Assessing the accuracy of a new hand hygiene monitoring device (SmartRub®): from the laboratory to clinical practice

Background We developed SmartRub® powered by iQati®, an electronic device composed of a wristband and an alcohol-based handrub pocket-sized dispenser that measures and provides feedback on the duration of hand friction and the volume poured during each hand hygiene action. We aimed to assess the accuracy of SmartRub®. Methods The specificity, sensitivity, positive and negative predictive values (PPV and NPV) of SmartRub® were assessed in a 3-phased experiment: (1) laboratory-controlled conditions with volunteers; (2) pre-planned clinical path with volunteers and (3) real clinical conditions with healthcare workers. The accuracy of SmartRub® was evaluated by quantifying its ability to correctly capture true hand hygiene actions and to not record other actions performed while wearing the device. Results In the laboratory, 7 volunteers performed 816 actions. Overall sensitivity was 94.1% (95% CI 91.4–96.2%) with a PPV of 99.0% (95% CI 97.3–99.6%) and specificity was 99.0% (95% CI 97.5–99.7%) with a NPV of 94.4% (95% CI 91.9–96.1%). During the pre-planned clinical path phase, 13 volunteers performed 98 planned paths and a total of 967 actions were performed. Overall sensitivity was 94.6% (95% CI 92.2–96.5%) with a PPV of 84.3% (95% CI 81.6–86.7%) and specificity was 82.4% (95% CI 78.7–85.7%) with a NPV of 93.9% (95% CI 91.3–95.7%). During the real clinical conditions phase, 17 healthcare workers were observed for a total of 15 h and 3 min while they performed 485 actions. Sensitivity was 96.8% (95% CI 93.8–98.6%) with a PPV of 98.3% (95% CI 95.6–99.3%) and specificity was 98.3% (95% CI 95.7–99.5%) with a NPV of 96.8% (95% CI 93.9–98.4%). Conclusions Smartrub® is a highly reliable device for capturing hand hygiene actions under a range of conditions, from the laboratory to clinical care activities.