Clinical Characterization and Prediction of Clinical Severity of SARS-CoV-2 Infection Among US Adults Using Data From the US National COVID Cohort Collaborative

Abstract

IMPORTANCEThe National COVID Cohort Collaborative (N3C) is a centralized, harmonized, highgranularity electronic health record repository that is the largest, most representative COVID-19 cohort to date. This multicenter data set can support robust evidence-based development of predictive and diagnostic tools and inform clinical care and policy.OBJECTIVES To evaluate COVID-19 severity and risk factors over time and assess the use of machine learning to predict clinical severity. DESIGN, SETTING, AND PARTICIPANTSIn a retrospective cohort study of 1 926 526 US adults with SARS-CoV-2 infection (polymerase chain reaction >99% or antigen <1%) and adult patients without SARS-CoV-2 infection who served as controls from 34 medical centers nationwide between January 1, 2020, and December 7, 2020, patients were stratified using a World Health Organization COVID-19 severity scale and demographic characteristics. Differences between groups over time were evaluated using multivariable logistic regression. Random forest and XGBoost models were used to predict severe clinical course (death, discharge to hospice, invasive ventilatory support, or extracorporeal membrane oxygenation). MAIN OUTCOMES AND MEASURESPatient demographic characteristics and COVID-19 severity using the World Health Organization COVID-19 severity scale and differences between groups over time using multivariable logistic regression. RESULTSThe cohort included 174 568 adults who tested positive for SARS-CoV-2 (mean [SD] age, 44.4 [18.6] years; 53.2% female) and 1 133 848 adult controls who tested negative for SARS-CoV-2 (mean [SD] age, 49.5 [19.2] years; 57.1% female). Of the 174 568 adults with SARS-CoV-2, 32 472(18.6%) were hospitalized, and 6565 (20.2%) of those had a severe clinical course (invasive ventilatory support, extracorporeal membrane oxygenation, death, or discharge to hospice). Of the hospitalized patients, mortality was 11.6% overall and decreased from 16.4% in March to April 2020 to 8.6% in September to October 2020 (P = .002 for monthly trend). Using 64 inputs available on the first hospital day, this study predicted a severe clinical course using random forest and XGBoost models (area under the receiver operating curve = 0.87 for both) that were stable over time. The factor most strongly associated with clinical severity was pH; this result was consistent across machine learning methods. In a separate multivariable logistic regression model built for inference, (continued) Key Points Question In a US data resource large enough to adjust for multiple confounders, what risk factors are associated with COVID-19 severity and severity trajectory over time, and can machine learning models predict clinical severity? Findings In this cohort study of 174 568 adults with SARS-CoV-2, 32 472 (18.6%) were hospitalized and 6565 (20.2%) were severely ill, and first-day machine learning models accurately predicted clinical severity. Mortality was 11.6%

Use of Hydroxychloroquine, Remdesivir, and Dexamethasone Among Adults Hospitalized With COVID-19 in the United States

“…Second, although the N3C is the largest U.S. cohort of patients with COVID-19 that has been assembled to date, the centers it includes are clustered in the Southeast, Mid-Atlantic, and Midwest regions, and because most data contributors are academic medical centers, our findings may not generalize to community-based hospitals ( 18 , 45 ). However, the distributions of age, sex, race, and ethnicity in the N3C indicate a cohort that is representative of many segments of the U.S. population ( 18 , 21 ). Third, our information on prescription drug exposure is a combination of information about medication ordering and medication administration, depending on the contributing center, and we do not have information about drug dose.…”

Associations between HIV infection and clinical spectrum of COVID-19: a population level analysis based on US National COVID Cohort Collaborative (N3C) data

“…18 All patients in the N3C Enclave include historical data within the same health-care system as of Jan 1, 2018, which provides information about pre-existing health conditions (eg, comorbidities) and other medical history (look back data). 19 We included all adult (aged ≥18 years) COVID-19 cases from 54 clinical sites across the USA with data being deposited into the N3C and harmonised into a data release set from Jan 1, 2020, to May 8, 2021. The data ingestion and harmonisation process are described in the appendix (p 2) .…”

“…Specifically, N3C defined a single index encounter for each laboratory-confirmed COVID-19 positive patient by selecting encounters that start up to 30 days before or 7 days after the positive test result, or a positive test result during the visit. 19 When multiple encounters met these criteria, the N3C Consortium broke ties by preferentially selecting the encounter in which the most severe outcome was observed, then the longest visit, and finally the most recent visit. 19 …”

“… 19 When multiple encounters met these criteria, the N3C Consortium broke ties by preferentially selecting the encounter in which the most severe outcome was observed, then the longest visit, and finally the most recent visit. 19 …”

Mortality from COVID-19 in Patients with COPD: A US Study in the N3C Data Enclave

“…The National Center for Advancing Translational Sciences (NCATS) established the National COVID Cohort Collaboration (N3C), a partnership among 81 academic hubs to share COVID-19 clinical data from electronic health records as part of a platform for answering critical research questions. 8,9 The N3C compiles data from patients who were tested for COVID-19, or who were symptomatic and highly suspected to have COVID-19. The data obtained from electronic health records include demographics, symptoms, lab test results, procedures, medications, medical conditions, and physical measurements.…”