- Email: [email protected]
Capnometry during High-Frequency Oscillatory Ventilation
_----.. ~IIII. t---I_ab_o_ra_to_r_y_an_.d_il----:..ni_ma_l_in_ve_s_ti_ga_t-_lon_s
_
Capnometry during High-Frequency Oscillatory Ventilation· . Masaji Nishimura, M.D.; Hideaki lmanaka, M.D.; Chikara Tashiro, Nobuyuki Taenaka, M.D.; and lkuto ¥os!liya, M.D.
~/.D.;
We used capnometry during high-frequency oscillatory ventilation (HFOV), and compared COl measurements at the distal and proximal ends of an endotracheal tube with arterial COl values. Ten white rabbits (mean weight, 2.00±O.2 [SD] kg) underwent tracheostomy under anesthesia with pentobarbital. The trachea was intubated with an endotracheal tube with a second lumen for sampling respiratory gas at the distal tip. Capnometry was performed through the lumen (COld) and the proximal end of the endotracheal tube (C0sP). The internal carotid artery was cannulated to sample blood for measuring arterial blood
gases. The differences between COld, COlp, and PaCOI were measured. Ooly the relation between COld and PaCO. was good (r=O.915). We concluded that capnometry can be used during HFOV to estimate PaCOs provided that respiratory gas is sampled from the distal tip of the endotracheal tube. (Chest 1992; 101:1681-83)
Capnometry and pulse oximetry are rapidly becom-I ing standard procedures in monitoring critically ill children. Capnometry provides a measurement of the CO2 concentration in the patient's respiratory gas. However, its use during high-frequency oscillatory ventilation (HFOV) has never been reported. In the present study, we used capnometry during HFO~ and determined the relationships between the CO2 concentration at the distal end of an endotracheal tube (C02d), the CO2 concentration at the proximal end of an endotracheal tube (C02P), and PaC02 •
METHODS
=
CO.d CO. concentration at the distal end of the endotracheal tube; COd» CO. concentration at the proximal end of the endotracheal tube; HFOV = high-frequency oscillatory ventilation; r-CO. CO. concentration of respiratory gas
*From the Department of Anesthesiology, Osaka Medical Center and Researcli Institute for Maternal and Child Health (Drs Nishimura and Tashiro), and the Intensive Care Unit, Osaka University Hospital (Drs Imanaka, Taenaka, and Yoshiya), Osaka, Japan. Reprint requests: Dr. Nishimura, De]J!Jrlment of Anesthesiology, Osaka Institute for ~Iatemal and Child Health, Osaka, japan 59002
(
= =
The experiment was performed on ten white rabbits weighing 1.16 to 2.21 kg (mean, 2.0±O.2 [SD] kg). The rabbits were anesthetized with intravenous pentobarbital, with supplemental doses as needed to suppress the corneal reHex. After tracheostonlY, the animals were intubated with a 4-mm-internal-diameter tracheal tube with a second lumen for sampling respiratory gas (Mallinckrodt, Glens Falls, NY). The distal tip of the tube was positioned just above the carina under x-ray guidance. The internal carotid artery was cannulated for samplin~ blood to analyze PaC02 • The analysis was performed with an ABL2 blood gas analyzer (Radiometer, Copenhagen). The rabbit's lungs were ventilated by using the HFOV mode of a IIummingbird B~tO 20N olechanical ventilator (~tera, lhkyo). The operating L'Onditions were set as follows: stroke volunle, 4, 6, 8, 10, and 12 ml; frequency, 15 cyclesls; inspiration-expiration ratio, 1: 1; mean airway pressure, 3 cm H 20; fresh gas flow rate, 1 or 2 Umin. At each setting, the PaC02 and the CO2 concentration of the respiratory gas (r-COJ were oleasured simultaneously. The respiratory gas was sampled through the second lumen of the tracheal tube from just above the carina (r-C02d) and from the proximal end
IHummingbird
"--------........
SMO 20N
Nellcor N-1000 Three - way stopcocks
FIGURE 1. Experimental schema. Rabbits underwent tracheostomy and intubation with an endotrclcheal tube with a se<..'Ond lumen for satnpling the respiratory gas. During "FOV with a llummingbird ventilator, capnography was perulrmed with a Nellcor monitor on the respiratory gas sampled froln the distal and proximal ends of the endotracheal tube.
CHEST / 101 / 6 / JUNE, 1992
1681
Endotrachea 1 tube
----.-..__~ Nellcor
N-l000
Three-way stopcock FICURE 2. Under HFOV the capnometric waveforms do not show fluctuation, and the capnometer cannot display partial pressures. To make this possible, a three-way stopcock was placed between the sampling tube and was connected directly to the endotracheal tube and the tube of the NeUcor monitor. The stopcock was turned several times to sample the respiratory gas and room air alternately. This technique produced square waveforms. of the tube (r-COdl). The r-C02 was measured with a Nellcor (Hayward, Calif) N-l000 monitor (Fig 1). The sampling rate was 50 mVmin; length of sampling line, 1.52 m; response time, 85 ms. During HFO~ r-C02 is constant, and the monitor, expecting the cyclic appearance and disappearance of CO2 in the respiratory gas, does not report numeric data but rather sounds an alarm if apnea is detected. To measure r-C02 during HFO~ a three-way stopcock was placed between the respiratory circuit and the sampling tube of the monitor. The cock was turned to sample the respiratory gas and the room air alternately (Fig 2). These arti6cial fluctuations in the CO2 concentration created the cyclic appearance and disappearan<'"e of CO2 ; thus, the monitor reported minimum and maximum CO2 concentrations with each respiratory cycle. We turned the cock to create arti6cial fluctuations of CO2 concentration at least ten times, and established the stability of the numeric values displayed by the monitor for end-tidal CO2 , RESULTS
In all rabbits, low stroke volume induced hypercarbia. The lower the stroke volume, the higher the rC02d and PaC02 • Because severe bradycardia or bigeminy pulse occurred at stroke volume settings under 4 or 6 ml, the experiment was discontinued at (mmHg)
70 60
50 ~ 40
0 <.;> 30 '-
20 10 0
0
60 (mmHg)
FIGURE 3. Relation between r-C02d and PaC02 as well as the line of identity.
1882
that stroke volume. The total number ofsample points was, therefore, 88 for PaC02 and r-C02d (PaC02 = 1.33 [r-C02d] -10.1 [r=0.915, p0.5). DISCUSSION
Capnometry is a very useful monitoring procedure in critically ill children. Capnometry provides measurements of the CO2 concentration in the patient's respiratory gas. In addition to guiding ventilatory management, capnography assists the clinician by rapidly detecting critical events such as breathing circuit disconnection and kinking of the endotracheal tube. 1,2 However, in small infants, if the respiratory gas is sampled at the slip joint of the endotracheal tube, the capnograms are distorted, and the expired CO2 concentration may not accurately reHect PaC02 • In small infants, the respiratory gas must be sampled from the distal tip of the endotracheal tube just above the carina to obtain close agreement between r-C02 and PaC02 .3,4 In the present study, respiratory gas was sampled from the distal tip of the tube. Therefore, it was possible to detect the capnometric waveforms accurately in our small rabbits when using HFO~ The response time is an important factor in monitoring end-tidal CO2 in small infants. N ellcor engineers creatively reduced the volume of the CO2 analyzing chamber such that the N-l000 monitor is able to use a sampling How rate of 50 mVmin and still achieve an instrument response time similar to that of other capnometers that use a higher sampling How rate. Its response time is shorter than 0.1 s. This makes it possible to detect the capnographic waveforms accurately even in small infants. High-frequency oscillatory ventilation is used especially for small pretenn infants. With this mode, Capnometry during High-Frequency 08ci1atory YenIIIation (Nishimura et aJ)
there are no inspiratory and expiratory phases such as occur during conventional mechanical ventilation. Therefore, there is no such thing as uend-tidal CO2'' during HFO~ Some capnographs, including those obtained with the Nellcor monitor, cannot measure the concentration or partial pressure of CO2 unless cyclic Huctuations are present, as there are during conventional mechanical ventilation. In the present stud~ we placed a three-way stopcock between the respiratory circuit and the sampling tube of the monitor, and turned it at least ten times to sample the respiratory gas and room air alternately at an interval of 1 s. This technique made it possible to produce square waveforms of CO2 concentration; the machine also displayed the partial pressure of the CO2 • Ifthe machine will work when waveforms are linear, we can monitor the partial pressure of CO2 even during HFOV without manipulation of the stopcock. At the proximal opening of the endotracheal tube, the respiratory gas is mixed with fresh gas. Its mixing rate is considerably affected by setting conditions such as fresh gas How rate and stroke volume. Therefore, the CO2 concentration at this point usually does not approximate PaC02 •
In the present study, r-C02P showed no significant relation with PaC02 , because eliminated gas is diluted by fresh gas significantly at the airway opening. However, r-C0 2 d showed good correlation with PaCO2 , probably because the dilution of eliminated gas by fresh gas is not significant at the level of the distal tip of endotracheal tube. It is therefore concluded that capnometry using distal sampling of respiratory gas accurately reveals the relationship between PaC02 and r-C02d even in small subjects during HFO~ REFERENCES
1 Murray I~ Modell JH. Early detection of endotracheal tube accidents by monitoring of carbon dioxide concentration in respiratory gas. Anesthesiology 1983; 59:334-46 2 Linko Ie, Paloheimo M, Tammisto 1: Capnographs for detection of accidental oesophageal intubation. Acta Anaesthesiol Scand 1983; 27:199-202 3 Badgwell JM, McLeod ME, Lerman J, Creighton RE. End-tidal Peo. measurements sampled at the distal and proximal ends of the endotracheal tube in infants and children. Anesth Analg 1987; 66:959-64 4 Badgwell JM, Heavener JE, May WS, Goldthorn JF, Lerman J. End-tidal Peo. monitoring in infants and children ventilated with either a partial rebreathing or a non-breathing circuit. Anesthesiology 1987; 66:405-10
CHEST I 101 I 6 I JUNE. 1992
1883
_
Capnometry during High-Frequency Oscillatory Ventilation· . Masaji Nishimura, M.D.; Hideaki lmanaka, M.D.; Chikara Tashiro, Nobuyuki Taenaka, M.D.; and lkuto ¥os!liya, M.D.
~/.D.;
We used capnometry during high-frequency oscillatory ventilation (HFOV), and compared COl measurements at the distal and proximal ends of an endotracheal tube with arterial COl values. Ten white rabbits (mean weight, 2.00±O.2 [SD] kg) underwent tracheostomy under anesthesia with pentobarbital. The trachea was intubated with an endotracheal tube with a second lumen for sampling respiratory gas at the distal tip. Capnometry was performed through the lumen (COld) and the proximal end of the endotracheal tube (C0sP). The internal carotid artery was cannulated to sample blood for measuring arterial blood
gases. The differences between COld, COlp, and PaCOI were measured. Ooly the relation between COld and PaCO. was good (r=O.915). We concluded that capnometry can be used during HFOV to estimate PaCOs provided that respiratory gas is sampled from the distal tip of the endotracheal tube. (Chest 1992; 101:1681-83)
Capnometry and pulse oximetry are rapidly becom-I ing standard procedures in monitoring critically ill children. Capnometry provides a measurement of the CO2 concentration in the patient's respiratory gas. However, its use during high-frequency oscillatory ventilation (HFOV) has never been reported. In the present study, we used capnometry during HFO~ and determined the relationships between the CO2 concentration at the distal end of an endotracheal tube (C02d), the CO2 concentration at the proximal end of an endotracheal tube (C02P), and PaC02 •
METHODS
=
CO.d CO. concentration at the distal end of the endotracheal tube; COd» CO. concentration at the proximal end of the endotracheal tube; HFOV = high-frequency oscillatory ventilation; r-CO. CO. concentration of respiratory gas
*From the Department of Anesthesiology, Osaka Medical Center and Researcli Institute for Maternal and Child Health (Drs Nishimura and Tashiro), and the Intensive Care Unit, Osaka University Hospital (Drs Imanaka, Taenaka, and Yoshiya), Osaka, Japan. Reprint requests: Dr. Nishimura, De]J!Jrlment of Anesthesiology, Osaka Institute for ~Iatemal and Child Health, Osaka, japan 59002
(
= =
The experiment was performed on ten white rabbits weighing 1.16 to 2.21 kg (mean, 2.0±O.2 [SD] kg). The rabbits were anesthetized with intravenous pentobarbital, with supplemental doses as needed to suppress the corneal reHex. After tracheostonlY, the animals were intubated with a 4-mm-internal-diameter tracheal tube with a second lumen for sampling respiratory gas (Mallinckrodt, Glens Falls, NY). The distal tip of the tube was positioned just above the carina under x-ray guidance. The internal carotid artery was cannulated for samplin~ blood to analyze PaC02 • The analysis was performed with an ABL2 blood gas analyzer (Radiometer, Copenhagen). The rabbit's lungs were ventilated by using the HFOV mode of a IIummingbird B~tO 20N olechanical ventilator (~tera, lhkyo). The operating L'Onditions were set as follows: stroke volunle, 4, 6, 8, 10, and 12 ml; frequency, 15 cyclesls; inspiration-expiration ratio, 1: 1; mean airway pressure, 3 cm H 20; fresh gas flow rate, 1 or 2 Umin. At each setting, the PaC02 and the CO2 concentration of the respiratory gas (r-COJ were oleasured simultaneously. The respiratory gas was sampled through the second lumen of the tracheal tube from just above the carina (r-C02d) and from the proximal end
IHummingbird
"--------........
SMO 20N
Nellcor N-1000 Three - way stopcocks
FIGURE 1. Experimental schema. Rabbits underwent tracheostomy and intubation with an endotrclcheal tube with a se<..'Ond lumen for satnpling the respiratory gas. During "FOV with a llummingbird ventilator, capnography was perulrmed with a Nellcor monitor on the respiratory gas sampled froln the distal and proximal ends of the endotracheal tube.
CHEST / 101 / 6 / JUNE, 1992
1681
Endotrachea 1 tube
----.-..__~ Nellcor
N-l000
Three-way stopcock FICURE 2. Under HFOV the capnometric waveforms do not show fluctuation, and the capnometer cannot display partial pressures. To make this possible, a three-way stopcock was placed between the sampling tube and was connected directly to the endotracheal tube and the tube of the NeUcor monitor. The stopcock was turned several times to sample the respiratory gas and room air alternately. This technique produced square waveforms. of the tube (r-COdl). The r-C02 was measured with a Nellcor (Hayward, Calif) N-l000 monitor (Fig 1). The sampling rate was 50 mVmin; length of sampling line, 1.52 m; response time, 85 ms. During HFO~ r-C02 is constant, and the monitor, expecting the cyclic appearance and disappearance of CO2 in the respiratory gas, does not report numeric data but rather sounds an alarm if apnea is detected. To measure r-C02 during HFO~ a three-way stopcock was placed between the respiratory circuit and the sampling tube of the monitor. The cock was turned to sample the respiratory gas and the room air alternately (Fig 2). These arti6cial fluctuations in the CO2 concentration created the cyclic appearance and disappearan<'"e of CO2 ; thus, the monitor reported minimum and maximum CO2 concentrations with each respiratory cycle. We turned the cock to create arti6cial fluctuations of CO2 concentration at least ten times, and established the stability of the numeric values displayed by the monitor for end-tidal CO2 , RESULTS
In all rabbits, low stroke volume induced hypercarbia. The lower the stroke volume, the higher the rC02d and PaC02 • Because severe bradycardia or bigeminy pulse occurred at stroke volume settings under 4 or 6 ml, the experiment was discontinued at (mmHg)
70 60
50 ~ 40
0 <.;> 30 '-
20 10 0
0
60 (mmHg)
FIGURE 3. Relation between r-C02d and PaC02 as well as the line of identity.
1882
that stroke volume. The total number ofsample points was, therefore, 88 for PaC02 and r-C02d (PaC02 = 1.33 [r-C02d] -10.1 [r=0.915, p
Capnometry is a very useful monitoring procedure in critically ill children. Capnometry provides measurements of the CO2 concentration in the patient's respiratory gas. In addition to guiding ventilatory management, capnography assists the clinician by rapidly detecting critical events such as breathing circuit disconnection and kinking of the endotracheal tube. 1,2 However, in small infants, if the respiratory gas is sampled at the slip joint of the endotracheal tube, the capnograms are distorted, and the expired CO2 concentration may not accurately reHect PaC02 • In small infants, the respiratory gas must be sampled from the distal tip of the endotracheal tube just above the carina to obtain close agreement between r-C02 and PaC02 .3,4 In the present study, respiratory gas was sampled from the distal tip of the tube. Therefore, it was possible to detect the capnometric waveforms accurately in our small rabbits when using HFO~ The response time is an important factor in monitoring end-tidal CO2 in small infants. N ellcor engineers creatively reduced the volume of the CO2 analyzing chamber such that the N-l000 monitor is able to use a sampling How rate of 50 mVmin and still achieve an instrument response time similar to that of other capnometers that use a higher sampling How rate. Its response time is shorter than 0.1 s. This makes it possible to detect the capnographic waveforms accurately even in small infants. High-frequency oscillatory ventilation is used especially for small pretenn infants. With this mode, Capnometry during High-Frequency 08ci1atory YenIIIation (Nishimura et aJ)
there are no inspiratory and expiratory phases such as occur during conventional mechanical ventilation. Therefore, there is no such thing as uend-tidal CO2'' during HFO~ Some capnographs, including those obtained with the Nellcor monitor, cannot measure the concentration or partial pressure of CO2 unless cyclic Huctuations are present, as there are during conventional mechanical ventilation. In the present stud~ we placed a three-way stopcock between the respiratory circuit and the sampling tube of the monitor, and turned it at least ten times to sample the respiratory gas and room air alternately at an interval of 1 s. This technique made it possible to produce square waveforms of CO2 concentration; the machine also displayed the partial pressure of the CO2 • Ifthe machine will work when waveforms are linear, we can monitor the partial pressure of CO2 even during HFOV without manipulation of the stopcock. At the proximal opening of the endotracheal tube, the respiratory gas is mixed with fresh gas. Its mixing rate is considerably affected by setting conditions such as fresh gas How rate and stroke volume. Therefore, the CO2 concentration at this point usually does not approximate PaC02 •
In the present study, r-C02P showed no significant relation with PaC02 , because eliminated gas is diluted by fresh gas significantly at the airway opening. However, r-C0 2 d showed good correlation with PaCO2 , probably because the dilution of eliminated gas by fresh gas is not significant at the level of the distal tip of endotracheal tube. It is therefore concluded that capnometry using distal sampling of respiratory gas accurately reveals the relationship between PaC02 and r-C02d even in small subjects during HFO~ REFERENCES
1 Murray I~ Modell JH. Early detection of endotracheal tube accidents by monitoring of carbon dioxide concentration in respiratory gas. Anesthesiology 1983; 59:334-46 2 Linko Ie, Paloheimo M, Tammisto 1: Capnographs for detection of accidental oesophageal intubation. Acta Anaesthesiol Scand 1983; 27:199-202 3 Badgwell JM, McLeod ME, Lerman J, Creighton RE. End-tidal Peo. measurements sampled at the distal and proximal ends of the endotracheal tube in infants and children. Anesth Analg 1987; 66:959-64 4 Badgwell JM, Heavener JE, May WS, Goldthorn JF, Lerman J. End-tidal Peo. monitoring in infants and children ventilated with either a partial rebreathing or a non-breathing circuit. Anesthesiology 1987; 66:405-10
CHEST I 101 I 6 I JUNE. 1992
1883