Methods used in reviews

 

Search strategies

Access to specialised register by reviewers

The editorial office of the Cochrane Anaesthesia Review Group maintains a specialized register of controlled trials on ProCite software. Our register (of over 26,000 trials) has now been submitted to CENTRAL. We are in the process of updating it. Reports are identified by the search methods described in 'Search strategies for the identification of studies'. The register is available to members of the Group upon request.

Additional search strategies

To ensure that as many relevant studies as possible are identified (and thereby also included in the group's specialized register), authors are encouraged to conduct (and document) additional searches including: searching other relevant bibliographic databases, reviewing reference lists and personally contacting experts in the area.

Study selection

Authors and their editors decide which study designs should be included in their reviews. The editorial team can provide guidance if non-randomized studies are the only source available for a specific clinical question.

The group encourages co-authors (at least two) to apply the inclusion criteria independently of each other for each potentially relevant study. Any disagreements may be resolved by involvement of a third person or the editorial team. Studies that, in the judgment of the authors, are seriously compromised by weaknesses in design, conduct or analysis should be excluded. Details of such studies are reported in the excluded studies table within reviews.

Assessment of methodological quality

Methodological quality is assessed through concealment of allocation and blinding. Apart from this, the diversity of outcome measures makes it very difficult to create standard criteria to assess the methodological quality of studies included in reviews (overall protection against bias). Reviewers may contact the editorial office for assistance. Methodological quality is assessed for random allocation, concealment of allocation, blinding, and handling of dropouts, but without using any quality index. Authors may contact the editorial office for assistance.

Data collection

Data regarding inclusion criteria (types of studies, participants/populations, interventions and outcomes), quality criteria and results should be extracted independently by two authors, as indicated previously. Other data can be extracted by one author and checked by another. When data are missing in a published report, the authors should contact the author of the report whenever possible. Authors should refer to the Cochrane Reviewers' Handbook, or contact the editorial office for further details.

Analysis

Authors should refer to the Cochrane Handbook for Systematic Reviews of Interventions for specific guidelines, or contact the editorial office for assistance.

Reporting of reviews

Further information may be obtained by contacting the editorial office or by consulting the Cochrane Handbook for Systematic Reviews of Interventions or 'Tips for Reviewers'