RT Journal Article
SR Electronic
T1 Mind the Gap: Identifying What Is Missed When Searching Only the Broad Scope with PubMed Clinical Queries
JF bioRxiv
FD Cold Spring Harbor Laboratory
SP 423111
DO 10.1101/423111
A1 Edwin V. Sperr, Jr.
YR 2018
UL http://biorxiv.org/content/early/2018/09/23/423111.abstract
AB Objective PubMed Clinical Queries are subdivided into “Broad” and “Narrow” versions. These versions are tuned to maximize either retrieval or sensitivity using two different sets of keywords and MeSH headings. While a searcher might assume that all items retrieved by Filter name/Narrow would also be found in the set Filter name/Broad, this is not explicitly guaranteed. It is the purpose of this study to quantify the overlap between these two sets and confirm whether Filter/Narrow is always a subset of Filter/Broad.Methods For each of the five PubMed Clinical Queries, PubMed was searched for citations matching the query Filter/Narrow NOT Filter/Broad. This number was compared with that for Filter/Broad to compute a “degree of discordance” between the two sets. This process was then repeated for the MeSH headings for “Medicine” and “Diseases” as well as for a set of test searches.Results Four of the five Clinical Queries returned citations using Filter/Narrow that were not found with Filter/Broad. Discordances between the sets Broad and Narrow were generally modest for “Etiology”, “Diagnosis” and “Clinical prediction guides”. “Prognosis” was notably more discordant – a searcher could easily miss one Prognosis/Narrow citation for every ten citations she retrieves when using Prognosis/Broad alone for a given search.Conclusions Users of the Clinical Queries apart from “Therapy” who are interested in retrieving as many relevant citations as possible should consider combining Filter/Narrow together with Filter/Broad in their search strategy. This is particularly true for “Prognosis”, as otherwise the risk of missing relevant citations is substantial.