Gdmedical.ch

Journal of Perinatology (2007) 27, 85–91 r 2007 Nature Publishing Group All rights reserved. 0743-8346/07 $30 ORIGINAL ARTICLEHigh flow nasal cannula versus nasal CPAP for neonatal respiratory disease: a retrospective study MT Shoemaker1, MR Pierce2, BA Yoder3 and RJ DiGeronimo11Wilford Hall USAF Medical Center, Lackland AFB, TX, USA; 2Pediatrix Medical Group, Santa Rosa Children’s Hospital, San Antonio,TX, USA and 3Primary Children’s Medical Center, University of Utah, Salt Lake City, UT, USA cost. One report put the cost at 4.4 billion dollars per year in the Objective: Our objective is to assess the frequency of usage, safety and United States, and this only includes the time in the neonatal clinical utility of humidified high flow nasal cannula (HHFNC) in two tertiary intensive care unit.1 The current practice of most neonatologists care hospitals and compare outcomes to a historical control group of premature reflects the belief that limited exposure to invasive mechanical infants who received nasal continuous positive airway pressure (NCPAP).
ventilation and careful use of oxygen support result in less lung Study design: The first part of the study describes the increased HHFNC injury and improved long-term pulmonary outcomes in premature usage in two tertiary neonatal intensive care units. The second part infants.2,3 Given this, there is now a concerted effort in many compares outcomes of infants, born at less than 30 weeks gestation, who practices to avoid the use of prolonged invasive ventilatory support received either NCPAP or HHFNC as an early respiratory support mode.
when treating acute respiratory distress in premature infants withearly application of nasal continuous positive airway pressure Results: HHFNC usage increased (64%) after its introduction in infants (NCPAP) either immediately after birth or following a brief period of all gestational ages whereas the usage of NCPAP decreased from 19 to of intubation, mechanical ventilation and dosing with 4%. Ninety-five percent of infants born at less than 30 weeks gestation received HHFNC at some point during their hospital stay whereas only An alternative to the use of NCPAP as a non-invasive modality 12% received NCPAP. There were no differences in death or to support respiratory distress in premature infants has been the bronchopulmonary dysplasia (BPD), but ventilator-days per patient were recent introduction of humidified high flow nasal cannula decreased (19.4 to 9.9) following introduction of HHFNC. Comparing the (HHFNC) devices in many units. Although its use has been widely cohort of infants who received either NCPAP or HHFNC as an early mode adopted, there have been only a few published abstracts describing of respiratory support, there were no differences in deaths, ventilator-days, its use in the neonatal population and no clinical trials using flows BPD, blood infections or other outcomes. More infants were intubated for >2 l/min.9–13 The effects of the introduction of HHFNC on failing early NCPAP compared to early HHFNC (40 to 18%).
outcomes such as duration of supplemental oxygen, mechanical Conclusions: HHFNC was well-tolerated by premature infants.
ventilation-days, bronchopulmonary dysplasia (BPD), hospital Compared to infants managed with NCPAP, there were no apparent length of stay (LOS) and mortality, among other measures, have differences in adverse outcomes following the introduction of HHFNC.
not been formally assessed. Furthermore, in the absence of an Additional research is needed to better define the utility and safety of accepted way to monitor end airway pressure, it is unclear if HHFNC is inadvertently generating high pressures and causing Journal of Perinatology (2007) 27, 85–91. doi:10.1038/sj.jp.7211647 unrecognized lung damage, particularly in the smaller preterminfants.
Keywords: humidified high flow nasal cannula; nasal continuouspositive airway pressure; bronchopulmonary dysplasia In this report we describe our recent experience with HHFNC following its widespread acceptance into practice in two largereferral medical centers, to include frequency of use and efficacyand safety of HHFNC compared to previous outcomes with non-invasive respiratory support consisting of NCPAP.
Respiratory failure in the neonatal period remains a difficultchallenge and is associated with high morbidity, mortality and Correspondence: Dr M Shoemaker, Division of Neonatology, Wilford Hall USAF Medical Center, 2200, Bergquest Dr, Lackland AFB, TX 78236, USA.
We performed a retrospective database review of infants admitted to E-mail: michael.shoemaker@lackland.af.milReceived 6 June 2006; revised 18 October 2006; accepted 10 November 2006 two regional referral medical centers (Wilford Hall USAF Medical Center, Lackland AFB, TX, USA and Christus Santa Rosa Children’s Hospital, San Antonio, TX, USA) during two study periods. The study periods were defined as August 2003 through June 2004 (Era 1) and August 2004 through June 2005 (Era 2). These timeperiods were based on the identification of June 2004 as a time point in both centers when the use of HHFNC became readilyavailable as an alternative therapy to NCPAP for neonates withrespiratory distress. The study was divided into two parts: (1) The first part describes the change in frequency of usage of HHFNC and NCPAP and compares selected outcomes of premature infants born during the two eras; and (2) the second part analyzes the outcomes of a defined cohort of infants less than 30 weeks gestational age(GA) who received either NCPAP or HHFNC as an early mode of respiratory support during the two eras (Figure 1).
Infants were included in the retrospective cohort (Part 2) if they were less than 30 weeks GA and inborn or transferred to one of thestudy centers within the first 24 h of life. To meet study criteria, infants had to be placed on one of the respiratory support modalitiesof interest (NCPAP or HHFNC) either initially following admission, Part 2 : Retrospective Cohort (Tables 4-7) as an escalation of support from oxyhood or low flow nasal cannula(p2 l/min (LPM), or immediately following extubation from Figure 1 Design of the two parts of the study.
mechanical ventilation within 96 h of birth. Nasal CPAP supportwas provided by either the Arabella (Hamilton Medical, Inc., than 0.05 were considered statistically significant. Logistic regression Bonaduz, Switzerland), InfantStar (Infrasonics, San Diego, CA, was used to analyze potential confounding variables. This study was USA), or Infant Flow Driver (Viasys Healthcare Inc., Warwick, approved by the Institutional Review Boards of both institutions.
England) utilizing pressures ranging from 3 to 8 cm H2O. Astraditionally the use of standard flow nasal cannulas in most neonatal units is limited to flow rates of 2 LPM or below, we defined high flow for rates greater than 2 LPM (range, 2.5 to 8 LPM).
Table 1 shows the frequency of usage of HHFNC or NCPAP for all Standard flows were delivered using the Vapotherm Neonate Nasal infants admitted during the two study eras. Percentages shown are Cannula (1.5 mm internal diameter), and HHFNC was generated the percent of infants who received either NCPAP or HHFNC during using the Vapotherm 2000i (Vapotherm Inc., Stevensville, MD, any point during their neonatal intensive care unit (NICU) stay.
USA). Both NCPAP pressure and HHFNC flow rates were adjusted as During Era 1, about 55% of premature infants less than 30 weeks needed by clinicians based on clinical exam, chest radiograph GA received NCPAP at some time during their NICU stay. This inflation and oxygen saturation levels. The fraction of inspired number dropped to only 12% in Era 2 whereas the use of HHFNC oxygen (FiO2) was adjusted according to center protocol targeting was nearly universal (95%). Similar trends were seen in those oxygen saturations between 85 and 92% Exclusion criteria included infants greater than 30 weeks GA, with NCPAP usage decreasing significant congenital heart disease, chromosomal abnormalities, from 12 to 2%, and HHFNC usage increasing from 11 to 57% genetic syndromes or other major congenital malformations.
during Era 1 versus Era 2, respectively.
Outcomes measures of interest included death, days on For those infants born at less than 30 weeks, regardless of mechanical ventilation, need for reintubation, air leak, infection, respiratory support received, there were no differences noted in BPD (defined as an ongoing requirement for supplemental oxygen mean gestation, birth weight, BPD and/or death between Era 1 and at 36 weeks corrected gestational age), necrotizing enterocolitis 2 (Table 2). Additionally, the numbers of infants admitted by week (NEC, either documented pneumatosis intestinalis or requiring of gestation from less than 24 to 30 weeks were similar between the surgical intervention), patent ductus arteriosus (PDA, either two eras (Table 2). There was a significantly higher number of receiving indomethacin or surgical ligation), severe ventilator-days per patient, however, observed in Era 1 as compared intraventricular hemorrhage (IVH, Papile’s grade 3 or 4), retinopathy of prematurity (ROP), days to full feeds (120 ml/kg/day) and hospital LOS.
Statistical analyses included Student’s t-test for continuous data We further analyzed selected outcomes for those babies placed on and Fisher’s exact test and w2 test for categorical data. P-values less either NCPAP or HHFNC as an early mode of respiratory support High flow nasal cannula versus NCPAPMT Shoemaker et al within the first 96 h of life. Thirty-six of 97 (37%) infants in Era 1 and the mean number of ventilator-days before receiving either and 65 of 103 (63%) infants in Era 2 born at less than 30 weeks NCPAP or HHFNC were similar (1.5 days; Table 4). There were no EGA received either early NCPAP or HHFNC and were included in significant differences in major clinical outcomes including death, the retrospective cohort (Figure 1). Of note, only two of 21 (10%) BPD, ventilator-days, NEC, PDA, severe IVH, LOS, ROP or time to infants born between 24 and 25 6/7 weeks gestation in Era 1 full feeds (Table 5). Although each group had the same percentage received early NCPAP compared to 11 of 25 (44%) managed with of infants with positive blood cultures, there were more cases of early HHFNC in Era 2. Reasons that infants did not receive the Gram-negative blood cultures documented in the HHFNC group respiratory mode of interest vary and are shown in Table 3. Over (not statistically significant). The Gram-negative organisms half of the infants excluded in each era resulted from either isolated in Era 1 were Escherichia coli (1) and Klebsiella prolonged intubation (>96 h) or death before extubation.
pneumonia (1) and in Era 2, K. pneumonia (2), E. coli (2), Characteristics of each group were similar as shown in Table 4.
E. cloaclae (3), Pseudomonas aeruginosa(1) and Ralstonia The majority of infants in both eras were initially supported by pickettii (1) No deaths occurred in either group of infants with mechanical ventilation for 1 to 2 days (61 and 66% respectively), documented Gram-negative bacteremia. Although there were nodifferences in outcomes between the two groups, more infants inthe NCPAP group were either intubated after initially receiving Table 1 Increase in usage of HHFNC after its introduction NCPAP or reintubated from NCPAP after an extubation attemptcompared to infants initially managed with or extubated to HHFNC Because there was a discrepancy in the two groups in patient numbers for the smallest babies (<26 weeks) included in the study (2 versus 11), additional analyses were performed comparinginfants 26 to 29 6/7 weeks EGA (Table 6). Although there was a trend toward less ventilator-days per patient (2.9 versus 4.5, P ¼ 0.25) and less BPD (20 versus 31%, P ¼ 0.23) in the HHFNC compared to the NCPAP group, these findings were not statisticallydifferent. However, there was a significantly lower intubation/ reintubation rate in Era 2 versus Era 1 (6 versus 35%, respectively, P<0.001), similar to that found for the less than 30 week gestation cohort as a whole. Outcomes are available for the groups stratified Abbreviations: GA, gestational age; HHFNC, humidified high flow nasal cannula; NCPAP, into 24 to 25 6/7, 26 to 27 6/7 and 28 to 29 6/7 weeks GA as well nasal continuous positive airway pressure.
As more infants in the HHFNC group had antenatal steroids Table 2 Characteristics and outcomes of infants <30 weeks GA before administered as well as delivered via cesarean section, we Table 3 Reasons for exclusion of <30 weeks GA infants from analysis Abbreviations: BPD, bronchopulmonary dysplasia; GA, gestational age; HHFNC, Abbreviations: GA, gestational age; HHFNC, humidified high flow nasal cannula; NCPAP, humidified high flow nasal cannula; s.d., standard deviation.
nasal continuous positive airway pressure.
Table 4 Characteristics of infants included in the analysis Table 5 Outcomes of infants in the analysis Total ventilator-days per patient (mean±s.d) Ventilator-days post NCPAP/HHFNCa (mean±s.d) Abbreviations: BPD, bronchopulmonary dysplasia; GA, gestational age; HHFNC, humidified high flow nasal cannula; IVH, intraventricular hemorrhage; NCPAP, nasal continuous positive airway pressure; NEC, necrotizing enterocolitis; PDA, patent ductus arteriosus; ROP, retinopathy of prematurity.
a Number of ventilator-days per patient after receiving the early respiratory mode of interest (NCPAP in Era 1 or HHFNC in Era 2).
bOne infant died in Era 1 and two were transferred before 36 weeks corrected GA while still receiving supplemental oxygen. In Era 2, 2 infants died and two were transferred.
cNumber of infants intubated after a trial on the respiratory mode of interest or reintubated after being extubated to the respiratory mode of interest.
dOnly ROP documented within the hospitalization. Information was not available for allinfants.
eDefined as stage 3 or greater in any zone or stage 2 in zone 1.
multivariate logistic regression was performed controlling for GA <26 weeks, male sex, outborn status, cesarean section and no antenatal steroid administration, there were more (re)intubations in the NCPAP compared to the HHFNC group (O.R. 10.7, 95% CI2.6 to 44, P ¼ 0.02). No statistical difference was seen for BPD (OR Abbreviations: HHFNC, humidified high flow nasal cannula; NCPAP, nasal continuous 2.43, 95% CI 0.81 to 7.2, P ¼ 0.23) or gram negative bacteremia positive airway pressure.
aInitial support when placed on the respiratory mode of interest (NCPAP or HHFNC).
(OR 0.52, 95% CI 0.09 to 2.79, P ¼ 0.5) between the two groups.
bNumber of ventilator-days per patient before receiving the early respiratory mode ofinterest (NCPAP in Era 1 or HHFNC in Era 2).
performed logistic regression analyses, which did not show an The use of HHFNC has increased in many NICUs over the past association between antenatal steroids or mode of delivery and BPD several years. Potential reasons for this increase include its ease of or (re)intubation. Only low birth weight, low GA and male sex were use and perceived improved tolerance with minimal nasal trauma positive predictors of BPD (Table 7), consistent with earlier compared to NCPAP. Clinical outcomes associated with the use of observations.4 Furthermore, infants who received antenatal steroids HHFNC are anecdotally perceived by some neonatologists to be at had a higher BPD rate compared to those who did not (32 versus least similar to those of NCPAP usage. Although HHFNC has been 19%, P ¼ 0.8) despite similar (re)intubation rates (19%). When widely accepted clinically, there is scant data regarding its efficacy High flow nasal cannula versus NCPAPMT Shoemaker et al Table 6 Outcomes of infants 26–29 6/7 weeks GA included in the 0.05 0.73 0.54–1.00 0.0001 0.53 0.37–0.74 Ventilator-days post NCPAP/HHFNCa (mean±s.d) Abbreviations: BPD, bronchopulmonary dysplasia; BW, birth weight; CI, confidence interval; GA, gestational age; NCPAP, nasal continuous positive airway pressure; OR, oddsratio.
aBirth weight in increments of 100 g; GA in increments of 1 week.
bOR+95% CI of infants in Era 1 compared to those in Era 2 for selected outcomes correcting for GA<26 weeks, male sex, outborn status, no antenatal steroids and cesarean Abbreviations: BPD, bronchopulmonary dysplasia; EGA, estimated gestational age; GA,gestational age; HHFNC, humidified high flow nasal cannula; IVH, intraventricularhemorrhage; NCPAP, nasal continuous positive airway pressure; NEC, necrotizingenterocolitis; PDA, patent ductus arteriosus.
mode during the study period. Specifically, there were no obvious aNumber of ventilator-days per patient after receiving the early respiratory mode of changes identified in ventilator management, intubation or interest (NCPAP in Era 1 or HHFNC in Era 2).
b extubation criteria, or antenatal steroid usage. Patient Number of infants intubated after a trial on the respiratory mode of interest or reintubated after being extubated to the respiratory mode of interest.
demographics and GA distribution were very similar between thetwo eras. In general, the majority of extremely low birth weight(ELBW) infants (<28 weeks GA) in our practice are and safety. One published trial exists comparing standard prophylactically intubated and administered exogenous surfactant high-flow nasal cannula (SHFNC) with NCPAP in which SHFNC with the goal of either immediate extubation to NCPAP or HHFNC, was shown to be as efficacious as NCPAP in preventing apnea of or early extubation following limited mechanical ventilation.
prematurity.14 Otherwise there have been only a few abstracts9–13 Although on a respiratory support modality, permissive hypercapnia with small patient numbers reported to date describing the safety of is usually tolerated with the goal of maintaining arterial pH>7.25 this modality compared to NCPAP in premature infants, and no coupled with a concerted effort to limit excessive oxygen exposure.
controlled clinical trials evaluating its utility in this population.
The above practice guidelines have been in place at both of our In this report we describe our experience in two large regional study centers for a number of years preceding the defined study medical centers where HHFNC has largely replaced NCPAP as the period; this is consistent with the approach adopted by many preferred mode of noninvasive respiratory support, particularly in institutions over the past decade in an effort to avoid or limit the those infants born at less than 30 weeks GA (Part 1). By comparison duration of mechanical ventilation in preterm infants with to historical data before its widespread introduction, we have shown that HHFNC appears to be well tolerated and to provide similar Historical problems associated with using an early NCPAP outcomes when compared to NCPAP. Death and BPD rates were strategy in ELBW infants include difficulty with comfortably similar for premature infants before and following the introduction maintaining a functional patient–device interface and the of HHFNC, and there was a decrease in ventilator-days.
associated nasal trauma that can occur with using this modality.
We are not aware of any significant shift in clinical practice The application of HHFNC in these small babies, however, is much other than the introduction of HHFNC as a respiratory support simpler, which is likely one of the main reasons for its widespread acceptance into a number of neonatal units. Additionally, the ease flow.21 In our study, no infant supported with HHFNC during their of use of HHFNC as compared to NCPAP (at least as perceived by hospitalization in either Era 1 or Era 2 (whether they were medical, nursing and respiratory care providers) may have included in the analysis or not) had a pneumothorax while on this contributed to a greater willingness to use HHFNC and bias toward respiratory mode. Further studies measuring airway pressure its more successful utility as a mode of respiratory support.
generated with HHFNC devices in premature neonates, to include When infants less than 30 weeks GA were compared by early how pressures vary with weight and at different flow rates, need to mode of respiratory support (NCPAP or HHFNC), outcomes were similar (Part 2). There were no statistically significant differences Recently the device we used to deliver high flow, the Vapotherm in ventilator-days, deaths, infections, IVH or LOS. There is debate 2000i, was recalled owing to concerns of increased Gram-negative whether early NCPAP decreases the risk of BPD in infants <30 bacteremia, specifically R. pickettii.22 In our study, although we weeks GA.4–8 Our data did not show a difference in BPD rates found the overall rate of bacteremia to be similar between the two between the NCPAP and HHFNC groups. Although there was a Eras, we did find a higher incidence of Gram-negative bacteremia lower rate of BPD in infants who received HHFNC as first intention in the infants who received HHFNC as an early mode of respiratory versus NCPAP, especially in the 26 to 29 6/7 week GA group, the support versus NCPAP. Only one infant grew R. pickettii. Although numbers were small and the power inadequate. The majority of we cannot directly attribute the increased Gram-negative infants in both eras included in the analysis were intubated for 1 to bacteremia to use of HHFNC, the relationship warrants further 2 days before receiving either NCPAP or HHFNC (Table 4) and investigation. Based on our data, approximately 250 infants would outcomes in these infants were not different. Six of the nine infants need to be enrolled in a prospective study to detect a significant who received NCPAP first (GA 28.8±1.1 weeks, birth weight (BW) difference in Gram-negative bacteremia (80% power). A recent 1205±197 g) versus only three of the 17 infants who received report described a positive association between nasal cannula HHFNC first (GA 28.8±0.7 weeks, BW 1094±202 g) were later continuous positive airway pressure (but not mechanical intubated for respiratory reasons, spending an average of 5.1±5.7 ventilation) and late onset Gram-negative blood infections in low and 1.6±2.9 days, respectively, mechanically ventilated. Although birth weight infants, which the authors attributed to increasing this finding is of interest, it is possible that some babies may have nasal mucosa damage from the cannulas.23 As HHFNC maintains a been placed on HHFNC in Era 2 who otherwise would have done normal mucosa better than standard high flow nasal cannula,24 just as well on less respiratory support (i.e., <2 LPM nasal cannula it remains to be seen if infection rates will be altered.
or room air), which would have contributed to relatively more Although this study is limited by its relatively small size and infants being intubated in the NCPAP cohort as well as to a inclusion of only those infants less than 30 weeks GA, the data reduction in the BPD rates reported for HHFNC babies in Era 2.
presented here indicate that HHFNC may represent a well-tolerated There are existing concerns among neonatologists regarding the and effective alternative respiratory support mode to NCPAP in the widespread application and usage of HHFNC in premature infants preterm infant population. Its potential advantages include its in the absence of sufficient published literature supporting its simplicity, improved tolerability with less injury to the nasal utility and safety. Particular concern has focused on the imprecise architecture and mucosa, and perhaps greater clinical utility in regulation and generation of pressure that may occur at higher managing respiratory distress in premature infants. However, owing flows, especially in the smallest of infants, as well as the potential to unresolved infection concerns and the paucity of published for a significant increased work of breathing with HHFNC devices outcomes to date, the safety and utility of HHFNC as compared to as compared to NCPAP.15–18 Of interest, a recently published small more traditional respiratory support modes remains unproven and randomized trial did not show increased work of breathing or needs to be further investigated. We believe our experience warrants respiratory rates of preterm infants <2 kg on HHFNC (3 to 5 LPM) a large, randomized controlled trial comparing the efficacy, safety compared to preterm infants receiving NCPAP set at 6 cm H2O.19 Additionally, recorded esophageal pressures were consistently <4 cm H2O at flow rates of 3 to 5 LPM, similar to delivered NCPAP pressures. Previously, Locke et al.20 reported that as much as 9 cm H2O pressure measured by esophageal balloon manometry can be The opinions expressed are of the authors only and do not necessarily represent generated with as little as 2 LPM SHFNC in 3 mm cannulas but not those of the Department of the Air Force or the Department of Defense.
in 2 mm cannulas. The cannulas used in our infants were 1.5 mm,and although we did not quantify pressure, pressures generated didnot appear to be excessive based on clinical evaluations including serial chest radiographs. One recent report looking at airway Angus DC, Linde-Zwirble WT, Clermont G, Griffen MF, Clark RH.
pressure generated in preterm infants with the Vapotherm device Epidemiology of neonatal respiratory failure in the United States. Am J Resp found a high pressure of only 4.5 cm H2O with up to 8 LPM of Crit Care Med 2001; 164: 1154–1161.
High flow nasal cannula versus NCPAPMT Shoemaker et al Hudson LD. Progress in understanding ventilator-induced lung injury. JAMA conventional nasal continuous positive airway pressure. Pediatrics 2001; Jobe AH, Ikegami M. Mechanisms initiating lung injury in the preterm Chang GC, Cox CC, Shaffer TH. Nasal cannula, CPAP, and Vapotherm: effect infant. Early Hum Dev 1998; 53: 81–94.
of flow on temperature, humidity, pressure and resistance. Pediatr Acad Soc Rich W, Finer NN, Vaucher YE. Ten-year trends in neonatal assisted ventilation of very low- birthweight infants. J Perinatol 2003; 28: 660–663.
Courtney SE, Pyon KH, Saslow JG, Arnold GK, Pandit PB, Habib RH. Lung Jegatheesan P, Keller RL, Hawgood S. Early variable-flow nasal continuous recruitment and breathing pattern during variable versus continuous flow positive pressure in infants < or ¼ 1000 grams at birth. J Perinatol 2006; nasal continuous positive airway pressure in premature infants: an evaluation of three devices. Pediatrics 2001; 107: 304–308.
Aly H, Milner JD, Patel K, El-Mohandes AA. Does the experience with the use Finer NN. Nasal cannula use in the preterm infant: oxygen or pressure? of nasal continuous positive airway pressure improve over time in extremely low birth weight infants? Pediatrics 2004; 114: 697–702.
Walsh M, Engle W, Laptook A, Kazzi SNJ, Buchter S, Rasmussen M et al.
Narendran V, Donovan EF, Hoath SB, Akinbi HT, Steichen JJ, Jobe AH. Early Oxygen delivery through nasal cannulae to preterm infants: can practice be bubble CPAP and outcomes in ELBW preterm infants. J Perinatol 2003; 23: improved? Pediatrics 2005; 116: 857–861.
Saslow JG, Aghai ZH, Nakhla TA, Hart JJ, Lawrysh R, Stahl GE et al. Work of Davis PG, Henderson-Smart DJ. Post-extubation prophylactic nasal breathing using high-flow nasal cannula in preterm infants. J Perinatol continuous positive airway pressure in preterm infants: systematic review and meta-analysis. J Paediatr Child Health 1999; 35: 367–371.
Locke RG, Wolfson MR, Shaffer TH, Rubenstein SD, Greenspan JS.
Nair G, Karna P. Comparison of the effects of Vapotherm and nasal CPAP in Inadvertent administration of positive end-distending pressure during nasal respiratory distress in preterm infants. Pediatr Acad Soc 2005; 57: 2054.
cannula flow. Pediatrics 1993; 91: 135–138.
Ovalle OO, Gomez T, Troncoso G, Palacios J, Ortiz E. High flow nasal Kubicka Z, Limauro J, Darnall R. High flow nasal cannula therapy with cannula after surfactant treatment for infant respiratory distress syndrome in Vapothem: yet another way to deliver CPAP? Pediatr Acad Soc 2006; 59: preterm infants <30 weeks. Pediatr Acad Soc 2005; 57: 3417.
Ramanathan A, Cayabyab R, Sardesai S, Siassi B, Seri I, Ramanathan R.
Center for Disease Control Convention. Ralstonia associated with Vapotherm High flow nasal cannula use in preterm and term newborns admitted to oxygen delivery device – United States, 2005. Morb Mortal Wkly Rep 2005; neonatal intensive care unit: A prospective, observational study. Pediatr Acad Graham III PL, Begg MD, Larson E, Della-Latta P, Allen A, Saiman L. Risk Sanchez F, Sabato K. Very high nasal cannula-alternative to NCPAP in select Factors for late onset Gram-negative sepsis in low birth weight infants ICN patients? Respir Care 2004; 49: 1373.
hospitalized in the neonatal intensive care unit. Pediatr Infect Dis J 2006; Holleman-Duray DL, Kaupie DL, Weiss MG. Safety and efficacy of the Vapotherms high flow humidification system and an early extubation Woodhead DD, Lambert DK, Clark JM, Christensen RD. Comparing two protocol. Pediatr Acad Soc 2006; 59: 517.
methods of delivering high-flow gas therapy by nasal cannula following Sreenan C, Lemke RP, Hudson-Mason A, Osiovich H. High-flow nasal endotracheal extubation: a prospective, randomized, masked, crossover trial.
cannulae in management of apnea of prematurity: a comparison with Supplementary Information accompanies the paper on the Journal of Perinatology website (http://www.nature.com/jp)

Source: http://gdmedical.ch/media/DIR_234330/2cfb7ee2a75cff6bffff9b6dffffffe7.pdf