The Science Journal of the American Association for Respiratory Care

2010 OPEN FORUM Abstracts

EFFECT OF SIGNAL LOSS ON RESPIRATORY RATE RECORDING WITH A CLINICAL OXYGEN DOSE RECORDER

Lauren R. Furnkase1, Dana E. Strollo1, Teresa A. Volsko1, Kathryn A. Tessmer2, Rachael J. Polhle-Krauza3,4, Michelle L. McCarroll4; 1Heatlh Professions, Youngstown State University, Youngstown, OH; 2Human Performance and Exercise Science, Youngstown State University, Youngstown, OH; 3Human Ecology, Youngstown State University, Youngstown, OH; 4Summa Health System, Akron, OH

BACKGROUND: The Clinical Oxygen Dose Recorder (CODR) is a non-invasive monitor which continuously measures respiratory rate (RR), heart rate and oxyhemoglobin saturation. This study sought to determine if signal loss affected respiratory rate averages calculated from recorded data. We hypothesized that signal loss would contribute to inaccurate RR calculations. METHODS: Adult patients who had been referred by a physician to a pulmonary rehabilitation program were recruited. A sampling cannula was used to connect patient to the CODR, and either a continuous or pulse-dose administration of oxygen, in accordance with manufacturer recommendations. A dual lumen cannula was used for those receiving oxygen by continuous flow and single lumen cannula for patients receiving oxygen by pulse-dose. Data were recorded over the course of 15 visits. Descriptive statistics were used to report mean values RR and signal loss. Clipped (without zero values) and unclipped RR data for each delivery system, continuous and pulse dose, were compared with Students T-test. Statistical significance was established at p < 0.05. RESULTS: CODR data were collected from four patients at a sampling rate of 20 milliseconds. Two of the patients received oxygen by pulse dose, and two by continuous flow. A total of 65,528 data points were collected by the CODR for the total cohort, 37,300 of which were from those receiving oxygen by pulse-dose. The percent signal loss for patients receiving oxygen therapy by continuous flow was 6.2%, and twice the rate of data collected from the pulse dose group. A value of zero was recorded for each signal loss. Upon review of aggregate data it was noted that a high value for the RR, between 40 and 100, was recorded immediately prior to every signal loss. Mean values for the unclipped data were lower for both the pulse dose 33.2, +SD 16.0, and the continuous flow 25.1, + SD 14.1 compared to clipped pulse dose 34.2, + SD 15.1 and continuous flow 26.8, + SD 12.9. Comparison of clipped and unclipped data for each type of oxygen delivery system was statistically significant, p < 0.001. CONCLUSIONS: Care should be taken to review RR data and determine signal loss occurrence. The inclusion of signal loss values, recorded as no detectable breath, may provide erroneously lower respiratory rates. Clinicians should evaluate the effect loss signal has on data prior to clinical decision making. Sponsored Research - None