© 2003 American Thoracic Society
Disrupted Sleep during Mechanical VentilationTo the Editor:We read with interest the study by Parthasarathy and Tobin (1) on the effects of the ventilator mode on sleep quality in patients submitted to invasive mechanical ventilation. Sleep fragmentation was significantly greater during pressure support ventilation (PSV in spontaneous mode) than during assist-control ventilation (ACV). Six of 11 patients developed central apneas during PSV but not during ACV. The authors stated that pressure support in the spontaneous mode may aggravate breathing instability and sleep fragmentation at commonly used ventilator settings, insisting on the incorrectness of setting a ventilator without taking into account the differences between sleep and wakefulness. In 1996, we reported central apneas in normal subjects mechanically ventilated with a two-level positive pressure ventilator in spontaneous mode (2). Sleep was fragmented and apneas were very frequent. In further studies, we found that central apneas and breathing instability were much more frequently observed during sleep than during wakefulness and were abolished when using the controlled mode. We concluded that the spontaneous-timed mode with a back-up frequency was preferred to the spontaneous mode (3). We also stated that ventilator settings adapted to the wakefulness state could be inappropriate during sleep (4). Although the differences between invasive and noninvasive ventilation are enormous and cannot be overlooked, it is striking to realize that the effects of the ventilator mode that we described in normal subjects are also observed in real life patients using invasive ventilation. In the study by Parthasarathy and Tobin (1), apneas were mainly seen in patients with heart failure, whereas in our studies, almost all normal subjects presented them while asleep. In a sense, it looks like invasive ventilation "protects" against apneas unless patients are in heart failure. This may be because of the stimulation of upper airway receptors by the tracheal tube, leading to the high mean inspiratory flow rates (indicating a great respiratory drive) seen by Parthasarathy and Tobin in their patients without apneas. Alternatively, altered respiratory mechanics in patients, but not in normal subjects, may have decreased the mechanical efficiency of assisted ventilation, leading to less hypocapnia and hence to a lesser propensity for central apneas. It seems important to call the attention of physicians working with mechanical ventilation, invasive and noninvasive, to the issue of breathing and sleep instability. Parthasarathy and Tobin rightly point to the potential deleterious consequences of central apneas and sleep fragmentation that should no longer be disregarded.
a Universidade Federal de Minas Gerais Belo Horizonte, Brazil REFERENCES
From the Authors: We agree with Parreira and colleagues that sleep disruption secondary to breathing instability during mechanical ventilation is an important problem. In our article (1), we cited two reports of central apneas occurring in healthy subjects during pressure support (2, 3). We regret not citing the studies of Parreira and colleagues that have also shown this effect in healthy subjects. Parreira and colleagues point out that almost all of their subjects developed central apneas while receiving pressure support, whereas this occurred in only 6 of our 11 subjects. The subjects of Parreira and colleagues were ventilated via a mask, whereas all our subjects had an endotracheal tube in place. It is naturally tempting to attribute the less frequent central apneas to the endotracheal tube. By what mechanism this could arise, however, is not clear to us. Indeed, a study in animals (4) suggests that irritation of the trachea can generate central sleep apneas through vagal mechanisms. We suspect the main reason our results differ from those of Parreira and colleagues is that we studied critically ill patients, whereas they studied healthy subjects. A given level of pressure support is more likely to cause arterial carbon dioxide to fall below the apnea threshold, and thus induce central apneas, in healthy subjects than in patients with respiratory disease. Critically ill patients also have less likelihood of developing central apneas because of decreased respiratory compliance, increased airway resistance, greater dead space, and greater respiratory drive, all of which can stabilize breathing (5, 6). Conversely, conditions such as heart failure predispose to central apneas. Before our study, it was possible only to speculate on the net effect of these countervailing factors in critically ill patients. We now know that central apneas and disrupted breathing is likely to occur in a substantial proportion of critically ill patients receiving pressure support.
Loyola University Medical Center Maywood, Illinois Edward Hines VA Hospital Hines, Illinois FOOTNOTES Conflict of Interest Statement: S.P. has received approximately $3,000 from Glaxo SmithKline Pharmaceuticals over the past 3 years toward speakers fees, and has received $1,300 from Orphan Medical, Inc., toward conference attendance. M.J.T. has no declared conflict of interest. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
