Development and Evaluation of BLE-Based Room-Level Localization to Improve Hand Hygiene Performance Estimation

Area of Science:

• Biomedical Engineering
• Infection Control
• Wireless Sensor Networks

Background:

• Hand hygiene (HH) is crucial for preventing healthcare-associated infections.
• Existing electronic monitoring systems have limitations in capturing complete HH adherence.
• Location data can significantly enhance the resolution and accuracy of HH monitoring.

Purpose of the Study:

• To propose and evaluate a Bluetooth Low Energy (BLE)-based solution for localizing healthcare workers within patient rooms.
• To enhance hand hygiene monitoring by integrating location information.
• To explore the potential of this localization method as a proxy for caregiving activity recognition.

Main Methods:

• Development of a BLE-based system for healthcare worker localization.
• Evaluation using proximity-based positioning (binary and multiclass) with 1-4 beacons.
• Collection of dynamic fingerprints from nine subjects performing 30 nursing activities.
• Application of the Extremely Randomized Trees algorithm for classification.

Main Results:

• The proposed BLE localization system achieved 81% accuracy in binary classification (entrance/patient zone).
• Multiclass classification (entrance/sink/bed sides) yielded an accuracy of 71%.
• Extremely Randomized Trees algorithm demonstrated superior performance in localization accuracy.

Conclusions:

• BLE-based localization offers a viable method to enhance hand hygiene monitoring in healthcare settings.
• Accurate localization of healthcare workers can serve as a proxy for recognizing caregiving activities.
• Improved monitoring can contribute to reducing the risk of infection transmission.