2003 OPEN FORUM Abstracts
EFFECTS OF HE/O2 MIXTURES ON THE PERFORMANCE OF SIEMENS SERVO 300 AND SERVO I VENTILATORS.
Steven T. Polston, RRT, Norton Healthcare, Inc., Kosair Children's Hospital, Louisville, KY.
BACKGROUND: This study was an investigation of how 50 psi 80/20 He/O2 (heliox) replacing 50 psi 79/21 N2/O2 (air) as source gas would effect these ventilators performance under test conditions.
METHOD: He and O2 tanks were adapted to a blender then to an air quick connect coupler. The output gas was analyzed and adjusted to provide an 80/20 He/O2 source mixture throughout the test. A Servo 300 and a Servo I ventilator were selected randomly from our in-house fleet. A simple circle test circuit was made using two Siemens 190 test lungs and 22mm tubing. A Novametrix Ventrak 1550 ventilation monitor placed inline between the circuit and test lungs was used as an external reference sensor. O2 % was monitored with the ventilators' internal and an external oxygen analyzer. The vents were set to a Rate of 20, VT 200 ml, PEEP of 5 cm/H20, Pause time % 10, Rise time % 5, and the trigger to -2. Pressure above PEEP was set to 7 in the PC mode to equal the VC and PRVC pressure. Data was collected from the ventilators internal systems and the Ventrak for VC, PC, and PRVC ventilation modes. Data was also collected with the ventilator in the 21% setting supplied first with 50-psi air then switched to 50-psi 80/20 He/O2, then with the ventilators set at 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 100%. The Ventrak was programmed for each gas composition tested.
RESULTS: The data collected was examined in relation to our facilities practice of when a ventilator is removed from service because of suspect calibration; Generally, 10 % + of the set value compared to the measured value of an expected result. The internal and external O2 analyzers were within tolerances through all FiO2's tested. The PIP was out of range by 2 cm H2O for the S300 VC mode at the highest concentration of He but recovered as FiO2 increased. Pause, Mean and PEEP remained stable as established or set through out the test. Ventrak measurements paralleled this. The output VTi was as set for the VC and PRVC mode in both ventilators. The VTe read low on a curve in the same modes and for both ventilators. The greatest variance was with the greatest concentration of He and recovered as the He concentration decreased. The VTe read high on a curve in the PC mode in the same manor with the PIP remaining as set. Static and dynamic compliance of the test system was much lower with the highest concentrations of He. Recovery followed curves as the He concentration decreased. Peak expiratory flow measured by the Ventrak was higher at the greater concentrations of He. The Servo I had the highest PEF in the PC mode, 85 LPM.
CONCLUSIONS: The Servo 300 and Servo I ventilators controls pressure, VTi, and FiO2 within acceptable accuracy. They cannot read VTe with the same level of confidence. Changes in the measured values are curved, not linear like the set output values. The Ventrak measured lower but showed similar results for all parameters. When the He concentration decreases as FiO2 increases, accuracy recovers. The Servo 300 and Servo I both recovered to within 10% accuracy when the ventilator FiO2 was between 40% and 60%. Running the Servo 300 and Servo I with 80/20 He/O2 instead of air should be done with great care and the understanding some displayed values are false. Delivery of He through the ventilator this way is also with the understanding that the air safety system is by-passed. Mixing He and O2 by blender to produce 80/20 He/O2 is not recommended because of the risk of accidental suffocation.