Comparison of nasal intermittent positive pressure ventilation and bubble CPAP with an in-line high-frequency interrupter in a premature infant lung model.

INTRODUCTION: Noninvasive ventilation has become a staple in the care of premature infants. However, failure rates continue to be high in this population. Modifications to noninvasive support, such as nasal intermittent positive pressure ventilation (NIPPV), are used clinically to reduce such failure. Previous in vitro studies have shown improved CO2 clearance when superimposing high-frequency oscillations onto bubble continuous positive airway pressure (BCPAP). OBJECTIVE: To compare the CO2 clearance of NIPPV to BCPAP with an in-line high-frequency interrupter (HFI) in a premature infant lung model. METHODS: A premature infant lung model was connected to either a Dräger VN500 for delivery of NIPPV or a BCPAP device with superimposed high-frequency oscillations generated by an in-line HFI. Change in end-tidal CO2  (ETCO2 ) and mean airway pressure at the simulated trachea were measured and compared for both noninvasive modalities. RESULTS: Superimposing HF oscillations onto BCPAP with an in-line HFI resulted in improved CO2 clearance relative to BCPAP alone for all tested oscillation frequencies at all CPAP levels (p < 0.001). NIPPV also resulted in improved CO2  clearance relative to nasal CPAP (NCPAP) alone (p < 0.001). Among the tested settings, BCPAP with an in-line HFI resulted in decreased ETCO2 relative to BCPAP ranging from -14% to -36%, while NIPPV resulted in decreased ETCO2  relative to NCPAP ranging from -2% to -12%. CONCLUSION: Superimposing high-frequency oscillations onto BCPAP using a novel in-line HFI was found to be more effective at clearing CO2 than NIPPV in a premature infant lung model.

Preserved pressure delivery during high-frequency oscillation of bubble CPAP in a premature infant lung model with both normal and abnormal lung mechanics.

BACKGROUND: Bubble continuous positive airway pressure (BCPAP) generates pressure oscillations which are suggested to improve gas exchange through mechanisms similar to high frequency (HF) ventilation. In a previous in-vitro lung model with normal lung mechanics, significantly improved CO2 washout was demonstrated using an HF interrupter in the supply flow of a BCPAP system. The effect of HF with BCPAP on delivered airway pressure (Paw) has not been fully investigated in a lung model having abnormal pulmonary mechanics. OBJECTIVE: To measure Paw in an infant lung model simulating normal and abnormal pulmonary compliance and resistance while connected to a BCPAP system with superimposed HF oscillations created using an in-line flow interrupter. DESIGN/METHODS: A premature infant lung model with either: normal lung mechanics, compliance 1.0 ml/cm H2 O, airway resistance 56 cm H2 O/(L/s); or abnormal mechanics, compliance 0.5 ml/cm H2 O, airway resistance 136 cm H2 O/(L/s), was connected to BCPAP with HF at either 4, 6, 8, 10, or 12 Hz. Paw was measured at BCPAPs of 4, 6, and 8 cm H2 O and respiratory rates (RR) of 40, 60, and 80 breaths/min and 6.0 ml tidal volume. RESULTS: Mean Paw averaged over all five frequencies showed no significant change from non-oscillated levels at all BCPAPs and RRs for both lung models. Paw amplitudes (peak-to-trough) during oscillation were significantly greater than the non-oscillated levels by an average of 1.7 ± 0.5 SD and 2.6 ± 0.5 SD cm H2 O (p < .001) for the normal and abnormal models, respectively. CONCLUSIONS: HF oscillation of BCPAP using a flow interrupter did not alter mean delivered Paw compared to non-oscillated BCPAP for both normal and abnormal lung mechanics models. This simple modification to BCPAP may be a useful enhancement to this mode of non-invasive respiratory support.

Carbon dioxide clearance using bubble CPAP with superimposed high-frequency oscillations in a premature infant lung model with abnormal lung mechanics.

BACKGROUND: High-frequency (HF) oscillatory ventilation has been shown to improve carbon dioxide (CO2 ) clearance in premature infants. In a previous in vitro lung model with normal lung mechanics we demonstrated significantly improved CO2 washout by HF oscillation of bubble continuous positive airway pressure (BCPAP). OBJECTIVE: To examine CO2 clearance in a premature infant lung model with abnormal lung mechanics via measurement of end-tidal CO2 levels (EtCO2 ) while connected to HF oscillated BCPAP. DESIGN AND METHODS: A 40 mL premature infant lung model with either: normal lung mechanics (NLM): compliance 1.0 mL/cm H2 O, airway resistance 56 cm H2 O/(L/s); or abnormal lung mechanics (ALM): compliance 0.5 mL/cm H2 O, airway resistance 136 cm H2 O/(L/s), was connected to BCPAP with HF oscillation at either 4, 6, 8, 10, or 12 Hz. EtCO2 was measured at BCPAPs of 4, 6, and 8 cm H2 O and respiratory rates (RR) of 40, 60, and 80 breaths/min and 6 mL tidal volume. RESULTS: HF oscillation decreased EtCO2 levels at all BCPAPs, RRs, and oscillation frequencies for both lung models. Overall mean ± SD EtCO2 levels decreased (P < .001) from nonoscillated baseline by 19.3 ± 10.2% for NLM vs 14.1 ± 8.8% for ALM. CO2 clearance improved for both lung models (P < .001) as a function of oscillation frequency and RR with greatest effectiveness at 40 to 60 breaths/min and HF at 8 to 12 Hz. CONCLUSIONS: In this in vitro premature infant lung model, HF oscillation of BCPAP was associated with improved CO2 clearance as compared with nonoscillated BCPAP for both NLM and ALM. The significant improvement in CO2 clearance in an abnormal lung environment is an important step towards clinical testing of this novel respiratory support modality.

An in-line high frequency flow interrupter applied to nasal CPAP: Improved carbon dioxide clearance in a premature infant lung model.

BACKGROUND: Noninvasive respiratory support continues to have high failure rates in small preterm infants. We previously demonstrated significantly improved in vitro CO2 washout by applying oscillations to a high flow nasal cannula system. OBJECTIVE: To develop a high frequency flow interrupter that could be applied to commonly used nasal continuous positive airway pressure (NCPAP) devices and to determine the effect of oscillations on end-tidal carbon dioxide (EtCO2 ) levels in an infant lung model. DESIGN/METHODS: NCPAP was applied to a premature infant lung simulator using either bubble (BCPAP) or variable-flow (VCPAP) CPAP. Supply gas was interrupted with a solenoid pinch valve. EtCO2 was measured before and during oscillation and repeated at 4, 6, 8, 10, and 12 Hz oscillation and CPAP pressures of 4, 6, and 8 cm H 2 O. RESULTS: BCPAP and VCPAP EtCO2 levels decreased with oscillation (P < .001). BCPAP EtCO2 was significantly dependent on oscillation frequency (P < .001) with decreases of 18% to 47% and maximum effect at 10 Hz. Optimum VCPAP CO2 clearance occurred at 6 Hz with reductions of 30% and 39% at 6 and 8 cm H2 O CPAP respectively. BCPAP and VCPAP mean airway pressures remained unchanged transitioning from nonoscillation to oscillation. Oscillated BCPAP and VCPAP average amplitudes were 8.3 ± 0.5 and 8.4 ± 2.3 SD cm H2 O, respectively. Power spectrum analysis of non-oscillated BCPAP showed bubbling-only dominant peaks at 10 to 12 Hz corresponding with the maximum BCPAP EtCO2 reductions. CONCLUSION: Application of high frequency oscillation to NCPAP improves CO2 clearance in a premature infant lung model. This simple modification to NCPAP delivery devices may prove to be an effective enhancement of this mode of noninvasive respiratory support.

Effect of high-frequency oscillation on pressure delivered by high flow nasal cannula in a premature infant lung model.

OBJECTIVE: This study describes the effect of high-frequency oscillation on airway pressure generated by high flow nasal cannula (HFNC) in a premature infant lung model. DESIGN/METHODS: A premature in 0.5 or 1.0 mL/cmH 2 O, respiratory rate (RR) of 40 or 60 breaths per min, and tidal volume of 6 mL. Oscillation was achieved by passing the HFNC supply flow through a 3-way solenoid valve operating at 4, 6, 8, or 10 Hz. Airway pressure at the simulated trachea was recorded following equilibration of end-tidal CO 2 both with and without oscillation. RESULTS: Superimposing high-frequency oscillations onto HFNC resulted in an average decrease in mean airway pressure of 17.9% (P = .011). The difference between the maximum and minimum airway pressures, ∆ P min-max, significantly increased as oscillation frequency decreased ( P < .001). Airway pressure during oscillation was 12.8% greater with the 1.0 vs the 0.5 mL/cmH 2 O compliance at flows > 4 L/min ( P = .031). CO 2 clearance was 13.1% greater with the 1.0 vs 0.5 mL/cmH 2 O compliance at oscillation frequencies less than 8 Hz ( P = .015). CONCLUSION: In this in-vitro study we demonstrate that delivered mean airway pressure decreases when applying high-frequency oscillation to HFNC, while still improving CO2 clearance. The combination of improved CO 2 clearance and reduced pressure delivery of this novel noninvasive modality may prove to be a useful improvement in the respiratory care of infants in respiratory distress.

Effect of high frequency oscillatory high flow nasal cannula on carbon dioxide clearance in a premature infant lung model: A bench study.

OBJECTIVE: This study compared CO2 clearance in a premature infant lung model connected to a high flow nasal cannula (HFNC) system supplied with oscillatory versus non-oscillatory flow. DESIGN/METHODS: The lung model was set to compliance 1.0 mL/cmH2 O, RR 60 breaths/min, and 6 mL tidal volume. A 100% CO2 was injected at a constant 15 mL/min. To create oscillation, HFNC flow was interrupted at rates of 4-6-8 and 10 Hz. equilibrated end-tidal CO2 (ETCO2 ) was recorded with and without oscillation at set flows of 2-8 L/min and repeated for each oscillation frequency. RESULTS: Overall ETCO2 decreased significantly (P < 0.001) during both non-oscillatory and oscillatory HFNC as set flow increased from 2 to 8 L/min by 26.3% and 60.8%, respectively. Oscillatory ETCO2 levels decreased linearly compared to non-oscillatory HFNC with negligible difference at 2 L/min and a 48.4% difference at 8 L/min (P < 0.001). There were no differences in ETCO2 levels between oscillation frequencies at any flow except at 6 Hz for which ETCO2 was significantly lower (P < 0.01) than at 4, 8, and 10 Hz for 5-8 L/min HFNC flows. Amplitude of volume oscillations increased with increasing flow from 0.5 mL at 2 L/min to 4.0 mL at 8 L/min (P < 0.001), and decreased with increasing oscillation frequency. CONCLUSION: Oscillatory HFNC as compared to non-oscillatory was associated with significantly improved CO2 clearance in this premature infant lung model. This simple modification of the HFNC system may prove to be a useful enhancement to this mode of non-invasive respiratory support for preterm infants at high risk for respiratory failure.

Carbon dioxide washout during high flow nasal cannula versus nasal CPAP support: An in vitro study.

OBJECTIVE: To compare CO2 washout time at different levels of HFNC versus NCPAP in a premature infant lung model with simulated mouth-closed and mouth-open conditions using two sizes of nasal cannula and full- and half-prong HFNC insertion depths. DESIGN/METHODS: A piston-cylinder lung simulator, having a fixed volume of 30 ml and a 4.8 ml dead space, simulated spontaneous breathing (6.5 ml tidal volume, 50 br/min, Ti = 0.5 sec). Two Fisher & Paykel™ cannulas (Fisher & Paykel Healthcare Ltd., Auckland, New Zealand) (2.8 and 3.2 mm O.D.) and two Infant-Flow™ (CareFusion, Yorba Linda, CA) NCPAP cannulas (3.4 and 4.1 mm O.D.) were applied to simulated airways having either 3.5 or 4.5 mm I.D. nares. Simulated mouth opening was a 5 mm I.D. side tap below the nasal interface. The lung was primed with 5% CO2 . Washout times were determined at HFNC settings of 3, 4, 5, 6, and 8 L/min and NCPAP at 3, 4, 5, 6, and 8 cm H2 O with simulated open and closed-mouth conditions and full- and half-inserted HFNC prongs. RESULTS: Overall combined mean washout times for NCPAP with mouth-closed were significantly longer than HFNC over all five pressure and flow device settings by 16.2% (P < 0.001). CO2 washout times decreased as flow or pressure device settings were increased. There were negligible differences in washout times between NCPAP and HFNC with mouth-open. Mouth-open washout times were significantly less than mouth-closed for all conditions. Overall closed-mouth washout times for HFNC half-prong insertion were longer than for full-prong insertion by 5.3% (P < 0.022). CONCLUSIONS: Significantly improved CO2 elimination using HFNC versus NCPAP should be a particularly important consideration in premature infants having very high dead space-to-tidal volume ratio compared to larger infants. Pediatr Pulmonol. 2017;52:792-798. © 2017 Wiley Periodicals, Inc.

Factors influencing delivered mean airway pressure during nasal CPAP with the RAM cannula.

OBJECTIVE: To measure mean airway pressure (MAP) delivered through the RAM Cannula® when used with a ventilator in CPAP mode as a function of percent nares occlusion in a simulated nasal interface/test lung model and to compare the results to MAPs using a nasal continuous positive airway pressure (NCPAP) interface with nares fully occluded. STUDY DESIGN: An artificial airway model was connected to a spontaneous breathing lung model in which MAP was measured at set NCPAP levels between 4 and 8 cmH2 O provided by a Dräger Evita XL® ventilator and delivered through three sizes of RAM cannulae. Measurements were performed with varying leakage at the nasal interface by decreasing occlusion from 100% to 29%, half-way prong insertion, and simulated mouth leakage. Comparison measurements were made using the Dräger BabyFlow® NCPAP interface with a full nasal seal. RESULTS: With simulated mouth closed, the Dräger interface delivered MAPs within 0.5 cmH2 O of set CPAP levels. For the RAM cannula, with 60-80% nares occlusion, overall delivered MAPs were 60 ± 17% less than set CPAP levels (P < 0.001). Further, MAP decreased progressively with decreasing percent nares occlusion. The simulated open mouth condition resulted in significantly lower MAPs to <1.7 cmH2 O. The one-half prong insertion depth condition, with closed mouth, yielded MAPs approximately 35 ± 9% less than full insertion pressures (P < 0.001). CONCLUSIONS: In our bench tests, the RAM interface connected to a ventilator in NCPAP mode failed to deliver set CPAP levels when applied using the manufacturer recommended 60-80% nares occlusion, even with closed mouth and full nasal prong insertion conditions.

Effect of HFNC flow rate, cannula size, and nares diameter on generated airway pressures: an in vitro study.

Increased use of non-invasive forms of respiratory support such as CPAP and HFNC in premature infants has generated a need for further investigation of the pulmonary effects of such therapies. In a series of in vitro tests, we measured delivered proximal airway pressures from a HFNC system while varying both the cannula flow and the ratio of nasal prong to simulated nares diameters. Neonatal and infant sized nasal prongs (3.0 and 3.7 mm O.D.) were inserted into seven sizes of simulated nares (range: 3-7 mm I.D. from anatomical measurements in 1-3 kg infants) for nasal prong-to-nares ratios ranging from 0.43 to 1.06. The nares were connected to an active test lung set at: TV 10 ml, 60 breaths/min, Ti 0.35 sec, compliance 1.6 ml/cm H2O and airway resistance 70 cm H2O/(L/sec), simulating a 1-3 kg infant with moderately affected lungs. A Fisher & Paykel Healthcare HFNC system with integrated pressure relief valve was set to flow rates of 1-6 L/min while cannula and airway pressures and cannula and mouth leak flows were measured during simulated mouth open, partially closed and fully closed conditions. Airway pressure progressively increased with both increasing HFNC flow rate and nasal prong-to-nares ratio. At 6 L/min HFNC flow with mouth open, airway pressures remained <1.7 cm H2O for all ratios; and <10 cm H2O with mouth closed for ratios <0.9. For ratios >0.9 and 50% mouth leak, airway pressures rapidly increased to 18 cm H2O at 2 L/min HFNC flow followed by a pressure relief valve limited increase to 24 cm H2O at 6 L/min. Safe and effective use of HFNC requires careful selection of an appropriate nasal prong-to-nares ratio even with an integrated pressure relief valve.

An open label comparison of calfactant and poractant alfa administration traits and impact on neonatal intensive care unit resources.

OBJECTIVE: To compare calfactant (CA) and poractant alfa (PA) administration traits, short-term clinical responses, and resource use in the neonatal respiratory distress syndrome (RDS) setting. METHODS: An open label series of 277 (213 PA and 64 CA) infants was evaluated for 445 administrations. Registered respiratory therapists collected patient, surfactant administration, and postadministration clinical data. Economic analysis involved labor costs of surfactant administration and usage, wastage, and product average wholesale price. Analysis utilized the Mann-Whitney rank sum test for differences in administration time and either the chi-square or Fisher's exact test for categorical variables. RESULTS: PA had a statistically lower bedside administration time than CA (3.8 minutes vs. 5.3 minutes; P = .006) and a higher percentage of doses administered in less than five minutes (58.9% vs. 4.3%; P < .001). Doses administered per patient were similar (1.67 vs. 1.72). PA and CA were similar in time to recovery (81.4% vs. 74.3%), percent desaturation (24.8% vs. 26.7%), and bradycardia (3.8% vs. 8.5%). Reflux was significantly higher (13.2% vs. 3.5%; P < .001) with CA. Economic analyses found total administration costs per dose were $2.21 for PA and $3.08 for CA. Mean wastage costs were $141.21 for PA and $337.34 for CA (P < .001). CONCLUSIONS: PA appeared to utilize fewer neonatal intensive care unit resources than CA due to reduced administration time and less wastage of drug product. Future studies should more closely evaluate time, resource, wastage, and post-administrative clinical effects to fully assess the impact of surfactant products in this setting.

Postnatal changes in pulmonary mechanics and energetics of infants with respiratory distress syndrome following surfactant treatment.

BACKGROUND: Postnatal alterations in pulmonary mechanics, energetics and functional residual capacity (FRC) describe the structural maturation of the preterm respiratory system. OBJECTIVE: To evaluate longitudinal changes in pulmonary function in infants with respiratory distress syndrome (RDS) treated with oxygen, positive pressure ventilation and synthetic surfactant (Exosurf). METHODS: Serial pulmonary function tests were performed in surfactant-treated infants [mean +/- SD birth weight (BW) = 1,112 +/- 276 g, gestational age (GA) = 29 +/- 3 weeks] at postnatal ages: <3 days, 1, 2, 3, 4 and 6-8 weeks until term postmenstrual age (PMA). Tidal volume, pulmonary compliance (C(L)), pulmonary resistance (R(T)) and flow-resistive work were analyzed following simultaneous measurements of airflow and transpulmonary pressure signals. Serial FRC measurements were made in a randomly selected group. RESULTS: Prior to 28 weeks' PMA, C(L) was unchanged irrespective of GA. At age 1 week the likelihood ratio (LR) for bronchopulmonary dysplasia (BPD) based on C(L), R(T) and GA was predicted to be >90% for those with BW <750 g (LR >100) as compared to <10% probability (LR = 0.3) for infants >1,500 g. Significant linear increase in C(L) to PMA was evident >28 weeks' PMA (r = 0.86, p < 0.01) at 0.17 ml/cm H2O/kg/week. By term PMA, mean C(L) was 2.60 +/- 0.07 ml/cm H2O. Improvements in FRC of preterm infants with RDS who recovered occur at a more rapid rate ( approximately 25 ml/kg) compared to those who developed BPD ( approximately 20 ml/kg). CONCLUSIONS: Slow but incremental postnatal pulmonary improvement, minimal <28 weeks' PMA, were comparable for all infants. Along with diminished FRC, these changes reflect persistent deleterious effects of positive pressure ventilation, alveolar hyperoxia and unrecognized pulmonary overdistension.

Innovative neonatal ventilation and meconium aspiration syndrome.

Respiratory failure remains a major cause of morbidity and mortality in the neonatal population. Infants with hypoxemic respiratory failure because of meconium aspiration syndrome (MAS), persistent pulmonary hypertension of the newborn (PPHN), and pneumonia/sepsis have a potential for increased survival with extracorporeal membrane oxygenation (ECMO). Other treatment options previously limited to inotropic support, conventional ventilatory management, respiratory alkalosis, paralysis and intravenousvasodilators have been replaced by high-frequency oscillatory ventilation (HFOV), surfactant, and inhaled nitric oxide (iNO). HFOV has been advocated for use to improve lung inflation while potentially decreasing lung injury through volutrauma. Other reports describe enhanced efficacy of HFOV when combined with iNO. Subsequent to studies reporting surfactant deficiency or inactivation may contribute to neonatal respiratory failure exogenous surfactant therapy has been implemented with apparent success. Recent studies have shown that iNO therapy in the neonate with hypoxemic respiratory failure can result in improved oxygenation and decreased need for ECMO. In this article, the authors place in context of a system-based strategy the prenatal, natal and postnatal management of babies delivered through meconium stained amniotic fluid (MSAF) so that adverse outcomes are minimized, and the least number of babies require innovative ventilatory support. At Pennsylvania Hospital, over a six-year period (1995 to 2000), 14.5% (3370/23,175 of live births babies were delivered with MSAF. These data show that 4.6% (155/3370) of babies with MSAF sustained MAS. Overall, 26% (40/155) of babies with MAS needed ventilatory support (or 0.17% of all live-births); of these only 20% (8/40 or 0.035% of live births) needed innovative ventilatory support. None died or needed ECMO. These data describe the impact of a system-based approach to prevent and manage adverse outcomes related to MSAF at regional Level III perinatal center.