Indian Pediatrics 1998; 35:631-640

Recent Advances in Assisted Ventilation for Neonatal Respira- tory Distress Syndrome

This manuscript provides an overview of recent advances in assisted ventilation in the neonate including continuous distending pressure (CDP), intermittent mandatory ventilation (IMV), synchronised ventilation, sedation and muscle relaxation, weaning and extubation, and high frequency ventilation. Although a great deal of research has been carried out on the treatment of respiratory distress syndrome (RDS), there is no 'best way' of clinical management as there remain many areas of uncertainty and often little objective evidence to indicate that one approach is better than another(1).

Continuous Distending Pressure

Techniques used to generate continuous positive airway pressure (CP AP) include a headbox with neck seal, face-mask, short twin nasal cannulae, long nasopharyngeal tube and endotracheal tube. Oropharyngeal pressures are similar whether monoor binasal cannulae are used(2). However, two short wide prongs should provide better pressure delivery than one long prong, as resistance is proportional to length and inversely proportion to radius. The Infant Flow Nasal CP AP System (flow driver), designed to assist expiratory flow, decrease respiratory resistance and decrease the work of breathing, is currently commonly used for CDP delivery. It must be applied correctly as nasal deformities including snubbing of the nose, flaring of the nostrils, and columella and septal necrosis have been described as complications from its use(3). A body-box which generates continuous negative extrathoracic pressure was used in the rnid-1970s but has recently been reintroduced for use in conjunction with IMV(4).

The benefits of CDP used for the treatment of RDS have been evaluated by meta-analysis(5). COP significantly improves oxygenation, reduces the need for subsequent IMV and lowers the mortality rate. However, the incidence of air leak is significantly increased and there is no reduction in chronic lung disease (CLD). The same meta-analysis show that, comparing early COP (when the FiO₂ has reached 0.3- 0.6) to late COP (when the FiO₂ has exceeded 0.6), early CDP results in a significant reduction in the need for subsequent IMV and duration of assisted ventilation, even though the reduction in the incidence of air leak and mortality did not reach statistical significance. A retrospective study comparing the incidence of CLD using different respiratory management policies and observational studies without a control group also suggested that the early use of nasal CDP in infants with RDS reduces the CLD rate(6). Although an observational study suggested that early nasal COP may be as good as initial endotracheal intubation and mechanical ventilation(7), a randomized controlled trial (RCT) is required to answer the question of which treatment is the best.

Intermittent Mandatory Ventilation

Pressure-limited ventilators remain by far the most commonly used in neonatal ventilation, and are also simpler to use and less expensive. However, with the current theory that ventilator induced lung injury does in fact result from 'volutrauma', there is renewed interest in volume-controlled ventilation using new ventilators designed with improved microprocessor technology(8).

Recent studies have demonstrated that if the peak inspiratory pressure (PIP) is held constant, this may decrease tidal volume and increase PaC02(9) as it reduces the difference between PIP and positive end-expiratory pressure (PEEP), and reduces lung compliance. PEEP is four times as potent as PIP in bringing about changes in tidal volume(1O). For example, reducing PEEP by 1 cm HP is twice as potent as increasing PIP by 2 cm Hp in achieving an increase in tidal volume. Similarly, a 1 cm H₂O increase in PEEP is twice as effective as a 2 cm H20 decrease in PIP in reducing tidal volume. Excessive PEEP also over inflates the lung which increases pulmonary vascular resistance and intra-pulmonary right-to-Ieft shunting as well as impairs venous return and cardiac output(11).

Controversy remains whether a slow or fast ventilator rate is better, and the optimal inspiratory time (IT), expiratory time (ET) and inspiratory: expiratory ratio (IE ratio). Recently, a large multicenter trial was published involving over 300 infants(12), in which slow rate (initial settings 30/min, IE ratio 1:1, IT 1 sec, ET 1 sec) was compared to fast rate (initial settings 60/ min, IE ratio 1:2, IT 0.33 see, ET 0.67 sec). With the fast rate regime, when the mean airway pressure (MAP) reached 12 cm H20 and the FiO₂ exceeded 0.6, the IE ratio was increased to 1:1 (IT 0.5 see, ET 0.5 sec) and the gas flow was increased from 6-8 1/ min to 8-10 1/ min. PIP and PEEP were similar in the two groups. The meta-analysis of all slow versus fast rate RCTs(5) showed a significant increase in the incidence in pneumothorax (PTX) with the slow rate regime for which the trend towards a lower incidence of CLD was offset by a trend towards a higher mortality rate. It appears that increasing the IT beyond 0.5 second on the slow rate regime may improve oxygenation by increasing MAP but the risk of PTX is also increased. However, infants with severe RDS cannot be adequately oxygenated with a short IT without resorting to a PIP of >30 cmH₂O. Neither the fast or slow rate regime suits all infants, probably due to the fact that it depends on the type and severity of their underlying lung disease. With the fast rate regime, rates of 120/min have been shown to be safe in the majority of non-paralysed infants but this should be avoided in paralysed infants especially those >31 weeks gestation because it causes air trapping(13).

Synchronized Ventilation

Neonatal microprocessor controlled ventilators have been designed which allow the delivery of mechanical breaths in response to a signal derived from the infant's spontaneous inspiratory effort (synchronised ventilation). The trigger signal can be flow or pressure (measured within the ventilator circuit nearest the infant's airway) or impedance (thoracic or abdominal). Successful triggering depends on two factors: magnitude (the amount of infant effort required to trigger) and delay (the time from initiation of infant effort and the rise in proximal airway pressure). Flow triggering is the best(14) and transthoracic impedence is worse of the available techniques(15). The modes of synchronised ventilation include the following:

Assist/Control Ventilation

Also known as patient triggered ventilation (PTV), this combines a ventilator delivered positive pressure breath in response to the infant's inspiratory effort (assist) with a guaranteed mechanical breath at a preset rate if no infant effort is detected (control). Earlier clinical experience has not shown it to be useful in infants < 28 weeks gestation, in the first 24 hours after birth, in those requiring a high Fi0₂ and ventilator rate, and in ventilator dependent infants with CLD(16). Its role might have been in I the larger, more mature infant with mild RDS during the recovery period especially when there is asynchronous breathing. One RCT has found that PTV is less successful in promoting synchrony than a fast rate regime on conventional ventilation(17). However, a RCT showed that infants of 1100-1500g with RDS on PTV utilizing a flow-derived signal to initiate and terminate a mechanical breath, had more rapid weaning that those on conventional pressure limited time-cycled ventilation(18). One study reported that PTV utilizing an airway pressure sensor is capable of providing long-term respiratory support from birth to extubation in even extremely preterm infants(19).

Synchronized Intermittent Mandatory Ventilation

In the SIMV mode, the mechanically delivered breaths are preset at a fixed rate but are synchronized to the onset of the infant's own breaths. For example, if the SIMV is 20 bpm and the infant is breathing at 60/min, the maximum number of triggered. breaths is 20. The synchrony produced by SIMV has been shown to result in a lower respiratory rate and a higher Pa0₂, and deliver larger and more consistent tidal volumes than during IMV(20-23). One RCT demonstrated benefits of SIMV over conventional IMV: a shorter duration of ventilation in infants> 2000 g, improved oxygenation in infants 1000-2000 g, and a lower CLD rate defined as supp.emental oxygen at 36 weeks postconceptional age in infants < 1000 g(24). The RCT comparing PTV and SIMV did not show any difference in wean- ing by pressure reduction (PTV) or rate reduction (SIMV)(25). Neither has it been shown that weaning by a combination of pressure and rate reduction, as can be achieved during SIMV, offer any advantage over PTV(26).

Pressure Support Ventilation (PSV)

PSV has been defined as infant initiated, pressure targeted and infant controlled ventilation, designed to assist the infant's spontaneous breathing with an inspiratory pressure 'boost'. Once the breath is triggered by the infant's inspiratory effort, a preset pressure is achieved and maintained throughout inspiration until the inspiratory flow falls below a preset value. Although, very similar to PTV, the infant has control of the inspiratory time. When combined with SIMV, PSV reduces the work of breathing which might be beneficial in weaning especially for infants with CLD(27,28). As new neonatal ventilators become available with advanced micro- processor based technology, ventilatory regimes based PTV, SIMV and PSV might prove one day to be superior to conventional IMV(29). Currently, RCTs comparing the newer ventilatory modes remain limited.

Sedation and Muscle Reiaxation

Muscle relaxants are often used in un- stable infants with RDS while on the ventilator. The concern is that spontaneous respiratory efforts which are out of phase with the ventilator breaths (active expiration) may increase the risk of PTX and intracranial hemorrhage. Sedation can be achieved with a narcotic such as morphine (100 µg/ kg q 4-6 h IM/IV, or 100 µg/kg/h over 2 hours followed by 25 µg/kg/h IV infusion), phenobarbitone, fentanyl or midazolam.(30-31). Morphine has been shown to encourage synchrony of the infant's breathing with the ventilator(32) but one RCT showed that ventilated infants given phenobarbitone had a higher incidence of air leak(33). Morphine results in small fall in heart and respiratory rates(34), but it has no significant effect on arterial blood pressure and cardiac and cerebral hemodynamics(35). Diamorphine is more potent than morphine and a loading dose of 50 µg/kg followed by an infusion of 15 µg/kg/h have been recommended in ventilated infants(36). Fentanyl (5 mg/kg loading and 1-2 µg/kg/h infusion)(37) and alfentanil (15-20 µg/kg loading and 3-5 µg/kg/h infusion)(38) are synthetic opioids which are even more potent and have the theoretical advantage of a shorter duration of action and greater cardiovascular stability. An alternative to sedation is to use fast ventilator rates (60-120/min) to achieve greater synchrony and less active expiration, or to use an IT and ET on the ventilator which matches those observed during spontaneous breating over a brief period on CDP(39). Preterm infants are capable of hormonal stress responses which might be detrimental to their outcome(40). Morphine or fentanyl infusion has been shown to reduce stress markers in preterm infants ventilated for RDS, and can be justified on humanitarian grounds for reducing the level of their pain and stress(37,40-42).

Neuromuscular paralysis is achieved with pancuronium (0.1 mg/kg q 3-4 h pm). The evidence that muscle paralysis prevents PTX, intraventricular haemorrhage (IVH) and CLD remains poor. Pancuronium should be used sparingly, one indication being infants> 34 weeks on mechanical ventilation who make vigorous respiratory efforts which are detrimental to gas exchange(43). The risks include an increase in heart rate and blood pressure, deterioration of Pa0₂ in 20% of cases, progressive worsening of lung compliance and resistance, fluid retention and disuse muscle atrophy. The deterioration after pancuronium in some infants might be a result of a reduction of minute ventilation following cessation of spontaneous respirations, but it has been shown that the decrease in oxygenation is caused by alveolar derecruitment and reduction of FRC from loss of expiratory braking nJchanisms(44). Studies which compared pancuronium with morphine or pethidine reported that neither paralysis nor sedation had an adverse effect on heart rate or blood pressure and showed no difference in the duration of ventilation and the incidence of air leak, IVH or CLD(41-45). One RCT indicated that routine paralysis offers no advantage to fast rate ventilation which is the preferred strategy for ventilating infants with RDS(46).

Weaning and Extubation

RCTs showed that PTV or SIMV can re- duce the duration of weaning compared with conventional IMV(18,24-47). Prophylactic aminophylline therapy (10 mg/kg IV followed by 2.5 mg/kg q 12h) has been shown to increase the success rate of extubation, decrease the re-intubation rate, or prevent post-extubation apnea(48). Caffeine is as efficacious as aminophylline, and has the advantages of a wider therapeutic to toxic ratio, the need for only once a day administration, and less problem with tachycardia and feed intolerance(49). Doxapram may aid in weaning when given with amonophylline in infants who are unresponsive to aminophylline alone(50).

Post-extubation nasal CP AP has been shown in two recent RCTs to result in a higher success rate of earlier extubation, lower respiratory rate, better blood gases and pH, less atelectasis and a lower re-intubation rate(51,52) but another three RCTs showed no benefit compared to extubation directly into a headbox(53-55), suggesting that the decision has been shown to reduce the problem of apnea of prematurity(56). There is no RCT evidence that dexamethasone is effective in preventing post-extubation laryngeal edema(57). However, one RCT reported earlier extubation in infants with severe RDS given dexamethasone therapy for the first 12 days after birth., starting on a dose of 1 mg/kg/ d(58). Studies which conducted pulmonary function testing to help determine the optimal timing of extubation in infants with RDS gave conflicting results as to its usefulness(59-62).

High Frequency Ventilation

Conventional mechanical ventilation uses rates up to 120/min and tidal volumes of 5 ml/kg, delivered by a ventilator with a pneumatic valve with a passive expiratory phase. High frequency jet ventilation (HFJV) uses rates of about 400/min and tidal volumes of about 3 ml/kg, delivered as compressed gas via a jet cannula with also a passive expiratory phase. High frequency oscillation (HFO) uses rates of 10- 15Hz (600-900/min) and tidal volumes of < 3ml/kg, delivered by a piston or loud- speaker to provide a continuous bias flow, and has an active expiratory phase. High frequency flow interrupted ventilation (HFFIV) mimics HFO but does not have an active expiratory phase. During high frequency ventilation, swings in alveolar pressure are extremely swing and pulmonary barotrauma is potentially reduced. Necrotising tracheobronchitis and gas trapping have been reported with HFJV but not With HFO. HFO can however result in a 14-18% reduction in left ventricular output(63). Compared to conventional ventilation, RCTs have not shown better(64), similar(65) and worse(66) outcomes with HFJV, and similar. outcome with HFFIV(67).

Six RCTs have been conducted with HFO and published in full: the first showed no benefits and a higher incidence of air leak, IVH and periventricular leukomalacia (PVL) with HFO(68); the second showed no difference in incidence of CLD but a less severe degree of pulmonary dysfunction among the CLD infants(69); the third in infants < 1750g with RDS showed a significant reduction in the incidence of CLD with no increased risk of IVH(70); the fourth in infants 750-2000g showed no difference in the rates of air leak, IVH, PVL, CLD or mortality(71); the fifth in infants with severe RDS showed a significant reduction in the incidence of air leak(72) and the sixth in infants < 36 weeks with RDS showed a number of benefits which include better oxygenation, less surfactant use, less CLD and NEC, and lower hospital costs(73), The first RCT was criticised for its low volume/ pressure strategy to HFO(74) as centers which use a high volume/pressure approach to HFO are reporting a reduction in air leak and CLD without the serious side effects such as IVH or PVL which was attributed to HFO(70,72,73). The MAP during HFO is known to be an important determinant of oxygenation(75) and earlier animal experiments have also indicated that maintenance of an adequate lung volume with a high MAP during HFO is essential to prevent atelectasis and lung injury. It was recommended that the MAP at the initiation of HFO be set at 2-3 cmH20 higher than that delivered by conventional mechanical ventilation. This MAP is measured at the junction of the ventilator circuit and the endotracheal tube, which is therefore an overestimate of mean alveolar pressure(76). Infants with atelectatic lungs who have the lowest lung volume require the greatest increase in MAP on transfer to HFO to optimise oxygenation(77).

HFJV has been shown in a RCT for treatment of pulmonary interstitial emphysema (PIE) to improve blood-gases, MAP, duration of PIE and mortality(78). An observational study also showed benefits of HFO in the rescue of infants with PIE(79). Reports also suggested benefits in using HFO for neonatal lobar emphysema, pulmonary hemorrhage and infants with respiratory failure secondary to increased intra-abdominal pressure(80-82). However, HFO is more effective in RDS or persistent pulmonary hypertension than in pulmonary airleak or pulmonary hypoplasia(83- 84). HFO has an important role as rescue . treatment in severe neonatal lung disease when conventional ventilation fails. How- ever, its routine use as a primary mode of ventilation is a contentious issue until further clinical trials are performed(85).
 

 References