OBJECTIVES. The goal were to characterize the flow dependence of bubble nasal continuous positive airway pressure delivery in a cohort of preterm infants and to compare the actual (delivered) intraprong continuous positive airway pressure with the intended (set) nasal continuous positive airway pressure for both ventilator-generated nasal continuous positive airway pressure and bubble nasal continuous positive airway pressure delivery. A range of set values and constant flow rates were studied in the same preterm infants.
METHODS. For 12 premature infants of <1500 g (birth weight: 1140 ± 267 g; gestational age: 28.5 ± 1.9 weeks; study age: 12.9 ± 8 days; all mean ± SD), intraprong pressures were measured at 3 increasing flow settings, repeated for set nasal continuous positive airway pressures (or desired immersion depths) of 4 and 6 cmH2O. Next, intraprong pressures were measured at bubble nasal continuous positive airway pressure expiratory tubing submersion depths and ventilator-generated nasal continuous positive airway pressure set expiratory pressures of 2, 3, 4, 5, and 7 cmH2O while the flow rate was held constant.
RESULTS. Actual (delivered) intraprong pressure during bubble nasal continuous positive airway pressure delivery was highly flow dependent and increased as the flow rate increased. During ventilator-generated nasal continuous positive airway pressure delivery, actual pressure at the nasal prongs closely approximated the pressure set at the ventilator. During bubble nasal continuous positive airway pressure delivery at constant flow rate, the average delivered prong pressure was 1.3 cmH2O (range: 0.5–2.2 cmH2O) higher than that set through submersion of the expiratory tubing, and the relative difference between the set and actual pressures increased at lesser immersion depths.
CONCLUSIONS. Prong pressure during bubble nasal continuous positive airway pressure delivery is highly variable and depends on the interaction of submersion depth and flow amplitudes.
- nasal continuous positive airway pressure
- bubble nasal continuous positive airway pressure
- ventilator-generated nasal continuous positive airway pressure
Nasal continuous positive airway pressure (NCPAP) therapy provides noninvasive respiratory support and is commonly used to maintain lung volume and to reduce apnea of prematurity. NCPAP therapy may reduce the need for intubation and possibly the incidence of chronic lung disease.1,2
NCPAP therapy is commonly provided by using an infant ventilator. During ventilator-generated NCPAP (V-NCPAP) delivery, the flow rate is constant and the continuous positive airway pressure (CPAP) level is maintained at the distal expiratory limb orifice, which adjusts to keep the delivered CPAP at the desired level. Another popular method of CPAP provision is bubble NCPAP (B-NCPAP) delivery. B-NCPAP delivery, which was used by Gregory et al3 in the initial report of CPAP application in infants, nearly disappeared from use as infant ventilators became available in the 1970s. During B-NCPAP delivery, the expiratory limb of the CPAP tubing is immersed in an underwater chamber to a depth (in centimeters) equal to the desired CPAP level (in centimeters of water). The flow through the system creates bubbling in the chamber (hence, the name). B-NCPAP delivery has regained popularity because it is favored by an institution reporting a low incidence of chronic lung disease.4 Benefits to gas exchange and lung recruitment during B-NCPAP delivery, because of the high-frequency oscillatory content of the bubbling, have been hypothesized.5–8
Most clinicians assume that the delivered mean intraprong pressure during B-NCPAP delivery is represented accurately by the submersion depth of the expiratory tubing. However, B-NCPAP delivery is characterized by substantial oscillations resulting from generation of the bubbles, and these oscillations may affect the delivered mean B-NCPAP, especially because they are dependent on the bias flow rate.6 We showed previously in a lung model that the pressure delivered to the nasal prongs in a B-NCPAP system was greater than the immersion depth of the expiratory tubing, even at the lowest flow rate causing gentle continuous bubbling. We also demonstrated that this pressure overshoot was systematically greater as the flow magnitude increased. By comparison, V-NCPAP-delivered pressures were equal to the set (desired) pressures and exhibited little flow dependence. In addition, as with any CPAP modality, the effective intrapulmonary NCPAP was substantially influenced by air leaks.9 The objectives of this study were to characterize the flow dependence of B-NCPAP delivery in a cohort of preterm infants and to compare the actual (delivered) intraprong CPAP with the intended (set) NCPAP for both V-NCPAP and B-NCPAP delivery over a range of set values and constant flow rates in the same preterm infants.
The infants studied were part of a trial comparing the work of breathing in premature infants during V-NCPAP and B-NCPAP therapy. Preterm infants of <1500-g birth weight who required NCPAP treatment because of mild respiratory distress (fraction of inspired oxygen: ≤0.40, to ensure tolerance of study procedures) were eligible for the study. Infants with airway anomalies, other major anomalies, or >28 days of life were excluded. The protocol was approved by the North Shore Long Island Jewish Health System institutional review board, and consent was obtained from both parents before study. Infants were studied with both devices, and the order of devices was randomized.
A calibrated pressure transducer (DP45-28; Validyne, Northridge, CA) was used to measure intraprong pressure. Data were collected by using a computerized data acquisition system (Biopac Systems, Goleta, CA), at a sampling rate of 1000 samples per second per channel. V-NCPAP treatment was administered by using a calibrated VIP Bird infant ventilator (Viasys Health Care, Yorba Linda, CA), with V-NCPAP delivery provided by using the positive end expiratory pressure control. The exhalation valve on this ventilator is under microcomputer control and works in conjunction with the exhalation valve pressure transducer to control CPAP.
Because no complete B-NCPAP devices are commercially available in the United States, we devised a system that allowed easy conversion from V-NCPAP delivery to B-NCPAP delivery and was similar to commonly used B-NCPAP systems. We submerged the expiratory limb of the ventilator tubing (length: 185 cm; internal diameter: 1 cm; noncorrugated tubing) in a chamber of water made specifically for B-NCPAP delivery (diameter: 7.5 cm; Airways Development, Kenilworth, NJ), to the level of NCPAP desired (in centimeters of water). We showed previously that this tubing length minimally affects the pressure at the prongs.9 Hudson prongs (Hudson Respiratory Care, Temecula, CA) were used for all infants with both devices. The largest prongs that fit comfortably in the infant's nares without blanching the surrounding tissue were used. Intraprong pressure monitoring was accomplished via the Hudson prong pressure-monitoring port, attached to the expiratory tubing. Minimal/no-leak conditions during data collection were ensured by gently holding the infant's mouth closed, if necessary.
To achieve our first objective (to characterize the flow dependence of B-NCPAP delivery in preterm infants), the B-NCPAP expiratory tubing was set to a fixed water immersion depth of 4 cm and the flow rate was slowly increased until the first sign of continuous bubbling was observed (lowest flow rate). The flow rate was then increased in increments of 2 L/min twice (lowest flow rate + 2 L/minute and lowest flow rate + 4 L/minute). After 5 minutes of stabilization at each setting, mean and peak-to-peak intraprong measurements were recorded by averaging the last 30 seconds of the measurement, during quiet breathing, for accuracy of analysis. These steps were repeated at a fixed expiratory tubing depth of 6 cm.
To achieve our second objective (to compare the actual [delivered] intraprong CPAP with the intended [set] NCPAP for both V-NCPAP and B-NCPAP delivery), each infant first received B-NCPAP treatment, to determine the lowest constant flow rate that provided continuous gentle bubbling at the highest NCPAP setting used in the study (7 cmH2O). Therefore, flow rates varied among infants but were constant in comparisons of V-NCPAP and B-NCPAP delivery for the same infant. Next, we measured intraprong pressures at 5 set NCPAP levels (3, 5, 7, 4, and 2 cmH2O, in that order), for 5 minutes at each setting. This was performed specifically as a positive end-expiratory pressure challenge, and results will be used for calculations of pulmonary mechanics, to be reported separately. Measurements were extended to obtain a period of quiet breathing as necessary. These measurements were repeated for B-NCPAP and V-NCPAP delivery; the order of devices was randomized. In all cases, the effective delivered NCPAP was estimated as the time-averaged pressure signal over the last 30 seconds of quiet breathing data at each CPAP setting.
Of the 19 infants enrolled in the trial, we were able to obtain complete data on flow dependence and actual versus intended pressures during B-NCPAP delivery for 12 infants. Demographic and baseline respiratory support data for all 12 enrolled infants are shown in Table 1. The birth weight (mean ± SD) was 1140 ± 267 g, gestational age 28.5 ± 1.9 weeks, and age at study 12.9 ± 8 days. Nine infants received NCPAP therapy at 5 cmH2O and 3 at 6 cmH2O before the study. The fraction of inspired oxygen before the study ranged from 0.21 to 0.28.
Flow Dependence of B-NCPAP Delivery
Figure 1 shows representative intraprong pressure values for B-NCPAP delivery, measured with the expiratory tubing set at a fixed immersion depth while the flow rate was increased from the lowest flow rate setting to higher flow rates in 2-L/min increments. The lowest flow rate for the measurements at 4-cm immersion depth varied between 4 and 9 L/min for the infants, and values were either the same or 1 to 2 L/min higher at 6-cm immersion depth. The intraprong pressure data illustrate the marked flow dependence of B-NCPAP delivery. Specifically, the delivered mean intraprong pressure was always higher than intended, even at the lowest flow rate, and increased systematically as flow rates increased. Also, B-NCPAP-delivered intraprong pressures were characterized by substantial noisy oscillations, whose magnitudes increased with increasing flow rates. Similar flow-dependence patterns for B-NCPAP delivery were observed for set B-NCPAP values (expiratory tubing immersion depth) of 4 and 6 cmH2O.
Figure 2A shows the actual (delivered) mean intraprong pressure as a function of increasing flow rate for each of the infants, whereas Fig 2B shows the combined averaged data. The delivered prong pressure during B-NCPAP delivery exhibited flow dependence for every infant at both 4 and 6 cmH2O. This effect was systematic and appreciable in all cases, and it was substantially greater than expected in some cases. The percentage overshoot was largest for the lower CPAP setting of 4 cmH2O at the highest flow rates (lowest flow rate + 4 L/min).
Comparison of B-NCPAP and V-NCPAP Delivery at Identical Flow Rates
Figure 3 depicts the actual (delivered) mean intraprong pressure as a function of multiple set pressure settings (2–7 cmH2O) for individual infants, as well as the corresponding averaged data (mean ± SD) for all 12 infants, with both B-NCPAP and V-NCPAP systems. For each infant, measurements with both devices and at all NCPAP settings (5 each) were performed at the same predetermined constant flow rate (see above). The within-infant flow rate setting (constant flow rate) used in this B-NCPAP versus V-NCPAP comparison varied between 4 and 9 L/min among the infants.
During V-NCPAP delivery, the actual pressure at the nasal prongs closely approximated the pressure set at the ventilator (approaching the line of unity) at every setting, for both individual values and average values for the 12 infants. During B-NCPAP delivery, the delivered prong pressure was higher for every infant and at every desired CPAP setting (range: 0.7–2.2 cmH2O higher; mean: 1.3 cmH2O higher), compared with the value set through submersion of the expiratory tubing. The average B-NCPAP overshoots were similar for all desired CPAP settings. This corresponded to the largest percentage overshoot (∼75%) at the lowest setting of 2 cmH2O. The overshoots were systematically less at higher settings, reaching ∼20% at a desired B-NCPAP delivery of 7 cmH2O.
Noninvasive respiratory support is widely used for infants with respiratory distress, in an attempt to decrease the complications of intubation and mechanical ventilation. Many forms of noninvasive ventilation are available, although limited data exist comparing the various devices. NCPAP treatment has been shown to decrease the need for mechanical ventilation,10 as well as the need for reintubation,11 in preterm infants.
B-NCPAP delivery, which was commonly used >30 years ago, has resurfaced as a popular method of NCPAP support, primarily because it is cost-effective and easy to use. Also, although a direct association between B-NCPAP therapy and low rates of bronchopulmonary dysplasia has not been rigorously demonstrated, this form of CPAP treatment has been championed by an institution reporting low incidences of bronchopulmonary dysplasia.4 Few studies of B-NCPAP delivery have been published; of those, 2 used lung models6,9 and 2 studied intubated subjects.5,8 Morley et al,7 in a study of 26 preterm infants receiving B-NCPAP therapy, found a significant difference in NCPAP prong pressure when slow bubbling at a low gas flow rate was compared with vigorous bubbling at a higher flow rate. Our study is the first to study B-NCPAP delivery provided to preterm infants with nasal prongs, compared with another commonly used form of NCPAP provision.
We showed previously, in a lung model study, that, whereas delivered V-NCPAP is largely flow independent, B-NCPAP is highly flow dependent.9 We have now confirmed that this is also true for preterm infants. The flow-independent property of V-NCPAP delivery is attributable to the fact that the ventilator adjusts the exhalation valve in the presence of changing flow rates, to maintain a constant delivered pressure. B-NCPAP delivery does not have a similar method of compensation. Because the delivered mean pressure is higher than the set B-NCPAP (Figs 2 and 3), any beneficial (or detrimental) effects noted by clinicians in comparisons of B-NCPAP and V-NCPAP delivery in infants are confounded by the fact that the B-NCPAP-delivered mean pressures are usually appreciably higher than the desired or set values.
Christensen et al12 showed that a bubble system using large-bore tubing (22 mm) and a large jar (12 × 18 × 24 cm) exhibited only minor flow dependence. However, this does not apply to the much-smaller components of systems used for neonates. We showed previously that the size of these components, including tubing length, tubing width, and chamber size, did not affect the flow dependence of B-NCPAP delivery at flow rates used for infants (5–10 L/min).9
Because B-NCPAP delivery does not have built-in mechanisms to stabilize intraprong pressures at varying flow rates, clinicians must carefully adjust the under-water tube immersion depth and flow rate combination and then confirm the delivered B-NCPAP through accurate intraprong pressure monitoring. Changes in air leaks at the nares-prong interface would further alter this balance. With V-NCPAP delivery, leaks decrease delivered pressure. During B-NCPAP delivery, the presence of a leak (leading to absent bubbling) often leads to operator increases in flow, thus increasing the delivered pressure. If head position or mouth closure decreases the leak, flow may not be readjusted, because bubbling would continue but would not alert the operator to increases in pressure. Together, these factors capture the unpredictable nature of B-NCPAP delivery as it is currently used in many nurseries across the United States.
Our data demonstrate that, in preterm infants, bubble intraprong pressure is highly variable, with mean values that overshoot the set CPAP and exhibit substantial flow dependence (greater overshoots at higher flow rates). Users should directly measure intraprong pressures to titrate B-NCPAP delivery in infants and should use the lowest flow rate at which bubbling occurs, to avoid substantial pressure overshoot. Studies comparing B-NCPAP and V-NCPAP delivery must ensure equivalent intraprong pressures for the comparisons to be meaningful. Otherwise, reported differences with B-NCPAP delivery, whether of clinical benefit or not, may be attributable at least in part to increased CPAP provision for a given set CPAP level.
- Accepted February 19, 2008.
- Address correspondence to Sherry E. Courtney, MD, MS, Division of Neonatology, Stony Brook University Medical Center, HSC, T-11, 060, Stony Brook, NY 11794-8111. E-mail:
This work was presented in part at the Society for Pediatric Research annual meeting; May 5–8, 2007; Toronto, CA.
Dr Kahn's current affiliation is Department of Neonatology, Joe DiMaggio Children's Hospital, Hollywood, FL.
The authors have indicated they have no financial relationships relevant to this article to disclose.
What's Known on This Subject
NCPAP provides noninvasive respiratory support and is commonly used to maintain lung volume and to reduce apnea of prematurity. NCPAP may reduce the need for intubation and perhaps the incidence of chronic lung disease.
What This Study Adds
With bubble NCPAP, intraprong pressure is highly variable. Mean values overshoot set values and exhibit substantial flow dependence. Users should directly measure intraprong pressures in infants and should use the lowest flow rate at which bubbling occurs.
- ↵Avery ME, Tooley WH, Keller JB, et al. Is chronic lung disease in low birth weight infants preventable? A survey of eight centers. Pediatrics.1987;79 (1):26– 30
- ↵Morley CJ, Lau R, De Paoli A, Davis PG. Nasal continuous positive airway pressure: does bubbling improve gas exchange? Arch Dis Child Fetal Neonatal Ed.2005;90 (4):F343– F344
- ↵Ho JJ, Subramaniam P, Henderson-Smart DJ, Davis PG. Continuous distending pressure for respiratory distress syndrome in preterm infants. Cochrane Database Syst Rev.2002;(2):CD002271
- ↵Davis PG, Henderson-Smart DJ. Nasal continuous positive airways pressure immediately after extubation for preventing morbidity in preterm infants. Cochrane Database Syst Rev.2003;(2):CD000143
- Copyright © 2008 by the American Academy of Pediatrics