Objective: To quantify the Hawthorne effect of hand hygiene performance among healthcare workers using direct observation.
Design: Prospective observational study.
Setting: Intensive care unit, university hospital.
Methods: Direct observation of hand hygiene compliance over 48 audits of 2 hours each. Simultaneously, hand hygiene events (HHEs) were recorded using electronic alcohol-based handrub dispensers. Directly observed and electronically recorded HHEs during the 2 hours of direct observation were compared using Spearman correlations and Bland-Altman plots. To quantify the Hawthorne effect, we compared the number of electronically recorded HHEs during the direct observation periods with the re-scaled electronically recorded HHEs in the 6 remaining hours of the 8-hour working shift.
Results: A total of 3,978 opportunities for hand hygiene were observed during the 96 hours of direct observation. Hand hygiene compliance was 51% (95% CI, 49%-53%). There was a strong positive correlation between directly observed compliance and electronically recorded HHEs (ρ=0.68 [95% CI, 0.49-0.81], P<.0001). In the 384 hours under surveillance, 4,180 HHEs were recorded by the electronic dispensers. Of those, 2,029 HHEs were recorded during the 96 hours in which direct observation was also performed, and 2,151 HHEs were performed in the remaining 288 hours of the same working shift that were not under direct observation. Healthcare workers performed 8 HHEs per hour when not under observation compared with 21 HHEs per hour during observation.
Conclusions: Directly and electronically observed HHEs were in agreement. We observed a marked influence of the Hawthorne effect on hand hygiene performance.