MIE 2012 workshop

Computational Methods in Pharmacovigilance


Palazzo dei Congressi di Pisa, Italy
Room A
August 29 2012


  1. Evaluation of automated term groupings for detecting upper gastrointestinal bleeding signals for drugs.
    Julien Souvignet, Gunnar Declerck, Marie-Christine Jaulent and Cédric Bousquet
  2. Automated identification of adverse events from case reports using machine learning.
    Luca Toldo, Sanmitra Bhattacharya and Harsha Gurulingappa
  3. A structured database of Adverse Drug Reaction based on information from SPC section 4.8.
    Jim Slattery, Benedicte Cappelli, Tomas Bergvall, Anngret Mallick, Xavier Kurz
  4. Building a database of articles from PubMed for documentation of adverse drug reactions related to Anti-HIV drugs.
    Cédric Bousquet, Nicolas Garcelon, Christine Le Beller, Mathilde Wiss-Thébault, Florence Amardheil, Julien Souvignet, Anita Burgun, and Agnès Lillo-Le Louët
  5. An attempt to expedite signal detection by grouping related adverse reaction terms.
    Richard Hill, Johan Hopstadius, Magnus Lerch, and G. Niklas Norén
  6. Inferring semantic relations between pharmacovigilance terms with the NLP methods
    Marie Dupuch, Laëtitia Dupuch, Thierry Hamon and Natalia Grabar
  7. Potential bias introduced when excluding concepts from different granularity terminologies: Application to the EU funded SOS Project.
    Frantz Thiessard, Nathalie Ong, Fleur Mougin, Andrea Arfè, Edeltraut Garbe, Ron Herings, Silvia Lucchi, Federica Nicotra, Gino Picelli, Silvana Romio, René Schade, Tania Schink, Huub Straatman, Vera E. Valkhoff, Marco Villa, Annie Fourrier-Réglat, Miriam Sturkenboom


Pharmacovigilance is the activity related to the collection, analysis and prevention of adverse drug reactions induced by drugs or biologics. This activity plays the fundamental role in monitoring the patient safety during their healthcare process and beyond it. Although this activity relies on a strong surveillance tradition within the pharmacoepidemiological area, some limitations appear and generate uncomfortable cases widely covered by media (such as those related to Mediator, VIOXX etc). First of all, these are due to the ever increasing amount of the data to collect and to process, and also to the recording and storage of these data in the digital form. Hence, the computational methods may provide with useful and intelligent solutions for a more rapid and systematic management of these data.

In this workshop, we favour the submission of various and novel experiments related to the computational methods in pharmacovigilance and describing different aspects of the pharmacovigilance process, as performed by acamedic researchers, institutional or industrial people:

  • Collection of the pharmacovigilance data: traditionnally, the pharmacovigilance cases are narrative documents spontaneously submitted by medical staff, by pharmacists and sometimes even by patients themselves. The actual concern is that the collection of these cases is far from being complete. In this situation, special efforts are oriented on the computational assistance when collecting and gathering the pharmacovigilance cases. Moreover, the reporting standards and frames are also addressed by the researchers.
  • Analysis of the pharmacovigilance data : this step relies on the encoding of the cases with the dedicated terminologies and then on the semantic and statistical analysis of these cases. The concerns are related to various aspects, such as:
    • the semantic correspondence between the cases and the terms;
    • the methods for the encoding of the cases;
    • the managing of semantic ambiguity and relatedness between the terms;
    • exploitation of the SMQs and of the related data;
    • terminological and ontological representation within the dedicated terminologies;
    • computational methods for the signal detection;
    • etc.
  • Prevention of adverse drug reactions: the prevention step aims at facilitating the exchange of information through the well established national and international networks. It manages specifically the grave adverse drug reactions and the guidelines which help countries to strengthen the pharmaceutical politics and decisions.

The objective of this workshop is to gather people working on various aspects of the pharmacovigilance. Experiments performed within academic, instritutional or industrial contexts are welcome.


Ismaïl Ahmed, Equipe Biostatistique, CESP, UMRS 1018 Inserm, Université Paris Sud, France
Andrew Bate, Visiting Professor, Department of Information Systems and Computing, Computing and Mathematics, Brunel University, London, UK, Adjunct Associate Professor of Medicine, Division of Clinical Pharmacology NYU School of Medicine, New York, USA and faculty of the NYU Center for Health Informatics and Bioinformatics, USA
Kevin Cohen, University of Colorado School of Medicine and University of Colorado at Boulder, USA
Mélanie Courot, BC Cancer Agency, Vancouver, Canada
Reinhard Fescharek, Behring Inc., Germany
Thierry Hamon, LIM&BIO, Université Paris 13, France
Mikaela Keller, CNRS LIFL UMR8022, Mostrare INRIA, Université Lille 1 and 3, France
Halil Kilicoglu, National Library of Medicine, NIH, USA
Niklas Noren, Uppsala Monitoring Centre, WHO Collaborating Centre for International Drug Monitoring, Uppsala, Sweden
Patrick Ruch, University of Applied Sciences, Geneva, Switzerland
Ozlem Uzuner, University at Albany, SUNY, USA