Implementing Early Detection Of Postoperative Deterioration With Enhanced Monitoring Using Sensors In Variably Resourced Contexts

Implementing Early Detection Of Postoperative Deterioration With Enhanced Monitoring Using Sensors In Variably Resourced Contexts

Afra Jiwa , Malcolm M Cameron

Early Warning Scores are used globally to detect patient deterioration. These are powered by intermittent physiological observations which are subject to user error, instrument failure and staffing availability, which can lead to escalation delays. Novel wearable sensors offer accurate, automated, and continuous patient monitoring. This study aims to perform feasibility testing of a wearable sensor in eight variably-resourced countries.

This prospective observational cohort study recruited patients undergoing abdominal incision of 5 cm or greater in the UK, India, Rwanda, Ghana, Nigeria, Benin, Guatemala, and Mexico. Continuous heart rate, respiratory rate, peripheral oxygen saturation, skin temperature, and accelerometery were recorded using a wearable sensor, worn for up to 10 days postoperatively. Sensor data was unavailable to clinical teams. Standard of care clinical data was simultaneously collected. Once uploaded, study data was available to local research teams via a Shiny dashboard for near real-time data quality monitoring. Focus groups with healthcare professionals and 1-1 interviews with patients explored barriers and facilitators to future implementation.

At interim analysis 786 patients were enrolled from 17 hospitals in eight countries. 232 complications were detected. The most common complication was pneumonia. Device maintenance and data quality monitoring was conducted locally at all sites. Individual case analysis highlighted early detection using wearable sensors compared with intermittent physiological observations, across every world bank income threshold.

Continuous monitoring with sensors can be successfully deployed across a range of global settings. These technologies may detect physiological deterioration earlier than the standard of care and are feasible to implement globally.

Keywords: Wearables, Implementation, Global surgery, Predictive algorithms