Table 2.

Overview of results

Method	Includes	What is the method used for	What the evidence says	Implications of evidence	Claimed advantages	Claimed disadvantages	Resource requirements
Contacting study authors	6 studies	Identify: unpublished, or on-going studies, missing or incomplete data, completed but unpublished studies Expert in field to review includes	Contact original investigators through study report contact details—mainly e-mail/telephone	E-mail considered effective with better responses from institutional addresses	Additional studies identified; additional study data provided	No guarantee of additional or all relevant information identified Challenging and time consuming Less successful for older studies	Additional resources needed (may need up to 3 contact attempts with authors)
Citation chasing	9 studies	Identify: further studies, clusters or networks or studies	Backward and forward citation chasing using 3 electronic citation databases	Effectiveness of electronic citation methods unclear and suggest using all 3 databases	Not limited by keywords or indexing as bibliographic database searching is	Reliant on the currency, accuracy and completeness of the underlying citation network	Citation chasing of 46 studies = 79 h or 40 studies = 5 days
Handsearching	12 studies	Identify: studies or publications not routinely indexed in, or identified by, searches of bibliographic databases, including recently published studies	Manual examination of the contents of topic relevant journals, conference proceedings and abstracts	Use experts to develop list of journals to handsearch	Unique study identification, increased sensitivity; identifying studies missed or not indexed in databases	Studies still missed by handsearching; time and access to resources; low precision	Range between 6 min and 1 h per journal
Searching trial registers	3 studies	Identify: unpublished, recently completed or on-going trialsFind adaptations to trial protocols reported study outcomes	Comprehensive list of registries to search	Should be completed as complementary and not in isolation	Unique study identification	Search interfaces lag behind major databases	None reported
Web searching	5 studies	Identify: studies not indexed in bibliographic databases. Retrieving grey literature, study protocols and on-going studies	Relevant websites and using search engines	Use advanced search functions where possible	Unique study identification, hints to on-going or recently completed studies	Difficulties in transparent search, quality and quantity of searches returned	429 results in 21 h; google searching 7.9 h; targeted web searching 9-11 h