Accuracy of a Noncontact Breathing Frequency Monitoring Technology in the Neonatal ICU

Background: Breathing frequency is a sensitive indicator of a neonate patient's condition and a crucial parameter in the neonatal ICU (NICU). However, conventional methods require sensors to be adhered to the newborn's sensitive skin. This study assesses a noncontact respiratory monitor, using linear frequency modulation continuous wave radar, to determine the monitor's accuracy when used in the NICU setting. Methods: An observational study was conducted on the noncontact respiration module produced by Mindray, connected to a BeneVisionN12 patient monitor. This unit collected data on breathing frequency; at the same time, the patient breathing frequency was also being measured using a CO₂ monitor (accepted standard). Statistical analyses were performed to assess the agreement between the 2 measurements and to assess whether this noncontact monitor is sufficiently accurate for clinical use. Results: 20 neonatal patients at 2 NICUs were enrolled in the study. The mean bias between the radar-derived and CO₂-derived breathing frequency was -0.29 bpm (95% CI: -0.53 to -0.04); root-mean-square error was 2.60 bpm; concordance correlation coefficient was 0.98; and Pearson correlation coefficient was 0.98. Subgroup analyses grouped by body weight show that the technology has particularly high accuracy and clinical utility for neonates with low body weight (<1.5 kg). Conclusions: The noncontact respiration monitoring technology and module developed by Mindray showed a high degree of agreement with the CO₂ reference standard and met the standard of accuracy required for clinical use.