Abstract
Ventilation–perfusion (V˙ a/Q˙) distributions were evaluated in 24 patients with acute respiratory distress syndrome (ARDS), during airway pressure release ventilation (APRV) with and without spontaneous breathing, or during pressure support ventilation (PSV). Whereas PSV provides mechanical assistance of each inspiration, APRV allows unrestricted spontaneous breathing throughout the mechanical ventilation. Patients were randomly assigned to receive APRV and PSV with equal airway pressure limits (Paw) (n = 12) or minute ventilation (V˙ e) (n = 12). In both groups spontaneous breathing during APRV was associated with increases (p < 0.05) in right ventricular end-diastolic volume, stroke volume, cardiac index (CI), PaO2 , oxygen delivery, and mixed venous oxygen tension (PvO2 ) and with reductions (p < 0.05) in pulmonary vascular resistance and oxygen extraction. PSV did not consistently improve CI and PaO2 when compared with APRV without spontaneous breathing. Improved V˙ a/Q˙ matching during spontaneous breathing with APRV was evidenced by decreases in intrapulmonary shunt (equal Paw: 33 ± 4 to 24 ± 4%; equal V˙ e: 32 ± 4 to 25 ± 2%) (p < 0.05), dead space (equal Paw: 44 ± 9 to 38 ± 6%; equal V˙ e: 44 ± 9 to 38 ± 6%) (p < 0.05), and the dispersions of ventilation (equal Paw: 0.96 ± 0.23 to 0.78 ± 0.22; equal V˙ e: 0.92 ± 0.23 to 0.79 ± 0.22) (p < 0.05), and pulmonary blood flow distribution (equal Paw: 0.89 ± 0.12 to 0.72 ± 0.10; equal V˙ e: 0.94 ± 0.19 to 0.78 ± 0.22) (p < 0.05). PSV did not improve V˙ a/Q˙ distributions when compared with APRV without spontaneous breathing. These findings indicate that uncoupling of spontaneous and mechanical ventilation during APRV improves V˙ a/Q˙ matching in ARDS presumably by recruiting nonventilated lung units. Apparently, mechanical assistance of each inspiration during PSV is not sufficient to counteract the V˙ a/Q˙ maldistribution caused by alveolar collapse in patients with ARDS.
Acute respiratory distress syndrome (ARDS) causes alveolar collapse primarily in dependent lung areas resulting in a mismatch between ventilation and perfusion (V˙a/Q˙) and severe arterial hypoxemia. An improvement in V˙a/Q˙ matching has been claimed an advantage of partial ventilatory support compared with controlled mechanical ventilation (1-3), because the diaphragmatic contraction presumably augments distribution of ventilation to dependent, well-perfused lung regions (3, 4).
Experimental data suggest that interfacing between spontaneous breathing and mechanical ventilation is a critical determinant of the effects of ventilatory support on V˙a/Q˙ matching (5, 6). Pressure support ventilation (PSV) provides mechanical assistance for each inspiratory effort (7). In contrast, spontaneous breathing in any phase of the mechanical ventilator cycle is possible with airway pressure release ventilation (APRV) that ventilates by periodic switching between two levels of continuous positive airway pressure (CPAP) (8, 9). In an oleic acid lung injury model, we observed better V˙a/Q˙ matching during APRV with spontaneous breathing than during controlled mechanical ventilation or PSV when delivered with equal airway pressure limits (1, 6). However, it is unknown whether spontaneous breathing during partial ventilatory support improves V˙a/Q˙ matching in patients with ARDS.
We hypothesized that in ARDS uncoupled spontaneous breathing during APRV provides better V˙a/Q˙ matching than assisted inspiration with PSV when compared with controlled mechanical ventilation. To test this hypothesis, we examined V˙a/Q˙ distributions in patients with ARDS during APRV with and without spontaneous breathing, and during PSV, delivered with equal airway pressure limits or minute ventilation (V˙e).
After approval by the Innsbruck University ethics committee, 24 mechanically ventilated patients with ARDS were studied. The criteria of the American–European Consensus Conference were used to define ARDS (10). In all patients computer tomography performed prior to the study indicated alveolar collapse in the dependent lung regions. Patients with a history of chronic lung or heart disease and those with unstable cardiopulmonary function were not included in the study. Organ Failure Score (11) and Simplified Acute Physiologic Score (12) were recorded at inclusion in the study.
Heart rate (HR) was obtained from the electrocardiogram. Systemic blood pressure (Psa), central venous, pulmonary artery, and pulmonary artery occlusion pressures were transduced (P50; Statham Gould, Oxnard, CA) and recorded. Cardiac output and right ventricular ejection fraction (RVEF) were estimated with the thermal dilution technique using an algorithm based upon an exponential curve analysis (Explorer; Baxter Edwards Critical-Care, Irvine, CA) (13). Ten milliliters of iced 0.9% saline solution was used as indicator, and an average of seven determinations were performed at random moments during the ventilatory cycle.
Gas flow was measured at the proximal end of the tracheal tube with a heated pneumotachograph (No. 2; Fleisch, Lausanne, Switzerland), connected to a differential pressure transducer (P130; Statham Gould). V˙e was derived from the integrated gas flow signal. Airway pressure (Paw) was measured at the proximal end of the tracheal tube with a pressure transducer (P130; Statham Gould). Esophageal pressure (Pes) was measured with a balloon catheter (Mallinckrodt, Argyle, NY) connected to a pressure transducer (P130; Statham Gould) as described by Baydur and coworkers (14). Intrinsic positive end-expiratory pressure (PEEPi) was estimated as negative deflection in Pes from the onset of inspiratory effort to the point of zero flow as described previously (15). Before baseline measurements a static pressure–volume curve of the total respiratory system was constructed in all patients during transient neuromuscular blockade with intravenous vecuronium bromide 0.1 mg/kg (16, 17).
Arterial and mixed venous blood gases and pH were determined immediately after sampling in duplicate with standard blood gas electrodes (STAT5Profil; Nova Biomedical, Waltham, MA). Oxygen saturation (So 2) and hemoglobin (Hb) in each sample were analyzed using spectrophotometry (OSM3; Radiometer, Copenhagen, Denmark). Fractions of inspired (Fi) and mixed expired (Fe) O2 and CO2 were continuously measured (Datex, Helsinki, Finland).
The method for estimating the distributions of continuous V˙a/Q˙ ratios was described by Wagner and coworkers (18, 19). Six inert gases (sulfur hexafluoride, ethane, cyclopropane, enflurane, diethyl ether, and acetone) were dissolved in lactated Ringer's solution and infused into a peripheral vein at a constant rate set at 0.05% of V˙e for at least 40 min (19). Arterial and mixed venous blood and expired gas samples were collected during stable conditions confirmed by constancy (± 5%) of V˙e, Fe O2 , Fe CO2 , and CI. Expired gas samples were collected with an appropriate time delay from a heated mixing chamber (19). Concentrations of the inert gases were measured with a gas chromatograph (HP 5890; Hewlett-Packard, Waltham, MA) and blood– gas partition coefficients were determined (20).
Arterial to mixed venous (retention) and mixed expired to mixed venous (excretion) concentration ratios of the inert gases were used to obtain retention–solubility and excretion–solubility relationships (18, 19). By formal mathematical analysis with enforced smoothing, these relationships were transformed into a 50-compartment distribution plot of blood flow and ventilation against V˙a/Q˙ (18, 21). Intrapulmonary shunt defined as fraction of pulmonary blood flow (Q˙t) perfusing essentially nonventilated alveoli (V˙1a/Q˙ < 0.005), low V˙a/Q˙ as fraction of Q˙t perfusing poorly ventilated lung areas (0.005 < V˙a/Q˙ < 0.1), high V˙a/Q˙ as fraction of V˙e ventilating poorly perfused lung areas (10 < V˙a/Q˙ < 100), dead space as fraction of V˙e ventilating nonperfused lung areas (V˙a/Q˙ > 100), mean V˙a/Q˙ ratio of perfusion (Q) and ventilation (V), and logarithmic standard deviations of perfusion (logSDq) and ventilation (logSDv) were derived from the 50-compartment model. Predicted values for PaO2 were calculated from the recovered V˙a/Q˙ distributions (19).
Transmural central venous pressure (Pcvtm), pulmonary artery pressure (Ppatm), and pulmonary artery occlusion pressure (Paotm) were derived by subtracting Pes from the respective pressures. Cardiac index (CI) was calculated as cardiac output/body surface area, stroke volume index (SVI) as CI/HR, right ventricular end-diastolic volume index (RVEDVI) as SVI/RVEF, right ventricular end-systolic volume index (RVESVI) as RVEDVI − SVI, systemic vascular resistance (SVR) as (mean Psa − mean Pcvtm) · 80 /CI, and pulmonary vascular resistance (PVR) as (mean Ppatm − Paotm) · 80 /CI. Oxygen content was determined for arterial (CaO2 ) and mixed venous (CvO2 ), blood as (1.34 · So 2 · Hb) + (0.0031 · Po 2), oxygen consumption (V˙o 2) as (V˙i · Fi O2 ) − (V˙e · Fe O2 ), oxygen delivery (Do 2) as CaO2 · CI, and oxygen extraction ratio (O2ER) as (CaO2 − CvO2 )/CaO2 .
After inclusion in the study patients remained supine and received continuous infusion of sufentanil and midazolam as clinically required to achieve a Ramsay sedation score of 4 (22). Fluid replacement and infusion of all drugs remained unchanged throughout the study.
Pressure-limited ventilatory support was provided with a demand valve CPAP circuit of a standard ventilator (Evita; Dräger, Lübeck, Germany). The low pressure level was set at 2 cm H2O above the inflection pressure on a static pressure/volume curve and the high pressure level was adjusted to the value that produced a tidal volume corresponding to the highest lung compliance during transient neuromuscular blockade (17, 23). The ventilator rate was set to maintain PaCO2 between 45 and 55 mm Hg and Fi O2 to maintain PaO2 above 60 mm Hg in the absence of spontaneous breathing. Then baseline measurements were performed. PSV was administered with the same ventilator circuit. Each inspiratory effort triggered an insufflation and the preset high pressure level was maintained until inspiratory gas flow decreased to 25% of its peak value.
Then, patients were randomly assigned to receive PSV and APRV with and without spontaneous breathing using identical low and high pressure levels (equal airway pressure limits) or with an equal V˙e (equal minute ventilation). To maintain V˙e constant the high airway pressure level was varied between the tested ventilatory modalities. All patients maintained spontaneous breathing during ventilatory support with the settings described previously. To assess cardiopulmonary function during APRV in the absence of spontaneous breathing patients were paralyzed with intravenous vecuronium bromide 0.1 mg/kg. Note that from a mechanical standpoint APRV without spontaneous breathing is identical to pressure-limited, time-cycled ventilation.
In 10 control patients ventilator rate was adjusted to lower PaCO2 below the apneic threshold when mechanically ventilated with APRV. When absence of spontaneous breathing during APRV was verified from the esophageal pressure tracing, measurements were performed in random order with and without intravenous vecuronium bromide 0.1 mg/kg. Recovery from neuromuscular blockade was considered sufficient when esophageal pressure deflections were observed during a transient increase in PaCO2 .
A 60-min equilibration period followed each intervention before measurements. Before each intervention, at least 30 min were allowed for cardiopulmonary variables to return to baseline values (± 10%) and the patients' lungs were inflated manually twice to an airway pressure of 40 cm H2O for 30 s to restore lung history.
Results are expressed as mean ± standard error of the mean (SE). Differences between groups were evaluated with the Mann-Whitney U test. Within-group differences were analyzed by repeated-measures analysis of variance. When a significant F ratio was obtained, differences between the means were isolated with the post hoc Duncan's multiple range test. Because of a lack of normal distribution, logSDq and logSDv were analyzed with Friedman's two-way analysis of variance and post hoc comparison with Wilcoxon's signed rank test. The relationship between measured and predicted PaO2 was assessed with a linear regression analysis. Differences were considered to be statistically significant if p < 0.05.
There were no statistically significant differences in the demographic and clinical data between patients who received the tested ventilatory modalities with equal airway pressure limits or V˙e (Table 1).
| Equal Airway Pressure Limits | Equal Minute Ventilation | |||
|---|---|---|---|---|
| Number of patients | 12 | 12 | ||
| Age, yr | 40 ± 5 | 42 ± 6 | ||
| Gender, M/F | 9/3 | 10/2 | ||
| SAPS | 16 ± 2 | 17 ± 3 | ||
| Survived, n (%) | 9 (75) | 8 (67) | ||
| Diagnosis | ||||
| Multiple trauma, n (%) | 10 (83) | 6 (50) | ||
| Postsurgery, n (%) | 2 (17) | 4 (33) | ||
| Others, n (%) | 0 (0) | 2 (17) | ||
| Organ failure, n (%)† | ||||
| 1 | 1 (8) | 0 (0) | ||
| 2 | 5 (42) | 5 (42) | ||
| ⩾ 3 | 6 (50) | 7 (58) | ||
| Sepsis, n (%) | 6 (50) | 7 (58) | ||
| Ventilatory support period, d | 10 ± 2 | 12 ± 3 |
Ventilatory variables are summarized in Table 2. During APRV spontaneous breathing accounted for more than 10% of the total V˙e in both groups. When airway pressure limits remained equal, V˙e was lowest and PaCO2 highest during APRV without spontaneous breathing (p < 0.05). During PSV delivered with equal airway pressure limits, ventilator rate and total V˙e increased (p < 0.05) whereas PaCO2 was unchanged when compared with APRV. To maintain V˙e equal between the tested ventilatory modalities the high Paw level had to be elevated during APRV without spontaneous breathing and PSV which was associated with an increase in mean Paw (p < 0.05) and no change in PaCO2 . Mean Pes was lowest during APRV with spontaneous breathing (p < 0.05).
| Equal Airway Pressure Limits† | Equal Minute Ventilation† | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Baseline‡ | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontaneous Breathing | Baseline‡ | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontaneous Breathing | |||||||||
| Paw high, cm H2O | 27 ± 1 | 27 ± 1 | 27 ± 1 | 27 ± 1 | 29 ± 1 | 28 ± 1 | 30 ± 1§ | 26 ± 1‖,§ | ||||||||
| Paw low, cm H2O | 13 ± 1 | 13 ± 1 | 13 ± 1 | 13 ± 1 | 13 ± 1 | 13 ± 1 | 13 ± 1 | 13 ± 1 | ||||||||
| Paw mean, cm H2O | 21 ± 1 | 20 ± 1 | 21 ± 1 | 21 ± 1 | 24 ± 1 | 22 ± 1 | 24 ± 1§ | 20 ± 1‖ | ||||||||
| Pes mean, cm H2O | 11.2 ± 1.3 | 9.2 ± 1.3 | 11.8 ± 0.9§ | 8.2 ± 0.7‖,§ | 12.0 ± 1.3 | 11.2 ± 1.3 | 12.2 ± 1.3§ | 9.2 ± 1.3‖,§ | ||||||||
| PEEPi, cm H2O | 1.1 ± 0.6 | 0.9 ± 0.6 | 0.9 ± 0.6 | 0.8 ± 0.5 | 0.8 ± 0.5 | 0.8 ± 0.5 | 1.0 ± 0.4 | 0.9 ± 0.6 | ||||||||
| RRspon, min−1 | — | — | — | 9 ± 1 | — | — | — | 10 ± 1 | ||||||||
| V˙ e spon, L · min−1 | — | — | — | 1.4 ± 0.3 | — | — | — | 1.6 ± 0.2 | ||||||||
| RR, min−1 | 18 ± 1 | 26 ± 1 | 18 ± 1§ | 18 ± 1§ | 18 ± 1 | 22 ± 1 | 18 ± 1§ | 18 ± 1§ | ||||||||
| V˙ e, L · min−1 | 8.5 ± 0.4 | 10.1 ± 0.3 | 8.4 ± 0.3§ | 10.1 ± 0.3‖ | 9.5 ± 0.4 | 9.9 ± 0.3 | 9.8 ± 0.3 | 9.9 ± 0.3 | ||||||||
| Ti/Ttot | 0.58 ± 0.3 | 0.52 ± 0.3 | 0.57 ± 0.2 | 0.57 ± 0.3 | 0.58 ± 0.3 | 0.51 ± 0.3 | 0.57 ± 0.3 | 0.57 ± 0.2 | ||||||||
| PaCO2 , mm Hg | 48 ± 1 | 43 ± 2 | 49 ± 1§ | 44 ± 1‖ | 48 ± 2 | 45 ± 1 | 46 ± 1 | 45 ± 1 | ||||||||
Changes in cardiovascular variables are shown in Table 3. Spontaneous breathing during APRV increased RVEDVI, RVEF, and CI (p < 0.05), while mean Ppatm and PVR decreased (p < 0.05) in both groups. Assisted inspiration with PSV increased RVEDVI and CI (p < 0.05) when compared with APRV without spontaneous breathing, but not when the high Paw level was increased to maintain V˙e constant. No change was observed in RVESVI, HR, mean Psa, Pcvtm, Paotm, or SVR.
| Equal Airway Pressure Limits† | Equal Minute Ventilation† | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Baseline‡ | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontaneous Breathing | Baseline | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontaneous Breathing | |||||||||
| HR, min−1 | 112 ± 5 | 112 ± 4 | 113 ± 4 | 106 ± 4 | 112 ± 5 | 106 ± 3 | 107 ± 3 | 102 ± 3 | ||||||||
| Psa, mm Hg | 88 ± 3 | 91 ± 3 | 90 ± 2 | 94 ± 3 | 88 ± 3 | 89 ± 2 | 87 ± 2 | 94 ± 2 | ||||||||
| Ppatm, mm Hg | 32 ± 2 | 31 ± 2 | 33 ± 1 | 29 ± 2§,‖ | 32 ± 2 | 30 ± 2 | 32 ± 2 | 28 ± 2§,‖ | ||||||||
| Pcvtm, mm Hg | 16 ± 1 | 16 ± 1 | 15 ± 1 | 16 ± 1 | 16 ± 1 | 16 ± 1 | 16 ± 1 | 15 ± 1 | ||||||||
| Paotm, mm Hg | 16 ± 1 | 16 ± 1 | 17 ± 1 | 16 ± 1 | 16 ± 1 | 16 ± 1 | 17 ± 1 | 15 ± 1 | ||||||||
| CI, L · min−1 · m−2 | 5.1 ± 0.2 | 5.3 ± 0.2 | 5.0 ± 0.2§ | 5.6 ± 0.2§,‖ | 5.1 ± 0.2 | 5.1 ± 0.2 | 4.8 ± 0.2 | 5.5 ± 0.2§,‖ | ||||||||
| SVR, dyn · s · cm−5 | 550 ± 25 | 550 ± 25 | 565 ± 31 | 521 ± 24 | 550 ± 25 | 535 ± 24 | 588 ± 31 | 538 ± 20§,‖ | ||||||||
| PVR, dyn · s · cm−5 | 126 ± 12 | 115 ± 15 | 130 ± 19 | 95 ± 17§,‖ | 127 ± 14 | 112 ± 17 | 125 ± 15 | 96 ± 15§,‖ | ||||||||
| RVEF, % | 33 ± 2 | 36 ± 1 | 32 ± 2§ | 41 ± 2§,‖ | 34 ± 2 | 38 ± 2 | 34 ± 2§ | 42 ± 2§,‖ | ||||||||
| RVEDVI, ml · m−2 | 122 ± 3 | 128 ± 4 | 123 ± 5§ | 136 ± 4§,‖ | 122 ± 3 | 126 ± 3 | 122 ± 4 | 133 ± 2§,‖ | ||||||||
| RVESVI, ml · m−2 | 86 ± 7 | 82 ± 2 | 87 ± 3 | 82 ± 2 | 82 ± 7 | 78 ± 2 | 83 ± 2 | 79 ± 2 | ||||||||
Spontaneous breathing during APRV was associated with an increase in PaO2 , Do 2, and PvO2 (p < 0.05) regardless of whether airway pressure limits or V˙e were maintained equal (Table 4). Despite spontaneous breathing, V˙o 2 remained unchanged and O2ER decreased (p < 0.05). PSV did not affect PaO2 or Do 2 significantly but increased V˙o 2 when V˙e was not maintained constant (p < 0.05). Arterial pH and Hb remained unchanged for all tested conditions.
| Equal Airway Pressure Limits† | Equal Minute Ventilation† | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Baseline‡ | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontaneous Breathing | Baseline‡ | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontanepous Breathing | |||||||||
| Fi O2 | 0.58 ± 0.02 | 0.58 ± 0.02 | 0.58 ± 0.02 | 0.58 ± 0.02 | 0.59 ± 0.02 | 0.59 ± 0.02 | 0.59 ± 0.02 | 0.59 ± 0.02 | ||||||||
| PaO2 , mm Hg | 83 ± 3 | 86 ± 4 | 82 ± 4 | 102 ± 4§,‖ | 82 ± 4 | 91 ± 4 | 84 ± 3§ | 104 ± 4§,‖ | ||||||||
| pHa, units | 7.34 ± 0.01 | 7.35 ± 0.02 | 7.34 ± 0.01 | 7.35 ± 0.02 | 7.34 ± 0.01 | 7.35 ± 0.02 | 7.35 ± 0.01 | 7.35 ± 0.01 | ||||||||
| PvO2 , mm Hg | 42 ± 1 | 43 ± 2 | 41 ± 1 | 45 ± 1‖ | 42 ± 2 | 44 ± 2 | 42 ± 2§ | 45 ± 2‖ | ||||||||
| Hb, g · dl−1 | 10.2 ± 0.3 | 10.1 ± 0.2 | 10.2 ± 0.3 | 10.2 ± 0.2 | 10.1 ± 0.1 | 10.1 ± 0.2 | 10.2 ± 0.3 | 10.1 ± 0.2 | ||||||||
| Do 2, ml · kg · min−1 · m−2 | 665 ± 34 | 684 ± 26 | 683 ± 30 | 782 ± 28§,‖ | 673 ± 32 | 700 ± 27 | 626 ± 28§ | 753 ± 23§,‖ | ||||||||
| V˙ o 2, ml · min−1 · m−2 | 155 ± 7 | 176 ± 7 | 157 ± 5§ | 163 ± 6 | 159 ± 7 | 173 ± 7 | 160 ± 6§ | 163 ± 6 | ||||||||
| O2ER, % | 24 ± 1 | 23 ± 1 | 23 ± 1 | 21 ± 1§,‖ | 23 ± 1 | 22 ± 1 | 24 ± 1 | 22 ± 1‖ | ||||||||
Results of the multiple inert gas elimination analysis are given in Table 5 and for representative patients in Figure 1. Spontaneous breathing during APRV with equal airway pressure limits or V˙e accounted for a decrease (p < 0.05) in the blood flow to shunt units (V˙a/Q˙ < 0.005) and an increase (p < 0.05) in the fraction of cardiac output to units with a normal V˙a/Q˙ ratio (0.1 < V˙a/Q˙ < 10). Pulmonary blood flow distribution to shunt and normal V˙a/Q˙ units remained essentially unchanged during PSV, compared with APRV without spontaneous breathing in both groups. Dead space (V˙a/Q˙ > 100) was lowest with spontaneous breathing during APRV (p < 0.05). Assisted inspiration with PSV produced no significant difference in dead space compared with APRV without spontaneous breathing. Dispersion of perfusion distribution (logSDq) was lowest during APRV with spontaneous breathing (p < 0.05). Dispersion of ventilation distribution (logSDv) was above the upper normal limit (logSDv > 0.6) for all ventilatory modalities but lowest during spontaneous breathing with APRV (p < 0.05). Predicted PaO2 was close to measured PaO2 for all tested ventilatory modalities. Mean residual sum of squares indicated acceptably small experimental error for all inert gas measurements.
| Equal Airway Pressure Limits† | Equal Minute Ventilation† | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Baseline‡ | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontaneous Breathing | Baseline‡ | PSV† | APRV†without Spontaneous Breathing | APRV†with Spontaneous Breathing | |||||||||
| RSS | 2.97 ± 0.87 | 3.31 ± 0.96 | 2.81 ± 0.90 | 3.11 ± 0.94 | 3.17 ± 0.95 | 3.21 ± 1.10 | 2.99 ± 0.90 | 3.21 ± 1.05 | ||||||||
| Shunt, %Q˙ t | 32 ± 3 | 29 ± 4 | 33 ± 4 | 24 ± 3‖,§ | 33 ± 3 | 28 ± 4 | 32 ± 4 | 25 ± 3‖ | ||||||||
| Low V˙ a/Q˙, %Q˙ t | 6 ± 2 | 8 ± 3 | 5 ± 3 | 6 ± 3 | 4 ± 2 | 8 ± 3 | 5 ± 3 | 6 ± 3 | ||||||||
| Normal V˙ a/Q˙, %Q˙ t | 62 ± 6 | 63 ± 5 | 62 ± 6 | 70 ± 5‖,§ | 63 ± 5 | 63 ± 6 | 62 ± 4 | 70 ± 5‖ | ||||||||
| High V˙ a/Q˙, %V˙ e | 6 ± 4 | 10 ± 6 | 10 ± 6 | 10 ± 6 | 6 ± 4 | 10 ± 6 | 10 ± 6 | 10 ± 6 | ||||||||
| Dead space, %V˙ e | 45 ± 6 | 41 ± 7 | 44 ± 9 | 38 ± 6‖,§ | 45 ± 6 | 41 ± 7 | 44 ± 9 | 38 ± 6‖ | ||||||||
| Q | 0.89 ± 0.16 | 0.88 ± 0.17 | 0.90 ± 0.13 | 0.96 ± 0.18 | 0.94 ± 0.20 | 0.93 ± 0.21 | 0.96 ± 0.16 | 0.98 ± 0.21 | ||||||||
| logSDq | 0.92 ± 0.11 | 0.81 ± 0.13 | 0.89 ± 0.12 | 0.72 ± 0.10‖,§ | 0.88 ± 0.15 | 0.91 ± 0.22 | 0.94 ± 0.19 | 0.78 ± 0.22‖,§ | ||||||||
| V | 1.73 ± 0.26 | 1.58 ± 0.30 | 1.59 ± 0.34 | 1.56 ± 0.37 | 1.72 ± 0.33 | 1.58 ± 0.37 | 1.60 ± 0.41 | 1.55 ± 0.27 | ||||||||
| logSDv | 0.95 ± 0.17 | 0.85 ± 0.21 | 0.96 ± 0.23 | 0.78 ± 0.22‖,§ | 0.95 ± 0.17 | 0.88 ± 0.21 | 0.92 ± 0.23 | 0.79 ± 0.20‖,§ | ||||||||
Fig. 1. Continuous distributions of ventilation and blood flow plotted versus V˙ a/Q˙ for representative patients.
[More] [Minimize]Cardiopulmonary variables during full ventilatory support with and without neuromuscular blockade are given in Table 6 for 10 control patients. In the absence of spontaneous breathing during APRV, blood flow to shunt units and low V˙a/Q˙ units, ventilation of high V˙a/Q˙ and dead space areas, Q, logSDq, V, logSDv, PaO2 , PaCO2 , CO, V˙o 2, and V˙e were not significantly affected by neuromuscular blockade.
| APRV without Spontaneous Breathing | ||||
|---|---|---|---|---|
| Without Neuromuscular Blockade† | With Neuromuscular Blockade† | |||
| RSS | 3.92 ± 0.29 | 4.05 ± 0.42 | ||
| Shunt, %Q˙ t | 25 ± 3 | 24 ± 2 | ||
| Low V˙ a/Q˙, %Q˙ t | 6 ± 3 | 8 ± 3 | ||
| Normal V˙ a/Q˙, %Q˙ t | 69 ± 7 | 68 ± 5 | ||
| High V˙ a/Q˙, %V˙ e | 9 ± 3 | 10 ± 3 | ||
| Dead space, %V˙ e | 42 ± 7 | 40 ± 7 | ||
| Q | 0.93 ± 0.27 | 0.95 ± 0.29 | ||
| logSDq | 0.86 ± 0.11 | 0.84 ± 0.13 | ||
| V | 1.98 ± 0.27 | 1.96 ± 0.30 | ||
| logSDv | 1.30 ± 0.29 | 1.29 ± 0.30 | ||
| PaO2 , mm Hg | 95 ± 7 | 92 ± 5 | ||
| PaCO2 , mm Hg | 35 ± 3 | 34 ± 3 | ||
| p − m PaO2 , mm Hg | 4.2 ± 0.8 | 5.0 ± 0.7 | ||
| V˙ e, L/min | 8.2 ± 0.4 | 8.3 ± 0.4 | ||
| CI, L/min/m2 | 4.1 ± 0.6 | 3.9 ± 0.7 | ||
| V˙ o 2, ml/min/m2 | 221 ± 25 | 205 ± 27 | ||
This study was designed to evaluate the effect of the interfacing of spontaneous and mechanical ventilation on gas exchange in patients with severe ARDS. The uncoupling of spontaneous and mechanical breaths during APRV improved V˙a/Q˙ matching, reflected by decreases in logSDq, logSDv, intrapulmonary shunt, and dead space. The concomitant increase in cardiac output and PaO2 improved the relationship between tissue oxygen supply and demand because V˙o 2 remained unchanged despite the work of spontaneous breathing. In contrast, inspiratory assistance with PSV did not improve V˙a/Q˙ distributions when compared with controlled mechanical ventilation.
Partial ventilatory support is commonly used, not only to separate patients from mechanical ventilation, but to provide stable ventilatory assistance of a desired degree during ventilatory failure (2, 3, 7, 24). We used PSV and APRV (8, 9) that were similar in their pressure-limited delivery of tidal volume, but differed in their interfacing with spontaneous ventilation. Spontaneous breathing in any phase of the mechanical ventilator cycle is possible with APRV that provides a constant degree of ventilatory support by time-cycled switching between two CPAP levels (8, 9). When spontaneous breathing is abolished, APRV is not different from a conventional pressure-controlled mechanical ventilation (8, 9). In contrast, PSV provides a breath-to-breath synchronized insufflation of the lungs. Consequently, relative degree of ventilatory support remains constant as the patient's ventilatory demand alters respiratory rate (7). Because insufflation during PSV is terminated when inspiratory gas flow decreases to 25% of the peak flow value (7), alveolar end-inspiratory pressure will not reach the preset pressure level. Thus, in our patients with a reduced lung compliance equal airway pressure limits resulted in a lower tidal volume during PSV compared with APRV, requiring a compensatory increase in respiratory rate during PSV to maintain alveolar ventilation. However, usually the pressure support level is adjusted to produce a desired tidal volume at an acceptable low respiratory rate (7). To deliver APRV and PSV in our patients with an equal V˙e and comparable tidal volume at a slower respiratory rate the pressure support level had to be significantly increased during PSV.
In agreement with previous findings V˙a/Q˙ distributions observed in our patients were essentially unimodal with 33 ± 4% of the pulmonary blood flow perfusing shunt units indicating severe ARDS (25). Intrapulmonary shunting during ARDS has been found to correlate directly with the quantity of nonaerated tissue observed by computer tomography in dependent lung regions adjacent to the diaphragm (26). These observations have been attributed to alveolar collapse caused by the superimposed pressure on the lung and a cephalad shift of the diaphragm most evident in dependent lung areas during mechanical ventilation (26). Persisting spontaneous breathing has been considered to improve distribution of ventilation to dependent lung areas and thereby V˙a/Q˙ matching, presumably by diaphragmatic contraction opposing alveolar compression (1, 3).
Several studies observed improved V˙a/Q˙ matching but no marked changes in intrapulmonary shunting during the change over from full to partial ventilatory support (5, 24, 27). In contrast to our patients with severe V˙a/Q˙ mismatch, previous investigations have compared different ventilatory support modalities in patients with essentially recovered pulmonary function with only mild V˙a/Q˙ inequality (5, 24, 27). Furthermore, it is difficult to evaluate the effect of different ventilatory support modalities on pulmonary gas exchange on the basis of previous nonrandomized trials, because the degree of mechanical lung inflation or ventilatory support was altered considerably during the course of these investigations (5, 24, 27). In this study, PSV and APRV were provided with equal airway pressure limits or V˙e. Therefore, our results should reflect essentially the effect of different interfacing between spontaneous and mechanical ventilation during partial ventilatory support on V˙a/Q˙ matching.
Spontaneous breathing during APRV resulted consistently in a marked decrease in blood flow to shunt units without creating low V˙a/Q˙ areas. These observations are in agreement with previous experimental findings that spontaneous breathing with APRV improves overall V˙a/Q˙ matching by decreasing intrapulmonary shunting (1, 6). The first explanation for this observation is that nonventilated lung units were completely recruited with spontaneous breathing. Because the normal mean V˙a/Q˙ of the distribution of pulmonary blood flow remained unchanged and its dispersion (logSDq) decreased, some perfused but essentially nonventilated lung units have become normal V˙a/Q˙ regions. Additionally, the absence of change in the mean V˙a/Q˙ of the alveolar ventilation distribution, its lower dispersion (logSDv), and the decreased dead space ventilation indicate improved ventilation of well-perfused lung regions. Previous radiographic observations have demonstrated that contractions of the diaphragm favor distribution of ventilation to dependent, well-perfused lung areas (4). In contrast, full ventilatory support in anesthetized and paralyzed patients has been shown to promote formation of atelectasis in dependent lung regions, which considerably contributes to V˙a/Q˙ mismatch and intrapulmonary shunting (4). Consistent with previous experimental findings (1) neuromuscular blockade did not affect V˙a/Q˙ distributions in patients who had been rendered apneic by lowering PaCO2 . Therefore, the use of neuromuscular blockade to guarantee controlled mechanical ventilation cannot explain the changes in the V˙a/Q˙ distributions associated with spontaneous breathing during APRV. The second explanation is related to the periodic decreases in intrathoracic pressure observed during spontaneous breathing with APRV, which are associated with an increase in cardiac output and may support the perfusion of nondependent high V˙a/Q˙ and dead space regions. Unfortunately, the results of the inert gas measurements only quantify V˙a/Q˙ distributions and do not allow us to distinguish between the two potential mechanisms.
In this study assisted spontaneous inspiration during PSV, regardless of the pressure support level, did not convert shunt to normal V˙a/Q˙ units. In agreement with these findings, alveolar recruitment has not been observed in studies comparing full with synchronized partial ventilatory support of each breath (5, 24, 27). Apparently the spontaneous contribution on a mechanically assisted breath was not sufficient to counteract the V˙a/Q˙ maldistribution of positive pressure lung insufflation in our patients with ARDS.
Decrease in dead space, in the presence of an unchanged mean V˙a/Q˙ of the alveolar ventilation and its decreased dispersion (logSDv) may indicate less hyperinflation during spontaneous breathing with APRV. Controlled mechanical ventilation and PSV appeared to worsen the V˙a/Q˙ inequality in areas with V˙a/Q˙ ratios above normal. This was reflected in unchanged mean V˙a/Q˙ of the alveolar ventilation and its higher dispersion (logSDv), and in increased dead space ventilation, compared with spontaneous breathing with APRV. Ventilation during PSV was distributed in the presence of alveolar collapse mainly to poorly or nonperfused lung units regardless of the level of ventilatory support. In agreement with previous experimental and clinical findings our observations indicate a lower inert gas dead space during spontaneous breathing with APRV when compared with PSV (6). Compatible with our observations, the efficiency of alveolar CO2 elimination during intermittent mandatory ventilation is significantly higher with unsupported spontaneous breaths than with mechanical cycles (28). Based on these results, it has been suggested that spontaneous breaths are distributed preferentially to well-perfused lung units, whereas mechanical cycles increase ventilation in already well ventilated and poorly perfused lung areas (28). Similarly, unsupported spontaneous breaths during APRV may have contributed to improved V˙a/Q˙ matching and decreased dead space ventilation in the presence of ARDS (1, 5, 6).
Small differences between PaO2 predicted from the recovered V˙a/Q˙ distribution and the measured PaO2 indicate that alveolar end-capillary equilibration was complete. Thus, observed changes in gas exchange can be attributed to the measured V˙a/Q˙ mismatch. Corresponding to improvement in overall V˙a/Q˙ distributions, PaO2 increased during spontaneous breathing with APRV. However, a small amount of the variation in PaO2 may be explained by extrapulmonary factors governing PaO2 in addition to V˙a/Q˙ mismatch such as alveolar ventilation, CI, and acid–base status (29).
In our patients spontaneous breathing was associated with an increase of RVEDVI, RVEF, and CI. Our findings support the concept that a fall in intrathoracic pressure during spontaneous inspiration may improve venous return and cardiac output (30). Compatible with previous studies during intermittent mandatory ventilation (28, 31), increase in RVEDVI, RVEF, and CI was highest during unassisted spontaneous breathing with APRV. The increase in CI observed during PSV when compared with controlled mechanical ventilation was a function of the pressure support level. This indicates that during assisted inspiration with PSV spontaneous respiratory activity may not decrease intrathoracic pressures sufficiently to counteract the cardiovascular depression of positive airway pressure.
Changes in cardiac output caused by mechanical ventilation have been reported to correlate positively with the intrapulmonary shunt fraction (32). In contrast, during spontaneous breathing with APRV the increased pulmonary blood flow was directed preferentially to normal ventilated lung units. Consequently, increased CI was associated with less intrapulmonary shunting, an increased PaO2 , and a considerably higher Do 2. In accordance with previous experimental (33) and clinical findings (34) total V˙o 2 was not measurably altered by spontaneous breathing in our patients with a low lung compliance. An increased Do 2 with unchanged V˙o 2 resulted in an improved relationship between tissue oxygen supply and demand as reflected by a significant decrease in O2ER and higher PvO2 , which also may have contributed to the higher PaO2 .
Mechanical ventilatory support techniques that allow unrestricted breathing throughout the mechanical cycle have been introduced to separate patients from mechanical ventilation. The results of this study demonstrate that uncoupling of spontaneous and mechanical breaths during APRV contributes to improved V˙a/Q˙ matching and increased systemic blood flow in patients with severe ARDS. Mechanical assistance of every breath during PSV did not provide any advantage in cardiopulmonary function or gas exchange compared with controlled mechanical ventilation. Long-term investigations are warranted to evaluate the validity of these results in ARDS patients.
| 1. | Putensen C., Räsänen J., Lopez F. A.Ventilation–perfusion distribution during mechanical ventilation with superimposed spontaneous breathing in canine lung injury. Am. J. Respir. Crit. Care Med.1501994101108 |
| 2. | Sydow M., Burchardi H., Ephraim E., Zielmann S., Crozier T. A.Long-term effects of two different ventilatory modes on oxygenation in acute lung injury: comparison of airway pressure release ventilation and volume-controlled inverse ratio ventilation. Am. J. Respir. Crit. Care Med.149199415501556 |
| 3. | Weisman I. M., Rinaldo J. E., Rogers R. M., Sanders M. H.Intermittent mandatory ventilation. Am. Rev. Respir. Dis.1271983641647 |
| 4. | Froese A. B., Bryan A. C.Effects of anesthesia and paralysis on diaphragmatic mechanics in man. Anesthesiology411974242255 |
| 5. | Valentine D. D., Hammond M. D., Downs J. B., Sears N. J., Sims W. R.Distribution of ventilation and perfusion with different modes of mechanical ventilation. Am. Rev. Respir. Dis.143199112621266 |
| 6. | Putensen C., Räsänen J., Lopez F. A., Downs J. B.Effect of interfacing between spontaneous breathing and mechanical cycles on the ventilation–perfusion distribution in canine lung injury. Anesthesiology811994921930 |
| 7. | Brochard L., Harf A., Lorino H., Lemaire F.Inspiratory pressure support prevents diaphragmatic fatigue during weaning from mechanical ventilation. Am. Rev. Respir. Dis.1391989513521 |
| 8. | Stock M. C., Downs J. B., Frolicher D. A.Airway pressure release ventilation. Crit. Care Med.151987462466 |
| 9. | Baum M., Benzer H., Putensen C., Koller W., Putz G.Biphasic positive airway pressure (BIPAP)—a new form of augmented ventilation. Anaesthesist381989452458 |
| 10. | Bernard G. R., Artigas A., Brigham K. L., Carlet J., Falke K., Hudson L., Lamy M., Legall J. R., Morris A., Spragg R.The American–European Consensus Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am. J. Respir. Crit. Care Med.1491994818824 |
| 11. | Knaus W. A., Draper E. A., Wagner D. P., Zimmermann J. E.Prognosis in acute organ system failure. Ann. Surg.111985685693 |
| 12. | Le Gall J. R., Loirat P., Alperovitch A., Glaser P., Granthil C., Mathieu D., Merceir P., Thomas R., Villers D.A simplified acute physiologic score for ICU patients. Crit. Care Med.121984975977 |
| 13. | Dhainaut J. F., Brunet F., Monsailier J. F., Villemant D., Devaux J. Y., Konno M., De Gournay J. M., Lanore J. J.Bedside evaluation of right ventricular performance using a rapid computer thermodilution method. Crit. Care Med.151987148152 |
| 14. | Baydur A., Behrakis P. K., Zin W. A., Jaeger M., Milic-Emili J.A simple method for assessing the validity of the esophageal balloon technique. Am. Rev. Respir. Dis.1261982788791 |
| 15. | Petrof B. J., Legare M., Goldberg P., Milic-Emili J., Gottfried S. B.Continuous positive airway pressure reduces work of breathing and dyspnea during weaning from mechanical ventilation in severe chronic obstructive pulmonary disease. Am. Rev. Respir. Dis.1411990281289 |
| 16. | Sydow M., Burchardi H., Zinserling J., Ische H., Crozier T. A., Weyland W.Improved determination of static compliance by automated single volume steps in ventilated patients. Intensive Care Med.171991108114 |
| 17. | Putensen C., Baum M., Hörmann C.Selecting ventilator settings according to variables derived from the quasi-static pressure/volume relationship in patients with acute lung injury. Anesth. Analg.771993436447 |
| 18. | Wagner P. D., Saltzman H. A., West J. B.Measurement of continuous distributions of ventilation–perfusion ratios: theory. J. Appl. Physiol.361974588599 |
| 19. | Wagner P. D., Laravuso R. B., Uhl R. R., West J. B.Continuous distributions of ventilation–perfusion ratios in normal subjects breathing air and 100 per cent O2. J. Clin. Invest.5419745468 |
| 20. | Wagner P. D., Naumann P. F., Laravuso R. B.Simultaneous measurement of eight foreign gases in blood by gas chromatography. J. Appl. Physiol.361974600605 |
| 21. | Evans J. W., Wagner P. D.Limits on V˙a/Q˙ distributions from analysis of experimental inert gas elimination. J. Appl. Physiol.421977889898 |
| 22. | Ramsay M. A., Savege T. M., Simpson B. R., Goodwin R.Controlled sedation with alphaxalone-alphadolone. B.M.J.21974656659 |
| 23. | Amato M. B., Barbas C. S., Medeiros D. M., Schettino G., Filho G. L., Kairalla R. A., Deheinzelin D., Morais C., Oliviera E., Fernandes, Yae T., Takagaki, De Carvalho C. R.Beneficial effects of the “open lung approach” with low distending pressures in acute respiratory distress syndrome. Am. J. Respir. Crit. Care Med.152199518351846 |
| 24. | Santak B., Radermacher P., Sandmann W., Falke K. J.Influence of SIMV plus inspiratory pressure support on V˙a/Q˙ distributions during postoperative weaning. Intensive Care Med.171991136140 |
| 25. | Dantzker D. R., Brook C. J., Dehart P., Lynch J. P., Weg J. G.Ventilation–perfusion distributions in the adult respiratory distress syndrome. Am. Rev. Respir. Dis.120197910391052 |
| 26. | Gattinoni L., Bombino M., Pelosi P., Lissoni A., Pesenti A., Fumagalli R., Tagliabue M.Lung structure and function in different stages of severe adult respiratory distress syndrome. J.A.M.A.271199417721779 |
| 27. | Beydon L., Cinotti L., Rekik N., Radermacher P., Adnot S., Meignan M., Harf A., Lemaire F.Changes in the distribution of ventilation and perfusion associated with separation from mechanical ventilation in patients with obstructive pulmonary disease. Anesthesiology751991730738 |
| 28. | Wolff G., Brunner J. X., Grädel E.Gas exchange during mechanical ventilation and spontaneous breathing. Chest9019861117 |
| 29. | Rodriguez-Roisin R., Roca J., Torres A.Interplay of intrapulmonary and extrapulmonary factors on pulmonary gas exchange during weaning. Intensive Care Med.171991249251 |
| 30. | Pinsky M. R.Determinants of pulmonary arterial blood flow during respiration. J. Appl. Physiol.56198412371245 |
| 31. | Downs J. B., Douglas M. E., Sanfelippo P. M., Stanford W., Hodges M. R.Ventilatory pattern, intrapleural pressure, and cardiac output. Anesth. Analg.5619778896 |
| 32. | Dantzker D. R., Lynch J. P., Weg J. G.Depression of cardiac output is a mechanism of shunt reduction in the therapy of acute respiratory failure. Chest771980636642 |
| 33. | Stock M. C., Downs J. B., Betts R. K., Frolicher D. A., Howie M. B., Tallman R. D.Oxygen consumption during spontaneous ventilation with acute lung injury in anesthetized pigs. Am. Rev. Respir. Dis.1371988790795 |
| 34. | Räsänen J., Puhakka K., Leijala M.Spontaneous breathing and total body oxygen consumption in children recovering from open-heart surgery. Chest1011992662667 |
Christian Putensen, M.D. was supported by the Lorenz Boehler Trauma Foundation.
