In secondary analysis of electronic health records, a crucial task consists in correctly identifying the patient cohort under investigation. In many cases, the most valuable and relevant information ...for an accurate classification of medical conditions exist only in clinical narratives. Therefore, it is necessary to use natural language processing (NLP) techniques to extract and evaluate these narratives. The most commonly used approach to this problem relies on extracting a number of clinician-defined medical concepts from text and using machine learning techniques to identify whether a particular patient has a certain condition. However, recent advances in deep learning and NLP enable models to learn a rich representation of (medical) language. Convolutional neural networks (CNN) for text classification can augment the existing techniques by leveraging the representation of language to learn which phrases in a text are relevant for a given medical condition. In this work, we compare concept extraction based methods with CNNs and other commonly used models in NLP in ten phenotyping tasks using 1,610 discharge summaries from the MIMIC-III database. We show that CNNs outperform concept extraction based methods in almost all of the tasks, with an improvement in F1-score of up to 26 and up to 7 percentage points in area under the ROC curve (AUC). We additionally assess the interpretability of both approaches by presenting and evaluating methods that calculate and extract the most salient phrases for a prediction. The results indicate that CNNs are a valid alternative to existing approaches in patient phenotyping and cohort identification, and should be further investigated. Moreover, the deep learning approach presented in this paper can be used to assist clinicians during chart review or support the extraction of billing codes from text by identifying and highlighting relevant phrases for various medical conditions.
The latent viral reservoir is the critical barrier for the development of a cure for HIV-1 infection. Previous studies have shown direct antiviral activity of potent HIV-1 Env-specific broadly ...neutralizing antibodies (bNAbs) administered when antiretroviral therapy (ART) was discontinued, but it remains unclear whether bNAbs can target the viral reservoir during ART. Here we show that administration of the V3 glycan-dependent bNAb PGT121 together with the Toll-like receptor 7 (TLR7) agonist vesatolimod (GS-9620) during ART delayed viral rebound following discontinuation of ART in simian-human immunodeficiency virus (SHIV)-SF162P3-infected rhesus monkeys in which ART was initiated during early acute infection. Moreover, in the subset of monkeys that were treated with both PGT121 and GS-9620 and that did not show viral rebound after discontinuation of ART, adoptive transfer studies and CD8-depletion studies also did not reveal virus. These data demonstrate the potential of bNAb administration together with innate immune stimulation as a possible strategy for targeting the viral reservoir.
Zika virus (ZIKV) is a flavivirus that is responsible for the current epidemic in Brazil and the Americas. ZIKV has been causally associated with fetal microcephaly, intrauterine growth restriction, ...and other birth defects in both humans and mice. The rapid development of a safe and effective ZIKV vaccine is a global health priority, but very little is currently known about ZIKV immunology and mechanisms of immune protection. Here we show that a single immunization with a plasmid DNA vaccine or a purified inactivated virus vaccine provides complete protection in susceptible mice against challenge with a strain of ZIKV involved in the outbreak in northeast Brazil. This ZIKV strain has recently been shown to cross the placenta and to induce fetal microcephaly and other congenital malformations in mice. We produced DNA vaccines expressing ZIKV pre-membrane and envelope (prM-Env), as well as a series of deletion mutants. The prM-Env DNA vaccine, but not the deletion mutants, afforded complete protection against ZIKV, as measured by absence of detectable viraemia following challenge, and protective efficacy correlated with Env-specific antibody titers. Adoptive transfer of purified IgG from vaccinated mice conferred passive protection, and depletion of CD4 and CD8 T lymphocytes in vaccinated mice did not abrogate this protection. These data demonstrate that protection against ZIKV challenge can be achieved by single-shot subunit and inactivated virus vaccines in mice and that Env-specific antibody titers represent key immunologic correlates of protection. Our findings suggest that the development of a ZIKV vaccine for humans is likely to be achievable.
Zika virus (ZIKV) is associated with severe neuropathology in neonates as well as Guillain-Barré syndrome and other neurologic disorders in adults. Prolonged viral shedding has been reported in ...semen, suggesting the presence of anatomic viral reservoirs. Here we show that ZIKV can persist in cerebrospinal fluid (CSF) and lymph nodes (LN) of infected rhesus monkeys for weeks after virus has been cleared from peripheral blood, urine, and mucosal secretions. ZIKV-specific neutralizing antibodies correlated with rapid clearance of virus in peripheral blood but remained undetectable in CSF for the duration of the study. Viral persistence in both CSF and LN correlated with upregulation of mechanistic target of rapamycin (mTOR), proinflammatory, and anti-apoptotic signaling pathways, as well as downregulation of extracellular matrix signaling pathways. These data raise the possibility that persistent or occult neurologic and lymphoid disease may occur following clearance of peripheral virus in ZIKV-infected individuals.
Display omitted
•ZIKV can persist in cerebrospinal fluid and lymph nodes of infected rhesus monkeys•ZIKV persistence correlates with modulation of specific transcriptomic pathways•ZIKV anatomic sanctuaries may contribute to viral pathogenesis
Persistence of Zika virus in the CNS and lymphoid tissues of NHPs suggests that ZIKV infection may have consequences to humans beyond the reported birth defects.
We describe a case of necrotizing fasciitis in the United Kingdom in which Pseudomonas guariconensis was isolated from multiple blood culture and tissue samples. The organism carried a Verona ...integron-encoded metallo-β-lactamase gene and evidence of decreased susceptibility to β-lactam antimicrobial agents. Clinicians should use caution when treating infection caused by this rare pathogen.
Documentation of care preferences within 48 hours of admission to an intensive care unit (ICU) is a National Quality Forum-endorsed quality metric for older adults. Care preferences are poorly ...captured by administrative data.
Using deep natural language processing, our aim was to determine the rate of care preference documentation in free-text notes and to assess associated patient factors.
Retrospective review of notes by clinicians using a deep natural language processing to identify care preference documentation, including goals-of-care and treatment limitations, within 48 hours of ICU admission within five ICUs (medical, cardiac, surgery, trauma surgery, and cardiac surgery) for adults 75 years and older. Covariates included demographics, ICU type, sequential organ failure assessment score, and need for mechanical ventilation.
Deep natural language processing reviewed 11,575 clinician notes for 1350 ICU admissions. Median patient age was 84.0 years (interquartile range 78.0–88.4). Overall, 64.7% had documentation of care preferences. Patients with documentation were older (85 vs. 83 years; P < 0.001) and more often female (53.8% vs. 43.4%; P < 0.001). In adjusted analysis, rates of care preference documentation were higher for older patients, females, nonelective admissions, and admissions to the medical vs. the cardiac or surgical ICUs (all P ≤ 0.01).
Care preference documentation within 48 hours was absent in more than one-third of ICU admissions among patients aged 75 years and older and was more likely to occur in medical vs. cardiac or surgical ICUs.
Clinicians lack reliable methods to predict which patients with congestive heart failure (CHF) will benefit from cardiac resynchronization therapy (CRT). Symptom burden may help to predict response, ...but this information is buried in free-text clinical notes. Natural language processing (NLP) may identify symptoms recorded in the electronic health record and thereby enable this information to inform clinical decisions about the appropriateness of CRT.
To develop, train, and test a deep NLP model that identifies documented symptoms in patients with CHF receiving CRT.
We identified a random sample of clinical notes from a cohort of patients with CHF who later received CRT. Investigators labeled documented symptoms as present, absent, and context dependent (pathologic depending on the clinical situation). The algorithm was trained on 80% and fine-tuned parameters on 10% of the notes. We tested the model on the remaining 10%. We compared the model's performance to investigators' annotations using accuracy, precision (positive predictive value), recall (sensitivity), and F1 score (a combined measure of precision and recall).
Investigators annotated 154 notes (352,157 words) and identified 1340 present, 1300 absent, and 221 context-dependent symptoms. In the test set of 15 notes (35,467 words), the model's accuracy was 99.4% and recall was 66.8%. Precision was 77.6%, and overall F1 score was 71.8. F1 scores for present (70.8) and absent (74.7) symptoms were higher than that for context-dependent symptoms (48.3).
A deep NLP algorithm can be trained to capture symptoms in patients with CHF who received CRT with promising precision and recall.
•Multidisciplinary team (MDT) model ensured a two-year long medical text processing (NLP) project was clinically relevant and technically solvable.•An academic setting helped connect diverse experts ...in databases, statistics, epidemiology, machine learning, NLP, engineering and medicine.•Up-front and regular time investment in cross-disciplinary learning and aligning research goals helped the MDT collaborated more effectively.•Domain knowledge from NLP and clinical experts was required to optimize the data sampling, annotation strategies and algorithm development.•Dynamic leadership, regular meetings and updates, clear task delegation, and an open learning environment helped maintain motivation and momentum.
Advancement of Artificial Intelligence (AI) capabilities in medicine can help address many pressing problems in healthcare. However, AI research endeavors in healthcare may not be clinically relevant, may have unrealistic expectations, or may not be explicit enough about their limitations. A diverse and well-functioning multidisciplinary team (MDT) can help identify appropriate and achievable AI research agendas in healthcare, and advance medical AI technologies by developing AI algorithms as well as addressing the shortage of appropriately labeled datasets for machine learning. In this paper, our team of engineers, clinicians and machine learning experts share their experience and lessons learned from their two-year-long collaboration on a natural language processing (NLP) research project. We highlight specific challenges encountered in cross-disciplinary teamwork, dataset creation for NLP research, and expectation setting for current medical AI technologies.
Latent knowledge can be extracted from the electronic notes that are recorded during patient encounters with the health system. Using these clinical notes to decipher a patient's underlying ...comorbidites, symptom burdens, and treatment courses is an ongoing challenge. Latent topic model as an efficient Bayesian method can be used to model each patient's clinical notes as "documents" and the words in the notes as "tokens". However, standard latent topic models assume that all of the notes follow the same topic distribution, regardless of the type of note or the domain expertise of the author (such as doctors or nurses). We propose a novel application of latent topic modeling, using multi-note topic model (MNTM) to jointly infer distinct topic distributions of notes of different types. We applied our model to clinical notes from the MIMIC-III dataset to infer distinct topic distributions over the physician and nursing note types. Based on manual assessments made by clinicians, we observed a significant improvement in topic interpretability using MNTM modeling over the baseline single-note topic models that ignore the note types. Moreover, our MNTM model led to a significantly higher prediction accuracy for prolonged mechanical ventilation and mortality using only the first 48 hours of patient data. By correlating the patients' topic mixture with hospital mortality and prolonged mechanical ventilation, we identified several diagnostic topics that are associated with poor outcomes. Because of its elegant and intuitive formation, we envision a broad application of our approach in mining multi-modality text-based healthcare information that goes beyond clinical notes. Code available at https://github.com/li-lab-mcgill/heterogeneous_ehr.
Among patients with acute respiratory failure requiring prolonged mechanical ventilation, tracheostomies are typically placed after approximately 7 to 10 days. Yet half of patients admitted to the ...intensive care unit receiving tracheostomy will die within a year, often within three months. Existing mortality prediction models for prolonged mechanical ventilation, such as the ProVent Score, have poor sensitivity and are not applied until after 14 days of mechanical ventilation. We developed a model to predict 3-month mortality in patients requiring more than 7 days of mechanical ventilation using deep learning techniques and compared this to existing mortality models. Retrospective cohort study. Setting: The Medical Information Mart for Intensive Care III Database. Patients: All adults requiring greater than or equal to 7 days of mechanical ventilation. Measurements: A neural network model for 3-month mortality was created using process-of-care variables, including demographic, physiologic and clinical data. The area under the receiver operator curve (AUROC) was compared to the ProVent model at predicting 3 and 12-month mortality. Shapley values were used to identify the variables with the greatest contributions to the model. There were 4,334 encounters divided into a development cohort (n = 3467) and a testing cohort (n = 867). The final deep learning model included 250 variables and had an AUROC of 0.74 for predicting 3-month mortality at day 7 of mechanical ventilation versus 0.59 for the ProVent model. Older age and elevated Simplified Acute Physiology Score II (SAPS II) Score on intensive care unit admission had the largest contribution to predicting mortality. We developed a deep learning prediction model for 3-month mortality among patients requiring greater than or equal to 7 days of mechanical ventilation using a neural network approach utilizing readily available clinical variables. The model outperforms the ProVent model for predicting mortality among patients requiring greater than or equal to 7 days of mechanical ventilation. This model requires external validation.