mixture and airway pressure can be adjusted. Variable flow CPAP systems may decrease the work of breathing and improve lung recruitment in infants on CPAP
but have not been shown to be clearly superior to conventional means of delivery. CPAP is usually delivered by means of nasal prongs, nasopharyngeal tube, or nasal mask. Endotracheal CPAP should not be used because the high resistance of the endotracheal tube increases the work of breathing, especially in small infants. Positive-pressure hoods and continuous-mask CPAP are not recommended.

a. CPAP is less invasive than mechanical ventilation and causes less lung injury.

b. When used early in infants with respiratory distress syndrome (RDS), CPAP can help prevent alveolar and airway collapse and thereby reduce the need for mechanical ventilation.

c. Use of immediate CPAP in the delivery room for spontaneously breathing immature infants ≥24 weeks’ gestation decreases the need for mechanical ventilation and administration of surfactant. Although individual trials comparing initial CPAP and mechanical ventilation and early surfactant treatment show similar rates of bronchopulmonary dysplasia (BPD), meta-analyses of the prospective randomized trials of early CPAP show that initial CPAP use is associated with a decreased risk of death or BPD.

d. CPAP decreases the frequency of obstructive and mixed apneic spells in some infants.

a. CPAP is not effective in patients with frequent apnea or inadequate respiratory drive.

b. CPAP provides inadequate respiratory support in the face of severely abnormal pulmonary compliance and resistance.

c. Maintaining nasal or nasopharyngeal CPAP in large, active infants may be technically difficult.

d. Infants on CPAP frequently swallow air, leading to gastric distension and elevation of the diaphragm, necessitating decompression by a gastric tube.

5. Indications (see section III.A)

1. Many units have switched to use of high flow nasal cannula (HFNC) as an alternative to conventional CPAP devices. HFNC allows delivery of distending pressure to the infant’s airway with a simpler patient interface.

2. General characteristics. HFNC usually refers to the delivery of blended, heated, and humidified oxygen at flows >1 L/minute via small binasal prongs. Two commercial devices for delivery of HFNC are available for use in newborns.

a. Reported advantages to HFNC include ease of use, a simpler patient interface, and a lower incidence of nasal breakdown compared with conventional CPAP.

b. Randomized trials to date comparing HFNC to CPAP as postextubation support in extremely preterm infants are limited but suggest that HFNC may be an acceptable alternative to CPAP in many infants.
Data suggest that the failure of HFNC may be higher than conventional CPAP in infants <26 weeks’ gestation.

a. Potential disadvantages include more variable distending pressure delivery (both low and high) and a tendency for a longer duration of respiratory support compared with CPAP.

C. Pressure-limited, time-cycled, continuous flow ventilators have historically been used in newborns with respiratory failure but have been replaced in most U.S. neonatal intensive care units (NICUs) by patient-triggered, volume-targeted ventilators (see the following text).

1. General characteristics of pressure-limited ventilation. A continuous flow of heated and humidified gas is circulated past the infant’s airway; the gas is a mixture of air, blended with oxygen to maintain the desired oxygen saturation level. Peak inspiratory pressure (P_I or PIP), positive end-expiratory pressure (PEEP), and respiratory timing (rate and duration of inspiration and expiration) are selected.

a. The continuous flow of fresh gas allows the infant to make spontaneous respiratory efforts between ventilator breaths (intermittent mandatory ventilation [IMV]).

b. Good control is maintained over respiratory pressures.

c. Inspiratory and expiratory time can be independently controlled.

d. The system is relatively simple and inexpensive.

a. V_T is poorly controlled.

b. The system does not respond to changes in respiratory system compliance.

c. Spontaneously breathing infants, who breathe out of phase with too many IMV breaths (“bucking” or “fighting” the ventilator), may receive inadequate ventilation and are at increased risk for air leak.

D. Synchronized and patient-triggered (assist/control or pressure support) ventilators are adaptations of conventional pressure-limited ventilators used for newborns.

1. General characteristics. These ventilators combine the features of pressure-limited, time-cycled, continuous flow ventilators with an airway pressure, airflow, or respiratory movement sensor. By measuring inspiratory flow or movement, these ventilators deliver intermittent positive-pressure breaths at a fixed rate, in synchrony with the baby’s inspiratory efforts (“synchronized IMV,” or synchronized intermittent mandatory ventilation [SIMV]). During apnea, SIMV ventilators continue to deliver the set IMV rate. In patient-triggered ventilation, a positive pressure breath is delivered with every inspiratory effort. As a result, the ventilator delivers more frequent positive pressure breaths, usually allowing a decrease in the inspiratory pressure (PIP) needed for adequate gas exchange. During apnea, the ventilator in patienttriggered mode delivers an operator-selected IMV (“control”) rate. In some ventilators, synchronized IMV breaths can be supplemented
by pressure-supported breaths in the spontaneously breathing infant. Ventilators equipped with a flow sensor can also be used to monitor delivered V_T continuously by integration of the flow signal. Two types of patient-triggered ventilation are commonly available to the following:

a. In assist/control (A/C) ventilation, the ventilator delivers a breath with each inspiratory effort. The clinician sets the inspiratory time and peak inflation pressure or target V_T. The clinician also sets a minimum mandatory ventilator rate to maintain adequate minute ventilation should the spontaneous respiratory rate fall below the minimum selected rate.

b. Pressure support ventilation (PSV) is similar to A/C mode in that each spontaneous patient breath results in a ventilator support breath. However, each breath is terminated when inspiratory gas flow falls to a predetermined proportion of peak flow (usually 15% to 20%). As a result, the patient determines the rate and pattern of breathing (inspiratory time or inspiratory:expiratory ratio). PSV counteracts the resistance imposed by the endotracheal tube and ventilator circuit by providing additional inspiratory flow that is limited to a preset pressure selected by the clinician. A higher inspiratory flow rate (shorter inspiratory rise time or slope) shortens the time to reach maximal airway pressure, which decreases the work of breathing.

a. Synchronizing the delivery of positive pressure breaths with the infant’s inspiratory effort reduces the phenomenon of breathing out of phase with IMV breaths (“fighting” the ventilator). This may decrease the need for sedative medications and aid in weaning mechanically ventilated infants.

b. Pronounced asynchrony with ventilator breaths, during conventional IMV, has been associated with the development of air leak and intraventricular hemorrhage. Whether the use of SIMV or A/C ventilation reduces these complications is not known.

a. Under certain conditions, the ventilators may inappropriately trigger a breath because of signal artifacts or fail to trigger because of problems with the sensor.

b. Limited data are available comparing patient-triggered ventilation to other modes of ventilation in newborns. PSV may not be appropriate for small premature infants with irregular respiratory patterns and frequent apnea because of the potential for significant variability in ventilation. However, some data suggest that use of patient-triggered modes of ventilation in premature infants may decrease markers of lung inflammation and facilitate earlier extubation, when used as the initial mode of mechanical ventilator support.

4. Indications. SIMV can be used when a conventional pressure-limited ventilator is indicated. If available, it is the preferable mode of ventilator therapy in infants who are breathing spontaneously while on IMV. The indications for A/C and PSV have not been established, although many NICUs use these modes because of perceived advantages of using lower peak inspired pressure and smaller V_Ts.

E. Volume-targeted ventilators. Advances in technology for measuring delivered V_Ts have made these ventilators first-line therapy for newborns with respiratory failure. Only volume-targeted ventilators specifically designed for newborns should be used. Volume-targeted ventilators are always patient-triggered.

1. General characteristics. Volume-targeted ventilators are similar to pressure-limited ventilators except that the operator selects the V_T delivered rather than the PIP. “Volume guarantee” is a mode of pressurelimited SIMV, in which the ventilator targets an operator-chosen V_T (usually 4 to 6 mL/kg) during mechanically delivered breaths. Volume guarantee allows rapid response of the ventilator pressures to changing lung compliance and may be particularly useful in infants with RDS who receive surfactant therapy. Pressure-regulated volume control (PRVC) is a modified pressure-targeted ventilatory mode, in which inspiratory pressure is sequentially adjusted to deliver a target inspiratory volume with the lowest possible pressures.

2. Advantages. The pressure automatically varies with respiratory system compliance to deliver the selected V_T, therefore minimizing variability in minute ventilation and avoiding wide swings in V_T frequently seen with pressure-limited ventilators. Recent data suggest that volume-targeted ventilation reduces the risk of death or BPD in extremely low birth weight infants, presumably by reduction of the risk of volutrauma.

a. The system can be complicated and requires more skill to operate.

b. Because V_Ts in infants are small, some of the V_Ts selected are lost in the ventilator circuit or from air leaks around uncuffed endotracheal tubes. Some ventilators compensate for these losses by targeting expired rather than inspired V_Ts or by accounting for dead space in the circuit.

4. Indications. Volume-targeted ventilators are particularly useful if lung compliance is rapidly changing, as in infants receiving surfactant therapy.

F. High-frequency ventilation (HFV) is an important adjunct to conventional mechanical ventilation in newborns. Three types of high-frequency ventilators are approved for use in newborns in the United States: a highfrequency oscillator (HFO), a high-frequency flow interrupter (HFFI), and a high-frequency jet (HFJ) ventilator.

1. General characteristics. Available high-frequency ventilators are similar despite considerable differences in design. All are capable of delivering extremely rapid rates (300 to 1,500 breaths per minute, 5 to 25 Hz; 1 Hz = 60 breaths per minute), with V_Ts equal to or smaller than anatomic dead space. These ventilators apply continuous distending pressure to maintain an elevated lung volume; small V_Ts are superimposed at a rapid rate. HFJ ventilators are paired with a conventional pressure-limited device, which is used to deliver intermittent “sigh” breaths to help prevent atelectasis. Sigh breaths are not used with HFO ventilation. Expiration is passive (i.e., dependent
on chest wall and lung recoil) with HFFI and HFJ machines, whereas expiration is active with HFO. The mechanisms of gas exchange are incompletely understood.

a. HFV can achieve adequate ventilation while avoiding the large swings in lung volume required by conventional ventilators and associated with lung injury. Because of this, HFV may be useful in pulmonary air leak syndromes (pulmonary interstitial emphysema [PIE], pneumothorax) or in infants failing conventional mechanical ventilation.

b. HFV allows the use of a high mean airway pressure (MAP) for alveolar recruitment and resultant improvement in ventilation-perfusion ([V with dot above]/[Q with dot above]) matching. This may be advantageous in infants with severe respiratory failure, requiring high MAP to maintain adequate oxygenation on a conventional mechanical ventilator.

3. Disadvantages. Despite theoretical advantages of HFV, no significant benefit of this method has been demonstrated in routine clinical use over more conventional ventilators. Only one rigorously controlled study found a small reduction in BPD in infants at high risk treated with HFO ventilation as the primary mode of ventilation. This experience is likely not generally applicable, however, because other studies have shown no difference. These ventilators are more complex and expensive, and there is less long-term clinical experience. The initial studies with HFO suggested an increased risk of significant intraventricular hemorrhage, although this complication has not been observed in recent clinical trials. Studies comparing the different types of high-frequency ventilators are unavailable; therefore, the relative advantages or disadvantages of HFO, HFFI, and HFJ, if any, are not characterized.

4. Indications. HFV is primarily used as a rescue therapy for infants failing conventional ventilation. Both HFJ and HFO ventilators have been shown to be superior to conventional ventilation in infants with air leak syndromes, especially PIE. Because of the potential for complications and equivalence to conventional ventilation in the incidence of BPD, we do not use HFV as the primary mode of ventilatory support in infants.

G. Noninvasive mechanical ventilation. Neonatal nasal intermittent positive pressure ventilation (NIPPV) provides noninvasive respiratory support to preterm infants who otherwise would require endotracheal intubation and ventilation. It is a supplement to CPAP. NIPPV superimposes inflations set to a peak pressure delivered through nasal prongs or mask. Some devices attempt to synchronize inflations with the infant’s spontaneous inspirations. It remains unclear if NIPPV is superior to conventional CPAP or prevents the need for mechanical ventilation.

1. NIPPV has been used for the following clinical settings:

a. Apnea of prematurity

b. Following extubation, NIPPV compared with nasal CPAP has been reported shown to reduce extubation failure in infants who required intubation and ventilation.

c. Primary mode of ventilation in preterm infants with RDS

A. Indications for CPAP in the preterm infant with RDS include the following:

1. Recently delivered preterm infant with minimal respiratory distress and low supplemental oxygen requirement (to prevent atelectasis)

2. Respiratory distress and requirement of FiO₂ above 0.30 by hood

3. FiO₂ above 0.40 by hood

4. Initial stabilization in the delivery room for spontaneously breathing, extremely preterm infants (25 to 28 weeks’ gestation)

5. Initial management of premature infants with moderate respiratory distress

6. Clinically significant retractions and/or distress after recent extubation

7. In general, infants with RDS who require FiO₂ above 0.35 to 0.40 on CPAP should be intubated, ventilated, and given surfactant replacement therapy. In some NICUs, intubation for surfactant therapy in infants with RDS is followed by immediate extubation to CPAP. We generally use mechanical ventilation for all infants who are given surfactant.

8. After extubation to facilitate maintenance of lung volume

9. HFNC is likely equivalent to CPAP in postextubation stabilization; it remains unclear if it is as effective in stabilization of infants with more severe respiratory distress or in infants <26 weeks’ gestation.

B. Relative indications for mechanical ventilation in any infant include the following:

1. Frequent intermittent apnea unresponsive to methylxanthine therapy