Evidence about which information resources should be searched to identify economic evaluations (EEs) of healthcare interventions when conducting a systematic review (SR) predates closure of the National Health Service Economic Evaluation Database (NHS EED) and Health Economic Evaluations Database (HEED). We assessed which databases are now the best sources of EEs and identify the most efficient combination of databases, taking into account the order in which databases could be searched.
We gathered a reference set of EEs from published reviews of EEs undertaken to inform Health Technology Assessments (HTA). We calculated yield and relative recall (RR) (number of reference set records identified / total number of records in reference set) for each database, and combination of databases. We assessed the order in which databases should be searched, to identify the most efficient combination of databases to identify the reference set. We report the characteristics of records not included in any database studied and implications for identifying this type of evidence.
To date, a reference set of fifty-five EEs from seven HTAs has been processed. Embase and Scopus each yielded 53/55 records (RR .96). MEDLINE yielded 52/55 (RR .95). Embase or Scopus included all of the journal publications in the reference set; no additional unique records were provided by MEDLINE, CEA Registry, EconLit, or Science and Social Science Citation Indexes. The two records that were not identified were unpublished evidence, one of which was included in the National Institute for Health Research (NIHR) HTA database. Processing will continue until we reach the threshold of a reference set of 350 records.
Preliminary results suggest that searching two or three databases may be most efficient, provided that resources are searched using appropriate strategies. Searchers should concentrate on developing search strategies that work well in those databases to ensure adequate sensitivity, and use freed time to identify grey literature.