iScrub: A Pilot Intervention with Feedback from a Companion Website

Background: To improve hand-hygiene compliance, many hospitals track hand-hygiene rates using human observations and feed these rates back to healthcare workers (HCWs). To facilitate and standardize the recording and collection of hand hygiene data, we developed an iPhone App, called iScrub, and coupled it with a companion website to quickly and easily generate reports in near-real time.

Objective: We report the experiences of a 34 week deployment with the companion website where results from human observations were fed back in near realtime to HCWs on a clearly visible monitor in a single unit at University of Iowa Hospital and Clinics.

Methods: iScrub was developed using Objective-C and Apple‘s iOS software development kit. The companion web-based application was implemented using the Python programming language, HTML, and the Django web framework. Analysis tools use JavaScript and jQuery to provide interactive visualizations of hand-hygiene compliance data. The iPhone app provides user authentication and communicates with the web application, over WiFi or through a cellular network, using Secure Socket Layer encryption to protect privacy. During the pilot, hand-hygiene observation data collected using iScrub was fed back to the unit on a rotating screen saver displaying summary statistics. The data and screen savers were updated every time human observers "synced" observations. We provided two iPod Touches to the unit, and nursing leadership was instructed about how to use the application. As part of the project, we specifically gave no instructions about when or how many observations should be done. To determine if there was a statistically significant trend in the hand-hygiene-compliance rate during the study period, we fit a logistic regression model using generalized estimating equations, assuming an autoregressive correlation structure.

Results: During the pilot project, HCWs who had never used an iPhone were comfortable using the application with minimal instruction. Nurse managers used iScrub for a total of 8982 observations over the period, for an average of 264.18 (s.d. 169.55) per week. Compliance rates for 12 types of healthcare workers -- both near-real-time and historical rates -- were clearly posted via the screen saver. Weekly compliance rates are given in Figure 1. The overall trend in the compliance rate during the study period was upward and statistically significant (trend estimate = 0.0260, model-based SE = 0.0096, p-value = 0.0065). ïz¼

Conclusions: iScrub and its companion web application provide a scalable, inexpensive, mobile, and inconspicuous method for recording and analyzing hand-hygiene observations. Use of the iScrub application and companion web application eliminates the need for transcription and provides near-real-time feedback. The results of our study indicate that using iScrub and providing near-realtime feedback had a positive impact on hand hygiene compliance.