O3222: A Continuous Respiratory Rate Monitor Intended for Low-Resource Settings: A Novel Device Pilot Study.

Background: Neonatal deaths account for nearly 50% of all under age 5 mortality, and over 80% of the annual neonatal deaths in the world occur in Low- and Middle-income countries (LMIC), largely from preventable causes. Continuous respiratory rate (RR) monitoring in high-income countries (HIC) allows for rapid detection and treatment of serious and life-threatening complications in premature and sick newborns such as respiratory distress syndrome, sepsis, apnea, and seizures. Early detection and treatment of these conditions is pivotal in reducing neonatal mortality and morbidity in HIC. Continuous RR monitors are not widely available in LMIC; thus the World Health Organization recommends manually counting breaths for 60 seconds to measure neonatal RR as the standard of care. However, manual counting is largely impractical due to high patient-to-nurse ratios in LMIC. To address the need for a low cost, continuous respiratory rate monitor in LMIC, a collaborative partnership of clinicians and engineers developed a novel monitor that identifies breaths by measuring the displacement of an inertial measurement unit sensor placed around a newborn’s abdomen. In this pilot study, we aim to compare the novel device’s performance against reference visual counting and existing patient monitors in HIC.

Methods: We enrolled newborns in a level 3 Neonatal Intensive Care Unit (NICU) in Houston, USA, and level 2 NICU in Blantyre, Malawi to receive continuous monitoring with the novel device and standard of care monitoring (GE Dash 4000/ Phillips intellivue MP 30). Webcam video monitoring (C925-E, Logitech) was used for manually counting breaths. Paired t-test and cross-correlation analysis was conducted between the RR from our novel device and impedance pneumography. Institutional review board approvals were obtained from all participating institutes.

Results: 20 newborns at the HIC NICU, and 4 newborns at the LMIC NICU have been enrolled. Total HIC patients average 31.9 weeks [28.4-36.1] with an enrollment weight of 1459g [1035-1975g] and LMIC patients average 31.25 weeks [30 – 33] gestational age with an enrollment weight of 1420g [1300-1555]. Patients were monitored for an average of 42 hours in HIC and 15 hours in LMIC respectively. LMIC enrollment and data analysis is currently ongoing. Thus far, data indicates that our novel device and reference manual counting show a very strong correlation (R = 0.97). RR bias and limits of agreement (LOA) were within the United Nations International Children's Emergency Fund (UNICEF) target product profile specifications for device validation. No adverse events have occurred.

Conclusion: Our novel device shows a high degree of accuracy in detecting breaths as compared to the gold standard methods in HIC and LMIC. It may offer a low-cost alternative to improve neonatal respiratory monitoring in LMIC, thereby potentially decreasing both the workload burden on nurses and the morbidity and mortality in sick and preterm newborns.