PubMed Clinical Queries and Google Scholar search engines set to narrow search are more efficient in returning relevant and high-quality research results than standard PubMed and Google search tools, a study in the Annals of Emergency Medicine reports.
Findings suggest that although a typical Google or PubMed search can produce results quickly, which is particularly relevant for use in the emergency department (ED), a narrower attempt to find relevant data may ultimately prove more useful.
Search terms related to the 3 clinical questions were entered into Google Web, Google Scholar, PubMed, PubMed Clinical Queries (set to narrow search), and PubMed Clinical Queries (set to broad search). Results were displayed in order of most recent first, and the first 60 hits were reviewed and examined for quality of evidence, relevance vs irrelevance, and number of high-quality hits. The search engines were then graded for overall readability, using a visual analog scale.
Using 5 search tools, researchers from the University of California, San Diego, entered search terms associated with 3 frequently encountered clinical questions in the ED. One of the clinical questions asked whether an antiemetic reduces recurrent vomiting and hospital admission while increasing hydration in young children admitted to the ED. Another asked whether noninvasive ventilation prevents intubation or hospital admission compared with standard treatment in patients who present to the ED with acute congestive heart failure exacerbation. The final question centered on identifying the predictive value of diagnostic criteria in assessing whether head computed tomography is necessary in patients who present to the ED with minor head injuries.
The PubMed narrow search resulted in the highest number of high-quality hits, totaling 20, whereas the Google Scholar search had the highest number relevant hits, totaling 50. The highest quality-to-relevance ratio was attributed to the PubMed narrow search tool, averaging at 0.85. PubMed Clinical Queries broad search featured the second highest mean quality relevance ratio of 0.52. Quality relevance ratio was lowest with Google Web search, averaging 0.14. In addition, a basic Google Web search yielded duplicate hits. In the first 60 results, Google Scholar provided the highest number of systematic reviews (n=6) and randomized controlled trials (n=12).
Limitations of this pilot study were the single-rater design, the lack of previous validation of the quality relevance ratio, and the analysis of only the abstracts of returned citations.
The study does not discount the utility of a standard Google or PubMed search for clinicians. “For a greater breadth of information,” the researchers wrote, “searches may be supplemented by a Google Scholar search.”
Morshed T, Hayden S. Google versus PubMed: comparison of Google and PubMed’s search tools for answering clinical questions in the emergency department [published online October 14, 2019]. Ann Emerg Med. doi:10.1016/j.annemergmed.2019.07.003