Study population and setting
The study included 928 adult patients (median age 66 years; 50% male) with current or past invasive cancer and with laboratory-confirmed COVID-19 treated at participating facilities in the United States, Canada, or Spain. Baseline information on SARS-CoV-2-infected patients was entered between March 17 and April 16, 2020, with follow-up through May 7, 2020. Median follow-up duration was 21 days. Logistic regression was used to examine associations between patient factors and outcomes (primary outcome: death within 30 days of COVID-19 diagnosis; secondary outcomes: hospital admission, ICU admission, mechanical ventilation, supplemental oxygen).
Summary of Main Findings
Overall, by the end of follow-up, 13% of patients died within 30 days of COVID-19 diagnosis; and 26% of patients experienced the composite outcome of death, ICU admission, or mechanical ventilation. Mortality among those admitted to the hospital was 23%. Factors associated with increased mortality included older age, male sex, history of smoking, more comorbidities, current or progressing disease (versus remission), and treatment with azithromycin and hydroxychloroquine. There was no observed impact of race/ethnicity, obesity, and cancer type on mortality.
The study included a large sample of patients with cancer. Geographically diverse study sites increase generalizability of findings.
Because the study relied on providers to input information into the database on cancer patients with COVID-19, those with severe disease may be more likely to be entered. Patients with mild or asymptomatic disease may not have presented for care or would be less likely to trigger data entry. This would result in overestimation of the likelihood of severe disease and death. Those who present with severe disease may have been more likely to be administered azithromycin/hydroxychloroquine; thus, we cannot conclude that use of the drug combination itself is independently associated with severe disease (confounding by indication). Patients with current cancer may have considerably different exposure-outcome relationships from those whose cancer was in remission or was successfully treated; if so, analysis with stratification would be more appropriate.
This is one of largest studies to date on COVID-19 in patients with cancer.