Inconsistent quality of reporting of searching clinical trials registries in Cochrane systematic reviews and protocols

Article type
Authors
Ko H1, Tai F1, Ghersi D2, Askie L1
1Australian New Zealand Clinical Trials Registry, Australia
2Research Policy and Cooperation, World Health Organisation, Switzerland
Abstract
Background: Searching of clinical trials registries (CTRs) for registered clinical trials has not been mandatory for systematic reviews (SRs) and protocols published in the Cochrane Database of Systematic Reviews (CDSR), however some Review Groups are encouraging this practice. In the 2010 Cochrane Colloquium, Ghersi et al and Hooft et al reported there was low searching of CTRs.

Objectives: To assess the quality and consistency of reporting of searching International Clinical Trials Registry Platform (ICTRP) and/or primary CTRs, as well as other methods to identify ongoing clinical trials, in SRs and protocols.

Methods: SRs and protocols that used the keywords 'trial registry’, 'trial registries', 'trial register’, or 'trial registers' were retrieved on 29 March 2011. Two authors independently evaluated methods sections' reporting of searching (1) the ICTRP for ongoing clinical trials, (2) primary CTRs listed in the WHO ICTRP, and/or clinicaltrials.gov, and whether (3) other sources were used to identify ongoing clinical trials. Disagreements were resolved by consensus.

Results: 414 SRs and 228 protocols were retrieved. Overall, 247 (60%) SRs and 167 (71%) protocols reported searching ICTRP and/or CTRs. Table 1 shows the breakdown of CTR searching methods reported. SRs with their latest search performed from 2008 show >70% of SRs reporting searching CTRs, with most preferring to search a variety of CTRs or searching both ICTRP and CTRs. The amount and names of CTRs used were variable. SRs that have not updated their search since 2007 have minimal or no reporting of searching CTRs. Searching other sources such as experts or pharmaceutical companies about ongoing trials was very common.

Conclusions: The quality and consistency of reporting of searching CTRs for ongoing clinical trials is highly variable. More explicit and consistent search protocols across all SRs and protocols are needed, as well as updating of search strategies of older SRs.