Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Naranjo Question Answering using End-to-End Multi-task Learning Model.

Bhanu Pratap Singh Rawat1, Fei Li2, Hong Yu2

  • 1UMass Amherst, Amherst, USA.

KDD : Proceedings. International Conference on Knowledge Discovery & Data Mining
|December 5, 2019
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Parameter Efficient Transfer Learning for Suicide Attempt and Ideation Detection.

Proceedings of the ... International Workshop on Health Text Mining and Information Analysis·2026
Same author

Intentional Self-Harm Among US Veterans With Traumatic Brain Injury or Posttraumatic Stress Disorder: Retrospective Cohort Study From 2008 to 2017.

JMIR public health and surveillance·2023
Same author

An Investigation of the Representation of Social Determinants of Health in the UMLS.

AMIA ... Annual Symposium proceedings. AMIA Symposium·2023
Same author

Knowledge Injected Prompt Based Fine-tuning for Multi-label Few-shot ICD Coding.

Proceedings of the Conference on Empirical Methods in Natural Language Processing. Conference on Empirical Methods in Natural Language Processing·2023
Same author

ScAN: Suicide Attempt and Ideation Events Dataset.

Proceedings of the conference. Association for Computational Linguistics. North American Chapter. Meeting·2023
Same author

Relation Classification for Bleeding Events From Electronic Health Records Using Deep Learning Systems: An Empirical Study.

JMIR medical informatics·2021
Same journal

A Survey on Unifying Large Language Models and Knowledge Graphs for Biomedicine and Healthcare.

KDD : proceedings. International Conference on Knowledge Discovery & Data Mining·2026
Same journal

Identifying Combinatorial Regulatory Genes for Cell Fate Decision via Reparameterizable Subset Explanations.

KDD : proceedings. International Conference on Knowledge Discovery & Data Mining·2025
Same journal

MentalChat16K: A Benchmark Dataset for Conversational Mental Health Assistance.

KDD : proceedings. International Conference on Knowledge Discovery & Data Mining·2025
Same journal

Graph ODEs and Beyond: A Comprehensive Survey on Integrating Differential Equations with Graph Neural Networks.

KDD : proceedings. International Conference on Knowledge Discovery & Data Mining·2025
Same journal

SatHealth: A Multimodal Public Health Dataset with Satellite-based Environmental Factors.

KDD : proceedings. International Conference on Knowledge Discovery & Data Mining·2025
Same journal

Deep Multi-Output Forecasting: Learning to Accurately Predict Blood Glucose Trajectories.

KDD : proceedings. International Conference on Knowledge Discovery & Data Mining·2025
See all related articles

This study introduces a novel method to automatically determine if a drug causes an adverse drug reaction (ADR) using electronic health records. The new model significantly improves upon existing approaches for ADR causality assessment.

Area of Science:

  • Clinical Informatics
  • Pharmacovigilance
  • Natural Language Processing

Background:

  • Assessing the causal relationship between medications and adverse drug reactions (ADRs) is crucial in clinical practice.
  • Clinical judgment studies are traditionally used for ADR causality assessment.
  • Electronic Health Records (EHRs) contain valuable data for understanding drug-induced adverse events.

Purpose of the Study:

  • To develop and evaluate an automated method for inferring drug-induced ADR causality from EHRs.
  • To adapt the Naranjo questionnaire, a validated clinical tool, for automated question answering.
  • To establish a computational approach for supporting pharmacovigilance and clinical decision-making.

Main Methods:

  • Development of a joint model employing multi-task learning to predict answers to a subset of the Naranjo questionnaire.
Keywords:
Attention based networkLSTMMulti-task learningNaranjo questionnaireQuestion answeringRNN

Related Experiment Videos

  • Utilizing physician annotations from EHRs as the gold standard for model training and evaluation.
  • Comparison of the proposed joint model against a baseline pipeline model.

    • Main Results:

    • The proposed joint model significantly outperformed the baseline pipeline model in ADR causality inference.
    • Achieved a macro-weighted F-score ranging from 0.3652 to 0.5271.
    • Achieved a micro-weighted F-score ranging from 0.9523 to 0.9918, indicating high precision and recall.

    Conclusions:

    • Automated inference of ADR causality from EHRs is feasible and can significantly aid clinical judgment.
    • The developed joint model demonstrates superior performance in assessing drug-ADR relationships.
    • This approach holds promise for enhancing pharmacovigilance and patient safety through improved ADR detection.