The search methodology comprised two components. The first, and primary, component was an electronic bibliographic database and website search. The second, supplementary, component was through personal com- munication with experts in trade in health services, to establish whether any work had been missed in the electronic search, or there was important work in progress. Both of these components are outlined briefly below.
Electronic bibliographic database and website search
An extensive range of health and economic databases were searched (EconLit, ISI, Medline, EMBASE, Grey Literature, OPAC, Cochrane Library Online and Web of Science), using a combination of search terms (FDI, Trade, Health and Health Services) over the period from January 1990 to December 2002, supplemented by a search of web sites relevant to trade in health services.
Where possible, search terms were used across fields concerning title, abstract and subject heading. The search strategy was tailored according to the specific database chosen, such that (combinations of) terms were used in order that the broadest range of literature was identified initially, and then narrowed down as necessary.
Literature search results
The bibliographic databases yielded 67 possibly relevant ‘hits’ which were downloaded (abstracts where possible, titles otherwise) and reviewed. Of these only 19 were deemed relevant and the full paper obtained and reviewed. The websites together yielded 49 potentially important papers, which were obtained and reviewed.15
Areviewofthereferencescontainedwithinthosepapers obtained yielded six additional papers for review. The expert contacts yielded two more relevant references. Thus, in total, 76 papers, books and reports were reviewed fully (marked by ? in the reference list). An annotated bibliography of all these 76 papers is available from the author upon request.
Criteria for prioritization of