Article type
Year
Abstract
Background: At least 12 Cochrane Review Groups use study-based registers to record where multiple papers report one specific study, with either MeerKat, ProCite, Reference Manager or RefTrak software (Trials Search Co-ordinators' survey 2005).
Study-based registers:
- contain the same information as reference-based registers but link together references that report on the same study;
- aim to have all references neatly grouped into studies but in reality they are constantly changing;
- often store 'coding' that describes the study reported. Coding involves identifying study information and storing it in either:
- user defined fields of reference records (reference-based register) OR
- unique study records which link to the related references (study-based register);
- can be a 'relational' database, allowing references, studies, reviews and authors to be inter-linked.
Reference-based registers are flat files of bibliographic data organized into fields and references.
Benefits:
- Prevents authors 're-inventing the wheel'. Search results show references grouped into studies. For example, the Cochrane Schizophrenia Group found over 120 references that report a unique study. They are linked together so register searches show they have already been identified as one study. Authors do not need to look at and group those references again.
- More efficient coding (for Trials Search Co-ordinators (TSCs) who code). Study data is coded once for each study, multiple reports are scanned for any further study data. In a reference-based register the study coding is repeated for each reference.
- Authors value TSCs for producing study searches.
- Number of studies in a given area can be accurately assessed, rather than the number of references.
- Efficient removal of references once a non-randomized controlled trial/controlled clinical trial study is identified; the study and linked reference records are removed together.
- Planned and ongoing trials can be more easily tracked to full publication.
- Organization and linking of references to studies is consistent with RevMan.
Study-based registers:
- contain the same information as reference-based registers but link together references that report on the same study;
- aim to have all references neatly grouped into studies but in reality they are constantly changing;
- often store 'coding' that describes the study reported. Coding involves identifying study information and storing it in either:
- user defined fields of reference records (reference-based register) OR
- unique study records which link to the related references (study-based register);
- can be a 'relational' database, allowing references, studies, reviews and authors to be inter-linked.
Reference-based registers are flat files of bibliographic data organized into fields and references.
Benefits:
- Prevents authors 're-inventing the wheel'. Search results show references grouped into studies. For example, the Cochrane Schizophrenia Group found over 120 references that report a unique study. They are linked together so register searches show they have already been identified as one study. Authors do not need to look at and group those references again.
- More efficient coding (for Trials Search Co-ordinators (TSCs) who code). Study data is coded once for each study, multiple reports are scanned for any further study data. In a reference-based register the study coding is repeated for each reference.
- Authors value TSCs for producing study searches.
- Number of studies in a given area can be accurately assessed, rather than the number of references.
- Efficient removal of references once a non-randomized controlled trial/controlled clinical trial study is identified; the study and linked reference records are removed together.
- Planned and ongoing trials can be more easily tracked to full publication.
- Organization and linking of references to studies is consistent with RevMan.