Can electronic search engines improve systematic reviews?
Author Information
Author(s): Sampson Margaret, Barrowman Nicholas J, Moher David, Clifford Tammy J, Platt Robert W, Morrison Andra, Klassen Terry P, Zhang Li
Primary Institution: Children's Hospital of Eastern Ontario Research Institute
Hypothesis
Can an Ultraseek search engine effectively rank MEDLINE records of included studies in systematic reviews?
Conclusion
The Ultraseek search engine's relevance ranking was better than expected by chance, indicating potential for improving systematic review efficiency.
Supporting Evidence
- Included studies were found in the top 500 retrieved studies more often than expected by chance.
- Recall of included studies was 0.70 across all reviews.
- Statistical search engine performance was better than random ranking in 7 out of 9 reviews.
Takeaway
This study looked at whether a special search engine could help find important studies faster when researchers are reviewing lots of medical information.
Methodology
The study tested the Ultraseek search engine's ability to rank MEDLINE records from systematic reviews by comparing included studies' rankings to expected random rankings.
Limitations
The study's performance depended on the selection and ordering of search terms, which could affect results.
Statistical Information
P-Value
<0.001
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website