Method: Four orthodontic residents, with training and experience with Clinical Queries and EviDents, evaluated the two search engines by searching for predetermined publications that answered eight PICO questions. Half of the searches answered PICO questions where there were high levels of evidence (randomized controlled trial) and half for questions where there were only low levels of evidence (case study & case series). The searches were done in a counterbalanced order. The search computers recorded the publication selected, the number of page views, and total time spent searching before selecting a reference.
Result: Evaluators located the target publications faster using EviDents (median 34.5 seconds for high levels of evidence and 60.5 seconds for low levels) compared to Clinical Queries (median 114 seconds for high levels and 133 seconds for low levels) (p=.018). EviDents also outperformed Clinical Queries in accurately locating the target publication under low levels of evidence (EviDents: 75% success versus Clinical Queries: 0% success, p<.001), but neither search engine outperformed the other under high evidence (EviDents: 75% success versus Clinical Queries: 71.4% success, (p=.081).
Conclusion: EviDents outperformed Clinical Queries in terms of speed for both low and high levels of evidence. EviDents also outperformed Clinical Queries in terms of accuracy under low but not under high levels of evidence. EviDents seems to be a time-efficient search engine for answering PICO questions about orthodontics.
Keywords: Evaluation, Learning, Orthodontics, Pubmed, EviDents and Technology