Model Development for Predicting the Crude Oil Price: Comparative Evaluation of Ensemble and Machine Learning Methods

The crude oil market is unstable, and its price is highly volatile. Due to the Covid-19 pandemic, the price of crude oils goes up and down in a short period of time. Future plans and projects' policies depend directly and indirectly on the future price of crude oil. So, the aim of this study is to predict the price of crude oil by using machine learning and ensemble algorithm, as well as to show the comparison of performance of Ada Boost, Bagging Lasso and Support Vector Regression model. The study uses crude oil price time series data for analysis and to form a model to predict future price. The actual vs. predicted curve is used to show the performance of each algorithm individually. Analysis shows that the ensemble AdaBoost algorithm displays better performance than other algorithms. The result is validated using mean square error (MSE), root mean square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), two accuracy score function variance score, and R2 score. This study will help the stakeholders of the crude oil industry in making decisions and formulating policies based on forecasted crude oil prices. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

Analysing Time-frequency Relationship between Oil price and Sectoral Indices in India using Wavelet Techniques

Oil is considered an essential factor of any economy. This paper examines the time-varying correlation between oil price return, BSE SENSEX, and 14 sectoral indexes in India using multiscale wavelet decomposition and wavelet coherence analysis. The maximal wavelet discrete wavelet transform analysis shows a feedback relationship between 13 sectors at higher time horizons (dC4, dC5, and dC6). Based on the wavelet coherence plot, the oil price and sectoral index return show a high co-movement at 32 to 128 s. The wavelet coherence plot shows that the oil price and sectoral index return show a high co-movement during the period of Mar-May 2020 (especially during the period of financial crisis widely spread due to the COVID19 pandemic and nationwide lockdown notification announced by the Government of India). We discuss the implications of these studies in detail. © 2022, Econjournals. All rights reserved.

Compositional modelling of immune response and virus transmission dynamics.

Transmission models for infectious diseases are typically formulated in terms of dynamics between individuals or groups with processes such as disease progression or recovery for each individual captured phenomenologically, without reference to underlying biological processes. Furthermore, the construction of these models is often monolithic: they do not allow one to readily modify the processes involved or include the new ones, or to combine models at different scales. We show how to construct a simple model of immune response to a respiratory virus and a model of transmission using an easily modifiable set of rules allowing further refining and merging the two models together. The immune response model reproduces the expected response curve of PCR testing for COVID-19 and implies a long-tailed distribution of infectiousness reflective of individual heterogeneity. This immune response model, when combined with a transmission model, reproduces the previously reported shift in the population distribution of viral loads along an epidemic trajectory. This article is part of the theme issue 'Technical challenges of modelling real-life epidemics and examples of overcoming these'.

DSP Gene Mutation – A Cause of Recurrent Myocarditis: A Case Series

Myocarditis has gained clinical awareness with the current COVID-19 pandemic. A recent ESC Expert Consensus document discussing management of acute myocarditis and chronic inflammatory cardiomyopathy has been published [1]. Although the document alludes to genetic predisposition, by stating that “patients with mutations responsible for arrhythmogenic cardiomyopathy may be at risk for acute myocarditis,” the growing clinical experience in this area suggests that perhaps pursuit of an inflammatory diagnosis has been at the cost of recognising an underlying genetic cause, with important implications for the patient and their family. Desmoplakin (DSP)-related arrhythmogenic cardiomyopathy (AC) is characterised by LV systolic dysfunction, subepicardial late gadolinium enhancement on cardiac magnetic resonance imaging (CMR) and frequent ventricular ectopy [2,3]. In ∼15% of DSP-related AC involving the LV, a clinical event indistinguishable from myocarditis can be the initial manifestation [4]. Myocarditis often precedes fibrosis and LV dysfunction. In this case series, we describe 3 unrelated young individuals (aged 21–28 years) diagnosed by our service with DSP pathogenic variants. In all cases the initial diagnosis was myocarditis and the diagnostic odyssey was characterised by multiple presentations with significant troponin elevations, including one in whom chronic inflammatory myocarditis was the diagnosis over many years. Since the COVID-19 pandemic, presentations for myocarditis have increased considerably [5]. This has increased the importance of the clinician considering genetic arrhythmogenic cardiomyopathy as an alternative diagnosis. Any case of myocarditis should prompt 3-generational family history, and recurrent presenters should have a CMR and be referred for consideration of genetic testing.

COMMUNITY-ENGAGED COVID-19 DISASTER SOLUTIONS (CCDS): COLLABORATIVE ACTION RESEARCH TO SUPPORT PSYCHO-SOCIALECOLOGICAL STRESSORS AND PROTECTIVE FACTORS AFFECTING THE HEALTH OF INDIGENOUS, VISIBLE MINORITIES, AND REFUGEE COMMUNITIES

This project is a community-led collaboration between an interdisciplinary research team (including Indigenous and visible minority academics, health professionals and students working in health, and community disaster researchers) and First Nation leadership and community members. The project addresses two critical issues that affect Indigenous, visible minorities, and refugee communities: negatively impact their health and poor access to healthcare. We will examine how ethnic and cultural identity, protective factors, and psycho-social stresses impact this pandemic. Given the growing ethnically diverse population of Alberta and Saskatchewan, Canada, it is essential to understand how these communities view, use, and experience health services to build their health resiliency. The findings will provide a high reward policy and programming recommendations to improve health services and deliver equitable, quality and ethnically conscious care during the COVID-19 disaster. Following relational, Indigenous, and antiracist theoretical frameworks, we will use a mixed-method approach of self-reported surveys, focus groups, individual interviews, and Indigenous story-sharing to collect data from the Indigenous, visible minorities, and refugee communities' overall health, factors that negatively impact their health, how they cope with adversity, and their usage of health services.

Detection of Pneumococcal Pneumonia during SARS-CoV-2 Infection

Background. Streptococcus pneumoniae (pneumococcus) is a common colonizer of the upper respiratory tract and can progress to cause invasive and mucosal disease. Additionally, infection with pneumococcus can complicate respiratory viral infections (influenza, respiratory syncytial virus, etc.) by exacerbating the initial disease. Limited data exist describing the potential relationship of SARS-CoV-2 infection with pneumococcus and the role of co-infection in influencing COVID-19 severity. Methods. Inpatients and healthcare workers testing positive for SARS-CoV-2 during March-August 2020 were tested for pneumococcus through culture-enrichment of saliva followed by RT-qPCR (to identify carriage) and for inpatients only, serotype-specific urine antigen detection (UAD) assays (to identify pneumococcal pneumonia). A multinomial multivariate regression model was used to examine the relationship between pneumococcal detection and COVID-19 severity. Results. Among the 126 subjects who tested positive for SARS-CoV-2, the median age was 62 years;54.9% of subjects were male;88.89% were inpatients;23.5% had an ICU stay;and 13.5% died. Pneumococcus was detected in 17 subjects (13.5%) by any method, including 5 subjects (4.0%) by RT-qPCR and 12 subjects (13.6%) by UAD. Little to no bacterial growth was observed on 21/235 culture plates. Detection by UAD was associated with both moderate and severe COVID-19 disease while RT-qPCR detection in saliva was not associated with severity. None of the 12 individuals who were UAD-positive died. Conclusion. Pneumococcal pneumonia (as determined by UAD) continues to occur during the ongoing pandemic and may be associated with more serious COVID-19 outcomes. Detection of pneumococcal carriage may be masked by high levels of antibiotic use. Future studies should better characterize the relationship between pneumococcus and SARS-CoV-2 across all disease severity levels.

Subcutaneous Sarilumab for the Treatment of Hospitalized patients with Moderate to Severe COVID19 Disease: A Pragmatic, Embedded, Multi-Center Randomized Clinical Trial

Background. The aim of this pragmatic, embedded adaptive trial was to measure the effectiveness of subcutaneous sarilumab in addition to an evolving standard of care for clinical management of inpatients with moderate to severe COVID-19 disease (NCT04359901). The study is also a real-world demonstration of the realization of a prospective learning healthcare system. Methods. Two-arm, randomized, open-label controlled 5-center trial comparing standard care alone to standard care (SOC), which evolved over time, with addition of subcutaneous sarilumab (200 mg or 400 mg anti-IL6R) among hospitalized patients with moderate to severe COVID-19 not requiring mechanical ventilation. The primary outcome was 14-day incidence of intubation or death. The trial used a randomized play-the-winner design and was fully embedded within the EHR system, including the adaptive randomization process. Results. Among 417 patients screened, 162 were eligible based on chart review, 53 consented, and 50 were evaluated for the primary endpoint of intubation or death ( >30% of eligible patients enrolled) (Figure 1). After the second interim review, the unblinded Data Monitoring Committee recommended that the study be stopped due to concern for safety: a high probability that rates of intubation or death were higher with addition of sarilumab to SOC (92.6%), and a very low probability (3.4%) that sarilumab would be found to be superior. Conclusion. This randomized trial of patients hospitalized with COVID-19 and requiring supplemental oxygen but not mechanical ventilation found no evidence of benefit from subcutaneous sarilumab in addition to an evolving standard-of-care. The numbers of patients and events were too low to allow independent conclusions to be drawn, but this study contributes valuable information about the role of subcutaneous IL-6 inhibition in the treatment of patients hospitalized with COVID-19. The major innovation of this trial was the advancement of embedded, point-of-care clinical trials for FDA-approved drugs;this represents a realization of the learning healthcare system. Methods developed and piloted during the conduct of this trial can be used in future investigations to speed the advancement of clinical science.

SoftHealth: Softwarized 5G-Driven Network Slicing for Real-time e-Healthcare Applications using ML

The recent years have whiteness the substandard situations of the modern healthcare system due to a fatal pandemic called COVID19. The rapid advancements of modern technology have disseminated the superficial benefits of medical infrastructure, but significant improvements are still extremely necessary over the massive e-healthcare system (mHS). Considering the fact of limited resources and unlimited demands, a highly stable end-to-end optimization model is required. Healthcare also struggles with real-time communication. The next-generation communication networks (e.g 5G and beyond) proficiently influence the network resource distribution for URLLC. In this work, we have envisioned a novel on-demand e-Healthcare dynamic network slice architecture that uses the ML algorithms at the edge server for real-time classification and access of the offloaded data from the central controller (vSDN-Control layer to Data plane layer). The comparative analysis over the datasets of patients consisting of special index parameters shows that our proposed model allows the end-user more efficient data accessibility over the conventional approaches. We have studied the model over the multi-classification ML models (kNN, DT and RF) and we have found an average improvement of 10% to 15% of average data offloading time efficiency from the local machines from the edge servers. This approach can be further extended as the QoS improvement of the healthcare data traffic over the dynamic network slice instances. We have kept the model simple but standard in nature. © 2022 IEEE.

Machine learning to identify symptoms associated with coronavirus disease 2019 (covid-19) in a national cohort of veterans

Rationale: Machine learning may augment conventional epidemiologic methods of detecting COVID-19. Prior studies of machine learning to screen for COVID-19 have relied largely upon radiographic and laboratory data. Few have used clinical symptoms alone to screen for COVID-19, and most are limited by sample size. We adapted an existing, open source machine learning pipeline to identify symptoms associated with COVID-19 using a large national cohort of Veterans with polymerase chain reaction (PCR)-confirmed COVID-19. Methods: We included members of the Veterans Affairs (VA) Birth Cohort (n=4,221,135) who had PCR-confirmed COVID- 19 between 01/01/2020-06/15/2020. Veterans with an outpatient VA visit from 03/01/2018-02/29/2020 were analyzed. To identify symptoms associated with COVID-19, we adapted the open source v3NLP platform. This symptom pipeline consists of several features (Figure) and was restricted to identify fever, cough, anosmia/ageusia, dyspnea, congestion, sore throat, diarrhea, nausea, loss of appetite, headache, dizziness, chills, confusion/altered mental status, myalgia, and fatigue. Before running this pipeline, a multidisciplinary team of 9 clinicians first developed an annotation guideline. Each will use the guideline to review of 100 random notes that include information about symptoms to evaluate interrater agreement. This guideline will be revised until a kappa of 0.85 for inter-rater reliability is achieved. The multidisciplinary team will then review a random subset of 900 notes to assess agreement between pipeline- and human-identified symptoms. Each clinician will review and annotate 150 notes so that each note is reviewed by at least 2 clinicians. We will run the symptom pipeline on text from clinical notes of Veterans with PCR-confirmed COVID-19 and evaluate the performance of symptom identification with snippet annotation. Results: The sample included all clinical notes (n=1,641,935) identified two weeks before through two weeks after a PCR-confirmed diagnosis of COVID-19. Initial testing found the opensource symptom pipeline achieved an F-measure (harmonic mean of positive predictive value and sensitivity) of 0.71 to detect symptoms and an overall F-measure of 0.87 indicating excellent performance when combined with detecting negative symptoms. Conclusions: This work demonstrates the method by which an existing open source machine learning pipeline can be adapted to identify symptoms associated with COVID-19. This work has broad applications and is distinguished by its focus on symptoms commonly reported in clinical notes. Future studies may consider applying this methodology to screen for symptoms in populations of interest (e.g., persons with prior COVID-19 or immunization against COVID-19) or to develop an electronic medical recordbased syndromic surveillance system.

Clinical and epidemiological features of healthcare workers detected with coronavirus disease

Background: Data early in the SARS-CoV-2 pandemic suggested frontline healthcare workers (HCW) may account for 10-20% of all infections. CDC estimated 600,000 infections in HCWs. Symptom screening is a strategy to prevent healthcare-associated transmission. This method may not identify asymptomatic or pre-symptomatic carriers. Methods: We conducted a prospective cohort study in asymptomatic or minimally symptomatic healthcare workers in a 1541-bed academic medical center. Although recruitment began in designated COVID-19 units, we expanded to all HCWs providing care to hospitalized patients during the pandemic. Data was gathered on demographics, work area in the hospital and daily questionnaires were sent listing symptoms of SARS-CoV-2. Protocol included twice weekly self-collected nasopharyngeal swab and saliva for SARS-CoV-2 N1 and N2.Those with positive PCR result, underwent telephone survey to assess symptomatology and severity of illness. Results: A total 525 HCWs began the study protocol and 16 were identified as PCR positive. Samples included concordant saliva and NP samples on 9 (56%), exclusively NP samples on 5 (31%) and 2 (12%) HCWs were positive by saliva PCR only. Majority were female, and all were nursing staff;with 19% reported not working in a designated COVID-19 unit. During the course of this active surveillance, universal masking was mandated in the institution. Rhinorrhea and headache were reported by 6 (38%), 5 (31%) reported cough and 3 (19%) developed myalgia. Changes in smell and taste preceded the positive PCR test in 2 (12%). One HCW reported developing a fever with acute illness. All were notified about their PCR positive status by institution's occupational health department and self-isolated to monitor for symptoms. Conclusion: The spectrum of disease in this HCW cohort is similar to mild disease in the community. Due to high incidence of asymptomatic or mildly symptomatic HCWs, active surveillance with routine testing proves be beneficial to prevent hospital transmission of SARS-CoV-2. Universal masking significantly decreased the HCW positive rate in our study, underscoring the need for universal efforts to mitigate healthcare-associated transmission with self-monitoring, face mask use, and other infection prevention behaviors like hand hygiene.

Implementing an at-home smell test for early assessment of COVID-19 in high-risk healthcare workers

Background: Smell loss has been recognized as an important, and potentially early, sign of COVID-19. However, to date smell loss has only been assessed in retrospective, COVID+ cohorts, and largely through self-report. The objective of this study was to implement a daily standardized behavioral test of smell sensitivity in healthcare workers (HCW) to capture changes in smell sensitivity over time and to assess whether these changes occur prior to positive COVID test. Methods: The study enrolled 500 high-risk COVID-negative HCW during the COVID-19 epidemic in Connecticut, beginning March 28, 2020 (80% F, mean age 38, 58% nurses). Initially, HCW received a daily symptom questionnaire with parosmia screening questions. On April 23 we introduced the “Jiffy”, a daily at-home psychophysical test of smell sensitivity, where olfactory stimuli are sampled and rated for perceived intensity. SARS-CoV-2 infection was tested every three days by PCR of nasopharyngeal swabs or saliva Results: Of the first 500 enrolled HCW, 376 HCW (75%) completed the Jiffy 4528 times (mean 12 times/HCW). 17/500 HCW (3.4%) had a COVID+ test, of which 9/17 (53%) reported smell loss through the Jiffy or the daily symptom survey. 6/9 (67%) reported smell loss that preceded or was concurrent with a COVID+ test. 8/17 COVID+ HCW completed the Jiffy, with 5/8 (63%) reporting reductions in smell versus 42/368 (11%) COVID- HCW (OR=13, 95% CI: 2.4-85, p=.001). COVID+ HCW rated their greatest reduction in smell sensitivity as slight (40%) and severe (60%), versus slight (88%) and moderate (12%) in COVID- HCW. 16/17 COVID+ HCW completed a daily symptom survey (mean 14 times/HCW), with 8/16 (50%) ever reporting parosmia versus 90/466 (19%) of COVID- HCW (OR=4.2, 95% CI: 1.3-13, p=.007). Overall, parosmia was the first reported symptom in 3/13 (23%) COVID+ HCW who reported symptoms. Conclusion: We conducted a prospective study of smell testing in a population at high risk for COVID-19 using two parallel approaches. Our results demonstrate the feasibility of at-home smell testing for assessing parosmia during COVID-19, in some cases even prior to a positive PCR result. Given the urgent need for widespread, lowcost, non-invasive testing for COVID-19, we are now developing an easy-to-use app to distribute this survey more widely to high-risk populations. (Table Presented).

Active monitoring of a healthcare worker cohort during the COVID-19 epidemic

Background: Initial CDC recommendations for passive monitoring of COVID-19 related symptoms among staff may not be sufficient in preventing the introduction and transmission of SARS-CoV-2 in healthcare settings. We therefore implemented active monitoring for SARS-CoV-2 infection in healthcare workers (HCWs) at an academic medical center during the COVID-19 epidemic in northeast US. Methods: We recruited a cohort of HCWs at Yale New Haven Hospital who worked in COVID-19 units and did not have COVID-19 related symptoms between March 28 and June 1, 2020. During follow-up, participants provided daily information on symptoms by responding to a web-based questionnaire, self-administered nasopharyngeal (NP) and saliva specimens every 3 days, and blood specimens every 14 days. We performed SARS-CoV-2 RT-PCR and an anti-spike protein IgM and IgG ELISA to identify virological and serological-confirmed infection, respectively. Results: We enrolled 525 (13%) amongst 4,136 HCW of whom daily information on symptoms and NP, saliva, and blood specimens were obtained for 66% (of 13208), 42% (or 1977), 44% (of 2071) and 65% (of 1099), respectively, of the follow-up measurement points. We identified 16 (3.0% of 525) HCWs with PCR-confirmed SARS-CoV-2 infection and an additional 12 (2.3% of 525) who were not tested by PCR or had negative PCR results but had serological evidence of infection. The overall cumulative incidence of SARS-CoV-2 infection was 5.3% (28 of 525) amongst HCWs. Cases were not identified by hospital protocols for passive staff self-monitoring for symptoms. Amongst 16 PCR-confirmed cases, 9 (56%) of the 16 PCR-confirmed HCW had symptoms during or after the date of initial detection. We did not identify an epidemiological link between the 28 confirmed cases. Conclusion: We found that a significant proportion (5.3%) of HCWs were infected with SARS-CoV-2 during the COVID-19 epidemic. In the setting of universal PPE use, infections were possibly acquired in the community rather than stemming from patient-HCW or HCW-HCW transmission. Passive monitoring of symptoms is inadequate in preventing introductions of SARS-CoV-2 into the healthcare setting due to asymptomatic and oligosymptomatic presentations.

Repurposing mefenamic acid in the management of covid-19

With the growing understanding of coronavirus disease-2019 (COVID-19) pathogenesis, different therapeutic targets are being considered for the management of COVID-19. The development of new drugs is a time-consuming process;hence, many drugs acting on similar therapeutic targets/sites in the COVID-19 treatment are repurposed in COVID-19. In this article, an expert panel deliberated on the existing evidence on the immunopathogenesis, therapeutic targets under consideration for treatment of COVID-19, and the place of mefenamic acid in the therapy landscape of COVID-19. The expert panel has also provided recommendations regarding the dose and regimen of mefenamic acid in different phases of the COVID-19 disease.

The COVID-19 Pandemic: An Immigrant Family Story on Reconnection, Resistance, and Resiliency

The COVID-19 pandemic has created a significant effect on the vulnerable portion of society, particularly on Indigenous and visible minority immigrants. We, as a minority family from Bangladesh who are on Indigenous land in Saskatchewan Canada, explore family-based pandemic resiliency, mainly focusing on Indigenous notions of resistance and reconnection. This article discusses our family-based resiliency on family interaction, social distancing, and isolation during the COVID-19 pandemic. This paper explores a family-based decolonizing autoethnography as a methodology for understanding health and wellness from an immigrant family's perspective. We discussed why Indigenous and immigrant stories matters for building resiliency and resistance within a family. How do we know it is effective? How can it be helpful for others? Here, we highlight how Indigenous Elders, Knowledge-Keepers, and ancestors' stories helped us for building our resistance and reconnection to be active, hopeful, and joyful during the COVID-19 pandemic.