2004 OPEN FORUM Abstracts
COMPARISION OF DUAL MODE VENTILATION AMONG SELECTED ADULT CRITICAL CARE VENTILATORS USING A PROGRAMMABLE LUNG SIMULATOR
Alotaibi, MS, RRT (University of Medicine and Dentistry of New
Jersey, Newark, NJ); Robert Kacmarek, PhD, RRT, FAARC (Massachusetts
General Hospital and Harvard Medical School, Boston, MA); Craig
Scanlan, EdD, RRT, FAARC (UMDNJ, Newark, NJ); Al Heuer,
PhD, RRT (UMDNJ, Newark, NJ).
BACKGROUND: Dual mode ventilation (DMV) is a variant of PCV in which the inspiratory pressure is adjusted breath-to-breath to accommodate changes in respiratory mechanics, with the goal a more stable minute volume. Currently, there are several adult critical care ventilators that provide DMV. Although it is assumed that this mode operates in a similar manner among these ventilators (Branson, Johannigman, Campbell & Davis, 2002), no one has empirically tested this proposition. To address this lack of knowledge, we compared the performance of four adult critical care ventilators set to deliver DMV using a programmable lung simulator. We hypothesized no significant differences among ventilators on any of 11 measures of respiratory mechanics during conditions of varying patient demand and impedances.
METHOD: The Active Servo Lung 5000 (Ingmar Medical, Pittsburgh, PA) simulator was programmed to produce 5 breathing conditions (a) normal resistance and compliance (baseline), (b) low compliance, (c) high resistance, (d) high patient demand, and (e) simulated leak. The DMV feature of the Puritan-Bennett 840 (PB840), Hamilton Galileo (Galileo), Drager Evita XL (Evita), and Siemens Servo 300 (Servo) ventilators was implemented using the following settings VT = 600mL, Ti = 1 sec, PEEP = 4 cmH2O, flow trigger = 2 L/min, and most rapid rise time. Ventilation parameters were acquired, stored, and analyzed using the ASL5000 software. Dependent variables included inspiratory trigger airway pressure (Pawt), trigger delay time (Dtrig), post-trigger delay time (Dptrig), total trigger delay time (Dtot), patient inspiratory work (Winsp), patient trigger work (Wtrig), inspiratory trigger muscle pressure (Pmust), inspiratory time 90% (IT90), VT, PIFR, and PIP.
RESULTS: Due to the high precision of the ASL 5000 measurements, we adopted a decision rule requiring variations in excess of 2 SD from the group mean to be considered of ‘practical’significance. At baseline, the Evita imposed higher Wtrig and Dtot than the other ventilators. When compliance was low, the Galileo exhibited longer delay times and higher work parameters that its counterparts. When resistance was high, the Evita imposed higher work and longer delay times than the Galileo, Servo or PB840. When patient demand and system impedances were increased, both the Galileo and Evita imposed higher Wtrig and Pmust than either the PB840 or the Servo. In comparison to the other ventilators, the Servo underperformed (more work and less synchrony) during the simulated leak test. The Galileo had the highest VT delivery under all conditions.
CONCLUSIONS: The assumption that DMV operates in a similar manner across adult critical care ventilators is not supported. The PB840 appears to offer the best overall synchrony and lowest imposed work. The Galileo and Evita imposed more work and exhibited variable synchrony under different tested conditions. These differences in performance during DMV are likely due to variations in the underlying algorithms used by each ventilator and/or the technical attributes of their sensor and measurement systems.