Novel citation-based search method for scientific literature: application to meta-analyses.

摘要: Finding eligible studies for meta-analysis and systematic reviews relies on keyword-based searching as the gold standard, despite its inefficiency. Searching based direct citations is not sufficiently comprehensive. We propose a novel strategy that ranks articles their degree of co-citation with one or more “known” before reviewing eligibility. In two independent studies, we aimed to reproduce results literature searches sets published meta-analyses (n = 10 n = 42). For each meta-analysis, extracted co-citations randomly selected ‘known’ from Web Science database, counted frequencies screened all score above selection threshold. second study, extended method by retrieving articles. first retrieved 82 % included in while screening only 11 % many were original publications. Articles missed non-English languages, 1975, very recently, available conference abstracts. 79 % half number Citation appears be an efficient reasonably accurate finding similar interest reviews.