Wireless Skin Sensors for Electrocardiogram and Heart Rate Monitoring in the Neonatal Intensive Care Unit: a prospective feasibility, safety, and accuracy study

Eva Senechal¹Daniel Radeschi²Shasha Lv³Emily Jeanne¹Ana Ruiz³Brittany Dulmage⁴Wissam Shalish^1,5Robert Edward Kearney²GUILHERME M SANT'ANNA^1,5*

• ¹Division of Experimental Medicine, Department of Medicine, Faculty of Medicine and Health Sciences, McGill University, Montreal, Quebec, Canada
• ²Department of Biomedical Engineering, School of Biomedical Sciences, Faculty of Medicine and Health Sciences, McGill University, Montreal, Quebec, Canada
• ³Research Institute, McGill University Health Center, Montreal, Quebec, Canada
• ⁴Wexner Medical Center, The Ohio State University, Columbus, Ohio, United States
• ⁵Montreal Children's Hospital, McGill University Health Centre, Montreal, Quebec, Canada

Objectives: Assess feasibility, safety, and accuracy of electrocardiogram (ECG) and heart rate (HR) monitoring in neonates, using a new wireless skin sensor.Methods: Prospective observational study in infants of any gestational age admitted in the neonatal intensive care unit. ECG/HR signals were simultaneously recorded from a standard wired and new wireless system with bedside annotations. Feasibility was evaluated as signal coverage, gap numbers/durations, and sources of gaps. Safety was appraised by changes in skin condition and pain after/upon wireless sensor removal. Accuracy was measured using bias and 95% limits of agreement, and the coefficient of determination. The ability of the wireless sensors to detect normal and abnormal HR values was evaluated using a Clark Error Grid. Additionally, user satisfaction from parents and nurses were appraised using a short questionnaire.Results: 25 infants had 757 hours of recorded signals over 96 days. ECG coverage was 99.9%[IQR: 99.9-99.95%] for the wired vs. 97.8% [IQR: 81.6-99.9%; p < 0.00] for the wireless system, while HR coverage was 99.4% [IQR: 98.6-99.9%] vs. 89.7% [IQR: 75.6-97.6%; p < 0.00]. Wireless ECG gaps were <5 seconds in 97% of cases, and HR gaps <30 seconds in 85%.All ECG gaps and 57% of HR gaps were due to Bluetooth disconnection (BD). 78% of BD in wireless HR were during kangaroo care (78%). Of 192 skin photographs (96 pairs), 98% were taken, showing increased but low skin scores post-removal, with median pain scores also low.Accuracy metrics showed strong agreement, with the Clark Error Grid indicating 97% of paired signals led to the same clinical outcome. Among 23 nurse and 18 parent responses, satisfaction with the wireless system was high.ECG and HR monitoring using a new wireless skin sensor was feasible, safe, and accurate when compared to the wired standard. Future adjustments in the technology are needed