Challenges encountered and lessons learned during a trial of an electronic hand hygiene monitoring system

Background: Automated hand hygiene monitoring systems (AHHMS) are being developed to supplement direct observations of hand hygiene (HH). We compared compliance rates generated by direct observations and by a badge-based AHHMS. Methods: Observations of HH compliance were conducted in a surgical intensive care unit (SICU) and a general medical ward (GMW) during a 7-month baseline period and a 13-week intervention period, when compliance was also estimated using an AHHMS. Compliance rates were analyzed using time-series analysis. Results: During the entire pre- and postintervention period, univariate analysis of observations revealed significant improvement in overall HH compliance in the SICU (P =.001) and the GMW (P =.03), beginning before implementation of the AHHMS. Initiation of the AHHMS was associated with a transient drop in entry and exit compliance on both units. During the intervention period, observations suggested that compliance on entry and exit continued to increase in the SICU, but not in the GMW. Time series analysis of AHHMS data revealed that entry and exit compliance did not change significantly in the SICU but decreased significantly in the GMW (P =.001). The average accuracy of the AHHMS was 60%. Conclusions: Based on observations, HH compliance increased during the baseline period. Observations and the AHHMS yielded different trends in compliance. The AHHMS's accuracy of HH events and health care personnel location were suboptimal.