We have created and launched a common dataset for suspected and confirmed COVID-19 patients.
The selected variables complement other existing hospitalized patient datasets using the WHO International Severe Acute Respiratory and emerging Infection Consortium forms which do not collect key ED variables, or variables on gender, vulnerability or follow-up.
Prospective Data Collection
ED nurses and physicians will complete a one-page prospective digital data collection form on consecutive eligible patients with suspected and confirmed COVID-19 to capture important clinical variables often not found in medical records or administrative data, such as the date of earliest symptom onset and information about vulnerability (e.g., substance use).
Retrospective Data Collection
Research assistants will review the patients’ records and abstract demographics, history, risk factors, comorbid conditions, medication use, physical examination findings, treatment, procedures and outcomes from the pre-hospital, ED and in-hospital records. Follow-up telephone calls at 30 days will confirm patients’ consent to participate in follow-up at 60 days, and 6 and 12 months, and to link our dataset with administrative data. At follow-up, we will ascertain gender, race, ethnicity and housing situation at the time of the ED visit.
REDCap database architecture
The database architecture we have developed is consistent with provincial laws and institutional policies. The University of British Columbia (UBC) hosts our national REDCap database of clinical data. Each provincial lead institute will hold the PHI and study ID on a unique REDCap database. This architecture will enable us to link our data with administrative datasets at a time when new data (e.g., COVID test results or serology) will emerge as the pandemic evolves. Our data infrastructure and executed agreement will also enable our collaborators to make secondary use of our network data, reducing their costs of clinical data capture.