15.3.1  Use of electronic search filters

Search methods for locating relevant health economics studies will differ depending on the scope of a critical review of such studies and the types of studies to be considered for inclusion (see also Sections 15.2.3 and 15.1.2). However, in all cases the first stage of the search strategy will have the same objective: to identify effectiveness studies retrieved for initial screening and potential inclusion in a Cochrane review which include relevant health economics studies.

 

Electronic records of effectiveness studies retrieved from electronic literature databases can be filtered using search strategies designed to capture health economics studies. This can precede visual screening of abstracts and full texts of studies, acting as an aid to location of economic studies by limiting the number of records to be assessed. Electronic filtering is most useful in reviews where the number of records retrieved from electronic literature databases is large (i.e. where this number is relatively small, use of electronic filters may not be judged necessary, but explicit criteria would still need to be applied).

 

The Centre for Reviews and Dissemination (CRD) has developed a series of electronic search strategies designed to capture potential economic evaluation studies for inclusion in the NHS Economic Evaluation Database (NHS EED). MEDLINE (Ovid CD-ROM), CINAHL (Ovid CD-ROM), EMBASE (Ovid online) and PsychINFO (Ovid online) versions are published in the NHS EED Handbook (Craig 2007) and online at www.york.ac.uk/inst/crd/nfaq2.htm. Each of these search strategies can be appended to review-specific search strategies of the corresponding database using the ‘AND’ operator, to filter search results for records which also contain ‘economics’ search terms.

 

These NHS EED search strategies are very broad and will capture economics methods studies and reviews of economics studies, as well as the full range of types of health economics studies (see Section 15.1.2). For more specific searches, narrower adaptations of the search strategies and close reading of the scope notes of MeSH are advised. The search strategies can also be adapted, in consultation with information retrieval specialists, for use in other electronic literature databases. Adaptation of the search strategies will need to take into account variations across databases in the indexing or classification of health economics studies. A useful annotated list of electronic literature databases that include coverage of health economics literature and details of internet sites containing relevant grey literature is available (Napper 2005).

 

An important procedural consideration when considering use of electronic search filters designed to capture health economics studies is that Cochrane reviews also frequently utilize other search filters designed to capture other specific study designs, such as randomized trials. These ‘study design search filters’ are also appended to review-specific search strategies using the ‘AND’ operator. Therefore, if the scope of the critical review is not restricted to health economics studies conducted alongside effectiveness studies included in the effectiveness component of the review (e.g. will also include model-based economic evaluations: see Section 15.2.3), then the ‘economics search filter’ should be appended to any other ‘study design search filter’ using the ‘OR’ operator, to ensure that all types of health economic studies to be considered are retrieved. Alternatively, if the scope of the critical review is limited to health economics studies conducted alongside effectiveness studies included in the effectiveness component of the review, then use of the ‘economics search filter’ is not required, since most of the economics studies to be considered will be retrieved using the ‘study design search filter’ (although, it is possible that in this case the search results may still omit some relevant economics studies, such as economic evaluations based on randomized trials but published separately from and usually after the trial results).