The Science Journal of the American Association for Respiratory Care

2012 OPEN FORUM Abstracts

COMPARATIVE RESPIROMETER STUDY FOR VOLUMETRIC ACCURACY ACROSS A PHYSIOLOGIC RANGE OF FLOW RATES.

Kevin Crowley, George Blaisdell, John Dziodzio; Pulmonary and Critical Care Medicine, Maine Medical Center, Portland, ME

Background: The frequent need to assess exhaled volume for non-ventilated patients is convieniently met with volumetric devices called respirometers, which have been in use for decades. Having now both mechanical and electronic variants of these devices, it was desired to validate their accuracy and to understand their limitations to best apply them in the clinical context of non-invasive ventilatory assessment. Method: The study was conducted with an experimental apparatus consisting of three in-line volumetric devices: the Wright Respirometer (Wright Model 4,Nspire,Inc), a Digital Respirometer (Anesthesia Associates, Inc. Model 00-295), and a calibrated spirometer. Ninety-nine serial three- liter syringe volumes where delivered through the devices at a variety of flow rates in representative physiologic ranges, and the volume from each device, as well as a mid-expiration flow measurement (FEF25-75), serving as proxy for the mean flow rate, were recorded. Mean error and standard deviation was assessed for the instruments in their overall response, and also separately in low-flow and high-flow ranges. Results: Both devices suffer from a drop off in sensitivity at low flow rates, especially under 1.2 LPS. Although the Wright displays a mean error of about 0%, the large 8.1% confidence interval (95% CI) allows for large deviations around that mean, while the digital device, with a smaller 6% CI, demonstrates a much larger fixed bias (mean error) of 5%. In the higher flows, there is a much more linear response, with the Wright displaying a high precision of volumetric determination (1.6%CI), with a fixed bias ( mean error) of about 5%. In the same range the digital device disappointingly demonstrates a fixed bias of almost 10%, with imprecision almost twice as large (3.4%CI) as the Wright. Conclusion: Successful application of these respirometers in the clinical context may be achieved only by application of a fixed correction factor, of about 5% for the Wright, and about 10% for the Digital, with the assumption that the measured flowrates over the expiratory phase typically exceed 1.2-1.5 LPS. Sponsored Research - None