A population-based study of COVID-19 mortality risk in US cancer patients

Abstract Background In this study, we provide the largest analysis to date of a US-based cancer cohort to characterize death from COVID-19. Methods A total of 4 020 669 patients across 15 subtypes living with cancer in 2020 and included in the National Cancer Institute’s Surveillance, Epidemiology, and End Results database were abstracted. We investigated prognostic factors for death due to COVID-19 using a Cox proportional hazards model and calculated hazard ratios (HRs). Standardized mortality ratios were calculated using observed mortality counts from Surveillance, Epidemiology, and End Results and expected mortality based on US mortality rates. Results A total of 291 323 patients died, with 14 821 (5.1%) deaths attributed to COVID-19 infection. The COVID-19 disease-specific mortality rate was 11.81/10 000-persons years, and the standardized mortality ratio of COVID-19 was 2.30 (95% confidence interval [CI] = 2.26 to 2.34; P < .0001). COVID-19 ranked as the second leading cause of death following ischemic heart disease (5.2%) among 26 noncancer causes of death. Patients who are older (80 years and older vs 49 years and younger: HR = 21.47, 95% CI = 19.34 to 23.83), male (vs female: HR = 1.46, 95% CI = 1.40 to 1.51), unmarried (vs married: HR = 1.47, 95% CI = 1.42 to 1.53), and Hispanic or non-Hispanic African American (vs non-Hispanic White: HR = 2.04, 95% CI = 1.94 to 2.14 and HR = 2.03, 95% CI = 1.94 to 2.14, respectively) were at greatest risk of COVID-19 mortality. Conclusions We observed that people living with cancer are at 2 times greater risk of dying from COVID-19 compared with the general US population. This work may be used by physicians and public health officials in the creation of survivorship programs that mitigate the risk of COVID-19 mortality.