Some data quality issues at ClinicalTrials.gov.

Area of Science:

• Clinical research informatics
• Data quality assessment
• Regulatory science

Background:

• Clinical trial registries aim for public accountability but face challenges with timely and accurate registration.
• Issues include non-registration, incomplete information, and delayed reporting of trial outcomes.
• This study focuses on data quality concerning Principal Investigator (PI) and Responsible Party information.

Purpose of the Study:

• To quantify data quality issues in ClinicalTrials.gov records.
• To identify specific problems with Principal Investigator (PI) and Responsible Party data.
• To provide insights for improving data accuracy in clinical trial registries.

Main Methods:

• Analysis of interventional trials registered on ClinicalTrials.gov.
• Selection criteria included trials from January 1, 2005, to December 31, 2014, involving drugs or biologics.
• Focus on trials registered with American authorities listing a real person as investigator and their role.

Main Results:

• Missing investigator names or roles occurred in 12% of analyzed trials.
• 17% of investigator names were identified as non-real or junk data.
• 3.5% of trials lacked a Responsible Party tag, and issues with PI name variations were prevalent.

Conclusions:

• Four categories of data problems were identified in ClinicalTrials.gov records.
• Quantification of these issues highlights significant data quality concerns.
• Regular audits of trial registries are crucial for identifying and rectifying data inaccuracies.