A comparative analysis of important public clinical trial registries, and a proposal for an interim ideal one.

BACKGROUND

It is an ethical and scientific obligation to register each clinical trial, and report its results, accurately, comprehensively and on time. The WHO recognizes 17 public registries as Primary Registries, and has also introduced a set of minimal standards in the International Standards for Clinical Trial Registries (ISCTR) that primary registries need to implement. These standards are categorized into nine sections-Content, Quality and Validity, Accessibility, Unambiguous Identification, Technical Capacity, Administration and Governance, the Trial Registration Data Set (TRDS), Partner registries and Data Interchange Standards. This study compared the WHO's primary registries, and the US's ClinicalTrials.gov, to examine the implementation of ISCTR, with the aim of defining features of an interim ideal registry.

METHODS AND FINDINGS

The websites of the 18 registries were evaluated for 14 features that map to one or more of the nine sections of ISCTR, and assigned scores for their variations of these features. The assessed features include the nature of the content; the number and nature of fields to conduct a search; data download formats; the nature of the audit trail; the health condition category; the documentation available on a registry website; etc. The registries received scores for their particular variation of a given feature based on a scoring rationale devised for each individual feature analysed. Overall, the registries received between 27% and 80% of the maximum score of 94. The results from our analysis were used to define a set of features of an interim ideal registry.

CONCLUSIONS

To the best of our knowledge, this is the first study to quantify the widely divergent quality of the primary registries' compliance with the ISCTR. Even with this limited assessment, it is clear that some of the registries have much work to do, although even a few improvements would significantly improve them.