Specialist neonatal respiratory care for babies born preterm

Why the committee made the recommendation

There was evidence that lower gestational age, lower birth weight, being small for gestational age, male sex, lower body temperature, sepsis, any formula feeding, surfactant use, treatment for a patent ductus arteriosus (PDA), cardiopulmonary resuscitation and mechanical ventilation, are all independent risk factors for bronchopulmonary dysplasia (BPD) in preterm babies.

No evidence was found to link antenatal steroids, chorioamnionitis, intrauterine growth restriction, ethnicity or race, or postnatal steroid use, and BPD. However, the committee did not prioritise these areas for further research.

The committee was concerned that including surfactant use and treatment for PDA as risk factors for BPD could lead to a reduction in surfactant use and PDA treatment. They agreed that there was unlikely to be a causal link – rather, the increased risk of BPD associated with these factors is more likely to reflect the severity of the baby's condition, and that surfactant should be used, and a PDA should be treated, where clinically appropriate.

The committee noted that there was an absence of evidence for certain risk factors for BPD; some evidence was for specific gestational ages at birth from which the committee was unable to extrapolate to other gestational ages, and for some risk factors, the evidence was underpowered to detect an effect. The committee therefore concluded that other gestational ages and other risk factors not listed here might also be associated with increased risk of BPD.

No evidence was found for some of the potential risk factors that had been suggested by the committee (such as necrotising enterocolitis and supplementary oxygen), but these were not prioritised by the committee for further research.

How the recommendation might affect services

Knowledge of BPD risk factors means healthcare professionals can identify preterm babies who are more likely to develop BPD, and prioritise treatment regimens accordingly. This may reduce the incidence of BPD, which will lead to long-term savings for the NHS.

Return to recommendation

Why the committee made the recommendation

The evidence did not show a clear difference between continuous positive airways pressure (CPAP) alone and invasive ventilation with surfactant when used in preterm babies in the delivery room, for any of the outcomes that the committee had prioritised (mortality, BPD and neurodevelopmental outcomes). However, the evidence showed a possible reduction in mortality before discharge, and a possible reduction in the incidence of BPD at 36 weeks' postmenstrual age with CPAP.

One large study found that just over half of the babies who received CPAP instead of intubation did need to be intubated at some point during their hospitalisation. However, the committee agreed that this was a very positive result, because around half of babies avoided all the risks of invasive intervention.

The committee agreed that it is preferable to avoid invasive ventilation wherever possible and agreed that when stabilising a preterm baby in the delivery room, the non-invasive ventilation technique of CPAP should be used rather than invasive ventilation. The committee agreed that this approach may not be suitable for some babies, for example, if the baby is not breathing and needs invasive ventilation. In addition, the committee agreed that this approach would probably not be suitable for preterm babies born very early, for example at less than 25 weeks, because these babies may not have the necessary respiratory drive, and because the failure rate of non-invasive ventilation is high in babies of this age. The committee agreed that for these very preterm babies, it may be more practical to use invasive ventilation with surfactant in the delivery room, but because this would be a clinical decision, it was not appropriate to set a particular age cut‑off.

Because there was not enough evidence to make recommendations on the use of CPAP with surfactant compared with CPAP without surfactant in the delivery room, the committee recommended that further research be done in this area.

How the recommendation might affect services

Current practice in most units is to routinely intubate preterm babies (below a certain gestation, often 27 to 28 weeks, but specific cut‑offs will vary) and give surfactant, so this will be a change in practice for these units. Because CPAP is associated with lower costs than invasive ventilation, this change is likely to lead to cost savings.

Return to recommendation

Why the committee made the recommendations

It is established clinical practice in the UK to give surfactant to preterm babies needing invasive ventilation in the early postnatal period, based on good evidence and extensive clinical experience, so the committee agreed to make a recommendation that reinforces this.

In preterm babies who do not require invasive ventilation, there was evidence that minimally invasive surfactant administration techniques reduce the incidence of BPD, the number of days on invasive ventilation, and the incidence of pneumothorax, compared with endotracheal administration.

However, not all neonatal units have the facilities to carry out minimally invasive surfactant administration techniques, and not all healthcare professionals have been trained to use them. The committee agreed that in these circumstances, endotracheal surfactant administration followed by early extubation should be used, because there was evidence that it reduces the incidence of BPD compared with conventional administration of surfactant with continued ventilation.

Because there was not enough good evidence to make recommendations on which minimally invasive administration technique leads to the best outcomes, or on different surfactant dosing regimens, the committee recommended that further research be done in these areas.

How the recommendations might affect services

Current practice for giving surfactant to preterm babies varies among neonatal units because of differences in available facilities and training. The recommendations may increase the trend towards using less invasive techniques of surfactant administration. Neonatal units that currently use conventional endotracheal administration of surfactant may therefore change practice to use minimally invasive techniques or to extubate earlier.

Return to recommendations

Why the committee made the recommendations

There was a small amount of evidence suggesting there is no difference in the effectiveness or safety of oxygen delivered by nasal cannula compared with oxygen delivered in the incubator. However, because the evidence was limited, the committee made a recommendation based on consensus that nasal cannula or incubator oxygen could be used. They noted that incubator oxygen may be preferred for short-term use and assessment, whereas nasal cannula oxygen is preferable for longer-term use because it allows the baby to be touched and cuddled, and provides a stable concentration of oxygen.

There was evidence that automated oxygen titration reduces the number of days on oxygen, reduces the number of manual adjustments for titration, and increases the time that preterm babies spend in the optimal target oxygen saturation range. However, the committee were concerned, based on their clinical knowledge, that the cumulative frequency oxygen curves for oxygen saturation achieved by automated titration may lead to the mean saturation level achieved by babies being reduced (because of the normal distribution of the frequency-saturation curve) compared with manual adjustments (where the frequency-saturation curve is skewed to the higher end of the target saturation range). The committee therefore made a recommendation for research to determine the optimal target oxygen saturation range for use in conjunction with an automated oxygen titration system.

There was no evidence comparing humidified to non-humidified oxygen, but the committee agreed that this is standard clinical practice at higher flow rates and made a consensus recommendation for the use of humidified oxygen. The committee also made a recommendation for research to assess the effectiveness of humidified and non-humidified supplemental low-flow oxygen.

How the recommendations might affect services

The recommendations reflect current clinical practice.

Return to recommendations

Why the committee made the recommendations

How the recommendations might affect services

The recommendations should reinforce current clinical practice and lead to greater consistency.

Return to recommendations

Why the committee made the recommendation

There was no evidence of benefit for inhaled nitric oxide in preterm babies who need respiratory support for respiratory distress syndrome (RDS). There was some evidence of adverse effects, and the treatment is unlikely to be cost effective. The committee agreed that there may be exceptions for preterm babies who have other conditions such as pulmonary hypoplasia and pulmonary hypertension, for whom there may be some survival benefits.

How the recommendation might affect services

The recommendation will reduce the use of inhaled nitric oxide for preterm babies who need respiratory support, which may lead to cost savings to the NHS given the high acquisition cost of inhaled nitric oxide.

Return to recommendation

Why the committee made the recommendations on dexamethasone

There was evidence that in babies 8 days or older, dexamethasone reduces the incidence of BPD, but dexamethasone is associated with an increased risk of hypertension. There were no clinically important differences in mortality before discharge between babies who received dexamethasone and those who did not.

There was some evidence suggesting that dexamethasone reduces the number of days on invasive ventilation, but this was difficult to interpret because of the different ways the studies reported the time on ventilation.

In babies 8 days or older, there was evidence that dexamethasone is not associated with an increased risk of cerebral palsy, neurodevelopmental delay, neurosensory impairment or gastrointestinal perforation. However, the committee emphasised that there was some uncertainty about the evidence for these outcomes.

In babies younger than 8 days, there was evidence that dexamethasone reduces the incidence of BPD but is associated with an increased risk of gastrointestinal perforation.

The committee therefore recommended that dexamethasone be considered for babies 8 days or older who still need ventilation for respiratory disease, after taking into account risk factors for BPD. This is in line with current practice, which is to use corticosteroids to assist weaning from ventilatory support when a baby is 8 days or older, rather than using corticosteroids as 'prophylaxis' for babies younger than 8 days old.

The committee agreed the importance of discussing the risks of gastrointestinal perforation, hypertension and cerebral palsy with parents and carers before starting dexamethasone therapy, because there may be lifelong implications for the baby and their family.

Although the combination of dexamethasone and non-steroidal anti-inflammatory drugs (NSAIDs) was not reviewed, the committee confirmed that they should not be used together because this increases the risk of gastrointestinal bleeding and perforation. The committee agreed that although this risk is widely recognised, it should be reinforced in the guideline to ensure that dexamethasone and NSAIDs are not used together in clinical practice.

Because of the increased risk of hypertension with dexamethasone, the committee recommended that babies' blood pressure should be monitored. There was no evidence about when or for how long to monitor blood pressure, so the committee agreed that this should be decided by the neonatologist responsible for the baby's care.

The evidence did not show any differences between different dosing strategies, and so the committee did not make any specific dosing recommendations.

Why the committee didn't make any recommendations on hydrocortisone and nebulised budesonide

Evidence comparing hydrocortisone and placebo was inconclusive, so the committee did not make any recommendations. The committee was aware that there is an ongoing, large multicentre randomised controlled trial investigating hydrocortisone compared with placebo in preterm babies who need respiratory support, so did not make a recommendation for research that would replicate this study. However, they agreed that a comparison of dexamethasone and hydrocortisone could provide useful guidance and so made a recommendation for research for this comparison.

There was very little evidence for the use of nebulised budesonide and therefore the committee made a recommendation for research.

How the recommendations might affect services

Current practice is to use corticosteroids in preterm babies to assist weaning or removal from ventilatory support, but they are not routinely used to prevent BPD in all preterm babies. The choice of dexamethasone or hydrocortisone varies among neonatal units. These recommendations are unlikely to affect how often corticosteroids are used, but they might prompt units who currently use hydrocortisone to consider dexamethasone as an alternative.

Return to recommendations

Why the committee did not make any recommendations

The evidence on the use of diuretics in preterm babies on respiratory support was very limited. None of the studies identified assessed critical outcomes such as mortality before discharge, BPD or neurodevelopmental outcomes. Although the studies looked at short-term adverse effects associated with diuretics, it was not clear whether there was an increased risk of adverse effects because of the small sample size of the studies.

Because of the limited evidence and lack of clinical consensus, the committee could not make any recommendations for or against diuretic use in preterm babies on respiratory support. Instead, the committee recommended that further research be done in this area.

How the recommendations might affect services

Although they did not make any recommendations, some of the committee members thought that the lack of evidence identified may lead to healthcare professionals reviewing their use of diuretics. This may lead to a reduction in the use of diuretics in preterm babies on respiratory support, at least until further evidence is available.

Return to recommendation

Why the committee made the recommendations

There was evidence that in preterm babies born before 31 weeks, caffeine citrate reduces the incidence of BPD, cerebral palsy (at 18 to 21 months' follow‑up) and blindness (at 11‑year follow‑up) compared with placebo. However, from their clinical experience, and based on the inclusion criteria of the CAP study, the committee agreed that it was appropriate to start babies, born at or before 30 weeks (equivalent to babies born at or below 1.25 kg), on caffeine citrate on admission to the neonatal unit, so they recommended earlier initiation. Based on their clinical experience, the committee agreed that administering caffeine citrate would also reduce apnoea in older preterm babies.

There was evidence that, compared with lower doses, higher doses of caffeine citrate reduce the incidence of BPD, continued apnoea and extubation failure.

Evidence showed that the treatment with caffeine citrate before 3 days of age may lead to a reduction in BPD. There was also evidence that treatment with caffeine citrate for 15 to 30 days reduces the incidence of BPD compared with a shorter duration, and that treatment for longer than 30 days reduces the incidence of necrotising enterocolitis compared with treatment for less than 15 days.

To determine when caffeine citrate should be stopped, the committee referred back to the studies and identified the age at which caffeine citrate was started, the duration of caffeine citrate, and hence the age at which it had been stopped. The committee noted that caffeine citrate had been stopped in the studies at between 33 and 35 weeks. This reflected the clinical experience of the committee as the age at which preterm babies were no longer expected to have apnoea, and so this figure was used by the committee to develop their recommendations.

The committee made their dosing recommendations based on evidence that a higher dose is more effective than a lower dose, and on currently recommended doses used in clinical practice. However, the variation in loading and maintenance doses used across different clinical trials made selecting an optimal dose difficult, and although higher doses appeared to improve early outcomes, there were few data on long-term outcomes. For this reason, the committee recommended further research to identify the maintenance dose of caffeine citrate needed to optimise neurodevelopmental outcomes.

The committee also discussed whether monitoring caffeine citrate levels was necessary and noted that the Evelina London Paediatric Formulary advises that babies can receive 10 mg/kg of caffeine citrate twice daily without monitoring blood plasma levels (Evelina London 2015). The committee noted that there are units that do not currently monitor blood levels, and increasing doses to higher than 20 mg/kg daily may be a concern if units did not test blood levels at these higher doses. Therefore, the committee made an additional recommendation that if apnoea persists and a baby receives more than 20 mg/kg daily, caffeine citrate levels should be tested.

How the recommendations might affect services

The recommendations will have a minimal impact on current practice. The committee noted that there is some variation in dosage regimens across the NHS, so these recommendations should lead to greater consistency in the choice of dosage regimens. The committee agreed that oral caffeine citrate is used in most cases, and has a low acquisition cost. The committee acknowledged that intravenous caffeine citrate has a substantially higher acquisition cost. However, because of the single-use vial sizes available, irrespective of dose, a single vial will be required for administration of each dose, so increasing the dose would not increase the costs. In addition, only a small number of babies would need the intravenous solution. Furthermore, there may be a small increase in the number of blood tests performed to assess caffeine citrate levels if higher doses are used, but again the number of babies who will require high doses is small. Overall, the recommendations may result in a slight increase in drug and monitoring costs, but this is not anticipated to be substantial.

Return to recommendations

Why the committee made the recommendation

There was no evidence of benefit from treating a PDA, and there was evidence for potential harms from treating it, with either medicines or surgery. However, the committee agreed that for some babies, treatment might be appropriate, for example, if there is difficulty weaning the baby from a ventilator.

How the recommendation might affect services

The recommendation will reduce the unnecessary treatment of PDA and the number of babies exposed to potential harms from its treatment. The recommendations may result in cost savings because fewer procedures will be carried out.

Return to recommendation

Why the committee made the recommendations

The evidence on the best method for measuring oxygen levels in diagnosing hyperoxia or hypoxia in preterm babies was very limited. There were no studies assessing the diagnostic accuracy of SpO₂ (peripheral capillary oxygen saturation) compared with the standard PaO₂ (partial pressure of arterial oxygen) that met the review's inclusion criteria. The committee agreed, based on clinical consensus and their experience of clinical practice, that SpO₂ should remain the first-line method for continuous monitoring of oxygen saturation levels in preterm babies because of its widespread availability and non-invasive nature. The committee agreed that arterial sampling of partial pressure of oxygen remained the 'gold standard', but is not always possible and can never be continuous.

The only evidence on tcPO₂ (transcutaneous oxygen) was 1 study from the 1970s, and the way this procedure is performed has changed substantially since then. However, tcPO₂ is currently used in clinical practice, and in the committee's experience it can provide useful information. This is particularly the case for preterm babies on invasive ventilation who are clinically unstable and need continuous monitoring to guide management, and in whom SpO₂ may not be accurate.

Because of the lack of good evidence, the committee agreed that further research needs to be conducted looking at the diagnostic accuracy of tcPO₂ and SpO₂ against the gold standard arterial oxygen saturation in diagnosing hyperoxia and hypoxia in a preterm baby population.

There was evidence that higher target oxygen saturation levels reduce mortality. Although a higher target is associated with an increase in retinopathy of prematurity and an increased risk of BPD, the evidence suggested no increase in severe visual impairment at 18 months, and the reduction in mortality was considered to offset the increased risk of BPD. The committee were aware that target oxygen levels (up to 97%) may be more beneficial but there was no evidence to support this, so they made a recommendation for research.

How the recommendation might affect services

The recommendations reflect current practice, where SpO₂ is generally used as routine continuous oxygen monitoring in preterm babies, and tcPO₂ is reserved for the more clinically unstable preterm babies as a continuous monitoring tool.

Many units already use 91% to 95% as their target saturation level for preterm babies, but for those that do not, this will be a change in practice. This will reduce the variation in clinical practice.

Return to recommendations

Why the committee made the recommendations

The evidence showed no differences in the outcomes measured between higher and lower target ranges for the partial pressure of carbon dioxide in preterm babies on invasive ventilation. The committee recognised that the higher target ranges specified in the studies were in line with the definition of permissive hypercapnia and would probably not have any detrimental effects on clinical outcomes and long-term neurodevelopmental outcomes. In view of this, the committee agreed that when healthcare professionals are monitoring carbon dioxide levels in preterm babies on invasive ventilation, a higher target range would be acceptable. This avoids the need for frequent adjustment of the ventilators to reach an extremely tight target range.

There was variation in the target ranges of carbon dioxide used by different studies, and the range of days at which at different permissive hypercapnia levels were tolerated. The committee agreed to make a recommendation in line with the largest and most recent study that looked at clinical and long-term neurodevelopmental outcomes, but simplified the 3-stage ranges (days 1 to 3, days 4 to 6 and day 7 onwards) used in this study, to a 2-stage range based on their clinical experience that the difference in upper limits tolerated would be negligible and would have minimal detrimental effects on a preterm baby on invasive ventilation.

There was no evidence on the action to be taken when a low carbon dioxide level was detected, but the committee were aware that this was a dangerous situation, so agreed the action to be taken based on their clinical knowledge and experience.

All the evidence for the optimal target range of carbon dioxide was in preterm babies on invasive ventilation. The committee recognised the lack of evidence in preterm babies on non-invasive ventilation, so they recommended further research in this area.

How the recommendations might affect services

The recommendations reflect current practice, both where permissive hypercapnia is accepted in the monitoring of carbon dioxide levels in preterm babies on invasive ventilation, and for the action to be taken if hypocapnia is detected.

Return to recommendations

Why the committee made the recommendation

There was no good evidence to define the normal range of blood pressure in preterm babies, or how blood pressure should be measured. The committee wanted to make healthcare professionals aware of this lack of evidence to prevent unnecessary treatment based on the level of blood pressure only. The committee advised, based on their clinical experience, that inadequate perfusion should be treated with the aim of increasing perfusion, and not to aim for a particular blood pressure target.

Because there was no good evidence, the committee made recommendations for research to determine both the optimal blood pressure target and method of measuring blood pressure in preterm babies.

How the recommendation might affect services

For units that routinely monitor blood pressure in preterm babies and treat when blood pressure falls outside certain limits, this will be a change in practice. The recommendation will lead to less unnecessary monitoring and treatment of blood pressure.

Return to recommendation

Why the committee made the recommendations

The evidence showed that there was no difference in mortality before discharge in babies who received morphine compared with placebo. Babies receiving morphine took longer to achieve full enteral feeding, and babies born at 27 to 29 weeks' gestation had an increased risk of severe intraventricular haemorrhage (IVH). There was some evidence that, when compared with placebo, morphine improves sedation and pain scores in preterm babies who need invasive respiratory support during infusion. However, moderate quality evidence from a larger study showed no difference in pain scores during endotracheal suctioning between babies who received morphine compared with placebo.

The only evidence available comparing morphine with fentanyl showed no clinically significant difference in rates of severe IVH.

There was some evidence that, when compared with midazolam, babies receiving morphine may have decreased rates of severe IVH.

Babies receiving morphine experienced less pain during infusion, but less sedation after infusion.

Because of the mixed evidence regarding the effectiveness of morphine and taking into account the risks, the committee agreed that morphine should not be used routinely, but may be considered when it is clear the baby is in pain, and that neonatal units would have their own guidelines or preferred scales to determine pain in preterm babies.

The committee discussed other concerns about using morphine, such as suppressed respiratory drive and opioid dependency. They agreed that regular reassessments are important to ensure that morphine is stopped as soon as appropriate.

The committee did not make any recommendations for paracetamol or non-pharmacological interventions because there was no evidence available. Instead, the committee recommended that further research be done to compare morphine with containment holding during respiratory support, because the committee agreed that containment holding may improve outcomes in preterm babies, with a reduced risk of adverse events compared with pharmacological therapy.

How the recommendations might affect services

Use of sedation and analgesia currently varies among units. The recommendations will have little impact in units that do not routinely use morphine, but other units may need to change practice and this may lead to a reduction in the use of morphine. The recommendations will make practice more consistent across the NHS.

Return to recommendations

Why the committee made the recommendations

There was some evidence from small, single studies that using an analgesic with a neuromuscular blocker, or an anaesthetic such as propofol used alone, is an effective regimen to achieve successful intubation in preterm babies, while avoiding adverse effects.

However, there was a lack of evidence to show exactly which medicines or classes of medicines form the best combination, so the committee recommended that healthcare professionals should consider premedication before elective intubation and recommended that further research be done in this area.

How the recommendations might affect services

Current practice of using premedication for elective intubation in preterm babies varies among units. Units that currently use single medicines (such as morphine or fentanyl) may need to change practice to follow the recommendation. The recommendation will make practice more consistent across the NHS.

Return to recommendations

Why the committee made the recommendations

There was good evidence that using a dummy (non-nutritive sucking) during nasogastric feeds reduces the length of the baby's hospital stay. In addition, there was some evidence that the Newborn individualized developmental care and assessment program (NIDCAP^®) improved neurodevelopmental outcomes relating to cognitive development and was a cost-effective intervention in babies born at less than 27 weeks. Although the evidence for skin-to-skin contact did not show any benefit, there was no evidence of harm. There was no evidence that Family integrated care (FIC) provided any additional benefits compared with standard care.

Based on their experience and the clinical evidence, the committee recommended explaining to parents and carers about the potential benefits of interacting with their baby because early social development and relationship-forming are key to successful emotional and behavioural development.

Because of the limited evidence available on FIC and NIDCAP^®, the committee made it a priority to recommend that further research be done to investigate the potential impact of NIDCAP^® and FIC on length of stay and BPD.

How the recommendations might affect services

The committee agreed that the recommendations on non-nutritive sucking and using positive touch (such as skin-to-skin contact) would not result in a major change in practice, but will help improve consistency in best practice.

Although there are cost implications for units to train professionals in NIDCAP^®, the recommendation to consider NIDCAP^® would lead to a more consistent approach across neonatal care networks to practice linked with neurodevelopmental care. It would also improve babies' access to this type of neurodevelopmental care and allow greater involvement of parents and carers in the care of their baby.

Return to recommendations

Why the committee made the recommendations

How the recommendations might affect services

The committee agreed that the recommendations would not result in a major change in practice, but will help improve consistency in best practice.

Return to recommendations

Why the committee made the recommendations

There was evidence that parents and carers value having 24‑hour access to the neonatal unit, which should be a homely environment with comfortable furniture and private areas. In a number of the support and information themes, parents and carers agreed that healthcare professionals who provide information and support should be trained and competent in this, so the committee made an overarching recommendation.

How the recommendations might affect services

The committee agreed that the recommendations would not result in a major change in practice, but will help improve consistency in best practice.

Return to recommendations

Why the committee made the recommendations

There was evidence about the importance of good communication with parents about their baby's discharge. The committee agreed that a designated neonatal discharge coordinator, as a single point-of-contact, would facilitate the communication of key information with parents and carers. The committee also agreed that early referral to community and continuing healthcare teams would also help parents prepare for their baby's discharge. Having the option to discharge to an alternative location, such as to another relative's home or a hospice, would enable parents and carers whose homes are not suitable for their preterm baby to be able to care for their baby outside the hospital.

The committee also recognised that some of the advice in the NICE guideline on postnatal care was also relevant to babies born preterm and so made a cross reference to this guideline.

However, because there were only 2 studies, and no evidence for a number of themes identified by the committee, the committee agreed that more research could better define best practice, and so made a recommendation for research.

How the recommendations might affect services

The committee agreed that the recommendations would not result in a major change in practice, but will help improve consistency in delivering best practice.

Return to recommendations

Why the committee made the recommendations

There was evidence that parents and carers value having support and information about their baby's routine care, being involved in preparing for the baby's discharge, and having information on equipment, identifying illness in their baby, and dealing with emergencies. Parents and carers also value information about future care, such as contact details, follow‑up appointments and immunisations, ongoing peer support and self-care for problems such as postnatal depression.

How the recommendations might affect services

The committee agreed that the recommendations would not result in a major change in practice, but will help improve consistency in delivering best practice.

Return to recommendations