Hospital admission risk stratification of patients with gout presenting to the emergency department

To characterise gout patients at high risk of hospitalisation and to develop a web-based prognostic model to predict the likelihood of gout-related hospital admissions. This was a retrospective single-centre study of 1417 patients presenting to the emergency department (ED) with a gout flare between 2015 and 2017 with a 1-year look-back period. The dataset was randomly divided, with 80% forming the derivation and the remaining forming the validation cohort. A multivariable logistic regression model was used to determine the likelihood of hospitalisation from a gout flare in the derivation cohort. The coefficients for the variables with statistically significant adjusted odds ratios were used for the development of a web-based hospitalisation risk estimator. The performance of this risk estimator model was assessed via the area under the receiver operating characteristic curve (AUROC), calibration plot, and brier score. Patients who were hospitalised with gout tended to be older, less likely male, more likely to have had a previous hospital stay with an inpatient primary diagnosis of gout, or a previous ED visit for gout, less likely to have been prescribed standby acute gout therapy, and had a significant burden of comorbidities. In the multivariable-adjusted analyses, previous hospitalisation for gout was associated with the highest odds of gout-related admission. Early identification of patients with a high likelihood of gout-related hospitalisation using our web-based validated risk estimator model may assist to target resources to the highest risk individuals, reducing the frequency of gout-related admissions and improving the overall health-related quality of life in the long term. Key points • We reported the characteristics of gout patients visiting a tertiary hospital in Singapore. • We developed a web-based prognostic model with non-invasive variables to predict the likelihood of gout-relatedhospital admissions.

SERIES

Requirement defects are one of the major sources of failure in any software development process, and the main objective of this chapter is to make requirement analysis phase exhaustive by estimating risk at requirement level by analyzing requirement defect and requirement inter-relationships as early as possible to using domain modeling to inhibit them from being incorporated in design and implementation. To achieve this objective, this chapter proposes a tool to assist software developers in assessing risk at requirement level. The proposed tool, software risk estimator, SERIES in short, helps in early identification of potential risk where preventive actions can be undertaken to mitigate risk and corrective actions to avoid project failure in collaborative manner. The entire process has been supported by a software case study. The results of the proposed work are promising and will help software engineers in ensuring that all business requirements are captured correctly with clear vision and scope.

Cardiovascular Risk Awareness and Calculated 10-Year Risk Among Female Employees at Taibah University 2019

Cardiovascular diseases (CVD) are the most common cause of death and disability worldwide. Saudi Arabia, one of the middle-income countries has a proportional CVD mortality rate of 37%. Knowledge about CVD and its modifiable risk factors is a vital pre-requisite to change the health attitudes, behaviors, and lifestyle practices of individuals. Therefore, we intended to assess the employee knowledge about risk of CVD, symptoms of heart attacks, and stroke, and to calculate their future 10-years CVD risk. An epidemiological, cross-sectional, community-facility based study was conducted. The women aged ≥40 years who are employees of Taibah University, Al-Madinah Al-Munawarah were recruited. A screening self-administrative questionnaire was distributed to the women to exclude those who are not eligible. In total, 222 women met the inclusion criteria and were invited for the next step for the determination of CVD risk factors by using WHO STEPS questionnaire: It is used for the surveillance of non-communicable disease risk factor, such as CVD. In addition, the anthropometric measurements and biochemical measurements were done. Based on the identified atherosclerotic cardiovascular disease (ASCVD) risk factors and laboratory testing results, risk calculated used the Framingham Study Cardiovascular Disease (10-year) Risk Assessment. Data were analyzed using GraphPad Prism 7 software (GraphPad Software, CA, USA). The result showed the mean age of study sample was 55.6 ± 9.0 years. There was elevated percentage of obesity and rise in abdominal circumference among the women. Hypertension (HTN) was a considerable chronic disease among the participants where more than half of the sample had it, i.e., 53%. According to the ASCVD risk estimator, the study participants were distributed into four groups: 63.1% at low risk, 20.2% at borderline risk, 13.5% at intermediate risk, and 3.2% at high risk. A comparison between these categories based on the CVD 10-year risk estimator indicated that there were significant variations between the low-risk group and the intermediate and high-risk groups (P = 0.02 and P = 0.001, respectively). The multivariate analysis detected factors related to CVD risk for women who have an intermediate or high risk of CVD, such as age, smoking, body mass index (BMI), unhealthy diet, blood pressure (BP) measurements, and family history of CVD (P < 0.05). The present study reports limited knowledge and awareness of CVD was 8.6 that is considered as low knowledge. In conclusion, the present study among the university sample in Madinah reported limited knowledge and awareness of CVD risk. These findings support the need for an educational program to enhance the awareness of risk factors and prevention of CVD.

The Facility Infection Risk Estimator™: A web application tool for comparing indoor risk mitigation strategies by estimating airborne transmission risk

The COVID-19 pandemic created needs for (a) estimating the existing airborne risk of infection from SARS-CoV-2 in existing facilities and new designs and (b) estimating and comparing the impacts of engineering and behavioural strategies for contextually reducing that risk. This paper presents the development of a web application to meet these needs, the Facility Infection Risk Estimator™, and its underlying Wells–Riley based model. The model specifically estimates (a) the removal efficiencies of various settling, ventilation, filtration and virus inactivation strategies and (b) the associated probability of infection, given the room physical parameters and number of individuals infected present with either influenza or SARS-CoV-2. A review of the underlying calculations and associated literature is provided, along with the model's validation against two documented spreading events. The error between modelled and actual number of additional people infected, normalized by the number of uninfected people present, ranged from roughly –18.4% to +9.7%. The more certain one can be regarding the input parameters (such as for new designs or existing buildings with adequate field verification), the smaller these normalized errors will be, likely less than ±15%, making it useful for comparing the impacts of different risk mitigation strategies focused on airborne transmission.

Learning from Complementary Labels via Partial-Output Consistency Regularization

In complementary-label learning (CLL), a multi-class classifier is learned from training instances each associated with complementary labels, which specify the classes that the instance does not belong to. Previous studies focus on unbiased risk estimator or surrogate loss while neglect the importance of regularization in training phase. In this paper, we give the first attempt to leverage regularization techniques for CLL. By decoupling a label vector into complementary labels and partial unknown labels, we simultaneously inhibit the outputs of complementary labels with a complementary loss and penalize the sensitivity of the classifier on the partial outputs of these unknown classes by consistency regularization. Then we unify the complementary loss and consistency loss together by a specially designed dynamic weighting factor. We conduct a series of experiments showing that the proposed method achieves highly competitive performance in CLL.

Evaluation of MRI Denoising Methods Using Unsupervised Learning

In this paper we evaluate two unsupervised approaches to denoise Magnetic Resonance Images (MRI) in the complex image space using the raw information that k-space holds. The first method is based on Stein’s Unbiased Risk Estimator, while the second approach is based on a blindspot network, which limits the network’s receptive field. Both methods are tested on two different datasets, one containing real knee MRI and the other consists of synthetic brain MRI. These datasets contain information about the complex image space which will be used for denoising purposes. Both networks are compared against a state-of-the-art algorithm, Non-Local Means (NLM) using quantitative and qualitative measures. For most given metrics and qualitative measures, both networks outperformed NLM, and they prove to be reliable denoising methods.

Is Hidden Safe? Location Protection against Machine-Learning Prediction Attacks in Social Networks

User privacy protection is a vital issue of concern for online social networks (OSNs). Even though users often intentionally hide their private information in OSNs, since adversaries may conduct prediction attacks to predict hidden information using advanced machine learning techniques, private information that users intend to hide may still be at risk of being exposed. Taking the current city listed on Facebook profiles as a case, we propose a solution that estimates and manages the exposure risk of users’ hidden information. First, we simulate an aggressive prediction attack using advanced state-of-the-art machine learning algorithms by proposing a new current city prediction framework that integrates location indications based on various types of information exposed by users, including demographic attributes, behaviors, and relationships. Second, we study prediction attack results to model patterns of prediction correctness (as correct predictions lead to information exposures) and construct an exposure risk estimator. The proposed exposure risk estimator has the ability not only to notify users of exposure risks related to their hidden current city but can also help users mitigate exposure risks by overhauling and selecting countermeasures. Moreover, our exposure risk estimator can improve the privacy management of OSNs by facilitating empirical studies on the exposure risks of OSN users as a group. Taking the current city as a case, this work offers insight on how to protect other types of private information against machine-learning prediction attacks and reveals several important implications for both practice management and future research.