As I've said before, "It's amazing how little evidence is required before infection prevention interventions are adopted." This phenomenon is particularly evident in the setting of automated hand hygiene monitoring systems. Few trials have been completed that have assessed the efficacy, effectiveness or cost-effectiveness of these systems, at least as of our recent systematic review. However, it appears that many facilities are purchasing these systems anyway. Why is this?
It appears that there are two potential limitations of directly-observed hand hygiene monitoring driving the purchase of automated systems: (1) the fear that the Hawthorne effect renders all direct observations invalid and (2) the idea that all hand hygiene opportunities must be counted, i.e. it's unacceptable to sample opportunities. There is a third potential reason for adopting these systems - anecdotal reports that automated systems improve compliance. However, there is little published evidence that compliance increases are sustained and some evidence that automated systems actually decrease compliance.
With the potential limitations of directly observed compliance monitoring in mind, our research group just completed a study published in ICHE that sought to determine if/when the Hawthorne effect appears in hand hygiene observation data. The idea being, if it takes some period of time for HCW to know they are being observed, shorter periods of hand hygiene monitoring could be less susceptible to bias from the Hawthorne effect. After 3,432 hours of observations and 11,444 witnessed opportunities in a multi-center study, we calculated that room-entry compliance increases after 38 minutes and room exit compliance increases after 14 minutes and then again after 50 minutes (Figure below). Thus, it appears that limiting direct observation periods to less than 15 minutes limits the impact of the Hawthorne effect.
The other aim of our study was to determine the number of hand hygiene opportunities that must be observed to have an adequate sample size for comparison. For example, if hand hygiene compliance in your ICU is 80% this month and you would like to get it to 90% next month, you would need to observe 108 opportunities in each month to have enough power to detect a difference. We provide a table (below) so that anyone can quickly determine the number of observations in a month/quarter/year needed to compare time-periods or units/wards or facilities. We hope you find this information useful and that you'll still give directly observed hand hygiene compliance monitoring a chance.