There are two fundamentally different ways to ventilate critically ill patients: constant flow, volume-preset modes (such as volume assist-control) and pressure-preset modes (such as pressure-control and pressure-support). Critically ill patients suffer mechanical derangements of the respiratory system that raise the work of breathing. Knowledge of these mechanical properties is useful diagnostically and as a measure of response to treatment over time. It has been proposed that only constant flow, volume-preset modes are able to offer diagnostic information about the changes in the subject's lungs in terms of resistance and elastance properties. This study proposes to examine if similar information can be extracted from pressure-preset modes by comparing information from both modes of ventilation.

Aim 1: To compare the respiratory system resistance and elastance obtained during constant-flow, volume-preset ventilation (using conventional means) and during pressure-preset ventilation (by analyzing the slope of the flow versus time waveform, as described below).

Aim 2: To determine whether patient effort and level of alertness impair the accuracy of resistance and elastance measurements during pressure-preset ventilation.

Hypothesis 1: Our primary hypothesis is that the flow versus time waveform contains information sufficient to calculate the respiratory system resistance and elastance. To test the primary hypothesis, we propose to measure resistance and elastance of subjects ventilated in the ICU during assist-control ventilation (a standard constant flow, volume-preset mode). Then we will record the flow versus time waveform during pressure-preset ventilation. By extrapolating the flow versus time waveform (which is generally linear) to the time axis, one can calculate elastance since at zero flow, the alveolar pressure equals the ventilator inspiratory pressure. Then Ers = (Pinsp - Total PEEP)/Extrapolated VT, where Pinsp is the set inspiratory pressure and extrapolated VT is the tidal volume if inspiratory time had been sufficient to allow equilibration between patient and ventilator (using trigonometry). Similarly, by extrapolating the flow versus time waveform to the flow axis (to find the maximal flow), one can calculate the resistance, assuming that flow depends on the pressure difference between ventilator and patient and the square of the resistance. We will compare the values derived during pressure-preset ventilation with those determined during assist-control (taken as the true values).

Hypothesis 2: We hypothesize that inspiratory effort will be sufficient in some subjects to distort the flow versus time waveform from that which would be seen if the patient were passive, leading to erroneous values for resistance and elastance. We will estimate the respiratory drive using a standard measure, the fall in Pao during a brief inspiratory occlusion 100ms following the onset of inspiration (P0.1). Further, we will measure each subject's alertness on the Richmond Agitation-Sedation Scale (RASS). We expect our estimations of resistance and elastance to less accurate (during pressure-preset ventilation compared with assist-control) in subjects with greater respiratory drive and higher levels of alertness.

Inclusion Criteria:

• Subjects may be included in the study if they meet the following inclusion criterion and no exclusion criteria:
• Adult medical or surgical ICU patients who are mechanically ventilated.

Exclusion Criteria:

• Subjects with planned extubation at time of screening
• Subjects who are judged too hemodynamically unstable by their attending physician to participate in the study
• Subjects on high-frequency ventilation
• Subjects whose oxyhemoglobin saturation cannot be maintained at 88% or greater
• Subjects with large air leaks around the endotracheal or tracheostomy tube
• Subjects younger than 18 years old
• Inability to obtain informed consent from the subject or the subject's authorized representative

Inclusion Criteria

Subjects may be included in the study if they meet the following inclusion criterion and no exclusion criteria:

[1] Adult medical or surgical ICU patients who are mechanically ventilated. Using a random number generator each subject will be assigned to pressure control or volume control for initial set of measurements.