PEDIATRICS Vol. 108 No. 3 September 2001, pp. 682-685

Work of Breathing During Constant- and Variable-Flow Nasal Continuous Positive Airway Pressure in Preterm Neonates

Paresh B. Pandit, MD^*, , Sherry E. Courtney, MD^{, §}, Kee H. Pyon, PhD^{, §}, Judy G. Saslow, MD^{, §}, and Robert H. Habib, PhD

From the ^* University of Medicine and Dentistry of New Jersey/Robert Wood Johnson Medical School, Camden, New Jersey;  Virtua-West Jersey Hospital, Voorhees, New Jersey; ^§ Department of Pediatrics, Division of Neonatology, The Children's Regional Hospital at Cooper Hospital/UMC, Camden, New Jersey; and  Mercy Children's Hospital at St. Vincent Mercy Medical Center and Department of Pediatrics, Medical College of Ohio, Toledo, Ohio.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Background.  Constant-flow nasal continuous positive airway pressure (NCPAP) often is used in preterm neonates to recruit and maintain lung volume. Physical model studies indicate that a variable-flow NCPAP device provides more stable volume recruitment with less imposed work of breathing (WOB). Although superior lung recruitment with variable-flow NCPAP has been demonstrated in preterm neonates, corroborating WOB data are lacking.

Objective.  To measure and compare WOB associated with the use of variable-flow versus constant-flow NCPAP in preterm neonates.

Methods.  Twenty-four preterm infants who were receiving constant-flow NCPAP (means, SD) and had birth weight of 1024 ± 253 g, gestational age of 28 ± 1.7 weeks, age of 14 ± 13 days, and FIO₂ of 0.3 ± 0.1 were studied. Variable-flow and constant-flow NCPAP were applied in random order. We measured changes in lung volume and tidal ventilation (V_T) by DC-coupled/calibrated respiratory inductance plethysmography as well as esophageal pressures at NCPAP of 8, 6, 4, and 0 cm H₂O. Inspiratory WOB (WOB_I) and lung compliance were calculated from the esophageal pressure and V_T data using standard methods. WOB was divided by V_T to standardize the results.

Results.  WOB_I decreased at all CPAP levels with variable-flow NCPAP, with a maximal decrease at 4 cm H₂O. WOB_I increased at all CPAP levels with constant-flow CPAP. Lung compliance increased at all NCPAP levels with variable-flow, with a relative decrease at 8 cm H₂O, whereas it increased only at 8 cm H₂O with constant-flow NCPAP. Compared with constant-flow NCPAP, WOB_I was 13% to 29% lower with variable-flow NCPAP.

Conclusion.  WOB_I is decreased with variable-flow NCPAP compared with constant-flow NCPAP. The increase in WOB_I with constant-flow NCPAP indicates the presence of appreciable imposed WOB with this device. Our study, performed in neonates with little lung disease, indicates the possibility of lung overdistention at CPAP of 6 to 8 cm H₂O with the variable-flow device. Further study is necessary to determine the efficacy of variable-flow NCPAP in neonates with significant lung disease and its use over extended periods of time.continuous-flow and variable-flow NCPAP, work of breathing, premature neonates, lung compliance.

Nasal continuous positive airway pressure (NCPAP) often is used in preterm neonates to recruit and maintain lung volume (V_L). NCPAP usually is provided by varying the resistance to exhalation while constant gas flow is delivered by a neonatal ventilator through nasal prongs (constant-flow NCPAP).

An NCPAP device that uses demand or variable gas flow (variable-flow NCPAP) is available.^1,2 Physical model studies comparing constant-flow and variable-flow NCPAP report a relative decrease in airway pressure variability during breathing with variable-flow NCPAP. This indicates a potential for superior lung recruitment and maintenance of V_L.^1,2 Moreover, Klausner et al² found that the imposed work of breathing (WOB) with the variable-flow NCPAP prongs was one fourth that of conventional constant-flow NCPAP prongs, ascribing this to differences in prong design. Infant data to support these findings are scarce. In particular, we are unaware of any published data on WOB during variable-flow NCPAP support.

We recently demonstrated that variable-flow NCPAP provides superior lung recruitment for similar CPAP levels in preterm neonates.³ However, the gas flows required to generate an equivalent CPAP with the variable-flow device generally are greater than with the constant-flow device. Although the design of the variable-flow NCPAP prongs allows for excess gas flow to be diverted away from the patient, it is unclear whether WOB is affected. Thus, our objective was to measure and compare WOB associated with the use of variable-flow versus constant-flow NCPAP in preterm neonates.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

This study was approved by the Institutional Review Board of Cooper Hospital/University Medical Center. Premature infants who weighed <1800 g at birth and were receiving constant-flow NCPAP for apnea or mild respiratory distress were eligible for enrollment, provided that they were otherwise medically stable. Informed parental consent was obtained before testing.

Data Acquisition and Analysis

The study design and data acquisition have been described in detail elsewhere.³ Briefly, NCPAP was provided to each infant with each of 3 devices applied in random order: 1) a modified nasal canula, 2) a standard nasal CPAP prongs (constant-flow), and 3) the new variable-flow NCPAP generator with prongs (variable-flow). Because a nasal canula is not used commonly to provide NCPAP, our WOB analysis was restricted to the 2 NCPAP delivery systems using prongs:

1. Constant-flow NCPAP was delivered by connecting Inca nasal prongs (Ackrad Laboratories, Cranford, NJ) to an infant ventilator set in CPAP mode. Adjusting the CPAP setting on the ventilator varied the amount of airway pressure applied. A continuous gas flow of 6 L/min was used. The largest prongs that fit the infant's nares without blanching the surrounding tissue were used.
2. Variable-flow NCPAP was delivered by the Aladdin/Infant-Flow system (Hamilton Medical, Reno, NV; manufactured by EME, Ltd, Brighton, UK; and currently distributed as the Infant-Flow Nasal CPAP system by SensorMedics Corp, Yorba Linda, CA). Changing the amount of gas flow varied the amount of CPAP. The largest prongs that fit easily into the nares were used for each infant.

Measurements were performed after a feeding to facilitate quiet sleep. Sedation was not used in any infant. Infants were instrumented and placed in the supine position. Instrumentation consisted of the following: 1) respiratory inductance plethysmography (RIP) bands that were fitted around the chest and the abdomen (Respiband Plus, SensorMedics Corp, Yorba Linda, CA; and Nims Inc, Miami Beach, FL); 2) insertion of a neonatal esophageal balloon catheter (Ackrad Laboratories, Cranford, NJ) at the level of the lower third of the trachea to estimate intrapleural pressure from the esophageal pressure (Pes); and 3) placement of a thermistor (BreathSensor, Nellcor Puritan Bennett, Eden Prairie, MN) to detect and continuously record (EdenTrace II Plus, EdenTec, Eden Prairie, MN) air leaks from the mouth.

RIP was used to measure lung volume changes (V_L) and tidal ventilation (V_T) from its direct current (DC) and alternating current components, respectively (Somnostar, SensorMedics Corp., Yorba Linda, CA).^4,5 After the RIP bands were placed, RIP was calibrated by direct comparison^6,7 to leak-free flow and volume data measured by face mask pneumotachography (Neonatal Flow Sensor #7218 [dead space 0.8 mL], Novametrix, Wallingford, CT). Before measurements began, proper placement of the Pes balloon catheter was checked by continuous on-line monitoring of Pes and adjusted until a high correlation (r² > 0.90) was obtained between airway opening pressure and Pes during spontaneous breathing by the infant against an occluded airway.⁸ When necessary, the infant's mouth was closed gently during data collection to stop any air leak, and data with air leak at the mouth were not used. Airway flow and Pes from a series of 10 to 15 leak-free breaths were integrated to calculate V_T. All signals were sampled at 100 Hz, monitored on-line, and stored on a computer for later analysis. Patients initially were placed on NCPAP of 8 cm H₂O to allow similar lung recruitment. With each device, V_L, V_T, and Pes were measured at NCPAP of 8, 6, 4, and 0 cm H₂O. Infants were kept for 3 to 5 minutes at each CPAP level. The leak-free breaths spanning the last 20 to 30 seconds at each setting were selected for subsequent analysis.

Lung compliance (C_L) was calculated from V_T and Pes data using standard methods.⁸ Inspiratory WOB (WOB_I) was calculated from the area subtended by the inspiratory limb of the Pes-V_T curve, according to Campbell's diagram.⁹ Resistive WOB (RWOB) was calculated as the area between the inspiratory and expiratory loops of the Pes-V_T curve. For standardizing the results from different infants and for varying breathing amplitudes, WOB was divided by V_T.

Sample Size Calculation/Statistical Analysis

Sample size was based on finding a clinically significant V_L between any 2 NCPAP devices as previously reported by us.³ A sample size of 28 to 32 patients is required to attain a statistical power of 0.8 to 0.85 when one assumes a 20% difference to be significant with a 0.05 significance criterion.¹⁰ Dependent variables were analyzed as mixed linear models in a randomized block factorial design.¹¹ Devices and CPAP levels were considered fixed effects, and patients were treated as random blocks. Differences between devices, within NCPAP levels, and devices-within-NCPAP levels were tested using least square, pair-wise mean comparisons. The effects of missing cells were adjusted for by using Satterthwaite approximations.¹¹ P < .05 was considered to be significant.

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

Thirty-five infants were recruited for study. Of these, Pes monitoring and RIP data that allowed accurate calculations of WOB were obtained in 24 infants. The patient characteristics of these 24 infants including baseline C_L are shown in Table 1.

Changes (means, SD) in respiratory rate (RR; min¹) and V_T with NCPAP are shown in Table 2. RR was similarly decreased as NCPAP increased for the 2 devices. V_T was not changed with NCPAP for either device but was greater for the variable-flow device at all CPAP levels (P < .001).

V_L as a function of NCPAP for each device are shown in Fig 1, top. Although V_L increased with NCPAP with both systems, V_L was significantly greater overall with the variable-flow device (P < .001). The corresponding C_L comparison is shown in Fig 1, bottom. C_L essentially was unchanged with constant-flow NCPAP but was slightly increased at CPAP of 4 and 6 cm H₂O, relative to 0 cm H₂O, for the variable-flow device. At CPAP of 8 cm H₂O, this tendency was reversed and C_L was slightly (not significantly) lower with variable flow.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Changes in VL (top) and CL (bottom) with NCPAP. *The overall VL (mean ± standard error [SE]) was significantly greater with variable-flow compared with constant-flow NCPAP (P < .001) and was associated with increased CL (mean ± SE). CL essentially was unchanged with constant-flow NCPAP. *CL generally was greater with variable-flow NCPAP (P < .05). Note the relative decrease in CL at 8 cm H2O with the variable-flow device, indicating the possibility of lung overdistension.

Changes in WOB_I and its component RWOB per milliliter of delivered V_T are shown in Fig 2. Overall, both WOB_I and RWOB were significantly lower with the variable-flow device compared with the constant-flow system (P < .001). WOB_I with variable-flow NCPAP was lower by 28.8%, 29.0%, and 13.6% at CPAP of 4, 6, and 8 cm H₂O, respectively. At CPAP of 0 cm H₂O, RWOB was significantly greater with constant-flow NCPAP (P < .05), possibly indicating the effects of the larger mechanical impedance of the Inca nasal prongs. This difference, however, did not affect WOB_I at 0 cm H₂O.

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Changes in WOBI (mean ± SE) and its component RWOB (mean ± SE). Overall, both WOBI and RWOB were significantly lower with the variable-flow device compared with constant-flow NCPAP (P < .001). Note that RWOB was greater at 0 cm H2O with constant-flow NCPAP (P < .05), possibly due to a larger mechanical impedance of the constant-flow nasal prongs.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

NCPAP is increasingly being used as the primary method of ventilatory support in preterm infants with respiratory distress syndrome. This trend is based on the perceived advantages of this relatively noninvasive approach, including avoiding the risks and complications of intubation and mechanical ventilation. More important, compared with intubation and mechanical ventilation, NCPAP is believed to decrease the likelihood of pulmonary barotrauma and volutrauma in infants and consequently the development of chronic lung disease.¹²

The conventional method used to provide NCPAP uses constant-flow CPAP delivered via nasal prongs connected to a neonatal ventilator. In a newer system that provides variable-flow NCPAP, a flow driver delivers gas via specially designed nasal prongs.¹ Studies comparing the efficacy of these 2 approaches remain scarce and are limited mostly to mechanical models.² Furthermore, no studies have compared the possible effects of constant- versus variable-flow NCPAP on WOB and lung mechanics in infants. This perhaps is due to the difficulty of obtaining the relevant WOB measurements in neonates, particularly without altering the delivery of NCPAP.

In mechanical model studies, constant-flow NCPAP has been associated with increased WOB compared with both mask CPAP¹³ and variable-flow NCPAP.² The design of the variable-flow NCPAP system offers several advantages during both inspiration and expiration. During inspiration, the design of the variable-flow nasal prongs results in high-velocity jet flows, allowing gas entrainment. This assists inspiration on demand, which, in turn, keeps the CPAP level constant.^1,2 A less variable CPAP would be expected to improve and maintain V_L recruitment with CPAP.

We recently reported that variable-flow NCPAP led to superior V_L compared with constant-flow NCPAP.³ Such increased V_L will improve lung mechanics and, hence, may decrease WOB, provided that parenchymal overdistention is avoided. Note that, on exhalation, the variable-flow nasal prongs allows any excess gas flow from the CPAP driver to be shunted away from the patient to ambient air, through an expiratory outlet. This is in contrast to constant-flow NCPAP whereby gas flow continues toward the nares during exhalation.^1,2,14 With constant-flow NCPAP, the patient's expiratory effort then must overcome this excess flow, leading to increased expiratory work.

Arguably, these advantages of variable-flow NCPAP should translate into superior clinical performance, including an improved success rate of extubation. Data supporting this were presented recently for ventilated preterm infants who were extubated prophylactically to CPAP. Here, investigators found that the rate of successful extubation was greater with the Infant-Flow system compared with constant-flow nasal or nasopharyngeal CPAP.^15,16 However, studies investigating the long-term or prophylactic use of variable-flow NCPAP in the treatment of preterm infants have not been reported.

Critique of Experimental Methods

Measurement of lung mechanics and WOB during NCPAP is not possible with standard methods, whereby flow and volume are measured at the airway opening, without affecting the results. This is because placement of an in-line flow-measuring device such as a pneumotachograph will interfere with CPAP delivery, will cause additional imposed WOB, and is cumbersome and impractical to use in a clinical setting. RIP provides an attractive alternative whereby measurements of volume changes (and hence flow) are done distally at the chest wall and therefore does not interfere with or alter how CPAP is delivered to patients. Moreover, with the use of DC-coupled RIP, careful measurements allow a simultaneous assessment of both V_L (DC component) and V_T (AC component). Measurement of changes in V_L is not possible from pneumotachography. A possible disadvantage of RIP is that it relies on an incomplete sampling of the thorax (1 abdominal and 1 rib-cage band; see Methods) to infer volume changes. This may limit its accuracy, particularly in the presence of significant chest wall distortion. Although we cannot discount completely the possible effects of this limitation of RIP, we believe that its importance is secondary as 1) we carefully marked the location of RIP bands to ensure identical placement, and 2) the devices were tested in random order with each patient serving as his or her own control.

In this study, performed in preterm neonates with minimal lung disease, we showed that, compared with constant-flow NCPAP, a variable-flow device is able to recruit more V_L (Fig 1, top) and increase the effective C_L (Fig 1, bottom) and that it is able to do this with relatively decreased WOB (Fig 2). The latter is due partly to the larger RWOB with the constant-flow device when no CPAP flow is provided (ie, CPAP of 0 cm H₂O). This indicates the presence of additional imposed WOB due to the Inca prongs (which were used exclusively in this study with the constant-flow device). Note that other commercially available nasal prongs used with conventional NCPAP may lead to less (or more) imposed WOB, depending on their specific design.

As mentioned above, lung recruitment was noticeably increased for 6 and 8 cm H₂O variable-flow NCPAP (Fig 1). Although the associated WOB at these NCPAP settings remained lower than for NCPAP = 0, they did not continue to decrease relative to NCPAP = 4 (Fig 2). In fact, despite greater V_L, WOB was slightlyalbeit not significantlygreater than for NCPAP = 4. Given that 1) gas compression compliance is necessarily increased at higher V_L and 2) RWOB was not increased as variable-flow NCPAP increased (Fig 2), our data might point to the possibility that lung overdistension at the higher NCPAP levels (or equivalent V_L) may have occurred in some patients with the variable-flow system. This is perhaps an indication of the increased efficacy of maintaining a constant airway pressure (and hence V_L recruitment) with variable-flow NCPAP.^1,2 Therefore, caution is advisable when titrating toward the higher NCPAP levels with this device.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

Variable-flow nasal CPAP potentially can provide superior respiratory support in neonates that is characterized by 1) greater and more stable volume recruitment, 2) improved C_L, and 3) decreased WOB. Our findings, however, are limited to its short-term use in 24 neonates with mild lung disease or apnea. Similar investigation of the use of 1) variable-flow nasal CPAP in a larger number of neonates, 2) for longer periods of time, 3) in comparison with other nasal prong designs/constant-flow NCPAP, and 4) in infants with more significant lung disease are needed for confirmation of the reported findings.

	ACKNOWLEDGMENTS

This work was supported in large part by a grant from the Cooper Faculty Practice Foundation. Hamilton Medical, Inc (Reno, Nevada) provided the variable-flow nasal CPAP device, manufactured by EME Ltd (Brighton, England).

We thank Gerald K. Arnold, PhD, for help in statistical analysis.

	FOOTNOTES

Received for publication Dec 7, 2000; accepted Feb 8, 2001.

This work was presented in part at the Pediatric Academic Societies' Annual Meeting; May 1-4, 1999; San Francisco, CA.

Reprint requests to (R.H.H.) Mercy Children's Hospital, 2213 Cherry St, ACC 309, Toledo, OH 43608. E-mail: robert_habib{at}mhsnr.org

	ABBREVIATIONS

NCPAP, nasal continuous positive airway pressure; V_L, lung volume; WOB, work of breathing; RIP, respiratory inductance plethysmography; Pes, esophageal pressure; V_T, tidal ventilation; V_L, change in lung volume; C_L, lung compliance; DC, direct current; RR, respiratory rate; SE, standard error.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Moa G, Nilsson K, Zetterstrom H, Jonsson LO A new device for the administration of nasal constant positive airway pressure in the newborn: an experimental study. Crit Care Med 1988; 16:1238-1242 [Medline]
2. Klausner JF, Lee AY, Hutchison AA Decreased imposed work with a new nasal continuous positive airway device. Pediatr Pulmonol 1996; 22:188-194 [CrossRef][Medline]
3. Courtney SE, Pyon KH, Saslow JG, Arnold GK, Pandit PB, Habib RH Lung recruitment and breathing pattern during variable versus continuous flow nasal continuous positive airway pressure in premature infants: an evaluation of three devices. Pediatrics 2001; 107:304-308 [Abstract/Free Full Text]
4. Locke R, Greenspan JS, Shaffer TH, Rubenstein SD, Wolfson MR Effect of nasal CPAP on thoracoabdominal motion in neonates with respiratory insufficiency. Pediatr Pulmonol 1991; 11:259-264 [Medline]
5. Sandberg KL, Lindstrom DP, Krueger ED, Sundell H, Cotton RB Measurement of tidal volume during high frequency ventilation by impedance plethysmography. Pediatr Res 1988; 23:253-256 [Medline]
6. Warren RH, Alderson SH Face mask application for calibration of respiratory inductive plethysmography in lambs. J Dev Physiol 1988; 10:175-178 [Medline]
7. Brooks LJ, DiFiore JM, Martin RJ, CHIME Study Group Assessment of tidal volume over time in preterm infants using respiratory inductance plethysmography. Pediatr Pulmonol 1997; 23:429-433 [CrossRef][Medline]
8. Pandit PB, Pyon KH, Courtney SE, England SJ, Habib RH Lung resistance and elastance in spontaneously breathing pre-term infants: effects of breathing pattern and demographics. J Appl Physiol 2000; 88:997-1005 [Abstract/Free Full Text]
9. Harris TR, Wood BR. Physiologic principles. In: Goldsmith JP, Karotkin EH, eds. Assisted Ventilation of the Neonate. 3rd ed. Philadelphia, PA: WB Saunders; 1996:21-68
10. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum; 1988:48-52, 274-280
11. Kirk R. Experimental Design: Procedures for the Behavioral Sciences. Pacific Grove, CA: Brooks/Cole; 1995:454-473
12. Van Marter LJ, Allred EN, Pagano M, Do clinical markers of barotrauma and oxygen toxicity explain interhospital variation in rates of chronic lung disease? The Neonatology Committee for the Developmental Epidemiology Network. Pediatrics 2000; 105:1194-1201 [Abstract/Free Full Text]
13. Goldman SL, Brady BP, Dumpit FE Increased work of breathing associated with nasal prongs. Pediatrics 1979; 64:160-164 [Abstract/Free Full Text]
14. Moa G, Nilsson K Nasal continuous positive airway pressure: experiences with a new technical approach. Acta Pediatr 1993; 82:210-211 [Medline]
15. Sun SC, Tein HC Randomized controlled trial of two methods of nasal CPAP (NCPAP): flow driver vs conventional NCPAP [abstract]. Pediatr Res 1999; 45:322A
16. Roukema H, O'Brien K, Nesbitt K, Zaw W A crossover trial of infant flow (IF) continuous positive airway pressure (CPAP) versus nasopharyngeal (NP) CPAP in the extubation of babies  1250 grams birthweight [abstract]. Pediatr Res 1999; 45:317A