The Effects of Long COVID-19, its Severity, and the Need for Immediate Attention: Analysis of Clinical Trials and Twitter data

The coronavirus disease 2019 (COVID-19) has been declared a pandemic since March 2020 by the World Health Organisation (WHO). The infection pathway follows symptoms of fever, cough, shortness of breath, dyspnea, and severe cases that lead to hospitalization, emergency life support, and even death. Identifying the disease progression and predicting patient outcomes early, precisely predicting the possibility of long-term adverse events through effective modeling, and use of real-world data such as longitudinal clinical trial data, electronic health records data, and health insurance data are of immense importance to effective treatment, resource allocation, and prevention of severe adverse events (SAE) of grades four or five. The main goal of the study is threefold. Firstly, we raise awareness about the different clinical trials that are being conducted concurrently on Long covid-19, and how these are beneficial. Secondly, we analyze the recent tweets on Long haul covid-19 and give an overview of the sentiments of the opinion of the people. Finally, we analyze the sentiment scores and find if they are associated with the demographics of the tweeters via a negative binomial regression model. The trials were selected with long Covid-19 available in ClinicalTrials.Gov. Also, the tweets obtained with the term #long covid-19 consisted of 8436 tweets. We utilized the NRC Emotion Lexicon method for sentiment analysis is a list of English words and their associations with eight basic emotions (anger, fear, anticipation, trust, surprise, sadness, joy, and disgust) and two sentiments (negative and positive) (11). We obtained a matrix of sentiment scores, as well as retweet counts and favorite counts which were analyzed. We regressed the retweet counts and the favorite counts with the sentiment scores and find if they are associated with the emotions and sentiments of the tweeters via a negative binomial regression model since the outcome variable is count data. Our results find that there are two types of clinical trials (a total of 298) being conducted 1)observational and b) interventional. The details about enrollment, time to completion, clinical trial phases, etc., are discussed. Sentiment analysis with the NRC method of the tweets shows that there are both positive and negative sentiments. The retweet counts and favorite counts are associated with the sentiments and emotions such as disgust, joy, sadness, surprise, trust, negative, positive, etc. Finally, to conclude we need resources, and further research needs to be conducted in this area of long Covid-19.

Bayesian Shrinkage Priors in Zero-Inflated and Negative Binomial Regression models with Real World Data Applications of COVID-19 Vaccine, and RNA-Seq

BackgroundCount data regression modeling has received much attention in several science fields in which the Poisson, Negative binomial, and Zero-Inflated models are some of the primary regression techniques. Negative binomial regression is applied to modeling count variables, usually when they are over-dispersed. A Poisson distribution is also utilized for counting data where the mean is equal to the variance. This situation is often unrealistic since the distribution of counts will usually have a variance that is not equal to its mean. Modeling it as Poisson distributed leads to ignoring under- or overdispersion, depending on if the variance is smaller or larger than the mean. Also, situations with outcomes having a larger number of zeros such as RNASeq data require Zero-inflated models. Variable selection through shrinkage priors has been a popular method to address the curse of dimensionality and achieve the identification of significant variables. MethodsWe present a unified Bayesian hierarchical framework that implements and compares shrinkage priors in negative-binomial and zero-inflated negative binomial regression models. The key feature is the representation of the likelihood by a Polya-Gamma data augmentation, which admits a natural integration with a family of shrinkage priors. We specifically focus on the Horseshoe, Dirichlet Laplace, and Double Pareto priors. Extensive simulation studies address the efficiency of the model and mean square errors are reported. Further, the models are applied to data sets such as the Covid-19 vaccine, and Covid-19 RNA-Seq data among others. ResultsThe models are robust enough to address variable selection, and MSE decreases as the sample size increases, having lower errors in p > n cases. The noteworthy results showed that the adverse events of Covid-19 vaccines were dependent on age, recovery, medical history, and prior vaccination with a remarkable reduction in MSE of the fitted values. No. of publications of Ph.D. students were dependent on the no. of children, and the no. of articles in the last three years. ConclusionsThe models are robust enough to conduct both variable selections and produce effective fit because of their high shrinkage property and applicability to a broad range of biometric and public health high dimensional problems.

Association of Covid-19 Vaccine Hesitancy with Demographics, Mental Health, and Disability

BackgroundThe world is witnessing a pandemic caused by the novel coronavirus named Covid-19 by WHO that has claimed millions of lives since its advent in December 2019. Several vaccine candidates and treatments have emerged to mitigate the effect of virus, along with came an increased confusion, mistrust on their development, emergency authorization and approval process. Increased job losses, jump in divorce rate, and the generic nature of staying home has also led to various mental health issues. MethodsWe analyzed two publicly available datasets to better understand vaccine hesitancy. The first dataset was extracted from ICPSR Covid-19 database (https://doi.org/10.3886/E130422V1).[1].This cross-sectional survey was conducted to assess the prevalence of vaccine hesitancy in the US, India, and China. The second dataset was obtained from the United States Census Bureaus Household Pulse Survey (HPS) Phase 3.2. For the ICPSR dataset, proportions and summary statistics are reported to give an overview of the global picture of vaccine hesitancy. The HPS dataset was analyzed using multinomial and binary logistic regression. Chi-square test of independence and exploratory data analysis supplemented provided insight into the casual factors involved in vaccine hesitancy. ResultsO_ST_ABSICPSR Global DataC_ST_ABSFor India, 1761 participants completed the survey as of November, 2020 of which 90.2% indicated acceptance of a Covid-19 vaccine. 66.4% are parents of 18 years old or younger, and 79.0% respondent has a parent 50 years or older. Vaccine acceptance rate was 99.8% among 928 out of 1761 participants who had a child. 1392 participants either had a parent or child of which 83.4% will encourage their parents and 90.5% will encourage their children to get the covid-19 vaccine. In this Indian survey, 16.2% identified as belonging to the rural population of which 51.2% showed vaccine hesitancy. A binary logistic regression model with vaccine hesitancy as a dichotomous variable showed that rural population had an odds ratio (OR) of 3.45 (p-value<0.05). Income seems to influence vaccine hesitancy, with income level of (7501-15,000 Indian Rupees (INR)/month) having an OR of 1.41 as compared to other income groups. In the US, 1768 individuals participated in the survey from August-November 2020. 67.3% respondents indicated the will to accept the vaccine. 1129 of them either had a parent or a child, of which 67.6% will take the vaccine; 66% will encourage their parents and 83% will encourage their children for taking the vaccination. 40.3% responded as vaccine hesitant, 31% identified as staying in rural areas, of which 52.5% are vaccine hesitant. In the binary logistic regression analysis, race, past flu shot history, rural living, income turned out to be significant. White race had OR >1 as compared to other races, low-income group (US dollar $2000-4999/month) had an OR of 1.03. In China, there were 1727 participants, of which 1551(90.0%) indicated that they will accept a vaccine. 90.1% of them who had either a parent or child will accept vaccine, 80.4% will influence parents, and 83.4% will encourage children to get vaccination needle in the arm. 30% had vaccine hesitancy. 262 belonged to the rural population, of which 34.8% are vaccine hesitant. Income and Northern region (OR = 3.17) were significant in saying "yes" to a vaccine. High income groups were least resistant (OR=0.96) as compared to other groups. HPS USA dataData used in this study was collected from United States Census Bureaus Household Pulse Survey (HPS) Phase 3.2 Weeks 34-39, which covers data collected from July 21, 2021, to October 11, 2021. The HPS data helped to understand the effect of several demographic and psychological, and health-related factors upon which responses were provided, thus helping to understand the social and economic effects during the COVID-19 pandemic. ConclusionAmong the three countries, it appears based on this survey that US has the highest rate of vaccine hesitancy. may contribute towards this result gender, education, religious beliefs, disbelief in science, government which remains unexplored due to data limitation.

Theta autoregressive neural network model for COVID-19 outbreak predictions

An unprecedented outbreak of the novel coronavirus (COVID-19) in the form of peculiar pneumonia has spread globally since its first case in Wuhan province, China, in December 2019. Soon after, the infected cases and mortality increased rapidly. The future of the pandemics progress was uncertain, and thus, predicting it became crucial for public health researchers. These future predictions help the effective allocation of health care resources, stockpiling, and help in strategic planning for clinicians, government authorities, and public health policymakers after understanding the extent of the effect. The main objective of this paper is to develop a hybrid forecasting model that can generate real-time out-of-sample forecasts of COVID-19 outbreaks for five profoundly affected countries, namely the USA, Brazil, India, UK, and Canada. A novel hybrid approach based on the Theta method and Autoregressive neural network (ARNN) model, named Theta-ARNN (TARNN) model, is developed. Daily new cases of COVID-19 are nonlinear, non-stationary, and volatile; thus a single specific model cannot be ideal for future prediction of the pandemic. However, the newly introduced hybrid forecasting model with an acceptable prediction error rate can help healthcare and government for effective planning and resource allocation. The proposed method outperforms traditional univariate and hybrid forecasting models for the test data sets on an average.