This, below, from Michele in Scotland, gave me “goose-bumps” (a thrill). Directly pertinent to my job, that was the smartest thing I’ve read in a long time. She nailed it! It’s lovely. I could not possibly agree more.
I do big literature-searches in support of systematic reviews. People bring me their thinking, in the form of a PICO chart/statement/outline (I prefer to scribble some overlapping circles on paper) — and then we must do a human-to-human conversation, and then I must do a huge amount of translating all that “intent” and presupposition to resource-selection and search-strategy. Frequently, we need to revisit the themes, words, controlled-vocabulary, mission, breadth or narrowness, exclusion-criteria, limits, and play play play. It’s a fascinating, dynamic process.
____
Tom Mead
Reference Librarian
Biomedical Libraries
Dartmouth College
Geisel School of Medicine at Dartmouth
Dartmouth-Hitchcock Medical Center
Dartmouth Institute for Health Policy and Clinical Practice
Dartmouth Center for Health Care Delivery Science
Hanover, New Hampshire, USA 03755
On May 2, 2014, at 6:03 AM, Hilton Boon Michele (HEALTHCARE IMPROVEMENT SCOTLAND - SD039) <[log in to unmask]> wrote:
> “Output from a search” is indeed the desired outcome of a PICO question, but it would not be an appropriate metric of the question’s quality. No matter how well-formulated, in order to produce search outputs and relevant information, the PICO question still has to be translated into search strategies for multiple databases, which will depend on knowledge-based activities that very quickly extend beyond the wording of the PICO question itself; then the search results have to be sifted, again an activity that depends in small part on the precision of wording of the PICO question and in large part on the information skills and tacit knowledge of the person sifting. This may also be an iterative process, with the question being revised in response to the literature identified (or not identified). The ‘quality’ of search output, then, is dependent upon the experience of the searcher, the choice of search terms, the precision and accuracy of indexing in the database, the sources used, and the application of inclusion/exclusion criteria, not just the wording of the PICO question – which could be viewed as simply a starting point for dialogue between the person with the information need and the search specialist, as well as part of the communication of the scope of the work with the intended audience. Viewed in this context, I would argue that it is more important to evaluate the quality of search strategies than of PICO questions, and there is a checklist for this: PRESS http://www.cadth.ca/en/publication/781 (Appendix G).
|