There are two main types of Specialised Registers that have been developed by Cochrane Information Specialists: Study-based and report/reference-based.
2.7.1 Study-based Specialised Registers
During the ‘life’ of a trial, reports of that trial may be recorded in a number of places. For example:
- In Clinical Trials Registries (eg ClinicalTrials.gov)
- In conference proceedings
- As a report in a journal
- As interim and secondary analyses
- In post-hoc analyses
- In follow-up studies
Each report of a trial may contain useful and different information that could contribute to a review. As review authors should cite all reports of studies in their reviews, CISs should add all identified reports of trials to their Registers, and not just the one considered to be the main report of a trial.
See an example of how studies with multiple reports are cited in a Cochrane review (Appendix 1).
A study-based register aims to link all these multiple reports together - either prospectively by somebody extracting enough information from the reports to be able to say that they are definitely the same trial, or retrospectively by using published reviews where review authors have done that linking.
Records in a 'study-based' register are organised primarily by studies and not by bibliographic reports. Each study record contains a list of bibliographic reports referring to that study.
See an example of a record in a Specialised Register that is study-based (Appendix 2)
Study-based registers (like reference-based registers) are a ‘work in progress’. They aim to contain a set of reports that are all neatly linked into distinct studies. The registers change constantly as reports are retrospectively grouped into studies, new reports are found, and so on.
You can work towards making a register study-based using simple means. For example, you could use a study name (eg Magpie Trial 2002) or a trial registration number (NCT00123456) to link all reports of the same study together.
You could also conduct a more complex extraction of all trial information into dedicated study fields in the register. For example, you could extract:
- the number of participants
- the method of randomization
- the setting
- the intervention
- the outcomes
This detailed data extraction is not generally something a CIS has time to do, and CISs are not generally expected to do it.
The Cochrane Register of Studies (CRS) has functionality that will help CISs who wish to develop a study-based register.
2.7.2 Reference-based
A reference or report-based register aims to contain all reports of a trial - but stores them as separate references with no attempt to link together reports of the same trial. Additional information is sometimes added to this type of Specialised Register.
See an example of a reference based study record (Appendix 3).