Predictors of Hand Hygiene Compliance Opportunities: A Retrospective Analysis of the Literature

Predictors of Hand Hygiene Compliance Opportunities: A Retrospective Analysis of the Literature Pages E49-E50 B. Whitaker*, J. Arbogast, M. Dolan, P. Taylor Close Close preview | Purchase PDF (71 K) | Related articles | Related reference work articles AbstractAbstract Abstract BACKGROUND: Historically, observational studies of hand hygiene compliance using trained observers have defined a hand hygiene “opportunity” as any event with a high risk of microbial transmission. As the definition of “opportunity” appears relatively standardized, one may expect a high degree of agreement across studies and subsequently analyze other study-reported variables as predictors of observed opportunities across studies. OBJECTIVE: The analysis objective was to determine the extent to which study-reported variables predict hand hygiene compliance opportunities in the healthcare setting. METHODS: Non-redundant predictors of observed opportunities across studies were theorized to be number of beds on each unit observed and number of hours observed. A literature review was conducted using the MEDLINE database from 1983 to 2003 in order to identify studies that reported these variables. Studies involving 17 units met inclusion criteria. Multiple regression was employed for analysis. RESULTS: Bed-size and hours observed accounted for 84.1% of the variance in observed opportunities (F2,16 = 43.38, p < 0.001). Further analysis indicated that the two predictor variables independently contribute to the prediction of observed opportunities. Content of observer opportunity training for the studies analyzed will be compared with the current CDC observer guidelines. CONCLUSIONS: Statistically, bed size predicts the number of trainee-observed opportunities for given hours observed. For the researcher, these results will assist in study planning in order to maximize power. Furthermore, these results indicate the feasibility of a statistically based, non-observational indexing method that may ultimately provide a means of addressing monitoring/feedback per the 2002 CDC guidelines.