On 2014 Aug 26, Elke Hausner commented:

The importance of study registries has increased markedly over the last few years and it is a highly relevant question whether such registries should regularly be used in Cochrane Reviews as an information source for additional relevant studies or data.

In the discussion section van Enst et al. describe the potential limitations of search portals: however, with their statement “... we found that many authors consult Clinicaltrials.gov which is also accessible through both the WHO ICTRP Search Portal and the MetaRegister” they seem to imply that searching directly in Clinicaltrials.gov (CT.gov) might be redundant. We disagree, as in our opinion it is inadequate to solely rely on the search results from the WHO ICTRP Search Portal (ICTRP).

In our daily work as information specialists in a German HTA agency we regularly search study registries and would like to provide some examples here of various problems encountered with ICTRP.

Limited functions: Commonly available functions of study registries, such as the use of brackets to structure searches, are lacking in ICTRP. The newly introduced function to search for synonyms is also disappointing. In comparison with CT.gov, the example of a search for “dolutegravir” shows that this function is also inadequate (synonyms found in CT.gov for dolutegravir: gsk-1349572, tivicay; synonyms found in ICTRP for dolutegravir: dolutegravir OR "DOLUTEGRAVIR").

Incomplete data: On many occasions we found that studies registered elsewhere (e.g. in CT.gov) cannot be found via ICTRP. In our opinion this calls into question the fundamental suitability of this information source. For example, the studies registered in CT.gov with the numbers NCT01006291, NCT00420212, and NCT00688688 currently cannot be found via ICTRP (status: 18 July 2014).

Error messages after complex search queries: When using ICTRP we often found that the system was unstable, thus making searches difficult or even impossible. For example, when entering more complex search queries the website often crashes. In addition, it frequently takes up to a minute for results to be shown, or after waiting an error message appears.

Summary

In our opinion, the problems described above show a structural problem of ICTRP. As long as these deficits persist, when searching for studies of drugs or non-drug interventions we recommend searching CT.gov separately. For studies of drugs, we additionally recommend searching EU-CT, the registry of the European Medicines Agency.

On 2014 Aug 26, Elke Hausner commented:

The importance of study registries has increased markedly over the last few years and it is a highly relevant question whether such registries should regularly be used in Cochrane Reviews as an information source for additional relevant studies or data.

In the discussion section van Enst et al. describe the potential limitations of search portals: however, with their statement “... we found that many authors consult Clinicaltrials.gov which is also accessible through both the WHO ICTRP Search Portal and the MetaRegister” they seem to imply that searching directly in Clinicaltrials.gov (CT.gov) might be redundant. We disagree, as in our opinion it is inadequate to solely rely on the search results from the WHO ICTRP Search Portal (ICTRP).

In our daily work as information specialists in a German HTA agency we regularly search study registries and would like to provide some examples here of various problems encountered with ICTRP.

Limited functions: Commonly available functions of study registries, such as the use of brackets to structure searches, are lacking in ICTRP. The newly introduced function to search for synonyms is also disappointing. In comparison with CT.gov, the example of a search for “dolutegravir” shows that this function is also inadequate (synonyms found in CT.gov for dolutegravir: gsk-1349572, tivicay; synonyms found in ICTRP for dolutegravir: dolutegravir OR "DOLUTEGRAVIR").

Incomplete data: On many occasions we found that studies registered elsewhere (e.g. in CT.gov) cannot be found via ICTRP. In our opinion this calls into question the fundamental suitability of this information source. For example, the studies registered in CT.gov with the numbers NCT01006291, NCT00420212, and NCT00688688 currently cannot be found via ICTRP (status: 18 July 2014).

Error messages after complex search queries: When using ICTRP we often found that the system was unstable, thus making searches difficult or even impossible. For example, when entering more complex search queries the website often crashes. In addition, it frequently takes up to a minute for results to be shown, or after waiting an error message appears.

Summary

In our opinion, the problems described above show a structural problem of ICTRP. As long as these deficits persist, when searching for studies of drugs or non-drug interventions we recommend searching CT.gov separately. For studies of drugs, we additionally recommend searching EU-CT, the registry of the European Medicines Agency.