Data Source and Study Design
Our study received Institutional Review Board Approval (HS#: 20-00875). The study was conducted at a large, New York City based, multi-hospital, academic medical institution. The institution’s Scientific Computing team has been maintaining a regularly updated, deidentified, database of patient-level data from our electronic medical record (EMR) system for encounters related to COVID19. At the time of data extraction for the present study, the database included encounters from March 1st, 2020 through August 10th, 2020.
We included all patients who had a healthcare encounter indicating a positive result for the severe acute respiratory syndrome coronavirus 2 (SARS-CoV2) real-time reverse transcriptase-polymerase-chain-reaction (RT-PCR) assay test during the study period. We identified patient characteristics including demographics, comorbidities (including historical cancer diagnoses), and outcomes from the EMR as well as the location of the encounter. If patients had multiple encounters, an encounter with a positive result was chosen, if applicable. If patients had multiple patient encounters with a positive result, the most severe encounter was chosen. Patients with missing data on any of the variables included in the analysis were excluded.