British Journal of Anaesthesia 1991; 67: 795-800
PRESSURE SUPPORT VENTILATION USING A NEW TRACHEAL GAS INJECTION TUBE L. BEYDON, D. ISABEY, G. BOUSSIGNAC, F. BONNET, P. DUVALDESTIN AND A. HARF
to 7.1 (1.7) litre min~') (P < 0.001). It reduced the total work of breathing (from 0.625 (0.223) to 0.263 (0.151) J litre-', respectively) (P < 0.01) and the occlusion pressure (from 2.62 (1.28) to 1.36 (0.74) cm H2O, respectively) (P < 0.01). It is concluded that this TGIT used with a specific system for sensing and triggering ventilation allows inspiratory pressure support during low frequency jet ventilation.
KEY WORDS Equipment: trachea/ gas injection tube. Ventilation: jet ventilation.
Mechanical ventilation is often required in the period after major surgery. Residual effects of anaesthesia [1], abdominal pain [2] and diaphragmatic dysfunction [3] may contribute to depressed ventilation after operation and impair spon-
PATIENTS AND METHODS
Patients and anaesthesia We studied 10 patients (five males) (mean age 61 (SD 11) yr; weight 65 (6) kg; height 167 (7) cm; non-smokers; ASA I or II) in the recovery room after abdominal surgery. Surgery was performed under general anaesthesia with fentanyl 3 ug kg"1, flunitrazepam 0.03 mg kg"1 and vecuronium 0.1 LAURENT BEYDON*, M.D.; DANIEL ISABEY, PH.D.; GEORGES BOUSSIGNAC, M.D.; FRANCOIS BONNET, M.D.; PHILIPPE DUVALDESTIN, M.D.; ALAIN HARP, M.D.; Unite INSERM
U296 and the Departement d'Anesthisie, H6pital Henri Mondor, 51 Avenue du Marechal de Lartre de Tassigny, 94010 Criteil, France. Accepted for Publication: May 9, 1991. •Address for correspondence: Department d'Anesthesie, Hdpital Henri Mondor, 94010 Creteil, France.
Downloaded from http://bja.oxfordjournals.org/ at University of Sydney on March 13, 2015
taneous breathing. To overcome these problems, conventional artificial ventilation may be used In order to explore new types of jet ventilation, after operation, followed by an assisted mode such we tested a trachea! gas injection tube (TGIT) as inspiratory pressure support (IPS), which is a which included six thin capillaries and provided pressure limited mode. IPS is provided by high pressure injection. The driving pressure was ventilators, using sophisticated servo control chosen to yield a plateau of inspiratory trachea/ techniques. We described a new tracheal gas pressure of 10 cm H2O. An original controller injection tube (TGIT) which was able to produce high velocity jets at its distal tip and consequently was built to monitor spirometry and trigger to generate positive pressure by air momentum injection in order to deliver both pressure contransfer down to the injection site [4, 5]. This trolled ventilation (PCVTGIT) and a new mode of system was used successfully for apnoeic veninspiratory pressure support jet ventilation tilation during cardiopulmonary resuscitation in (IPSJQU). The PVCTGn mode maintained the same animals [6], and man [7] and to prevent oxygen end-tidal carbon dioxide concentration as con- desaturation during apnoeic periods in patients ventional ventilation with the same tidal and undergoing mechanical ventilation [8]. We have minute ventilation. We studied 10 patients after also developed a new electronic system which may abdominal surgery. During spontaneous breath- be coupled to the TGIT to produce either ing, the patients were allowed to breathe through pressure controlled ventilation (PCV TGIT) or inthe tube, successively with and without IPSTGn. spiratory pressure support (IPSTCrT). The aim of IPSTGIT' compared with spontaneous breathing this study was to examine high flow jet ventilation increased minute ventilation (from 5.7 (SD 1.6) in both controlled and spontaneous ventilation. SUMMARY
BRITISH JOURNAL OF ANAESTHESIA
796
Compressed gas source
t Entrained air • Proximal end
Capillaries.
Site of injection
FIG. 2. The tracheal gas injection tube (TGIT).
C7
Downloaded from http://bja.oxfordjournals.org/ at University of Sydney on March 13, 2015
A TGIT (fig. 2) (Porges laboratories, Le Plessis Robinson, France) of 8 mm i.d. with a low pressure cuff. In its wall, six thin capillaries (diameter: 720 urn) are moulded by extrusion. These are distributed evenly around the circumference and emerge on the internal aspect of the TGIT, 1 cm before its distal tip. These capillaries are connected to an adjustable high pressure air-oxygen source (driving pressure 0—3 bar) to produce high velocity gas injection at the TGIT distal end, as described previously [8]. The performances of the TGIT were tested in our laboratory before the clinical study. The relationship between the driving pressure and the FIG. 1. System used with the TGIT. A pneumotachograph flow of the injectors was found to be constant for measured the flow (V) passing through the lumen of the all TGIT. The pressure generated by the TGIT TGIT. The flow signal was adapted by an analog system in the trachea (measured at the carinal level) (Processor) to trigger injection via electronic valves (1, 2). The ranged between 0 and 27 cm H2O according to the system also included unidirectional valves on both branches of driving pressure used (0-3.5 bar). For instance, a the Y-piece (3), a pneumatic valve with solenoid for occlusion of 1.2 bar caused a flow of pressure measurement (4), a rubber reservoir bag (5), a driving pressure 1 through the injectors and this humidifier (6) and an air-oxygen mixer and pressure regulator 350 ml s" (7). Tracheal pressure (Paw) was measured by a catheter inside generated a pressure of 10 cm H2O in the trachea. the trachea and oesophagcal pressure (Poe) via a balloon An inspiratory line with an air-oxygen mixer, a tipped catheter. pressure regulator, a cascade humidifier (FisherPaykell, MR 428, New Zealand), a 3-litre rubber mg kg 1 for induction and tracheal intubation. reservoir which served as a mixing chamber to Anaesthesia was maintained with 70% nitrous allow a constant Fio,> a n d a unidirectional valve. oxide in oxygen and administration of increments From the air-oxygen mixer a line supplied high of the same drugs. Patients' lungs were ventilated pressure gas to the capillaries of the TGIT via a perioperatively by a volume cycled ventilator solenoid valve (opening—closing delay: 11 ms). An (Draeger, SAX) in a conventional mode. occluding pneumatic valve was used to measure the occlusion pressure 0.1 s after the beginning of Material and signal processing inspiration (P0.i) [9]. This valve comprised a latex The system we used comprised the following balloon mounted inside a short, 22-mm diameter tube placed in the inspiratory line. This balloon components (fig. 1):
VENTILATORY SUPPORT BY A NEW TRACHEAL TUBE
Vent
Vent
was linked to a 0.5-bar pressure source via a threeway electric solenoid valve. The valve was operated by the servo controller. Occlusion was triggered by pressing an occlusion button to inflate automatically the balloon located on the inspiratory limb, during the next expiration. The controller kept the valve inflated during the first 100 ms of the following inspiration. The expiratory line included a unidirectional valve. An analog controller was used to measure the gas volumes and to trigger injection in the TGIT, via the solenoid valve, in the mode specific for each of the two ventilation methods tested (PCVTCIT or IPSTCIT). Measurement of gas volumes necessitated signal processing. Indeed, during injection (inspiration) the flow received by the patient (FPt) corresponded to the algebraic sum of the flow of the jets (Fjet), which is constant throughout inspiration, and that of entrained gas (Kent) (fig. 3) [10-12]. Therefore, during injection, we measured Kent at the TGIT proximal opening (Fleish No. 1 pneumotacho-
graph Lausanne, Switzerland, and Validyne MP45, ±2.5cmH 2 O, Northridge, California) and added to this signal a constant voltage representing Fjet in order to obtain FPt. Tidal volume was obtained by integration of FPt (Gould, ES1000). In contrast, expiration was passive and Fjet nil. At this time, FPt corresponded to the flow measured by the pneumotachograph. The accuracy of the tidal volume reconstruction from Fjet and Kent had been assessed previously. Error did not exceed ± 1.5 %. In PCVTGIT the controller triggered injection according to the ventilatory frequency rate and inspiratory time selected, by means of an internal clock controlling the injection solenoid valve. Expiration was passive and started from the end of injection. IPSTG1T required spontaneous breathing. In this case, injection was started after the onset of inspiration—that is, when FPt exceeded 100 ml s"1, and stopped when it decreased to less than 25 % of its peak inspiratory value (fig. 4). The controller monitored tidal volume (integration of FPt) in the same way as in PCV. In both modes, expiration occurred freely and tracheal pressure (Paw) returned to atmospheric before injection resumed at the next breath. No self-cycling was observed during the triggering of l r oTGIT.
The mixing of the jets in the trachea has been studied previously on a tracheal model and appeared to be complete within the first 2 cm distal to the TGIT tip. Hence, in order to measure Paw in conditions unaffected by injection, it was measured via a 1-mm i.d. catheter inserted down the trachea, 5 cm below the TGIT tip, which was connected to a pressure transducer (Validyne MP45, ±30cmH 2 O). Oesophageal pressure (Poe) was measured using a balloon catheter and a transducer (Validyne MP 45, ±30 cm H2O). The position of the oesophageal catheter was checked by an occlusion test [13]. Respiratory gas was sampled continuously (ISOmlmin"1) at the Ypiece for end-tidal (PE' CO ,) measurement (Gould, Mark IV). All signals were recorded on an eightchannel paper chart recorder (Gould, ES1000) at 25 mm s~l during each set and at 200 mm s"1 during measurement of P0A. Procedure
The experimental study was conducted in the recovery room during the early postoperative period. Informed consent was obtained from all
Downloaded from http://bja.oxfordjournals.org/ at University of Sydney on March 13, 2015
FIG. 3. Flow regimen during ventilation by the TGIT. Top: The flow of the jets (Kjet) and that measured at the airway opening (Kent) are displayed against time. The different phases of the breathing cycle are: (1) onset of spontaneous inspiration when Kent increases until the injection (2) is triggered, resulting in a constant flow (Kjet) produced by the injectors (3). During the injection phase (3) Kent becomes negative resulting from opening of Kjet to atmosphere via the TGIT. When injection stops (4) at the onset of expiration, the flow recorded by the pneumotachograph is measured by Kent. Bottom: The true flow signal corresponds to the algebraic sum of that measured at the airway opening (Kent) and the constant flow of the jets (Kjet) represented by a calibrated electronic signal. During expiration, the flow of the jets is nil and KPt equals Kent.
797
BRITISH JOURNAL OF ANAESTHESIA
798
using the TGIT without a conventional ventilator. The IPS pressure was set to reach a plateau of 10cmH 2 O by adjusting the driving Flow pressure. At the end of each period the flow, pressure and carbon dioxide signals were reTime corded. They were averaged over 3 min in steady state conditions in order to obtain: ventilatory frequency (/), FPt, FT, Paw, Poe and PE' C O I . Expiration Five occlusion tests were performed and averaged FIG. 4. Comparison between KPt with and without (sponto assess P 0 1 . From these signals, minute extaneous breathing) ( # ) I P S ^ ^ . Fjct induces a brisk increase of inspiratory flow (a to b). It ceases when FPt is 25 % (d) of piratory volume (FE) was computed as the its peak inspiratory value (c). Expiratory flow is greater in product of / and FT and PE' C O , was calculated IPSXC.T because of a larger volume of the breath. from the carbon dioxide trace. The work of breathing was computed automatically (Apple patients and the study was approved by the local lie) as described previously [14], using Poe and Ethics Committee. It comprised two parts. FPt signals to calculate the corresponding work Controlled mode ventilation. The first part was (l^oe). performed before the return of spontaneous breathing. First the patient received controlled Statistical analysis mode ventilation (CMV) and a CPUj ventilator Results are expressed as mean (SD). Com(Ohmeda, France) with standard settings, during a resting period of 30 min. Then the patient was parisons between mean values of each mode of allocated randomly to one 15-min period of CMV ventilation were performed using paired t tests on the CPUX ventilator and to one period of [15]. PCVTQJT generated by gas injection in the TGIT RESULTS with the patient's lungs disconnected from the ventilator. The settings chosen for the two In the first part of the study, when the trachea! gas modes were identical ( F T = 8 ml kg"1, ventilatory injection tube was used in PCVTCT mode, it frequency 14 b.p.m., TI/TT = 33%, no PEEP, appeared to be as efficient as a conventional Fi o , = 0.40). Therefore, FT in PCVTGIT was ventilator. Indeed, using the same ventilatory controlled by adjusting the driving pressure. At settings, the alveolar ventilation achieved was the end of each period and after steady state was identical in the two modes, as assessed by equal achieved, Paw, FPt, F T and PE' COJ signals were values of PE' C 0 ] (4.48 (0.77) kPa compared with recorded and averaged over 3 min. All patients 4.64 (0.77) kPa (ns)). In contrast, the peak underwent ventilation passively at the selected pressures differed slightly between PCVTGIT and settings without spontaneous ventilation whilst CMV (11.7 (6.0) cm H2O and 14.1 (7.0) cm H2O, still under the influence of residual anaesthesia. respectively), but this was not statistically Inspiratory pressure support. The second part of significant. This difference in peak pressures the study was initiated when spontaneous breath- could be caused by differences in flow profile ing resumed. Then 15-min periods of spontaneous which would cause different pressure changes breathing and IPSTGIT were allowed at random, across the respiratory resistance. Inspiration
f
Tl/Tl (%)
Tl (»)
TE
VT
VE
(s)
(ml)
(litre min"1)
14 5
33 9
1.4 0.2
3.2 1.3
442 188
5.7 1.6
5.65 0.72
13
29 9 ns
1.3 0.2 ns
3.5 1.5 ns
557 181
PRESSURE SUPPORT VENTILATION USING A NEW TRACHEAL GAS INJECTION TUBE L. BEYDON, D. ISABEY, G. BOUSSIGNAC, F. BONNET, P. DUVALDESTIN AND A. HARF
to 7.1 (1.7) litre min~') (P < 0.001). It reduced the total work of breathing (from 0.625 (0.223) to 0.263 (0.151) J litre-', respectively) (P < 0.01) and the occlusion pressure (from 2.62 (1.28) to 1.36 (0.74) cm H2O, respectively) (P < 0.01). It is concluded that this TGIT used with a specific system for sensing and triggering ventilation allows inspiratory pressure support during low frequency jet ventilation.
KEY WORDS Equipment: trachea/ gas injection tube. Ventilation: jet ventilation.
Mechanical ventilation is often required in the period after major surgery. Residual effects of anaesthesia [1], abdominal pain [2] and diaphragmatic dysfunction [3] may contribute to depressed ventilation after operation and impair spon-
PATIENTS AND METHODS
Patients and anaesthesia We studied 10 patients (five males) (mean age 61 (SD 11) yr; weight 65 (6) kg; height 167 (7) cm; non-smokers; ASA I or II) in the recovery room after abdominal surgery. Surgery was performed under general anaesthesia with fentanyl 3 ug kg"1, flunitrazepam 0.03 mg kg"1 and vecuronium 0.1 LAURENT BEYDON*, M.D.; DANIEL ISABEY, PH.D.; GEORGES BOUSSIGNAC, M.D.; FRANCOIS BONNET, M.D.; PHILIPPE DUVALDESTIN, M.D.; ALAIN HARP, M.D.; Unite INSERM
U296 and the Departement d'Anesthisie, H6pital Henri Mondor, 51 Avenue du Marechal de Lartre de Tassigny, 94010 Criteil, France. Accepted for Publication: May 9, 1991. •Address for correspondence: Department d'Anesthesie, Hdpital Henri Mondor, 94010 Creteil, France.
Downloaded from http://bja.oxfordjournals.org/ at University of Sydney on March 13, 2015
taneous breathing. To overcome these problems, conventional artificial ventilation may be used In order to explore new types of jet ventilation, after operation, followed by an assisted mode such we tested a trachea! gas injection tube (TGIT) as inspiratory pressure support (IPS), which is a which included six thin capillaries and provided pressure limited mode. IPS is provided by high pressure injection. The driving pressure was ventilators, using sophisticated servo control chosen to yield a plateau of inspiratory trachea/ techniques. We described a new tracheal gas pressure of 10 cm H2O. An original controller injection tube (TGIT) which was able to produce high velocity jets at its distal tip and consequently was built to monitor spirometry and trigger to generate positive pressure by air momentum injection in order to deliver both pressure contransfer down to the injection site [4, 5]. This trolled ventilation (PCVTGIT) and a new mode of system was used successfully for apnoeic veninspiratory pressure support jet ventilation tilation during cardiopulmonary resuscitation in (IPSJQU). The PVCTGn mode maintained the same animals [6], and man [7] and to prevent oxygen end-tidal carbon dioxide concentration as con- desaturation during apnoeic periods in patients ventional ventilation with the same tidal and undergoing mechanical ventilation [8]. We have minute ventilation. We studied 10 patients after also developed a new electronic system which may abdominal surgery. During spontaneous breath- be coupled to the TGIT to produce either ing, the patients were allowed to breathe through pressure controlled ventilation (PCV TGIT) or inthe tube, successively with and without IPSTGn. spiratory pressure support (IPSTCrT). The aim of IPSTGIT' compared with spontaneous breathing this study was to examine high flow jet ventilation increased minute ventilation (from 5.7 (SD 1.6) in both controlled and spontaneous ventilation. SUMMARY
BRITISH JOURNAL OF ANAESTHESIA
796
Compressed gas source
t Entrained air • Proximal end
Capillaries.
Site of injection
FIG. 2. The tracheal gas injection tube (TGIT).
C7
Downloaded from http://bja.oxfordjournals.org/ at University of Sydney on March 13, 2015
A TGIT (fig. 2) (Porges laboratories, Le Plessis Robinson, France) of 8 mm i.d. with a low pressure cuff. In its wall, six thin capillaries (diameter: 720 urn) are moulded by extrusion. These are distributed evenly around the circumference and emerge on the internal aspect of the TGIT, 1 cm before its distal tip. These capillaries are connected to an adjustable high pressure air-oxygen source (driving pressure 0—3 bar) to produce high velocity gas injection at the TGIT distal end, as described previously [8]. The performances of the TGIT were tested in our laboratory before the clinical study. The relationship between the driving pressure and the FIG. 1. System used with the TGIT. A pneumotachograph flow of the injectors was found to be constant for measured the flow (V) passing through the lumen of the all TGIT. The pressure generated by the TGIT TGIT. The flow signal was adapted by an analog system in the trachea (measured at the carinal level) (Processor) to trigger injection via electronic valves (1, 2). The ranged between 0 and 27 cm H2O according to the system also included unidirectional valves on both branches of driving pressure used (0-3.5 bar). For instance, a the Y-piece (3), a pneumatic valve with solenoid for occlusion of 1.2 bar caused a flow of pressure measurement (4), a rubber reservoir bag (5), a driving pressure 1 through the injectors and this humidifier (6) and an air-oxygen mixer and pressure regulator 350 ml s" (7). Tracheal pressure (Paw) was measured by a catheter inside generated a pressure of 10 cm H2O in the trachea. the trachea and oesophagcal pressure (Poe) via a balloon An inspiratory line with an air-oxygen mixer, a tipped catheter. pressure regulator, a cascade humidifier (FisherPaykell, MR 428, New Zealand), a 3-litre rubber mg kg 1 for induction and tracheal intubation. reservoir which served as a mixing chamber to Anaesthesia was maintained with 70% nitrous allow a constant Fio,> a n d a unidirectional valve. oxide in oxygen and administration of increments From the air-oxygen mixer a line supplied high of the same drugs. Patients' lungs were ventilated pressure gas to the capillaries of the TGIT via a perioperatively by a volume cycled ventilator solenoid valve (opening—closing delay: 11 ms). An (Draeger, SAX) in a conventional mode. occluding pneumatic valve was used to measure the occlusion pressure 0.1 s after the beginning of Material and signal processing inspiration (P0.i) [9]. This valve comprised a latex The system we used comprised the following balloon mounted inside a short, 22-mm diameter tube placed in the inspiratory line. This balloon components (fig. 1):
VENTILATORY SUPPORT BY A NEW TRACHEAL TUBE
Vent
Vent
was linked to a 0.5-bar pressure source via a threeway electric solenoid valve. The valve was operated by the servo controller. Occlusion was triggered by pressing an occlusion button to inflate automatically the balloon located on the inspiratory limb, during the next expiration. The controller kept the valve inflated during the first 100 ms of the following inspiration. The expiratory line included a unidirectional valve. An analog controller was used to measure the gas volumes and to trigger injection in the TGIT, via the solenoid valve, in the mode specific for each of the two ventilation methods tested (PCVTCIT or IPSTCIT). Measurement of gas volumes necessitated signal processing. Indeed, during injection (inspiration) the flow received by the patient (FPt) corresponded to the algebraic sum of the flow of the jets (Fjet), which is constant throughout inspiration, and that of entrained gas (Kent) (fig. 3) [10-12]. Therefore, during injection, we measured Kent at the TGIT proximal opening (Fleish No. 1 pneumotacho-
graph Lausanne, Switzerland, and Validyne MP45, ±2.5cmH 2 O, Northridge, California) and added to this signal a constant voltage representing Fjet in order to obtain FPt. Tidal volume was obtained by integration of FPt (Gould, ES1000). In contrast, expiration was passive and Fjet nil. At this time, FPt corresponded to the flow measured by the pneumotachograph. The accuracy of the tidal volume reconstruction from Fjet and Kent had been assessed previously. Error did not exceed ± 1.5 %. In PCVTGIT the controller triggered injection according to the ventilatory frequency rate and inspiratory time selected, by means of an internal clock controlling the injection solenoid valve. Expiration was passive and started from the end of injection. IPSTG1T required spontaneous breathing. In this case, injection was started after the onset of inspiration—that is, when FPt exceeded 100 ml s"1, and stopped when it decreased to less than 25 % of its peak inspiratory value (fig. 4). The controller monitored tidal volume (integration of FPt) in the same way as in PCV. In both modes, expiration occurred freely and tracheal pressure (Paw) returned to atmospheric before injection resumed at the next breath. No self-cycling was observed during the triggering of l r oTGIT.
The mixing of the jets in the trachea has been studied previously on a tracheal model and appeared to be complete within the first 2 cm distal to the TGIT tip. Hence, in order to measure Paw in conditions unaffected by injection, it was measured via a 1-mm i.d. catheter inserted down the trachea, 5 cm below the TGIT tip, which was connected to a pressure transducer (Validyne MP45, ±30cmH 2 O). Oesophageal pressure (Poe) was measured using a balloon catheter and a transducer (Validyne MP 45, ±30 cm H2O). The position of the oesophageal catheter was checked by an occlusion test [13]. Respiratory gas was sampled continuously (ISOmlmin"1) at the Ypiece for end-tidal (PE' CO ,) measurement (Gould, Mark IV). All signals were recorded on an eightchannel paper chart recorder (Gould, ES1000) at 25 mm s~l during each set and at 200 mm s"1 during measurement of P0A. Procedure
The experimental study was conducted in the recovery room during the early postoperative period. Informed consent was obtained from all
Downloaded from http://bja.oxfordjournals.org/ at University of Sydney on March 13, 2015
FIG. 3. Flow regimen during ventilation by the TGIT. Top: The flow of the jets (Kjet) and that measured at the airway opening (Kent) are displayed against time. The different phases of the breathing cycle are: (1) onset of spontaneous inspiration when Kent increases until the injection (2) is triggered, resulting in a constant flow (Kjet) produced by the injectors (3). During the injection phase (3) Kent becomes negative resulting from opening of Kjet to atmosphere via the TGIT. When injection stops (4) at the onset of expiration, the flow recorded by the pneumotachograph is measured by Kent. Bottom: The true flow signal corresponds to the algebraic sum of that measured at the airway opening (Kent) and the constant flow of the jets (Kjet) represented by a calibrated electronic signal. During expiration, the flow of the jets is nil and KPt equals Kent.
797
BRITISH JOURNAL OF ANAESTHESIA
798
using the TGIT without a conventional ventilator. The IPS pressure was set to reach a plateau of 10cmH 2 O by adjusting the driving Flow pressure. At the end of each period the flow, pressure and carbon dioxide signals were reTime corded. They were averaged over 3 min in steady state conditions in order to obtain: ventilatory frequency (/), FPt, FT, Paw, Poe and PE' C O I . Expiration Five occlusion tests were performed and averaged FIG. 4. Comparison between KPt with and without (sponto assess P 0 1 . From these signals, minute extaneous breathing) ( # ) I P S ^ ^ . Fjct induces a brisk increase of inspiratory flow (a to b). It ceases when FPt is 25 % (d) of piratory volume (FE) was computed as the its peak inspiratory value (c). Expiratory flow is greater in product of / and FT and PE' C O , was calculated IPSXC.T because of a larger volume of the breath. from the carbon dioxide trace. The work of breathing was computed automatically (Apple patients and the study was approved by the local lie) as described previously [14], using Poe and Ethics Committee. It comprised two parts. FPt signals to calculate the corresponding work Controlled mode ventilation. The first part was (l^oe). performed before the return of spontaneous breathing. First the patient received controlled Statistical analysis mode ventilation (CMV) and a CPUj ventilator Results are expressed as mean (SD). Com(Ohmeda, France) with standard settings, during a resting period of 30 min. Then the patient was parisons between mean values of each mode of allocated randomly to one 15-min period of CMV ventilation were performed using paired t tests on the CPUX ventilator and to one period of [15]. PCVTQJT generated by gas injection in the TGIT RESULTS with the patient's lungs disconnected from the ventilator. The settings chosen for the two In the first part of the study, when the trachea! gas modes were identical ( F T = 8 ml kg"1, ventilatory injection tube was used in PCVTCT mode, it frequency 14 b.p.m., TI/TT = 33%, no PEEP, appeared to be as efficient as a conventional Fi o , = 0.40). Therefore, FT in PCVTGIT was ventilator. Indeed, using the same ventilatory controlled by adjusting the driving pressure. At settings, the alveolar ventilation achieved was the end of each period and after steady state was identical in the two modes, as assessed by equal achieved, Paw, FPt, F T and PE' COJ signals were values of PE' C 0 ] (4.48 (0.77) kPa compared with recorded and averaged over 3 min. All patients 4.64 (0.77) kPa (ns)). In contrast, the peak underwent ventilation passively at the selected pressures differed slightly between PCVTGIT and settings without spontaneous ventilation whilst CMV (11.7 (6.0) cm H2O and 14.1 (7.0) cm H2O, still under the influence of residual anaesthesia. respectively), but this was not statistically Inspiratory pressure support. The second part of significant. This difference in peak pressures the study was initiated when spontaneous breath- could be caused by differences in flow profile ing resumed. Then 15-min periods of spontaneous which would cause different pressure changes breathing and IPSTGIT were allowed at random, across the respiratory resistance. Inspiration
f
Tl/Tl (%)
Tl (»)
TE
VT
VE
(s)
(ml)
(litre min"1)
14 5
33 9
1.4 0.2
3.2 1.3
442 188
5.7 1.6
5.65 0.72
13
29 9 ns
1.3 0.2 ns
3.5 1.5 ns
557 181
