Attention-Gated Graph Convolutions for Extracting Drug Interaction Information from Drug Labels

Preventable adverse events as a result of medical errors present a growing concern in the healthcare system. As drug-drug interactions (DDIs) may lead to preventable adverse events, being able to extract DDIs from drug labels into a machine-processable form is an important step toward effective dissemination of drug safety information. Herein, we tackle the problem of jointly extracting mentions of drugs and their interactions, including interaction outcome, from drug labels. Our deep learning approach entails composing various intermediate representations, including graph-based context derived using graph convolutions (GCs) with a novel attention-based gating mechanism (holistically called GCA), which are combined in meaningful ways to predict on all subtasks jointly. Our model is trained and evaluated on the 2018 TAC DDI corpus. Our GCA model in conjunction with transfer learning performs at 39.20% F1 and 26.09% F1 on entity recognition (ER) and relation extraction (RE), respectively, on the first official test set and at 45.30% F1 and 27.87% F1 on ER and RE, respectively, on the second official test set. These updated results lead to improvements over our prior best by up to 6 absolute F1 points. After controlling for available training data, the proposed model exhibits state-of-the-art performance for this task.

Poor appetite and overeating reported by adults in Australia during the coronavirus-19 disease pandemic: a population-based study

OBJECTIVE

As a result of the coronavirus-19 disease (COVID-19) pandemic, Australia adopted emergency measures on 22 March 2020. This study reports the effect of the COVID-19 lockdown on appetite and overeating in Australian adults during the first month of emergency measures.

DESIGN

This study reports analysis of data from the population-based, self-completed survey. The main outcome measure was an item from the Patient Health Questionnaire 9 asking: 'Over the past 2 weeks, how often have you been bothered by poor appetite or overeating?'. Data on sociodemographic factors, symptoms of anxiety and depression, and the impact of COVID-19 and lockdown were also collected. Multivariable logistic regression was used to examine associations with poor appetite or overeating.

SETTING

An anonymous online survey available from 3 April to 2 May 2020.

PARTICIPANTS

A total of 13 829 Australian residents aged 18 years or over.

RESULTS

The weighted prevalence of being bothered by poor appetite or overeating in the past 2 weeks was 53·6 %, with 11·6 % (95 % CI 10·6, 12·6) of the cohort reporting poor appetite or overeating nearly every day. High levels of anxiety, concern about contracting COVID-19, being in lockdown with children and reporting a severe impact of the lockdown were associated with increased odds of poor appetite or overeating.

CONCLUSIONS

Given the widespread prevalence of being bothered by poor appetite or overeating, universal public health interventions to address emotion-focused or situational eating during periods of lockdown may be appropriate.

Identifying current Juul users among emerging adults through Twitter feeds

INTRODUCTION

Juul is the most popular electronic cigarette on the market. Amid concerns around uptake of e-cigarettes by never smokers, can we detect whether someone uses Juul based on their social media activities? This is the central premise of the effort reported in this paper. Several recent social media-related studies on Juul use tend to focus on the characterization of Juul-related messages on social media. In this study, we assess the potential in using machine learning methods to automatically identify Juul users (past 30-day usage) based on their Twitter data.

METHODS

We obtained a collection of 588 instances, for training and testing, of Juul use patterns (along with associated Twitter handles) via survey responses of college students. With this data, we built and tested supervised machine learning models based on linear and deep learning algorithms with textual, social network (friends and followers), and other hand-crafted features.

RESULTS

The linear model with textual and follower network features performed best with a precision-recall trade-off such that precision (PPV) is 57 % at 24 % recall (sensitivity). Hence, at least every other college-attending Twitter user flagged by our model is expected to be a Juul user. Additionally, our results indicate that social network features tend to have a large impact (positive) on classification performance.

CONCLUSION

There are enough latent signals from social feeds for supervised modeling of Juul use, even with limited training data, implying that such models are highly beneficial to very focused intervention campaigns. This initial success indicates potential for more involved automated surveillance of Juul use based on social media data, including Juul usage patterns, nicotine dependence, and risk awareness.

Toward on-board microchip synthesis of CdSe vs. PbSe nanocrystalline quantum dots as a spectral decoy for protecting space assets

Fabrication of the reaction chamber using silicon carbide. (A) A schematic sketch of the fabrication flow; (B) a photograph of a transparent 6 inch SiC-on-glass wafer; (C) the surface morphology of the SiC film.

Occurrence and molecular characteristics of Listeria monocytogenes isolated from ready-to-eat meats in Hanoi, Vietnam

Listeria monocytogenes represents one of the most serious threats to food safety. Several studies have shown that Ready-To- Eat (RTE) meats are an important vehicle responsible for listeriosis in human. In Vietnam, little is known about the occurrence and molecular characteristics of L. monocytogenes in meat products, which are essential for developing monitoring plans and control measures. In the present study, we investigated the occurrence of L. monocytogenes in 258 sausage and sliced meat samples collected during the period of 2013-2015 and determined the genetic diversity of the isolates using multi-locus sequence typing (MLST). Overall, L. monocytogenes was present in 19/129 (14.7 %) and 40/129 (31.0 %) sausage and sliced meat samples respectively, with the peak of occurrence being in summer. Furthermore, a minimum spanning tree was constructed based on MLST data of 47 isolates. A total of 15 sequence types were found, with five being novel. Notably, the majority of the isolates (34/47) belonged to the hypervirulent clonal complexes 1, 2, and 3.

Urine dicarboxylic acids change in pre-symptomatic Alzheimer's disease and reflect loss of energy capacity and hippocampal volume

Non-invasive biomarkers will enable widespread screening and early diagnosis of Alzheimer’s disease (AD). We hypothesized that the considerable loss of brain tissue in AD will result in detection of brain lipid components in urine, and that these will change in concert with CSF and brain biomarkers of AD. We examined urine dicarboxylic acids (DCA) of carbon length 3–10 to reflect products of oxidative damage and energy generation or balance that may account for changes in brain function in AD. Mean C4-C5 DCAs were lower and mean C7-C10 DCAs were higher in the urine from AD compared to cognitively healthy (CH) individuals. Moreover, mean C4-C5 DCAs were lower and mean C7-C9 were higher in urine from CH individuals with abnormal compared to normal CSF amyloid and Tau levels; i.e., the apparent urine changes in AD also appeared to be present in CH individuals that have CSF risk factors of early AD pathology. In examining the relationship between urine DCAs and AD biomarkers, we found short chain DCAs positively correlated with CSF Aβ₄₂, while C7-C10 DCAs negatively correlated with CSF Aβ₄₂ and positively correlated with CSF Tau levels. Furthermore, we found a negative correlation of C7-C10 DCAs with hippocampal volume (p < 0.01), which was not found in the occipital volume. Urine measures of DCAs have an 82% ability to predict cognitively healthy participants with normal CSF amyloid/Tau. These data suggest that urine measures of increased lipoxidation and dysfunctional energy balance reflect early AD pathology from brain and CSF biomarkers. Measures of urine DCAs may contribute to personalized healthcare by indicating AD pathology and may be utilized to explore population wellness or monitor the efficacy of therapies in clinical trials.

Social media surveillance for perceived therapeutic effects of cannabidiol (CBD) products

BACKGROUND

CBD products have risen in popularity given CBD's therapeutic potential and lack of legal oversight, despite lacking conclusive scientific evidence for widespread over-the-counter usage for many of its perceived benefits. While medical evidence is being generated, social media surveillance offers a fast and inexpensive alternative to traditional surveys in ascertaining perceived therapeutic purposes and modes of consumption for CBD products.

METHODS

We collected all comments from the CBD subreddit posted between January 1 and April 30, 2019 as well as comments submitted to the FDA regarding regulation of cannabis-derived products and analyzed them using a rule-based language processing method. A relative ranking of popular therapeutic uses and product groups for CBD is obtained based on frequency of pattern matches including precise queries that entail identifying mentions of the condition, a CBD product, and some "trigger" phrase indicating therapeutic use. We validated the social media-based findings using a similar analysis on comments to the U.S. Food and Drug Administration's (FDA) 2019 request-for-comments on cannabis-derived products.

RESULTS

CBD is mostly discussed as a remedy for anxiety disorders and pain and this is consistent across both comment sources. Of comments posted to the CBD subreddit during the monitored time span, 6.19% mentioned anxiety at least once with at least 6.02% of these comments specifically mentioning CBD as a treatment for anxiety (i.e., 0.37% of total comments). The most popular CBD product group is oil and tinctures.

CONCLUSION

Social media surveillance of CBD usage has the potential to surface new therapeutic use-cases as they are posted. Contemporary social media data indicate, for example, that stress and nausea are frequently mentioned as therapeutic use cases for CBD without corresponding evidence, that affirms or denies, in the research literature. However, the abundance of anecdotal claims warrants serious scientific exploration moving forward. Meanwhile, as FDA ponders regulation, our effort demonstrates that social data offers a convenient affordance to surveil for CBD usage patterns in a way that is fast and inexpensive and can inform conventional electronic surveys.

European Position Paper on Rhinosinusitis and Nasal Polyps 2020

The European Position Paper on Rhinosinusitis and Nasal Polyps 2020 is the update of similar evidence based position papers published in 2005 and 2007 and 2012. The core objective of the EPOS2020 guideline is to provide revised, up-to-date and clear evidence-based recommendations and integrated care pathways in ARS and CRS. EPOS2020 provides an update on the literature published and studies undertaken in the eight years since the EPOS2012 position paper was published and addresses areas not extensively covered in EPOS2012 such as paediatric CRS and sinus surgery. EPOS2020 also involves new stakeholders, including pharmacists and patients, and addresses new target users who have become more involved in the management and treatment of rhinosinusitis since the publication of the last EPOS document, including pharmacists, nurses, specialised care givers and indeed patients themselves, who employ increasing self-management of their condition using over the counter treatments. The document provides suggestions for future research in this area and offers updated guidance for definitions and outcome measurements in research in different settings. EPOS2020 contains chapters on definitions and classification where we have defined a large number of terms and indicated preferred terms. A new classification of CRS into primary and secondary CRS and further division into localized and diffuse disease, based on anatomic distribution is proposed. There are extensive chapters on epidemiology and predisposing factors, inflammatory mechanisms, (differential) diagnosis of facial pain, allergic rhinitis, genetics, cystic fibrosis, aspirin exacerbated respiratory disease, immunodeficiencies, allergic fungal rhinosinusitis and the relationship between upper and lower airways. The chapters on paediatric acute and chronic rhinosinusitis are totally rewritten. All available evidence for the management of acute rhinosinusitis and chronic rhinosinusitis with or without nasal polyps in adults and children is systematically reviewed and integrated care pathways based on the evidence are proposed. Despite considerable increases in the amount of quality publications in recent years, a large number of practical clinical questions remain. It was agreed that the best way to address these was to conduct a Delphi exercise . The results have been integrated into the respective sections. Last but not least, advice for patients and pharmacists and a new list of research needs are included. The full document can be downloaded for free on the website of this journal: http://www.rhinologyjournal.com.

Impact of osteoporotic fracture type and subsequent fracture on mortality: the Tromsø Study

Less is known about the impact of non-hip non-vertebral fractures (NHNV) on early death. This study demonstrated increased risk of dying following hip and NHNV fractures which was further increased by a subsequent fracture. This highlights the importance of early intervention to prevent both initial and subsequent fractures and improve survival.

INTRODUCTION

Osteoporotic fractures are a major health concern. Limited evidence exists on their impact on mortality in ageing populations. This study examined the contribution of initial fracture type and subsequent fracture on mortality in a Norwegian population that has one of the highest rates of fractures.

METHODS

The Tromsø Study is a prospective population-based cohort in Norway. Women and men aged 50+ years were followed from 1994 to 2010. All incident hip and non-hip non-vertebral (NHNV) fractures were registered. NHNV fractures were classified as either proximal or distal. Information on self-reported co-morbidities, lifestyle factors, general health and education level was collected. Multivariable Cox models were used to quantify mortality risk with incident and subsequent fractures analysed as time-dependent variables.

RESULTS

Of 5214 women and 4620 men, 1549 (30%) and 504 (11%) sustained a fracture, followed by 589 (38%) and 254 (51%) deaths over 10,523 and 2821 person-years, respectively. There were 403 (26%) subsequent fractures in women and 68 (13%) in men. Hip fracture was associated with a two-fold increase in mortality risk (HR 2.05, 95% CI 1.73-2.42 in women and 2.49, 95% CI 2.00-3.11 in men). Proximal NHNV fractures were associated with 49% and 81% increased mortality risk in women and men (HR 1.49, 95% CI 1.21-1.84 and 1.81, 95% CI 1.37-2.41), respectively. Distal NHNV fractures were not associated with mortality. Subsequent fracture was associated with 89% and 77% increased mortality risk in women and men (HR 1.89, 95% CI 1.52-2.35 and 1.77, 95% CI 1.16-2.71), respectively.

CONCLUSION

Hip, proximal NHNV and subsequent fractures were significantly associated with increased mortality risk in the elderly, highlighting the importance of early intervention.