Search strategies to retrieve studies related to lowand middle-income countries (LMICs)

Article type
Authors
Johansen M1
1Department of Evidence-Based Health Services, Norwegian Knowledge Centre for the Health Services, Oslo, Norway
Abstract
Background: The Norwegian Satellite of the Cochrane Effective Practice and Organisation of Care (EPOC) Group supports the production of Cochrane reviews that address questions regarding health systems relevant to LMICs. To our knowledge there is currently no LMIC filter to help identify relevant studies for these reviews. Objectives: To develop LMIC filters for MEDLINE, EMBASE, PubMed and CENTRAL Methods: Filters are based on the World Bank list of countries, classified as low-income, lower-middle-income or upper-middle-income economies. Countries either currently or formerly classified as an LMIC are included. Countries are listed alphabetically. However, for a country that formerly had another name, the alphabetic listing is violated in order to group names together. Countries are not grouped according to economies since classification can vary from one year to another. Each filter is a combination of index terms and text words. The following fields are searched: Subject Heading, Key Words, Subject Heading Word, Keyword Heading Word, Title, Abstract, Other Terms, Country of Publication, Place of Publication. Results: Filters have been developed for 4 databases. They have not been tested for sensitivity and precision. Number of records retrieved by mid March 2010 is about: MEDLINE (Ovid from 1950): 2 million of 17 million records; EMBASE (Ovid from 1980): 100 thousand of 8 million records; PubMed: 2.5 million of 19 million records; CENTRAL: 18 000 of 600 000 records. Of these 18 000, 295 are submitted from the EPOC Register. Filters can be found by following the link to LMIC Databases on the Norwegian EPOC satellite web site: http://epocoslo.cochrane.org/. Conclusions: Reviews that address questions relevant to LMICs might benefit from adding an LMIC filter to the search strategies used to identify studies for inclusion. Further evaluation of these filters is necessary.