Article type
Year
Abstract
Background: Mapping the evidence regarding the current state of research in medical treatments or procedures is a quite new methodological approach. This kind of overview may be helpful for the development of research questions and decision making about their relevance and priority. However, the amount of data available today leads to challenges for scientists who are sometimes confronted with literature searches that retrieve over 30,000 results for screening and 4000 full texts for further examination.
Objectives: Our aim was to produce an evidence map on the topic ‘diabetes and driving’ for investigating its feasibility for a national clinical guideline. In addition, we compared a highly sensitive search strategy with a highly specific one.
Methods: We conducted systematic literature searches in the databases Embase, MEDLINE, and the Cochrane Library in April 2014. Based on a systematic review, the searches were limited to publications from the year 2002 through to the present, and to the English and German languages.
Results: Due to the strongly focused topic and the search limits, our highly sensitive string identified a manageable number of references including sufficient evidence for a decision in favour of creating a guideline. Using the specific search strategy, we achieved a reduction of citations by 25%, concurrently identifying 88% of relevant references.
Conclusions: The purpose of evidence maps is to provide an overview that is a prerequisite for decisions, but not to create the high level of accuracy as produced by systematic reviews.
Our results confirm those of Parkhill 2011 concerning a clinical question about whether highly specific searches lead to a sufficient or even the same number of relevant citations for decision making compared to sensitive searches, while avoiding a mass of extraneous information.
Objectives: Our aim was to produce an evidence map on the topic ‘diabetes and driving’ for investigating its feasibility for a national clinical guideline. In addition, we compared a highly sensitive search strategy with a highly specific one.
Methods: We conducted systematic literature searches in the databases Embase, MEDLINE, and the Cochrane Library in April 2014. Based on a systematic review, the searches were limited to publications from the year 2002 through to the present, and to the English and German languages.
Results: Due to the strongly focused topic and the search limits, our highly sensitive string identified a manageable number of references including sufficient evidence for a decision in favour of creating a guideline. Using the specific search strategy, we achieved a reduction of citations by 25%, concurrently identifying 88% of relevant references.
Conclusions: The purpose of evidence maps is to provide an overview that is a prerequisite for decisions, but not to create the high level of accuracy as produced by systematic reviews.
Our results confirm those of Parkhill 2011 concerning a clinical question about whether highly specific searches lead to a sufficient or even the same number of relevant citations for decision making compared to sensitive searches, while avoiding a mass of extraneous information.