Impact of observing hand hygiene in practice and research: a methodological reconsideration

Summary Healthcare-associated infection is often spread by direct contact, and the importance of hand hygiene to break the chain of infection is recognised internationally. In many countries hand hygiene is regularly audited as part of quality assurance based on recommendations issued by the World Health Organization (WHO). Direct observation is the recommended audit method but is associated with a number of disadvantages, including potential for being observed to alter usual behaviour. The Hawthorne effect in relation to hand hygiene is equated with productivity by increasing the frequency that hand hygiene is undertaken. Unobtrusive and/or frequent observation to accustom staff to the presence of observers is considered an acceptable way of reducing the Hawthorne effect but little has been written about how to implement these techniques or assess their effectiveness. There is evidence that awareness of being watched can disrupt the usual behaviour of individuals in complex and unpredictable ways other than simple productivity effect. In the presence of auditors health workers might defer or avoid activities that require hand hygiene, but these issues are not addressed in guidelines for practice or research studies. This is an important oversight with implications for the validity of hand hygiene audit findings. It needs to be considered if such findings are taken as indicators of quality of care and if the results of hand hygiene research are used to inform future policy and practice. Measuring hand hygiene product use overcomes avoidance tactics. It is cheaper and generates data continuously to give a 24 hour picture of compliance for all clinicians without disrupting patient care. Disadvantages are the risk of over-estimating uptake through spillage, wastage or use by visitors and non-clinical staff entering patient care areas. Electronic devices can overcome Hawthorne and avoidance effects but are costly and are not widely used outside research studies.