Search

Diagnostic uncertainty is prevalent throughout medicine and significantly impacts patient care, especially when it goes unrecognized. However, we lack a reliable clinical means of identifying uncertainty. This study evaluates the narrative discourse within clinical notes in the Electronic Health Record as a means of identifying diagnostic uncertainty. Recognizing that discourse producers use language ‘semi-automatically’ (Partington et al. 2013), we hypothesized that clinicians include distinct indications of uncertainty in their written assessments, which could be elucidated by linguistic analysis. Using a cohort of patients prospectively identified as having an uncertain diagnosis (UD), we conducted a detailed corpus-assisted discourse analysis. The analysis revealed a set of linguistic indicators constitutive of diagnostic uncertainty including terms of modality, register-specific terms, and linguistically identifiable clinical behaviours. This dictionary of UD indicators was thoroughly tested, and its performance was compared with a matched-control dataset. Based on the findings, we built a machine learning classification algorithm with the ability to predict UD patient cohorts with 87.0% accuracy, effectively demonstrating the feasibility of using clinical discourse to classify patients and directly impact the clinical environment.

BACKGROUND Theme extraction by way of topic modeling has increasingly become a central interest in a variety of disciplines seeking to better understand large, unstructured text datasets without the burden of manual text analysis or the need for mechanisms and language experts to make sense of those analyses. Its popularity has resulted in the development of numerous algorithmic approaches and a revolutionary view on the value of studying natural language in many different contexts. Although these machine learning methods have brought a wealth of knowledge, there remains a certain degree of elusiveness in defining the language processing that occurs therein and therefore in the interpretability of its product. OBJECTIVE This study sought to capitalize on the synergy of linguistic analysis and traditional computational models to extend the capabilities of topic modeling, providing more refined and accessible results that will enhance our understanding of thematic elements and their relationships. METHODS To demonstrate this synergy, a corpus of over 22,000 abstracts from the American Society of Anesthesiologists Annual Meeting (ASAAM) between the years 2000-2013 was used. The corpus was analyzed using our novel method, “model of models” (MoM), which encompasses a blend of information retrieval, machine learning, sophisticated data visualizations and human verification. Then, these outputs were compared to both a gold standard of topic classification (i.e., ASAAM’s manual organization of accepted abstracts) and a classical clustering approach. RESULTS Our novel analytical method, MoM, indicated a high level of agreement with the gold standard of ASAAM classification. The MoM’s information-rich yet interpretable network visualization produced an interactive way of exploring the subtle relationships existent in the model, including cluster to cluster and cluster to model relationships, allowing for straightforward verification of validity and coherence as well as convenient interpretation of its results. The clusters produced were logically oriented, reflective of the abstracts’ latent connections with one another, and mostly paralleled those designated by the ASAAM. There were some gaps between the findings of the MoM approach and the traditional text clustering approach, resulting in discrepancies in terms of the best performing model or algorithm. CONCLUSIONS This study illustrated the value of re-introducing human input into our machine learning methodologies. Our findings demonstrate an extended functionality of tried-and-true topic modeling approaches, maintaining the computational efficiency offered through traditional techniques while adding a delicate nuance that allows for a deeper exploration of extracted themes, a better understanding of why those themes and their corresponding relationships exist, and the ability to verify the coherence of machine output. The study also reveals the disparities between our proposed method and classic techniques, opening a scholarly dialogue for future refinements of theme extraction and a more effective way to integrate linguistic analysis in an external evaluation process.

In the social media marketplace of ideas during the 2020 coronavirus pandemic, epidemiologists and other scientific and medical experts competed for attention with news media, government agencies, politicians, celebrities, and rank conspiracy theorists. However, everyone with a Twitter account was not equally qualified to speak knowledgeably about critical issues related to the outbreak, such as prevention and treatment. And, accurate information from informed sources can mean the difference between life and death. Our exploratory study addresses a simple, but important question: whose messages about the efficacy of hydroxychloroquine as a treatment for coronavirus were getting the most attention on Twitter? We provide a data visualization of Twitter activity for the period of 21 January through 21 May 2020 that shows users who tweeted about hydroxychloroquine, as well as who interacted with each of them (through likes, comments, retweets, etc.) to determine who were the most prominent voices on the network during a critical juncture of the outbreak. From our analysis, it appears that President Donald Trump’s handle ( @realDonaldTrump ) and other pro-Trump related accounts were the most influential voices on Twitter during this time of the crisis, rather than those from relevant experts, such as the Centers for Disease Control and Prevention ( @CDCgov ) or the National Institute of Allergy and Infectious Diseases ( @NIAIDnews ).

Research on viral outbreaks at the pandemic scale responds to heightened social urgency and the need to expedite scientific discovery from the “bench” to the “bedside” to the wider population. We sought to better understand translational research within the context of pandemics, both historical and present day, by tracking publication trends in the immediate aftermath of virus outbreaks. We used a blend of natural language processing (NLP), social network analysis and human annotation approaches to analyze the 85,663 articles in the COVID-19 Open Research Dataset (CORD-19). We found stable and repeated characteristics throughout subsets of peer-reviewed published literature corresponding to seven different viral outbreaks over the last several decades. Three distinct groups or “neighborhoods” recurred across all of the model networks – (1) bench science, (2) clinical treatments, and (3) broader public health trends. Notably, in each historical virus model, small “bridge” nodes representing translational research connect the three otherwise disconnected neighborhoods. These bridging studies embody research convergence by both integrating the vocabulary and methods of different disciplines and bodies of previous work and by citing other papers beyond their narrow field. In the case of COVID-19, the literature continues to evolve apace along with the virus, and we can witness the phases of response unfold as the science progresses. This study demonstrates how the different sectors of biomedical research respond independently to public health emergencies and how translational research can facilitate greater information synthesis and exchange between disciplinary silos.

Social media platforms have broadened the scope of voices responding to social justice movements, significantly impacting public conversations of important social justice issues. This social network analysis examined hashtags that were invoked on Twitter in the aftermath of the Mike Brown shooting in the St. Louis suburb of Ferguson in 2014. From the millions of tweets globally, the use of specific hashtags appeared to focus the conversation on Twitter toward the personal meaning of story events and framed the shooting as something relatable to the posters’ own lives and experiences.

Children with long-term ventilator dependence are a growing population that generate substantial cost to the healthcare system and require very lengthy admissions to initiate support. Respiratory therapy notes contain free-text descriptions of key respiratory events during these admissions but are underutilized. Using a retrospective electronic health record data set from 101 patients, we identified more clinically concerning patients, extracted key features from the free-text notes that differentiated these patients, and displayed those features in a timeline visualization that has implications for clinical decision support.