Observing is influencing: How hand disinfection compliance observations affect hand disinfection rates; specifics derived from an electronic monitoring system.

Abstract

Background: Monitoring hand disinfection (HD) with direct observation is considered the gold standard but is flawed by the Hawthorne effect (HE). In our project in a German university hospital, we quantified the HE and evaluated possible room-specific differences.

Methods: Data were collected on an intensive care unit (ICU), an intermediate care unit, and a normal ward (NW) each categorized into "patient," "utility/waste," and "other" rooms. While dispensers were monitored by sensors of a full-time active electronic monitoring system, the infection prevention and control staff performed step-by-step observations according to the World Health Organization "5 moments." Periods with and without direct observations were compared regarding the number of HDs performed per patient day (HD/PD) using descriptive evaluation and a linear mixed model.

Results: The HE was detected in "patient" rooms on the NW and ICU (+ 31.4%, + 27.4% HD/PD; both P < .01) during observation periods. Opposite effects, expressed by significant decreases in HD/PD during observations, occurred in "utility/waste" rooms on the NW (-57.7%), intermediate care unit (-99.4%), and ICU (-32.9%).

Conclusions: In our setting, the HE varied considerably between wards and room types. As HE occurred in patient rooms, the usual HD compliance during patient care may be lower than directly observed.