Adverse Drug Events Detection in Clinical Notes by Jointly Modeling Entities and Relations Using Neural Networks

Bharath Dandala, Venkata Joopudi, Murthy Devarakonda

    Research output: Contribution to journalArticle

    3 Citations (Scopus)

    Abstract

    Background and Significance: Adverse drug events (ADEs) occur in approximately 2–5% of hospitalized patients, often resulting in poor outcomes or even death. Extraction of ADEs from clinical narratives can accelerate and automate pharmacovigilance. Using state-of-the-art deep-learning neural networks to jointly model concept and relation extraction, we achieved the highest integrated task score in the 2018 Medication and Adverse Drug Event (MADE) 1.0 challenge. Methods: We used a combined bidirectional long short-term memory (BiLSTM) and conditional random fields (CRF) neural network to detect medical entities relevant to ADEs and a combined BiLSTM and attention network to determine relations, including the adverse drug reaction relation between medication and sign or symptom entities. Using these models, we conducted three experiments: (1) separate and sequential modeling of entities and relations; (2) joint modeling where relations between medications and sign or symptoms determined ADE and indication entities; (3) use of information from external resources such as the US FDA’s adverse event database as additional input to the second method. Results: Joint modeling improved the overall task accuracy from 0.62 to 0.65 F measure, and the additional use of external resources improved the accuracy to 0.66 F measure. Given the gold-standard medical entity labels, the joint model plus external resources method achieved F measures of 0.83 for ADE-relevant medical entity detection and 0.87 for relation detection. Conclusion: It is important to use joint modeling techniques and external resources for effectively detecting ADEs from clinical narratives in electronic health record (EHR) systems. While the extraction of entities and relations individually achieved high accuracy, the integrated task still has room for further improvement.

    Original languageEnglish (US)
    JournalDrug Safety
    DOIs
    StateAccepted/In press - Jan 1 2019

    Fingerprint

    Drug-Related Side Effects and Adverse Reactions
    Neural networks
    Pharmaceutical Preparations
    Joints
    Long-Term Memory
    Short-Term Memory
    Signs and Symptoms
    Pharmacovigilance
    Electronic Health Records
    Gold
    Labels
    Health
    Learning
    Databases

    ASJC Scopus subject areas

    • Toxicology
    • Pharmacology
    • Pharmacology (medical)

    Cite this

    Adverse Drug Events Detection in Clinical Notes by Jointly Modeling Entities and Relations Using Neural Networks. / Dandala, Bharath; Joopudi, Venkata; Devarakonda, Murthy.

    In: Drug Safety, 01.01.2019.

    Research output: Contribution to journalArticle

    @article{2d7288c9d14249af9fe96d5cf5e5d9f5,
    title = "Adverse Drug Events Detection in Clinical Notes by Jointly Modeling Entities and Relations Using Neural Networks",
    abstract = "Background and Significance: Adverse drug events (ADEs) occur in approximately 2–5{\%} of hospitalized patients, often resulting in poor outcomes or even death. Extraction of ADEs from clinical narratives can accelerate and automate pharmacovigilance. Using state-of-the-art deep-learning neural networks to jointly model concept and relation extraction, we achieved the highest integrated task score in the 2018 Medication and Adverse Drug Event (MADE) 1.0 challenge. Methods: We used a combined bidirectional long short-term memory (BiLSTM) and conditional random fields (CRF) neural network to detect medical entities relevant to ADEs and a combined BiLSTM and attention network to determine relations, including the adverse drug reaction relation between medication and sign or symptom entities. Using these models, we conducted three experiments: (1) separate and sequential modeling of entities and relations; (2) joint modeling where relations between medications and sign or symptoms determined ADE and indication entities; (3) use of information from external resources such as the US FDA’s adverse event database as additional input to the second method. Results: Joint modeling improved the overall task accuracy from 0.62 to 0.65 F measure, and the additional use of external resources improved the accuracy to 0.66 F measure. Given the gold-standard medical entity labels, the joint model plus external resources method achieved F measures of 0.83 for ADE-relevant medical entity detection and 0.87 for relation detection. Conclusion: It is important to use joint modeling techniques and external resources for effectively detecting ADEs from clinical narratives in electronic health record (EHR) systems. While the extraction of entities and relations individually achieved high accuracy, the integrated task still has room for further improvement.",
    author = "Bharath Dandala and Venkata Joopudi and Murthy Devarakonda",
    year = "2019",
    month = "1",
    day = "1",
    doi = "10.1007/s40264-018-0764-x",
    language = "English (US)",
    journal = "Drug Safety",
    issn = "0114-5916",
    publisher = "Adis International Ltd",

    }

    TY - JOUR

    T1 - Adverse Drug Events Detection in Clinical Notes by Jointly Modeling Entities and Relations Using Neural Networks

    AU - Dandala, Bharath

    AU - Joopudi, Venkata

    AU - Devarakonda, Murthy

    PY - 2019/1/1

    Y1 - 2019/1/1

    N2 - Background and Significance: Adverse drug events (ADEs) occur in approximately 2–5% of hospitalized patients, often resulting in poor outcomes or even death. Extraction of ADEs from clinical narratives can accelerate and automate pharmacovigilance. Using state-of-the-art deep-learning neural networks to jointly model concept and relation extraction, we achieved the highest integrated task score in the 2018 Medication and Adverse Drug Event (MADE) 1.0 challenge. Methods: We used a combined bidirectional long short-term memory (BiLSTM) and conditional random fields (CRF) neural network to detect medical entities relevant to ADEs and a combined BiLSTM and attention network to determine relations, including the adverse drug reaction relation between medication and sign or symptom entities. Using these models, we conducted three experiments: (1) separate and sequential modeling of entities and relations; (2) joint modeling where relations between medications and sign or symptoms determined ADE and indication entities; (3) use of information from external resources such as the US FDA’s adverse event database as additional input to the second method. Results: Joint modeling improved the overall task accuracy from 0.62 to 0.65 F measure, and the additional use of external resources improved the accuracy to 0.66 F measure. Given the gold-standard medical entity labels, the joint model plus external resources method achieved F measures of 0.83 for ADE-relevant medical entity detection and 0.87 for relation detection. Conclusion: It is important to use joint modeling techniques and external resources for effectively detecting ADEs from clinical narratives in electronic health record (EHR) systems. While the extraction of entities and relations individually achieved high accuracy, the integrated task still has room for further improvement.

    AB - Background and Significance: Adverse drug events (ADEs) occur in approximately 2–5% of hospitalized patients, often resulting in poor outcomes or even death. Extraction of ADEs from clinical narratives can accelerate and automate pharmacovigilance. Using state-of-the-art deep-learning neural networks to jointly model concept and relation extraction, we achieved the highest integrated task score in the 2018 Medication and Adverse Drug Event (MADE) 1.0 challenge. Methods: We used a combined bidirectional long short-term memory (BiLSTM) and conditional random fields (CRF) neural network to detect medical entities relevant to ADEs and a combined BiLSTM and attention network to determine relations, including the adverse drug reaction relation between medication and sign or symptom entities. Using these models, we conducted three experiments: (1) separate and sequential modeling of entities and relations; (2) joint modeling where relations between medications and sign or symptoms determined ADE and indication entities; (3) use of information from external resources such as the US FDA’s adverse event database as additional input to the second method. Results: Joint modeling improved the overall task accuracy from 0.62 to 0.65 F measure, and the additional use of external resources improved the accuracy to 0.66 F measure. Given the gold-standard medical entity labels, the joint model plus external resources method achieved F measures of 0.83 for ADE-relevant medical entity detection and 0.87 for relation detection. Conclusion: It is important to use joint modeling techniques and external resources for effectively detecting ADEs from clinical narratives in electronic health record (EHR) systems. While the extraction of entities and relations individually achieved high accuracy, the integrated task still has room for further improvement.

    UR - http://www.scopus.com/inward/record.url?scp=85060126752&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85060126752&partnerID=8YFLogxK

    U2 - 10.1007/s40264-018-0764-x

    DO - 10.1007/s40264-018-0764-x

    M3 - Article

    C2 - 30649738

    AN - SCOPUS:85060126752

    JO - Drug Safety

    JF - Drug Safety

    SN - 0114-5916

    ER -