Semantic Queries Expedite MedDRA Terms Selection Thanks to a Dedicated User Interface: A Pilot Study on Five Medical...

Area of Science:

• Pharmacovigilance and Medical Terminology
• Health Informatics
• Natural Language Processing in Healthcare

Background:

• Traditional MedDRA (Medical Dictionary for Regulatory Activities) searches rely on hierarchical or string-based methods.
• These conventional approaches can be time-consuming and may not fully capture the nuances of medical conditions.
• There is a need for more efficient and precise methods for MedDRA terminology exploration.

Purpose of the Study:

• To compare user performance and term selection accuracy in MedDRA using a novel semantic query interface versus classical search methods.
• To evaluate the effectiveness of the OntoADR Query Tools (OQT) in improving MedDRA search capabilities for pharmacovigilance.

Main Methods:

• Implementation of OntoADR Query Tools (OQT), a web interface utilizing OntoADR, which describes MedDRA terms with SNOMED CT concepts and semantic relations.
• Comparison of user time, precision, and recall when using OQT versus the MedDRA web-based browser (MWB) for five medical conditions.
• Assessment of user experience using the System Usability Scale (SUS).

Main Results:

• OQT significantly reduced search time by 35% (average 4 min 30 s vs. 7 min for MWB, p < 0.001).
• Precision increased by 27% (p = 0.01) and recall by 34% (p = 0.02) with OQT.
• User performance (correct terms per minute) was over three times better with OQT, which received a good SUS score of 62.19.

Conclusions:

• The OntoADR Query Tools (OQT) demonstrate the feasibility of using terminological reasoning for expedited and improved MedDRA term selection in pharmacovigilance.
• This semantic query approach shows significant potential for enhancing search capabilities for end-users.
• Further evaluation with more users and conditions is recommended to confirm OQT's broader applicability and utility across different user profiles.