How and when do healthcare workers remember to sanitize upon patient visits? Researchers from the section Statistics and Data Analysis, DTU Compute, at the Technical University of Denmark worked on analyzing the data from the Sani nudge system to investigate this question.
Collaboration is the way forward

In the beginning of the year, DTU Compute, a division of the Technical University of Denmark (DTU), granted the IoT company Sani nudgeTM the opportunity to have some of their best researchers dive deeper into the world of hand hygiene data. This was part of the new AI entrepreneurial initiative which aimed to bring machine learning and predictive analytics to understand hand hygiene trends.

Sani nudge is an electronic hand hygiene monitoring system which continuously records the hand hygiene performance of healthcare workers according to hospital guidelines, CDC and the World Health Organization standards. The system has been developed together with two Danish University Hospitals that clinically tested and validated the system and its data.

Why is this collaboration urgently needed?

Millions of patients are affected by hospital-acquired infections worldwide each year, leading to significant financial losses for hospitals, development of antibiotic resistance and mortality. For this reason, the prevention of these infections is a first priority for hospitals committed to making healthcare safer.

Research has shown that improving hand hygiene among healthcare workers is the single most effective way of reducing hospital-acquired infections and yet most hospitals don’t have systems in place to monitor hand hygiene compliance. Those that do perform hand hygiene monitoring (many actually don’t!) often struggle with the reliability of data because they perform direct observations using manual auditing which overestimates performance by 300%.

The future is big data

The potential for analyzing hand hygiene behavior based on data from multiple sensor inputs is endless. The researchers from the section Statistics and Data Analysis, DTU Compute, at the Technical University of Denmark worked on analyzing the data from the Sani nudge system to investigate how and when healthcare workers remembered to sanitize upon patient visits and used this to create predictive models that showed when a healthcare worker was most likely to forget to sanitize.

The large amount of data that Sani nudge gathers, revealed some interesting patterns. These patterns are simply not visible when measuring hand hygiene using direct observations because of the severe lack of data points. Results from the collaboration is expected to be published soon.

Sani nudge and the Technical University of Denmark are currently investigating on the possibility to continue the collaboration due to the great success of the project. They hope to further develop algorithms that can predict high risk areas of infections at hospitals in real-time and thereby intervene before infections actually occur. This will put preventive medicine to a whole new level.