Background: Wheezing is a critical symptom in children with acute asthma exacerbations, often indicating need for intervention. Traditionally, wheezing is diagnosed and monitored through physical examination by healthcare providers. However, monitoring for this symptom regularly in a busy emergency department with high patient-to-provider ratio is challenging. A device for continuous and accurate wheeze monitoring could improve care quality by aiding in prompt decision-making, guiding interventions, and potentially reducing frequency of physical examinations. This study assessed the AeviceMD, a cloud-connected smart wearable stethoscope as a solution for wheeze monitoring in children presenting with respiratory complaints at the emergency department. Methods: We enrolled thirty-six children, aged 2 to 17 years, who presented with respiratory symptoms at the Pediatric Emergency Department of National University Hospital, Singapore. Each patient underwent five lung sound recordings using the AeviceMD, applied sequentially to all five lobes of the lungs. The device used an advanced detection algorithm to analyze and annotate each recording for wheeze presence, and these annotations were compared with clinical assessments by healthcare providers to evaluate the accuracy of the device. We calculated the sensitivity, specificity, and overall accuracy of the AeviceMD in detecting wheeze in a busy, often noisy Emergency Department environment. Results: The AeviceMD demonstrated strong performance in detecting wheeze in enrolled patients, with a sensitivity of 85.3% and a specificity of 80.8%, resulting in an overall accuracy of 83%. When evaluated only with data that conforms to the device's prescribed environment (home environment) per its intended use, the sensitivity and specificity improves to 93.2% and 91.1% respectively, with accuracy of 92%. Notably, the device can identify wheeze in intervals as short as five seconds, allowing for near-real-time tracking of respiratory status. When used continuously, AeviceMD provided detailed trends patient's wheeze patterns across long periods of wear, which could assist healthcare providers in making timely decisions. Furthermore, each wheeze recording was stored in cloud, enabling later review by physicians when needed. Conclusion: Our findings show that AeviceMD is a promising tool for wheeze monitoring in emergency settings. In an often overwhelmed Paediatric Emergency Department, devices like this can elevate care delivery while reducing the burden on healthcare providers by removing the need for constant physical examination of asthma patients undergoing therapy or in post-therapy observation. Continuous wheeze monitoring with AeviceMD may support better informed treatment decisions, improving workflow efficiency in pediatric emergency care and patient outcome.
