Journal of Perinatology (2014) 34, 464–467 © 2014 Nature America, Inc. All rights reserved 0743-8346/14 www.nature.com/jp
ORIGINAL ARTICLE
Bringing back the old: time to reevaluate the high-frequency ventilation strategy A Mukerji1, J Belik1 and M Sanchez-Luna2 OBJECTIVE: To examine the role of frequency in high-frequency ventilation (HFV) on carbon-dioxide (CO2) elimination and lung injury, independent of its effect on tidal volume. STUDY DESIGN: An anatomically representative lung model was attached to a mechanical ventilator capable of providing HFV with a constant volume. CO2 was infused directly into the lung, and a commercially available end-tidal CO2 detector was used to determine CO2 elimination. CO2 elimination and amplitude of pressure transmissions were evaluated using frequencies ranging from 5 to 15 Hz. The pressure–volume index (PVI) was described as the product of the volume and pressures delivered to the lung, a surrogate for lung injury. RESULT: The use of increasing frequencies directly correlated with improved CO2 clearance when keeping the tidal volume fixed, expressed as percent CO2 remaining in the lung at 25 s (66.5 (±1.1)%, 50.5 (±0.1)% and 37.8 (±0.3)% at 5, 10 and 15 Hz, respectively, P o0.05). With a fixed tidal volume, there was a decrease in pressure amplitudes transmitted to the lung with a decline in the PVI (53.9 (±2.7) mmHg ml − 1, 41.1 (±0.9) mmHg ml − 1 and 23.4 (±3.6) mmHg ml − 1, at 5, 10 and 15 Hz, respectively, P o 0.05). CONCLUSION: Frequency has a direct relationship with CO2 elimination when tidal volume is fixed. Using low delivered tidal volumes and high frequencies may allow for improved ventilation efficacy, while minimizing lung injury. Journal of Perinatology (2014) 34, 464–467; doi:10.1038/jp.2014.39; published online 13 March 2014 Keywords: pressure–volume index; ventilation efficacy; CO2 elimination; tidal volume
INTRODUCTION High frequency ventilation (HFV) has been used in neonatal intensive care units for more than two decades, mostly in infants with severe respiratory failure as a rescue mode of ventilation.1,2 It uses low tidal volumes at supra-physiological respiratory frequencies,3 and has been proposed as an effective respiratory strategy that may reduce ventilator-induced lung injury.4,5 Yet in the clinical setting, the superiority of HFV over conventional ventilation in the prevention of chronic lung disease remains questionable.6 The factors accounting for the apparent discrepancy between the theoretical advantage of HFV and available clinical data showing at best a rather small reduction in the chronic lung disease prevalence amongst infants ventilated with this modality7,8 are unclear. HFV is more effective than conventional ventilation for carbondioxide (CO2) elimination.9 Animal and bench studies conducted over 30 years ago showed that the efficacy of CO2 elimination in HFV is related to the tidal volume (VT) and frequency (f), best expressed as V2T × f,10,11 often described as the diffusion coefficient of CO2 (DCO2), a measure of CO2 elimination. As is evident by the aforementioned relationship, an increase in either VT or f would result in improved CO2 elimination. The original data on which the equation was based clearly indicate that the higher the f the greater the CO2 elimination. In spite of the long-standing evidence of the clear benefit of using the highest possible f, the ventilatory strategy's most commonly utilized starting frequency in various studies and clinical guidelines is 10 Hz.7,8,12 One obvious reason for such an approach
relates to the fact that until recently none of the commercially available ventilators allowed for the independent adjustment of VT and f. In these ventilators an increase in f leads to a concomitant reduction in VT, thereby reducing DCO2.13 Furthermore, a number of unproven and mostly theoretical concerns have been raised against the use of frequencies higher than 12 Hz. These concerns mostly relate to the fact that in order to sustain VT at high frequencies an incremental pressure amplitude in the circuit is needed, raising fears that these would be transmitted to the alveoli and result in lung injury. The current strategy of limiting f is geared at minimizing alveolar pressure swings and peak alveolar pressures,14 as well as to reduce the pressure cost of ventilation.15 The advent of a commercially available ventilator able to sustain a fixed VT allows one to overcome the decrease in VT with increasing frequencies, thus enabling an actual increase in DCO2 with higher f. In addition, currently available ventilators measure volume at the distal tubing, providing a much more reliable measure of the VT, as compared with previous ventilator prototypes used in experimental models that depended on a ‘stroke volume’ at the ventilator end.16,17 The goal of the present study was to evaluate CO2 elimination at a fixed VT and varying frequencies. We hypothesized that the proximal pressure amplitude is dampened such that the distal values do not significantly increase with the rise in frequency. As such a study is not feasible in humans, we developed an artificial lung model where CO2 elimination could be measured and amplitude pressures monitored.
1 Division of Neonatology, Department of Pediatrics, Hospital for Sick Children, University of Toronto, Toronto, ON, Canada and 2Division of Neonatology, Instituto de Investigación Sanitaria Gregorio Marañón, Hospital General Universitario “Gregorio Marañón”, University Complutense of Madrid, Madrid, Spain. Correspondence: Dr M Sanchez-Luna, Neonatology Division, Instituto de Investigación Sanitaria Gregorio Marañón, Hospital General Universitario “Gregorio Marañón”, C/ Dr. Esquerdo 46, Madrid E-28007, Spain. E-mail [email protected] Received 21 November 2013; revised 19 January 2014; accepted 5 February 2014; published online 13 March 2014
Reevaluating the high-frequency ventilation strategy A Mukerji et al
465 METHODS The experimental setup consisted of an artificial lung model with a volume of 500 ml, which had a measured compliance of 4.2 ml cm − 1 H2O, and resistance of 8 cm H2O l − 1 s − 1. This lung model was attached to connectors simulating the trachea (measured volume 35 ml, measured diameter 1.8 cm). The trachea was directly attached to flow tubings from a mechanical ventilator capable of providing HFV both with and without a constant VT (VN500, Dräger Medical, Lubeck, Germany). The ventilator designated a constant VT as ‘volume guarantee’ (VG). The ventilator’s flow sensor was placed between the ventilator tubing (RT series, Fisher and Paykel Healthcare, Laval, Canada) and the trachea. A schematic of the experimental setup is shown in the online Supplementary Figure. The amplitude of pressure transmission (pulse pressure) was measured quasi-simultaneously with a commercially available pressure transducer (Model 041-508-001, Cobe Laboratories, CO, USA) and a rigid catheter at three positions denoted as P1, P2 and P3, as depicted in the online Supplementary Figure. P1 was located at the most distal end of the ventilator tubing, P2 referred to the anatomically equivalent trachea and P3 was measured inside the model lung, analogous to the distal lung parenchyma. CO2 was infused directly into the lung continuously from a source with 100% CO2 at a flow rate of 100 ml min − 1. Once the mode of ventilation to be tested was initiated and the CO2 level in the test lung reached a state of equilibrium, the infusion was turned off and the rate of CO2 clearance was measured, as a surrogate for ventilation efficacy. A catheter at P2 connected to an end-tidal CO2 detector (Model CD-102-45-02 PuritanBennett Corporation, Los Angeles, CA, USA) acquired the CO2 levels at P2 continuously. This analog data, along with the pressure measurements, were acquired at a rate of 20 data points per second and stored for further analysis by a custom-made computer software program. The concentration of CO2 remaining in the lung at 25 s after the infusion was turned off, expressed as a percentage of the initial concentration of CO2 recorded, was analyzed as a measure of ventilation efficacy. The mean airway pressure was set at 10 cm H2O for all experiments. When using HFV without a constant VT, the amplitude was set to 30 cm H2O and in HFV with a constant VT, the VT was set at 15 ml, with a maximum allowable pressure amplitude of 80 cm H2O. The frequencies were tested from 5 to 15 Hz (highest frequency provided by the manufacturer). Measurements from each ventilator setting were performed in triplicate to assure precision, and DCO2 values were calculated for each experiment run from the frequency and VT measurements from the ventilator. Continuous positive airway pressure with a positive-end expiratory pressure of 5 cm H2O was used as a control. The I:E (inspiratory:expiratory) time ratio was kept constant at 1:1 for all the experiments. Data are presented as mean ± s.e. and were analyzed by two-way analysis of variance and Tukey–Kramer multiple comparison testing (NCSS, LLC Kaysville, UT, USA). P o0.05 was accepted as statistically significant.
RESULTS The DCO2 and VT values were compared against frequencies of 5 to15 Hz both with and without a constant VT (‘VG’ on and ‘VG’ off, respectively), as shown in Figure 1. When plotted against HFV without a constant VT, the DCO2 values decreased as the frequencies increased, whereas the opposite was seen with a constant VT (Figure 1a). The VT volumes decreased with higher frequencies in regular HFV, and as expected, remained constant when the VT was set at a fixed value (Figure 1b). Interestingly, however, the VT values showed a decline with frequencies ⩾ 14, indicative of the actual maximum frequency for which VT was maintained constant with this device. Ventilation efficacy, determined by CO2 elimination, was compared for the various modes tested. When the lung was connected to the ventilator set on continuous positive airway pressure, no CO2 elimination was detected (data not shown). The CO2 clearance, as determined by the percentage of CO2 remaining in the lung at 25 s, was compared for HFV with and without a constant VT for frequencies from 5 to 15 Hz, as shown in Figure 2. With a constant VT, CO2 elimination improved with increasing frequencies (66.5 (±1.1)%, 50.5 (±0.1)% and 37.8 (±0.3)% at 5, 10 and 15 Hz, respectively, P o0.05), whereas the opposite was seen © 2014 Nature America, Inc.
Figure 1. (a) DCO2 values plotted against frequency with and without a fixed VT. With a fixed VT, DCO2 values increased until 13 Hz showing a drop-off at 14 and 15 Hz. Without a fixed VT, the DCO2 values declined with higher frequencies. (b) Delivered VT values plotted against frequency with and without a fixed VT. As expected with a fixed VT on the ventilator, the delivered VT remained constant, whereas they decreased with increasing frequencies when not fixed. **P o0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. DCO2, diffusion coefficient of CO2; VG off, no constant VT; VG on, constant VT; VT, tidal volume.
for HFV without a constant VT (28.1 (±0.5)%, 43.8 (±0.1)%, and 51.9 (±0.4)% at 5, 10 and 15 H z, respectively, P o 0.05). The pulse pressure delivered to the lung was also evaluated, as measured at P3. There was a decrease in the pressure amplitudes at P3 with increasing frequencies, as depicted in Figure 3, and this was statistically significant, as compared with the data obtained while VT was not kept constant. The pressure transmission to the lung (P3) was combined with VT to describe a pressure–volume index (pressure × VT)—denoted as PVI. The PVI was compared at frequencies of 5 to 15 Hz both with and without a constant VT, as shown in Figure 4. The PVI values were 262.9 (±3.1) mmHg ml − 1, 56.1 (±0.7) mmHg ml − 1 and 13.1 (±1.3) mmHg ml − 1 without a constant VT (P o 0.05), and were 53.9 (±2.7) mmHg ml − 1, 41.1 (±0.9) mmHg ml − 1 and 23.4 (±3.6) mmHg ml − 1 with a fixed VT (P o 0.05), all at 5, 10 and 15 Hz, respectively. DISCUSSION In this study of HFV using a constant VT, we found frequencies directly correlates with CO2 elimination, and at higher frequencies, distal lung equivalent transmission of pulse pressure was negligible. In fact, the transmitted pulse pressures decreased with Journal of Perinatology (2014), 464 – 467
Reevaluating the high-frequency ventilation strategy A Mukerji et al
466
Figure 2. CO2 elimination efficacy, as determined by percentage of CO2 remaining in the lung at 25 s after ventilation initiation, plotted against frequencies with and without fixed VT. CO2 elimination was directly proportional to f when VT was fixed. **Po0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. VG off, no constant VT; VG on, constant VT; VT, tidal volume.
Figure 3. Amplitude of pressure transmission to the lungs plotted against frequency with and without a fixed VT. Despite higher pressure amplitudes at the ventilator with increasing frequency, the distal pressure amplitude showed a decline with a fixed VT. **P o0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. VG off, no constant VT; VG on, constant VT; VT, tidal volume.
increasing frequencies. We also demonstrated a decrease in PVI, a theoretical surrogate for ventilation injury that incorporates barotrauma and volutrauma, across all frequencies with a constant VT. In contrast, without a constant VT, maximal ventilation efficacy was achieved at the lowest frequencies, but this was associated with a significantly higher PVI. This study confirms the previously described DCO2 equation11 indicating that CO2 clearance with constant VT is directly proportional to frequency. However, in this study, we show for the first time that improving ventilation efficacy with higher frequencies in HFV is feasible with newer generation ventilators that reliably deliver a constant VT, and this may be achievable with reduced lung injury. We did, however, note that the ventilator utilized for the present study is unable to sustain VT when frequencies were increased beyond 14 and 15 Hz, under the conditions used in this lung model. Journal of Perinatology (2014), 464 – 467
Figure 4. Pressure–volume index plotted against frequency with and without fixed VT. The PVI dropped sharply without a constant VT until frequency of 11 Hz, then remained relatively constant. On the other hand, with a fixed VT, the PVI remained stable at all frequencies. **P o0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. PVI, pressure–volume index; VG off, no constant VT; VG on, constant VT; VT, tidal volume.
In their study of HFV on dogs, Watson et al.17 found a similar relationship between frequency and ventilation efficacy as ours. But there was a lack of reliability of the delivered volume, which was being measured at the ventilator end. Moreover, there was concern regarding the use of higher frequencies, as it was presumed that higher frequencies would lead to higher airway and lung pressures to deliver the desired volume and increase the risk of barotrauma.17 Despite many other studies similarly demonstrating a clear relationship between increasing frequencies and improved ventilation,16 the lack of reliability of delivering tidal volumes and the inability to measure the pressure transmission to the airways and lungs in the animal studies led to uncertainty regarding the ideal frequency for HFV. Until recently, most commercially available ventilators did not allow for a constant deliverable VT. Some recent clinical studies of HFV have utilized starting frequencies of 10 Hz, adjusting the amplitude to optimize ventilation efficacy7,8 despite previous recommendations to use higher set frequencies.3,9 Our data suggest that if VT can be held constant and reliably delivered, the highest frequency should be used to maximize ventilation efficacy while minimizing barotrauma, challenging the current practice of using a starting frequency of 10 Hz in HFV. A meta-analysis of these and other studies demonstrate marginal benefit of HFV on ventilator-induced lung injury and chronic lung disease in preterm infants.6 The relationship between tidal volumes, alveolar stretch and lung injury is well described,18–20 and utilizing higher frequencies with lower VT in HFV therefore may result in reduced lung injury. We postulate that the limited impact of HFV in reducing lung morbidity may be related to a failure in exploiting the well-known relationship of ventilator frequency and tidal volume when using this ventilation modality. We here propose a new strategy for HFV ventilation taking into account the aforementioned previous studies and our current findings. We suggest using HFV with a constant VT and identifying the minimum VT that results in adequate ventilation at a frequency of 10 Hz, noting the DCO2 value. We then suggest increasing the frequency as high as allowable while decreasing the VT that maintains the same DCO2. We speculate that newer generation ventilators may optimize delivery of VT at frequencies higher than the currently allowed. Using HFV with a constant VT may also confer other advantages such as better control of ventilation, as the DCO2 will be under direct control of the clinician. In fact, some studies have © 2014 Nature America, Inc.
Reevaluating the high-frequency ventilation strategy A Mukerji et al
467 demonstrated significant variations in actual delivered VT using flow sensors when HFV is used without a constant set VT.21,22 This study does have several limitations. It employs a lung model with fixed mechanical properties, lacking the parenchymal inhomogeneity that characterizes the distinct neonatal lung conditions. Also, given the lack of branching airways in our model, the pressure transmission to the parenchyma may be further attenuated in a human lung. Further, our test lung had a volume representative of a 10 kg infant and the measured compliance and resistance were not representative of typical neonatal values. The ventilator employed in this study was designed for neonates, which may explain the limitation of delivering the set VT at frequencies of 14 and 15 Hz. Further testing utilizing more advanced lung models with a wide range of lung mechanics and volumes, along with branching airways with progressively smaller diameters, is warranted. We suggest a critical reevaluation of the current clinical practice of HFV use in neonatal intensive care units and propose the use of the highest possible frequency that achieves adequate ventilation in all infants to minimize lung injury. The clinical benefit of utilizing such a strategy is unknown and requires further clinical investigation. Such studies, however, are difficult to design given the challenge of measuring the impact of changes in f, VT and DCO2 on CO2 elimination and ventilation-induced lung injury in vivo. Yet the present data warrant future clinical studies targeted at evaluating the potential beneficial impact of the proposed ventilatory strategy on the reduction of severity and/or prevalence of bronchopulmonary dysplasia in preterm infants. CONFLICT OF INTEREST MSL declares having received advisory board consulting fees from Drager. The remaining authors declare no conflict of interest.
ACKNOWLEDGEMENTS This study received no specific funding.
REFERENCES 1 Ventre KM, Arnold JH. High frequency oscillatory ventilation in acute respiratory failure. Paediatr Respir Rev 2004; 5: 323–332. 2 Moriette G, Brunhes A, Jarreau PH. High-frequency oscillatory ventilation in the management of respiratory distress syndrome. Biol Neonate 2000; 77: 14–16. 3 Pillow JJ. High frequency oscillatory ventilation: Mechanisms of gas exchange and lung mechanics. Crit Care Med 2005; 33: S135–S141. 4 McCulloch PR, Forkert PG, Froese AB. Lung volume maintenance prevents lung injury during high frequency oscillatory ventilation in surfactant-deficient rabbits. Am Rev Respir Dis 1988; 137: 1185–1192.
5 Yoder BA, Siler-Khodr T, Winter VT, Coalson JJ. High-frequency oscillatory ventilation: effects on lung function, mechanics, and airway cytokines in the immature baboon model for neonatal chronic lung disease. Am J Respir Crit Care Med 2000; 162: 1867–1876. 6 Cools F, Askie LM, Offringa M, Asselin JM, Calvert SA, Courtney SE et al. Elective high-frequency oscillatory versus conventional ventilation in preterm infants: a systematic review and meta-analysis of individual patients’ data. Lancet 2010; 375: 2082–2091. 7 Courtney SE, Durand DJ, Asselin JM, Hudak ML, Aschner JL, Shoemaker CT et al. High-frequency oscillatory ventilation versus conventional mechanical ventilation for very-low-birth-weight infants. N Eng J Med 2002; 347: 643–652. 8 Johnson AH, Peacock JL, Greenough A, Marlow N, Limb ES, Marston L et al. Highfrequency oscillatory ventilation for the prevention of chronic lung disease of prematurity. N Eng J Med 2002; 347: 633–642. 9 Froese AB, Bryan AC. High frequency ventilation. Am Rev Respir Dis 1987; 135: 1363–1374. 10 Weinmann GG, Mitzner W, Permutt S. Physiological dead space during high-frequency ventilation in dogs. J Appl Physiol Respir Environ Exerc Physiol 1984; 57: 881–887. 11 Boynton BR, Hammond MD, Fredberg JJ, Buckley BG, Villanueva D, Frantz ID3rd. Gas exchange in healthy rabbits during high-frequency oscillatory ventilation. J Appl Physiol 1989; 66: 1343–1351. 12 High Frequency Oscillatory Ventilation (HFOV) in Newborn Services Clinical Guideline http://www.adhb.govt.nz/newborn/guidelines/respiratory/hfov/hfov. htm. Accessed 19 January 2014. 13 Hatcher D, Watanabe H, Ashbury T, Vincent S, Fisher J, Froese A. Mechanical performance of clinically available, neonatal, high-frequency, oscillatory-type ventilators. Crit Care Med 1998; 26: 1081–1088. 14 Ghazanshahi SD, Khoo MCK. Optimal application of high-frequency ventilation in infants: A theoretical study. IEEE Trans Biomed Eng 1993; 40: 788–796. 15 Venegas J, Fredberg JF. Understanding the pressure cost of ventilation: why does high-frequency ventilation work?. Crit Care Med 1994; 22: S49–S57. 16 Rieke H, Hook C, Meyer M. Pulmonary gas exchange during high-frequency ventilation in dogs. Respir Physiol 1983; 54: 1–17. 17 Watson JW, Jackson AC, Gillespie JR. CO2 elimination and airway opening pressure during high frequency oscillation in dogs. Respir Physiol 1984; 58: 235–244. 18 Chan A, Jayasuriya K, Berry L, Roth-Kleiner M, Post M, Belik J. Volutrauma activates the clotting cascade in the newborn but not adult rat. Am J Physiol Lung Cell Mol Physiol 2006; 290: L754–L760. 19 Copland IB, Martinez F, Kavanagh BP, Engelberts D, McKerlie C, Belik J et al. High tidal volume ventilation causes different inflammatory responses in newborn versus adult lung. Am J Respir Crit Care Med 2004; 169: 739–748. 20 Copland IB, Kavanagh BP, Engelberts D, McKerlie C, Belik J, Post M. Early changes in lung gene expression due to high tidal volume. Am J Respir Crit Care Med 2003; 168: 1051–1059. 21 Zimová-Herknerová M, Plavka R. Expired tidal volumes measured by hot-wire anemometer during high-frequency oscillation in preterm infants. Pediatr Pulmonol 2006; 41: 428–433. 22 Dimitriou G, Greenough A, Kavvadia V, Laubscher B, Milner AD. Volume delivery during high frequency oscillation. Arch Dis Child Fetal Neonatal Ed 1998; 78: F148–F150.
Supplementary Information accompanies the paper on the Journal of Perinatology website (http://www.nature.com/jp)
© 2014 Nature America, Inc.
Journal of Perinatology (2014), 464 – 467
ORIGINAL ARTICLE
Bringing back the old: time to reevaluate the high-frequency ventilation strategy A Mukerji1, J Belik1 and M Sanchez-Luna2 OBJECTIVE: To examine the role of frequency in high-frequency ventilation (HFV) on carbon-dioxide (CO2) elimination and lung injury, independent of its effect on tidal volume. STUDY DESIGN: An anatomically representative lung model was attached to a mechanical ventilator capable of providing HFV with a constant volume. CO2 was infused directly into the lung, and a commercially available end-tidal CO2 detector was used to determine CO2 elimination. CO2 elimination and amplitude of pressure transmissions were evaluated using frequencies ranging from 5 to 15 Hz. The pressure–volume index (PVI) was described as the product of the volume and pressures delivered to the lung, a surrogate for lung injury. RESULT: The use of increasing frequencies directly correlated with improved CO2 clearance when keeping the tidal volume fixed, expressed as percent CO2 remaining in the lung at 25 s (66.5 (±1.1)%, 50.5 (±0.1)% and 37.8 (±0.3)% at 5, 10 and 15 Hz, respectively, P o0.05). With a fixed tidal volume, there was a decrease in pressure amplitudes transmitted to the lung with a decline in the PVI (53.9 (±2.7) mmHg ml − 1, 41.1 (±0.9) mmHg ml − 1 and 23.4 (±3.6) mmHg ml − 1, at 5, 10 and 15 Hz, respectively, P o 0.05). CONCLUSION: Frequency has a direct relationship with CO2 elimination when tidal volume is fixed. Using low delivered tidal volumes and high frequencies may allow for improved ventilation efficacy, while minimizing lung injury. Journal of Perinatology (2014) 34, 464–467; doi:10.1038/jp.2014.39; published online 13 March 2014 Keywords: pressure–volume index; ventilation efficacy; CO2 elimination; tidal volume
INTRODUCTION High frequency ventilation (HFV) has been used in neonatal intensive care units for more than two decades, mostly in infants with severe respiratory failure as a rescue mode of ventilation.1,2 It uses low tidal volumes at supra-physiological respiratory frequencies,3 and has been proposed as an effective respiratory strategy that may reduce ventilator-induced lung injury.4,5 Yet in the clinical setting, the superiority of HFV over conventional ventilation in the prevention of chronic lung disease remains questionable.6 The factors accounting for the apparent discrepancy between the theoretical advantage of HFV and available clinical data showing at best a rather small reduction in the chronic lung disease prevalence amongst infants ventilated with this modality7,8 are unclear. HFV is more effective than conventional ventilation for carbondioxide (CO2) elimination.9 Animal and bench studies conducted over 30 years ago showed that the efficacy of CO2 elimination in HFV is related to the tidal volume (VT) and frequency (f), best expressed as V2T × f,10,11 often described as the diffusion coefficient of CO2 (DCO2), a measure of CO2 elimination. As is evident by the aforementioned relationship, an increase in either VT or f would result in improved CO2 elimination. The original data on which the equation was based clearly indicate that the higher the f the greater the CO2 elimination. In spite of the long-standing evidence of the clear benefit of using the highest possible f, the ventilatory strategy's most commonly utilized starting frequency in various studies and clinical guidelines is 10 Hz.7,8,12 One obvious reason for such an approach
relates to the fact that until recently none of the commercially available ventilators allowed for the independent adjustment of VT and f. In these ventilators an increase in f leads to a concomitant reduction in VT, thereby reducing DCO2.13 Furthermore, a number of unproven and mostly theoretical concerns have been raised against the use of frequencies higher than 12 Hz. These concerns mostly relate to the fact that in order to sustain VT at high frequencies an incremental pressure amplitude in the circuit is needed, raising fears that these would be transmitted to the alveoli and result in lung injury. The current strategy of limiting f is geared at minimizing alveolar pressure swings and peak alveolar pressures,14 as well as to reduce the pressure cost of ventilation.15 The advent of a commercially available ventilator able to sustain a fixed VT allows one to overcome the decrease in VT with increasing frequencies, thus enabling an actual increase in DCO2 with higher f. In addition, currently available ventilators measure volume at the distal tubing, providing a much more reliable measure of the VT, as compared with previous ventilator prototypes used in experimental models that depended on a ‘stroke volume’ at the ventilator end.16,17 The goal of the present study was to evaluate CO2 elimination at a fixed VT and varying frequencies. We hypothesized that the proximal pressure amplitude is dampened such that the distal values do not significantly increase with the rise in frequency. As such a study is not feasible in humans, we developed an artificial lung model where CO2 elimination could be measured and amplitude pressures monitored.
1 Division of Neonatology, Department of Pediatrics, Hospital for Sick Children, University of Toronto, Toronto, ON, Canada and 2Division of Neonatology, Instituto de Investigación Sanitaria Gregorio Marañón, Hospital General Universitario “Gregorio Marañón”, University Complutense of Madrid, Madrid, Spain. Correspondence: Dr M Sanchez-Luna, Neonatology Division, Instituto de Investigación Sanitaria Gregorio Marañón, Hospital General Universitario “Gregorio Marañón”, C/ Dr. Esquerdo 46, Madrid E-28007, Spain. E-mail [email protected] Received 21 November 2013; revised 19 January 2014; accepted 5 February 2014; published online 13 March 2014
Reevaluating the high-frequency ventilation strategy A Mukerji et al
465 METHODS The experimental setup consisted of an artificial lung model with a volume of 500 ml, which had a measured compliance of 4.2 ml cm − 1 H2O, and resistance of 8 cm H2O l − 1 s − 1. This lung model was attached to connectors simulating the trachea (measured volume 35 ml, measured diameter 1.8 cm). The trachea was directly attached to flow tubings from a mechanical ventilator capable of providing HFV both with and without a constant VT (VN500, Dräger Medical, Lubeck, Germany). The ventilator designated a constant VT as ‘volume guarantee’ (VG). The ventilator’s flow sensor was placed between the ventilator tubing (RT series, Fisher and Paykel Healthcare, Laval, Canada) and the trachea. A schematic of the experimental setup is shown in the online Supplementary Figure. The amplitude of pressure transmission (pulse pressure) was measured quasi-simultaneously with a commercially available pressure transducer (Model 041-508-001, Cobe Laboratories, CO, USA) and a rigid catheter at three positions denoted as P1, P2 and P3, as depicted in the online Supplementary Figure. P1 was located at the most distal end of the ventilator tubing, P2 referred to the anatomically equivalent trachea and P3 was measured inside the model lung, analogous to the distal lung parenchyma. CO2 was infused directly into the lung continuously from a source with 100% CO2 at a flow rate of 100 ml min − 1. Once the mode of ventilation to be tested was initiated and the CO2 level in the test lung reached a state of equilibrium, the infusion was turned off and the rate of CO2 clearance was measured, as a surrogate for ventilation efficacy. A catheter at P2 connected to an end-tidal CO2 detector (Model CD-102-45-02 PuritanBennett Corporation, Los Angeles, CA, USA) acquired the CO2 levels at P2 continuously. This analog data, along with the pressure measurements, were acquired at a rate of 20 data points per second and stored for further analysis by a custom-made computer software program. The concentration of CO2 remaining in the lung at 25 s after the infusion was turned off, expressed as a percentage of the initial concentration of CO2 recorded, was analyzed as a measure of ventilation efficacy. The mean airway pressure was set at 10 cm H2O for all experiments. When using HFV without a constant VT, the amplitude was set to 30 cm H2O and in HFV with a constant VT, the VT was set at 15 ml, with a maximum allowable pressure amplitude of 80 cm H2O. The frequencies were tested from 5 to 15 Hz (highest frequency provided by the manufacturer). Measurements from each ventilator setting were performed in triplicate to assure precision, and DCO2 values were calculated for each experiment run from the frequency and VT measurements from the ventilator. Continuous positive airway pressure with a positive-end expiratory pressure of 5 cm H2O was used as a control. The I:E (inspiratory:expiratory) time ratio was kept constant at 1:1 for all the experiments. Data are presented as mean ± s.e. and were analyzed by two-way analysis of variance and Tukey–Kramer multiple comparison testing (NCSS, LLC Kaysville, UT, USA). P o0.05 was accepted as statistically significant.
RESULTS The DCO2 and VT values were compared against frequencies of 5 to15 Hz both with and without a constant VT (‘VG’ on and ‘VG’ off, respectively), as shown in Figure 1. When plotted against HFV without a constant VT, the DCO2 values decreased as the frequencies increased, whereas the opposite was seen with a constant VT (Figure 1a). The VT volumes decreased with higher frequencies in regular HFV, and as expected, remained constant when the VT was set at a fixed value (Figure 1b). Interestingly, however, the VT values showed a decline with frequencies ⩾ 14, indicative of the actual maximum frequency for which VT was maintained constant with this device. Ventilation efficacy, determined by CO2 elimination, was compared for the various modes tested. When the lung was connected to the ventilator set on continuous positive airway pressure, no CO2 elimination was detected (data not shown). The CO2 clearance, as determined by the percentage of CO2 remaining in the lung at 25 s, was compared for HFV with and without a constant VT for frequencies from 5 to 15 Hz, as shown in Figure 2. With a constant VT, CO2 elimination improved with increasing frequencies (66.5 (±1.1)%, 50.5 (±0.1)% and 37.8 (±0.3)% at 5, 10 and 15 Hz, respectively, P o0.05), whereas the opposite was seen © 2014 Nature America, Inc.
Figure 1. (a) DCO2 values plotted against frequency with and without a fixed VT. With a fixed VT, DCO2 values increased until 13 Hz showing a drop-off at 14 and 15 Hz. Without a fixed VT, the DCO2 values declined with higher frequencies. (b) Delivered VT values plotted against frequency with and without a fixed VT. As expected with a fixed VT on the ventilator, the delivered VT remained constant, whereas they decreased with increasing frequencies when not fixed. **P o0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. DCO2, diffusion coefficient of CO2; VG off, no constant VT; VG on, constant VT; VT, tidal volume.
for HFV without a constant VT (28.1 (±0.5)%, 43.8 (±0.1)%, and 51.9 (±0.4)% at 5, 10 and 15 H z, respectively, P o 0.05). The pulse pressure delivered to the lung was also evaluated, as measured at P3. There was a decrease in the pressure amplitudes at P3 with increasing frequencies, as depicted in Figure 3, and this was statistically significant, as compared with the data obtained while VT was not kept constant. The pressure transmission to the lung (P3) was combined with VT to describe a pressure–volume index (pressure × VT)—denoted as PVI. The PVI was compared at frequencies of 5 to 15 Hz both with and without a constant VT, as shown in Figure 4. The PVI values were 262.9 (±3.1) mmHg ml − 1, 56.1 (±0.7) mmHg ml − 1 and 13.1 (±1.3) mmHg ml − 1 without a constant VT (P o 0.05), and were 53.9 (±2.7) mmHg ml − 1, 41.1 (±0.9) mmHg ml − 1 and 23.4 (±3.6) mmHg ml − 1 with a fixed VT (P o 0.05), all at 5, 10 and 15 Hz, respectively. DISCUSSION In this study of HFV using a constant VT, we found frequencies directly correlates with CO2 elimination, and at higher frequencies, distal lung equivalent transmission of pulse pressure was negligible. In fact, the transmitted pulse pressures decreased with Journal of Perinatology (2014), 464 – 467
Reevaluating the high-frequency ventilation strategy A Mukerji et al
466
Figure 2. CO2 elimination efficacy, as determined by percentage of CO2 remaining in the lung at 25 s after ventilation initiation, plotted against frequencies with and without fixed VT. CO2 elimination was directly proportional to f when VT was fixed. **Po0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. VG off, no constant VT; VG on, constant VT; VT, tidal volume.
Figure 3. Amplitude of pressure transmission to the lungs plotted against frequency with and without a fixed VT. Despite higher pressure amplitudes at the ventilator with increasing frequency, the distal pressure amplitude showed a decline with a fixed VT. **P o0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. VG off, no constant VT; VG on, constant VT; VT, tidal volume.
increasing frequencies. We also demonstrated a decrease in PVI, a theoretical surrogate for ventilation injury that incorporates barotrauma and volutrauma, across all frequencies with a constant VT. In contrast, without a constant VT, maximal ventilation efficacy was achieved at the lowest frequencies, but this was associated with a significantly higher PVI. This study confirms the previously described DCO2 equation11 indicating that CO2 clearance with constant VT is directly proportional to frequency. However, in this study, we show for the first time that improving ventilation efficacy with higher frequencies in HFV is feasible with newer generation ventilators that reliably deliver a constant VT, and this may be achievable with reduced lung injury. We did, however, note that the ventilator utilized for the present study is unable to sustain VT when frequencies were increased beyond 14 and 15 Hz, under the conditions used in this lung model. Journal of Perinatology (2014), 464 – 467
Figure 4. Pressure–volume index plotted against frequency with and without fixed VT. The PVI dropped sharply without a constant VT until frequency of 11 Hz, then remained relatively constant. On the other hand, with a fixed VT, the PVI remained stable at all frequencies. **P o0.01 as compared with VG on values at same frequency by two-way analysis of variance and Tukey–Kramer multiple comparison testing. PVI, pressure–volume index; VG off, no constant VT; VG on, constant VT; VT, tidal volume.
In their study of HFV on dogs, Watson et al.17 found a similar relationship between frequency and ventilation efficacy as ours. But there was a lack of reliability of the delivered volume, which was being measured at the ventilator end. Moreover, there was concern regarding the use of higher frequencies, as it was presumed that higher frequencies would lead to higher airway and lung pressures to deliver the desired volume and increase the risk of barotrauma.17 Despite many other studies similarly demonstrating a clear relationship between increasing frequencies and improved ventilation,16 the lack of reliability of delivering tidal volumes and the inability to measure the pressure transmission to the airways and lungs in the animal studies led to uncertainty regarding the ideal frequency for HFV. Until recently, most commercially available ventilators did not allow for a constant deliverable VT. Some recent clinical studies of HFV have utilized starting frequencies of 10 Hz, adjusting the amplitude to optimize ventilation efficacy7,8 despite previous recommendations to use higher set frequencies.3,9 Our data suggest that if VT can be held constant and reliably delivered, the highest frequency should be used to maximize ventilation efficacy while minimizing barotrauma, challenging the current practice of using a starting frequency of 10 Hz in HFV. A meta-analysis of these and other studies demonstrate marginal benefit of HFV on ventilator-induced lung injury and chronic lung disease in preterm infants.6 The relationship between tidal volumes, alveolar stretch and lung injury is well described,18–20 and utilizing higher frequencies with lower VT in HFV therefore may result in reduced lung injury. We postulate that the limited impact of HFV in reducing lung morbidity may be related to a failure in exploiting the well-known relationship of ventilator frequency and tidal volume when using this ventilation modality. We here propose a new strategy for HFV ventilation taking into account the aforementioned previous studies and our current findings. We suggest using HFV with a constant VT and identifying the minimum VT that results in adequate ventilation at a frequency of 10 Hz, noting the DCO2 value. We then suggest increasing the frequency as high as allowable while decreasing the VT that maintains the same DCO2. We speculate that newer generation ventilators may optimize delivery of VT at frequencies higher than the currently allowed. Using HFV with a constant VT may also confer other advantages such as better control of ventilation, as the DCO2 will be under direct control of the clinician. In fact, some studies have © 2014 Nature America, Inc.
Reevaluating the high-frequency ventilation strategy A Mukerji et al
467 demonstrated significant variations in actual delivered VT using flow sensors when HFV is used without a constant set VT.21,22 This study does have several limitations. It employs a lung model with fixed mechanical properties, lacking the parenchymal inhomogeneity that characterizes the distinct neonatal lung conditions. Also, given the lack of branching airways in our model, the pressure transmission to the parenchyma may be further attenuated in a human lung. Further, our test lung had a volume representative of a 10 kg infant and the measured compliance and resistance were not representative of typical neonatal values. The ventilator employed in this study was designed for neonates, which may explain the limitation of delivering the set VT at frequencies of 14 and 15 Hz. Further testing utilizing more advanced lung models with a wide range of lung mechanics and volumes, along with branching airways with progressively smaller diameters, is warranted. We suggest a critical reevaluation of the current clinical practice of HFV use in neonatal intensive care units and propose the use of the highest possible frequency that achieves adequate ventilation in all infants to minimize lung injury. The clinical benefit of utilizing such a strategy is unknown and requires further clinical investigation. Such studies, however, are difficult to design given the challenge of measuring the impact of changes in f, VT and DCO2 on CO2 elimination and ventilation-induced lung injury in vivo. Yet the present data warrant future clinical studies targeted at evaluating the potential beneficial impact of the proposed ventilatory strategy on the reduction of severity and/or prevalence of bronchopulmonary dysplasia in preterm infants. CONFLICT OF INTEREST MSL declares having received advisory board consulting fees from Drager. The remaining authors declare no conflict of interest.
ACKNOWLEDGEMENTS This study received no specific funding.
REFERENCES 1 Ventre KM, Arnold JH. High frequency oscillatory ventilation in acute respiratory failure. Paediatr Respir Rev 2004; 5: 323–332. 2 Moriette G, Brunhes A, Jarreau PH. High-frequency oscillatory ventilation in the management of respiratory distress syndrome. Biol Neonate 2000; 77: 14–16. 3 Pillow JJ. High frequency oscillatory ventilation: Mechanisms of gas exchange and lung mechanics. Crit Care Med 2005; 33: S135–S141. 4 McCulloch PR, Forkert PG, Froese AB. Lung volume maintenance prevents lung injury during high frequency oscillatory ventilation in surfactant-deficient rabbits. Am Rev Respir Dis 1988; 137: 1185–1192.
5 Yoder BA, Siler-Khodr T, Winter VT, Coalson JJ. High-frequency oscillatory ventilation: effects on lung function, mechanics, and airway cytokines in the immature baboon model for neonatal chronic lung disease. Am J Respir Crit Care Med 2000; 162: 1867–1876. 6 Cools F, Askie LM, Offringa M, Asselin JM, Calvert SA, Courtney SE et al. Elective high-frequency oscillatory versus conventional ventilation in preterm infants: a systematic review and meta-analysis of individual patients’ data. Lancet 2010; 375: 2082–2091. 7 Courtney SE, Durand DJ, Asselin JM, Hudak ML, Aschner JL, Shoemaker CT et al. High-frequency oscillatory ventilation versus conventional mechanical ventilation for very-low-birth-weight infants. N Eng J Med 2002; 347: 643–652. 8 Johnson AH, Peacock JL, Greenough A, Marlow N, Limb ES, Marston L et al. Highfrequency oscillatory ventilation for the prevention of chronic lung disease of prematurity. N Eng J Med 2002; 347: 633–642. 9 Froese AB, Bryan AC. High frequency ventilation. Am Rev Respir Dis 1987; 135: 1363–1374. 10 Weinmann GG, Mitzner W, Permutt S. Physiological dead space during high-frequency ventilation in dogs. J Appl Physiol Respir Environ Exerc Physiol 1984; 57: 881–887. 11 Boynton BR, Hammond MD, Fredberg JJ, Buckley BG, Villanueva D, Frantz ID3rd. Gas exchange in healthy rabbits during high-frequency oscillatory ventilation. J Appl Physiol 1989; 66: 1343–1351. 12 High Frequency Oscillatory Ventilation (HFOV) in Newborn Services Clinical Guideline http://www.adhb.govt.nz/newborn/guidelines/respiratory/hfov/hfov. htm. Accessed 19 January 2014. 13 Hatcher D, Watanabe H, Ashbury T, Vincent S, Fisher J, Froese A. Mechanical performance of clinically available, neonatal, high-frequency, oscillatory-type ventilators. Crit Care Med 1998; 26: 1081–1088. 14 Ghazanshahi SD, Khoo MCK. Optimal application of high-frequency ventilation in infants: A theoretical study. IEEE Trans Biomed Eng 1993; 40: 788–796. 15 Venegas J, Fredberg JF. Understanding the pressure cost of ventilation: why does high-frequency ventilation work?. Crit Care Med 1994; 22: S49–S57. 16 Rieke H, Hook C, Meyer M. Pulmonary gas exchange during high-frequency ventilation in dogs. Respir Physiol 1983; 54: 1–17. 17 Watson JW, Jackson AC, Gillespie JR. CO2 elimination and airway opening pressure during high frequency oscillation in dogs. Respir Physiol 1984; 58: 235–244. 18 Chan A, Jayasuriya K, Berry L, Roth-Kleiner M, Post M, Belik J. Volutrauma activates the clotting cascade in the newborn but not adult rat. Am J Physiol Lung Cell Mol Physiol 2006; 290: L754–L760. 19 Copland IB, Martinez F, Kavanagh BP, Engelberts D, McKerlie C, Belik J et al. High tidal volume ventilation causes different inflammatory responses in newborn versus adult lung. Am J Respir Crit Care Med 2004; 169: 739–748. 20 Copland IB, Kavanagh BP, Engelberts D, McKerlie C, Belik J, Post M. Early changes in lung gene expression due to high tidal volume. Am J Respir Crit Care Med 2003; 168: 1051–1059. 21 Zimová-Herknerová M, Plavka R. Expired tidal volumes measured by hot-wire anemometer during high-frequency oscillation in preterm infants. Pediatr Pulmonol 2006; 41: 428–433. 22 Dimitriou G, Greenough A, Kavvadia V, Laubscher B, Milner AD. Volume delivery during high frequency oscillation. Arch Dis Child Fetal Neonatal Ed 1998; 78: F148–F150.
Supplementary Information accompanies the paper on the Journal of Perinatology website (http://www.nature.com/jp)
© 2014 Nature America, Inc.
Journal of Perinatology (2014), 464 – 467
