Other limits: language, date

Author(s): 
Last revised: 
2016-03-31

Introduction

This summary is based on the Cochrane Handbook for Systematic Reviews of Interventions (1), the guidance for undertaking systematic reviews produced by the Centre for Reviews and Dissemination (CRD) (2) and the methods guide for effectiveness and comparative effectiveness reviews produced by the Agency for Healthcare Research and Quality (AHRQ) (3). The Cochrane Handbook and AHRQ methods guide are based on the best available evidence and the CRD guidance is recommended as a source of good practice by agencies such as the National Institute for Health and Clinical Excellence (NICE).

The revised and updated searching chapter of the Cochrane Handbook is in preparation. To avoid duplication of effort with the development of the Cochrane Handbook, appraisals have not been prepared for studies in this chapter. Once the revised Cochrane Handbook is available it will be used to update this chapter.

Limiting a search strategy by date

Limiting a search strategy by date may reduce the number of records retrieved for screening, but date limits should only be applied if there is a robust rationale for doing so.  For example, if a healthcare intervention was introduced at a certain date, limiting the search strategy to only retrieve studies reported from this date would be appropriate (3,4). 

Applying date limits to the search strategy can also be an option if an existing search is being updated (2,4). When conducting an update search, searchers should be cautious about how date limits are applied (2,4). If attempting to limit by date, an appropriate field (or fields) such as update date rather than publication date should be used.  Limiting searches by publication date risks missing relevant records (2). For databases where there is no update field, running the search without date limits is advised, using reference managing software to de-duplicate the returned records against the original search results. (2)

Limiting a search strategy by language

Including non-English language studies in a review can add to the resources necessary to complete the review (for example, time needed to identify results, translation costs, time needed to data-extract) (4).  By only including English language studies however, language bias is potentially introduced (2,3,5).  For topic areas where research from, or relating to, non-English speaking regions is of increased significance, the issue of limiting by language may be a particular concern. Shenderovich et al studied methodological issues in systematic reviews which aimed to include evidence from low- and middle-income countries, using the example of a review of risk factors for child conduct problems and youth violence (6).  The authors reported that 15 % of the eligible studies were in a language other than English and therefore would not have been retrieved if English language search limits were applied.  The impact of omitting the non-English studies on the conclusions of the review was not investigated.

Current guidance recommends that search strategies should not be restricted by language (2,3,4).  This is advised even if translation is not feasible (2, 3).  Reviewers may exclude non-English language studies from the review, but make a list of potentially relevant studies which were excluded on the basis of language.  This can help inform an assessment of the potential risk for language bias (2,3).

Reference list