© 2002 American Thoracic Society
When Ventilator and Patient's End of Inspiration Don't CoincideWhat's the Matter?Henri Mondor Hospital University of Paris 12 Paris, France During assisted mechanical ventilation, the time at which the ventilator terminates inspiration and opens the exhalation valve defines the beginning of the expiration process, for the ventilator, but also for the patient. In this issue of the AJRCCM, Younes and collaborators (pp. 21–30) assessed the impact of prolonging the inspiratory time of a ventilator breath beyond the end of the patient's inspiratory time (i.e., after the end of neural inspiratory time) (1). The question addressed here was therefore: what is the response of the patient's breathing controller when the onset of the natural (neural) expiration is impeded? Do the respiratory centers prolong expiratory time to defend against the risk of dynamic hyperinflation? Are expiratory muscles going to be activated to accelerate expiration? This was a difficult experiment to conduct and an important question to address. Why is this an important question? Many attempts have been previously made to design new modes of ventilation with optimal synchrony between a patient's inspiratory effort and assistance from a ventilator. The onset of the breath has been the subject of much attention on optimizing trigger sensitivity. Also, the time elapsing before recognition of patient's effort in cases of dynamic hyperinflation and auto or intrinsic positive end-expiratory pressure has been addressed extensively (2). The criteria used for terminating the breath, also called the cycling criteria, have been much less oriented toward a specific recognition of the end of a patient's effort. Up to now, the most frequently used mode of ventilation, assist-control ventilation (3), terminates the breath on a time criterion independent of a patient's effort. What the ventilator primarily controls is flow, and insufflation time depends on the peak flow setting. The total inspiratory time can also be prolonged by the addition of a plateau at end-insufflation. Another popular mode of ventilation is pressure support ventilation. This mode proposes an original solution for terminating the breath closer to a patient's neural signal than would be achieved using a preset inspiratory time (4). Indeed, the decelerating flow signal is used to determine the time at which the ventilator switches to expiration. Because the inspiratory flow should be influenced by the patient's effort at any time of the breath, this criterion is influenced by a signal coming from the patient. It has, however, long been recognized that this off switch criterion is influenced by complex interferences: time constant of the respiratory system, the value of flow used as a threshold criterion, the level of pressure applied at the opening of the respiratory system, and the remaining inspiratory effort at the end of the breath (5). Many examples of dysynchrony occurring at the end of the breath have been described, especially for high levels of pressure support ventilation and in patients with high respiratory resistance. Therefore, for the two most popular modes of assisted ventilation, the question of inadequate timing addressed in the study by Younes and coworkers is very relevant (1). As explained previously here, this inadequacy may influence the level of dynamic hyperinflation and thus interact with the amount of patient's inspiratory effort. Could every effort made to optimize the triggering systems and the adequacy of inspiratory support be ruined by neglecting the timing of end-inspiration? The question was difficult to address precisely because the neural inspiratory time cannot be controlled or even predicted with standard assisted modes of ventilation. Younes and coworkers (1) took advantage of the unique feature of proportional assist ventilation to terminate the inspiratory support at the exact end of patient's inspiratory neural time (or flow). They intermittently added a pause (occlusion) of variable duration at the end of inspiration, and assessed its consequences by matching the studied breaths with similar breaths without a pause. Ironically, the results of this study are very relevant for standard modes of assisted ventilation, whereas they are only of marginal interest for proportional assist ventilation, in which this situation is rather artificial. The authors found that most patients (45 of 50) lengthened their expiratory time because of the delayed onset of expiration. In other words, their respiratory centers activated the inspiratory muscles with a slight delay compared with a nonoccluded breath. This time compensation, however, represented a highly variable and often very limited fraction of the delay imposed by the occlusion. Patients with chronic obstructive pulmonary disorder and expiratory flow limitation, who are more prone to develop hyperinflation, had a very poor compensatory response. Lastly, the authors did not observe any substantial expiratory muscle activation, at least as far as they could measure it. The study has limitations. We do not know how, for instance, the acute response to delayed expiration gets modified over a more prolonged period of time. After a change in ventilator settings, it may take more than a few breaths to find a new equilibrium, because many different stimuli can be progressively involved (6). This study tells us, however, that delayed expiration, as frequently observed with inadequate settings during assisted ventilation, may influence the level of dynamic hyperinflation and worsen patient-ventilator interaction. It always takes time to explain to students why obstructive lung diseases, which are primarily defined as an expiratory problem, cause enormous problems to patients during inspiration. This study by Younes and coworkers (1) illustrates the same kind of relationship between expiration and inspiration. It shows that failing to detect precisely the correct onset of expiration may interact with the next inspiration and put the patient at much higher stress to inspire. Finally, to determine the true onset of inspiratory effort, one of the investigators analyzed each breath by looking for the abrupt deflection observed on the flow and airway pressure tracings. In most patients, this clearly occurred before the ventilator was able to detect patient's effort, because of the presence of dynamic hyperinflation. This well-known phenomenon has prompted other investigators to design complex electromyographic recording systems to trigger the ventilator (7). It is striking to see that the same information is conveyed by a careful examination of flow and airway pressure tracing signals available in every ventilator. To move from readings performed by an investigator with the skill, experience, and background of Younes to an automatic recognition of these signals by a specially designed algorithm is not simple. This study shows, however, that this important information is available noninvasively at the bedside in most ventilated patients. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
