Aspiration of Airway Dead Space A New Method to Enhance CO2 Elimination EDOARDO DE ROBERTIS, SIGURDUR E. SIGURDSSON, BJÖRN DREFELDT, and BJÖRN JONSON Departments of Clinical Physiology and Anaesthesia and Intensive Care, University Hospital of Lund, Lund, Sweden; and Department of Anaesthesia and Intensive Care, University “Federico II”, Naples, Italy
Alveolar ventilation and CO2 elimination during mechanical ventilation can be enhanced by reducing dead-space ventilation. Aspiration of gas from the dead space (ASPIDS) is a new principle, according to which gas rich in CO2 during late expiration is aspirated through a channel ending at the distal end of the tracheal tube. Simultaneously, fresh gas injected into the inspiratory line fills the airway down to the same site. We hypothesized that ASPIDS would allow a reduction of tidal volume (VT) and airway pressure (Paw). To test our hypothesis we studied six anaesthetized and mechanically ventilated pigs (24 6 4 kg). The intention was to decrease VT while keeping PaCO2 constant by using . . ASPIDS. VT was reduced by decreasing the minute ventilation (VE) in two steps, of 1.8 L/min (VE 2 . 1.8) and 2.2 L/min (VE 2 2.2), respectively, and by increasing respiratory rate (RR) from 20 to 46 . . breaths/min. At ASPIDS, peak Paw was reduced by 35% at VE 2 1.8 and at VE 2 2.2 (p , 0.001), and by 20% at an RR of 46 (p , 0.01). PaCO2 was maintained or reduced at ASPIDS. No intrinsic positive end-expiratory pressure developed. Arterial blood pressure and heart rate were unaffected. The results show that ASPIDS allows a reduction in VT and Paw while PaCO2 is kept constant. ASPIDS does not lead to problems associated with jet streams of gas or with gas humidification, and can be developed as a safe technique. De Robertis E, Sigurdsson SE, Drefeldt B, Jonson B. Aspiration of airway dead space: a new method to enhance CO2 elimination. AM J RESPIR CRIT CARE MED 1999;159:728–732.
Inadequate alveolar ventilation is a common problem in critical lung disease. Efforts to avoid CO2 retention during mechanical ventilation may lead to high tidal volumes and airway pressures, which may cause additional lung damage. Methods directed at reducing tidal volume (VT) and airway pressure (Paw) are extracorporeal CO2 removal, high-frequency ventilation, partial liquid ventilation, and permissive hypercapnia. So far, none of these methods has proved to be of clinical benefit. Dead-space ventilation may be reduced by expiratory flushing of airways (1) or tracheal gas insufflation (2–4). By increasing alveolar ventilation, these methods may increase CO2 clearance. However, in patients with an expiratory flow continuing until the end of expiration, the positive effect of these methods is limited, since the insufflated gas will be mixed with CO2 in alveolar gas. Another undesired effect is that tracheal gas insufflation during expiration increases Paw in a way that is difficult to control and which may induce or amplify dynamic hyperinflation. Such an effect could also counteract the goal of reducing high airway pressures, which may lead to barotrauma and hemodynamic compromise (5). Moreover, (Received in original form December 30, 1997 and in revised form August 17, 1998) Supported by the Swedish Institute, grant 02872 from the Swedish Medical Research Council, the Swedish Heart Lung Foundation, and the Medical Faculty of Lund, Sweden. Correspondence and requests for reprints should be addressed to Björn Jonson, Department of Clinical Physiology, Lund University Hospital, S-221 85 Lund, Sweden. E-mail:
[email protected] Am J Respir Crit Care Med Vol 159. pp 728–732, 1999 Internet address: www.atsjournals.org
drying of airway secretions and damage to the airway mucosa are problems associated with tracheal gas insufflation (6). An alternative to tracheal gas insufflation would be to aspirate dead-space gas from the trachea (ASPIDS) and simultaneously replace it with new gas through the ordinary inspiratory tubing. This would permit gas in the ventilator tubing, Y-piece, filter, and tracheal tube to be aspirated from the tip of the tracheal tube during the late part of expiration and be replaced with fresh gas. The resulting reduction in volume of airway dead-space gas that returns to the alveoli during inspiration will increase alveolar ventilation and thereby allow the use of a smaller VT and a lower Paw. The aim of this study was to present a system for ASPIDS, to investigate its technical feasibility, and to evaluate the extent to which VT could be reduced in an animal model involving healthy pigs.
METHODS The study was done with six domestic pigs (Swedish land race) weighing 23.9 6 1.9 kg. Permission for the study was given by the Ethics Board of Animal Research of the University of Lund. Animals were fasted overnight but allowed free access to water. At 30 min before induction of anesthesia, the pigs were premedicated with azaperon (Stresnil; Janssen, Beerse, Belgium) 6 mg/kg intramuscularly. Anesthesia was induced with sodium pentothal (Pentothal; Abbott, North Chicago, IL) 12.5 mg/kg intravenously and was maintained by the continuous infusion of fentanyl (Leptanal; Janssen) 75 mg/kg/h, pancuronium bromide (Pavulon; Organon Teknika, Boxtel, The Netherlands) 0.4 mg/kg/h, and midazolam (Dormicum; Hoffmann-La Roche, Basel, Switzerland) 0.25 mg/kg/h. A catheter was inserted in the left carotid artery for blood sampling and monitoring of mean arterial blood pressure ( Pa) and heart rate (HR). Body temperature was kept constant by covering the ani-
De Robertis, Sigurdsson, Drefeldt, et al.: Aspiration of Airway Dead Space
729
Figure 1. The ASPIDS system. A pump and a damping reservoir serve as a vacuum source. From the moment during expiration when the solenoid valve opens, gas is aspirated from the tracheal tube. Simultaneously, gas from the oxygen mixer is injected into the inspiratory line at a flow rate slightly higher than the flow for aspiration. Thereby, during the later part of expiration, new gas will flush the line from the Y-piece down to the distal end of the tracheal tube. A flap valve prevents an accidental negative pressure in the circuit. For the present study, the flow of gas from the expiratory port of the ventilator and the flow · · from the aspiration pump, V ASPIDS, are blended to yield a total expired flow, V EXPtot, which is fed to the sys· tem for measurement of V CO2.
mal and by heating the operating table as required. Animals were hydrated with Ringer’s glucose at 5 ml/kg/h. A “Hi-lo jet” endotracheal tube (NCC; Mallinckrodt) with an I.D. of 7 mm was introduced orally. In addition to the ordinary lumen, this cuffed tube has two extra channels that open at 10 mm and 60 mm from the distal end of the tube. The tube cuff was inflated and frequently tested to avoid air leakage. A moisture exchanger, a bacterial/ viral filter (Light-S Filter; Humid-Vent, Gibeck, Sweden), and a connector were used. The total volume of the filter, the connector, and the tracheal tube was 92 ml. Volume-controlled ventilation was provided with a ServoVentilator 900 C (Siemens-Elema AB, Sweden) in a square inspiratory flow pattern at a respiratory rate (RR) of 20 breaths/min, an inspiratory time of 25% of the respiratory cycle, and a postinspiratory pause of 5% of the cycle. Expiratory CO2 concentration was measured with a· model 930 CO2 analyzer (Siemens-Elema AB). Minute ventilation ( VE) was adjusted to give an arterial carbon dioxide concentration (PaCO2) of 5 to 5.5 kPa (37 to 41 mm Hg). The positive end expiratory pressure was 4 cm H2O and the inspiratory oxygen fraction (FIO2) was 0.21. Signals from the ventilator representing Paw in the expiratory line and the inspiratory and expiratory flow were fed together with the CO2 signal to an IBM-compatible personal computer and converted to digital format at 50 samples per second. Intrinsic positive end expiratory pressure (PEEPi) was calculated as the difference between Paw measured at the end of expiration and after an end-expiratory pause of 3 s.
The ASPIDS System The ASPIDS system comprises the Servo Ventilator 900C, an electronic control unit, and two solenoid valves that connect the airway to a vacuum source and to a source for replacement of the aspirated dead-space gas (Figure 1). Through use of the control unit, the operator sets the moments during expiration at which the ASPIDS valves should open and close. When the valves open, gas is aspirated from the auxiliary port inside the tracheal tube at 60 mm from its tip. The vacuum source consists of a membrane pump (MP-2; Alitea, Sweden) with a regulated power supply to control the subatmospheric pressure,
and a 3-L damping reservoir. By modifying the duration of the period of aspiration and the subatmospheric pressure, the operator controls the volume of gas aspirated per breath. Simultaneously with gas aspiration, fresh gas is injected into the inspiratory line. This gas is tapped from a second outlet of the gas mixer, which controls the oxygen fraction of inspired gas; the gas passes a pressure-regulating valve that allows adjustment of the volume of injected gas during the period when the valve is open. The gas is injected into the inspiratory line upstream of an optional humidifier that was not used in the present study. A flap valve in the inspiratory line serves as a safety measure against accidental development of a negative pressure in the circuit. The baseline RR was 20 breaths/min. At this frequency the ASPIDS pulse started at 0.51 s after onset of expiration and lasted 0.9 s (Figure 2). The volumes of gas injected and aspirated were measured by first activating the injection valve and reading the increase in expired VT on the digital ventilator display. The aspiration system was then activated and the change in VT was read again. During each ASPIDS pulse, about 140 ml of gas was aspirated, while 160 ml of gas was injected. This implies that gas was slightly oversupplied. Because the expiratory valve is open, this does not affect patient ventilation. At the higher RR studied (46 breaths/min), the timing of the ASPIDS pulse was adjusted to cover the same relative period during expiration. The gas leaving the expiratory port of the ServoVentilator and the gas from the suction pump were fed to a laboratory flow meter (L5PVC; H. Wohlgroth & Co., Zürich, Switzerland) via damping and mixing bags. With this system, the meter measured the total expired · VE. The fraction of CO2 in the mixed expired gas was measured with a blood gas analyzer (ABL 505; Radiometer, ·Copenhagen, Denmark). V The volume of CO2 eliminated per · minute ( CO2), was determined by multiplying the total expiratory VE by the expired fraction of CO2. As a first approximation, we estimated that ASPIDS would fully clear the volume of the connecting tubes of CO2 (i.e., 92 ml per breath). In theory, · at 20 breaths/min this would lead to a reduced requirement for VE of 20 · 92 ml or 1.8 L/min. However, in preliminary tests we showed that ASPIDS cleared · an additional volume of 20 ml per breath. This would imply that VE during ASPIDS might be reduced by 2.2 L/min without compromising CO2 elimination.
730
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
VOL 159
1999
RESULTS Pa, HR, and body temperature remained stable under all conditions (Table 1). The results of baseline studies before and after each period of ASPIDS showed no significant differences. Therefore, the data obtained during ASPIDS were compared only with those from the baseline study preceding the ASPIDS period. Figure 2 shows typical pressure and flow patterns at baseline ventilation and with ASPIDS. In pressure tracings, trivial oscillations were observed at the moments when ASPIDS was switched on and off (Figure 2). The set PEEP level was maintained. During the ASPIDS pulse a slight increase in expiratory flow was observed. This increase was caused by the surplus of gas injected over the gas aspirated. No PEEPi developed at ASPIDS, not even at an RR of 46 breaths/min (Table 2). At each of the ASPIDS settings, the reduction in VT led to a substantial decrease in peak airway pressure (Pawpeak) and postinspiratory· airway plateau pressure (Pawplat) (Table 2). At indicating hypervenASPIDS with VE 2 1.8, PaCO2 decreased, · tilation. A concomitant increase in VCO2 may indicate that the washout of CO2 · to a new stable level was not completely achieved (7). At VE 2 2.2, nonsignificant changes in PaCO2 and · VCO2 indicated that an isocapnic condition was maintained. When RR was increased to 46 breaths/min, ASPIDS induced a decrease in PaCO2, indicating slight hyperventilation.
Figure 2. Tracings of Paw and flow rate from a representative ani· mal at 20 breaths/min. Curves representing ASPIDS at a V E reduced by 2.2 L/min are superimposed over those for baseline ventilation. The former curves are interrupted. The dotted lines indicate when, at ASPIDS, the aspiration and injection valves open (on) and close (off).
DISCUSSION Tracheal gas insufflation may enhance CO2 elimination in different ways, depending on when the gas is insufflated during the breathing cycle (4, 8, 9). Tracheal gas insufflation during expiration will lead to washout of CO2 contained in the upper airway dead space. ASPIDS may be regarded as a development of the latter principle, with the aim of reducing or eliminating problems associated with tracheal gas insufflation. In accordance with our hypothesis, ASPIDS allowed a substantial reduction in VT at preserved isocapnic conditions. Accordingly, airway pressures were reduced. · · The results at VE 2 1.8 and VE 2 2.2 verified findings in pilot studies that ASPIDS would clear CO2 from a space about 20 ml larger than the dead space measured from the Y-piece to the tip of the tracheal tube of the ventilator circuit used in our study. It is known that flow causes turbulence in the Y-piece and adjacent tubes. A mixing of inspired and expired gas will occur in the tubes. This phenomenon contributes a volume of about 24 ml to the dead space at an RR of 10 breaths/min (11). During the ASPIDS pulse the gas injected into the inspiratory line will clear the inspiratory tubing of the amount of CO2 that was mixed into the inspiratory line through turbulence during the preceding expiration. The surplus of the injected gas over the aspirated gas will simultaneously push the CO2-rich ex-
Protocol Our primary hypothesis was that a constant PaCO2 could be maintained while VT was reduced by a volume equal to the dead-space volume cleared by ASPIDS.· At a frequency of 20 breaths/min we re· · duced VE by 1.8 L/min (VE 2 1.8) and by 2.2 L/min (VE 2 2.2). A secondary issue tested was that a further benefit · of ASPIDS might be obtained by increasing the RR at a constant VE. Such settings lead to smaller tidal volumes and lower pressures, but also to increased deadspace ventilation. We hypothesized that the latter increase could be balanced by the use of ASPIDS. It was estimated that 46 breaths/min was the highest RR that would be compatible with an unaltered CO2 elimination given the total airway dead-space volume in pigs and the fraction thereof that might be cleared with ASPIDS. The reduction in · VE by 1.8 and 2.2 L, and the increase in RR to 46 breaths/min, were studied in random order. Before and after each period of application of ASPIDS, a measurement was made under baseline conditions. To reach a steady state, 20 min were allowed to pass after changing the mode of ventilation before data collection was begun (7).
Statistical Analysis All data are expressed as mean 6 SD. The two-tailed Student’s t test was used to compare findings from different study periods. Differences were considered significant at p , 0.01.
TABLE 1 HEMODYNAMIC VARIABLES · Constant V E
Constant RR · V E 2 1.8 L/min
HR, beats/min Pa, mm Hg T, 8 C
· V E 2 2.2 L/min
RR 5 46 Breaths/min
Baseline
ASPIDS
Baseline
ASPIDS
Baseline
ASPIDS
75 6 9 74 6 9 37.8 6 0.5
75 6 8 74 6 8 37.8 6 0.6
68 6 10 76 6 7 37.7 6 0.4
67 6 10 77 6 10 37.7 6 0.4
71 6 10 75 6 11 37.9 6 0.3
78 6 14 76 6 14 37.8 6 0.3
· Definition of abbreviations: HR 5 heart rate; Pa 5 mean arterial blood pressure; RR 5 respiratory rate; T 5 body temperature; V E 5 minute ventilation. No significant changes were observed between baseline and ASPIDS periods.
731
De Robertis, Sigurdsson, Drefeldt, et al.: Aspiration of Airway Dead Space TABLE 2 VENTILATORY AND GAS EXCHANGE VARIABLES · Constant V E
Constant RR · V E 2 1.8 L/min
· V E, L/min RR, breaths/min VT, ml Pawpeak, cm H2O Pawplat, cm H2O PEEPi, cm H2O · V CO2, ml/min PaCO2, kPa PaO2, kPa pH
· V E 2 2.2 L/min
Baseline
ASPIDS
6.0 6 0.2 20 299 6 7 19.7 6 3.6 12.4 6 3.0 0.58 6 0.3 113 6 14 5.4 6 0.4 12.3 6 0.9 7.48 6 0.02
4.2 6 0.2 20 208 6 5† 12.8 6 1.1* 9.1 6 1.9† 0.52 6 0.3 124 6 14* 4.9 6 0.4† 12.7 6 0.8 7.52 6 0.02† †
RR 5 46 Breaths/min
Baseline
ASPIDS
6.0 6 0.2 20 299 6 9 19.4 6 3.2 13.5 6 1.0 0.55 6 0.3 115 6 15 5.5 6 0.4 12.0 6 0.7 7.47 6 0.02
3.8 6 0.2 20 191 6 9† 12.7 6 1.6† 9.4 6 0.6† 0.50 6 0.4 121 6 16 5.4 6 0.4 12.0 6 0.6 7.48 6 0.03 †
Baseline
ASPIDS
6.1 6 0.2 20 303 6 11 20.1 6 2.9 14.5 6 1.4 0.69 6 0.23 116 6 12 5.6 6 0.4 11.7 6 1 7.47 6 0.03
6.2 6 0.1 46.3 6 2.3† 133 6 8† 16 6 3.8* 8.6 6 1.2† 0.22 6 0.1 127 6 13 5.0 6 0.6* 13.2 6 2.2 7.52 6 0.04
Definition of abbreviations: Pawpeak 5 peak airway pressure; Pawplat 5 plateau airway pressure; PEEPi 5 intrinsic positive end-expiratory · · pressure; RR 5 respiratory rate; V CO2 5 CO2 elimination; V E 5 minute ventilation; VT 5 tidal volume. * p , 0.01; † p , 0.001 for significance of difference between baseline and ASPIDS.
pired gas down into the expiratory line. This prevents turbulence during the following inspiration from leading to entrapment in the inspired gas of CO2 from the expired gas. The finding that ASPIDS has an effect beyond clearing the space from the Y-piece to the end of the tracheal tube is accordingly explained, in part by the surplus of injected gas volume over aspirated gas volume. When RR was adjusted to 46 breaths/min without changing · VE, Paw, and particularly the plateau pressure, was efficiently reduced. A strategy that remains to be explored is to combine · a decrease in VE with an increase in RR. In the present study, PEEPi was nearly zero at an RR of 46. This implies that expiratory flow had ceased before the succeeding inspiration. PEEPi is associated with an expiratory flow that continues until the end of expiration. In order to remain efficient, the ASPIDS system must then clear both the volumes of the upper dead space and the volume of gas coming from deeper airways during the ASPIDS pulse. This pulse should also continue until the very end of expiration. The ASPIDS system has a capacity to produce pulses of aspiration and injection with flow rates from two to three times greater than that used in this study. The pulses can be set to begin and end at any time during expiration. In theory, ASPIDS would function even in the presence of PEEPi. However, this remains to be tested. In ASPIDS, the extra volume of gas delivered has the same composition as the gas used for basal ventilation. It may be humidified by an ordinary humidifier in the inspiratory line. It does not cause any jet effects in the airway. As expected, the system did not interfere to any extent with inspiration from the ventilator, and did not interfere significantly with the expiratory pressure. During ASPIDS, the built-in monitoring and alarm systems of the ventilator are functioning as they do during basal ventilation, except that the surplus of injected gas over aspirated gas will appear in the measurements of expired gas. The present system should be regarded as an experimental system to be used only with constant surveillance by a wellinformed operator. It is not suitable for assisted ventilation. Further safety aspects should be considered for routine clinical use of ASPIDS. Should the gas-injection side of the ASPIDS system be accidentally blocked, the flap valve opens, which eliminates the risk for negative airway pressure. Air will then enter the inspiratory line and reduce the FIO2 of the inspired gas. To eliminate this risk, a spacer that is continuously flushed with new respiratory gas may be placed between the
safety valve and the room air. Systems providing alarm and automatic stopping of the ASPIDS system, on the basis of flow sensors in the aspiration and injection lines, may also be warranted. A risk with a system for providing ASPIDS is that a subatmospheric intrapulmonary pressure may develop if the tracheal tube is blocked above the port for aspiration. This would hinder gas from entering the lung but leave the aspiration port open. If a grave subatmospheric airway pressure is then to be avoided, the ASPIDS system must immediately be brought to a halt. This can be achieved with systems using the combined information contained in the signals from the inspiratory flow and Paw sensors. In addition, one may continuously measure the intratracheal pressure. The tracheal tube that we use has a further channel that ends at its tip, distal to the port that is used for aspiration. This channel can be connected to a pressure transducer for detection of a negative tracheal pressure and automatic interruption of ASPIDS. We have shown that ASPIDS is technically feasible and allows an important decrease in VT and airway pressures. It does not impede expiration. The injected gas passes through the normal inspiratory line and a humidifier. These advantages over known systems for tracheal gas injection merit tests of ASPIDS in patients who are particularly difficult to ventilate because of critical lung disease. Benefits may be achieved in terms of lower tidal volumes and peak pressures, or in terms of improved CO2 elimination. In the respiratory distress syndrome, one may wish to increase PEEP without drastically increasing peak pressures or PaCO2. An improved efficiency of ventilation provided by ASPIDS may make this possible. Additionally, with supported ventilation, ASPIDS, by reducing dead-space ventilation, may reduce the need for total ventilation and thereby the work of breathing. Acknowledgment : The authors are grateful to Gerth-Inge Jönsson for technical assistance. Johan Thörne, Sten Blomquist, and Peter L. Dahm gave valuable help.
References 1. Jonson, B., T. Similowski, P. Levy, N. Viires, and R. Pariente. 1990. Expiratory flushing of airways: a method to reduce deadspace ventilation. Eur. Respir. J. 3:1202–1205. 2. Nahum, A., W. C. Burke, S. A. Ravenscraft, T. W. Marcy, A. B. Adams, P. S. Crooke, and J. J. Marini. 1992. Lung mechanics and gas exchange during pressure controlled ventilation in dogs: augmentation of CO 2 elimination by an intratracheal catheter. Am. Rev. Respir. Dis. 146:
732
AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
965–973. 3. Nahum, A., S. A. Ravenscraft, G. Nakos, A. B. Adams, W. C. Burke, and J. J. Marini. 1993. Effect of catheter flow direction on CO 2 removal during tracheal gas insufflation in dogs. J. Appl. Physiol. 75: 1238–1246. 4. Ravenscraft, S. A., W. C. Burke, A. Nahum, A. B. Adams, G. Nakos, T. W. Marcy, and J. J. Marini. 1993. Tracheal gas insufflation augments CO2 clearance during mechanical ventilation. Am. Rev. Respir. Dis. 148:345–351. 5. Nakos, G., S. Zakinthinos, A. Kotanidou, H. Tsagaris, and C. Roussos. 1994. Tracheal gas insufflation reduces the tidal volume while PaCO2 is maintained constant. Int. Care Med. 20:407–413. 6. Marini, J. J. 1994. Tracheal gas insufflation—a useful adjunct to ventilation? Thorax 49:735–737. 7. Taskar, V., J. John, A. Larsson, T. Wetterberg, and B. Jonson. 1995. Dynamics of carbon dioxide elimination following ventilatory resetting.
VOL 159
1999
Chest 108:196–202. 8. Burke, W. C., A. Nahum, S. A. Ravenscraft, G. Nakos, A. B. Adams, T. W. Marcy, and J. J. Marini. 1993. Modes of tracheal gas insufflation: comparison of continuous and phase-specific gas injection in normal dogs. Am. Rev. Respir. Dis. 148:562–568. 9. Ravenscraft, S. A., R. S. Shapiro, A. Nahum, W. C. Burke, A. B. Adams, G. Nakos, and J. J. Marini. 1996. Tracheal gas insufflation: catheter effectiveness determined by expiratory flush volume. Am. J. Respir. Crit. Care Med. 153:1817–1824. 10. Marini, J. J. 1996. Adjunctive ventilation with tracheal gas insufflation— good vibrations? Crit. Care Med. 24:375–377. 11. Fletcher, R., O. Werner, L. Nordström, and B. Jonson. 1983. Sources of error and their correlation in the measurement of carbon dioxide elimination using the Siemens-Elema CO2 analyser. Br. J. Anaesth. 55:177–185.
