Marie-Christine Jaulent Iulian Alecu SPIM- UMRS872,EQ20 - France - - PowerPoint PPT Presentation
Marie-Christine Jaulent Iulian Alecu SPIM- UMRS872,EQ20 - France MIE 2009 Sarajevo - 31 aout 2009 Agenda History of the project Pharmacovigilance and Signal detection The ontological resource: OntoEIM The web service PharmARTS
MIE 2009 – Sarajevo - 31 aout 2009
History of the project
Pharmacovigilance and Signal detection The ontological resource: OntoEIM The web service PharmARTS
Evaluation of the ontological resource
Context Design of the study Methodology of evaluation Current results Evolution and maintenance of the ontology
Pharmacovigilance
Set of procedures for the identification, evaluation and prevention of
ADR (adverse drug reaction) risk
ADR are coded in databases
Signal detection
Signal ↔ possible cause {Drug → ADR} (OMS) Data mining - automatic analysis need prior automatic grouping
Limits of terminology’ structure for grouping [Bousquet et a. 2005]
Lack of polyhierarchy MSSO : manual answer to the problem, SMQ
synonyms (Cullen's sign, Hereditary pancreatitis,Ischaemic pancreatitis, Blood amylase increased, Pancreatic enzymes increased, Urine amylase abnormal, Blood bilirubin increased, Gastrointestinal pain, Nausea, …)
90000 concepts (45000 SNOMED_CT
WHO-ART MedDRA Snomed_CT
* Alecu I, Bousquet C, Jaulent MC. A case report: using SNOMED CT for grouping Adverse Drug Reactions Terms. BMC Med Inform Decis Mak. 2008 Oct 27;8 Suppl 1:S4.
Browse the resource for coding purpose,
Build queries to group terms
Retrieve cases in a pharmacovigilance database
Context: Define « on the fly » a medical condition by
querying the ontology and grouping concepts
Question: Is the ontology appropriate to propose relevant
set of concepts for a given medical condition?
Evaluation of ontologies [Portzel 2004; Rogers 2006; Cornet 2008]
internal and external
the relevance of the vocabulary describing the concepts the relevance of the “is-a” hierarchy the relevance of the semantic relations
OntoEIM is evaluated according to its purpose
Rogers 2006:
One significant problem for ontology quality assurance is the
lack of a gold standard against which to determine the correctness of the ontology and its suitability for the given purpose
Definition of a precise gold standard: SMQ
SMQ
return
OntoEIM
Modelisation for each medical condition
Group of individuals: Set of possible terms (MedDRA) Definition of the condition: term present in SMQ (SMQ
+)
Definition of the test (a query). We say that the test is:
SMQ + SMQ - Q+ a b Q- c d
a b c
Correspondance
Example
Composition
Example : 1) SMQ = Asthma/bronchospasm 2) C1 = Asthma 3) C2 = bronchospasm
C2 C1
Enlargment
Example : 1) SMQ = CONVULSIONS 2) C = CONVULSIONS 3) C’ = Seizures_incl_subtypes
C C’ C
For each candidate concept, we calculate
If there are several candidate concepts,
C1, C2 Q1=C1; Q2=C2; Q3=C1+C2
Final query
Q = Q such that max (a) and min (b+c) < λ λ is an heuristic
a b c
Number of terms in the SMQ Sensitivity Mean Range Mean Range 23,5 [7 ; 47] 0,82 [0,45 ; 1]
List of 24 SMQs
pancreatitis; Acute renal failure ; Agranulocytosis ; Anaphylactic reaction; Asthma/bronchospasm ; Cardiac failure; Haemorrhagic cerebrovascular condition; Convulsions; Guillain-Barre syndrome ; Dyslipidaemia ; Hostility/aggression ; Interstitial lung disease; Lack of efficacy/effect; Neuroleptic malignant syndrome; Pseudomembranous colitis ; Peripheral neuropathy ; Pulmonary hypertension ; Retroperitoneal fibrosis ; Rhabdomyolysis/ myopathy ; Severe cutaneous adverse reactions; Torsade de pointes/QT prolongation ; hyperglycemie/ diabete mellitus
b≠0. The term is missing in the SMQ.
SMQ « angioedema », term
b≠0. There is a taxonomic error in the
SMQ “acute pancreatitis”, term
c≠0. Formal concept definitions are not
c≠0. Semantic relations are missing in
the relation “evoke” between an exam result
Evaluation of an existing resource
Specific purpose : prior grouping of terms to improve
the performances of signal detection algorithms in pharmacivigilance
Evaluation by comparison with the existing SMQs
Implementation of the approach Reusability
Ontology evolution : The method does not support the
development of the ontology but allows identifying how it has to evolve by finding the corrections that have to be made (missing definitions or missing relations, wrong definitions or wrong relations)
Ontology maintenance: measure if the ontology
remains adequate when there is a new context (construction of new SMQs by the MSSO)
Continue the evaluation for the other
Collect manually the classification and
Design tools and interfaces to assist the
Define medical conditions « on the