Article type
Year
Abstract
Background: A growing number of systematic reviews are written in areas outside clinical medicine. Thus, reviewers increasingly search non-medical as well as medical databases and consider a variety of types of research evidence. This is particularly true in the area of public health. Being able to search effectively by study design as well as subject topic can reduce the volume of records that need to be assessed for relevance and improve the efficiency of the review process.
Objectives: To provide practical guidance for searchers, a collaboration was established between the Health Development Agency, the Centre for Reviews & Dissemination and the Health Evidence Bulletins Wales to look at the indexing terms used to describe different evidence types in databases of relevance to public health.
Methods: A number of databases were examined from the fields of health care, social science, education, environment, ongoing research & grey literature, including specialist public health/health promotion databases. The indexing policies of each database were assessed by searching both the subject headings (controlled vocabulary), if available, and text words. A large number of research terms were examined, covering randomised trials, non-randomised trials and observational studies as well as reviews of these studies.
Results: Tables of the most appropriate subject heading and text words were drawn up for each evidence type within each database, along with other pragmatic advice to information searchers in developing a sensitive search strategy within a research area. This guidance has been written up and is also available as a searchable database on the Health Development Agencys web-site (**url to be added).
An example table of some of the social science databases and research terms is given below (Table 1)
Conclusions: Thorough and systematic searching for evidence from research of relevance to public health requires involves a range of databases and a variety of approaches. As far as the authors are aware, there is little evidence on the reliability of indexing of study design in the databases which offer subject headings. Until such evidence becomes available the authors suggest that searchers should combine the available subject headings with free text terms to ensure sensitive searching.
This project offers practical advice for reviewers. The next step will be to encourage searchers to make use of, and evaluate, this information so that lessons learnt can be fed back to further develop, update and expand the resource.
Objectives: To provide practical guidance for searchers, a collaboration was established between the Health Development Agency, the Centre for Reviews & Dissemination and the Health Evidence Bulletins Wales to look at the indexing terms used to describe different evidence types in databases of relevance to public health.
Methods: A number of databases were examined from the fields of health care, social science, education, environment, ongoing research & grey literature, including specialist public health/health promotion databases. The indexing policies of each database were assessed by searching both the subject headings (controlled vocabulary), if available, and text words. A large number of research terms were examined, covering randomised trials, non-randomised trials and observational studies as well as reviews of these studies.
Results: Tables of the most appropriate subject heading and text words were drawn up for each evidence type within each database, along with other pragmatic advice to information searchers in developing a sensitive search strategy within a research area. This guidance has been written up and is also available as a searchable database on the Health Development Agencys web-site (**url to be added).
An example table of some of the social science databases and research terms is given below (Table 1)
Conclusions: Thorough and systematic searching for evidence from research of relevance to public health requires involves a range of databases and a variety of approaches. As far as the authors are aware, there is little evidence on the reliability of indexing of study design in the databases which offer subject headings. Until such evidence becomes available the authors suggest that searchers should combine the available subject headings with free text terms to ensure sensitive searching.
This project offers practical advice for reviewers. The next step will be to encourage searchers to make use of, and evaluate, this information so that lessons learnt can be fed back to further develop, update and expand the resource.