Common Format for Searching Literature and Synthesizing the Evidence

Common format for searching literature and synthesizing the evidence

1 Formulating the key clinical questions
Perform a multidisciplinary problem analysis
Formulate structured key clinical questions using the PICO format
Formulate these structured key clinical questions with input from clinical experts. ‘
Document the key clinical question in the guidelineand identify the excluded key questions (if present).
2 Searching studies
Define selection criteria (e.g. study design, language and date) for studies a priori for each key clinical question.
Develop a search strategy for each key clinical question.
Include the selection of bibliographic databases (at least 2 databases including Medline) and websites to be searched in the search strategy.
Document any strategies used or decisions made to reduce/minimise/manage the amount of information retrieved, for example: selection criteria, use of methodological filters, making a trade off between exhaustive and more modest searching (sensitivity and specificity).
Carry out the search strategy by qualified information specialists or experts trained in literature search. Methodological and clinical input is highly recommended.
Document the search strategy including search terms in a reproducible way.
Document bibliographic sources/databases.
3 Selecting the evidence
Document the study selection process.
Make a first selection of the abstracts by sifting literature that meets the key clinical question (by
2 independent members of the guideline development group, if resources are available).
Make a second selection of literature on full text of selected papers.
Use previously defined selection criteria to select studies.
Validate the papers to be included and their classification according to clinical questions.
If resources are available, perform a double sifting of a random selection of abstracts periodically, because there is always an element of bias in selecting the evidence.
Consider, using specific methodology including formal consensus, to identify current best practice if no evidence was found to answer the clinical question.
Document the results of the selection process.
4 Critical appraisal of the evidence: Methodological quality of selected studies evaluated
Base critical appraisal on full text of selected papers.
Assess the quality of studies using a pre-defined documented appraisal criteria specific to the study.
Appraise each study preferably by 2 members of the guideline development group independently.
If a double sifting has been carried out, discuss any differences in assessment. Where differences cannot be resolved, an independent reviewer or an experienced member of the staff will arbitrate to reach an agreed quality assessment.
Validation by a third researcher experienced in literature review is highly recommended as part of the quality control process.
Document the results of the critical appraisal explicitly, detailing reasons for inclusion or exclusion.
5 Synthesising the evidence
Data extraction, tables of evidence, level of evidence
Extractrelevant data from appraised studies.
Summarize all studies identified from the systematic literature review relating to each key search question into standardized evidence tables.
Use pre-defined documented levels of evidence.
Involve all those reviewing the evidence in relation to each specific question on the assignment of a level of evidence.
Use an appropriate clinical expert input into summarising the evidence tables and in defining the levels of evidence.
Increase or lower the level of evidence depending on the quality of the study, where the study design is the major determinant for the level of evidence.
Consider using consensus to identify current best practice if no evidence was found to answer the clinical question.
Document evidence summary and levels of evidence.

Optional items are indicated in orange.

Page 1 of 2