Brief report
Pilot evaluation of a ward-based automated hand hygiene training system

https://doi.org/10.1016/j.ajic.2012.03.034Get rights and content

A novel artificial intelligence (AI) system (SureWash; GLANTA, Dublin, Ireland) was placed on a ward with 45 staff members for two 6-day periods to automatically assess hand hygiene technique and the potential effectiveness of the automated training system. Two human reviewers assessed videos from 50 hand hygiene events with an interrater reliability (IIR) of 88% (44/50). The IIR was 88% (44/50) for the human reviewers and 80% (40/50) for the software. This study also investigated the poses missed and the impact of feedback on participation (+113%), duration (+11%), and technique (+2.23%). Our findings showed significant correlation between the human raters and the computer, demonstrating for the first time in a clinical setting the potential use of this type of AI technology in hand hygiene training.

Section snippets

Methods

A computer cart fitted with the SureWash system (GLANTA, Dublin, Ireland) automatically measured compliance with the WHO hand hygiene protocol for alcohol-based hand rub and provided training feedback in real time (Fig 1A). The artificial intelligence (AI) software compared the user's hand movements with a database containing examples identified by members of the research team. To pass each pose, the user needed to achieve 1 second of correct technique, or 1 second for each part in a pose with

Results

The IIR agreement between each human reviewer and the computer was 88% (44 of 50), Kα = 0.74 and 80% (40 of 50), Kα = 0.56, respectively. The IIR agreement between human reviewers was 88% (44 of 50), Kα = 0.76. In phase 2, using the real-time on-screen feedback resulted in a 113% increase in participation (from 16 to 34). The pass rate for the hand hygiene events increased from 62.5% (95% confidence interval [CI], 62.8-62.2) in phase 1 to 64.7% (95% CI, 64.6-64.9) in phase 2, a small but

Discussion

The Kα values in the pilot study showed substantial agreement (α = 0.61-0.80)7 and moderate agreement (α = 0.41-0.60)7 between each reviewer and the software, respectively. These initial results strongly suggest that the SureWash software is capable of reliably measuring hand hygiene technique; however, studies with larger sample sizes are needed to verify this. Assessment of the videos required significant concentration by the reviewers, who found that after 20 minutes, fatigue significantly

Acknowledgment

We thank Toney Thomas, John Walsh, and all the Beaumont Hospital staff who participated in the study.

References (9)

There are more references available in the full text version of this article.

Cited by (0)

Partially funded by Enterprise Ireland Project CP-2008-204 under the National Development Plan 2007-13 and Health Research Board Project TRA/2006/4.

Conflict of interest: H.H. has recent research collaborations with Steris Corporation, Inov8 Science, Pfizer, and Cepheid, and has also recently received lecture and other fees from Novartis, AstraZeneca, and Astellas. G.L. is a director of GLANTA Ltd.

View full text