Keywords

1 Introduction

The use of noninvasive positive pressure ventilation (NPPV) to treat both acute respiratory failure (ARF) and chronic respiratory failure (CRF) has been tremendously expanded in the last two decades in terms of spectrum of diseases to be successfully managed, settings of application/adaptation, and achievable goals [1–3]. The choice of a ventilator may be crucial for the outcome of NPPV in the acute and chronic setting as poor tolerance and excessive air leaks are significantly correlated with the failure of this ventilatory technique [3, 4]. Patient-ventilator dyssynchrony and discomfort may occur when the clinician fails to adequately set NPPV in response to the patient’s ventilatory demands during acute distress, wakefulness, and sleep [3–5]. Technical properties of the ventilator (i.e., type of circuit, efficiency of trigger and cycling systems, speed of pressurization, air leak compensation, CO2-rebreathing, blender for O2, monitoring accuracy, transportability) play a key role in helping NPPV to achieve the goals of mechanical ventilation in ARF (unloading respiratory muscles, improved gas exchange) and CRF (improved gas exchange, sleep quality, quality of life, survival) [3, 4].

With the growing implementation of NPPV, a wide range of ventilators have been produced to deliver noninvasive support, both in randomized controlled trials and in “real-life scenarios.” This chapter examines the key points concerning the technology of ventilators for NPPV and their main impact in clinical practice. Because of constraints in length, ventilators for negative pressure ventilation (i.e., iron lung, cuirass, poncho-wrap) are not covered.

2 Classification of Ventilators

Even if any ventilator can be theoretically used to start NPPV both in ARF and in CRF, success is more likely if the ventilator is able to (a) adequately compensate for leaks; (b) let clinician continuously monitor patient-ventilator synchrony and ventilatory parameters due to a display of pressure-flow-volume waveforms and a double-limb circuit; (c) adjust the fraction of inspired oxygen (FiO2) with a blender to assure stable oxygenation; and (d) adjust inspiratory trigger sensitivity and expiratory cycling as an aid to manage patient-ventilator asynchronies [3, 4].

Ventilators may be classified in four categories, whose features are briefly summarized below and in Table 6.1 [3]:

Table 6.1 Characteristics of the four categories of ventilators for NPPV
Full size table
  1. 1.

    Volume-controlled home ventilators were the first machines used to deliver NPPV, mostly for a domiciliary care. Even if well equipped with alarms, monitoring system, and inner battery, their usefulness to apply NPPV is largely limited by their inability to compensate for air leaks. Consequently, their NPPV application is today restricted to home-based selected cases of neuromuscular disorders, although they still play a role in the safe invasive support of ventilatory-dependent tracheostomized patients [3].

  2. 2.

    Bi-level ventilators are the evolution of home-based continuous positive airway pressure (CPAP) devices and derive their name from their capability of supporting spontaneous breathing with two different pressures: an inspiratory positive airway pressure (IPAP) and a lower expiratory positive airway pressure (EPAP) or positive end-expiratory pressure (PEEP). These machines were specifically designed to deliver NPPV, thanks to their efficiency in compensating air leaks. Because of their easy handling, transportability, lack of alarms and monitoring system, and low costs, the first generation of bi-level ventilators matches the needs for nocturnal NPPV in chronic patients with a large ventilatory autonomy. However, traditional bi-level ventilators showed important technical limitations (risk of CO2 rebreathing due to their single-limb circuit in non-vented masks; inadequate monitoring; lack of alarms and O2 blending; limited generating pressures; poor performance to face the increase in respiratory system load, lack of battery), which have been largely overcome by more sophisticated machines. The newer generations of bi-level ventilators have gained popularity in clinical practice to apply acute NPPV, especially in higher levels of care settings, as well as to invasively support ventilatory-dependent chronic patients at home. These new devices are capable of delivering a large extent of more advanced pressometric modalities of ventilation, with the inclusion of “hybrids modes” such as volume-target pressure-preset ventilation (i.e., volume-assured pressure support ventilation, or VAPS), which can dynamically change the level of pressure assistance depending on the measured tidal volume according to different algorithms. Despite their physiological benefits, the real clinical advantages of these “hybrids modes” have yet to be demonstrated compared with the traditional pressometric modalities [3, 4].

  3. 3.

    ICU ventilators were initially designed to deliver invasive ventilation via a cuffed endotracheal tube or tracheal cannula to either sick patients in the intensive care unit (ICU) or to allow surgical procedures in the theatre room. Despite good monitoring of ventilatory parameters and of flow-pressure-volume waves, as well as a satisfactory setting of FiO2 and of ventilation, performance of conventional ICU ventilators in delivering NPPV is poor because they are not able to cope with leaks. Thus, a new generation of ICU ventilators has been developed to efficiently assist acute patients with NPPV with to the option of leak compensation (i.e., “NPPV mode”), which allows a partial or total correction of air leak-induced patient-ventilator asynchrony, even with large intermachine variability [5].

  4. 4.

    Intermediate ventilators combine some features of bi-level, volume-cycled, and ICU ventilators (dual-limb circuit, sophisticated alarm and monitoring systems, inner battery, both volumetric and pressometric modes, wide setting of inspiratory and expiratory parameters). “Hybrid modes” of ventilation, such as VAPS, are available with the great majority of newer intermediate ventilators. These new machines are designed to meet the patients’ needs, both at home and in the hospital, and for the safe transport of critically ill patients [3, 4].

3 Technological Issues (Table 6.2)

Table 6.2 Key points of the performance of ventilators for NPPV [3]
Full size table

3.1 Source of Gas and Oxygen Supply

ICU ventilators are equipped with high pressure air sources and with a blender in which O2 from high-pressure sources and room air are variably mixed, making the FiO2 controlled and stable. Conversely, bi-level and several intermediate ventilators are provided with either a compressor or an electrically supplied turbine pump to pressurize the room air. which may not assure constant stability in the pressurization. Moreover, these machines do not have a blender, so O2 is delivered from low pressure sources and the FiO2 during NPPV is not easily predictable because it is dependent on several variables: site of O2 enrichment, type of exhalation port, ventilator setting, O2 flow, breathing pattern, and amount of leakage [3]. It was calculated that the highest FiO2 is achieved with the leak port in the circuit and O2 added into the mask using low IPAP levels [6]. Assuring a preset precise FiO2 is a great help in managing acute hospitalized patients suffering from either severe hypoxemia (i.e., requirement of high O2 inhaled gas) or hypercapnic decompensated CRF (i.e., avoidance of CO2 rebound due to inappropriately high O2 delivery).

3.2 Circuit

With ventilators having a single-limb circuit there are two possibilities, depending on the type of exhalation system: (1) intentional leak or “vented circuit” and (2) anti-rebreathing expiratory valve or “non-vented circuit” [7]. The original Respironics BiPAP, like most of the first-generation bi-level ventilators, was provided with a vented single-limb circuit. With this device, the exhalation of the expired air occurs through the whisper swivel, a fixed resistance, variable flow, leak port situated either in the circuit proximal to the interface or within the interface itself. According to physiologic bench studies, this type of equipment may theoretically expose the patient to the risk of CO2 rebreathing, which may be detrimental when treating hypercapnic patients [3]. The CO2 rebreathing is also influenced by the site of the exhalation port, being significantly lower by using a facial mask with the exhalation port inside compared with a facial mask with the exhalation port in the circuit and a total face mask with the exhalation port inside [8]. The options that the clinician has to prevent this risk are (a) to keep the conventional whisper swivel and apply high EPAP levels, such as 8 cm H2O, which may be, therefore, poorly tolerated or (b) to use specific devices such as the plateau exhalation valve, which has a diaphragm that limits air leaks during inspiration and allows a unidirectional air flow during expiration [3]. However, it should be noted that the clinical impact of the potential risk of CO2 rebreathing using ventilators equipped with a vented single-limb circuit is probably overestimated.

Ventilators with a non-vented single-limb circuit are provided with a non-rebreathing valve (mushroom, diaphragm, or balloon valve), which works a true valve. During inspiration, the diaphragm or its balloon is inflated, with full occlusion of the expiratory circuit limb, whereas during expiration, as the valve is deflated, air is allowed to be exhaled throughout it [7]. According to physiological bench studies, even with large variability, these valves may interfere with resistance and expiratory work and, therefore, may increase lung hyperinflation (i.e., intrinsic PEEP) [3, 4]. However, the clinical significance of these physiological findings is unknown.

With ventilators having a dual-limb circuit in which a complete separation exists between inspiratory and expiratory lines (i.e., ICU, last generation bi-level, and intermediate ventilators) there is no risk of rebreathing. Conversely, dual-limb circuit ventilators are less user friendly and more cumbersome compared with single-limb circuit devices. The latter may be preferable for home-based noninvasive ventilatory treatment of the clinical patterns of CRF patients [3].

3.3 Inspiratory Trigger and Expiratory Cycle

The optimization of patient-ventilator interaction during NPPV is essentially based on the technological efficiency of the machine in detecting the patient’s minimum inspiratory effort as quickly as possible (i.e., inspiratory trigger) and in ending the delivery of mechanical support as close as possible to the beginning of the patient’s expiration (i.e., expiratory cycling), independent from respiratory system impedance and air leaks [3, 4]. Ideally, the inspiratory trigger should be set at a higher sensitivity capable of reducing the patient’s effort required to activate the mechanical support. Bi-level ventilators equipped with flow triggers are associated with lower work of breathing and shorter triggering delay time compared with those equipped with pressure triggers [9]. As a matter of a fact, the chance of patient-ventilator dyssynchrony due to “wasted efforts” for a too “tough trigger” is likely to be lower with the former devices. On the other hand, a too sensitive trigger, especially if flow-based, may induce auto-triggering during NPPV with substantial air leaks, and, consequently, ventilator dyssynchrony due to “unwanted efforts” [5]. Inspiratory trigger function may significantly differ, not only among the different categories of ventilators but also within the same category, because of the structural features of the circuit (i.e., single-limb circuit with high resistive valves, “incomplete dual-limb circuit” and a PEEP valve in the short expiratory limb) and the heterogeneity in their performance (pressure-time and flow-time waveforms, trigger delay, leak-induced auto-triggering during NPPV with flow-triggered ventilators) [3].

The cycling to expiration optimizes the synchrony between the inspiratory time (Ti) of the patient and that of the machine. During pressure support ventilation (PSV), cycling to expiration is flow-dependent and occurs at a threshold, which is the decrease in flow either to a default or changeable percentage (usually 25 %) of inspiratory peak flow or to an absolute flow [3, 4]. Patient-ventilator dyssynchrony with expiratory muscle activation and “wasted efforts” due to incomplete lung emptying may happen under NPPV in the case of excessive air leaks that delay or prevent the inspiratory flow from reaching the threshold (i.e., “inspiratory hang-up”) [5]. Successful strategies for preventing inspiratory hang-up that may be applied with some ventilators include (a) setting a suitable threshold and/or a maximum Ti; (b) use of special algorithms (i.e., “auto-track system”); and (c) switching to pressure control ventilation (PCV) mode, in which expiratory cycling is time-dependent [3]. The opportunity to finely set the threshold for expiratory cycling because of the display of mechanics waveforms available in some newer ventilators may be helpful in improving patient-ventilator synchrony during NPPV, as well as comfort and the possibility of success [4, 10].

As observed with the inspiratory trigger, the behavior of different ventilators varies in terms of cycling to expiration, and a marked heterogeneity is reported for a given ventilator in response to various conditions of respiratory mechanics and air leaks. Generally speaking, most of the bi-level ventilators use a cut-off at a higher fraction of inspiratory flow than most of the ICU ventilators to avoid the mask leak-induced deleterious prolongation of Ti. Newer bi-level ventilators tend to prematurely cycle to expiration under normal respiratory system mechanics, and this tendency is exaggerated in restrictive conditions. Conversely, under obstructive conditions, most of the older bi-level ventilators show a delayed cycling, and this behavior is greatly exaggerated by the presence of air leaks. Consequently, at their default setting, bi-level ventilators seem to be better adapted for supporting obstructed patients [3]. Opposite to bi-level ventilators, in the absence of leaks and at their default setting, newer ICU ventilators present some degree of delay in cycling to expiration that is worsened by obstructive conditions, whereas restrictive mechanics lead to premature cycling. The addition of leaks increases the delayed cycling in normal and obstructive conditions and partially corrects premature cycling in restrictive status. This dyssynchrony in expiratory cycling may be prevented by using NPPV modes in normal and obstructive mechanics [3–5].

3.4 Inspiratory Flow

It is known that severely dyspneic patients with chronic obstructive pulmonary disease (COPD) cope better with higher inspiratory flow and neuromuscular patients do better with lower inspiratory flow (i.e., pressure rise times of 0.05–0.1 and 0.3–0.4 s, respectively) [3]. In most bi-level ventilators, this parameter is unchangeable; conversely, in more advanced bi-level ventilators, as well as in most intermediate and ICU ventilators, the rise time may be set with a potential profound effect on unloading of respiratory muscles, tolerance, and leaks. In a physiologic study, the highest pressurization rate was associated with an increased air leakage and poorer NPPV tolerance, even though the diaphragmatic effort was reduced more compared with lower speeds without significant differences in blood gases or breathing pattern. As patient comfort was not different at the lower pressurization speeds, the authors suggested that the individual titration should be targeted to achieve a good tolerance and to minimize air leaks, keeping a relatively high pressurization rate [11].

3.5 Back-up Respiratory Rate

Some bi-level ventilators do not have the option of setting a back-up respiratory rate (f), which raises the costs. Conversely, the majority of newer bi-level ventilators and all intermediate and ICU ventilators are equipped with a back-up f. This option is particularly advantageous in sicker patients with instability of their respiratory drive because it prevents the phenomena of apneas and of periodic breathing, such as Cheyne-Stokes in chronic heart failure. Back-up f may also be useful when a cautious sedation is administered to improve patient compliance to NPPV in expert intensive acute care settings [3].

3.6 Air Leak Compensation

Because of the kind of interface used, air leak is almost a constant feature of NPPV and may interfere with patient comfort, patient-ventilator synchrony and, eventually, the likelihood of success both in acute and chronic patients [1, 5, 10]. During NPPV delivered by ventilators equipped with a vented single-limb circuit, one must consider both intentional (due to the presence of the exhalation system) and unintentional leaks (throughout the mouth during nasal ventilation and/or between the interface and the face with both nasal and oronasal masks) [7]. Excessive unintentional leaks are strongly correlated with NPPV failure as a consequence of alveolar hypoventilation, discomfort, patient-ventilator asynchronies, and sleep fragmentation. On the other hand, attempting to tightly fit the straps of headgear to reduce air leaks should be avoided, because thus may reduce the patient’s tolerance and predispose to skin damage [3, 4]. Consequently, it is important to have a ventilator capable of well compensating air leaks during NPPV. Air leak compensation is greater using bi-level than volume-target home ventilators, with the fall in tidal volume (Vt) >50 % with the latter. Conversely, the fall in Vt is <10 % and in IPAP <8 % with bi-level ventilators in case of leaks because of an adequate increase in the inspiratory flow and in the Ti. However, the effects of air leaks during NPPV are more complex than the simple fall in IPAP and Vt, due to the role played by further variables such as Ti, expiratory cycling and inspiratory trigger sensitivity. Mathematical models that analyze the complex interaction between air leaks and PSV in the obstructive conditions and their potential clinical implications have been recently implemented. Even though all bi-level ventilators and most intermediate and newer ICU ventilators equipped with NPPV modes were able to compensate air leaks, their performance was not uniform [3, 4].

3.7 Battery

For both acute and chronic patients with a high level of dependency on NPPV, a battery power source is mandatory in case of electricity supply failure at home and in case of the need to transport the patient within the hospital or to another hospital. However, clinician must be aware that battery duration differs greatly among the different portable ventilators and may be shorter than that reported in the operator's manual. Moreover, portable ventilator battery duration is affected by the setting, the lung impedance characteristics, and the ventilator features [3].

As an alternative or in addition to internal batteries for NPPV ventilators, it is also possible to use external batteries that guarantee a prolonged autonomy of the ventilator in case of loss of electricity. It has to be considered that external batteries may make the ventilator too heavy when it is to be transported.

3.8 Alarm and Monitoring System

The need for sophisticated alarms and monitoring systems during NPPV is based on clinical practice because, to date, there is no scientific evidence of their clinical utility. This is especially true for patients on home ventilation. Care must be taken when setting the alarms on the ventilator to ensure that they will only function when a genuine need arises, as frequent, often spurious alarms can significantly disturb the sleep of the patient. The prototype of Respironics BiPAP did not have either alarm or monitoring features, with an advantage in cost and transportability in the home care setting. In the acute setting, the availability of newer bi-level, intermediate, and ICU ventilators with more sophisticated alarms (i.e., low and high pressure, Vt, f, FiO2, leaks) and monitoring graph (i.e., flow, Vt, and pressure curves) may be useful in terms of safety and in improving patient-ventilator interaction [4, 5]. Conversely, too elaborate alarms may be counter-productive in the clinical practice because they frequently indicate minor air leaks during NPPV [3].

In the context of patient-ventilator interaction, even if some asynchronies may be suspected at bedside by a careful observation of chest and abdomen movements during NPPV, the interpretation of ventilator curves is helpful to noninvasively assess patient-ventilator interaction [7] (Fig. 6.1). The correct identification of the type of asynchrony during NPPV is helpful in choosing the best strategy to improve the degree of patient-ventilator interaction (e.g., choosing a different interface to reduce leaks and/or changing the setting of the ventilator). A randomized controlled trial clearly demonstrated that a curve-driven setting of the ventilator is capable of achieving a correction of acidosis in a shorter time compared with traditional settings in severe COPD exacerbations [10].

Fig. 6.1
figure 1

A typical patient-ventilator asynchrony pattern resulting from ineffective efforts during noninvasive ventilation that may be easily suspected by looking at the flow-pressure curves of sophisticated more advanced ventilators

Full size image

The keys parameters to be monitor during NPPV are expiratory Vt and f, the determinants of the breathing pattern. Concerning the former, the excessive air leaks may cause a significant discrepancy between inspiratory and expiratory Vt. Expiratory Vt assessment is feasible only with ventilators equipped with a dual-limb circuit where expiratory Vt is obtained by subtracting leaks from inspiratory Vt. With these ventilators, monitoring of expiratory Vt is more reliable with machines that take the measurement at the level of the expiratory branch of the Y-tube than with those that take the measurement at the inlet of the expiratory tube into the ventilator [3]. With ventilators having a single-limb circuit there are two possibilities, depending on the type of exhalation system. In presence of a non-vented circuit, the ventilator gives a inspiratory Vt value that is always an actual measurement of volume delivered by the ventilator. The values are computed at the beginning of inspiration, so that, in the presence of leaks, the leaks are considered as part of the delivered inspiratory Vt. In this case, the ventilator is not able to measure and provide an estimation of leaks and of expiratory Vt as well. In presence of a vented circuit, the ventilator provides an estimation (and not a measure) of expiratory Vt that should be the real volume inspired by the patient without the intentional leaks. In this case, the ventilator is able to provide an estimation of leaks. The leak value displayed may be the total value of leak (intentional + unintentional) or only the unintentional leaks according to the algorithm used by the ventilator. Unfortunately, for a very large leak, its estimation, as well as the estimation of expiratory Vt, may become unreliable [7] (Fig. 6.2).

Fig. 6.2
figure 2

Differences in Vt monitoring during noninvasive ventilation delivered by means of ventilators equipped with double (a) or single-limb circuit provided with a non-vented (b) or vented (c) exhalation system with permission from reference [7]

Full size image

Concerning f assessment during NPPV, there may be a gap between the rate of ventilator-assisted and patient-triggered breaths. Looking at the ventilator-f, ineffective efforts and auto-triggering may cause, respectively, underestimation and overestimation of the effective patient’s f [7].

4 Controversial Issues

Because of the huge gap between the increasing number of newer ventilators that are commercially available and their physiologic and clinical careful evaluation, there is no published data about several sophisticated ventilators routinely used in the clinical practice. With few in vivo investigations, the majority of data about the performance of the different available ventilators comes from in vitro studies conducted on lung models. Therefore, some doubts remain about the real clinical significance of the technical differences observed in the bench studies among the vary types of ventilators. Consequently, every extrapolation of these experimental data to the clinical setting must be done cautiously because no lung model can simulate the ventilatory variability observed in patients. This is particularly true when the findings of in vitro studies have to be applied to acute patients under NPPV in the presence of leaks.

Based on the published data of the literature, despite a wide heterogeneity found in each category of machines, several bi-level ventilators demonstrated a better performance than several ICU ventilators [3, 4]. However, no study has shown greater NPPV clinical success for one type of ventilator than another both in the acute and chronic settings. Nevertheless, some points should be clear when the clinician must choose a ventilator [3].

Key Major Recommendations

  • As excessive air leaks are correlated with treatment failure, the clinician should choose ventilators designed for NPPV with leak compensation capability (i.e., bi-level, some intermediate and new ICU ventilators). Moreover, the capability of setting several parameters and looking at flow-volume-pressure waveforms with newer ventilators may be helpful in improving patient-ventilator synchrony, comfort, gas exchange and, hopefully, clinical outcome.

  • The choice of ventilator should be tailored to the pathophysiology and the severity of ARF and CRF. In the acute setting, for hypoxemic patients, ventilators with an O2 blender are recommended, whereas in those with hypercapnia, ventilators with a dual-limb circuit have an advantage in lowering PaCO2. In patients with mild COPD exacerbation, the use of home ventilators may be appropriate, particularly if the patient is already on home NPPV. In contrast, patients with life-threatening ARF at risk of intubation should be treated with more sophisticated machines. In the chronic setting, conditions where respiratory drive is good (e.g., COPD) could use a simple ventilator that works in a spontaneous mode, whereas those in whom respiratory drive is impaired must have a mandatory back-up rate. Conversely, in case of fast-progressing neuromuscular diseases, a more sophisticated ventilator with adequate monitoring equipment and an inner battery is recommended. The clinical superiority of new hybrid modes of NPPV (e.g., VAPS) over the traditional pressometric modes has yet to be demonstrated in both the acute and chronic settings.

  • The selection of a ventilator should also take into account costs and staff experience. The more sophisticated a ventilator is, the longer the training for clinicians is required. With the tremendous growth of the ventilator market in terms of complexity, some of the new bi-level ventilators are not user-friendly even for trained ICU clinicians. The smaller the variety of devices used, the greater the likelihood that all team members will acquire enough experience in NPPV set-up, with positive repercussions in costs and workload.

  • The clinician should be aware of the multiple interferences of the accessories for NPPV (interfaces, exhalation systems, pressure settings, and humidification devices) with the performance of the different categories of ventilators. For example, concerning humidification during NPPV, heated humidifiers show great clinical and physiological advantages compared with heat-moisture exchangers, even though the former is more time-consuming.