Predictors of adherence to safe handling practices for antineoplastic drugs: A survey of hospital nurses

Despite growing awareness of the hazards of exposure to antineoplastic drugs (ADs), surveys continue to find incomplete adherence to recommended safe handling guidelines. A 2011 survey of healthcare workers presents an opportunity to examine factors associated with adherence among 1094 hospital nurses who administered ADs. Data for these hypothesis-generating analyses were taken from an anonymous, web-based survey of healthcare workers. Regression modeling was used to examine associations between a number of predictors (engineering controls, work practices, nurse perceptions, and nurse and hospital characteristics) and three outcomes reported by nurses: use of personal protective equipment (PPE); activities performed with gloves previously worn to administer ADs; and spills of ADs. Adherence to safe handling guidelines was not universal, and AD spills were reported by 9.5% of nurses during the week prior to the survey. Familiarity with safe handling guidelines and training in safe handling were associated with more reported PPE use. Nurse-perceived availability of PPE was associated with more reported PPE use and lower odds of reported spills. Use of closed system drug-transfer devices and luer-lock fittings also decreased the odds of self-reported AD spills, while more frequent AD administration increased the risk. AD administration frequency was also associated with performing more activities with gloves previously worn to administer ADs, and nurse perception of having adequate time for taking safety precautions with fewer such activities. The results suggest that training and familiarity with guidelines for safe handling of ADs, adequate time to adhere to guidelines, and availability of PPE and certain engineering controls are key to ensuring adherence to safe handling practices. Further assessment of training components and engineering controls would be useful for tailoring interventions targeting these areas.