A framework for determining the applicability and practical feasibility of a systematic review: experience from the area of risk assessment for food and feed safety

Poster
2010 Keystone
Aiassa E1, Higgins J2, Frampton G3, Greiner M4, Afonso A5, Amzal B6, Deeks J7, Dorne J8, May-Glanville J9, L vei G10, Nienstedt K11, O Connor A12, Pullin A13, Raji A14, Verloo D3
1Assessment Methodology Unit, European Food Safety Authority, Parma, Italy
2MRC Biostatistics Unit, University of Cambridge, Cambridge, UK
3Southampton Health Technology Assessments Centre, School of Medicine, University of Southampton, Southampton, UK
4Scientific Services Unit 33 - Epidemiology, Biostatistics and Mathematical Modelling, Federal Institute for Risk Assessment (BfR), Berlin, Germany
5Animal Health and Welfare Unit, European Food Safety Authority, Parma, Italy
6Program for HIV Prevention and Treatment, Institut de Recherche pour le D éveloppement, Chiang Mai, Thailand
7Public Health, Epidemiology and Biostatistics, University of Birmingham, Birmingham, UK
8Unit on the Contaminants in the food chain, European Food Safety Authority, Parma, Italy
9York Health Economics Consortium, York, UK
10Faculty of Agricultural Sciences, Dept. of Integrated Pest Management, Flakkebjerg Research Centre, Aarhus University, Slagelse, Denmark
11Plant Protection Products and their residues, European Food Safety Authority, Parma, Italy
12Veterinary Diagnostic and Production Animal Medicine, College of Veterinary Medicine, Iowa State University, Ames, Iowa, United States
13Centre for Evidence-Based Conservation School of the Environment, Natural Resources and Geography, Bangor University, Bangor, UK
14Public Health Agency of Canada, Universities of Saskatchewan and Guelph, Ottawa and Guelph, Ontario, Canada
Background: Systematic reviews (SRs) are suitable for summarising evidence for clearly specified clinical questions. Food and feed safety risk assessments (FFSRA), however, address broader topics. No methodological framework exists for refining such broad problems into specific, reviewable questions, and for assessing the practical feasibility of SRs. Objectives: To develop a framework for refining broad problems, determining questions suitable for SRs, and prioritising questions for formal SRs in the area of FFSRA. Methods: The European Food Safety Authority (EFSA) formed a working group of SR methodologists from The Cochrane Collaboration, other research groups, and experts in FFSRA, and developed, through multiple group discussions, an initial generic framework for the application of SRs to broad food and feed safety problems. Results: The first step in assessing suitability for SRs involves determining the question type, which may include interventions or exposures with the well-known PICO structure; test accuracy questions ( PIT structure Population(s), Index test(s), Target condition(s)); and descriptive questions ( PO structure Population(s), Outcome or condition(s) of interest). The key elements (e.g. P, I, C and O) form either closed-framed questions (where all key elements are specified) or open-framed questions (one or more key elements are unspecified). Open-framed questions are less suitable than closed-framed questions for SR, but may be refined into closed-framed questions by specifying the missing key elements (e.g. through further research) or by revising the question. We also propose a procedure for prioritising which questions to answer by SR, through considering the relative importance of different parameters and issues of practical feasibility. Conclusions: A framework for planning and prioritising SRs in FFSRA is being developed, and will continue to evolve as experience is gained through initial implementation. It may have useful implications for improving transparency in FFSRA and for early stages of planning Cochrane reviews.