Discussion
Critically ill newborns and infants, admitted to the PICU for severe bronchiolitis, are typically affected by severe hypoxemic respiratory failure, generally refractory to HFNC oxygen administration17-18. Even if invasive mechanical ventilation can be considered the most aggressive approach for the severe cases 19, NIV support has been repeatedly proposed in order to assist such patients and to reduce intubation rate20-21. The Italian Network of Pediatric Intensive Care Units (TIPNET) reported a national intubation rate of 15% among critically ill children with bronchiolitis in 2010-201622, regardless which NIV interface was applied16. These findings were confirmed by the present study. Nevertheless, there is still a lack of multicenter, prospective, and randomized studies to determine the most effective and well-tolerated type of non-invasive support for these patients. Furthermore, among many studies on NIV in children with bronchiolitis admitted to the PICU, information about which optimal inspiratory and expiratory pressure levels should be administered is currently lacking23. Finally, HCPAP treatment in children with respiratory failure has rarely been evaluated, potentially due to the fact that few centers are specifically skilled at this technique and routinely apply it 24.
Although this study was interrupted due to organizational issues further complicated by the Coronavirus Disease 2019 pandemic, results carry several implications. We found that during HCPAP a PEEP level of 10 is not likely to reduce intubation rates compared to PEEP 5. However, children with respiratory insufficiency due to severe bronchiolitis in the PICU seem to require intubation when bacterial coinfection worsens respiratory insufficiency. These patients likely represent a subgroup of particularly ill subjects who may be resistant to CPAP support. The theoretical premise of this study was that the use of a PEEP level of 5 cmH2O might be ineffective as a respiratory strategy in comparison to a level of 10 cmH2O. As a matter of fact, PEEP escalation proved effective in 38 over 47 P5 patients (75%) and clinicians reported improvement of symptoms in all such cases. The remaining 9 patients who did not show a benefit from escalation were all eventually intubated. These patients may be represented by a subgroup of particularly severe cases and were interestingly a very similar number then the P10 group, mostly being bacterial coinfections. There are many potential explanations for this finding. Positive pressure during HCPAP is delivered through a high flow of an air/oxygen mixture and a mechanical spring valve and without the use of the ventilator. Furthermore, the plastics of the helmet have relatively high compliance that, in part, explain the high tolerance of this method25. With these premises, the system may not accurately deliver the exact pressure level that is set with the valve. Therefore, it is possible that a setting of 5 cmH2O does not confer enough pressure to effectively recruit distal bronchioles. The rapid relief provided by an adequate HCPAP setting is due to the resolution of ‘air trapping’, after provision of a positive pressure level equivalent to intrinsic PEEP. Oxygenation, CO2 levels and respiratory efforts rapidly improve; hence, the weakening of respiratory muscles and patients’ exhaustion are avoided. Since the clinicians optimized the respiratory parameters of the two groups in this study, it is not surprising that, over time, collected clinical variables were not different.
Another point that is worth remarking about our partial results is that despite being insufficient in most patients with a failure rate of 78%, an initial PEEP of 5 was not detrimental in the sense that it did not lead to a higher intubation rate. In other words, this study suggests that although it is likely not enough support, starting with a low PEEP and titrating the support may be at least not harmful.
This study also confirmed that HCPAP, in general, and a PEEP level of 10 cmH2O, in particular, are not associated to pneumothorax. Furthermore, enteral feeding intolerance was low. The efficacy of a PEEP above 6 cmH2O to manage acute respiratory failure has been previously described 26, and its use in the pediatric population has been also confirmed 27. However, several concerns have been expressed regarding side effects of high PEEP on smaller children; our study provides a significant contribution against them. It can be speculated that a lower level of PEEP during HCPAP could be delivered to non-critical bronchiolitis children in pediatric wards or emergency rooms in a timely manner and that ‘escalated therapy’ could be reserved for cases of clinical worsening 28. Further studies exploring this strategy are warranted to explore this strategy in order to improve the treatment of these patients and anticipate the ventilatory support29.