The Science Journal of the American Association for Respiratory Care

2005 OPEN FORUM Abstracts


Dave Crotwell, RRT, Cary Jackson RRT, John Salyer RRT. Respiratory Care Department. Children's Hospital and Regional Medical Center, Seattle WA.

Introduction: Disposable flow sensors for the Viasys Avea are costly ($40 each). The reusable sensor is less costly per use, but prone to inaccuracy due to accumulation of condensate & secretions during use. The disposable sensor is less prone to these issues. We sought to determine if disposable sensors could be reprocessed and retain accuracy and precision.

Methods: Five disposable neonatal variable orifice flow sensors were gathered after a single clinical use. Cleaning procedure was provided by our outside recycling vendor: 1) pressure sensing lines & sensor inner surface were flushed with 70% isopropyl alcohol then flushed with sterile H2O, 2) exterior surfaces were wiped the alcohol solution, 3) sensor placed in vacuum sealed bag, 4) bag placed in pasteurizer at 160° F for 30 minutes, 5) sensor then in air dried for 1 hour at 100° F. This was repeated 5 times. Sensors were then returned to us. We attached each recycled sensor to Viasys Avea. A calibration syringe (Hans-Rudolf) was set at 2 mL and attached to the sensor and held in a vice. The volume in the calibration syringe had been verified to within 0.1 mL, using a finely graduated glass cylinder to measure the water displaced by the syringe when set at 2 & 10 mL. The plunger was activated manually 10 times and exhaled VT (VTE) displayed by the ventilator was recorded. This process was then repeated with the syringe volume set at 10mL. The entire testing procedure was then repeated on 5 new sensors. Data for recycled sensors and new sensors were pooled respectively. Descriptive statistics were computed. Mean values for volumes were compared using Mann Whitney-U test for unpaired means with significance established as P< 0.05. Percent error was calculated as [(true value-observed value)/true value], using the true values of 2 & 10 mL respectively, and the observed values being the means for VTE.

Results: The table shows statistically significant differences between mean VTE from the new versus the recycled sensors. Single sample means tests using hypothesized means of 2 & 10 mL respectively revealed that at 2 ml, the recycled sensor was not statistically different than the true value for VT (P = 0.32), while new sensor was statistically different (P< 0.001). Percent error for new sensors at 2 & 10 mL were 15% & 3%, and for recycled sensors were 0% and 4%.

Discussion: Statistical significance and clinical importance can be confused. We do not consider these differences clinically important. The standard deviations were larger in both recycled sensor sets, but clinical decisions about VT are not made on the volumes from individual breaths but instead on multiple breaths over time. It is surprising that at the low volume test, the mean for the recycled sensor was closer to the true mean than the new sensor. Potential weaknesses in our design include manual movement of the plunger to simulate tidal breathing. It is difficult to keep plunger speed constant manually, and thus flow rates may have been erratic, which could affect measurement accuracy if flows exceeding the linearity of the sensor. We also tested the sensors at ambient temperature and humidity. Mimicking heated humidified conditions encountered clinically may change the results also. We intend to study this issue further, but it may be possible to reprocessing single use sensors with the Avea, and reduce costs of mechanical ventilation.
You are here: » Past OPEN FORUM Abstracts » 2005 Abstracts »