Exploitation de la distance sémantique pour la création de groupements de termes en pharmacovigilance

Pharmacovigilance is the activity related to the collection, analysis and prevention of adverse drug reactions (ADRs) induced by drugs. Beside other methods, statistical methods are used to detect new ADRs in the framework of signal detection. Groupings of terms containing similar ADRs allow to increase the signal intensity and to improve the detection of new ADRs. SMQs have become reference groupings in the field of pharmacovigilance. They are built on the MedDRA terminology and thanks to the study of scientific literature. Even if SMQs are built manually by experts, they still show some shortcomings: they tend to be over-inclusive and thus become too sensitive, although they can also miss several relevant terms. Moreover, the spectrum of available SMQs is limited. The objective of this work is to propose an automated method for a flexible creation of groupings of terms. This method is based on exploitation of the semantic distance between MedDRA terms. In a first experience, we used ARD terms alone and obtained results with a high precision (mean 74% within the interval [49; 91]). In a second experience, we used ADR terms and their formal definitions and this worsened the results because the semantic information within definitions may be missing. We assume that more exhaustive definitions will have a positive effect on results.

[1]  Marie-Christine Jaulent,et al.  A case report: using SNOMED CT for grouping Adverse Drug Reactions Terms , 2008, BMC Medical Informatics Decis. Mak..

[2]  Reinhard Fescharek,et al.  Medical Dictionary for Regulatory Activities (MedDRA): Data retrieval and presentation , 2004 .

[3]  Patrice Degoulet,et al.  Implementation of automated signal generation in pharmacovigilance using a knowledge-based approach , 2005, Int. J. Medical Informatics.

[4]  Jürgen Kübler,et al.  Medical Dictionary for Regulatory Activities (MedDRA) , 2012, International Journal of Pharmaceutical Medicine.

[5]  Patrice Degoulet,et al.  Clustering WHO-ART Terms Using Semantic Distance and Machine Learning Algorithms , 2006, AMIA.

[6]  Kent A. Spackman,et al.  SNOMED clinical terms: overview of the development process and project status , 2001, AMIA.

[7]  A. Bate,et al.  A Bayesian neural network method for adverse drug reaction signal generation , 1998, European Journal of Clinical Pharmacology.

[8]  E. Brown,et al.  The Medical Dictionary for Regulatory Activities (MedDRA) , 1999, Drug safety.

[9]  Kent A. Spackman,et al.  Compositional concept representation using SNOMED: towards further convergence of clinical terminologies , 1998, AMIA.

[10]  A Bate,et al.  Decision support methods for the detection of adverse events in post-marketing data. , 2009, Drug discovery today.

[11]  M. Lindquist,et al.  Signal Selection and Follow-Up in Pharmacovigilance , 2002, Drug safety.

[12]  Michael Pignone,et al.  Bmc Medical Informatics and Decision Making the Effect of Offering Different Numbers of Colorectal Cancer Screening Test Options in a Decision Aid: a Pilot Randomized Trial , 2008 .

[13]  Roy Rada,et al.  Development and application of a metric on semantic nets , 1989, IEEE Trans. Syst. Man Cybern..

[14]  David Madigan,et al.  Influence of the MedDRA® hierarchy on pharmacovigilance data mining results , 2009, Int. J. Medical Informatics.

[15]  A. Pariente,et al.  Data mining on electronic health record databases for signal detection in pharmacovigilance: which events to monitor? , 2009, Pharmacoepidemiology and drug safety.