Identifying adverse drug event information in clinical notes with distributional semantic representations of context

Abstract

For the purpose of post-marketing drug safety surveillance, which has traditionally relied on the voluntary reporting of individual cases of adverse drug events (ADEs), other sources of information are now being explored, including electronic health records (EHRs), which give us access to enormous amounts of longitudinal observations of the treatment of patients and their drug use. Adverse drug events, which can be encoded in EHRs with certain diagnosis codes, are, however, heavily underreported. It is therefore important to develop capabilities to process, by means of computational methods, the more unstructured EHR data in the form of clinical notes, where clinicians may describe and reason around suspected ADEs. In this study, we report on the creation of an annotated corpus of Swedish health records for the purpose of learning to identify information pertaining to ADEs present in clinical notes. To this end, three key tasks are tackled: recognizing relevant named entities (disorders, symptoms, drugs), labeling attributes of the recognized entities (negation, speculation, temporality), and relationships between them (indication, adverse drug event). For each of the three tasks, leveraging models of distributional semantics - i.e., unsupervised methods that exploit co-occurrence information to model, typically in vector space, the meaning of words - and, in particular, combinations of such models, is shown to improve the predictive performance. The ability to make use of such unsupervised methods is critical when faced with large amounts of sparse and high-dimensional data, especially in domains where annotated resources are scarce.

Learning from heterogeneous temporal data in electronic health records

“…Research on mining both structured and unstructured EHR data for ADE detection is nascent, see e.g. [37,38,39,40,41]. Unfortunately, most approaches to ADE detection in EHRs do not take into account the temporality of clinical events, which is critical for this task.…”

A corpus for mining drug-related knowledge from Twitter chatter: Language models and their utilities

“…These models can be used to explore a variety of research tasks that can benefit from drug related knowledge generated directly from the users, such as exploring associations between drugs and adverse reactions, identifying drug abuse related signals, and, from a more NLP/data science perspective, the effects of different parameters such as context window and vector sizes in capturing semantic and syntactic properties. Such distributed word representation models are already being applied for research utilizing other sources of noisy health-related data, such as clinical reports [5] and such models have been generated from texts from other domains such as published literature [6] and generic social media [7]. However, perhaps due to the absence of available drug-related chatter data from social media, there are currently no such models available for this domain.…”

Finding Important Terms for Patients in Their Electronic Health Records: A Learning-to-Rank Approach Using Expert Annotations

“…It has shown to improve several biomedical and clinical NLP tasks, such as biomedical named entity recognition [54,55], protein-protein interaction detection [56], biomedical event extraction [57,58], adverse drug event detection [59,60], ranking biomedical synonyms [61], and disambiguating clinical abbreviations [62,63]. …”