Search

Background Adverse drug reactions (ADRs) are statistically characterized within randomized clinical trials and postmarketing pharmacovigilance, but their molecular mechanism remains unknown in most cases. This is true even for hepatic or skin toxicities, which are classically monitored during drug design. Aside from clinical trials, many elements of knowledge about drug ingredients are available in open-access knowledge graphs, such as their properties, interactions, or involvements in pathways. In addition, drug classifications that label drugs as either causative or not for several ADRs, have been established. Methods We propose in this paper to mine knowledge graphs for identifying biomolecular features that may enable automatically reproducing expert classifications that distinguish drugs causative or not for a given type of ADR. In an Explainable AI perspective, we explore simple classification techniques such as Decision Trees and Classification Rules because they provide human-readable models, which explain the classification itself, but may also provide elements of explanation for molecular mechanisms behind ADRs. In summary, (1) we mine a knowledge graph for features; (2) we train classifiers at distinguishing, on the basis of extracted features, drugs associated or not with two commonly monitored ADRs: drug-induced liver injuries (DILI) and severe cutaneous adverse reactions (SCAR); (3) we isolate features that are both efficient in reproducing expert classifications and interpretable by experts (i.e., Gene Ontology terms, drug targets, or pathway names); and (4) we manually evaluate in a mini-study how they may be explanatory. Results Extracted features reproduce with a good fidelity classifications of drugs causative or not for DILI and SCAR (Accuracy = 0.74 and 0.81, respectively). Experts fully agreed that 73% and 38% of the most discriminative features are possibly explanatory for DILI and SCAR, respectively; and partially agreed (2/3) for 90% and 77% of them. Conclusion Knowledge graphs provide sufficiently diverse features to enable simple and explainable models to distinguish between drugs that are causative or not for ADRs. In addition to explaining classifications, most discriminative features appear to be good candidates for investigating ADR mechanisms further. Supplementary Information The online version contains supplementary material available at 10.1186/s12911-021-01518-6.

International audience; Our objective was to improve the accuracy of bacteria and resistance coding in a hospital case mix database. Data sources consisted of 50,074 files on bacteriological susceptibility tests transmitted with the HPRIM protocol from laboratory management system to electronic health record of the University hospital of Saint Etienne in July 2017. An algorithm was implemented to detect susceptibility tests containing information corresponding to codes whose addition in the case mix database was susceptible to increase the severity level of a diagnosis related group. Among 132 hospital stays fulfilling the conditions, 27 were lacking bacteria and/or resistance codes, and the tariff was increased for 9 stays, with earnings of €54,612. Analyzing Antimicrobial susceptibility tests helps to improve clinical coding and optimize the financial gain.

International audience; Since December 2019 and the first reported cases of COVID-19 in Wuhan, China, there have been 199,466,211 confirmed cases of COVID-19 in the World. The WHO defined vaccination hesitancy as one of the top ten threats to global health in 2019. Our objective was thus to identify topics and trends about COVID-19 vaccines from French web forums to understand the perception of the French population on these vaccines before the vaccination campaign started. We performed a topic model analysis on 485 web forums' posts. 10 topics were found. We reviewed 120 posts from 6 of these 10 topics. One topic was about vaccine hesitancy, refusal, and mistrust, and two topics were related to what the users think about the government, the political and economic choices made towards this epidemic.

Resume Objectif Evaluer l’informativite et la qualite des descriptions d’effets indesirables susceptibles d’etre lies au baclofene, dans les forums de discussion francais. Methodes Nous avons evalue la qualite des cas de pharmacovigilance potentiels associes au baclofene dans 22 forums de discussion. Nous avons ensuite compare ces donnees aux cas de pharmacovigilance issus de la base nationale de pharmacovigilance (BNPV), en termes d’informativite concernant le patient, le traitement et l’effet indesirable ainsi que la gravite et le caractere inattendu des effets. Resultats Nous avons identifie 782 cas de pharmacovigilance potentiels parmi 2621 messages issus de forums francais. L’informativite des messages des forums etait significativement inferieure a celle des cas issus de la BNPV, pour l’ensemble des donnees concernant le patient (3 %/6 % vs 88 % pour l’âge/la classe d’âge et 46 % vs 99 % pour le sexe, respectivement), pour la duree de traitement (9 % vs 24 %) et pour l’evolution de l’effet indesirable (1 % vs 64 %, respectivement). Mais l’indication du traitement et sa posologie etaient plus frequemment renseignees dans les forums que dans la BNPV (67 % vs 24 % et 27 % vs 9 %, respectivement). Les cas d’effets indesirables issus des forums sont tres majoritairement non graves, contrairement a ceux issus de la BNPV (38 % vs 0,7 %). La proportion de cas rapportant des effets indesirables inattendus est significativement plus importante dans les forums que dans la BNPV (43,8 % vs 11,6 %). Conclusion L’indication et la posologie etaient plus souvent renseignees dans les messages que dans la BNPV, ce qui fait des forums une ressource interessante pour etudier les conditions d’utilisation du baclofene. Bien que les forums comportent plus d’effets inattendus, leur moindre informativite rend difficile l’evaluation de l’imputabilite. Nous estimons que les forums et la BNPV sont deux sources de donnees complementaires pour la securite d’emploi du baclofene.

International audience; Introduction and Background: Since June 2011, the European legislation allows patients and patients' organizations to report adverse drug reactions (ADRs). In France, in 2017, more than 32,000 ADRs were reported by patients representing 42% of all ADRs collected yearly and an unexpected increase in patient reports (5.8% in 2016). Most of these cases involved the new Levothyrox® formulation (available since March 2017), leading to the "French Levothyrox® Crisis" [1]. To evaluate whether such crises can be predicted by analyzing social media data, the PHARES multidisciplinary consortium carried out the Levothyrox® case study that is described in this abstract. The PHARES project was funded from 2017 to 2019 by the French National Agency for the Safety of Medicines and Health Products (ANSM [2]) with the objective to implement a computerized tool for social media mining. Objectives: Quantitative and qualitative analysis of Levothyrox® mentions in a French web site dedicated to thyroid pathology to investigate whether social media may be useful to predict ADR-related crises. Methods: All posts published in the forums of the "Vivre sans thyroïde" website were collected and annotated automatically using machine learning and natural language processing methods to identify mentions of drugs and potential Adverse Events (AE) [3, 4]. Only posts containing at least one mention of Levothyrox® or levothyroxine were considered as our analysis collection. From this data sample, we first counted the number of posts per each verbatim detected by the annotation algorithm to identify those AEs most mentioned with Levothyrox®. The 100 most mentioned verbatims were then manually reviewed and classified into categories by a pharmacovigilance expert (cf. Table 1). This manual step generated an association list (verbatim -> Category) that we used as a reference to generate the following statistics over the whole collection: the number of posts per category, the number of posts per AE in each category, and the evolution over time of the number of posts per AE and per category and a time-series analysis using change-point analysis method (CPA) [5].Results: As for July 2018, the studied website contained over 900 thousand posts published since 2001. The evolution of the number of AE mentions associated to Levothyrox® from June 2005 to July 2018 shows no particular rise was present before the first declaration to the pharmacovigilance network. The qualitative analysis of the categories selected for this study resulted in: 25% of General AE, 25% of Levothyrox® indication, 18% of neurological and psychiatric effects, 7% of seriousness description ("emergency", "consultation", "crisis") and 5% of cardiovascular effects. The main AEs associated to general effects are "tiredness", "weight gain" and "pain". The CPA method shows a change point one month before the beginning of the crisis.Conclusion and perspectives: To our knowledge, this study is the first retrospective analysis of social media data following a drug health crisis. It concerns a new formulation of a drug used by more than 3 million people in France, leading to thousands of patients' complaints about ADRs. Most of the collected AEs were expected, some of them may be difficult to classify and a human evaluation is still required. The detected change point corresponds to a low variation and further investigations need to be performed.

International audience; Drug Safety (DS) is a domain with significant public health and social impact. Knowledge Engineering (KE) is the Computer Science discipline elaborating on methods and tools for developing "knowledge-intensive" systems, depending on a conceptual "knowledge" schema and some kind of "reasoning" process. The present systematic and mapping review aims to investigate KE-based approaches employed for DS and highlight the introduced added value as well as trends and possible gaps in the domain. Journal articles published between 2006 and 2017 were retrieved from PubMed/MEDLINE and Web of Science ® (873 in total) and filtered based on a comprehensive set of inclusion/exclusion criteria. The 80 finally selected articles were reviewed on full-text, while the mapping process relied on a set of concrete criteria (concerning specific KE and DS core activities, special DS topics, employed data sources, reference ontologies/terminologies, and computational methods, etc.). The analysis results are publicly available as online interactive analytics graphs. The review clearly depicted increased use of KE approaches for DS. The collected data illustrate the use of KE for various DS aspects, such as Adverse Drug Event (ADE) information collection, detection, and assessment. Moreover, the quantified analysis of using KE for the respective DS core activities highlighted room for intensifying research on KE for ADE monitoring, prevention and reporting. Finally, the assessed use of the various data sources for DS special topics demonstrated extensive use of dominant data sources for DS surveillance, i.e., Spontaneous Reporting Systems, but also increasing interest in the use of emerging data sources, e.g., observational healthcare databases, biochemical/genetic databases, and social media. Various exemplar applications were identified with promising results, e.g., improvement in Adverse Drug Reaction (ADR) prediction, detection of drug interactions, and novel ADE profiles related with specific mechanisms of action, etc. Nevertheless, since the reviewed studies mostly concerned proof-of-concept implementations, more intense research is required to increase the maturity level that is necessary for KE approaches to reach routine DS practice. In conclusion, we argue that efficiently addressing DS data analytics and management challenges requires the introduction of high-throughput KE-based methods for effective knowledge discovery and management, resulting ultimately, in the establishment of a continuous learning DS system.

The aim of this study is to analyze drug mentions in web forums to evaluate the utility of this data source for drug post-marketing studies. We automatically annotated over 60 million posts extracted from 21 French web forums. Drug mentions detected in this corpus were matched to drug names in a French drug database (Theriaque®). Our analysis showed that a high proportion of the most frequent drug mentions in the selected web forums correspond to drugs that are usually prescribed to young women, such as combined oral contraceptives. The most mentioned drugs in our corpus correlated weakly to the most prescribed drugs in France but seemed to be influenced by events widely reported in traditional media. In this article, we conclude that web forums have high potential for post-marketing drug-related studies, such as pharmacovigilance, and observation of drug utilization. However, the bias related to forum selection and the corresponding population representativeness should always be taken into account.

International audience; Background: The Food and Drug Administration (FDA) in the United States and the European Medicines Agency (EMA) have recognized social media as a new data source to strengthen their activities regarding drug safety. Objective: Our objective in the ADR-PRISM project was to provide text mining and visualization tools to explore a corpus of posts extracted from social media. We evaluated this approach on a corpus of 21 million posts from five patient forums, and conducted a qualitative analysis of the data available on methylphenidate in this corpus. Methods: We applied text mining methods based on named entity recognition and relation extraction in the corpus, followed by signal detection using proportional reporting ratio (PRR). We also used topic modeling based on the Correlated Topic Model to obtain the list of the matics in the corpus and classify the messages based on their topics. Results: We automatically identified 3443 posts about methylphenidate published between 2007 and 2016, among which 61 adverse drug reactions (ADR) were automatically detected. Two pharmacovigilance experts evaluated manually the quality of automatic identification, and a f-measure of 0.57 was reached. Patient's reports were mainly neuro-psychiatric effects. Applying PRR, 67% of the ADRs were signals, including most of the neuro-psychiatric symptoms but also palpitations. Topic modeling showed that the most represented topics were related to Childhood and Treatment initiation, but also Side effects. Cases of misuse were also identified in this corpus, including recreational use and abuse. Conclusion: Named entity recognition combined with signal detection and topic modeling have demonstrated their complementarity in mining social media data. An in-depth analysis focused on methylphenidate showed that this approach was able to detect potential signals and to provide better understanding of patients' behaviors regarding drugs, including misuse.

International audience; Objective: To propose a method to build customized sets of MedDRA terms for the description of a medical condition. We illustrate this method with upper gastrointestinal bleedings (UGIB).Research design and methods: We created a broad list of MedDRA terms related to UGIB and defined a gold standard with the help of experts. MedDRA terms were formally described in a semantic resource named OntoADR. We report the use of two semantic queries that automatically select candidate terms for UGIB. Query 1 is a combination of two SNOMED CT concepts describing both morphology ‘Hemorrhage’ and finding site ‘Upper digestive tract structure’. Query 2 complements Query 1 by taking into account MedDRA terms associated to SNOMED CT concepts describing clinical manifestations ‘Melena’ or ‘Hematemesis’.Results: We compared terms in queries and our gold standard achieving a recall of 71.0% and a precision of 81.4% for query 1 (F1 score 0.76); and a recall of 96.7% and a precision of 77.0% for query 2 (F1 score 0.86).Conclusions: Our results demonstrate the feasibility of applying knowledge engineering techniques for building customized sets of MedDRA terms. Additional work is necessary to improve precision and recall, and confirm the interest of the proposed strategy.

International audience; Literature, specifically MEDLINE, is among the main sources of information used to detect whether a drug may be responsible for Adverse Drug Reactions cases. The aim of our work is to automate the search of publications that correspond to a given Adverse Drug Reactions case: (i) by defining a general pattern for the queries used to search MEDLINE and (ii) by determining the threshold number of publications capable to confirm or infirm the Adverse Drug Reaction. We applied our algorithm to a set of 620 cases from a French pharmacovigilance database. We obtained a precision of 93%, recall 70%. We determined a threshold of 3 publications to confirm an Adverse Drug Reaction case.

National audience; Ce travail est motivé par la problématique de recherche d'informations spécifiques au domaine de la pharmacovigilance. Un signal de pharmacovigilance est la relation détectée statistiquement entre un médicament et un groupe d'effets indésirables exprimant des conditions cliniques similaires. Le regroupement pertinent des termes cliniques normalisés rend un signal plus spécifique et sa détection plus sensible. Notre objectif est de trouver les regroupements pertinents des effets indésirables en employant une méthode d'apprentissage automatique sur l'UMLS - un grand metathèsaurus de la médecine. A partir des relations pertinentes regroupant des termes cliniques normalisés et non définies explicitement dans MedDRA, le thesaurus d'origine, nous avons extrait des règles permettant de prédire ces relations à l'intérieur d'UMLS. En appliquant ces règles nous avons réussi à prédire les relations pour deux couples de concepts. Ces résultats démontrent l'intérêt de l'utilisation des méthodes d'apprentissage automatique sur l'UMLS pour l'extraction de relations non définies explicitement dans un système terminologique.