Article type
Year
Abstract
Background: The reporting of search strategies and results has implications for critical appraisal and reproducibility of systematic reviews (SR).
Objectives: To compare reported and recommended searching practices.
Methods: Data sources: Rating scales addressing the conduct or reporting of the search strategies in SRs, health technology assessments (HTA) or guidelines were identified through a systematic review. These were compared to the suggested reporting elements in The Cochrane Handbook for Systematic Reviews of Interventions. Current reporting practice data are from a cohort of 297 SRs that were newly indexed in MEDLINE in November 2004.
Results: Of the 11 scales examined (including the Cochrane Handbook), eight addressed SRs and three were developed for HTA
reports or guidelines. Six scales were reporting standards and five were critical appraisal tools. Three scales addressed only the search; the rest addressed the search and other aspects of the report. Seven cited supporting evidence but only three were validated. Four scales provided or cited data on current practice or provided examples. Across all scales, 18 different elements were suggested. The only element common to all scales was databases used. Eight elements appeared on at least five scales. There was a trend toward including more elements in more recent scales (r=0.41). Scales focused on the search included slightly more search-related elements (median=9) than did global scales (median=7.35). Current reporting practices for elements appearing on five or more scales ranged from a low of 43% for reporting the full electronic search strategy to 99% for databases used. Four of the elements were reported in less than 80% of reviews. Cochrane reviews show more complete reporting than non-Cochrane on all elements except presence or absence of language restrictions reported.
Conclusions: There is no clear consensus regarding optimum reporting of SR search methods and commonly recommended elements show suboptimal reporting.
Objectives: To compare reported and recommended searching practices.
Methods: Data sources: Rating scales addressing the conduct or reporting of the search strategies in SRs, health technology assessments (HTA) or guidelines were identified through a systematic review. These were compared to the suggested reporting elements in The Cochrane Handbook for Systematic Reviews of Interventions. Current reporting practice data are from a cohort of 297 SRs that were newly indexed in MEDLINE in November 2004.
Results: Of the 11 scales examined (including the Cochrane Handbook), eight addressed SRs and three were developed for HTA
reports or guidelines. Six scales were reporting standards and five were critical appraisal tools. Three scales addressed only the search; the rest addressed the search and other aspects of the report. Seven cited supporting evidence but only three were validated. Four scales provided or cited data on current practice or provided examples. Across all scales, 18 different elements were suggested. The only element common to all scales was databases used. Eight elements appeared on at least five scales. There was a trend toward including more elements in more recent scales (r=0.41). Scales focused on the search included slightly more search-related elements (median=9) than did global scales (median=7.35). Current reporting practices for elements appearing on five or more scales ranged from a low of 43% for reporting the full electronic search strategy to 99% for databases used. Four of the elements were reported in less than 80% of reviews. Cochrane reviews show more complete reporting than non-Cochrane on all elements except presence or absence of language restrictions reported.
Conclusions: There is no clear consensus regarding optimum reporting of SR search methods and commonly recommended elements show suboptimal reporting.