Article type
Year
Abstract
Background: The effective search for data is primordial for the development of evidence-based clinical practice guidelines. In the setting of the French Federation of Comprehensive Cancer Centres guidelines programme (SOR) a series of generic search strategies have been elaborated to facilitate this aspect of guideline development. Clinical problems can be formulated using the PICO model ( Patient or Pathology, Intervention, Compared intervention and Outcome) for facilitating the elaboration of a search strategy.
Objective: To optimise literature searching in the guidelines development process by providing search strategy modules for MEDLINE, EMBASE, and the Cochrane Controlled Trials Database. The main aims are to provide a more systematic and homogenous literature searching step within the guidelines programme, and to avoid unnecessary duplication of effort between working groups.
Methods: Standardised search strategies were elaborated and tested for the different biomedical databases. These had a modular structure, based on the PICO model, which enabled the different modules to be combined to define a specific search strategy. As outcomes are rarely indexed as keywords in biomedical databases, this module was not taken into account. However, a module corresponding to the methodological aspects was elaborated.
Results: Modules defining patient or pathology (disease site + disease stage), intervention (management step(s) + intervention(s) evaluated) and methodology (study design relevant to the question + other methodological aspects) have been elaborated and tested. These modules are stored in a database which can be accessed when a guidelines working group needs to run a literature search. These tools are progressively created according to the litterature search needs. When a module is improved or a new module is elaborated, the database is updated.
Conclusion: The development of evidence-based guidelines is an ideal setting for this modular approach to literature searching. The storage of the modules in a central database facilitates access by the guidelines working groups for elaborating specific search strategies, building on previous work. This approach avoids unnecessary duplication of effort and provides a useful tool for supporting the work of guidelines working groups and leads to more systematic, homogeneous search strategies for the guidelines produced within our guideline programme.
Objective: To optimise literature searching in the guidelines development process by providing search strategy modules for MEDLINE, EMBASE, and the Cochrane Controlled Trials Database. The main aims are to provide a more systematic and homogenous literature searching step within the guidelines programme, and to avoid unnecessary duplication of effort between working groups.
Methods: Standardised search strategies were elaborated and tested for the different biomedical databases. These had a modular structure, based on the PICO model, which enabled the different modules to be combined to define a specific search strategy. As outcomes are rarely indexed as keywords in biomedical databases, this module was not taken into account. However, a module corresponding to the methodological aspects was elaborated.
Results: Modules defining patient or pathology (disease site + disease stage), intervention (management step(s) + intervention(s) evaluated) and methodology (study design relevant to the question + other methodological aspects) have been elaborated and tested. These modules are stored in a database which can be accessed when a guidelines working group needs to run a literature search. These tools are progressively created according to the litterature search needs. When a module is improved or a new module is elaborated, the database is updated.
Conclusion: The development of evidence-based guidelines is an ideal setting for this modular approach to literature searching. The storage of the modules in a central database facilitates access by the guidelines working groups for elaborating specific search strategies, building on previous work. This approach avoids unnecessary duplication of effort and provides a useful tool for supporting the work of guidelines working groups and leads to more systematic, homogeneous search strategies for the guidelines produced within our guideline programme.