451 - Neonatal Dual Pulse Oximeter Algorithm Improves SpO2 Monitoring Performance

Poster Number: 451
Publication Number: 451.228

Dale R. Gerstmann, Timpanogos Regional Hospital, Orem, UT, United States

Background: False or spurious pulse oximeter alarms remain a significant burden for NICU caretakers and continue to create confusion in identifying significant desaturation and bradycardia events from as many as 250 alarms per infant per day (McClure et al., J Neonatal Perinatal Med, 2016). Novel schemes to improve SpO2 monitoring performance are needed (van Der Eijk et al., J Perinat Neonatal Nurs, 2012).

Objective: To compare standard single pulse oximeter (SPO) monitoring to monitoring using a novel dual pulse oximeter algorithm (DPOA).

Design/Methods: The local hospital IRB approved this pilot study. Infants admitted to the NICU were eligible for enrollment. In addition to the usual bedside SpO2 monitor, study monitors consisted of a pair of Nellcor PM1000N pulse oximeters with alarms disabled, connected to standard neonatal sensors placed on one upper (UE) and one lower extremity (LE). Data ports on each pulse oximeter used blue-tooth transmitters to connect to a notebook computer. Output values were error codes, SpO2 (SAT), and pulse rate (PR). The study used proprietary software to perform data collection, algorithm logic, and compilation. For analysis purposes, we assigned low, in-range, and high limits as SAT ≤87, 90-95, ≥98; PR ≤85, 95-155, ≥180, and results presented as the ratio of DPOA to SPO counts (DS ratio).

Results: We enrolled and studied 16 patients, recording 513 hours of dual pulse oximetry data (32.1 hours per patient). The computer software analyzed 907,560 complete time-matched records collected at 2-second intervals. Loss of Pulse error occurred in 0.0947% of records for SPO values and 0.0008% of records for DPOA. Motion error occurred in 10.7967% of SPO records and 3.2236% of records for DPOA. The UE/LE count ratio was 1.62 for Loss of Pulse error and 1.57 for Motion error. Paired values for SAT and PR were within ±1 44.1% and 62.4% of the time, respectively. The DS ratios for low, in-range, and high SAT were 0.686, 1.424, 1.341; and for PR were 0.341, 1.049, 1.130, respectively. The DS ratio for combined low SAT and low PR was 0.575.Conclusion(s): DPOA decreased Loss of Pulse error by a factor of 124.9 and Motion error by a factor of 3.3. Loss of Pulse error and Motion error were more frequent in UE. DPOA reduced findings of low SAT by 31%, increased in-range SAT by 42%, and improved high SAT identification by 34%. Similarly, low PR decreased by 66%, and high PR increased by 13%. The combination of low SAT and low PR occurred 42% less frequently. Using a dual pulse oximeter algorithm in this pilot study appeared to improve the safety profile of SpO2 monitoring.