A Web-based Platform to Share Harmonized Results from COVID-19 Clinical Studies

COVID-19 has caused a worldwide pandemic, accompanied by a high number of deaths and hospitalizations. Multiple preventative vaccines and variety of COVID-19 treatments have been developed and explored. This large volume of scientific work led to an extensive number of COVID-19 publications, which resulted in the necessity to standardize, store, share, and investigate research results in a harmonized manner. Attempts to standardize and share COVID-19 research data have been lacking. The purpose of the ReMeDy platform is to provide an intelligent informatics solution of integrating diverse COVID-19 trial outcomes and omics data across COVID-19 research studies. To test the platform, we utilized 48 COVID-19 observational retrospective studies. The robustness of the platform was validated through the ability to efficiently organize the diverse data elements. Next steps include expanding our database through the inclusion of all published COVID-19 studies. ReMeDy is located at https://remedy.mssm.edu/.

Characteristics of Electronic Informed Consent Platforms for Consenting Patients to Research Studies: A Scoping Review.

Informed consent process assures that research study participants are properly informed about the study prior to their consent. Due to the increasing significance of electronic informed consent (eIC) platforms, particularly during the COVID-19 pandemic, we conducted a scoping review of eIC systems to address the following characteristics: 1) technological features of current eIC platforms, 2) eIC platforms usability and efficacy, and 3) areas for future eIC research. We performed a literature search using publically available PubMed repository, where we included studies discussing an eIC platform or multimedia educational module given to patients prior to signing a consent form. In addition, we tracked first author, year of publication, sample size, study location, eIC procedure, methodology, and eIC's comparison to paper consent. Our results showed that with a few noted exceptions, electronic consent improves patient usability, satisfaction, knowledge, and trust scores when compared to traditional paper consent.

NLP-Assisted Pipeline for COVID-19 Core Outcome Set Identification Using ClinicalTrials.gov.

Core outcome sets (COS) are necessary to ensure the systematic collection, metadata analysis and sharing the information across studies. However, development of an area-specific clinical research is costly and time consuming. ClinicalTrials.gov, as a public repository, provides access to a vast collection of clinical trials and their characteristics such as primary outcomes. With the growing number of COVID-19 clinical trials, identifying COSs from outcomes of such trials is crucial. This paper introduces a semi-automatic pipeline that can efficiently identify, aggregate and rank the COS from the primary outcomes of COVID-19 clinical trials. Using Natural language processing (NLP) techniques, our proposed pipeline successfully downloads and processes 5090 trials from all over the world and identifies COVID-19-specific outcomes that appeared in more than 1% of the trials. The top-of-the-list outcomes identified by the pipeline are mortality due to COVID-19, COVID-19 infection rate and COVID-19 symptoms.

Assessing Disparities in COVID-19 Testing Using National COVID Cohort Collaborative.

With NCATS National COVID Cohort Collaborative (N3C) dataset, we evaluated 14 billion medical records and identified more than 12 million patients tested for COVID-19 across the US. To assess potential disparities in COVID-19 testing, we chose ten US states and then compared each state's population distribution characteristics with distribution of corresponding characteristics from N3C. Minority racial groups were more prevalent in the N3C dataset as compared to census data. The proportion of Hispanics and Latinos in N3C was slightly lower than in the state census. Patients over 65 years old had higher representation in the N3C dataset and patients under 18 were underrepresented. Proportion of females in the N3C was higher compared with the state data. All ten states in N3C showed a higher representation of urban population versus rural population compared to census data.

Using Big Data to Identify Impact of Asthma on Mortality in Patients with COVID-19.

The goal of this paper was to assess if mortality in COVID-19 positive patients is affected by a history of asthma in anamnesis. A total of 48,640 COVID-19 positive patients were included in our analysis. A propensity score matching was carried out to match each asthma patient with two patients without history of chronic respiratory diseases in one stratum. Matching was based on age, comorbidity score, and gender. Conditional logistics regression was used to compute within each strata. There were 5,557 strata in this model. We included asthma, ethnicity, race, and BMI as risk factors. The results showed that the presence of asthma in anamnesis is a statistically significant protective factor from mortality in COVID-19 positive patients.

Use of Artificial Intelligence for Predicting COVID-19 Outcomes: A Scoping Review.

During the COVID-19 pandemic, artificial intelligence has played an essential role in healthcare analytics. Scoping reviews have been shown to be instrumental for analyzing recent trends in specific research areas. This paper aimed at applying the scoping review methodology to analyze the papers that used artificial intelligence (AI) models to forecast COVID-19 outcomes. From the initial 1,057 articles on COVID-19, 19 articles satisfied inclusion/exclusion criteria. We found that the tree-based models were the most frequently used for extracting information from COVID-19 datasets. 25% of the papers used time series to transform and analyze their data. The largest number of articles were from the United States and China. The reviewed artificial intelligence methods were able to predict cases, death, mortality, and severity. AI tools can serve as powerful means for building predictive analytics during pandemics.

Latent COVID-19 Clusters in Patients with Opioid Misuse.

The goal of this paper is to apply unsupervised machine learning techniques in order to discover latent clusters in patients who have opioid misuse and also undergone COVID-19 testing. Target dataset has been constructed based on COVID-19 testing results at Mount Sinai Health System and opioid treatment program (OTP) information from New York State Office of Addiction Service and Support (OASAS). The dataset was preprocessed using factor analysis for mixed data (FAMD) method and then K-means algorithm along with elbow method were used to determine the number of optimal clusters. Four patient clusters were identified among which the fourth cluster constituted the maximum percentage of positive COVID-19 test results (20%). Compared to the other clusters, this cluster has the highest percentage of African Americans. This cluster has also the highest mortality rate (16.52%), hospitalization rate after receiving the COVID-19 test result (72.17%, use of ventilator (7.83%) and ICU admission rate (47.83%). In addition, this cluster has the highest percentage of patients with at least one chronic disease (99.13%) and age-adjusted comorbidity score more than 1 (83.48%). Longer participation in OTP was associated with the highest morbidity and mortality from COVID-19.

Identifying Core Outcome Sets in COVID-19 Clinical Trials Using ClinicalTrials.gov.

Introduction of core outcome sets (COS) facilitates evidence synthesis, transparency in outcome reporting, and standardization in clinical research. However, development of COS may be a time consuming and expensive process. Publicly available repositories, such as ClinicalTrials.gov (CTG), provide access to a vast collection of clinical trial characteristics including primary and secondary outcomes, which can be analyzed using a comprehensive set of tools. With growing number of COVID-19 clinical trials, COS development may provide crucial means to standardize, aggregate, share, and analyze diverse research results in a harmonized way. This study was aimed at initial assessment of utility of CTG analytics for identifying COVID-19 COS. At the time of this study, January, 2021, we analyzed 120 ongoing NIH-funded COVID-19 clinical trials initiated in 2020 to inform COVID-19 COS development by evaluating and ranking clinical trial outcomes based on their structured representation in CTG. Using this approach, COS comprised of 25 major clinical outcomes has been identified with mortality, mental health status, and COVID-19 antibodies at the top of the list. We concluded that CTG analytics can be instrumental for COVID-19 COS development and that further analysis is warranted including broader number of international trials combined with more granular approach and ontology-driven pipelines for outcome extraction and curation.

Impact of COVID-19 Pandemic on Use of Telemedicine Services in an Academic Medical Center.

The COVID-19 pandemic changed the landscape of telehealth services. The goal of this paper was to identify demographic groups of patients who have used telemedicine services before and after the start of the pandemic, and to analyze how different demographic groups' telehealth usage patterns change throughout the course of the pandemic. A de-identified study dataset was generated by querying electronic health records at the Mount Sinai Health System to identify all patients. 129,625 patients were analyzed. Demographic shifts in patients seeking telemedicine service were identified. There was significant increase in the middle age and older population using telehealth services. During the pandemic use of telemedicine services was increased among male patients and racial minority patients. Furthermore, telehealth services had expanded to a broader spectrum of medical specialties.

Machine Learning to Predict Mortality and Critical Events in a Cohort of Patients With COVID-19 in New York City: Model Development and Validation.

BACKGROUND: COVID-19 has infected millions of people worldwide and is responsible for several hundred thousand fatalities. The COVID-19 pandemic has necessitated thoughtful resource allocation and early identification of high-risk patients. However, effective methods to meet these needs are lacking. OBJECTIVE: The aims of this study were to analyze the electronic health records (EHRs) of patients who tested positive for COVID-19 and were admitted to hospitals in the Mount Sinai Health System in New York City; to develop machine learning models for making predictions about the hospital course of the patients over clinically meaningful time horizons based on patient characteristics at admission; and to assess the performance of these models at multiple hospitals and time points. METHODS: We used Extreme Gradient Boosting (XGBoost) and baseline comparator models to predict in-hospital mortality and critical events at time windows of 3, 5, 7, and 10 days from admission. Our study population included harmonized EHR data from five hospitals in New York City for 4098 COVID-19-positive patients admitted from March 15 to May 22, 2020. The models were first trained on patients from a single hospital (n=1514) before or on May 1, externally validated on patients from four other hospitals (n=2201) before or on May 1, and prospectively validated on all patients after May 1 (n=383). Finally, we established model interpretability to identify and rank variables that drive model predictions. RESULTS: Upon cross-validation, the XGBoost classifier outperformed baseline models, with an area under the receiver operating characteristic curve (AUC-ROC) for mortality of 0.89 at 3 days, 0.85 at 5 and 7 days, and 0.84 at 10 days. XGBoost also performed well for critical event prediction, with an AUC-ROC of 0.80 at 3 days, 0.79 at 5 days, 0.80 at 7 days, and 0.81 at 10 days. In external validation, XGBoost achieved an AUC-ROC of 0.88 at 3 days, 0.86 at 5 days, 0.86 at 7 days, and 0.84 at 10 days for mortality prediction. Similarly, the unimputed XGBoost model achieved an AUC-ROC of 0.78 at 3 days, 0.79 at 5 days, 0.80 at 7 days, and 0.81 at 10 days. Trends in performance on prospective validation sets were similar. At 7 days, acute kidney injury on admission, elevated LDH, tachypnea, and hyperglycemia were the strongest drivers of critical event prediction, while higher age, anion gap, and C-reactive protein were the strongest drivers of mortality prediction. CONCLUSIONS: We externally and prospectively trained and validated machine learning models for mortality and critical events for patients with COVID-19 at different time horizons. These models identified at-risk patients and uncovered underlying relationships that predicted outcomes.

Retrospective cohort study of clinical characteristics of 2199 hospitalised patients with COVID-19 in New York City.

OBJECTIVE: The COVID-19 pandemic is a global public health crisis, with over 33 million cases and 999 000 deaths worldwide. Data are needed regarding the clinical course of hospitalised patients, particularly in the USA. We aimed to compare clinical characteristic of patients with COVID-19 who had in-hospital mortality with those who were discharged alive. DESIGN: Demographic, clinical and outcomes data for patients admitted to five Mount Sinai Health System hospitals with confirmed COVID-19 between 27 February and 2 April 2020 were identified through institutional electronic health records. We performed a retrospective comparative analysis of patients who had in-hospital mortality or were discharged alive. SETTING: All patients were admitted to the Mount Sinai Health System, a large quaternary care urban hospital system. PARTICIPANTS: Participants over the age of 18 years were included. PRIMARY OUTCOMES: We investigated in-hospital mortality during the study period. RESULTS: A total of 2199 patients with COVID-19 were hospitalised during the study period. As of 2 April, 1121 (51%) patients remained hospitalised, and 1078 (49%) completed their hospital course. Of the latter, the overall mortality was 29%, and 36% required intensive care. The median age was 65 years overall and 75 years in those who died. Pre-existing conditions were present in 65% of those who died and 46% of those discharged. In those who died, the admission median lymphocyte percentage was 11.7%, D-dimer was 2.4 µg/mL, C reactive protein was 162 mg/L and procalcitonin was 0.44 ng/mL. In those discharged, the admission median lymphocyte percentage was 16.6%, D-dimer was 0.93 µg/mL, C reactive protein was 79 mg/L and procalcitonin was 0.09 ng/mL. CONCLUSIONS: In our cohort of hospitalised patients, requirement of intensive care and mortality were high. Patients who died typically had more pre-existing conditions and greater perturbations in inflammatory markers as compared with those who were discharged.

Latent COVID-19 Clusters in Patients with Chronic Respiratory Conditions.

The goal of this paper was to apply unsupervised machine learning techniques towards the discovery of latent COVID-19 clusters in patients with chronic lower respiratory diseases (CLRD). Patients who underwent testing for SARS-CoV-2 were identified from electronic medical records. The analytical dataset comprised 2,328 CLRD patients of whom 1,029 were tested COVID-19 positive. We used the factor analysis for mixed data method for preprocessing. It performed principle component analysis on numeric values and multiple correspondence analysis on categorical values which helped convert categorical data into numeric. Cluster analysis was an effective means to both distinguish subgroups of CLRD patients with COVID-19 as well as identify patient clusters which were adversely affected by the infection. Age, comorbidity index and race were important factors for cluster separations. Furthermore, diseases of the circulatory system, the nervous system and sense organs, digestive system, genitourinary system, metabolic diseases and immunity disorders were also important criteria in the resulting cluster analyses.

Usability of Remote Assessment of Exercise Capacity for Pulmonary Telerehabilitation Program.

Pulmonary rehabilitation [PR] has been successfully carried out via telemedicine however initial patient assessment has been traditionally conducted in PR centers. The first step in PR is assessment of patient's exercise capacity which allows individualized prescription of safe and effective exercise program. With COVID-19 pandemics assessment of patients in PR centers has been limited resulting in significant reduction of patients undergoing life-saving PR. The goal of this pilot study was to introduce approaches for remote assessment of exercise capacity using videoconferencing platforms and provide initial usability assessment of this approach by conducing cognitive walkthrough testing. We developed a remote assessment system that supports comprehensive physical therapy assessment necessary for prescription of a personalized exercise program tailored to individual fitness level and limitations in gait and balance of the patient under evaluation. Usability was assessed by conducting cognitive walkthrough and system usability surveys. The usability inspection of the remote exercise assessment demonstrated overall high acceptance by all study participants. Our next steps in developing user-centered interface should include usability evaluation in different subgroups of patients with varying socio-economic background, different age groups, computer skills, literacy and numeracy.

Unsupervised Machine Learning for the Discovery of Latent Clusters in COVID-19 Patients Using Electronic Health Records.

The goal of this paper was to apply unsupervised machine learning techniques towards the discovery of latent clusters in COVID-19 patients. Over 6,000 adult patients tested positive for the SARS-CoV-2 infection at the Mount Sinai Health System in New York, USA met the inclusion criteria for analysis. Patients' diagnoses were mapped onto chronicity and one of the 18 body systems, and the optimal number of clusters was determined using K-means algorithm and the elbow method. 4 clusters were identified; the most frequently associated comorbidities involved infectious, respiratory, cardiovascular, endocrine, and genitourinary disorders, as well as socioeconomic factors that influence health status and contact with health services. These results offer a strong direction for future research and more granular analysis.