Time Course Evolution of Ventilatory Responses to ... - ATS Journals

14 downloads 1211 Views 97KB Size Report
Pressure support ventilation (PSV) is commonly used to un- load the respiratory ... time course of inspiratory muscle inhibition induced by un- loading inspiratory ...
Time Course Evolution of Ventilatory Responses to Inspiratory Unloading in Patients JEAN PAUL VIALE, SERGE DUPERRET, PHILIPPE MAHUL, BERTRAND DELAFOSSE, CLAUDE DELPUECH, DIETER WEISMANN, and GUY ANNAT Department of Anesthesia and Intensive Care, Lyon; Department of Anesthesia and Intensive Care, Saint Etienne; INSERM U280, Lyon, France; and Drägerwerk AG, Lübeck, Germany

Inspiratory muscle unloading decreases ventilatory drive. In this study, we examined the time course of this effect in patients with chronic obstructive pulmonary disease receiving two modes of ventilatory support: pressure support ventilation (PSV), during which each cycle was assisted, and biphasic positive airway pressure (BIPAP), set up in such a manner that one spontaneous breath took place between two consecutive pressure-assisted breaths. The first breath following the switch from spontaneous breathing to PSV was associated with an increase in tidal volume (VT) and a drop in mean transdiaphragmatic pressure (mean Pdi) and inspiratory work (WI) performed per liter but with unchanged values of esophageal occlusion pressure at 100 ms (Pes 0.1), diaphragmatic electrical activity (EMGdi), and WI performed by breath. The same phenomena were observed for the assisted breath of BIPAP as compared with the preceding spontaneous breath. During the subsequent breaths of PSV, Pes 0.1, EMGdi, and WI performed per breath decreased progressively up to the sixth to eighth breaths, and VT returned to pre-PSV values. We conclude that in patients with chronic obstructive pulmonary disease the decrease in ventilatory drive associated with PSV takes place from the first breath onwards but requires six to eight breaths to be fully achieved. During BIPAP, as a consequence of the kinetics of the PSV-induced downregulation of ventilatory drive, assisted breaths following spontaneous breaths are characterized by an enhanced inspiratory efficiency. Viale JP, Duperret S, Mahul P, Delafosse B, Delpuech C, Weismann D, Annat G. Time course evolution of ventilatory responses to inspiratory unloading in patients. AM J RESPIR CRIT CARE MED 1998;157:428–434.

Pressure support ventilation (PSV) is commonly used to unload the respiratory muscles during mechanical ventilation while allowing the patient to keep control of his/her ventilatory pattern. Several studies have shown that PSV is associated with a reduction in the patients’ ventilatory drive and work of breathing, the magnitude of which is dependent on the level of pressure applied (1–3). Thus, there is general agreement concerning the pressure dependency of the ventilatory response on pressure support. However, less is known about its time dependency, i.e., the time course of the inspiratory inhibition upon respiratory muscle unloading. A knowledge of this time dependency could be particularly helpful in patients undergoing types of partial mechanical ventilation in which spontaneous breaths are intermixed with assisted breaths. The objective of the present study was, first, to examine the time course of inspiratory muscle inhibition induced by unloading inspiratory muscles in tracheally intubated patients

(Received in original form January 21, 1997 and in revised form September 10, 1997) Supported in part by a grant from the Ministère de l’Education Nationale, de l’Enseignement et de la Recherche, EA 1896. Correspondence and requests for reprints should be addressed to J. P. Viale, Hôpital E. Herriot, Pavillon D, 5 Place d’Arsonval, 69437 Lyon, France. E-mail: jp-viale@ cismsun.univ-lyon1.fr Am J Respir Crit Care Med

Vol 157. pp 428–434, 1998

undergoing pressure support ventilation, and second, to evaluate the effects on respiratory motor output of introducing spontaneous breaths between pressure-assisted breaths. Such a partial ventilatory support, characterized by pressure-assisted breaths intermixed with spontaneous breaths, represents an intermittent ventilatory support as opposed to the permanent ventilatory support provided by the PSV. This intermittent support can be achieved in clinical practice by a new ventilatory mode called biphasic positive airway pressure (BIPAP).

METHODS Patients After institutional approval and informed consent were obtained, eight tracheally intubated patients were enrolled in the study. All of them were admitted to the intensive care unit for acute respiratory failure of chronic obstructive pulmonary disease (COPD). Before the study, each patient underwent mechanical ventilation on a controlled mode for several days. At the time of the study, all patients were receiving partial ventilatory support (PSV or BIPAP; Dräger Evita II) and were able to sustain spontaneous breathing for at least 5 min without clinical signs of respiratory distress. During PSV, each breath is assisted by an inspiratory pressure generated by the respirator at a preset value. This is synchronized with the patient’s efforts to breathe. The assisted breath is initiated when the spontaneous inspiratory flow reaches a preset threshold value of 3 L/min (flow-triggering level). The insufflation is stopped when the instantaneous flow is , 25% of the maximal value of the inspiratory flow. During BIPAP ventilation, two types of airway pressure are generated, the levels and durations of

429

Viale, Duperret, Mahul, et al.: Kinetics of Respiratory Unloading which may be adjusted independently. Unrestricted spontaneous breathing is possible for the patient whatever the level of airway pressure (4). However, as the duration of the level of high airway pressure is set close to that observed during PSV, spontaneous breaths are observed only during low airway pressure levels. In this case, BIPAP may be described as an intermittent pressure support ventilation, which allows spontaneous breaths among pressure-assisted breaths. Moreover, as with PSV, changes in airway pressures are synchronized with the patient’s own breathing efforts using the same flow-triggering system. Throughout the study, the applied positive end-expiratory pressure was set by the attending physician and was not modified for the study. All patients were in a semi-recumbent position during measurements. Only two of them had an indwelling arterial line at the time of the study, allowing repeated samples to be taken for gas analysis.

Protocol Three sets of measurements were taken for each patient. Phase 1: Mechanics and transdiaphragmatic pressure measurement. In order to measure the respiratory mechanics, all patients underwent a period of passive mechanically controlled ventilation. This was achieved by increasing the respiratory rate of the controlled ventilation period above the baseline value observed during partial mechanical ventilation. At the end of the study, while patients were spontaneously breathing, an occlusion of the airway at the disconnected Y piece level was performed for at least 10 breaths in order to measure maximal transdiaphragmatic pressure (Pdimax) (5). Phase 2: Time course of inspiratory inhibition. To investigate the time course of inspiratory muscle inhibition induced by unloading the respiratory muscles, the transient period following the onset of PSV was studied as follows: measurements were performed while patients were spontaneously breathing for at least 2 min and then after the onset of PSV, during the first 15 breaths (see Figure 1). Although several challenges were performed in order to obtain reliable tracings, only one was included in the study. The pressure level was the one chosen by the attending physician (see Table 1). In order to be sure that patients were not assisted beyond the point of total respiratory muscle unloading, an esophageal pressure recording (see below) was used to confirm that the depression was sustained after the triggering of the ventilator (3). Phase 3: Intermittent ventilatory support. To evaluate the role of intermixing spontaneous breaths with assisted breaths on the respiratory output, we studied the same patients undergoing BIPAP ventilation at a particular setting. In our study, this ventilatory mode was set in such a manner that one spontaneous breath took place between two pressure-assisted breaths (see Figure 6). The pressure level of the assisted breath was the same as that used during the PSV.

Measurements Esophageal (Pes) and gastric (Pga) pressures were measured with a micro pressure transducer-tipped catheter (MTC P3FC 3F; Dräger

ME, Best, The Netherlands). The airway pressure was recorded 1 cm from the oral end of the endotracheal tube by using another external transducer (Sims, Kirchseeon, Germany). The validity of Pes measurements was assessed by performing “occlusion tests” as proposed by Baydur and colleagues (6). The gas flow was measured by using a pneumotachograph (Fleish No. 2) connected to a differential pressure transducer (Validyne MP45; 6 2 cm H2O). The diaphragmatic electromyogram (EMGdi) was recorded by an esophageal probe positioned at the level of the gastro-esophageal junction (MCT-Cond.M 8F; Dräger ME). This probe consisted of eight steel rings at a distance of 9 mm from each other, two adjacent electrodes forming a pair. With the esophageal probe in place, the optimal pair giving the best-quality signal was chosen. The raw EMG was amplified and then bandpassfiltered between 20 and 500 Hz. All pressure, flow, and EMGdi signals were digitized by an analog-to-digital converter with a 16-bit resolution at a sampling frequency of 1,000 Hz (MP100 Biopac System, Inc., Santa Barbara, CA).

Signal Analysis and Calculation Respiratory mechanics were measured according to Rossi and associates (7) during constant flow inflation, from signals recorded during the period of passive controlled mechanical ventilation. Only the linear portion of the airway pressure- or esophageal pressure–volume curve was taken into account for the computation. This allowed us to calculate the compliance and resistance of the respiratory system as a whole as well as the compliance of the chest wall. Transdiaphragmatic pressure (Pdi) was obtained by subtracting the Pes from the Pga. The peak Pdi was the maximum value of transdiaphragmatic pressure generated during inspiration, while the integration of the Pdi over inspiratory time divided by the inspiratory time, determined on the flow curve, gave the mean Pdi. The digital integration of flow over time provided the volume data. The duration of inspiration, the total duration of breaths, and thus the duration of the inspiratory time relative to the total duration (TI/Ttot) were derived from the flow curves. From these data, we computed the tension-time index of the diaphragm while the patients were on spontaneous breathing (8). During PSV or BIPAP, the inspiratory work of breathing (WI) of each breath was calculated using the area delimited by the Pes-volume curve during inspiration and the Pes-volume curve derived from the chest wall compliance computation. In order to take into account the presence of an intrinsic positive end-expiratory pressure (PEEPi), the beginning of inspiration, and thus of the integration of the Pes-V curve, was considered as the onset of a rapid decrease of Pes. Therefore, the inspiratory threshold load due to the presence of PEEPi was taken into account in the calculation of the work of breathing (9). The PEEPi itself was measured as the difference in Pes between the beginning of inspiration and the zero point flow. On the same curve, the occlusion pressure at 100 ms (Pes 0.1) was determined as the difference between Pes at the beginning of inspiration and that obtained 100 ms later (10). For each breath, WI was expressed as the WI performed per breath or

TABLE 1 DEMOGRAPHIC DATA

Patient No.

Age (yr)

Weight (kg)

Height (cm)

Duration of Mechanical Ventilation before the Study (d )

1 2 3 4 5 6 7 8

64 71 60 66 77 72 70 63

80 53 79 64 88 65 50 45

175 155 170 160 160 170 160 175

14 53 8 38 15 19 30 13

14.7 20.5 20.5 25.2 31.0 16.7 25.7 28.5

41.2 39.0 50.4 52.5 48.7 45.0 43.5 41.2

12 22 16 18 16 12 15 21

3 3 3 2 3 4 3 6

68 6 2

66 6 6

166 6 3

24 6 5

22.9 6 2.0

45.3 6 1.7

16.5 6 1.3

3.3 6 0.4

Values are mean 6 SE. * Measured before the beginning of the study. † Pressure support level used in the study.

Spontaneous Respiratory Rate (breaths/min)

PaCO2* (mm Hg)

Pressure Support Level† (cm H2O)

PEEP (cm H2O)

430

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE TABLE 2 MECHANICAL VENTILATORY DATA Patient No. 1 2 3 4 5 6 7 8

Crs (ml · cm H2O21)

Ccw (ml · cm H2O21)

Rrs (cm H2O · s · L21)

Peepi (cm H2O)

Pdimax (cm H2O)

TTI

51 27 45 38 47 27 73 24

78 94 59 178 93 92 99 30

28 33 30 20 27 18 8 26

5.1 5.3 6.1 2.8 5.5 6.5 5.2 8.5

57 64 79 61 68 76 79 51

0.109 0.102 0.082 0.122 0.123 0.114 0.115 0.192

42 6 6

90 6 15

24 6 3

5.6 6 0.6

69 6 3

0.119 6 0.011

Definition of abbreviations: Crs 5 compliance of the respiratory system; Ccw 5 compliance of the chest wall; Rrs 5 resistance of the respiratory system; Peepi 5 intrinsic positive end-expiratory pressure; Pdi max 5 maximum transdiaphragmatic pressure; TTI 5 tension-time index. Values are mean 6 SE. Crs, Ccw, and Rrs were measured during controlled ventilation. Peepi, Pdimax, and TTI were measured during spontaneous breathing.

Figure 1. Individual tracings of airway pressure (Paw), esophageal pressure (Pes), flow (Flow), and rectified integrated electromyodiaphragmatic activity (∫EMGdi) in arbitrary units (AU) in a patient (Patient 8) during spontaneous breathing ventilation (five breaths) and the first 14 breaths following application of inspiratory pressure ventilation. After the abrupt change in Paw, the Pes swing and ∫EMGdi required five to six breaths to stabilize.

per liter of inspired gas. Regarding the EMGdi, on the digitized electrical signal the removal of the artifacts from the EMG tracing was done manually on the computer screen. Then, the signal was rectified and a moving time average was obtained using a time constant of 100 ms. From these processed signals, for each cycle, the maximum value (EMGdi-max) and the area under the curve (∫EMGdi.dt) were derived.

Data Analysis Results are expressed as mean 6 SE. For Phase 2, the values of the five breaths preceding the institution of PSV were averaged and com-

Figure 2. Evolution of tidal volume (VT), respiratory rate (RR), ratio of inspiratory time to total duration of the breath (TI/Ttot), inspiratory work of breathing expressed per breath [W I(J/breath)] or per liter [WI(J/L)] during spontaneous breathing (SB) and the first 14 breaths of inspiratory pressure support ventilation (PSV). *p , 0.05 compared with the mean of the five spontaneous breaths preceding the institution of pressure support.

VOL 157

1998

431

Viale, Duperret, Mahul, et al.: Kinetics of Respiratory Unloading pared to the following first 15 assisted breaths by analysis of variance for repeated measures. When the F value was significant, comparisons between means were carried out by a Tukey’s test (11). In a second comparison concerning Phase 2, the mean value of the five spontaneous breaths, and of the last five breaths of PSV, were considered. During BIPAP ventilation, for each patient, five unassisted and five adjacent assisted breaths were averaged. Comparisons between data acquired during Phases 2 or 3 were carried out by Student’s t test. Statistical significance was assumed at a p value , 0.05.

RESULTS The salient characteristics of the eight patients are given in Table 1. The mean duration of the period of ventilation pre-

Figure 3. Evolution of the esophageal occlusion pressure at 100 ms (Pes 0.1), the peak Pdi (peak Pdi), the mean Pdi (mean Pdi), the maximal value of the rectified and integrated electromyodiaphragmatic signal [Max(EMGdi)], and the area under this signal (∫EMGdi.dt) during spontaneous breathing (SB) and the first 14 breaths following onset of pressure support ventilation (PSV). *p , 0.05 compared with the mean of the five spontaneous breaths preceding the institution of pressure support.

ceding the recordings was 24 d. After the study day, the mean length of partial mechanical ventilation was 5 6 3 d. The respiratory mechanics are reported in Table 2. Time Course of Inspiratory Inhibition Induced by PSV

A typical polygraph recording representative of the institution of PSV is shown in Figure 1. Following the institution of PSV, the tidal volume was significantly increased. However, this increase was transient. The tidal volume was no longer different from baseline value after the eighth breath (Figures 2 and 4). Other changes associated with the institution of PSV were as follows: a decrease in mean Pdi and WI per liter, significant from the first breath onwards; a reduction of WI per breath, significant from the second breath onwards; and a slower decrease of Pes 0.1, peak Pdi, EMGdi-max, and ∫EMGdi.dt, significant from the sixth to the eighth breaths (Figures 2 and 3). Conversely, the respiratory rate did not change after the beginning of PSV. When comparing the mean of the tenth to fourteenth breaths after the onset of PSV with the five preceding spontaneous breaths, no difference was observed for the respiratory rate, the tidal volume (VT), and the ventilation per minute

Figure 4. Breathing data of patients undergoing spontaneous breathing (black column), pressure support ventilation (white column), or BIPAP ventilation. For the spontaneous breathing data, the means of the last five breaths preceding the pressure support onset are reported. For the inspiratory pressure ventilation, the means from the tenth to the fourteenth cycle following the onset of pressure support are reported. For the BIPAP ventilation, the means of the unassisted (dark hatched column) and assisted (clear hatched column) cycles are presented. VT 5 tidal volume; WI 5 inspiratory work of breathing expressed per breath or per liter; mean Pdi 5 area under the inspiratory Pdi curve. *p , 0.05 versus spontaneous breathing for PSV or versus unassisted breaths for BIPAP.

432

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

Figure 5. Breathing data of patients undergoing spontaneous breathing, pressure support ventilation or BIPAP ventilation (same symbols as in Figure 4). Pes 0.1 5 esophageal occlusion pressure at 100 ms; Max(EMGdi) 5 maximum value of rectified integrated electrical activity of diaphragm in arbitrary units (AU); ∫EMGdi.dt 5 area under the curve of rectified integrated electrical activity of diaphragm in arbitrary units (AU). *p , 0.05 versus spontaneous breathing for PSV or versus unassisted breaths for BIPAP.

(8.3 6 2.7 L/min versus 7.6 6 2.6 L/min). Conversely, a significant decrease was observed for the other measured variables (Figures 4 and 5). Finally, arterial partial pressure of carbon dioxide, as measured in two patients, decreased by 4 and 6 mm Hg between spontaneous ventilation and stable pressure support ventilation periods. Comparison of Spontaneous Breath and the Adjacent Assisted Breath during BIPAP Ventilation

Figure 6 shows a polygraph recorded during a BIPAP ventilation set on a mode allowing a spontaneous breath between two pressure-assisted breaths. The VT of the assisted breath was higher than that of the spontaneous breath (Figure 4). However, its duration and TI/Ttot were not significantly changed. The WI per breath was not modified by the pressure assistance, whereas the WI expressed on a per-liter basis was significantly decreased when the breaths were assisted. The value of the Pes 0.1 was not influenced by the pressure assistance nor the EMGdi-max and the ∫EMGdi.dt (Figure 5).

DISCUSSION The purpose of this study was to examine the respiratory response to inspiratory muscle unloading in patients suffering from COPD. The major findings relevant to this issue were the following. First, PSV was associated with a decrease in respiratory drive. Second, although the downregulation occurred soon after the onset of PSV, it needed six to eight breaths to settle down. Third, this transient period could explain the characteristics of the pressure-assisted breaths intermixed with spontaneous breaths during BIPAP ventilation: an increased VT for an unchanged work per breath, suggesting an enhanced inspiratory efficiency.

VOL 157

1998

Figure 6. Individual tracings of the same patient as in Figure 1 of airway pressure (Paw), esophageal pressure (Pes), airway flow (Flow), and rectified integrated electrical diaphragmatic activity (∫EMGdi) during ventilation on the BIPAP mode. This mode was set in order to allow a spontaneous breath between two assisted breaths. Note that the Pes swing and the ∫EMGdi were quite the same in assisted as well as unassisted breaths.

Assessment of the Inhibitory Effects of PSV on the Respiratory Drive

We assessed the inhibitory effect of PSV on respiratory drive by considering certain characteristics of breaths recorded beyond the eighth breath after the onset of PSV, when relatively steady-state conditions were achieved. As no direct measurement of the true respiratory motor output is available, we used several indirect estimates, namely, pressure changes generated during the breath, WI, and EMGdi. However, for these measured or calculated variables to represent valid indices of the respiratory drive, the most important factors transforming neural discharge into pressure variations, then pressure variations into inspiratory work, must remain constant. This is a particularly important factor to consider in COPD patients who have significant levels of PEEPi. Indeed, in such patients, any variation of the dynamic hyperinflation that results from airflow limitation may represent a change in the inspiratory threshold load that they have to face before initiating an inspiratory flow. In our patients, VT and respiratory rate were found to be unchanged during spontaneous breathing and beyond the eighth breath after the onset of PSV. This indicates that end-expiratory volume and thus dynamic hyperinflation were probably unchanged. Therefore, pressures and work measurements did represent reliable estimates of the command signal. The same merit could be conferred to the indices derived from EMG recordings, measured in conditions likely to allow the EMG probe to be in the same position between spontaneous and PSV breathing. Due to a reduced respiratory motor output facing a PSVinduced decreased respiratory load, the ventilatory pattern was unchanged between spontaneous and PSV breathing. This unchanged VT and respiratory rate secondary to the institution of PSV has already been observed in ventilated patients (2, 3, 12). Hence, as has already been observed in previous studies, when pressure support was offered, COPD patients under mechanical ventilation did not use this extra pressure to increase their ventilation for the same amount of work of breathing but to reduce their work of breathing. Unlike healthy subjects (13), patients experiencing airflow limitation adopted a strategy aimed at sparing the work of their respiratory muscles.

433

Viale, Duperret, Mahul, et al.: Kinetics of Respiratory Unloading Kinetics of the PSV-induced Downregulation of Respiratory Motor Output

During the first six to eight breaths following the onset of PSV, patients’ ventilatory patterns were transiently modified. There was an increase in VT associated with an unchanged respiratory rate, which indicated that patients’ end-expiratory volume, and thus dynamic hyperinflation, was likely to be increased. This in turn may have modified the value of some of the variables we used to estimate the respiratory drive. On one hand, hyperinflation, on its own, is a factor of increased work of breathing. Moreover, it might bring into play expiratory muscle activity, which is liable to further increase the calculated WI (14). Similarly, an increase in end-expiratory volume was likely to cause an artifactual increase in EMGdi recording (15). Thus, the direction of change of WI and EMGdi would have led to an overestimation of the respiratory motor output. On the other hand, increased end-expiratory volume is a factor of decreased Pes 0.1, as shortening inspiratory muscles renders them less effective as pressure generators (10). In that case, the direction of change would have led to an underestimation of the respiratory motor output. Keeping in mind these limitations, the step changes in mean Pdi and WI indicate that an inhibition of patients’ respiratory motor output took place within the first two breaths following the institution of PSV. However, two lines of evidence indicate that this inhibitory effect is not fully achieved within the first two breaths. First, all measured estimates of patients’ respiratory motor output continue to decline beyond the second breath. This is particularly obvious for Pes 0.1 and EMGdi, six to eight breaths being necessary for their value to stabilize at a low level following the onset of PSV. Second, beyond the eighth breath of PSV, VT values were lower than those measured for the very first breaths of PSV. As the respiratory load was similar between the two periods, or if modified by a change in end-expiratory volume, even higher during the initial period following PSV, this decrease in VT suggests a further reduction of the motor output. To summarize, from a kinetic point of view, the downregulation of the respiratory motor output associated with PSV is a rapid but gradual phenomenon, which can be detected from the first breath following the onset of PSV, but needs six to eight breaths to be fully established. According to the literature, the mechanisms of inhibition of the ventilatory drive by a mechanical ventilation are intricate (16). Experimental (17, 18) as well as clinical (19, 20) studies suggest that mechanical ventilation depresses the respiratory drive, partially by induced hypocapnia and partially by pulmonary vagal afferent fiber stimulation. The relative importance of vagal influences and hypocapnia on breathing in humans remains a topic of debate. Hypocapnia is widely recognized as having a major role (21), whereas inhibitory sensory input from the lung and chest wall play a minor role (19, 20, 22). In our study, although not designed to address the issue of the mechanism of respiratory inhibition induced by PSV, both mechanisms might have been brought into play: a hypocapnia as usually observed after PSV institution and in two patients of this study, and a vagally mediated inhibition due to the increase in VT initially induced by the PSV (20). BIPAP Ventilation

The particular kinetics of the downregulation of ventilatory drive upon institution of PSV may explain the apparently paradoxical results we observed during BIPAP ventilation: the work of breathing of the assisted breath was similar to the one observed during the spontaneous breath. Indeed, an immedi-

ate and fully achieved downregulation occurring on a breathby-breath basis should have induced an oscillation of the ventilatory drive, but in opposite phase with the pressure support level, the assisted breath leading to a decrease in drive of the following breath. This should have induced an oscillation in parameters such as Pes 0.1 and WI. The higher VT and unchanged Pes 0.1 could be the result of an inspiratory neural output partially but not fully downregulated by the assisted breath following a spontaneous breath. Therefore, the applied pressure was used to increase the efficiency of the assisted breath rather than to decrease the work of breathing, as has been observed for the first breaths following the onset of PSV. These results are consistent with the observations of Marini and coworkers (23) and Imsand and colleagues (24), who noted that the work accomplished per breath was not lessened in assisted versus spontaneous breath during synchronized intermittent mechanical ventilation. However, in this mode, the flow delivered during the assisted breath was constant and may not have been adapted to the inspiratory demand of the patient, thereby increasing the work of breathing. The present study, using a mode of partial ventilation characterized by an inspiratory flow adapted to the patient’s demand, confirms that the high WI observed by previous investigators during machine-assisted breathing was not the consequence of an imbalance between the inspiratory flow delivered by the machine and the patient’s demand. It was rather the consequence of an increase in the efficiency of the assisted breath due to a partially preserved ventilatory drive observed during this breath (23, 24). The potential advantage of such an increased efficiency of breathing does not rest on the increased VT but rather on the possibility it might offer of further reducing the pressure assistance without causing hypoventilation. This advantage was obtained in ready-to-wean patients, as indicated by the fact that their ratio of respiratory rate to VT was below the point at which a high likelihood of a successful weaning is expected (25), and further confirmed by their mean time of ventilation after the measurement period. As previously stressed, improving the ventilatory drive during the weaning process, whatever the means, could maximize the chances of successful ventilator withdrawal (26) or could reduce the weaning time (27). However, the speculative advantage of BIPAP support in the clinical setting of difficult-to-wean patients remains to be established. In summary, for patients with airflow limitation, unloading the respiratory muscles by a pressure support led to a reduced work of breathing with a similar ventilation. This was achieved by a decrease in the respiratory drive. Although present from the very first breaths upon institution of PSV, the downregulation occurred gradually, with a transient period extending over six to eight breaths. The kinetics of the response to inspiratory unloading could explain the increased efficiency of the work of breathing occurring during assisted breath following spontaneous breath, as performed by a partial mode of mechanical ventilation such as BIPAP. References 1. Brochard, L., F. Pluskwa, and F. Lemaire. 1987. Improved efficacy of spontaneous breathing with inspiratory pressure support. Am. Rev. Respir. Dis. 136:411–415. 2. MacIntyre, N., and N. Leatherman. 1990. Ventilatory muscle loads and the frequency-tidal volume pattern during inspiratory pressure-assisted (pressure-support) ventilation. Am. Rev. Respir. Dis. 141:327–331. 3. Berger, K., I. Sorkin, R. Norman, D. Rapoport, and R. Goldring. 1996. Mechanism of relief of tachypnea during pressure support ventilation. Chest 109:1320–1327. 4. Hörmann, C., M. Baum, C. Putensen, N. Mutz, and H. Benzer. 1994. Bi-

434

5.

6.

7.

8. 9. 10. 11. 12.

13.

14.

15.

16.

AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE

phasic positive airway pressure (BIPAP)—a new mode of ventilatory support. Eur. J. Anaesth. 11:37–42. Marini, J., T. Smith, and V. Lamb. 1986. Estimation of inspiratory muscle strength in mechanically ventilated patients: the measurement of maximal inspiratory pressure. J. Crit. Care 1:32–38. Baydur, A., P. K. Behrakis, W. A. Zin, M. Jaeger, and J. Milic-Emili. 1982. A simple method for assessing the validity of the esophageal balloon technique. Am. Rev. Respir. Dis. 126:788–791. Rossi, A., S. B. Gottfried, B. D. Higgs, L. Zocchi, A. Grassino, and J. Milic-Emili. 1985. Respiratory mechanics in mechanically ventilated patients with respiratory failure. J. Appl. Physiol. 58:1849–1858. Bellemare, F., and A. Grassino. 1982. Effect of pressure and timing of contraction on human diaphragm fatigue. J. Appl. Physiol. 53:1190–1195. Rossi, A., G. Polese, G. Brandi, and G. Conti. 1995. Intrinsic positive end-expiratory pressure (PEEPi). Intensive Care Med. 21:522–536. Whitelaw, W., and J. Derenne. 1993. Airway occlusion pressure. J. Appl. Physiol. 74:1475–1483. Winer, B. 1971. Statistical Principles in Experimental Design, McGrawHill, New York. 196–202. Banner, M., R. Kirby, A. Gabrielli, P. Blanch, and A. Layon. 1994. Partially and totally unloading respiratory muscles based on real-time measurements of work of breathing. Chest 106:1835–1842. Morrell, M., S. Shea, L. Adams, and A. Guz. 1993. Effects of inspiratory support upon breathing in humans during wakefulness and sleep. Respir. Physiol. 93:53–70. Lessard, M., F. Lofaso, and L. Brochard. 1995. Expiratory muscle activity increases intrinsic positive end-expiratory pressure independently of dynamic hyperinflation in mechanically ventilated patients. Am. J. Respir. Crit. Care Med. 151:562–569. Gandevia, S. C., and D. K. McKenzie. 1986. Human diaphragmatic EMG: changes with lung volume and posture during supramaximal phrenic stimulation. J. Appl. Physiol. 60:1420–1428. Simon, P. 1994. Effect of mechanical ventilation on the control of breathing. In M. Tobin, editor. Principles and Practice of Mechanical Ventilation. McGraw-Hill, New York. 631–645.

VOL 157

1998

17. Tan, C., and D. Simmons. 1981. Effect of assisted ventilation on respiratory drive of normal anesthetized dogs. Respir. Physiol. 43:287–297. 18. Shams, H., and P. Scheid. 1995. Respiratory response to positive inspiratory pressure in the cats: effects of CO 2 and vagal integrity. J. Appl. Physiol. 79:1704–1710. 19. Tobert, D., P. Simon, R. Stroetz, and R. Hubmayr. 1997. The determinants of respiratory rate during mechanical ventilation. Am. J. Respir. Crit. Care Med. 155:485–492. 20. Manchada, S., A. Leevers, C. Wilson, P. Simon, J. Skatrud, and J. Dempsey. 1996. Frequency and volume thresholds for inhibition of inspiratory motor output during mechanical ventilation. Respir. Physiol. 105: 1–16. 21. Georgopoulos, D., I. Mitrouska, K. Webster, Z. Bshouty, and M. Younes. 1997. Effects of inspiratory muscle unloading on the response of respiratory motor output to CO2. Am. J. Respir. Crit. Care Med. 155:2000– 2009. 22. Lofaso, F., D. Isabey, H. Lorino, A. Harf, and P. Scheid. 1992. Respiratory response to positive and negative inspiratory pressure in humans. Respir. Physiol. 89:75–88. 23. Marini, J., T. Smith, and V. Lamb. 1988. External work output and force generation during synchronized intermittent mechanical ventilation. Am. Rev. Respir. Dis. 138:1169–1179. 24. Imsand, C., F. Feihl, C. Perret, and J. Fitting. 1994. Regulation of inspiratory neuromuscular output during synchronized intermittent mechanical ventilation. Anesthesiology 80:13–22. 25. Yang, K., and M. Tobin. 1991. A prospective study of indexes predicting outcome of trials of weaning from mechanical ventilation. N. Engl. J. Med. 324:1445–1450. 26. Marini, J., C. Roussos, M. Tobin, N. MacIntyre, M. Belman, and J. Moxham. 1988. Weaning from mechanical ventilation. Highlights: ATS Symposia summaries and topics. Am. Rev. Respir. Dis. 138:1043–1047. 27. Holliday, J., and T. Hyers. 1990. The reduction of weaning time from mechanical ventilation using tidal volume and relaxation biofeedback. Am. Rev. Respir. Dis. 141:1214–1220.