- Email: [email protected]
Technique of lung ventilation through an injector
Resuscitation, 6,91-95
Technique
of lung ventilation
through an injector
E. POPA Anaesthesiology Department of the Ministry of Transport, Hospital No. 1, Witting Street 37, Bucharest 12, Romania
Summary Clinical and experimental evidence that the risks of ventilation through an injector can be diminished if the rate of gas flow in the tracheal probe is kept slow is presented. In practice those inlet devices for oxygen which permit an acceleration of the flow at the end of insufflation give better ventilation. The oxygen and air mixture is more uniform and the distribution is more physiological if there are side holes in the tube; the intake of additional air is then not due to the Bernoulli effect but to the effect of propulsion. Lung ventilation through an injector is worthwhile and is the only possible method in some cases, but it demands skill and care. Introduction Laryngeal microsurgery involves difficult anaesthetic problems, such as free laryngeal access for the surgeon, relaxed vocal cords (immobile and without spasm), relaxation of the neck and mandible, adequate period of anaesthesia for 1 S-50 min, and rapid recovery of reflexes at the end of operation. The myorelaxants meet the above requirements. Ventilation with an intermittent positive pressure by injection of oxygen originating from the endoscopical methods, is considered to be the best technique so far.
Technical methods and applications (a) With tubes of 2 mm internal diameter with a nozzle without side holes and with oxygen admission through a pneumatic trigger A balloon of volume 1500 ml in which oxygen is injected through a 0.5 m long probe with an internal diameter of 2 mm through a spring device at a pressure of 3 barr is filled, and reaches a pressure on the walls of 15 mmHg in 3 s. The flow rate through the tube at 3-4 barr from a static pressure source is 700-1000 ml/s. One has to take into consideration the Bernoulli effect in ventilation by injecting oxygen in the trachea: a fluid jet with a certain speed stimulates the environment by means of molecular friction, in this case air from the room into the insufflated lung (Lee & Atkinson, 1975). The effect is maximum if the internal diameter of the probe (injector) is l-2 mm, with a diameter of the trachea (diffuser) of 8-15 mm, if the probe has a conical end and a 91
92
E. POPA
nozzle at the distal end, and is placed in the tracheal axis. In these conditions the Bernoulli effect adds 4000 ml of additional air to the flow through the probe. A lung model, also of 1500 ml, opened into the atmosphere fills in 0.75 s instead of 3 s, and an insufflation may reach 5000 ml of oxygen and air mixture (almost the entire lung capacity), which is obviously excessive. In practice, the probe never remains in the tracheal axis; it is closed in upon the trachea wall, outside Kleinsasser’s laryngoscope tube, usually in the posterior commissure of the vocal cords. The same pattern is seen if the injector is closed in upon the internal wall of the trachea and not in its axis; it is filled in 1 s. The Bernoulli effect does not disappear due to the side movement of the probe; it is diminished from four-fifths to two-thirds of the previous rate in the probe. The diminution of pressure created at the entrance in the trachea is reduced from -10 cm water in the axially placed probe to -3 cm/water when the probe is placed sideways. For the system to operate properly the respiratory tract should have free communication with the atmosphere, particularly because the expiration is performed passively. The gas flow speed in the probes of 2 mm internal diameter and with rates of about 1000 ml/s may exceed 260 km/h: this jet concentrated on a small surface constitutes a dangerous force, which may easily cause lung rupture (Sauvage, Klein, Beutter, Fleury & Desmonts, 1976) particularly if the injection is performed by means of a spring knob (pneumatic trigger) and the maximum acceleration takes place at the beginning of the insufflation with sudden drastic increase of output. On the transparent plastic models the probe floating freely in the trachea can be observed; the air mixed with oxygen is unequally distributed to the two lungs, according to eddy currents, the form of the carena and the different angularity of the bronchi; ventilation is not uniform from one insufflation to another. If the probe is placed in front of the right bronchi, the right lung received oxygen predominantly at a higher pressure than the left lung. A nozzle probe without side holes and with an internal diameter of 2 mm with its end 2-3 cm below the inferior end of the Kleinsasser’s laryngoscope tube, when it is placed on the right side without free movement, is more dangerous than a free probe. Experimentally in lung models, the pressure in the right bronchi may reach 750 g/cm2 (about 150 g/cm2 for each bronchial segment). Any tendency towards excess pressure may easily burst the alveoli; the compulsory condition for eliminating the danger is that the admission pressur of 4000 g/cm2 is to be diminished 200 to 400 times at the alveolar surface. The theoretical pressure which is not dangerous to the alveolar walls is lo-20 g/cm2. The energy which hits the alveolar walls, when the probe is on the right and cannot move, may be greater than the latter, particularly during acceleration of the flow at the beginning of injection. If the pressure of 15-25 mmHg is established in the lung and the air duct .is free, the insufflation phase reverses automatically; when the gas is administered, the mixture of oxygen and air comes out; the phase shift behaves as if it relieves a pressure (Rioux, Guerrier & du Cailer, 1975). (b) With tubes of 4 mm internal diameter without a nozzle, with side holes, and with the oxygen entry through a T tube
LUNG VENTILATION
93
The gas s.ource pressure remained unchanged within 3-4 barr. With the T branch closed, the filling time of the lung model was not significantly changed, although the rate was doubled when the section increased from 3 mm2 to 12.5 mm2. The Bernoulli effect was almost cancelled both because of lack of the nozzle and because the insufflating speed wd reduced to one-third of the values found with the thinner probes. With greater probes the distribution was also not uniform but the initial pressure of 3-4 barr was more equally distributed over the alveolar surface. It is important clinically that one could adjust the rate by means of partial or complete closure of the T branch. The main advantage of the adjustment of the pressure at entry is the acceleration of flow at the end rather than at the beginning of the insufflation. This resulted in an intra-lung pressure which, at a frequency of 8-10 cycles/min, and with an I/E rate of 2/4 s, provided an alveolar contact time of 0.4-0.5 s; this is sufficient for adequate diffusion. With one side hole on the probe the size and orientation of the walls of the hole are very important; the hole should point downwards and obliquely towards the branch of the trachea; its orientation in respect of the anterior or posterior tracheal wall is not important. If the hole is directed towards the lateral tracheal wall, it produces a secondary control jet, similar to a Coanda fluid bistable element; the jet reflected downwards from the tracheal wall intercepts and directs the principal terminal jet towards the opposite lung; such probes are therefore not usually taken into consideration (Popa, 1976). Four side holes in the four cardinal points 2 cm from the tip of the probe, centre it automatically and produce an ‘air piston’, which compresses and pushes the gas column in front of it in the respiratory tract. This ‘propulsion effect’ is combined with the air which is sucked from the space behind the piston. The entry of the outside air is due to a mechanism completely different from the Bernoulli effect (Barava, 1975). Naturally, the air ducts must be free for this system to operate.
Clinical results Premeditation consisted of: atropine, 0.50 mg; diazepam, 20 mg; dihydrobenzperidol, 0.3-0.4 mg/kg. The anaesthetic was nitrous oxide (66%) and fluothane until the level of anaesthesia was satisfactory. Nasal and pharyngeal anaesthesia was achieved with local lignocaine (4%). Succinylcholine (50 mg) and then perfusion of succinylcholine solution (0.2%) at a rate of 0.1 mg/kg per min were given. This was equivalent to 50 drops/min for a 70 kg patient (Bouche, Pech, Piquet, F&he & Pardes, 1973). The laryngoscope with a curved blade was inserted, the larynx and nose were sprayed with lignocaine, the probe was guided under direct vision to 5 cm below the vocal cords. Installation of the Kleinsasser laryngoscope is often difficult. Fentanyl(O.01 mg/kg) was given before the operation began. The probe was fed from a constant pressure source. The system appears as a flux generator in which the pressure gradient between the apparatus and patient did not decrease, the lung volume increased proportionally to the pressure and the intrapulmonary pressure was a function of compliance. The circuit may be compared to a semi-open circuit with active inspiration and passive expiration. We have eliminated any manual or automatic mechanical device for insufflation. We use a T-piece branch on the probe circuit; progressive closure of its free orifice allows an adequate rate and the acceleration of the flux at the end of the insufflation.
94
E. POPA
As the oxygen flux in the probe does not stop even during the expiratory pause, the insufflation occurs as positive pressure peaks, while the intra-lung pressure is also positive. The intra-oesophageal curves, recorded in ten cases, showed a continuous positive pressure which was less convenient from the haemodynamic point of view and for the carbon dioxide diffusion, but more convenient for ventilation; dead space and arterial PO2 were increased, which would prevent atelectasis. When a pressure in the lungs has been reached, the gas flow direction changes in the system by itself; it reverses at a certain ‘rejecting’ pressure which is different from one case to another. This constant pressure must be sought by sounding and when it has been reached, the expiration should be left free. ‘Level pressure’, important for a good and correct ventilation and for the safety of the patient, is valid only with the glottis open. Insufflation overcomes the dynamic resistance rather easily: the elasticity of the lungs and chest, the flow resistance in the air ducts and the deformation of the lungs. Use of increased pressure is not recommended if these resistances are not easily eliminated. Emphysema is an absolute contraindication for jet-ventilation; fibrous lung with a rigid chest and small compliance, pleurisy, and bronchopneumonia with bubbles due to localized emphysema, are relative contraindications. The static resistance represented by the weight of the Kleinsasser’s laryngoscope support, which presses rather heavily on the patient’s sternum is additional, unwanted and rather difficult to overcome, especially if the curarizing is slightly excessive. The most convenient control is to observe the expansion of the chest, a convenient inspiratory position and, later on, auscultation with a stethoscope is used to assess the efficacy of the insufflation. Our experience covers 55 operations. In one of the first 10 cases, the Kleinsasser’s laryngoscope slipped at the moment of lung insufflation, creating almost complete respiratory obstruction, there followed a peribronchical alveolar rupture, interstitial emphysema in the lungs, emphysema in the mediastinum and subcutaneously, pneumothorax. The patient recovered. In the following 12 surgical interventions we started to insufflatc with the T piece, but also continued to use 2 mm probes. We could not explain a case of atelectasis but postulated that the oxygen jet dragged a clot which blocked a branch of an air duct. In the last 33 surgical interventions we have worked with 4 mm interior diameter probes without nozzle, with or without side holes. In one of the cases, during the radiological control, we discovered an hyperaerated area of the lung without any other clinical signs. Probably, a valve effect trapped some air and gave a slight excess pressure which progressed insidiously during the operation; the valve effect could be due to secretions or thrombi pushed into the respiratory ducts. Discussion
It is generally recognized that there is a risk of local rise of pressure in spite of all precautions. In our opinion, the first step should be to slow the flow, the main force of which can cause rupture of the lung. The chest which finds itself under atmospheric pressure draws in air with avidity (Souron, Ginguene, Legent & Nicolas, 1974). A nasal wound while introducing the probe, or in the throat while introducing the suspended laryngoscope (Herman, 1972), allows air to be sucked in. It is not possible to establish the role of critical excess pressure in the case of the most insignificant respiratory obstruction.
LUNG VENTILATION
95
We continued to find the Bernoulli effect when the injector was close to the tracheal wall, although it diminished. The molecular dragging of the air continued to take place on one side of the jet from the injector. There are little data in the literature about the dynamic behaviour of the gases when the injector has side holes as well; we have found that they can be favourable if directed correctly and if their sizes are chosen judiciously; their diameters should be about onequarter of that of the injector. Any aspect less studied is the effect of propulsion and final distribution of pressure on the alveolar surface in these cases. As there can be no discussion about the laminar currents, the eddy currents must necessarily occur, and in our opinion these may help to make the air/oxygen mixture uniform. Finally, the unequal gas distribution in the lungs at each insufflation did not have nay significant negative effect upon Pa,02 and Pa,CO;l. Conclusions The replacement of an injection device with flux acceleration at the beginning of the insufflation with a device which allows final acceleration, avoids the sudden very rapid flow, allows an adjustable rate without modifying the frequency and improves the ventilation conditions. Reduction of the flow rate by increasing the area of the injector and by eliminating the nozzle, and the predominance of the propulsion effect compared with the Bernoulli effect, proved to be useful manoeuvres. If patients are selected, ventilation through a jet remains valuable. References Barava, A. (1975) The challenge of anaesthesia for microlaryngoscopic of Anaesthesiology
procedures. Middle East Journal
4, 1-5.
Bouche, J., Pech, A., Piquet, J. J., F&he, Ch. & Pardes, P. (1973) La Micro-chirurgie Laryngke sous Suspension, pp. 72 and 74. Arnette, Paris. Herman, B. (1972) Un cause peu frdquante d’emphyseme souscutand. Cuhiers d’Anesth&iologie 20, 277-218. Lee, J. A. & Atkinson, R. S. (1975) A Synopsis of Anaesthesia, p. 65. Wright, Bristol.
a la technique de la respiration artificielle sur sonde mince. First Abstracts. Bucharest, USSM. p. 49. Rioux, J., Guerrier, B. & du Cailar, J. (1975) Ventilation par injection (jet-ventilation) sous anesthesie gem!rale pour endoscopie oto-rhino-laryngologique. Ann. Anesth. Franc. 16, lJ-9 J. Sauvage, J. P., Klein, J. P., Beutter, P., Fleury, P. & Desmonts, J. M. (1976) Notre experience de I’oxygene pulse et de l’anesthesie gdnt%ale dans les laryngoscopies directes en suspension. Ann. Oto.
Popa, E. (1976) Contributions
National Congress of Anaesthesia and Intensive are.
Laryng. (Paris) 93, 577-578.
Souron, R., Ginguene, P. H., Legent, F. & Nicholas, F. (1974) Les epanchements gazeux dans le mddiastin, la cavite pleurale et le tissu sous-cutand en periode operatoire. Anesth. Analg. Rbn. 31,739-751.
Technique
of lung ventilation
through an injector
E. POPA Anaesthesiology Department of the Ministry of Transport, Hospital No. 1, Witting Street 37, Bucharest 12, Romania
Summary Clinical and experimental evidence that the risks of ventilation through an injector can be diminished if the rate of gas flow in the tracheal probe is kept slow is presented. In practice those inlet devices for oxygen which permit an acceleration of the flow at the end of insufflation give better ventilation. The oxygen and air mixture is more uniform and the distribution is more physiological if there are side holes in the tube; the intake of additional air is then not due to the Bernoulli effect but to the effect of propulsion. Lung ventilation through an injector is worthwhile and is the only possible method in some cases, but it demands skill and care. Introduction Laryngeal microsurgery involves difficult anaesthetic problems, such as free laryngeal access for the surgeon, relaxed vocal cords (immobile and without spasm), relaxation of the neck and mandible, adequate period of anaesthesia for 1 S-50 min, and rapid recovery of reflexes at the end of operation. The myorelaxants meet the above requirements. Ventilation with an intermittent positive pressure by injection of oxygen originating from the endoscopical methods, is considered to be the best technique so far.
Technical methods and applications (a) With tubes of 2 mm internal diameter with a nozzle without side holes and with oxygen admission through a pneumatic trigger A balloon of volume 1500 ml in which oxygen is injected through a 0.5 m long probe with an internal diameter of 2 mm through a spring device at a pressure of 3 barr is filled, and reaches a pressure on the walls of 15 mmHg in 3 s. The flow rate through the tube at 3-4 barr from a static pressure source is 700-1000 ml/s. One has to take into consideration the Bernoulli effect in ventilation by injecting oxygen in the trachea: a fluid jet with a certain speed stimulates the environment by means of molecular friction, in this case air from the room into the insufflated lung (Lee & Atkinson, 1975). The effect is maximum if the internal diameter of the probe (injector) is l-2 mm, with a diameter of the trachea (diffuser) of 8-15 mm, if the probe has a conical end and a 91
92
E. POPA
nozzle at the distal end, and is placed in the tracheal axis. In these conditions the Bernoulli effect adds 4000 ml of additional air to the flow through the probe. A lung model, also of 1500 ml, opened into the atmosphere fills in 0.75 s instead of 3 s, and an insufflation may reach 5000 ml of oxygen and air mixture (almost the entire lung capacity), which is obviously excessive. In practice, the probe never remains in the tracheal axis; it is closed in upon the trachea wall, outside Kleinsasser’s laryngoscope tube, usually in the posterior commissure of the vocal cords. The same pattern is seen if the injector is closed in upon the internal wall of the trachea and not in its axis; it is filled in 1 s. The Bernoulli effect does not disappear due to the side movement of the probe; it is diminished from four-fifths to two-thirds of the previous rate in the probe. The diminution of pressure created at the entrance in the trachea is reduced from -10 cm water in the axially placed probe to -3 cm/water when the probe is placed sideways. For the system to operate properly the respiratory tract should have free communication with the atmosphere, particularly because the expiration is performed passively. The gas flow speed in the probes of 2 mm internal diameter and with rates of about 1000 ml/s may exceed 260 km/h: this jet concentrated on a small surface constitutes a dangerous force, which may easily cause lung rupture (Sauvage, Klein, Beutter, Fleury & Desmonts, 1976) particularly if the injection is performed by means of a spring knob (pneumatic trigger) and the maximum acceleration takes place at the beginning of the insufflation with sudden drastic increase of output. On the transparent plastic models the probe floating freely in the trachea can be observed; the air mixed with oxygen is unequally distributed to the two lungs, according to eddy currents, the form of the carena and the different angularity of the bronchi; ventilation is not uniform from one insufflation to another. If the probe is placed in front of the right bronchi, the right lung received oxygen predominantly at a higher pressure than the left lung. A nozzle probe without side holes and with an internal diameter of 2 mm with its end 2-3 cm below the inferior end of the Kleinsasser’s laryngoscope tube, when it is placed on the right side without free movement, is more dangerous than a free probe. Experimentally in lung models, the pressure in the right bronchi may reach 750 g/cm2 (about 150 g/cm2 for each bronchial segment). Any tendency towards excess pressure may easily burst the alveoli; the compulsory condition for eliminating the danger is that the admission pressur of 4000 g/cm2 is to be diminished 200 to 400 times at the alveolar surface. The theoretical pressure which is not dangerous to the alveolar walls is lo-20 g/cm2. The energy which hits the alveolar walls, when the probe is on the right and cannot move, may be greater than the latter, particularly during acceleration of the flow at the beginning of injection. If the pressure of 15-25 mmHg is established in the lung and the air duct .is free, the insufflation phase reverses automatically; when the gas is administered, the mixture of oxygen and air comes out; the phase shift behaves as if it relieves a pressure (Rioux, Guerrier & du Cailer, 1975). (b) With tubes of 4 mm internal diameter without a nozzle, with side holes, and with the oxygen entry through a T tube
LUNG VENTILATION
93
The gas s.ource pressure remained unchanged within 3-4 barr. With the T branch closed, the filling time of the lung model was not significantly changed, although the rate was doubled when the section increased from 3 mm2 to 12.5 mm2. The Bernoulli effect was almost cancelled both because of lack of the nozzle and because the insufflating speed wd reduced to one-third of the values found with the thinner probes. With greater probes the distribution was also not uniform but the initial pressure of 3-4 barr was more equally distributed over the alveolar surface. It is important clinically that one could adjust the rate by means of partial or complete closure of the T branch. The main advantage of the adjustment of the pressure at entry is the acceleration of flow at the end rather than at the beginning of the insufflation. This resulted in an intra-lung pressure which, at a frequency of 8-10 cycles/min, and with an I/E rate of 2/4 s, provided an alveolar contact time of 0.4-0.5 s; this is sufficient for adequate diffusion. With one side hole on the probe the size and orientation of the walls of the hole are very important; the hole should point downwards and obliquely towards the branch of the trachea; its orientation in respect of the anterior or posterior tracheal wall is not important. If the hole is directed towards the lateral tracheal wall, it produces a secondary control jet, similar to a Coanda fluid bistable element; the jet reflected downwards from the tracheal wall intercepts and directs the principal terminal jet towards the opposite lung; such probes are therefore not usually taken into consideration (Popa, 1976). Four side holes in the four cardinal points 2 cm from the tip of the probe, centre it automatically and produce an ‘air piston’, which compresses and pushes the gas column in front of it in the respiratory tract. This ‘propulsion effect’ is combined with the air which is sucked from the space behind the piston. The entry of the outside air is due to a mechanism completely different from the Bernoulli effect (Barava, 1975). Naturally, the air ducts must be free for this system to operate.
Clinical results Premeditation consisted of: atropine, 0.50 mg; diazepam, 20 mg; dihydrobenzperidol, 0.3-0.4 mg/kg. The anaesthetic was nitrous oxide (66%) and fluothane until the level of anaesthesia was satisfactory. Nasal and pharyngeal anaesthesia was achieved with local lignocaine (4%). Succinylcholine (50 mg) and then perfusion of succinylcholine solution (0.2%) at a rate of 0.1 mg/kg per min were given. This was equivalent to 50 drops/min for a 70 kg patient (Bouche, Pech, Piquet, F&he & Pardes, 1973). The laryngoscope with a curved blade was inserted, the larynx and nose were sprayed with lignocaine, the probe was guided under direct vision to 5 cm below the vocal cords. Installation of the Kleinsasser laryngoscope is often difficult. Fentanyl(O.01 mg/kg) was given before the operation began. The probe was fed from a constant pressure source. The system appears as a flux generator in which the pressure gradient between the apparatus and patient did not decrease, the lung volume increased proportionally to the pressure and the intrapulmonary pressure was a function of compliance. The circuit may be compared to a semi-open circuit with active inspiration and passive expiration. We have eliminated any manual or automatic mechanical device for insufflation. We use a T-piece branch on the probe circuit; progressive closure of its free orifice allows an adequate rate and the acceleration of the flux at the end of the insufflation.
94
E. POPA
As the oxygen flux in the probe does not stop even during the expiratory pause, the insufflation occurs as positive pressure peaks, while the intra-lung pressure is also positive. The intra-oesophageal curves, recorded in ten cases, showed a continuous positive pressure which was less convenient from the haemodynamic point of view and for the carbon dioxide diffusion, but more convenient for ventilation; dead space and arterial PO2 were increased, which would prevent atelectasis. When a pressure in the lungs has been reached, the gas flow direction changes in the system by itself; it reverses at a certain ‘rejecting’ pressure which is different from one case to another. This constant pressure must be sought by sounding and when it has been reached, the expiration should be left free. ‘Level pressure’, important for a good and correct ventilation and for the safety of the patient, is valid only with the glottis open. Insufflation overcomes the dynamic resistance rather easily: the elasticity of the lungs and chest, the flow resistance in the air ducts and the deformation of the lungs. Use of increased pressure is not recommended if these resistances are not easily eliminated. Emphysema is an absolute contraindication for jet-ventilation; fibrous lung with a rigid chest and small compliance, pleurisy, and bronchopneumonia with bubbles due to localized emphysema, are relative contraindications. The static resistance represented by the weight of the Kleinsasser’s laryngoscope support, which presses rather heavily on the patient’s sternum is additional, unwanted and rather difficult to overcome, especially if the curarizing is slightly excessive. The most convenient control is to observe the expansion of the chest, a convenient inspiratory position and, later on, auscultation with a stethoscope is used to assess the efficacy of the insufflation. Our experience covers 55 operations. In one of the first 10 cases, the Kleinsasser’s laryngoscope slipped at the moment of lung insufflation, creating almost complete respiratory obstruction, there followed a peribronchical alveolar rupture, interstitial emphysema in the lungs, emphysema in the mediastinum and subcutaneously, pneumothorax. The patient recovered. In the following 12 surgical interventions we started to insufflatc with the T piece, but also continued to use 2 mm probes. We could not explain a case of atelectasis but postulated that the oxygen jet dragged a clot which blocked a branch of an air duct. In the last 33 surgical interventions we have worked with 4 mm interior diameter probes without nozzle, with or without side holes. In one of the cases, during the radiological control, we discovered an hyperaerated area of the lung without any other clinical signs. Probably, a valve effect trapped some air and gave a slight excess pressure which progressed insidiously during the operation; the valve effect could be due to secretions or thrombi pushed into the respiratory ducts. Discussion
It is generally recognized that there is a risk of local rise of pressure in spite of all precautions. In our opinion, the first step should be to slow the flow, the main force of which can cause rupture of the lung. The chest which finds itself under atmospheric pressure draws in air with avidity (Souron, Ginguene, Legent & Nicolas, 1974). A nasal wound while introducing the probe, or in the throat while introducing the suspended laryngoscope (Herman, 1972), allows air to be sucked in. It is not possible to establish the role of critical excess pressure in the case of the most insignificant respiratory obstruction.
LUNG VENTILATION
95
We continued to find the Bernoulli effect when the injector was close to the tracheal wall, although it diminished. The molecular dragging of the air continued to take place on one side of the jet from the injector. There are little data in the literature about the dynamic behaviour of the gases when the injector has side holes as well; we have found that they can be favourable if directed correctly and if their sizes are chosen judiciously; their diameters should be about onequarter of that of the injector. Any aspect less studied is the effect of propulsion and final distribution of pressure on the alveolar surface in these cases. As there can be no discussion about the laminar currents, the eddy currents must necessarily occur, and in our opinion these may help to make the air/oxygen mixture uniform. Finally, the unequal gas distribution in the lungs at each insufflation did not have nay significant negative effect upon Pa,02 and Pa,CO;l. Conclusions The replacement of an injection device with flux acceleration at the beginning of the insufflation with a device which allows final acceleration, avoids the sudden very rapid flow, allows an adjustable rate without modifying the frequency and improves the ventilation conditions. Reduction of the flow rate by increasing the area of the injector and by eliminating the nozzle, and the predominance of the propulsion effect compared with the Bernoulli effect, proved to be useful manoeuvres. If patients are selected, ventilation through a jet remains valuable. References Barava, A. (1975) The challenge of anaesthesia for microlaryngoscopic of Anaesthesiology
procedures. Middle East Journal
4, 1-5.
Bouche, J., Pech, A., Piquet, J. J., F&he, Ch. & Pardes, P. (1973) La Micro-chirurgie Laryngke sous Suspension, pp. 72 and 74. Arnette, Paris. Herman, B. (1972) Un cause peu frdquante d’emphyseme souscutand. Cuhiers d’Anesth&iologie 20, 277-218. Lee, J. A. & Atkinson, R. S. (1975) A Synopsis of Anaesthesia, p. 65. Wright, Bristol.
a la technique de la respiration artificielle sur sonde mince. First Abstracts. Bucharest, USSM. p. 49. Rioux, J., Guerrier, B. & du Cailar, J. (1975) Ventilation par injection (jet-ventilation) sous anesthesie gem!rale pour endoscopie oto-rhino-laryngologique. Ann. Anesth. Franc. 16, lJ-9 J. Sauvage, J. P., Klein, J. P., Beutter, P., Fleury, P. & Desmonts, J. M. (1976) Notre experience de I’oxygene pulse et de l’anesthesie gdnt%ale dans les laryngoscopies directes en suspension. Ann. Oto.
Popa, E. (1976) Contributions
National Congress of Anaesthesia and Intensive are.
Laryng. (Paris) 93, 577-578.
Souron, R., Ginguene, P. H., Legent, F. & Nicholas, F. (1974) Les epanchements gazeux dans le mddiastin, la cavite pleurale et le tissu sous-cutand en periode operatoire. Anesth. Analg. Rbn. 31,739-751.