MONITORING OF NIV

Dr Claudio Rabec

NIV is predominantly applied during sleep. Sleep greatly influences ventilatory behaviour by

inducing modifications of ventilatory control, upper airway patency and respiratory muscle

recruitment, in particular in patients with respiratory insufficiency. Therefore, NIV settings

chosen empirically during daytime may not predict optimal nocturnal ventilatory support.

Consequently, NIV effectiveness might be more correctly assessed by sleep studies than

through daytime assessment [1]

A patient with NIV could be considered as well ventilated when the ventilator provides a

proportional assistance to his needs without limiting the expression of its own respiratory

activity and when there are obvious signs of improvement or correction of alveolar

hypoventilation but also an improvement or at least a preservation of sleep quality [2] (Table

1). However, until now, neither a univocal definition of effective ventilation nor a codified

strategy to evaluate its effectiveness have been established. The optimal monitoring of

patients treated with long-term NIV is still a matter of debate. Hence, physicians caring for

these patients may vary greatly in their methods of assessing the effects of NIV, from a

single blood gas measurement to full PSG.

Oximetry, capnography or poly(somno)graphy?

Initially, long-term mechanical ventilation was set according patient tolerance and

improvement of arterial blood gases. Although the long-term results were rather satisfactory

with improvement of life expectancy, a significant percentage of the patients with chronic RF

were not correctly ventilated and it was difficult to understand why, because adequate tools

to explore the patient/ventilator interactions were lacking at that time.

More recently, it has been suggested that NIV evaluation may be done by clinical

assessment, and simple tools such as arterial blood gases, and nocturnal SpO2 under NIV

[3]. There is an agreement on SpO2 recording as a minimal requisite but some studies have

shown that overnight PtcCO2 is also mandatory as patients may remain hypercapnic despite

normal SpO2 levels [4,5]. In addition, simplified in-built monitoring systems coupled to some

ventilators may allow to obtain interesting additional data [6-8]. These devices administer

trends that could provide useful information about the global quality of ventilation.

Furthermore, some of these systems could also provide raw data such as flow and pressure

traces from the ventilator. Hence, it is possible to complete recordings of the respiratory

pattern and therapeutic compliance of patients on long-term NIV. A plugged interface can

even permit to obtain SpO2 curve and also transcutaneous carbon dioxide (PtcCO2) on the

same recording. Furthermore, derived pulse plethysmographic parameters can provide

useful autonomic markers of sympathetic tone and information on sleep fragmentation.

These data may be sent through internet to the prescriber on demand or automatically in the

context of telemonitoring.

But as these data lack some critical signals (i.e thoraco-abdominal movements), others

[1,2,9] suggested that this strategy may not be sufficient. During synchronous ventilation,

triggering and cycling occur at the same time as the beginning and end of inspiratory efforts

on thoracoabdominal signals. Therefore, combined with flow and pressure recording, thoraco

abdominal signals are crucial to understand patient-ventilator synchrony. For instance, in a

ventilated patient, recognising thoraco-abdominal movements without synchronous

pressurisation is a good marker of “unrewarded” inspiratory efforts [2]. Additionally, a

qualitative estimation effectiveness of ventilation can be obtained from the two bands (and

eventually also from the sum of an inductive plethysmograph). Thus, when these signals are

recorded together, one can obtain quite an accurate picture of the adaptation between the

patient and the ventilator and of the efficacy of NIV. Using full PSG gives additional

information on the sleep efficiency and sleep architecture during NIV treatment and could
give more insight in the occurrence of specific respiratory events during different sleep

stages.

All these data, together with the patient’s clinical status, permit to define the quality of NIV.

Frequently used therapeutic goals include: clinical improvement and reduction of daytime

PaCO2, mean nocturnal SpO2 > 90% more than 90% of the recording time without residual

oscillations and use of nocturnal NIV > 4hours without discomfort (fragmented use or multiple

short periods of ventilator use).

Ventilator built-in software

Home ventilators have evolved rapidly since the first cohort studies, with an increasing use of

pressure-cycled bi-level ventilators over the past twenty years. As already discussed,

monitoring of the efficacy of long-term NIV usually relies on medical history, daytime arterial

blood gases, nocturnal pulse-oximetry, sometimes coupled with transcutaneous

capnography. Recently, an increasing awareness of nocturnal respiratory events occurring

under NIV has led to a wider use of respiratory polygraphy and PSG to better understand,

define and detect these events. The most frequently detected problems are unintentional

leaks (UL), PVA, obstructive or central events (either residual or induced by NIV), and those

have been recently extensively reviewed [3]. The latest generations of home ventilators have

built-in software which provide the clinician with potentially valuable information such as

compliance, estimation of leaks, VT, VE, respiratory rate, percentage of inspirations triggered

by the patient, percentage of pressurizations interrupted by the patient (cycling), and indices

of apnea and/or apnea-hypopnea. The first study to explore the validity (reliability) of this

information in one home ventilator is the study of Rabec et al [8]. Authors tested one

ventilator and data of the built-in software were highly correlated (for leak r = .947; p < .001

for VE; r = 0.959; p < 0.001).

Over the last four years, three studies have evaluated the reliability of the data collected by

built-in software.
Leaks and patient-ventilator asynchronies

Leaks are inherent to the process of NIV. They are subdivided into intentional leaks, i.e.

those associated with the exhalation valve placed either on the tubing or built into the

interface, and UL, i.e. leaks occurring anywhere between the ventilator and the patient’s

airways, but not through the exhalation valve. Intentional leaks are mandatory for elimination

of CO2 from the ventilation circuit and avoiding rebreathing. They can vary considerably from

one interface to another and choice of interface may affect capacity to achieve the preset

pressure support. UL always occur to some extent during NIV. Bi-level pressure support

ventilators, or intensive care unit (ICU) ventilators with an “NIV” mode are designed to detect

and compensate UL to a certain level.

PVA refers to the presence of an asynchrony between the patient’s neural respiratory drive

(NRD) and effective ventilation or pressurization. It encompasses respiratory events such as

ineffective or unrewarded inspiratory efforts, auto-triggering, double or multiple triggering and

prolonged dissociation between pressurizations and inspiratory efforts. Intra-cycle PVA

(ineffective or delayed triggering, premature or late cycling) may also result from UL. [27]

These events have been shown to affect sleep structure and may affect work of breathing.

[26,27]

PVA per se increases Stage I and II sleep, and microarousal index and decreases slow wave

and REM sleep in stable obesity-hypoventilation patients under long-term NIV. [25]

Ineffective efforts may also affect efficacy of NIV and lead to a lesser control of nocturnal

hypoventilation and arterial blood gases. [25]

Leaks have a clinically relevant deleterious impact in NIV. In volume-cycled NIV, they lead to

insufficient compensation of hypoventilation, lower VT, lower positive end-expiratory

pressure, and affect the ability of the patient to trigger the ventilator. In patients with

kyphoscoliosis, under bi-level pressure support, leaks through the mouth have been

associated with frequent microarousals and sleep fragmentation. [11]. Leaks may decrease
the fraction of inspired oxygen when supplemental oxygen is administered during NIV. [11]

Important UL may lead to significant pressure drops, poor inspiratory triggering, increase in

duration of inspiratory pressurization, and even inversion of I:E ratio. [8].

Leaks and PVA are related: in ICU patients, leaks have been shown to be significantly

correlated with ineffective breath efforts, delayed cycling, and presence of an asynchrony

index above 10% of total recording time. [27] In patients with long-term NIV for

neuromuscular disorders, PVA was shown to occur in relation with leaks. [25] In this setting,

PVA events were mainly ineffective inspiratory efforts, auto-triggering and prolonged

insufflations. PVA rate was correlated with arousals and awakenings.

In summary, in acute or chronic care settings, UL are frequent during NIV. They have a

major impact on efficacy of ventilation in volume-cycled devices which do not compensate for

UL. In bilevel positive airway pressure ventilation (or “NIV mode” for ICU ventilators), UL are

partially or totally compensated according to devices used. However, UL are associated with

PVA. PVA may in turn affect efficacy of NIV, and sleep structure, although considerably PVA

may occur without any adverse effect on arterial blood gases and correction of nocturnal

hypoventilation [25]. The relationship between sleep disruption resulting from PVA and leaks

and clinical outcomes such as compliance, HRQL or even survival remains to be assessed.

Scoring of respiratory events during NIV

Compared with invasive ventilation, NIV has 2 unique characteristics: the nonhermetic nature

of the system and the fact that the ventilator-lung assembly cannot be considered as a

single-compartment model because of the presence of variable resistance represented by

the upper airway.

As already discussed, unintentional leaks may be absent or minimal when the patient is

awake but may worsen during sleep as a result of the loss of voluntary control and

decreased muscle tone [11]. For that, NIV itself may induce undesirable respiratory events.
During NIV, a variable resistance constituted by the upper airway (UA) is interposed between

the ventilator and the lungs. This resistance may vary, compromising the delivery of an

effective tidal volume to the lungs. Intermittent obstruction of the UA is something common

under NIV and may be related to two mechanisms. The first corresponds to obstructive

events at the oropharyngeal level because of UA collapse. Another mechanism corresponds

to episodes of intermittent obstruction at the glottic level reflecting active cyclic glottic

closure.

Respiratory events during NIV

Recent observations have shown that standard definitions for nocturnal respiratory events in

spontaneous breathing do not lend themselves well to the description of respiratory events

occurring during positive pressure ventilation. Indeed, one major difference during NIV is the

continuous interaction between the ventilator, generating an intermittent positive pressure,

and the patient’s NRD [2]. During NIV the patient is assisted by a ventilator and events can

result from the patient, the ventilator or from poor patient-ventilator coordination. These

different events have been recently described by Gonzalez-Bermejo et al [2] (Table 2)

A respiratory event was defined as the occurrence of a modification, discontinuity or

instability of ventilation which had deleterious consequences on SpO2, PtcCO2 and/or sleep.

The SomnoNIV task group considers as a minimal prerequisite for polygraphic (PG) analysis

of these events the presence of signals for pressure, flow, abdominal and thoracic

movements and SpO2 [2]

These tools are generally sufficient to identify, with an adequate specificity, different types of

events that could be identified in patients on NIV and their consequences

The prevalence of these events is variable depending on the clinical situation, etiology of

respiratory failure, type of mask or ventilator settings. Few studies have tried to relate

untoward respiratory events to alterations in sleep structure and/or reductions in oxygen

saturation. Studying a population of children on long-term NIV, Caldarelli et al [12]

demonstrated that respiratory events are common, even in well-compliant children who do

not complain of symptoms. A recent study analyzing 125 polygraphies under NIV in an adult

population, showed that the mean nighttime spent in event was greater than 20%. [13] (Fig

1). In this study, time spent on events was significantly correlated with lower nocturnal SpO2

and higher diurnal PaCO2 levels (Figure 2). In both studies, applying pathophysiological-

based adjustment of ventilator settings led to a significant improvement in the quality of

ventilation [12,13].This underlines the interest of performing PG in ventilated patients to

optimize efficacy of NIV.

Unintentional leaks

As discussed in the previous section, the impact of leaks on ventilatory efficacy depends not

only on the absolute amount of leak but also on the capability of the device to compensate

them [6,17]. It is also possible that the influence of leaks could vary regarding underlying

disease (i.e difference in respiratory mechanics) [18]. Therefore, detecting unintentional

leaks and particularly their impact on quality of ventilation is of major importance when

monitoring NIV.

When performing a PG [2] the importance of the leak and the ability of the ventilator to

compensate for them determines whether the pressure signal amplitude remains stable or

decreases. A fall in positive pressure (inspiratory and expiratory) indicates major

unintentional leaks. With pressure-controlled ventilators, an increase in ventilator flow signal

during insufflation with a simultaneous decrease in thoracic and abdominal belt signal

amplitude is suggestive of unintentional leaks (Fig 3). Ventilator flow increases to

compensate for drop in pressure, but leaks result in decreased effective VT delivered to the

patient. Because of the inability of volume-controlled ventilators to compensate for leaks, a

decrease in thoracic and abdominal belt signal amplitude can occur even in the presence of

small leaks without any increase in flow signal. However, a decrease in pressure signal is

usual. Furthermore, as volume targeted ventilation is usually provided by a circuit in which

inspiration and expiration are separated by a true expiratory valve, if present on baseline

tracing (pneumotachograph between mask and expiratory valve), an amputation of the

expiratory part of the flow curve indicates the loss of expiratory flow in the circuit and thus

leaks (Fig.4).

Decrease of ventilatory drive

During sleep the voluntary controller is abolished and ventilation becomes exclusively

dependent on automatic control. In addition, ventilatory control is physiologically altered

during sleep with a decreased responsiveness to chemical, mechanical, and cortical inputs. If

NIV settings lead to hyperventilation, bursts of central apnea or hypopnea can occur,

particularly during transitions between sleep onset and wakefulness [19]. Adduction of the

vocal cords resulting in progressive closure of the glottis has also been described in

response to ventilator-induced hyperventilation [20].

On PG [2] the essential feature of a decrease in ventilatory command is a proportional and

simultaneous reduction in flow and in thoracic and abdominal belt signal amplitudes without

phase opposition. In the absence of a back-up respiratory rate (ventilator in ‘spontaneous’

mode), thoracic and abdominal belt signals may disappear completely, generating a pattern

of recurrent central apnoeas (figure 5A). With a backup respiratory rate, the length of the

events is limited by the preset maximal interval between breaths (figure 5B).

Partial or total upper airway obstruction with or without reduction in ventilatory drive

Intermittent obstruction of the upper airway (UA) is something commun under NIV and may

be related to two mechanisms. The first corresponds to obstructive events at the

oropharyngeal level because of UA collapse, as a result of insufficient expiratory airway

pressure (EPAP). This mechanism may be present in patients with an unstable UA, such as

patients with OSA or obesity hypoventilation syndrome (OHS). Another mechanism

corresponds to episodes of intermittent obstruction at the glottic level, reflecting cyclic glottic

closure induced by hyperventilation. Positive pressure ventilation-induced hyperventilation

has been shown to promote active glottic closures in normal subjects when awake or asleep

[20]. Glottic closure was shown to be proportional to total ventilation and inversely

proportional to end-tidal CO2. By using simple tools, such as nocturnal SpO2, these 2

mechanisms are both indistinctly expressed as desaturation dips under NIV. Both tools can

not distinguish between apneas at oropharyngeal and glottic levels [8].

Moreover, both mechanisms are not only located at different stages in the airway and

recognize different pathophysiological mechanisms. They have also distinct semiology in the

polygraphic traces and their therapeutic approach remains different. While both cases are

characterized by a sudden reduction in flow amplitude during insufflation, obstruction at the

oropharyngeal level will be accompanied by progressively increased inspiratory activity,

indicating a struggle against UA collapse. When performing a PG, it will be expressed as a

progressive increase of abdominal and thoracic belt signals, with or without a phase

opposition or a phase angle between thoracic and abdominal belt signals, suggesting partial

or total closure of the upper airways [2] (fig 6A). In this case, the strategy is to increase the

level of EPAP to stabilize the UA.

Conversely, if the underlying mechanism is glottic closure, the essential feature of a

decrease in ventilatory command is a decrease in flow with a simultaneous reduction or

disappearence of thoracic and abdominal belt signal, that occurs without phase opposition,

as a result of an excessive level of ventilation promoting respiratory pauses (Fig 6B) [2]. In

this case, the suggested approach is to reduce minute ventilation [20,21]. Once again, this

underlines the interest of analysing thoraco abdominal belts to assess NIV quality and guide

therapy.

Asynchrony
Synchrony between patient’s spontaneous breathing activity and ventilator’s set parameters

is one of the key factors determining efficacy and tolerance to NIV. Asynchrony during

triggering and cycling is quite common during sleep in patients during long-term NIV.

According to different published series the estimated prevalence of significant asynchronies

varies from 17 to 55% [12,25,26].

During NIV, leaks may greatly affect patient ventilator synchrony and most asynchronies in

patients on NIV are related to unintentional leakage [15]. Different types of leak-induced

asynchronies can be seen in Fig 7.

 
 
 
 
 
 Patient’s satisfaction
Symptoms improvement
Perceived good quality of ventilation
Improving quality of life

Respecting / improving sleep structure

Efficacy of ventilatory support

Improving daytime PaCO2
Improving nocturnal hypoventilation
Ensuring appropriate nocturnal SpO2

No respiratoy events under NIV

No apneas under NIV
Low level of unintentional leaks
Optimal patient ventilator synchrony

Improving prognosis
Reducing respiratory morbidity, and disease-related
burden
Improving survival..
 
 
               Table 1 : NIV: therapeutical goals
 
 
 
 
 
 
 
 
 
 
 
 
 
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Proper to NIV
Unintentional leaks
Decrease in ventilatory drive
Upper airway obstruction
-­‐ With reduced ventilatory drive
-­‐ With increased ventilatory drive

Common to NIV and invasive ventilation

Asynchronies
-­‐ Ineffective effort
-­‐ Double triggering
-­‐ Autotriggering
-­‐ Late cycling
-­‐ Premature cycling

Residual hypoventilation
   
 
 
Table  2:  Respiratory  events  during  NIV  
 
 % time spent in event

ARF CRF LTMV

 
Figure 1: Mean night time spent in respiratory events during NIV in three
different clinical situations: ARF (acute respiratory failure): NIV introduced
in the acute setting, CRF (chronic respiratory failure): NIV initiated
electively, in patients for whom home MV was indicated; LTMV: patients
ventilated at home for at least 3 month (*indicates p <0.05 compared to ARF
group) (from [13])
 
 
 %TSpO2 < 90%
Mean SpO2

% time spent in events % time spent in events

Diurnal PaCO2

% time spent in events

Figure 2 : Correlation between % of time spent in respiratory events during NIV and :
A: mean SpO2 (r = - 0,281, p < 0,001) .
B : % temps spent with SpO2 < 90% (r = 0,365, p <0,001).
C : Diurnal PaCO2 (r = 0,186, p < 0,0393) (from [13])
  

Pressure

Flow

Thorax

Abdomen

SpO2

Leak

Figure 3: Unintentional leaks with a pressure-controlled ventilator (2 ‘ page)

Typical patterns: (1) increase in machine flow amplitude and reduction in
thoracic and abdominal belt signals. (2) mild reduction in inspiratory
pressure. At the bottom of the graph are represented unintentional leaks as
estimated by the ventilator software coupled to the polygraph
 
 
 

Pressure

Flow

Thorax

Abdomen

SpO2
 
  
Figure 4: Unintentional leaks with a volume-controlled ventilator (1’
page): (1) inspiratory pressure is not maintained (full arrow); (2)
inspiratory flow amplitude is maintained with reduction in thoracic and
abdominal belt signals; (3) note the amputation of the expiratory part of
the flow curve (dashed arrow)

Pressure

Flow

Thorax
Abdomen

SpO2

Pressure

Flow

Thorax
Abdomen

SpO2

Figure 5: Decrease in ventilatory drive (pressure controlled ventilation).

(A) In the absence of back-up ventilation: loss of pressure signal, flow signal
and thoracic and abdominal belt signals (2’ page) B) With back-up respiratory rate :
reduction in abdominal and thoracic belt signal amplitudes without phase
opposition.Note the switch to back-up ventilation.
 A

Pressure

Flow

Thorax

Abdomen

SpO2

Pressure

Flow

Thorax

Abdomen

SpO2

Figure 6 Upper airway obstruction (pressure controlled ventilation. (A) Without

reduction of ventilatory drive (1’ page): sudden reduction in flow amplitude; (2)
phase opposition or phase anglein thoracic and abdominal belts; (3) increase in
respiratory rate at the end of the event due to increase in patient efforts to open
the airways; (4) patient ventilator asynchrony.
B) With decrease in ventilatory command (1’ page) : (1) progressive decrease in
flow; (2) pressure signal unchanged; (3) disappearance of thoracoabdominal
movements and switch to back-up respiratory rate without thoracoabdominal
movements indicating recurrent airway closure with reduced ventilatory command:
(4) entirely synchronous resumption of flow and thoracoabdominal movements
without fighting movements.
 A

Pressure

Flow

Thorax

Abdomen

SpO2

Leak

Pressure

Flow

Thorax

Abdomen

SpO2

Leak
 C

Pressure

Flow

Thorax

Abdomen

SpO2

Leak

Figure 7 Example of leak-induced aynchrony (pressure controlled ventilation)

A) Autotriggering (2 ‘ page): autotriggering is defined as the occurrence of at
least three consecutive pressurisations at a ventilator frequency of >40/min not
synchronised with patient inspiratory effort, as defined by Guo et al [26]
B) Double triggering (2’page): double-triggering is defined as two cycles
separated by a very short expiratory time, defined as less than half of the mean
inspiratory time and a concomitant inspiratory activity at the thoraco-abdominal
bands [27]
C) Ineffective inspiratory efforts (1’ page): ineffective inspiratory effort are
defined by the presence of both: a respiratory movement on thoraco-abdominal
bands and/or a positive deflection in expiratory flow [24] (full arrow), without a
concomitant breath delivered by the ventilator (dashed arrow). In this example, as
leakage impede the detection of patient inspiratory effort, the ventilator switch to
backup respiratory rate In each graph, at the bottom of the graph are represented
unintentional leaks as estimated by the ventilator software coupled to the polygraph
References
 
1. Gonzalez MM, Parreira VF, Rodenstein DO: Non-invasive ventilation and sleep. Sleep Med Rev
2002;6:29-44.
2. Gonzalez-Bermejo J, Perrin C, Janssens JP, Pepin JL, Mroue G, Leger P, Langevin B, Rouault S, Rabec
C, Rodenstein D: Proposal for a systematic analysis of polygraphy or polysomnography for identifying
and scoring abnormal events occurring during non-invasive ventilation. Thorax 2012;67:546-552.
3. [Non-invasive positive pressure ventilation for neuromuscular diseases. Practice guidelines]. Rev Mal
Respir 2006;23:14S15-14S40.
4. Nardi J, Prigent H, Adala A, Bohic M, Lebargy F, Quera-Salva MA, Orlikowski D, Lofaso F: Nocturnal
oximetry and transcutaneous carbon dioxide in home-ventilated neuromuscular patients. Respir
Care;57:1425-1430.
5. Paiva R, Krivec U, Aubertin G, Cohen E, Clement A, Fauroux B: Carbon dioxide monitoring during
long-term noninvasive respiratory support in children. Intensive Care Med 2009;35:1068-1074.
6. Contal O, Vignaux L, Combescure C, Pepin JL, Jolliet P, Janssens JP: Monitoring of noninvasive
ventilation by built-in software of home bilevel ventilators: a bench study. Chest;141:469-476.
7. Pasquina P, Adler D, Farr P, Bourqui P, Bridevaux PO, Janssens JP: What does built-in software of
home ventilators tell us? An observational study of 150 patients on home ventilation.
Respiration;83:293-299.
8. Rabec C, Georges M, Kabeya NK, Baudouin N, Massin F, Reybet-Degat O, Camus P: Evaluating
noninvasive ventilation using a monitoring system coupled to a ventilator: a bench-to-bedside study.
Eur Respir J 2009;34:902-913.
9. Berry RB, Chediak A, Brown LK, Finder J, Gozal D, Iber C, Kushida CA, Morgenthaler T, Rowley JA,
Davidson-Ward SL: Best clinical practices for the sleep center adjustment of noninvasive positive
pressure ventilation (NPPV) in stable chronic alveolar hypoventilation syndromes. J Clin Sleep
Med;6:491-509.
10. Rabec C, Rodenstein D, Leger P, Rouault S, Perrin C, Gonzalez-Bermejo J: Ventilator modes and
settings during non-invasive ventilation: effects on respiratory events and implications for their
identification. Thorax;66:170-178.
11. Rabec CA, Reybet-Degat O, Bonniaud P, Fanton A, Camus P: [Leak monitoring in noninvasive
ventilation]. Arch Bronconeumol 2004;40:508-517.
12. Caldarelli V, Borel JC, Khirani S, Ramirez A, Cutrera R, Pepin JL, Fauroux B: Polygraphic respiratory
events during sleep with noninvasive ventilation in children: description, prevalence, and clinical
consequences. Intensive Care Med;39:739-746.
13. Vagner A, Rabec C, Georges M, Monnet C, Kabeya N, Metchedjin A, Camus P: Nocturnal polygraphic
monitoring in patients on noninvasive ventilation: epidemiology of abnormal events, prognostic and
therapeutic implications. Rev Mal Respir 2013;30:A34.
14. Calderini E, Confalonieri M, Puccio PG, Francavilla N, Stella L, Gregoretti C: Patient-ventilator
asynchrony during noninvasive ventilation: the role of expiratory trigger. Intensive Care Med
1999;25:662-667.
15. Carteaux G, Lyazidi A, Cordoba-Izquierdo A, Vignaux L, Jolliet P, Thille AW, Richard JC, Brochard
L: Patient-ventilator asynchrony during noninvasive ventilation: a bench and clinical study.
Chest;142:367-376.
16. Racca F, Squadrone V, Ranieri VM: Patient-ventilator interaction during the triggering phase. Respir
Care Clin N Am 2005;11:225-245.
17. Lujan M, Sogo A, Pomares X, Monso E, Sales B, Blanch L: Effect of leak and breathing pattern on the
accuracy of tidal volume estimation by commercial home ventilators: a bench study. Respir
Care;58:770-777.
18. Carlucci A, Pisani L, Ceriana P, Malovini A, Nava S: Patient-ventilator asynchronies: may the
respiratory mechanics play a role? Crit Care;17:R54.
19. Douglas NJ, White DP, Pickett CK, Weil JV, Zwillich CW: Respiration during sleep in normal man.
Thorax 1982;37:840-844.
20. Jounieaux V, Aubert G, Dury M, Delguste P, Rodenstein D: Effects of nasal positive pressure
ventilation on the glottis in normal asleep subjects. J Appl Physiol 1995;79:186-193.
21. Parreira VF, Jounieaux V, Aubert G, Dury M, Delguste PE, Rodenstein DO: Nasal two-level positive-
pressure ventilation in normal subjects. Effects of the glottis and ventilation. Am J Respir Crit Care
Med 1996;153:1616-1623.
22. Adler D, Perrig S, Takahashi H, Espa F, Rodenstein D, Pepin JL, Janssens JP: Polysomnography in
stable COPD under non-invasive ventilation to reduce patient-ventilator asynchrony and morning
breathlessness. Sleep Breath;16:1081-1090.
23. Crescimanno G, Canino M, Marrone O: Asynchronies and sleep disruption in neuromuscular patients
under home noninvasive ventilation. Respir Med;106:1478-1485.
24. Fanfulla F, Delmastro M, Berardinelli A, Lupo NDA, Nava S: Effects of Different Ventilator Settings
on Sleep and Inspiratory Effort in Patients with Neuromuscular Disease. Am. J. Respir. Crit. Care Med.
2005;172:619-624.
25. Fanfulla F, Taurino AE, Lupo NDA, Trentin R, D'Ambrosio C, Nava S: Effect of sleep on
patient/ventilator asynchrony in patients undergoing chronic non-invasive mechanical ventilation.
Respiratory Medicine 2007;101:1702-1707.
26. Guo YF, Sforza E, Janssens JP: Respiratory patterns during sleep in obesity-hypoventilation patients
treated with nocturnal pressure support: a preliminary report. Chest 2007;131:1090-1099.
27. Vignaux L, Vargas F, Roeseler J, Tassaux D, Thille AW, Kossowsky MP, Brochard L, Jolliet P:
Patient-ventilator asynchrony during non-invasive ventilation for acute respiratory failure: a multicenter
study. Intensive Care Med 2009;35:840-846.