Route of administration of redback spider bite antivenom: Determining clinican beliefs to facilitate Bayesian analysis of a clinical trial.

Objective: To determine current beliefs of Australasian emergency physicians, to form the basis of 'stopping rules' for a clinical trial of intravenous (i.v.) versus intramuscular (i.m.) redback spider antivenom.Methods: An email survey of fellows and trainees of the Australasian College for Emergency Medicine.Results: There were 218 responses; 30% used the i.v. route exclusively, 16% used the i.m route exclusively, 17% used i.m. followed by i.v. if there was a poor initial clinical response, and 38% stated that they had no particular preference. The probability given by respondents that the i.v. route is superior allowed us to differentiate 'i.v. enthusiasts' from 'i.v. sceptics'. Median predicted response rates were 90% versus 80% for the i.v. route and 60% versus 75% for the i.m. route in the enthusiastic and sceptical groups, respectively. The median expected absolute advantage of i.v. compared with i.m. antivenom was 20% versus 5%, respectively. The median number-needed-to-treat threshold that would lead respondents to choose the i.v. route in preference to the i.m. was 5.Conclusion: Australasian emergency physicians have polarized views on the optimal route for administering redback spider antivenom. We were therefore able to define both sceptical and enthusiastic priors for a fully Bayesian trial analysis. Our findings support using a number needed to treat of 5 (20% absolute advantage) for powering a clinical study and for determining the point at which it should be stopped.