The purpose of this retrospective study is to determine whether frailty is predictive of 30-day readmission in adults 50 years of age and older who were admitted with a psychiatric diagnosis to a behavioral health hospital, 2013-2017. A total of 1,063 patients were included. A 26-item frailty risk score (FRS-26-ICD10) was constructed from electronic health record (EHR) data. There were 114 readmissions.Cox regression modeling for demographic characteristics, emergent admission, comorbidity, and FRS-26-ICD determined prediction of time to readmission was modest (iAUC=0.671; the FRS-26-ICD was a significant predictor of readmission alone and in models with demographics and emergent admission; however, only the ECI was significantly related to hazard of readmission adjusting for other factors (adj.HR = 1.26, 95% CI=(1.17, 1.37), p<0.001) while FRS-26-ICD became non-significant. Frailty is a relevant syndrome in behavioral health that should be further studied in risk prediction and incorporated into care planning to prevent readmissions.
The purpose of the current study was to investigate the predictive properties of five definitions of a frailty risk score (FRS) and three comorbidity indices using data from electronic health records (EHRs) of hospitalized adults aged ≥50 years for 3-day, 7-day, and 30-day readmission, and to identify an optimal model for a FRS and comorbidity combination. Retrospective analysis of the EHR dataset was performed, and multivariable logistic regression and area under the curve (AUC) were used to examine readmission for frailty and comorbidity. The sample (N = 55,778) was mostly female (53%), non-Hispanic White (73%), married (53%), and on Medicare (55%). Mean FRSs ranged from 1.3 (SD = 1.5) to 4.3 (SD = 2.1). FRS and co- morbidity were independently associated with readmission. Predictive accuracy for FRS and comorbidity combinations ranged from AUC of 0.75 to 0.77 (30-day readmission) to 0.84 to 0.85 (3-day readmission). FRS and comorbidity combinations performed similarly well, whereas comorbidity was always indepen- dently associated with readmission. FRS measures were more associated with 30-day readmission than 7-day and 3-day readmission.
Streaming social media provides a real-time glimpse of extreme weather impacts. However, the volume of streaming data makes mining information a challenge for emergency managers, policy makers, and disciplinary scientists. Here we explore the effectiveness of data learned approaches to mine and filter information from streaming social media data from Hurricane Irma’s landfall in Florida, USA. We use 54,383 Twitter messages (out of 784 K geolocated messages) from 16,598 users from Sept. 10–12, 2017 to develop 4 independent models to filter data for relevance: 1) a geospatial model based on forcing conditions at the place and time of each tweet, 2) an image classification model for tweets that include images, 3) a user model to predict the reliability of the tweeter, and 4) a text model to determine if the text is related to Hurricane Irma. All four models are inde- pendently tested, and can be combined to quickly filter and visualize tweets based on user-defined thresholds for each submodel. We envision that this type of filtering and visualization routine can be useful as a base model for data capture from noisy sources such as Twitter. The data can then be subsequently used by policy makers, environmental managers, emergency managers, and domain scientists interested in finding tweets with specific attributes to use during different stages of the disaster (e.g., preparedness, response, and recovery), or for detailed research.
psi-collect is a command line tool for collecting post storm imagery from National Geodetic Survey (NGS) Remote Sensing Division of the US National Oceanographic and Atmospheric Administration. The tool enables reproducible computational workflows in downstream learning and labeling tasks and uses parallel processing to capture over 100,000 images each with an average size of 7.7 Mb from several different sources.
This study investigates Twitter usage during Hurricane Sandy following the survey of the general population and exploring communication dynamics on Twitter through different modalities. The results suggest that Twitter is a highly valuable source of disaster-related information particularly during the power outage. With a substantial increase in the number of tweets and unique users during the Hurricane Sandy, a large number of posts contained firsthand information about the hurricane showing the intensity of the event in real-time. More specifically, a number of images of damage and flooding were shared on Twitter through which researchers and emergency managers can retrieve valuable information to help identify storm damages and plan relief efforts. The social media analysis revealed the most important information that can be derived from twitter during disasters so that authorities can successfully utilize such data. The findings provide insights into the choice of keywords and sentiments and identifying the influential actors at different stages of disasters. A number of key influencers and their followers from different domains including political, news, weather, and relief organizations participated in Twitter-based discussions related to Hurricane Sandy. The connectivity of the influencers and their followers on Twitter plays a vital role in information sharing and dissemination throughout the hurricane. These connections can provide an effective vehicle for emergency managers towards establishing better bi-directional communication during disasters. However, while government agencies were among the prominent Twitter users during the Hurricane Sandy, they primarily relied on one-way communication rather than engaging with their audiences, a challenge that need to be addressed in future research.
Background: Diabetes and cardiovascular disease are two of the main causes of death in the United States. Identifying and predicting these diseases in patients is the first step towards stopping their progression. We evaluate the capabilities of machine learning models in detecting at-risk patients using survey data (and laboratory results), and identify key variables within the data contributing to these diseases among the patients.
Methods: Our research explores data-driven approaches which utilize supervised machine learning models to identify patients with such diseases. Using the National Health and Nutrition Examination Survey (NHANES) dataset, we conduct an exhaustive search of all available feature variables within the data to develop models for cardiovascular, prediabetes, and diabetes detection. Using different time-frames and feature sets for the data (based on laboratory data), multiple machine learning models (logistic regression, support vector machines, random forest, and gradient boosting) were evaluated on their classification performance. The models were then combined to develop a weighted ensemble model, capable of leveraging the performance of the disparate models to improve detection accuracy. Information gain of tree-based models was used to identify the key variables within the patient data that contributed to the detection of at-risk patients in each of the diseases classes by the data-learned models.
Results: The developed ensemble model for cardiovascular disease (based on 131 variables) achieved an Area Under – Receiver Operating Characteristics (AU-ROC) score of 83.1% using no laboratory results, and 83.9% accuracy with laboratory results. In diabetes classification (based on 123 variables), eXtreme Gradient Boost (XGBoost) model achieved an AU-ROC score of 86.2% (without laboratory data) and 95.7% (with laboratory data). For pre-diabetic patients, the ensemble model had the top AU-ROC score of 73.7% (without laboratory data), and for laboratory based data XGBoost performed the best at 84.4%. Top five predictors in diabetes patients were 1) waist size, 2) age, 3) self-reported weight, 4) leg length, and 5) sodium intake. For cardiovascular diseases the models identified 1) age, 2) systolic blood pressure, 3) self-reported weight, 4) occurrence of chest pain, and 5) diastolic blood pressure as key contributors.
Conclusion: We conclude machine learned models based on survey questionnaire can provide an automated identification mechanism for patients at risk of diabetes and cardiovascular diseases. We also identify key contributors to the prediction, which can be further explored for their implications on electronic health records.
As Internet-based communications have expanded, online debating has become a significant form of political participation. This work examines online discussions around health care in the United States by analysing tweets about Obamacare and then assessing the degrees of polarisation in social media. The results indicate that highly influential entities in social media have an important capacity to polarise the public. Another relevant finding is that ideology is a powerful mechanism to frame online discussions by relegating policy arguments in online debates. Finally, this work shows that social media can easily promote negative sentiments towards ‘the other’, confirming group homogeneity in online communities.
We introduce a family of authenticated data structures — Ordered Merkle Trees (OMT) — and illustrate their utility in security kernels for a wide variety of sub-systems. Specifically, the utility of two types of OMTs: a) the index ordered merkle tree (IOMT) and b) the range ordered merkle tree (ROMT), are investigated for their suitability in security kernels for various sub-systems of Border Gateway Protocol (BGP), the Internet’s inter-autonomous system routing infrastructure. We outline simple generic security kernel functions to maintain OMTs, and sub-system specific security kernel functionality for BGP sub- systems (like registries, autonomous system owners, and BGP speakers/routers), that take advantage of OMT .
South African students across numerous university campuses joined together in the second half of 2015 to protest the rising cost of higher education. In addition to on-campus protesting, activists utilized Twitter to mobilize and communicate with each other, and, as the protests drew national attention, the hashtag# FeesMustFall began trending on Twitter. Then, what began as a localized movement against tuition increases became a global issue when a court interdict was granted by a South African court against the use of the# FeesMustFall hashtag. This paper traces that global spread of the# FeesMustFall hashtag on Twitter as a response to the extraordinary attempt to limit online free speech. In this paper, we analyze the global flow and geographic spread of the# FeesMustFall hashtag on Twitter. Our evidence supports the argument that the attempt to censor and curtail the protestors’ right to organize and share the hashtag in fact propelled the# FeesMustFall movement onto the international stage.
A cloud storage assurance architecture (CSAA) for providing integrity, privacy and availability assurances regarding any cloud storage service is presented. CSAA is motivated by the fact that the complexity of components (software / hardware and personnel) that compose such a service, and lack of transparency regarding policies followed by the service makes conventional security mechanisms insufficient to provide convincing assurances to users. As it is impractical to rule out hidden undesired functionality in every component of the service, CSAA bootstraps all desired assurances from simple transformation procedures executed inside a low complexity trustworthy module; no component of the cloud storage service is trusted.
As social media tools become more popular at all levels of government, more research is needed to determine how the platforms can be used to create meaningful citizen–government collaboration. Many entities use the tools in one-way, push manners. The aim of this research is to determine if sentiment (tone) can positively influence citizen participation with government via social media. Using a systematic random sample of 125 U.S. cities, we found that positive sentiment is more likely to engender digital participation but this was not a perfect one-to-one relationship. Some cities that had an overall positive sentiment score and displayed a participatory style of social media use did not have positive citizen sentiment scores. We argue that positive tone is only one part of a successful social media interaction plan, and encourage social media managers to actively manage platforms to use activities that spur participation.
Devices participating in mobile ad hoc networks (MANET) are expected to strictly adhere to a uniform routing protocol to route data packets among themselves. Unfortunately, MANET devices, composed of untrustworthy software and hardware components, expose a large attack surface. This can be exploited by attackers to gain control over one or more devices, and wreak havoc on the MANET subnet. The approach presented in this paper to secure MANETs restricts the attack surface to a single module in MANET devices a trusted MANET module (TMM). TMMs are deliberately constrained to demand only modest memory and computational resources in the interest of further reducing the attack surface. The specific contribution of this paper is a precise characterization of simple TMM functionality suitable for any distance vector based routing protocol, to realize the broad assurance that “any node that fails to abide by the routing protocol will not be able to participate in the MANET”.