| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Anaesthesiology, Emergency, and Critical Care Medicine (Drs. Moerer, Quintel, and Neumann), University of Göttingen, Göttingen, Germany; University of Göttingen (Mr. Fischer and Mr. Hartelt), Göttingen, Germany; Department of Anaethesiology and Critical Care Medicine (Dr. Kuvaki), Balkan Dokuz Eylül University School of Medicine, Izmir, Turkey.
Correspondence to: Peter Neumann, MD, PhD, Department of Anaesthesiology, Emergency and Intensive Care Medicine, Georg-August-University of Göttingen, Robert Koch Str. 40, D-37075 Göttingen, Germany; e-mail: pneuman{at}gwdg.de
Abstract
Background: Noninvasive ventilation (NIV) is increasingly used in intensive care medicine, but only little information is available how different NIV interfaces affect the performance of a ventilatory system. Therefore, we compared delay times, pressure time products (PTPs), and wasted efforts during inspiration among patients receiving invasive ventilation and NIV with a helmet (NIV-H) or a face mask (NIV-FM).
Methods: Using an in vitro lung model capable of simulating spontaneous breathing, gas flow and airway pressure were measured with varying positive end-expiratory pressure and pressure support (PS) levels. Wasted efforts were determined while lung compliance, respiratory rate (RR), continuous positive airway pressure (CPAP), and PS levels were changed.
Results: Delay times were more than twice as long with a helmet compared to NIV-FM or invasive ventilation (p < 0.001), but decreased during NIV-H with increasing CPAP (p < 0.001) and PS levels (p < 0.001). During the initial inspiratory phase, PTP was smaller with NIV-H compared to NIV-FM or invasive ventilation, but not so when a complete inspiration with PS was evaluated. Wasted efforts occurred earlier during NIV-H and were aggravated with rising PS, RR, and compliance.
Conclusions: Although delay times are prolonged during NIV-H, PTP is initially smaller compared to NIV-FM and invasive ventilation, indicating less work of breathing due to the high volume the patient can access. Increasing the CPAP or PS level decreases delay times in NIV-H and should therefore be considered whenever possible. Wasted inspiratory efforts occurred at higher RRs and should carefully be monitored during NIV.
Key Words: facemask • helmet • invasive ventilation • noninvasive ventilation • pressure support ventilation • trigger
Noninvasive ventilation (NIV) is increasingly used in the treatment of acute and chronic respiratory failure and during weaning from invasive ventilation.1234 NIV has been applied in patients with acute hypoxemic respiratory failure,456 severe cardiogenic pulmonary edema,7 or acute exacerbation of COPD3 in order to decrease the need and the complications of endotracheal intubation.
Noninvasive respiratory support can be applied either as continuous positive airway pressure (CPAP) alone or as NIV with inspiratory pressure support (NPSV) by means of a nasal or face mask. Problems with the widely used face masks result partially from air leakage,89 discomfort of the patient,10 and pressure-related ulcerations of the nose.1112 These problems may limit the duration of use and account for a large proportion of NIV failures. In an attempt to improve NIV tolerance, a helmet was developed that has been successfully used in different clinical situations.513141516
The increased importance of NIV in intensive care medicine and the new interfaces for its application led to the development of special modes of intensive care ventilators to overcome the problems related to higher gas leakage and dead space. Only limited data are available on how the performance of a ventilatory system is affected by such a helmet or a face mask compared to standard invasive ventilation. With demand flow systems, an inspiratory effort of the patient is necessary to trigger gas flow from the respirator. Thus, the trigger sensitivity of the whole respiratory system is of major importance for the work of breathing.1718 The helmet for NIV may affect the trigger sensitivity due to its large compressible volume. Furthermore, besides the specific settings of the ventilator, individual patient characteristics like compliance and resistance of the respiratory system and respiratory rate (RR) as well as the amount of gas leakage may affect the performance of the system and could potentially result in desynchronization between ventilator and patient. Consequently, during NPSV, the patient may inspire with delayed or even without any support from the ventilator (wasted efforts). Therefore, the aim of this study was to characterize the effects of two commonly used interfaces for NIV, face mask and helmet, on the performance of an ICU ventilator in comparison to invasive ventilation, by the following means: (1) measuring delay times after an inspiratory effort; (2) calculating pressure time products (PTPs) for different inspiratory phases; and (3) analyzing the occurrence of wasted inspiratory efforts during varying trigger sensitivities, pressure support (PS) levels, RRs, and lung compliance values.
Materials and Methods
Equipment
CPAP and NPSV were performed with a helmet (Starmed Castar R; Mirandola; Modena, Italy) or a face mask (King Systems Corporation; Noblesville, IN) put on a glass head that was connected to a lung model (Fig 1
). The Castar R Helmet (size medium) has an internal volume of 7.5 L with inflated cuffs. When the head is inserted into the helmet, the internal volume is reduced to approximately 2.4 L due to the volume of the glass head used in this study. Two underarm laces attached to a ring at the lower site of the helmet should prevent the helmet from lifting when it is inflated. A plastic collar fitting around the neck prevents leakage during ventilation. Inspiratory and expiratory tube connectors are fitted in the upper part of the helmet.
|
For invasive ventilation, the lung model was connected to the ventilator via an endotracheal tube (Portex, 7.5 mm; Portex Ltd.; Kent, UK). CPAP and PS ventilation were performed using a conventional ICU ventilator used in our ward capable of invasive ventilation and NIV (Evita 4; Dräger Medical; Lübeck, Germany).
Lung Models and Measurements
Gas flow was measured with a pneumotachometer (Fleisch II; Fleisch; Lausanne, Switzerland) at the inspiratory side of the helmet (Fig 1). The ventilator and lung model were connected by standard disposable ventilator tubes (B&B Beatmungsprodukte GmbH; Neunkirchen, Germany). Flow signals were stored on a personal computer using an analog-digital converter, and the signals were integrated to obtain volume during off-line evaluation. The pneumotachometer was calibrated by a mass flowmeter (TSI 4040 D; TSI Inc.; Shoreview, MN). Airway pressure (Paw) was measured at the inspiratory side before the helmet with differential pressure transducers (Sensortechnics; Puchheim, Germany), adjusted meticulously during zero flow conditions before each measurement.
Measurements of Time Delay
To simulate spontaneous breathing, we used a modified lung model (LS1500; Dräger Medical). This lung model consists of an electrically driven pneumatic lung simulator that allows the adjustment of tidal volume, RR, compliance, and resistance.
To analyze the effect of the different interfaces on trigger sensitivity, effort, and the resulting PTPs, we defined three phases during the pressure curve (Fig 2 ): (1) the time interval (DelayTRIGGER), and the corresponding PTP interval between the initiation of an inspiration and the time point when the deflection of the Paw-time curve showed no further decrease in Paw (PTPTRIGGER); (2) the time interval from the initiation of an inspiration until the preset positive end-expiratory pressure (PEEP) was reached again (DelayPEEP), and the corresponding PTP from the initiation of an inspiration until the preset PEEP was reached again (PTPPEEP); and (3) complete inspiration: during PS ventilation (PSV), the complete inspiration including the unassisted (DelayTRIGGER) and the pressurization segment was added to get the total PTP (PTP calculated over complete inspiration [PTPTOT]).
|
Measurements of Missed Inspiratory Efforts
The influence of the inspiratory PS, RR, and lung compliance on the occurrence of wasted inspiratory efforts was measured using a two-chamber, single-compartment lung model (TTL model 5600I; Michigan Instruments; Grand Rapids, MI). This lung model consists of two chambers connected by a lifting bar (Fig 1). Ventilator and chamber were connected by standard disposable ventilator tubes (B&B Beatmungsprodukte-GmbH; Neunkirchen, Germany). The driving chamber, representing the respiratory muscles, was connected to a ventilator (Evita 4; Dräger Medical), adjusted in volume-controlled mode with an inspiratory time of 1 s, tidal volume of 300 mL, inspiratory flow of 20 L/min, and a varying RR. The pressurization phase (ramp) was set to the shortest possible time interval. The compliance of the driving chamber was set at 20 mL/cm H2O. The resistance of the whole system was kept constant, defined by the used ventilatory tubes and pneumotachometers.
The second chamber, representing the "lung," was connected to the tested ventilator (Evita 4; Dräger Medical). The lung is passively displaced when the driving chamber is inflated by the ventilator. The following settings were studied: (1) RRs of 10, 20, 30, and 40 breaths per minute; (2) PS levels raised in 1 cm H2O steps every minute until wasted inspiratory efforts occurred or up to a maximum of 30 cm H2O; (3) flow trigger set to 0.5 L/min and 15 L/min, respectively; and (4) respiratory compliance set at 30, 60, and 90 mL/cm H2O, respectively. The occurrence of wasted efforts was detected visually (failure of the ventilator-driven chamber to activate the passively driven lung chamber) and by analyzing the pressure volume curves off-line.
Statistical Analysis
Statistical analysis was performed using self-programmed and commercially available software. Data are given as mean ± SD if not stated otherwise. Using a two-way analysis of variance, delay times and PTPs were analyzed for differences between the used interfaces (helmet vs mask vs endotracheal tube) and the applied PEEP or PS levels. If a significant difference was detected, a post hoc analysis using the Scheffe test was performed. A p level < 0.05 was considered to be significant.
Results
Delay Times
During CPAP, DelayTRIGGER was significantly longer (p < 0.001) if a helmet for NIV was used (range, 144 to 174 ms) compared to the face mask (range, 70 to 74 ms) or invasive ventilation (range, 66 to 76 ms) [Fig 3
]. DelayPEEP was even longer with NIV-H (287 to 397 ms) compared to the other two settings (NIV-FM, 171 to 184 ms; invasive ventilation, 151 to 156 ms, p < 0.001). During NIV-H, delay times were reduced by increasing PEEP from 4 to 8 cm H2O (p < 0.001), whereas an increase > 8 cm H2O had only little further effect on delay times (Fig 3). In contrast, no such effect was observed during NIV-FM or invasive ventilation. Different PS levels were studied using a PEEP of 8 cm H2O because increasing PEEP > 8 cm H2O had only a minor effect on delay times (see above). In similar to CPAP, delay times were significantly longer during NIV-H compared to NIV-FM and invasive ventilation (p < 0.001). However, raising the PS caused a significant reduction of delay times, and this effect was most pronounced in NIV-H (reduction when PS was increased from 5 to 20 cm H2O: DelayTRIGGER, 37%; DelayPEEP, 47%), compared to NIV-FM (DelayTRIGGER, 19%; DelayPEEP, 38%) or invasive ventilation (DelayTRIGGER, 5%; DelayPEEP, 16%).
|
|
|
Wasted Efforts
The frequency of wasted inspiratory efforts increased with higher RRs, when the level of PS was increased, when the sensitivity of the flow trigger was decreased, and when the respiratory compliance increased (Table 1
). In addition, the interface used for ventilation influenced the occurrence and frequency of wasted efforts, since wasted efforts were detected at lower PS levels and lower RRs in NIV-H compared to NIV-FM or invasive ventilation (Table 1).
|
The performance of our modern ICU ventilator was significantly affected by the use of a helmet designed for NIV because delay times were more than twice as long with a helmet compared to a face mask or invasive ventilation. The level of PS and CPAP had a major effect on delay times when a helmet was used, but not with a face mask or during invasive ventilation. In addition, PTP was smaller with NIV-H compared to NIV-FM or invasive ventilation, but not so when a complete inspiration with PS was evaluated. Wasted efforts occurred earlier during NIV-H and were aggravated with rising PS, RR, and compliance.
Technical Aspects
The data presented are based on trials in a mechanical lung model and are highly influenced by the compliance of the used interface and the degree of leakage. Repeated measurements with different helmets showed a relative wide range of measured time intervals due to varying compliance values of these helmets. Thus, delay times and PTPs reported in this study may vary if different helmets or tubing are used. The DelayTRIGGER was defined as the time interval between the initiation of an inspiration until the deflection of the Paw-time curve showed no further decrease in Paw, because the latter time point can easily and reliably be identified. However, this definition leads to an overestimation of DelayTRIGGER, since the support of the ventilator actually starts before the nadir of the Paw curve is reached. Even with a sampling rate of 200 Hz, this inflection point could only inconsistently be identified in a preliminary evaluation, and when identified it occurred approximately 5 to 10 ms before the nadir of Paw. Nevertheless, the delay times measured for invasive ventilation are in the range of the quality control data of the manufacturer (Dräger Medical; personal communication, 2004).
Delay Times
The performance of a demand flow system is affected by the interface used: DelayTRIGGER and DelayPEEP were considerably delayed in NIV-H, while the use of a face mask caused only a small increase of the delay times compared to invasive ventilation. This effect was most pronounced with PEEP levels < 8 cm H2O, while increasing PEEP > 8 cm H2O had little further effect on delay times (Fig 3). In contrast to the helmet, the effect of PEEP on delay times was marginal with a face mask or an endotracheal tube. This observation can be explained by the increasing elastance of the helmet with higher PEEP levels, which causes a better pressure transmission and thereby increasing the trigger sensitivity when PEEP was increased with this interface.
With a helmet, the use of PS decreased delay times significantly, whereas this effect was relatively small during NIV-FM or invasive ventilation. These results suggest that the highest PEEP and PS levels clinically indicated and tolerated by the patient should be used when NIV with a helmet is used in order to enhance the trigger sensitivity.
The DelayPEEP seems to be important for patients with acute hypoxemic lung failure who continuously need a high alveolar distending pressure. Lungs collapse, and recruitment may occur rapidly on a breath-by-breath basis with time constants not longer than 400 ms.19 Depending on the PEEP and PS levels provided, DelayPEEP averaged between 81 ms and 397 ms. This time interval would allow collapse of unstable lung units with fast time constants. However, a decrease in Paw during a spontaneous inspiration is always secondary to an increase in the transpulmonary pressure, and should therefore not cause atelectasis.
Inspiratory Effort
Despite the longer delay times, NIV-H was associated with smaller negative PTPs during the initial phase of inspiration (PTPTRIGGER and PTPPEEP), indicating less work of breathing, than NIV-FM and invasive ventilation. This rather surprising finding is—at first sight—in contrast to the results presented by Chiumello and coworkers,13 who found smaller PTPs indicating less ventilatory support with a helmet than a face mask. However, they calculated PTP after initiation of the inspiratory flow. Thus, the initial pressure drop during DelayTRIGGER is disregarded by their calculation.13 In addition, PTP associated with the trigger phase may vary with the respiratory drive of the patient and the specific ventilator used. Without an accessible gas reservoir, the patients’ inspiratory effort causes a negative airway pressure swing before the ventilator responds with an adequately high gas flow. In contrast to an endotracheal tube and a face mask, the helmet contains a large gas reservoir, which can be utilized by the patient during the beginning of an inspiration. Consequently, even though the delay times were shorter during NIV-FM and invasive ventilation compared to NIV-H, the resulting PTPs were higher. This finding could be clinically relevant in patients threatened by respiratory fatigue (eg, patients with neuromuscular diseases or COPD), in whom it is mandatory to minimize the work of breathing.2021 However, PTP calculated over the complete inspiration (PTPTOT) was similar for all three interfaces, since the pressurization during the later phase of inspiration occurred more rapidly with a face mask or invasive ventilation compared to the helmet, as already shown by Chiumello and coworkers.13
Racca et al22 recently studied the effectiveness of the helmet in a human model of resistive breathing, finding a higher inspiratory effort if the helmet was used compared to the face mask. In line with our data (Fig 5), the less effective unloading of the respiratory muscles in NIV-H was partially explained by the underassistance due to the long inspiratory delay time and the impaired pressurization rate during NIV-H. The pressurization may, however, be dependent on the amount of gas leakage and the maximal inspiratory flow of the respirator. Since gas leakage could almost completely be avoided in our lung model study, this may also offer an explanation for our favorable results regarding PTP with the use of a helmet.
In addition, rebreathing of CO2 with a helmet resulted in an almost doubled minute ventilation during resistive breathing22 and was therefore most likely the major cause for the increased work of breathing with the helmet. The problem of limited CO2 elimination has also been described in COPD patients,23 and was analyzed in an experiment24 showing increased CO2 concentrations within the helmet especially with demand flow systems. Both studies2324 imply that a continuous flow or flow-by system may be beneficial to reduce the inspiratory CO2 concentration and thus the risk of CO2 rebreathing.
Wasted Efforts
The occurrence of unassisted, wasted efforts depends on the RR, level of PS, level of PEEP, and respiratory compliance. Unfortunately, high PS levels, which have been advocated in the previous section, facilitate wasted efforts. In the clinical routine, the majority of patients may be managed with PS levels < 15 cm H2O, a range where wasted efforts were only observed at RR > 30 breaths/min in our lung model. With a respiratory compliance of 60 mL/cm H2O, which is rather typical for ventilated patients, a PS < 8 cm H2O resulted in no unassisted efforts regardless of the RR with NIV-H. In NIV-FM, a PS < 10 cm H2O and in invasive ventilation < 11 cm H2O was possible without wasted efforts with a compliance of 60 mL/cm H2O. In line with these results, Chiumello and coworkers13 did not find missed respiratory efforts using either helmet or face mask in a study with human volunteers. However, the RRs did not exceed 14.9 ± 4.1 breaths/min in this study. At RRs > 20 breaths/min, patient ventilator asynchrony occurred in NIV and invasive ventilation with high PS levels, but increasing the trigger sensitivity reduced the frequency of wasted efforts. In addition, respirators used for NIV should be capable to provide high inspiratory flow rates, since delay times and wasted efforts phenomena may increase if the maximal inspiratory flow is reduced. The ventilator used in our study provides a peak flow of 180 L/min, which is rarely exceeded during spontaneous breathing.
Conclusion
DelayTRIGGER and DelayPEEP are considerably increased in NIH-H compared to a face mask or invasive ventilation. However, the PTPs during the triggering phase were the lowest in NIV-H, presumably as a result of the large air reservoir within the helmet. In order to improve the performance of the system, a minimum PEEP of 6 cm H2O might be helpful in the clinical setting. Adding PS may further shorten the delay times and will hardly promote the occurrence of wasted efforts in most clinical settings. Nevertheless, at high RRs, wasted efforts occurred with all interfaces used. Therefore, a close and careful clinical monitoring of patients during NIV is recommended.
Footnotes
Abbreviations: CPAP = continuous positive airway pressure; DelayPEEP = time interval from the initiation of an inspiration until the preset positive end-expiratory pressure was reached again; DelayTRIGGER = time interval between the initiation of an inspiration until the deflection of the airway pressure-time curve showed no further decrease in airway pressure; NIV = noninvasive ventilation; NIV-FM = noninvasive ventilation with a face mask; NIV-H = noninvasive ventilation with a helmet; NPSV = noninvasive ventilation with inspiratory pressure support; Paw = airway pressure; PEEP = positive end-expiratory pressure; PS = pressure support; PSV = pressure support ventilation; PTP = pressure time product; PTPPEEP = corresponding pressure time product from the initiation of an inspiration until the preset positive end-expiratory pressure level was reached again; PTPTOT = pressure time product calculated over complete inspiration; PTPTRIGGER = corresponding pressure time product interval between the initiation of an inspiration and the time point when the deflection of the airway pressure-time curve showed no further decrease in airway pressure; RR = respiratory rate
This study was supported by grants from B & P Beatmungsprodukte-GmbH, Neunkirchen, Germany, and departmental funds. Technical support was provided by TIM GmbH, Göttingen, Germany.
Received for publication November 3, 2005. Accepted for publication February 23, 2006.
References
This article has been cited by other articles:
|
|
 |
|
  D. Chiumello Is the Helmet Different Than the Face Mask in Delivering Noninvasive Ventilation? Chest, June 1, 2006; 129(6): 1402 - 1403. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
