Automated hand hygiene count devices may better measure compliance than human observation

Background: Hand hygiene is considered a critical factor in the prevention of health care-associated infections, and there have been many studies on ways to measure hand hygiene compliance. Objective: Our objective was to evaluate the utility of estimating hand hygiene compliance using automated count technology versus direct human observation before and after a feedback intervention. We used a before and after quasi-experimental study over 30 weeks, in the setting of one 12-bed neurocare intensive care unit (NCICU) and one 15-bed cardiac intensive care unit (CCU) in a university, tertiary care hospital. Methods: We assessed hand hygiene through a quasi-experimental study comparing estimated compliance using automated count technology and direct observation by a secret shopper with a feedback intervention at month 3. We used Poisson segmented regression and χ² tests to compare trends before and after the intervention. Results: Over 30 weeks, we collected 424,682 dispenser counts and 338 hours of human observations that included 1,783 room entries. Electronic hand hygiene dispenser counts increased significantly in the post-intervention period relative to the pre-intervention period (average count/patient-day increased 22.7 in the NCICU and 7.3 in the CCU, both P <.001), but direct observation of compliance did not change significantly (percent compliant increased by 2.9% in the NCICU and decreased by 6.7% in the CCU, P =.47 and P =.07, respectively). Conclusion: Passive electronic monitoring of hand hygiene dispenser counts does not closely correlate with direct human observation and was more responsive than observation to a feedback intervention.