Pressure-controlled inverse ratio ventilation after cardiac surgery

Abstract

Background and objective 

Pressure-controlled inverse ratio ventilation was compared with controlled mechanical ventilation in patients after cardiac surgery.

Methods 

Ten patients were ventilated after sternal closure using a Siemens Servo 900C ventilator to a target end-tidal PCO₂ of 4.0 kPa. They were randomized to receive controlled mechanical ventilation or pressure-controlled inverse ratio ventilation. CO₂-based data were recorded on a laptop personal computer, which together with arterial PCO₂ permitted measurement of the respiratory dead space. Once measurements were complete the ventilator was switched to the other mode and new measurements taken.

Results 

PaCO₂ and CO₂ were virtually the same in both modes. Peak airway pressure (17.2 ± 2.7 vs. 20.8 ± 2.5 cmH₂O, P < 0.01) and minute ventilation (4.9 ± 1.1 vs. 5.3 ± 1.1 cmH₂O, P < 0.01) were less during pressure-controlled inverse ratio ventilation. Physiological dead space fraction (0.39 ± 0.06 vs. 0.51 ± 0.05, P < 0.001), airway dead space (56 ± 15 vs. 81 ± 15 mL, P < 0.001) and alveolar dead space fraction (0.25 ± 0.07 vs. 0.31 ± 0.09, P < 0.01) were all less during pressure-controlled inverse ratio ventilation. There were no differences in heart rate or mean arterial pressure.

Conclusions 

The prolonged inspiratory period and pressure-controlled flow pattern of pressure-controlled inverse ratio ventilation reduce the alveolar and airway dead spaces, and give lower peak airway pressures, compared with conventional ventilation, in cardiac surgical patients.

Introduction

Pressure-controlled inverse ratio ventilation (PC-IRV) has been used in the treatment of patients with acute lung injuries (ALI). During PC-IRV, airway pressure is held at a constant level throughout the major, inspiratory, part of the respiratory cycle, with a short release period for expiration. In some situations, PC-IRV is more efficient than conventional controlled mechanical ventilation (CMV), allowing lower peak airway pressures and minute volumes for any given arterial PCO₂, and reducing the risk of baro- and volume trauma [¹]. During acute lung injury, where there is decreased compliance, areas of collapse, a reduced functional residual capacity and an increased right-to-left shunt [^2,3], this is of benefit.

After cardiopulmonary bypass, the lung exhibits changes that resemble those seen in acute lung injury to some extent. We wished to analyse the differences in the components of respiratory dead space between CMV and PC-IRV after cardio-pulmonary bypass, which has not been previously investigated. We used a Siemens Servo 900C ventilator (Siemens–Elema AB, Solna, Sweden), and an on-line computer program [⁴], which integrated expired CO₂ and flow to produce a CO₂ single breath curve (SBT-CO₂). This, when combined with arterial blood–gas analysis, enabled us to analyse the components of the respiratory dead space and to observe the changes produced by PC-IRV. The method gives essentially the same results as the classic Douglas bag method for measuring respiratory dead space, with the additional benefit of offering subdivision into alveolar and airway dead spaces, and giving information about the quality of gas mixing.

Materials and methods

After local ethics committee approval and after gaining informed consent, 10 consecutive patients (seven of whom were male), undergoing coronary artery bypass grafting (when both authors were available), were enrolled in the study. Patients receiving additionally valve replacements, and those who were considered high risk, were excluded. Their median age was 65 years (range 45–74), height 168 cm (152–179 cm) and weight 77 kg (62–93 kg). They were premedicated with lorazepam 2–3.5 mg orally. Anaesthesia was induced with fentanyl 30–35 µg kg⁻¹. Pancuronium 6–9 mg was given to facilitate tracheal intubation. Patients' lungs were ventilated with 0.5–1% isoflurane in an air/oxygen mixture, FiO₂ = 0.5. Cardiopulmonary bypass was conducted with a 2-L crystalloid prime. Donor blood was given as necessary to achieve a post-bypass haemoglobin concentration of 8.5–9.5 g L⁻¹.

Lung ventilation from induction of anaesthesia to commencement of cardiopulmonary bypass (CPB) was with CMV, 15 breaths min⁻¹, I/E ratio 1 : 2 and 7.5 cmH₂O positive end-expiratory pressure (PEEP) delivered by an Ohmeda 7810 ventilator (Datex–Ohmeda Ltd, Hatfield, Herts, UK). During CPB the patients remained connected to the ventilator (switched off). On recommencement of ventilation all patients received two re-expansion breaths from the reservoir bag, each held for 5 s. Ventilation post-CPB was the same as pre-CPB.

After the sternum had been wired, the ventilator was changed to a Servo 900C ventilator with a Siemens 930 capnograph [⁵], the accuracy of which has been analysed in detail [⁶]. CMV was delivered using volume-controlled ventilation at 15 breaths min⁻¹, I/E ratio 1:2 and 7.5 cmH₂O PEEP. PC-IRV was delivered using pressure-controlled ventilation at 15 breaths min⁻¹, with an I/E ratio of 2:1 and 7.5 cmH₂O of PEEP. In both modes, minute volume was adjusted to produce an end-tidal PCO₂ (PETCO₂) of about 4.0 kPa. This was achieved in CMV by altering the tidal volume and in PC-IRV by altering the inspiratory pressure. The breathing system and the endotracheal cuff had been previously tested for leaks. The first patient was randomized to receive CMV initially; thereafter, PC-IRV and CMV were initially applied in alternate patients. Data collection started once steady state occurred as judged by constant PETCO₂ and CO₂ minute elimination (CO₂). PETCO₂ and CO₂ were recorded from the capnograph display over the following 5–10 breaths. Also recorded, were expired minute volume (_E), peak and mean airway pressures (P_paw, P_maw), heart rate (HR) and mean arterial pressure (MAP). Arterial blood gases were taken for PaCO₂. Signals from the ventilator for expired PCO₂ and airway flow and pressure were recorded on a laptop computer. The computer produced SBT-CO₂, the plot of expired CO₂ vs. expired minute volume [⁷]. The program corrects for compressed volume and re-breathing in the Y-piece [^6,8]. It allows a supplied value for PaCO₂ to be inserted, and from this, calculates the various dead-spaces and dead space fractions.

After recording SBT-CO_{2 and} other variables, the ventilator was then switched to the other mode and adjusted to produce a similar PETCO₂ and CO₂. Once a steady state was re-established, new readings and blood gases were taken. Steady state took 2–5 min to achieve. The patients' lungs were ventilated for approximately 10 min in each mode.

The results were tested using a normal probability plot and found to conform to a normal distribution. No period effect was found when tested. The results are summarized as mean ± one standard deviation (SD) and were analysed using Student's paired t-test. P-values of < 0.05 were regarded as significant.

Results

During PC-IRV, CO₂ was 2 mL min greater⁻¹, and mean PaCO₂ was 0.1 kPa lower, than with CMV Table 1; these were statistically (but not biologically) significant changes. Minute ventilation and P_paw were significantly less, and P_maw greater, during PC-IRV. There were no differences in PaO₂Table 1 or haemodynamics (HR or MAP) between the two modes Table 2.

Table 1: Summary of data reflecting ventilation and airway mechanics during PC-IRV and CMV

Table 2: Summary of cardiovascular parameters during CMV and PC-IRV

The airway dead space was significantly less with PC-IRV (mean 56 vs. 81 mL). The alveolar dead space (expressed here as a fraction of the alveolar tidal volume, V_Dalv/V_Talv [⁷]) was also reduced (mean 0.25 vs. 0.31). Thus, the physiological dead space fraction (V_Dphys/V_T) was much reduced (0.39 vs. 0.50). The slope of phase III, which is a measure of homogeneity of alveolar gas mixing [⁷], was also less in PC-IRV Table 3. Figure 1 shows CO₂ single breath tests [⁷] obtained from one of the patients. It shows that the PC-IRV breath, although containing a similar amount of CO₂, had a smaller tidal volume, airway dead space, and lesser phase III slope than the CMV breath.

Table 3: Summary gas exchange data reflecting the components of dead space in PC-IRV and CMV

Figure 1.: CO₂ single breath tests obtained from one of the patients. The CMV breath contained 9.2 mL CO₂, and the PC-IRV breath, 9.5 mL CO₂. However, the PC-IRV breath had a smaller tidal volume, airway dead space, and lesser phase III slope than the CMV breath. Arterial PCO₂ was the same, 5.3 kPa (which roughly equals 5.3% CO₂) for both modes. All measurements in BTPS (body temperature and pressure saturated).

Figure 2 shows the airway flow and pressure patterns produced by CMV (left) and PC-IRV (right). It can be noted that the airway flow patterns were appreciably different between the two modes. During PC-IRV there is a high initial flow, followed by an exponential decline, whereas in CMV the inspiratory flow is square wave. When PC-IRV was applied to this patient, inspiratory flow ended about half way through the inspiration. Peak airway pressure was clearly lower during PC-IRV. It can be noted that the termination of pressure release occurs at a time when active exhalation is still in progress.

Figure 2.: Airway flow and pressure patterns produced by CMV (left) and PC-IRV (right). When PC-IRV was applied to this patient, inspiratory flow ended about half way through the inspiration. Peak airway pressure was clearly lower during PC-IRV.

Discussion

This study compares PC-IRV with conventional mechanical ventilation after cardiac surgery, and shows the expected improvement in the efficiency of CO₂ elimination, achieved at lower peak airway pressures and a significant increase in P_maw. Similar findings have been reported elsewhere [^1,10,13].

The effectiveness of PC-IRV in patients with adult respiratory distress syndrome is well documented. It has beneficial effects on oxygenation by recruitment of collapsed alveoli, decreases intrapulmonary shunt and venous admixture and increases functional residual capacity [^1,9]. The time scale for the experiments was too short to accurately study changes in PaO₂ as alveolar recruitment occurs over a longer time period than is required to achieve steady state for CO₂ elimination. We did not, however, observe any changes in PaO₂.

The Servo 900C ventilator allows longer inspiratory phases of the respiratory cycle of 67% and 80% of cycle. We used the former, i.e. I/E ratio 2:1, as the 4:1 ratio resulted in detrimental breath stacking during a pilot study. At an I/E ratio 2:1, the termination of pressure release occurred at a time when active exhalation was still in progress Figure 2; air trapping that occurred in this late stage of expiration may have prevented small airways closure resulting in intrinsic positive end expiratory pressure (PEEP_I). In theory, this could improve arterial oxygen tension. Air trapping in PC-IRV has been noted previously [^10,11]. The occurrence of PEEP_I in CMV will result in higher P_paw and P_maw. In PC-IRV, PEEP_I results in a loss of tidal volume, which can be compensated for by increasing P_paw. The protocol did not, however, include the use of an end-expiratory occlusion manoeuvre to measure PEEP_I, with hindsight it would have been useful. This study was primarily interested in the differences in the components of respiratory dead space between CMV and PC-IRV. A previous study by Smith and Fletcher [¹²] has demonstrated no effect on CO₂ elimination by the application of PEEP during CMV.

During PC-IRV, CO₂ was 2 mL min⁻¹ greater, and mean PaCO₂ was 0.1 kPa lower, than with CMV Table 1. These differences were statistically (but not biologically) significant, and are at the limits of the capnograph's stated accuracy. In this experiment, CO₂ recorded at the airway opening is a measure of instantaneous CO₂ elimination, rather than tissue production. We attempted to keep CO₂, thus, constant when changing ventilatory mode. We cannot, of course, exclude the possibility that tissue CO₂ was changing; however, there was only a short-time interval between measurements, and at each, tissue CO₂ and CO₂ elimination appeared to be equal as witnessed by constant CO₂ and PETCO₂.

Previous studies have implied, but not measured, a decrease in V_Dphys during PC-IRV compared with CMV [^1,14,15]. In this study, our equipment allowed us to quantify the changes in the components of the physiological dead space (V_Dphys), the airway dead space (V_Daw) and the alveolar dead space (V_Dalv), all of which were reduced in PC-IRV. The mean alveolar dead space fraction [⁷], V_Dalv/V_Talv, was reduced from 0.31 to 0.25. Compared with CMV, the increased inspiratory time and rapid initial inspiratory flow allow a longer mean distribution time, leading to better alveolar mixing, both between units and also, by diffusion, within units [⁷]. In addition, the decelerating inspiratory flow pattern may allow the fresh gas/alveolar gas interface to move proximally during inspiration, by diffusion, and thus reduce V_Daw, as shown in Figure 1. Figure 1 also shows the reduction in phase I volume (airway dead space gas) and phase III slope (an index of alveolar gas homogeneity) in PC-IRV, evidence of improved alveolar gas mixing. This was a cross-over trial; therefore patients with co-morbidities, such as COPD, which increases phase III slope, would have exerted an equal effect in either CMV or PC-IRV modes. Several were smokers or ex-smokers.

This type of analysis of the individual components of respiratory dead space [⁷] permits a greater understanding of the mechanism by which the reduction in minute ventilation required during PC-IRV occurs. When changing a patient's ventilation from CMV to PC-IRV, a reduction in minute ventilation may be necessary to avoid hypocapnia. This could be achieved by reducing respiratory rate or P_paw, thus decreasing the likelihood of baro- or volutrauma.

The effects of PC-IRV (I/E ratio of 2:1) on cardiovascular stability are inconclusive. Some studies found no change in heart rate (HR) or mean arterial pressure (MAP) but a small decrease in cardiac output (CO) [^10,13]. Another [⁹], found no change in HR, MAP or CO with PC-IRV (I/E ratio 2:1). A transoesophageal echocardiographic and Doppler study [¹⁶] showed no change in HR or MAP, but there was augmentation of CO with PC-IRV (I/E ratio 2:1) compared with CMV. We were unable to measure cardiac output as the insertion of a pulmonary artery flotation catheter in cardiac surgery patients is not routine in our Institution and was not indicated in any of the patients in the study. In our study, cardiovascular stability was not compromised by PC-IRV, as evidenced by the lack of its effect on HR or MAP.

In summary, we have shown in patients undergoing cardiac surgery, PC-IRV was found to reduce both the airway and alveolar dead space secondary to prolongation of inspiration, when compared with CMV. Although mean airway pressure is greater in PC-IRV, peak airway pressure is lower, and there is no negative effect on the circulation.

Acknowledgments

Research funding received from the University Department of Anaesthesia, Manchester University, Manchester, UK.