Creating an answerable and searchable question
In evidence-based medicine, there is emphasis on using high-quality research to inform clinical decision making. To effectively search the literature one must first create an answerable question. Several tools exist to help create answerable questions, the most frequently cited of which is PICO, which stands for population/patient, intervention/indicator, comparison/control, and outcome.
This mnemonic is effective in formulating focused research questions, but using PICO alone to construct a search strategy has some pitfalls. First, the disease concept may not be fully captured by population/patient if one focuses too tightly on demographics. Second, including all four PICO facets as search terms (e.g., adding too many comparisons) can eliminate many relevant results.
One solution is to build research questions using PICO, then extract search terms using the concepts underpinning the Medline database’s tree structure: disease, body part, and process. While disease and body part are largely self-explanatory, process encompasses both interventions (e.g., drugs, surgical techniques) and nonclinical processes (e.g., quality-assurance measures). A question such as “What is the best diagnostic modality in patients with traumatic brain injury?” becomes answerable and more effectively searched when stated: “In patients with suspected TBI (population/disease + body part) what is the diagnostic value of CT (intervention/process) compared with MRI (intervention/process)?”[1]
Whether a literature search is the end of a research process or the beginning of a larger project, College librarians are happy to help you build your search or to search the literature for you. Please contact the College Library for assistance.
—Paula Osachoff
College Librarian
hidden
This article is the opinion of the Library of the College of Physicians and Surgeons of BC and has not been peer reviewed by the BCMJ Editorial Board.
References
1. Asking good clinical research questions and choosing the right study design. Injury 2010;41(Suppl1):S3-S6.