Drivers of choice of resuscitation fluid in the intensive care unit: a discrete choice experiment.

OBJECTIVE To understand the fundamental drivers, and their relative importance, of doctors' and nurses' choice of resuscitation fluid in critically ill patients in Australia and New Zealand. DESIGN A discrete choice experiment (DCE) administered via an online survey. Respondents were presented with one of four randomly selected DCE choice sets, each including five patient scenarios. The respondent chose between two types of hypothetical resuscitation fluid. The fluid type was characterised by several attributes and each attribute had pre-specified levels. PARTICIPANTS Convenience sample of 367 Australian and New Zealand intensive care unit doctors and nurses. MAIN OUTCOME MEASURES The dependent variable was fluid choice, and a regression equation was used to estimate the effect of each fluid attribute on the probability of observing the sequence of choices made over the five patient scenarios. The relative importance of each of the respective fluid attributes was calculated based on the percentage contribution to overall utility (ie, fluid preference). RESULTS For doctors, safety concerns, patient type and fluid type were collectively responsible for almost three-quarters of decision-making utility (71%). The volume of intravenous fluid administered was the only clinical parameter not reaching statistical significance as a driver of fluid choice (P = 0.06). For nurses, decision making was influenced to a greater extent by the same three attributes (90%), although other unmeasured attributes may have been driving choice. CONCLUSIONS Doctors and nurses rely on different information when choosing resuscitation fluids, although both cohorts are heavily influenced by safety concerns, patient type and fluid type. This information can be used to modify prescribing behaviour.