N3C CTR DURs

We will perform a retrospective analysis of COVID-19 data that includes confirmed cases and persons under investigation (PUI) from all MaineHealth institutions to understand demographic distribution, spectrum of disease severity as well as hospitalization and mortality rates. \nIn the state of Maine, over 31,000 people have been infected with SARS-CoV2-virus and more than 453 people have died. Th e counties in Maine served by a MaineHealth member hospital or clinic include: Cumberland, York, Oxford, Franklin, Sagadahoc, Lincoln, Knox and Waldo, and Aroostook, which combined account for 66% of reported cases in Maine according to Maine CDC data as of January 13, 2021.\nThe risks of serious disease and death in COVID-19 cases increases with advancing age and presence of comorbid health conditions such as chronic obstructive pulmonary disease (COPD), diabetes mellitus, heart failure immunosuppressive states such as post solid organ transplant patients. To prevent the spread and death from this virus, identifying symptoms early, performing tests to detect cases, and isolating infected individual has been critical. National and local data have shown that COVID-19 disproportionately affects the elderly, those living in congregate care settings, and racial and ethnic minorities. For those of us in the MaineHealth system, knowing how this condition has presented in our patient population, what the demographic profile, medical complexity and service utilization has been will help us to focus our educational efforts for staff, learners and community. There has been no comprehensive descriptive report of the first six months of the pandemic and we aim to describe MaineHealth?s early experiences. \nWe plan to perform a retrospective analysis of all patients with identified with COVID-19 across the MaineHealth system submitted to the N3C from January 2020 through December 31, 2020. \nOur specific Aims include: 1. Analyze N3C data available to MaineHealth for COVID-19 from January 1 ? December 31, 2020, including demographic, medical diagnostic and service utilization data. 2. Perform descriptive data analyses of COVID-19 cases by geographic area. 3. Describe hospitalization, disease severity and mortality rates among confirmed cases by geographic area.\nOur study will add knowledge about COVID-19 across the MaineHealth system. By analyzing and summarizing the demographic distribution of COVID-19 infection, understanding the spectrum of disease severity, as well as having the knowledge of hospitalization and mortality rates, we will learn which segment of the population is affected and who is at risk of developing severe disease and dying from COVID-19 in the MaineHealth service area.

Heavy alcohol use (>4 drinks/day for women and >5 drinks/day for men) is a known risk factor for bacterial pneumonia and the development of ARDS. Certain respiratory viruses such as RSV are also more severe in those that drink heavily. It is not currently known whether alcohol use increases the susceptibility to infection with SARS-CoV2 or worsens disease severity.

We aim to investigate IgG antibody response to SARS-CoV-2 in patients with multiple sclerosis who are being treated with B-cell depleting therapies such as rituximab and ocrelizumab given B-cell deleting therapies may impair or blunt antibody response to SARS-CoV-2 infection and impact future vaccine readiness.

This project will compare comorbidity rates across both the COVID-19 negative and COVID-19 positive patients in N3C. The utility\nof this analysis is to determine if comorbidity incidence, which will facilitate more trips to the hospital for routine care and condition management, is indicative of a higher or lower incidence of COVID-19 in the N3C population. Preliminary evidence suggests that pati ents with comorbidities, particularly those with immunosuppressive and/or low survival over 10 years, may have a lower incidence of COVID-19 due to better adherence to social distancing practices.

The purpose of the study is provide information to help guide clinical decision making regarding respiratory support for CoVID-19 patients, specifically the effects of treating patients with high levels of non-invasive respiratory support (> 6L supplemental oxygen) vs early intubation (< 72 hours). This is an observational retrospective cohort study to determine the potential benefit of avoiding e ndotracheal intubation in CoVID-19 patients. Information used pertaining to the primary outcome of interest includes rate of intubation of those trialed on non-invasive methods, overall time on ventilator support, and length of stay in ICU.

Malnutrition is a global health crisis that effects up to one out of two older adults and results in an estimated annual cost of $51.3 billion per year in the United States. Malnutrition is linked to weaker immune systems, reduced cardiac output, poor diaphragmatic and respiratory muscle function and impaired gastrointestinal function. Early studies have explored the prevalence and severity of mal nutrition in COVID-19 patients. However, small sample sizes and single-site studies limit the generalizability of results. Currently, a gap in knowledge exists regarding the relationship of malnutrition in hospital admissions in COVID-19 patients and the impact of malnutrition on clinical outcomes

With the progress of COVID pandemic, reoccurrence of COVID cases have been reported not only among people without symptoms or with mild symptoms, but also have occurred with patients who have been hospitalized, recovered with two consecutive negative tests and then infected again with severe symptoms. Understanding the mechanism behind the reoccurrence of COVID and the severity of the disease rela ted to it is important for fighting COVID in the long term. We will develop novel competing risk models and machine learning algorithms to understand this question.

With the COVID-19 pandemic ongoing and the flu season approaching, there will be a challenge for people to differentiate the two diseases based on their symptoms. Therefore, the COVID-19 testing centers will potentially face challenges of reaching testing capacity and higher risks for infection due to increased number of people going on site tests. Our goal is to use machine learning method to bui ld a classification model to establish a pre-test screening system to help people self-evaluate their risks with COVID-19 during the flu season to decrease their possibility with COVID-19 infection. We are requesting Level 2 de-identified data for this purpose.

The SARS-CoV-2 enters the human gastrointestinal tract via the ACE-2 receptors which are predominantly expressed within the gastrointestinal tract, from gastric to colonic tissue. The pathophysiology of the SARS-CoV-2 infection on the gastrointestinal tract is commonly manifested as anorexia, nausea, vomiting, and diarrhea in COVID-19 patients. These clinical manifestations usually (have been repo rted to) occur in combination with respiratory symptoms in 50% of the affected patients and independently without respiratory symptoms in 10% of patients. Patients also present with abnormalities of pancreatic and hepatic enzymes. Whether COVID presentation including gastrointestinal symptoms portends worse outcomes remains poorly undefined. The project will examine whether COVID clinical presentation with gastrointestinal symptoms is associated with ICU length of stay, total length of hospitalization, and inpatient survival. The relationship between gastrointestinal symptoms at clinical presentation and other known or candidate risk factors for severity of illness and poor outcomes such as obesity, diabetes mellitus, and hypertension will also be examined.

Background: Children are at risk for severe SARS-Cov-2 infection. Epidemiology of SARS-COVID-19 in children is evolving and poorly characterized. There is paucity of literature regarding outcomes and predictive factors of hospitalized children. Projections indicate that incidence of COVID-19 may persist for several months. Clinical characteristics and outcomes of hospitalized children will be help ful for future resource utilization. \nObjectives:\nTo study the characteristics and outcomes of children presenting with COVID-19 using a large multi-institutional dataset and a validated identification method. To study the factors predictive of adverse outcomes in hospitalized children.

Patients with inflammatory bowel disease are frequently prescribed biologic agents and steroids for their gastrointestinal disease. Some preliminary data has suggested that IBD patients may be at risk of severe disease. We wish to study the clinical outcomes of COVID 19 in these patients including those with ulcerative colitis and Inflammatory bowel disease. We will compare risk of severe disease in patients on different biologic agents and those on no biologic medication.

Preliminary data showed that cancer and heart diseases are amongst common comorbidities in patients who died of COVID-19. However, data on patients suffering from both cancer and heart diseases is unknown. Recent and concomitant chemotherapy use can mitigate immune response and make patients with cancer more prone to infection. Certain cardiac conditions such as heart failure could also worsen in flammatory state and subsequently blunting immune response. Thus, outcomes of patients with cancer and cardiac conditions could be worse. To date, no large scale data has looked at the outcomes of patients with cancers and co-existing cardiac conditions.

The purpose of this request is to establish a group of experts who will examine the N3C data in terms of COVID’s impact on individuals (of all ages) who have at least one disability. Disability types would include: intellectual, developmental, vision impairment, deaf or hard of hearing, mental health conditions, acquired brain injury, and mobility/physical disabilities. We anticipate disparities b etween individuals based on ability differences in terms of the prevalence of COVID, symptomology, and service utilization. Individuals with disabilities and their families have documented challenges to care and an increased vulnerability to infections and complications due to their disability and other medical conditions. With more than 35% of the nation’s population reporting at least one disability, it is important to examine the impact of COVID-19 across the lifespan. Information garnered from this effort would potentially provide opportunities for public health, infectious disease, and other approaches.

There are no clear data on the best treatments for COVID-19 to minimize mortality in various populations. As an example, an NIH randomized clinical trial concluded that remdesivir was superior to placebo among hospitalized patients, significantly shortening time to recovery (10 vs. 15 days).1 Though mortality was significantly lower in the remdesivir group at day 15 (6.7% vs. 11.9%), mortality at day 29 trended lower (11.4% vs. 15.2%) but did not reach statistical significance. In contrast, a recent World Health Organization guidelines panel recommended against remdesivir use citing that there is currently no evidence that it improves survival. The WHO guideline was based on analysis of four clinical trials conducted among 7000 patients.2 This project utilizing the N3C database seeks to evaluate what therapies are associated with the lowest mortality among various populations.

The purpose of the project is to further investigate increased incidence of thrombotic events post COVID-19 vaccine administration. We aim to identify potential risk factors including but not limited to age, sex, co-morbidities, etc. We also plan to do a cross study comparing rates of clotting between COVID-19 patients and those individuals who received the COVID-19 vaccine.

Patients who have acquired coronavirus disease 2019 (COVID-19) have displayed numerous cardiovascular complications during the course of their infection, where these complications have included acute myocardial injury, ischemic heart disease, heart failure, stroke, and arrhythmias. While these manifestations are responsible for several complications during the course of the disease, numerous patie nts have reported continued cardiovascular abnormalities following their recovery. Our study will use the data present in the N3C Enclave to investigate the pharmacological management of patients prior to and post COVID19 exposure to determine whether there is an association between COVID-19 and an increased incidence or progression of different cardiovascular diseases that required increased or additional pharmacological management.

In our study, we wish to compare outcomes of COVID-19 infection in patients with and without chronic liver disease. Subgroup analyses will be conducted to compare these outcomes in patients with viral hepatitides B, C, alcoholic liver disease and non-alcoholic fatty liver disease (NASH), with and without cirrhosis. The effect of different antiviral treatments for Hepatitis B or C in patients with viral hepatitis on clinical course of COVID-19 will also be explored. We also aim to look at predictors of severe disease in this cohort.

After Covid diagnosis, there are changes to the pro-inflammatory/pro-thrombotic state. In elective joint replacement surgery, this inflammation/thrombosis could lead to increased postoperative risks. Will be utilizing data to understand the safest timing for elective surgery following Covid infection and complication rates compared to cohorts.

Our project aims to determine the prevalence of gastrointestinal hemorrhage in patients with COVID-19 infection. We also aim to compare treatment pathways and outcomes in patients with or without COVID-19 infection who present with gastrointestinal hemorrhage. Our secondary aim is to further explore the risk factors for gastrointestinal hemorrhage in patients with COVID-19 infection. Lastly, the t reatment of gastrointestinal hemorrhage involves potentially aerosol-generating procedures such as endoscopic evaluation. Therefore, we wish to explore the concern that these procedures may be used less frequently in patients with COVID-19 infection.

Aim:\nThe purpose of this study is to study the outcomes of COVID-19 infection in patients with elevated cardiac biomarkers (D-Dimer, BNP, and Troponin)\n\nMethods:\nUsing ICD-10 codes we will identify all patients over the age of 18 years with COVID-19 infection. The identified cohort will then be further stratified into two groups based on the presence or absence of elevated Cardiac Biomarkers. Patients with COVID-19 infection and normal cardiac biomarkers will serve as the control population. \n\nStatistical plan: To be determined\n\nOutcomes:\n1.\tThe primary outcome will be to study in-hospital and short term (30 days) all-cause mortality \n2.\tSecondary outcomes include need for mechanical circulatory support devices, new-onset arrhythmias, need for ventilator support, cardiovascular mortality, re-admission rates, total length of stay, and hospitalization cost

The effect of severe COVID-19 infection in pregnancy is limited to a small number of case reports. The use of extracorporeal membrane oxygenation (ECMO) in pregnancy is rare, typically reserved for salvage therapy in acute respiratory distress syndrome (ARDS). Specifically, veno-venous ECMO (V-V ECMO) may be used for refractory hypoxemia in patients who cannot be safely supported with conventional mechanical ventilation. The use of ECMO for COVID-19 infection in pregnancy has not yet been reported in a large series as individual institutions only have few cases. This limits the information available for analysis. We seek to investigate the maternal and neonatal outcomes of ECMO use in pregnancy-associated COVID-19 infection in a large national cohort.

The OU Data lab is interested in developing methods to discover statistical anomalies across COVID-19 data. Statistical anomalies can lead to premature or incorrect conclusions. Analysts who are not experts need tools and methodologies to help them avoid harmful or misleading results. We will develop the methods and evaluate the results across various sections of data. Goals of the project: a ) Implement Simpson’s Paradox discovery pipeline for data sets. b) Identify Simpson’s paradox across attributes of the data. c) Define data quality measures for available data sets. d) Implement visualizations to clearly show the strength of Simpson’s paradox across data sets.

The Rural Health Domain Team seeks to understand the epidemiology, utilization, treatment, and outcomes of the COVID-19 pandemic in rural communities. The overarching goal is to develop better evidence for potential differences in the COVID-19 epidemic response and outcomes for rural health care centers and rural dwellers. Studying the at-risk populations and unique challenges faced by rural commu nities during and after the pandemic may provide opportunities for improving rural health and health care.\n\nAdapted from https://covid.cd2h.org/rural-health

This will help us establish the OU workflow for obtaining enclave data, restructuring the format, and modeling outcomes. Only synthetic data are involved.

In this project, we aim to study the role of the COVID-19 pandemic on the weight-gain trends in the US. Informed by the existing studies, our hypothesis is that the pandemic had a meaningful adverse impact on the already alarming weight-gain trends. We plan to investigate these trends separately in children and adult populations. While trying to identify such trends at a large scale and on the ind ividual level, we also aim to study the disparities with respect to COVID-19 status, age, gender, race, and social determinant of health (SDOH).

Mathematical models currently used to explain the progression of pandemic-inducing pathogens are wholly deficient at explaining and accurately predicting pandemics with ever-evolving characteristics. The current basic SEIR model (susceptible, exposed, infectious, and recovered) can only account for one wave of infection because it does not account for dynamically changing susceptibility, infectiou sness, etc. With this project, we aim to develop a mathematical modeling tool developed from chemical reaction kinetics equations that can better predict the dynamic evolution of the COVID-19 pandemic among others, as well as predict the effect of certain policies on transmission (such as mask mandating, social distancing, vaccination, etc.) This model already shows promise with adult populations, as validated by recursive forward prediction. Using patient data on COVID test type, result, date as well as recovery or death dates, a mathematical model can be built to represent changes in infection over time. Also, using deprivation indices and ZIP code, this model may be able to predict infection at a localized level.

Coronavirus disease 2019 (COVID-19) is characterized by a male predominance in severity and mortality worldwide, but reasons underlying the sex-bias are unclear. Sex differences in comorbidities, sex hormones, and inflammatory biomarkers may all be factors involved in the sex differences of COVID-19 outcomes. However, few studies examined the role of these factors in COVID-19 outcomes stratified b y sex. In our recent study in a COVID cohort of adults in Louisiana, we observed sex differences in comorbidities and biomarkers as predictors of COVID-19 severe outcomes. Our study underscores the need for a larger study with sex-stratification and robust analysis to determine if our findings are region-specific or hold generalizability to other populations. We propose to use the rich and population-based National COVID Cohort Collaborative (N3C) data to examine to what extent comorbidities, female sex hormones, and inflammatory biomarkers represent sex-specific determinants of COVID-19 outcomes.