Suppression of Spontaneous Breathing During High-Frequency Jet Ventilation Separate Effects of Lung Volume and Jet Frequency

Intensive Care Med (1987) 13:315-322 Intensive Care Medicine © Springer-Verlag 1987

Suppression of spontaneous breathing during high-frequency jet ventilation Separate effects of lung volume and jet frequency

A. J. van Vught 1'2, A. Versprille 1 and J. R. C. Jansen 1

1Pathophysiological Laboratory, Department of Pulmonary Diseases, Erasmus University, Rotterdam, and 2Pediatric Intensive Care Unit, University Children's Hospital "Het Wilhelmina Kinderziekenhuis",Utrecht, The Netherlands

Received: 15 October 1986; accepted: 2 December 1986

Abstract. The effect of ventilatory frequency of high- High-frequency ventilation can suppress central frequency jet ventilation (HFJV) from 1 to 5 Hz, apart respiratory activity in animals [2, 7, 10, 12, 15, 23, 24] from changes in thoracic volume, on spontaneous as well as in humans [6]. Although the exact mecha- breathing activity was studied in Yorkshire piglets nism is unknown the phenomenon has been attributed under pentobarbital anesthesia. The highest PaCO 2 at mainly to afferent inhibiting signals from pulmonary which the animals did not breathe against the ven- [2, 3, 10, 12, 15, 23], thoracic [10, 15, 23, 26] and tilator (apnea point) was established either by chang- laryngo-tracheal [8] mechano-receptors. ing minute volume of ventilation or by adding CO2 to Four different kinds of mechanical receptor activa- the respiratory gas. The higher the apnea point, the tion in the respiratory system can be postulated during higher the suppression of spontaneous breathing ac- high-frequency ventilation: (a) the static component tivity was assumed to be. If the apnea point was sear- of the stretch stimulus depending on the level of tissue ched for by changing minute volume a progressive in- stretch at end-expiration, (b) the change in stretch crease of suppression of spontaneous respiratory ac- stimulus depending on tidal volume, (c) the velocity of tivity was found at ventilatory rates of 3 Hz or more, stretch depending on the flow of insufflation and (d) concomitantly with a rise in end-expiratory pressure the repetitive activation of stretch receptors depending (PEE). In case the tidal volume was kept constant, in- on ventilatory frequency. crease of ventilatory rate resulted in a tremendous in- From previous experiments [24] we concluded that crease of lung volume, together with considerably suppression of breathing activity during high-fre- higher levels of PEE. When under these conditions the quency jet ventilation (HFJV) in piglets was positively apnea point was searched for by adding CO 2 to the related to jet frequency as well as to end-expiratory in- respiratory gas a much higher CO2-drive was needed tratracheal pressure, increasing lung volume and thus for spontaneous breathing and therefore a much the basic level of tissue stretch. Neither the change in stronger inhibition of spontaneous breathing was con- stretch nor the velocity of stretch of pulmonary or cluded. By placing the animals in a body box in which thoracic mechano-receptors seemed to be a mecha- pressure could be varied, thoracic volume could be nism for the suppression of spontaneous breathing. In kept constant during HFJV. When thoracic volume an attempt to differentiate between jet frequency and was kept constant in this way a constant tidal volume end-expiratory lung volume, we have examined the ef- at increasing jet frequencies resulted in only a slight fect of repetitive activation of stretch receptors on increase in suppression of spontaneous breathing. breathing activity during HFJV by studying the rela- We conclude that the increase in lung volume is a tionship between ventilatory frequency and the major factor in suppressing central respiratory activity PaCO2 necessary to provoke spontaneous breathing during HFJV. Jet frequency by itself might be an addi- movements, with and without concomitant changes in tional suppressive factor. Airway CO2 did not seem to lung volume. Because CO 2 is a potent drive for respir- have an important effect. atory activity, a higher PaCO 2 level without spontaneous breathing indicates a higher degree of suppression. Key words: PEEP - Lung stretch - Respiratory drive The highest PaCO 2 at which the animals did not - Carbon dioxide - Piglets breathe against the ventilator was denoted as the apnea point (PaCOz-apnea). In addition, based on ex- 316 A.J. van Vught et al.: HFJV and spontaneous breathing

perimental evidence [4, 16, 17], that airway CO2 tions in mixed expiratory air were measured by a mass could act upon pulmonary stretch receptors we have spectrometer (Perkin-Elmer MGAll00) from the tried to analyse the effect of mixed expiratory CO 2 mixing box. PO2, PCO 2 and acid-base variables in (FECO2) in suppressing respiratory activity during blood were determined by means of an automatic HFJV. blood gas analyser (Radiometer ABL3) and oxygen saturation and hemoglobin with an oxymeter (Radiometer OSM2). Methods In six piglets changes in thoracic volume were estimated from changes in resistance of a mercury strain The methods have been previously described [24]. gauge. Within certain limits the electrical resistance of Therefore, only the essentials of the technique and its a mercury strain gauge has a linear relationship to its modifications will be given here. The experimental set- stretch; more stretch will give a higher resistance when up is shown in Figure 1. the mercury filled column becomes longer and thinner Yorkshire piglets (5-7 weeks old, 7-10 kg) were [5, 21]. The frequency response is linear up to 30 Hz anesthetized with pentobarbital sodium (30 mg.kg -1 [14]. The resistance range of the mercury strain gauge i.p.). Anesthesia was maintained by a continuous infu- in our experiments was between 700 and 800 m~ with sion of pentobarbital (7.5-10 mg. kg- 1. h-l), suffi- a temperature dependence of 1 mQ-°C-1 which was cient to eliminate pain reflexes, but allowing the neglected. animals to breathe spontaneously. Central tempera- The strain gauge was calibrated and checked on ture was kept at approximately 39 °C. linearity at the end of the experiment by stepwise in- After tracheostomy and connection via a Y-can- flation of air with a syringe after paralysing the nula to inspiratory and expiratory tubes, catheters animals with d-tubocurarine hydrochloride were inserted into the right common carotid artery, the (0.1mg.kg-1). Changes in thoracic volume of superior caval vein and the pulmonary artery for 4-8 ml could so be detected. No absolute values of blood pressure monitoring, blood sampling and infu- lung volume were measured. However, the strain gauge sions. During venous cannulation the animals were was primarily used as a zero-method, which means ventilated in order to prevent air embolism. Tracheal that its length and the corresponding thoracic volume pressure was measured deep in the trachea with a fluid were restored to its initial values. For such a method filled catheter provided with side holes at the tip. linearity is not relevant and frequency response less Blood pressures and tracheal pressure were measured critical. with Statham transducers P23De. Blood pressures Experimental procedures were measured relative to atmospheric pressure at manubrium level. For intratracheal measurement end- The animals were ventilated with frequencies from 1 to expiratory pressure during spontaneous breathing was 5 Hz using a high-frequency jet ventilator. Inspiratory taken as zero-reference. Heparin was administered in- time was kept constant at 0.1 s. The inspiratory gas termittently (250 IU" kg-l-h- l). The electrocardio- contained 40% oxygen, the fraction of inspired carbon gram was monitored continuously. dioxide could be varied from 0-0.20 as indicated in Ventilatory (tidal) volume was calculated by in- Figure 1. Spontaneous breathing was allowed via an tegration of mean airway flow, measured with a Fleish inlet valve in the inspiratory tube. During HFJV this pneumotachograph (type 0 Godart) behind a mixing valve was clamped off in order to prevent entrainment box in the expiratory tube. CO 2 and 0 2 concentra- of air. PaCO2 could be varied either by changing the

airflow

O2-N 2 CL,MP ./I ~ 40-60°~ Fig. 1. Experimental set-up. During ventilation with ventilator the CO2-O2-N2 / I I the jet 20-40-40% expiratory valve is closed at each insufflation. Changes in thoracic volume were measured with a mer- 1. MIXING BOX MASS SPECTROMETER cury strain gauge. Thoracic volume 2. FLOW METER was varied by varying pressure in the EXP. VALVE body box, using a water seal and a r::£7 ,uoE pbox continuous air flow, in order to eliminate effects of slight air leakages TRACHEAL PRESSURE under pressure A. J. van Vught et al.: HFJV and spontaneous breathing 317

minute volume or by adding CO 2 to the inspiratory PaCO2-apnea at 1 Hz (41.6___3.7 mmHg, mean_+ 1 SD) gas. Spontaneous inspiratory activity during HFJV as a reference point. These animals breathed was detected by concomitant decreases in intra- spontaneously at a PaCO 2 of 44.2_+3.2 mmHg (mean tracheal, central venous and pulmonary artery pres- _+ 1 SD). PaCO2-apnea, searched for by changing sures. minute volume of ventilation, increased with increas- In a first group of five piglets (group I) the apnea ing ventilatory frequency. As we reported previously point for five successive jet frequencies was searched [24] tidal volume under these conditions has been for by changing tidal volume. Thereafter, in a second decreased to find the apnea point. Concomitantly series of observations in the same group, tidal volume end-expiratory intratracheal pressure rose (Fig. 2b). (VT) was kept constant for all frequencies and was However, when tidal volume was kept constant at a equal to V T at the apnea point at 1 Hz. This implied level just sufficient to get apnea at 1 Hz a linear increase of minute volume with frequency. (4.9_+0.5 ml.kg -1, mean -+ 1 SD) and the apnea point CO2 was added to the inspiratory gas at the higher was searched for by adding CO 2 to the respiratory gas frequencies until the apnea point was found. Keeping higher apnea points were found, indicating a much V r constant led to a tremendous increase in lung stronger inhibition of spontaneous breathing activity. volume at higher ventilatory rates. At higher frequencies PaCO 2 levels of more than In a second group of six piglets (group II), ven- twice the PaCOz-apnea at 1 Hz were insufficient to tilated with a constant V T and varying frequencies, elicit breathing activity. At the same time end-ex- thoracic volume was kept constant during the search piratory tracheal pressures reached much higher levels. for the apnea point with CO2, using a body box in In all animals of group I there was a positive correla- which the pressure could be varied. By varying this tion between end-expiratory intratracheal pressure and pressure thoracic volume could be varied. In these PaCOz-apnea in both series. In the series with a con- animals jet frequencies from 2 to 5 Hz were imposed stant tidal volume at increasing frequencies this rela- in a random order. In between each observation the tion was within the range of the first series with apnea point at 1 Hz was determined to check stability. adapted tidal volumes up till an end-expiratory pres- The apnea point at each frequency was measured at sure of about 10 cmH20 in all animals except one. At the same end-expiratory strain gauge resistance (zero- higher PEEP values, corresponding with higher ven- method), i.e. the same end-expiratory thoracic volume tilatory rates, there was more suppression in the sec- as at the preceding 1 Hz. ond series of experiments than in the first series (Fig. In three of these animals a second series of ex- 2c). The ventilatory frequency and the end-expiratory periments was performed, identical to the second pressure at which even a very high PaCO2 could not series in the first five animals: constant VT, successive elicit breathing movements was rather different for increase of jet frequency and thus of lung volume and different animals. Therefore data of this series could search for the apnea point using CO2. not be averaged and all individual results obtained All animals were autopsied at the end of the ex- under these conditions are presented (Fig. 2). periment to check the position of the catheters and the degree of atelectasis. Catheter position was correct in all animals. They all had a variable degree of atelec- HFJV at constant thoracic volume tasis. Results of an individual experiment out of group II, in Statistical analysis which thoracic volume was kept equal to the level of 1 Hz at apnea and the apnea point was searched for by Differences were tested by the t-test for paired and un- adding CO 2 to the respiratory gas, are presented in paired small samples [1]. The relation between Figure 3. Subsequently the apnea points were deter- PaCO2-apnea and successive frequencies of ventila- mined, allowing the thorax to expand freely. When tion was further analysed using a test for trend [25]. thoracic expansion was prevented by increasing pressure in the body box, the apnea point at a ventilatory Results rate of 3 Hz or more was much lower and thus at this level suppression of spontaneous breathing was considerably less. The apnea points at 1 Hz revealed a HFJV and PEEP stable condition throughout the experiment. However, The relationship between PaCO2-apnea and jet fre- a gradual diminution in thoracic volume at the apnea quency in the five piglets of group I is presented in Fig- point occurred. ure 2a. For mutual comparison of all individual data Figure 4 summarizes the results of the six piglets of PaCOz-apnea was expressed as a fraction of group II in which the apnea point was established in 318 A.J. van Vught et al.: HFJV and spontaneous breathing

Pa C02- apnea

Pa C02-apnea at 1Hz 2.5.

PEE cmH20

2.0.

1.5-

1.0- P ns (005 (oos (002

, l i i i i i i i 1 2 3 4 5Hz 1 2 3 4 5Hz frequency of ventilation b frequency of ventilation

Pa CO 2 - apnea Pa C02-apnea at 1Hz 2.5-

Fig. 2 a-c. Apnea point as a fraction of the apnea point at 1 Hz (a) and end-expiratory intratracheal pressure, PEE (h), both as func- tions of ventilatory rate in five piglets. From a and b a relationship has been constructed between the apnea point and PEE (c). Thick 1,5 lines summarize the results (mean___ 1 SD), when apnea point was searched for by changing minute volume of ventilation. Thin lines represent individual results in these five animals when tidal volume was kept constant and apnea point was estimated by adding CO 2 to the inspiratory gas. Arrows indicate an apnea point at a higher 1.0- PaCO2 than the measured value, so in reality the last segment of the concerning curve lies more to the left. Circles indicate overlapp-

1~0 115 210 ing points, p values indicate the level of probability of difference C PEE cm H20 from 1 Hz as calculated by the paired t-test, ns: not significant

the same way. These animals breathed spontaneously found with increasing frequencies. This rise was signif- at a PaCO 2 of 40.1 +2.7 mmHg (mean+_ 1 SD). Apnea icantly less than in the animals of group I (17 < 0.02 at at 1 Hz occurred with a tidal volume of 5 Hz). The apnea point increased considerably when 4.8+_0.8ml.kg -1 (mean+_1 SD). When thoracic vol- the thorax was allowed to expand which was demon- ume was kept constant a slight but significant rise strated in three animals. These individual results were (p < 0.01, test for trend) in the apnea point and there- identical to those of the five animals from Figure 2a fore in suppression of spontaneous breathing was with the same procedure. A. J. van Vught et al.: HFJV and spontaneous breathing 319

Changes in thoracic volume in ml. 100 -

Pa CO2- apnea mmHg 75-

90- T 50-

80-

25- 70- D

60- O- 1 • • 2e Fig. 3 a and b. A representative individual ex- ample of the apnea point in absolute values 50- 3e -25- (a) and the change in thoracic volume at the 4e • apnea point (b) as a function of ventilatory 40- rate. Tidal volume was constant for each fre- -50 - quency. Apnea point was searched for by ad- 30- 7 ding CO 2 to the inspiratory gas. Thick lines represent results when thoracic volume was 6a 20- -75- kept constant by changing pressure in the body box. Thin lines represent results with pressure in the body box at atmospheric level. 10- Numbers indicate the sequence in which the values of thoracic volume at 1 Hz were ob- i i i i i I I I I I tained. Note the decrease in thoracic volume, 1 2 3 4 5 Hz 1 2 3 4 5 Hz corresponding with a small decrease in apnea frequency of ventilation b frequency of ventilation point. Arrow: see Figure 2a

Pa CO 2- apnea

Pa CO2-apnea at 1Hz

F~. C02

F~* CO2 at 1 Hz Fig. 4. Apnea point as a function of jet fre- I ! 4.0. quency in six piglets ventilated with a constant tidal volume for each frequency. The thick line summarizes the results (mean _+ 1 SD), when thoracic volume was kept constant, p values, relative to the reference point at 1 Hz, were calculated by t-test for paired 3.0- samples. Thin lines represent individual results in 3 piglets out of this group when thorax was allowed to expand freely. Arrows: see Figure 2a

Fig. 5. Ratio of mixed expiratory fraction of 2.0- CO 2 (Ft~CO2) and FI~CO 2 at 1 Hz as point of reference plotted as a function of jet frequency. The thick line represents the combined results (mean_+ 1 SD), when the apnea point was searched for by changing tidal volume of ventilation in 5 piglets of group I. The thin

1.0- lines represent the individual results in these five animals when tidal volume was kept con- P P P p stant and PaCO2-apnea was estimated by ad- <0.05 (0.05 (O.O1 <0.01 ding CO 2 to the respiratory gas. The broken line summarizes the results (mean_+ 1 SD) in the 6 piglets of group II in which this relationship was found with a constant tidal I I t I i i i i i i volume, a constant end-expiratory thoracic 1 2 3 4 5 Hz 1 2 3 4 5 Hz volume, whereas PaCO2-apnea was searched frequency of ventilation frequency of ventilation for by adding CO 2 to respiratory gas 320 A.J. van Vught et al.: HFJV and spontaneous breathing

C02 in the airways slowly developing atelectasis, co-incided with a small decrease in the apnea point (Fig. 3a). Therefore not The influence of airway CO 2 on the inhibitory effect the thoracic volume at the beginning of the experiment of HFJV on respiration was analysed from Figure 5. was taken as a reference but thoracic volume at apnea This figure shows the ratio of mixed expiratory CO2 at 1 Hz, preceding the jet frequency which was tested. fraction (F~CO2) and FzCO2 at 1 Hz versus jet fre- In addition the jet frequencies from 2 to 5 Hz were im- quency in the group I animals under the two condi- posed in a random order to eliminate a systematic ef- tions of constant and adapted tidal volume. F~CO z fect on our results by this slowly progressive atelec- was measured at the apnea point. When tasis. PaCO2-apnea was searched for by changing minute When end-expiratory thoracic volume was kept volume of ventilation apnea at higher frequencies oc- constant a small but increasing suppression with ven- curred at lower F~CO 2. However, when the tilatory frequency could be observed. This suppression PaCO2-apnea was estimated by adding CO2 to the in- seemed to be due to the effect of the increasing fre- spiratory gas apnea occurred at increased levels of quency with which a tidal volume of about 40 ml is in- FzCO2. When in the group II animals thoracic sufflated, or in terms of mechano-receptor activation, volume was kept constant, F~CO z was in between the the frequency at which receptors are stretched from a two ranges and did not change significantly except for constant basic level of stretch. a small increase from 1 to 2 Hz. Most studies on the suppressive effect of high-frequency ventilation on respiration were conducted dur- Discussion ing high-frequency oscillation with only one frequency, either 15 Hz [2, 10, 23] or 25 Hz [15]. Summarizing We have previously shown [24] that HFJV suppresses the literature on this topic, there is increasing evidence spontaneous breathing activity in piglets. The amount that suppression of spontaneous breathing activity of suppression increased at higher jet frequencies. In- during HFO is due to an increasing lung volume as creasing ventilatory frequencies entails shortening of well as to the rapid phasic effect of oscillations on expiratory time and elevation of end-expiratory mechano-receptors in lung and thoracic wall. How- pressure and lung volume. Increase in lung volume was ever, only in the study of Banzett et al. [2] lung volume an important factor in suppression of breathing activi- was actually measured and its effects separated from ty. Above a critical lung volume spontaneous breath- those of frequency. In the other studies [10, 15, 23] on- ing was completely abolished. ly mean airway pressure was kept constant and in- The present study confirmed these results. Ventila- creases in lung volume during oscillation could be in- tion with a constant tidal volume but increasing fre- volved in the induction of apnea at least partially [3]. quencies elicited elevated end-expiratory pressures and Up till now the studies concerning high-frequency an increase in the arterial PCO2 necessary to provoke jet ventilation are not conclusive in this way. Jonzon spontaneous respiratory movements. This positive cor- [12] observed disappearance of phrenic nerve activity relation between end-expiratory intratracheal pressure in cats when the ventilatory rate reached 56 per and apnea point was found in the series with constant minute. But in this very first study on the suppressive tidal volume and freely expandable thorax, where the effect of high-frequency ventilation on central apnea point was searched for by addition of CO2 to respiratory activity neither blood gases nor lung: the respiratory gas as well as in the series with a tidal volume were reported. Chakrabarti et al. [7] could volume just adapted to reach the apnea point. How- reverse suppression of breathing at each jet frequency ever, the same end-expiratory pressure appeared to in- just by diminishing end-expiratory intratracheal duce more suppression when the animals were ven- pressure. tilated with a constant tidal volume and freely expan- Our study indicates that also during HFJV lung dable thorax, especially for PEEP values above 10 cm volume as well as frequency are factors involved in H20 and thus for the corresponding higher frequen- suppressing respiratory activity. However, the sup- cies (Fig. 2c). This increased suppression may reflect pressive effect of increasing frequencies was only an increased lung volume at the same end-expiratory small. A restriction to this latter observation could be intratracheal pressure under these conditions. It is well that ventilation with a constant tidal volume at rising recognized that during high-frequency ventilation air- frequencies gives a small rise in mean lung volume by way pressure may underestimate alveolar pressure and itself when end-expiratory lung volume is kept cons- that changes in intratracheal pressure not simply in- tant. At a tidal volume of 40 ml and a constant indicate proportional changes in lung volume [3, 20, 22]. spiratory time of 0.1 s this increase in mean lung Diminution of thoracic volume at 1 Hz (Fig. 3b) in volume from 1 to 5 Hz could maximally be 16ml. the course of the experiment, presumably due to a Presumably this increase is a negligible change. A. J. van Vught et al.: HFJV and spontaneous breathing 321

Measurements of end-expiratory thoracic volume pression of respiration should be even stronger under were taken during apnea, undisturbed by breathing isocarbic conditions. The effect of jet frequency on movements. Under this condition we suppose the respiratory activity in our second group of ex- respiratory system has one degree of freedom, that is periments in which thoracic volume was kept constant to say we suppose an equal proportional contribution was not influenced by airway CO2, because mixed ex- of rib cage and abdomen to changes in lung volume piratory CO2 fraction, apart from a small increase [13]. This supposition is sustained by the work of from 1 Hz to 2 Hz, did not change. Rouby et al. [19]. From our study we conclude that suppression of The frequency response of the mercury strain spontaneous breathing activity during high-frequency gauge was sufficient to measure thoracic volume dur- jet ventilation in the most used frequency range is ing HFJV up to 5 Hz adequately [14]. On the other mainly due to increase in lung volume. Provided that hand, more relevant are the combined properties of lung volume is sufficiently controlled, there is no strain gauge and thoracic wall [19]. We are not inform- ground for the concern of Drazen et al. [9], that the ed about the frequency response of chest wall and ab- additional suppressive effect of ventilatory frequency domen in piglets. In our experiments we had no in- could be detrimental if high-frequency ventilation is dication that a damping occurred. But should this used as an adjunctive means of ventilatory support in happen, it would have lead to overestimation of end- spontaneous breathing patients. However, respiratory expiratory thoracic volume and therefore to overcor- activity is the result of a complicated multi-input rection. Such an overcorrection would have influenced system in which mechanical, chemical and higher negatively the effects of ventilatory frequencies. Thus cerebral stimuli are integrated in an only partially it would strengthen rather than jeopardize our conclu- understood fashion. High-frequency ventilation acts sion, that increasing frequencies suppress spontaneous upon this system and the effects are also critically breathing during HFJV apart from changes in dependent on the functional state of the central thoracic volume. respiratory controller as it will be determined by sleep Although most attention has been paid to the ef- state [11], arousal and level of anaesthesia or sedation fects of mechanical stimuli of high-frequency ventila- [2, 24]. tion in suppressing central respiratory activity non- mechanical effects must be considered too. Philipson Acknowledgement. The authors would like to thank A. Drop for his technical assistance during the experiments. et al. [18] found respiratory rhythm generation to be critically dependent on CO 2 elimination. In awake sheep normocapnic apnea was achieved when CO 2 References elimination by a carbon dioxide membrane lung equalled CO2 production. In the same way HFO 1. Bancroft H (1966) Introduction to biostatistics. Harper and mediated CO2 elimination could stop breathing Row, New York 2. Banzett R, Lehr J, Geffroy B (1983) High-frequency ventilation movements. All the studies on HFO reported nor- lengthens expiration in the anesthetized dog. J Appl Physiol mocapnic apnea. Only our study indicates that HFJV 55:329 can give rise to hypercapnic apnea. 3. Barnas GM, Banzett RB, Reid MB, Lehr J (1986) Pulmonary Furthermore evidence exists that pulmonary afferent activity during high-frequency ventilation at constant stretch receptor activity is influenced by airway CO2 mean lung volume. J Appl Physiol 61:192 4. Bartoli A, Cross BA, Guz A, Jain SK, Noble MIM, Trenchard [4, 16, 17]. Increases in airway CO 2 could inhibit DW (1974) The effect of carbon dioxide in the airways and pulmonary stretch receptor activity, whereas decreases alveoli on ventilation; a vagal reflex studied in the dog. J in airway CO 2 could facilitate this activity. But ac- Physiol 240:91 cording to Pack [17] CO2 sensitivity is unlikely to be 5. Brakkee AJM, Vendrik AJH (1966) Strain-gauge plethysmogra- phy; theoretical and practical notes on a new design. J Appl of physiological importance. Physiol 21:701 In preliminary experiments we found a decrease in 6. Butler WJ, Bohn DJ, Bryan AC, Froese AB (1980) Ventilation mixed expiratory PCO 2 as respiratory frequency in- by high-frequency airway oscillation in humans. Anesth Analg creased (see also Figs. 2a and 5). So a fall in airway 59:577 CO 2 might contribute to the suppressive effect of 7. Chakrabarti MK, Whitwam JG (1983) Evaluation of a new valveless all purpose ventilator: effect of ventilating frequency, high-frequency ventilation by deblocking stretch re- PEEP, PaCO 2 and PaO2 on phrenic nerve activity. In: Scheck ceptor activity. In our experiments with constant tidal PA, SjOstrand UH, Smith RB (eds) Perspectives in high fre- volume and freely expandable thorax Ft2CO 2 increas- quency ventilation. Martinus Nijhoff, The Hague, p 140 ed by increasing FICO 2. Thus if CO 2 in the airways 8. DeWeeseEL, Sullivan TY, Yu P (1985) Ventilatory response to high-frequency airway oscillation in humans. J Appl Physiol acts negatively on the suppression of breathing, also in 58:1099 piglets, this effect was overruled by the static stretch 9. Drazen JM, Kamm RD, Slutsky AS (1984) High-frequency ven- stimulus of lung distension and the tremendous sup- tilation. Physiol Rev 64:505 322 A.J. van Vught et al.: HFJV and spontaneous breathing

10. England SJ, Onayemi A, Bryan AC (1984) Neuromuscular 20. Saari AF, Rossing ThH, Solway J, Drazen JM (1984) Lung in- blockade enhances phrenic nerve activity during high-frequency flation during high-frequency ventilation. Am Rev Respir Dis ventilation. J Appl Physiol 56:31 129:333 11. England SJ, Sullivan C, Bowes G, Onayemi A, Bryan AC (1985) 21. Sigdell J (1969) A critical review of the theory of the mercury State related incidence of spontaneous breathing during high strain-gauge plethysmograph. Med Biol Eng 7:365 frequency ventilation. Respir Physiol 60:357 22. Simon BA, Weinmann GG, Mitzner W (1984) Mean airway 12. Jonzon A (1977) Phrenic and vagal nerve activities during spon- pressure and alveolar pressure during high-frequency ventila- taneous respiration and positive-pressure ventilation. Acta tion. J Appl Physiol 57:1069 Anaesth Scand [Suppl] 64:29 23. Thompson WK, Marchak BE, Bryan AC, Froese AB (1981) 13. Konno K, Mead J (1967) Measurement of the separate volume Vagotomy reverses apnea induced by high-frequency oscillatory changes of rib cage and abdomen during breathing. J Appl ventilation. J Appl Physiol 51:1484 Physiol 22:407 24. Van Vught A J, Versprille A, Jansen JRC (1986) Suppression of 14. Lawton RW, Collins CC (1959) Calibration of an aortic cir- spontaneous breathing during high-frequency jet ventilation: cumference gauge. J Appl Physiol 14:465 influence of dynamic changes and static levels of lung stretch. 15. Man GCW, Man SFP, Kappagoda CT (1983) Effects of high- Intensive Care Med 12:26 frequency oscillatory ventilation on vagal and phrenic nerve ac- 25. Wallenstein S, Zucker CL, Fleiss JL (1980) Some statistical tivities. J Appl Physiol 54:502 methods useful in circulation research. Circ Res 47:1 16. Mustafa MEKY, Purves MJ (1972) The effect of CO 2 upon dis- 26. Ward HE, Armengol J, Jones RL (1985) Ventilation by external charge from slowly adapting stretch receptors in the lungs of high-frequency oscillation in cats. J Appl Physiol 58:1390 rabbits. Respir Physiol 16:197 17. Pack AI (1981) Sensory inputs to the medulla. Ann Rev Physiol 43:73 18. Phillipson EA, Duffin J, Cooper JD (1981) Critical dependence Dr. A. J. van Vught of respiratory rhythmicity on metabolic CO 2 load. J Appl Pathophysiological Laboratory Physiol 50:45 Department of Pulmonary Diseases 19. Rouby J J, Simonneau G, Benhamou D, Sartene R, Sardnal F, Erasmus University Deriaz H, Duroux P, Viars P (1985) Factors influencing P. O. Box 1738 pulmonary volumes and CO 2 elimination during high-fre- NL-3000 DR Rotterdam quency jet ventilation. Anesthesiology 63:473 The Netherlands