Use of hand-held computers to determine the relative contribution of different cognitive, attitudinal, social, and organizational factors on health care workers' decision to decontaminate hands.

Background Observational and survey methods have limitations in measuring hand hygiene behavior. The ability of a personal digital assistant to anonymously gather data at the point of decision making could potentially address these. Methods Participants were provided with a personal digital assistant to be used for three 2-hour periods and asked to rate influential factors of the Health Belief Model (HBM). Participants were also required to enter what they thought they should do and what they actually did. Results A total of 741 hand hygiene opportunities was recorded. All HBM constructs were higher for hand hygiene opportunities where there was compliance versus noncompliance, with a significant difference for patient pressure, my risk, perceived benefits, perceived seriousness, and availability of good facilities. Only 20% of doctors, 28% of nurses, and 66% of physiotherapists always did what they thought they should. There was no correlation between self-reported and actual compliance. Conclusion The HBM appeared to be a useful theoretical framework. Surprisingly, participants rated their compliance as high despite having recorded instances where they did not do what they thought they should do. This suggests that staff may have a different definition of compliance than strict observation of the guidelines.