Article type
Year
Abstract
Background: For systematic reviews (SR) searching clinical trials registries is now recommended and becoming commonplace. Although this is a step in the right direction for ensuring as many sources of information are found for the conclusions drawn by the SR, there still seems to be little done with this information except see if there are any trials currently underway that may be completed soon. We have devised a step by step method that will enable people to not only find all relevant registered trials but to also find any publication information regarding those trials.
Objectives: To test the feasibility and information gain from our step by step protocol to locate any published information regarding registered clinical trials.
Methods: We searched clinical trials registries to find trials on two topics. These clinical trials were then run through a five-step process to identify any publications deriving from them. This covered not only trials published in standard databases such as PubMed and Embase but also looking for publications in other locations, such as personal websites of authors or theses of Research Higher Degree students.
Results: We piloted the process looking for all registered homeopathy trials. We found six trials in PubMed, 18 potentially relevant results in Embase, and an additional 11 publications outside of the medical databases. We then tested the method in a full review of the adverse effects of macrolides. For this review we found 54 potential publications, 48 of which were in PubMed. Of these, 17 were not found in the original search and eight of them made it through the title/abstract screening process. When the same technique was used in Embase, 112 records were returned, suggesting identification of a high number of potentially relevant conference proceedings.
Conclusions: The method appears to enhance the quality of SRs by ensuring any trials are easily found and included in the final analysis as well as creating an initial results set that can be used to create and validate a search strategy. This process is a fast and effective way to enhance the results of SRs of clinical trials.
Objectives: To test the feasibility and information gain from our step by step protocol to locate any published information regarding registered clinical trials.
Methods: We searched clinical trials registries to find trials on two topics. These clinical trials were then run through a five-step process to identify any publications deriving from them. This covered not only trials published in standard databases such as PubMed and Embase but also looking for publications in other locations, such as personal websites of authors or theses of Research Higher Degree students.
Results: We piloted the process looking for all registered homeopathy trials. We found six trials in PubMed, 18 potentially relevant results in Embase, and an additional 11 publications outside of the medical databases. We then tested the method in a full review of the adverse effects of macrolides. For this review we found 54 potential publications, 48 of which were in PubMed. Of these, 17 were not found in the original search and eight of them made it through the title/abstract screening process. When the same technique was used in Embase, 112 records were returned, suggesting identification of a high number of potentially relevant conference proceedings.
Conclusions: The method appears to enhance the quality of SRs by ensuring any trials are easily found and included in the final analysis as well as creating an initial results set that can be used to create and validate a search strategy. This process is a fast and effective way to enhance the results of SRs of clinical trials.