Related Experiment Videos
Optimizing search strategies to identify randomized controlled trials in MEDLINE
Li Zhang1, Isola Ajiferuke, Margaret Sampson
1Natural Sciences Library, University of Saskatchewan, Saskatoon, Canada. li.zhang@usask.ca
BMC Medical Research Methodology
|May 11, 2006
Summary
This study found that adding "versus" to the Cochrane Highly Sensitive Search Strategy (HSSS) slightly improved sensitivity for identifying Randomized Controlled Trials (RCTs) but missed relevant articles. Using the full three-phase HSSS is recommended for comprehensive systematic reviews.
Area of Science:
- Medical Informatics
- Systematic Review Methodology
- Evidence-Based Medicine
Background:
- The Cochrane Highly Sensitive Search Strategy (HSSS) is a three-phase approach for identifying Randomized Controlled Trials (RCTs) in MEDLINE.
- Four revisions of the HSSS were proposed, involving combinations of free text terms and the Medical Subject Heading CROSS-OVER STUDIES with the initial phases of the HSSS.
Purpose of the Study:
- To validate four proposed revisions of the Cochrane HSSS for identifying RCTs.
- To assess the sensitivity and precision of different HSSS search strategies.
Main Methods:
- The study replicated subject searches for 61 Cochrane reviews.
- Included studies indexed in MEDLINE were pooled and combined with proposed search strategies, the top two phases of HSSS, and all three phases of HSSS.
- Sensitivity and precision were calculated for six distinct search strategies for each review.
Main Results:
- The search term "versus" combined with the top two phases of HSSS identified 3 more studies than the top two phases alone but missed 56 relevant articles compared to the full three-phase HSSS.
- Estimated time savings were noted for revised strategies (e.g., "versus" strategy at 823 hours) compared to the full three-phase HSSS (1086 hours).
Conclusions:
- Adding "versus" to the first two phases of HSSS offers a minor balance in sensitivity and precision but does not fully compensate for missing relevant studies.
- Reviewers must carefully consider the trade-offs between search comprehensiveness and efficiency when selecting HSSS strategies to minimize bias in systematic reviews.