Continuous wireless sensor monitoring with applied diagnostics: Clinical Sensor Pain Scale and Automated Sensor Pain Scale in the NICU.

Abstract

Inappropriately treated pain can have deleterious outcomes in infants. Current tools rely on intermittent, subjective observation requiring specialised paediatric skills. This study aimed to diagnose infant pain through continuous monitoring with wireless sensors using Neonatal Pain and Agitation Sedation Scale (NPASS)-derived Clinical Sensor Pain Scale (CSPS) and Automated SPS (ASPS). Clinically stable neonatal intensive care unit infants undergoing phlebotomy were recorded with wireless sensors and video, capturing vital signs, extremity movement and vocalisations. Clinicians and non-clinicians scored the sensor data with CSPS and videos with NPASS; ASPS was applied to the sensor data. Median scores were compared, inter-rater reliability assessed with intraclass correlation coefficients (ICC) and cross-scale comparisons performed using Wilcoxon signed-rank and Kruskal-Wallis tests. CSPS and ASPS closely aligned with NPASS scores, supporting their validity for continuous infant pain assessment. In 32 infants, the median CSPS score was 3 (IQR 2, 5), with excellent reliability (ICC, 95% CI 92 to 97), high internal consistency (Cronbach's α=0.99) and 95% absolute agreement, comparable to NPASS (p=0.95). Clinician and non-clinician scores were more consistent using CSPS than NPASS. ASPS also performed well, with a median score of 3 (IQR 1, 5), yielding results similar to CSPS (p=0.94) and NPASS (p=0.56). Wireless biosensors enabled objective monitoring of infant pain. CSPS and ASPS showed validity and reliability for diagnosing acute procedural pain, and feasibility for clinical use. Findings support the development of automated, real-time tools to reduce subjectivity and improve infant pain management, with the potential to advance treatment models and outcomes.