2009 OPEN FORUM Abstracts
COMPARISON OF CAPNOGRAPHY DERIVED RESPIRATORY RATE ALARM FREQUENCY USING THE SARA ALGORITHM VERSUS A ESTABLISHED NONADAPTIVE RESPIRATORY RATE ALARM MANAGEMENT ALGORITHM IN BARIATRIC SURGICAL PATIENTS
Steven Hockman1, Troy Glembot2, Kathleen Niebel3; 1Department of Respiratory Care, Winchester Medical Center, Winchester, VA; 2Department of Bariatric Surgery, Winchester Medical Center, Winchester, VA; 3Department of Clinical Research, Oridion Capnography, Inc, Needham, MA
BACKGROUND: Obstructive Sleep Apnea and Obesity Hypoventilation Syndrome are common in severely obese patients and often exacerbated by sedatives and narcotics. At our institution, continuous capnography and pulse oximetry were mandated as a standard of care on the sub-acute Bariatric Unit. We tested if alarm frequencies from a conventional alarm (CONV) differed from the Smart Alarm for Respiratory Analysis (SARA), an FDA-cleared alarm management algorithm designed to improve alarm vigilance among the clinical staff by diminishing the frequency of clinically insignificant respiratory rate (RR) alarms caused by talking, snoring, and cardiogenic artifacts superimposed on the CO2 waveform. The purpose of this clinical evaluation is to determine if SARA reduces the frequency of clinically insignificant RR alarms. METHODS: The RR alarm frequencies for two alarm algorithms were compared sequentially in 24 participants with BMI â¥ 40 but less than 400 pounds (n = 12 per group) as part of an ongoing trial. Respiratory alarm frequency data and patient data (etCO2, RR, FiCO2, SpO2, PR) were collected using the BernoulliÂ® MSM central station wireless connected to bedside Capnostreamâ¢ 20 monitors (Oridion Capnography Inc.) on patients monitored more than seven hours. Capnographic RR alarm threshold settings were fixed in both groups at a High RR of 36 and a Low RR of 8. The Mann-Whitney U-test for nonparametric comparisons was used to compare the frequency of RR alarms in the two algorithm groups. RESULTS: Review of the patient parameter data and alarm event log indicated that no significant ventilation adverse events were missed in either RR alarm algorithm evaluation group. There was a highly significant difference between SARA and CONV for the High Alarm Frequency (U = 136, p< 0.001) but no significant difference between the CONV and SARA Low Alarm Frequency (U = 76.5, pâ¥0.05). CONCLUSIONS: When using capnography to monitor ventilation, the SARA alarm management algorithm triggered the High RR alarm significantly less frequently than the CONV for clinically insignificant events (false alarms) as noted by the corresponding patient data. The limitation of not being able to silence audible alarms did not allow for evaluation of RR alarm frequency with both algorithms in parallel and may have influenced the frequency of Low RR alarms by diminishing patient uninterrupted rest. Sponsored Research - Oridion Capnography, Inc. funded data collection with a device donation to WMC.
Table 1 Summary of the Alarm Data within Each Alarm Algorithm Group