Online Supplement
Title:
A SCOPING REVIEW OF INTERSECTORAL ACTION FOR HEALTH EQUITY INVOLVING GOVERNMENTS
Journal:
International Journal of Public Health
Ketan Shankardass1,2
Orielle Solar3
Kelly Murphy1
Lorraine Greaves4
Patricia O’Campo1
Author Affiliations:
1 The Centre for Research on Inner City Health, The Li Ka Shing Knowledge Institute of St. Michael's Hospital, Toronto, Canada
2 Department of Psychology, Wilfrid Laurier University, Waterloo, Canada
3 University of Chile, Santiago, Chile
4 Senior investigator, British Columbia Centre of Excellence for Women’s Health, Vancouver, Canada
Corresponding Author:
Ketan Shankardass
Email:
Telephone: 416-864-6060 ext. 77360
Fax: 416-864-5485
Methods
Scoping review process
The scoping review process involved four stages for winnowing down from an initial collection of 5,343 articles that possibly described government-centred intersectoral action for health equity into distinct case-articles describing specific examples, and a fifth stage for summarizing information from across case-articles.
(1) Searching: A strategy for comprehensively searching scholarly and grey literature for information about cases of government-centred intersectoral action for health equity was designed by an information specialist on the research team. Appropriate wildcards were used when searching for literature in order to account for plurals and variations in spelling.
a. Scholarly Literature: Between June 2-4, 2010 the following research databases were searched with no limitation on dates: International Bibliography of the Social Sciences, Medline, EMBASE, Worldwide Political Science Abstracts, PAIS International and PAIS Archive were searched and retrieved 1468, 1031, 431, 415, 362, 81 articles (including some books), respectively. Since one member of our research team was fluent in Spanish and Portuguese, Scientific Electronic Library Online (SCIELO) and Literatura latinoamericana en ciencias de la salud (LILACS) were also searched for articles in these languages with all search terms, on June 7, 2010 and retrieved 4 and 92 articles, respectively. In total, 4833 articles were identified during the scholarly literature search, including 964 duplicates and 52 articles for which a full text version could not be obtained in English, Spanish or Portuguese.
A few potentially relevant journals that were not originally included by the databases searched for scholarly literature were manually searched using three key search terms: “intersectoral” and “policy or collaboration or action or cooperation” and “health or equity or inequity”; Health in all policies; and, “joined up government” and “health”. This process resulted in 1 additional article.
b. Grey Literature Search: Keyword combinations developed for the search of scholarly literature were searched in Google and the first 30 hits per search term were captured and saved in a Microsoft Word document. Seventeen keyword combinations were used in this manner, meaning that 510 Google links were recovered for screening. Literature was also obtained through mass communication with four relevant email listserves, including: The Spirit of 1848: A caucus of the American Public Health Association, the Asia Pacific Health Impact Assessment Listserv, the Pan-American Health Organization Listserv, and the Social Determinants of Health listserv; as well as participants of the 2010 Health in All Polices International Meeting in Adelaide (co-sponsored by the South Australian Government and the World Health Organization). In these cases, individuals were requested to provide information regarding international examples of intersectoral efforts to address health equity or Health in all policies initiatives specifically, or any scholarly/grey literature that offer theoretical perspectives on these approaches. Examples of intersectoral action for health equity or Health in all policies subsequently brought to the attention of the research team were compared to examples already identified by earlier strategies and no novel examples were identified, thus supporting the inclusive nature of the primary search strategy.
(2) Screening: In order to be inclusive, and to draw lessons from failed attempts as well as projects that are in progress, we did not exclude potential cases of intersectoral action based on whether or not they were characterized as successful, nor recent or new examples of intersectoral action that had not reached an endpoint. Six members of the research team reviewed abstracts (or in the case of some grey literature, full documents) in order to classify them into one of three folders representing the level of agreement that the abstract reflected a case of intersectoral action for health equity: Yes, No, and Maybe. An abstract was classified as ‘Yes’ if it satisfied all three criteria described above and ‘Maybe’ if the abstract wasn’t completely informative (e.g., if the number of government sectors involved was not readily apparent). Finally, abstracts were classified as ‘No’ if they clearly did not satisfy all three criteria. Some records in Refworks were found to only contain article titles, lacking an abstract. In these cases, records were screened initially based on article title only, and articles that were suggestive of intersectoral action for health equity were classified using a fourth category: Promising. During this stage of the process, each abstract and title was reviewed by a single researcher.
Once all abstracts were screened, the full-text articles for ‘Promising’ titles and ‘Maybe’ abstracts were downloaded or accessed at the University of Toronto library. The research team was divided into three groups of 2 – 3 members, who proceeded to read the full text version of ‘Promising’ and ‘Maybe’ articles and classify as either ‘Yes’ or ‘No’ according to the inclusion criteria. This step of the screening phase was performed in groups to ensure multiple members reviewed each article. In the case that there was non-consensus among a 2- or 3-person screening group, the article was included if at least one group member classified it as ‘Yes’.
After screening the abstracts and titles there were, 59 ‘Yes’, 65 ‘Promising’, 146 ‘Maybe’ and 4470 ‘No’. At the end of the screening stage 122 articles or books were identified as referring to potential cases of intersectoral action for health equity. A further 381 articles, white papers, speeches, etc. were recovered from the grey literature search. Over the course of project, a further 30 pieces of literature were found (e.g., from a listserv response) that met the screening criteria.
(3) Sorting: Literature referring to potential cases of intersectoral action for health equity was then sorted by the research team by country and sub-national region. When ‘families’ of articles were found describing the same or related intersectoral actions in the same setting to address the similar or related problems (cf. Kingdon’s (1984) notion of the “problem stream”), these were sorted together into a single case. A sorting table was developed which identified the following information for all articles: the specific setting and geographic level (e.g., city, province/state, country) of intersectoral action for health equity; the case country (which was sometimes the same as the setting); and, the number of ‘Yes’ documents per country.
(4) Scoping: The “scoping table” for extracting information from articles that described cases of government-centred intersectoral action for health equity was designed to indicate the presence or abstract of information about the initiation and implementation of specific initiatives, as well as some categories for which the extraction of more detailed information would be feasible and could eventually contribute to the identification of strong cases of Health in all policies initiatives during a subsequent phase of this study.
Of the 83 countries with potential cases of intersectoral action for health equity identified during the Sorting stage, 43 were confirmed to have at least one government-centred approach following the scoping process. This reflected information from 128 unique articles, although there were 194 case-articles in the scoping table as a result of some articles describing multiple cases of intersectoral action for health equity.
(5) Summary: A summary of scoping categories was prepared to describe how and why government-centred intersectoral approaches to health equity have been introduced and implemented across various settings. The proportion of case-articles for which insufficient information prevented classification was also described in terms of implications for more intensive knowledge synthesis based on this literature.
References
Kingdon JW (1984) Agendas, Alternatives, and Public Policies. Little, Brown & Co., Boston
1