Non-invasive Continuous Positive Airway Pressure (nCPAP) is widely recognized as an efficient respiratory support in infants with mild to moderate Acute Hypoxemic Respiratory Failure (AHRF).

Its application results in alveolar recruitment, inflation of collapsed alveoli, and reduction of intrapulmonary shunt. nCPAP is traditionally delivered with nasal prongs, nasal/facial mask. CPAP by helmet was introduced more recently in the clinical practice.

The helmet circuit was described in details in previously published studies. From a physiological point of view the helmet circuit could be considered the best system to deliver CPAP because of the following: 1) it is characterized by the lowest amount of leaks around the interface and mouth opening 2) airways are free from potentially obstructing devices (cannula) thus the resistance is minimized and 3) theoretically the pressure is more stable minimizing the leaks 4) it is comfortable and usually sedation is not needed.

High Flow Nasal Cannula (HFNC) is increasing in use both in adults and pediatric population. HFNC could result in several clinical benefits by reducing inspiratory effort and work of breathing, increasing end-expiratory volume and CO2 wash-out for upper airways and creating a CPAP effects of 2-3 cmH2Oin the upper airways. This CPAP effect combined with an increase in CO2 wash-out and optimal airways humidification could decrease the respiratory work of breathing and improve gas exchange. However little is known about the optimal flow rate setting to improve the respiratory mechanics and gas exchange. Recent studies have reported that HFNC in nonintubated children improves oxygenation, reduces the respiratory drive and prevent reintubation in high patient risk. However all these physiological effects during HFNC therapy are only speculative.

To address the question on the more efficient devices to support the child in the early phase of mild to moderate AHRF, the Authors designed a physiological randomized crossover study aimed at measuring the physiological effects of HFNC 2 and 3 l/Kg and helmet CPAP on the work of breathing (estimated by the esophageal Pressure Time Product, PTPes) in pediatric AHRF.

Primary end point of the study is the variation in esophageal Pressure Time Product (PTPes) across the trials of HFNC at different flow rates (2and 3 l/kg/min) and CPAP by helmet.

Study design

Physiological crossover prospective study comparing three study trials (20 mins) delivered in computer generated random order:

1- HFNC flow 2l/Kg; 2- HFNC flow 3l/Kg; 3- Helmet CPAP (fresh gas flow 35 l/min, CPAP 6 cmH2O)

Four 20 mins wash out period between trials are imposed to avoid the carry over effect one treatment on the following.

Sample size calculation and statistical analysis

The sample size was established to detect a difference at least of 30% in the primary end point, i.e. the PTP esophageal, with a Type 1 error of 0.05 and a desired power of 90% between HFNC trials compared with baseline value on oxygen mask (PTPes mean value 250 ± 65 cmH2O/s with oxygen mask vs PTPes 173 ± 62 cmH2O/s with HFNC).

The distribution data were determined using the Kolmogorov-Smirnov analysis. Normally distributed variables are expressed as mean (SD) while median and interquartile range are used to report non-normally distributed variables. Differences between variables across different treatment are tested by one-way ANOVA for repeated measures with post hoc Bonferroni comparison. Significance was taken as p < .05.

Demographic data collection and patients monitoring

At enrolment the following variable are collected: sex, age, weight, PRISM III, etiology of AHRF, comorbidities, hours before study, PICU and hospital outcome. All patients are monitored as follow: tcPO2, tcpCO2, SpO2, and EKG continuously; arterial blood pressure every 15 min; COMFORT score.

Experimental protocol.

Patients are kept in semirecumbent position. Sedation, if needed, is provided according to PICU protocol (dexmedetomidine 0.5-0.7mcg/Kg/hour) to maintain a COMFORT score between 17 and 26. The attending physician evaluated treatment failure or success during stabilization period and a PICU senior consultant not involved in the study was always present for monitoring and treating potential adverse events.

Inspired Oxygen Fraction (FiO2). FiO2 is set to obtain a peripheral oxygen saturation > 94% and then kept constant during all the study for each devices.

High Flow Nasal Cannula. In all patients, HFNC is delivered through specific pediatrics nasal prongs. FiO2 is chosen by the attending physician to target a peripheral saturation of 90-96% during oxygen facial mask breathing and kept constant during all phases. The set FiO2 during each phase is measured using a dedicated system connected to nasal cannulas.

Helmet CPAP. Helmet CPAP is delivered by high fresh gas flow circuit with helmet. The pediatric helmet is made of transparent latex-free polyvinyl chloride and is secured to a soft collar that adheres to the infant's neck. One helmet port is connected to gas source and the other to an underwater CPAP valve. There are two safety systems: a pressure monitoring device with overpressure safety valve and an anti-asphyxia valve. High fresh-gas flow (.35 L/ minute) was used to avoid CO2 rebreathing.

Esophageal Pressure monitoring A nasogastric tube equipped with an esophageal balloon is advanced through the nose to reach the stomach and inflated with 1 ml air.

The intragastric position is confirmed by the positive pressure deflections during spontaneous inspiration. The catheter is then withdrawn into the esophagus, as indicated by the appearance of cardiac artifacts and negative swings of pressure tracings during inspiration, and fixed. Waveforms of the esophageal pressure were recorded for 5 min at the end of each study phase and before starting the next one by a dedicated data acquisition system.

Randomization. Concealed randomization is conducted centrally through a computer generated block-randomization schedule. A phone-call service is available h 24/7 for patients' assignments to related group. The attending physician is not involved in the study. Medical treatment for infants with acute bronchiolitis remains unchanged for the study purpose as per standard hospital protocol.

Protocol interruption criteria. The experimental protocol will be interrupted in case of treatment failure and patient will be managed according to attending physician judgement (thus including an approach with non-invasive pressure support ventilation as intermediate step before endotracheal intubation). Criteria for endotracheal intubation includes: a-failure to maintain paO2>60mmHg with FiO2<0.6; b-clinical signs of exhaustion; c- need to protect airways and/or manage copious tracheal secretions; d-hemodynamic impairment.

Inclusion Criteria:

P/F<300 with oxygen mask (FiO2=0.4) for 15 minutes plus two of the followings:

• Respiratory Rate (RR)>2SD according to age
• Active contraction of respiratory muscles
• Paradoxical abdominal motion

Exclusion Criteria:

• Emergency need for intubation
• Glasgow Coma Scale <12
• Hypercapnia with pH <7.25
• Cough reflex impairment
• Upper-airway obstruction
• Facial/gastric surgery
• Recurrent apnoeas
• Hemodynamic instability (need for vasopressor or inotropes)
• Pneumothorax on lung echo or chest x ray
• esophageal surgery.