Endotracheal intubation with positive pressure ventilation causes lung damage in very preterm infants. Newer ventilatory strategies have made little impact on rates of chronic lung disease.1 Alternative less invasive techniques such as nasal intermittent positive pressure ventilation (NIPPV)—that is, continuous positive airway pressure (CPAP) with superimposed inflations, are becoming increasingly popular.2 3

NIPPV has been shown to reduce reintubation after extubation in premature infants4 but many other aspects of this treatment have not been evaluated in good clinical trials—for example, its mechanism of action, clinical indications, best mode of delivery, best interface, need for synchronisation, optimal settings, complications, weaning and long-term effects.

Without good evidence, neonatal units have used NIPPV based on personal experience and anecdotal reports. This survey aimed to determine how widespread the use of neonatal NIPPV has become in England and to describe the range of practice used.

METHODS

An email questionnaire was sent to the lead neonatologist of 95 level three neonatal intensive care units in England (identified from the BLISS database (the Premature Baby Charity, London, UK)). If no response was received, repeat emails were sent followed by a postal questionnaire, finally a second neonatologist was approached. We asked about intensive care cot numbers and use of NIPPV. We asked:

• Was NIPPV used? What pressure source and interface were used?

• Was a synchronisation device used for NIPPV delivery?

• For which clinical indications was NIPPV used and did guidelines apply?

• What complications were seen?

• What were the typical settings?

• How was NIPPV weaned?

Answers were tick boxes or numerical, with free text space for comments.

RESULTS

Ninety-one (96%) of 95 questionnaires were returned—44/91 (48%) after the initial email, increasing to 60 (66%) after a second email, 72 (79%) after a postal questionnaire and 96% after contact with a second neonatologist. Although all units contacted were identified as level 3 by the database used, 38 (42%) described themselves as level 2 and seven (8%) did not identify their level. Eighty-nine respondents (98%) supplied cot numbers, 73/89 (82%) units had at least four intensive care cots, and 26/89 (29%) had at least eight.

NIPPV was used by 44/91 (48%). Table 1 shows the numbers of units using different NIPPV pressure sources, interfaces and systems designed to synchronise with infant inspirations. Most units used custom-designed devices for NIPPV delivery; Infant Flow Advance, or SiPAP (both Viasys Healthcare, CA, USA) together with a synchronisation device. All but one used binasal prongs, many also used nasal masks. Some units were using more than one pressure source and interface.

NIPPV was used routinely after extubation by 26/44 (59%), it was used for “rescuing” babies for whom CPAP failed by 35/44 (80%), including treatment of apnoea, and as a first-line mode of support in the treatment of respiratory distress by 7/44 (16%). Four (9%) units had developed formal guidelines for NIPPV use, two used NIPPV for infants below a particular weight or gestation and three said they had only just started to use NIPPV and were still establishing their indications.

Forty units (91%) reported no complications from NIPPV, three (7%) did not provide this information and one unit, using non-synchronised NIPPV, reported that very small infants appeared to develop abdominal distension during NIPPV, spending longer on parenteral nutrition while establishing full enteral feeds. No reports of gastrointestinal perforations were received.

Units were asked to specify the typical values they set, when using NIPPV. During NIPPV the end expiratory pressure (PEEP), peak pressure (PIP), inflation rate and inflation time can all be manipulated. Inflation rates may simply be “back up” rates should apnoea occur—for example, with the Infant Flow Advance and SiPAP machines, or “true” set rates during ventilator-delivered NIPPV. A wide range of values were seen (table 2).

When weaning NIPPV, 21 (48%) units reduced the PIP towards PEEP and 29 (66%) reduced the number of mechanical inflations per minute, six (14%) used a combination of both methods. One unit simply stopped NIPPV, from peak pressures of 20 cm H₂0, PEEP of 5 cm H₂0 and a rate of 20/min.

DISCUSSION

NIPPV is used by almost half of the level 3 neonatal intensive care units in England, with several units having started to use it recently.

Despite little evidence about the relative benefits of different types of delivery device the majority used variable flow devices designed to deliver NIPPV via nasal prongs. Short binasal prongs are likely to be more effective than nasopharyngeal prongs for CPAP delivery as they allow greater transmission of the applied pressure and less resistance to flow.5 Clinical practice was consistent with these findings, with virtually all units using short binasal prongs. Nasal masks were also used in half of the units surveyed, though data about their efficacy or safety are minimal.

Although there is no evidence about the relative efficacy of synchronised versus non-synchronised NIPPV, the majority used devices designed to synchronise with the infants’ breathing movements, detected by an abdominal capsule. It has been reported that non-synchronised NIPPV is associated with a high rate of gastrointestinal perforation.6 Neither this complication, nor other serious complications, were reported in this survey.

NIPPV was most commonly used for treating apnoea; however, conflicting evidence exists about its use for this indication. Systematic review showed no advantage over the use of CPAP,4 but the studies evaluated were small (total 54 infants) and used non-synchronised NIPPV. More data are needed to answer this question. Sixty per cent used NIPPV after extubation to prevent reintubation. Evidence for this indication is from a systematic review4 analysing three randomised controlled trials (RCTs), totalling 159 infants. One in six units was using NIPPV as a primary mode of treatment for respiratory distress. No RCTs have evaluated NIPPV for this indication, the only published evidence has been observational.2 3

There are no satisfactory data in the literature about the optimal rate or pressure settings during NIPPV, or how to wean infants from NIPPV. This survey showed that a wide range of settings are used, particularly in the choice of PIP (7–20 cm H₂O) and inflation rate (5–60/min, or assist/control mode). It was uncommon for the PEEP to be above 6 cm H₂O or for the PIP to be above 14 cm H₂O. The latter may be because PIP is limited by some delivery devices (Infant Flow Advance: maximum 11 cm H₂O, SiPAP: maximum 15 cm H₂O). There is no evidence that respiratory support with these pressures is better than CPAP, as the RCTs investigating apnoea and reintubation rates used a PIP in the range 16–30 cm H₂O.4

Little is known about the mechanism of action of NIPPV. No randomised studies have looked at long-term outcome or been powered to look for complications. Further research is also urgently needed in this area.

CONCLUSION

This survey reports the way in which NIPPV is being used in English neonatal intensive care units. These results cannot be used as a guide to NIPPV management.

Historically, neonatal medicine has many examples of new treatments enthusiastically espoused without sufficient evidence. Where good evidence exists current practice appears to be consistent, but the wide range of clinical approaches in areas identified by this survey highlights the paucity of evidence available in this field. Research studies and clinical trials of NIPPV are urgently needed to establish best practice and reduce the chance of harm.