Since the Hering-Breuer reflex is variably active in humans, ventilator-delivered tidal volume variation could affect Voff in some individuals. Tidal volume delivered during one breath would affect the subsequent breath, altering Voff. Our analysis relies heavily on meaned data, reducing the chance of showing a statistically significant effect of tidal volume (or flow rate) on Voff. Because of this, although we were unable to demonstrate it, the possibility of delivered tidal volume affecting Voff still exists. A combined effect of tidal volume and flow rate variation greater than their individual effects may exist as well. A larger number of patients would be necessary to address these issues.

We noted variability in the Voff both within and between patients. This variability may be attributable in part to the central respiratory drives response to stimuli such as stretch receptor input related to pulmonary mechanical properties, body temperature, and cortical and chemoreceptor input. An effect on Voff from both elastic and resistive loading has been suggested in animals; the vagal influence on medullary control of inspiratory off-switch has been shown to be greater in restrictive than in obstructive disease. The peak inspiratory electromyographic activity precedes maximal volume during spontaneous breathing; the amount of this lag has not been determined in humans but is around 200 to 300 ms in cats and increases with increasing airway resistance. Cortical input may be reduced by random sequencing of tidal volume and flow rate changes; our patients were acclimated to the ventilator, relaxed, and accustomed to changes in their ventilator settings. Variability in the Voff response to ventilator inflations between patients is, therefore, not unexpected. Intrapatient differences in Voff were reduced by expressing Voff as a fraction of spontaneous tidal volume.

Since diaphragmatic electromyographic activity indicates diaphragmatic contractile activity and tension development during assisted breaths, certainly metabolic “work,” and probably mechanical work, is being performed. Marini et al demonstrated that ventilator-determined inspiratory flow rate significantly influences a patients mechanical work performed during assist-control ventilation, with increasing flow rates decreasing inspiratory work. Our data also suggest that even high flow rates may not completely rest the respiratory muscles during assist-control ventilation. In addition, diaphragmatic contraction and performance of work may be seen with controlled ventilation at inappropriate tidal volumes and flow rates. Because the ventilator performs all of the inspiratory work of breathing after Voff, the quantity of mechanical work performed during assisted ventilation is likely dependent on tidal volume as a portion of the total minute ventilation as well as ventilator-delivered flow rate. At low flow rate and tidal volume settings, the diaphragm may contract against a ventilator-imposed inspiratory load, increasing inspiratory work. The diaphragm should be rested more effectively during controlled ventilation than with assist-control, as there is no electromyographic activity during controlled ventilation as long as delivered tidal volume and flow rate are above the patients spontaneous tidal volume and flow rate. Alternatively, diaphragmatic contraction beyond triggering the ventilator may be beneficial and facilitate weaning from mechanical ventilation. Neuromuscular control loop integrity would be maintained, perhaps forestalling muscle atrophy and asynchronous breathing described in patients ventilated for prolonged periods.