We thank Dr. Romanoff for his interest in our article. First, we will clarify how the modulation program controls the respiratory cycle. Dr. Romanoff is correct in his understanding that the computer changes tidal volume (VT) and respiratory rate (RR) according to a file of 376 sequential (not random) settings of normalized rate. The operator sets minute ventilation and RR while in control mode. VTchanges are reciprocal to those of RR as the ventilator functions as a volume divider. Mean VT(VTm) is the same as in control mode but has a range of VTmin= 0.64VTmand VTmax= 1.46VTm. The duty cycle of the ventilator (TT) is fixed while in variable ventilation mode at TI:TE(inspiratory time:expiratory time) equal to 1:2. For any given breath within the 376-breath file, the actual respiratory cycle duration = 60/TTwith TI= 60/TT/3 and TE = 60/2TT/3. Dr. Romanoff’s calculations for the duration of inspiration and expiration for the largest and smallest breaths are correct. He errs, however, in the assumption that TE(or TI, for that matter) is influenced by subsequent breaths. As the above equation shows, respiratory cycle time for any breath varies only according to VT, and every breath at that same VTwill have the same TIand TE, independent of succeeding or preceding breath size. The “file reading” time for the microprocessor is less than 1 ms as it “reads” the normalized variability file with no discernible effect on duration of either TIor TEat any rate.

Could there be an effect of “stacking” leading to “auto–positive end-expiratory pressure” (auto-PEEP) in some of the patients in our study? Certainly this is possible, especially in the case of a series of rapid cycling breaths—when RR is high. However, in this study, the mean RR was set to 10 breaths/min, so the most rapid instantaneous RR was only 14.6 breaths/min. Under the conditions of our experiment, lung function was expected to be abnormal, so it is possible that, as Dr. Romanoff states, there may be alveoli with slow time constants that develop an auto-PEEP effect leading to improved oxygenation. Our protocol did not provide for extrinsic PEEP, and we did not see evidence of auto-PEEP in post hoc  analysis. Previous work has not demonstrated any auto-PEEP by variable ventilation in animal models.1If such an effect is occurring, it is at low levels not associated with increases in mean intrathoracic pressures leading to the deleterious effects on cardiac output and oxygen levels seen with higher levels of extrinsic PEEP.

* University of Manitoba, Winnipeg, Manitoba, Canada. gdoak@hsc.mb.ca

Boker A, Graham MR, Walley KR, McManus BM, Girling LG, Walker E, Lefevre GR, Mutch WA: Improved arterial oxygenation with biologically variable or fractal ventilation using low tidal volumes in a porcine model of acute respiratory distress syndrome. Am J Respir Crit Care Med 2002; 165:456–62