Article type
Year
Abstract
Background: When searching for intervention trials, free medical databases such as Medline are universally utilized. However, little is known about the utility of extending the search to fee-based databases, such as Embase. Ulrich s database lists more than 260 specialty journal titles in the field of gastroenterology (GI) worldwide, and less than 40% are covered either by Medline or Embase. The journal overlap of Medline to Embase have been reported to be in the 70% range, suggesting the need to extend a search to Embase, particularly when performing systematic reviews.
Relying on journal count or impact factor related gold standards to define the maximum utility of a medical database is problematic, since both approaches are biased. Instead, we would argue that for intervention related questions, the maximum utility is derived from randomized controlled trials and controlled clinical trials. The Cochrane Central Register of Controlled Trials (CENTRAL) is the most comprehensive source of controlled trials in health care. We therefore defined CENTRAL as the gold standard for the purpose of this study.
Objectives: Compared to Medline, what is the incremental benefit of the fee-based medical database Embase, utilizing the Cochrane Central Register of Controlled Trials (CENTRAL) as the gold standard, to answer questions related to medical interventions in gastroenterology?
Methods: CENTRAL (Issue 1, 2004) was searched with 81 keywords or key phrases derived from 308 GI subspecialty journal titles identified through various lists (e.g., Ulrich s, Medline and Embase). Citations were then assigned according to the time period of journals indexed in Medline and Embase. GI trial citations were only associated with the database Embase if they would not have been found in Medline. We calculated the incremental benefit of Embase in terms of percentage of trial citations not identifiable through Medline.
Results: A total of 24,931 trials citations published in more than 300 journal titles in the field of gastroenterology were identified. About one third of all GI journals were listed in Medline and one sixth of GI journals furnished trial citations from Embase not identifiable through Medline. However, the absolute number of Embase specific trial citations were only 699 of 24,931 citations, or less than 3%. 16,807 (68%) citations had Medline coverage.
Discussion: Using CENTRAL as the gold standard, Medline covers about two-thirds of controlled trials in gastroenterology, hepatology and nutrition. From a journal coverage perspective, the addition of Embase would increase this coverage by approximately 40%. However, Embase incremental benefit in identifying intervention studies in gastroenterology is minimal (<3%), although this may vary from specialty to specialty.
Implications: Because of this limited additional benefit, an Embase search for intervention studies in gastroenterology may not always be necessary or cost-effective. Further studies are needed to assess the extend of impact of Embase specific trials on the results of systematic reviews.
Relying on journal count or impact factor related gold standards to define the maximum utility of a medical database is problematic, since both approaches are biased. Instead, we would argue that for intervention related questions, the maximum utility is derived from randomized controlled trials and controlled clinical trials. The Cochrane Central Register of Controlled Trials (CENTRAL) is the most comprehensive source of controlled trials in health care. We therefore defined CENTRAL as the gold standard for the purpose of this study.
Objectives: Compared to Medline, what is the incremental benefit of the fee-based medical database Embase, utilizing the Cochrane Central Register of Controlled Trials (CENTRAL) as the gold standard, to answer questions related to medical interventions in gastroenterology?
Methods: CENTRAL (Issue 1, 2004) was searched with 81 keywords or key phrases derived from 308 GI subspecialty journal titles identified through various lists (e.g., Ulrich s, Medline and Embase). Citations were then assigned according to the time period of journals indexed in Medline and Embase. GI trial citations were only associated with the database Embase if they would not have been found in Medline. We calculated the incremental benefit of Embase in terms of percentage of trial citations not identifiable through Medline.
Results: A total of 24,931 trials citations published in more than 300 journal titles in the field of gastroenterology were identified. About one third of all GI journals were listed in Medline and one sixth of GI journals furnished trial citations from Embase not identifiable through Medline. However, the absolute number of Embase specific trial citations were only 699 of 24,931 citations, or less than 3%. 16,807 (68%) citations had Medline coverage.
Discussion: Using CENTRAL as the gold standard, Medline covers about two-thirds of controlled trials in gastroenterology, hepatology and nutrition. From a journal coverage perspective, the addition of Embase would increase this coverage by approximately 40%. However, Embase incremental benefit in identifying intervention studies in gastroenterology is minimal (<3%), although this may vary from specialty to specialty.
Implications: Because of this limited additional benefit, an Embase search for intervention studies in gastroenterology may not always be necessary or cost-effective. Further studies are needed to assess the extend of impact of Embase specific trials on the results of systematic reviews.