In an urban emergency department on weekend nights in 2010 and 2011, 105 interviews assessed feasibility of collecting alcohol brand consumption data from injured patients who drank within 6 h of presentation, with responses to the orally administered survey specifying 331 alcohol brands recorded on a netbook computer.
A Kruskal–Wallis test adjusted for tied ranks assessed demographic differences; confidence intervals were created around comparisons with national brand shares.
The study found collection of such information feasible; limitations include comparison of national brand market share data with a local sample of drinkers.
Read Full Abstract
Request Reprint E-Mail: djernigan@jhsph.edu