Article
What proportion of ethically approved randomised clinical trials can be found in a trial registry?
Search Medline for
Authors
Published: | February 12, 2020 |
---|
Outline
Text
Background/research question: Clinical trial registries are supposed to give a comprehensive overview of all ongoing randomised clinical trials (RCTs) which helps to avoid duplication of research and to estimate and control publication bias as well as selective reporting. As it remains unclear if all RCTs are actually registered we assessed what proportion of RCTs that received ethical approval in 2012 can be found in a clinical register.
Methods: We had access to a total of 369 study protocols of RCTs that were approved by research ethic committees in Switzerland (n=183), the United Kingdom (n=108), Germany (n=38), and Canada (n=40). In duplicate, it was searched between March 2019 and September 2019 if the RCTs were registered and published. The proportion of RCTs that were registered and published were calculated. We present sub-group analyses for the type of sponsor (i.e. industry vs. non-industry). Furthermore, we assessed if the results from these RCTs are available (in study register or as a published full text article).
Results: From the 369 RCTs a total of 337 (91%) were registered (industry 188 of 196; 96%; non-industry: 149 of 173; 86%). Considering the 369 RCTs, a total of 261 (71%) made their results available (industry 175 of 196; 89%; non-industry: 86 of 173; 50%). Results could be found for 166 (45%) of RCTs directly in the register (industry 148 of 196; 76%; nonindustry: 18 of 173; 10%). We identified for 228 (62%) of RCTs a full text publication containing primary results (industry 144 of 196; 73%; non-industry: 84 of 173; 49%). While for 108 (29%) of RCTs no results were identified (neither in a study register nor as a full text publication; industry 21 of 196; 11%; non-industry: 87 of 173; 50% Table 1). However, one has to consider that it is possible that some of the trials which were approved might be still ongoing. In fact, for 81 (22%) RCTs we either gathered some information that they might be still ongoing or did not find any evidence that they were completed. Hence, we plan to contact the Principal Investigators of those RCTs to find out about the status of those trials.
Conclusion: Even though it is an obligation to register a clinical trial, we still found that 9% of assessed RCTs were not registered. Room for improvement exists especially for non-industry sponsored trials. Furthermore, results were more often available for industry initiated RCTs. Especially on clinical trial websites hardly any results of non-industry RCTs were available.
Competing interests: The authors declare that they have no conflict of interest