Factors associated with deaths due to COVID-19 versus other causes: population-based cohort analysis of UK primary care data and linked national death registrations within the OpenSAFELY platform

Bhaskaran, K.; Bacon, S. C.; Evans, S. J.; Bates, C. J.; Rentsch, C. T.; Brian, M.; Tomlinson, L.; Walker, A. J.; Schultze, A.; Morton, C. E.; Grint, D.; Mehrkar, A.; Eggo, R. M.; Inglesby, P.; Douglas, I. J.; McDonald, H. I.; Cockburn, J.; Williamson, E. J.; Evans, D.; Curtis, H. J.; Hulme, W. J.; Parry, J.; Hester, F.; Harper, S.; Spiegelhalter, D.; Smeeth, L.; Goldacre, B.

Background: Mortality from COVID-19 shows a strong relationship with age and pre-existing medical conditions, as does mortality from other causes. However it is unclear how specific factors are differentially associated with COVID-19 mortality as compared to mortality from other causes. Methods: Working on behalf of NHS England, we carried out a cohort study within the OpenSAFELY platform. Primary care data from England were linked to national death registrations. We included all adults (aged [≥]18 years) in the database on 1st February 2020 and with >1 year of continuous prior registration, the cut-off date for deaths was 9th November 2020. Associations between individual-level characteristics and COVID-19 and non-COVID deaths were estimated by fitting age- and sex-adjusted logistic models for these two outcomes. Results: 17,456,515 individuals were included. 17,063 died from COVID-19 and 134,316 from other causes. Most factors associated with COVID-19 death were similarly associated with non-COVID death, but the magnitudes of association differed. Older age was more strongly associated with COVID-19 death than non-COVID death (e.g. ORs 40.7 [95% CI 37.7-43.8] and 29.6 [28.9-30.3] respectively for [≥]80 vs 50-59 years), as was male sex, deprivation, obesity, and some comorbidities. Smoking, history of cancer and chronic liver disease had stronger associations with non-COVID than COVID-19 death. All non-white ethnic groups had higher odds than white of COVID-19 death (OR for Black: 2.20 [1.96-2.47], South Asian: 2.33 [2.16-2.52]), but lower odds than white of non-COVID death (Black: 0.88 [0.83-0.94], South Asian: 0.78 [0.75-0.81]). Interpretation: Similar associations of most individual-level factors with COVID-19 and non-COVID death suggest that COVID-19 largely multiplies existing risks faced by patients, with some notable exceptions. Identifying the unique factors contributing to the excess COVID-19 mortality risk among non-white groups is a priority to inform efforts to reduce deaths from COVID-19.