Association of Age, Gender, Deprivation, Urbanicity, Ethnicity, and Smoking with a Positive Test for COVID-19 in an English Primary Care Surveillance Network: Cross Sectional Study of the First 500 Cases
Lusignan SD., Dorward J., Correa A., Jones N., Akinyemi O., Amirthalingam G., Andrews N., Byford R., Dabrera G., Elliot A., Ellis J., Ferreira F., Bernal JL., Okusi C., Ramsay M., Sherlock J., Smith G., Howsam G., Zambon M., Hobbs R., Joy M., Williams J.
Background: There are few epidemiological studies of community cases in the current coronavirus-2019 (COVID-19) pandemic. We report on the first 500 COVID-19 cases identified through United Kingdom primary care surveillance and describe risk factors for testing COVID-19 positive. Methods: The Oxford-Royal College of General Practitioners (RCGP) Research and Surveillance Centre (RSC), is a nationally representative primary care sentinel network sharing pseudonymised data, including virological test data for COVID-19. We used multivariable logistic regression models with multiple imputation to identify risk factors for positive COVID-19 tests within this surveillance programme. Findings: We identified 3,802 COVID-19 results between 28/01/20 and 04/04/2020, 587 were positive. Greater odds of testing COVID-19 positive included: working-age people (40-64years) and older age, (≥75 years) versus 0-17 year olds (adjusted odds ratio [aOR] 5.26, 95%CI:3.26-8.49 and 5.17,95%CI:2.99-8.92, respectively); male gender (aOR 1.56, 95%CI:1.28-1.90); black and mixed ethnicity compared with white (aOR 4.55, 95%CI:2.55-8.10 and 1.84 95%CO:1.1-3.14, respectively)); urban compared with rural areas (aOR 4.58, 95%CI:3.57-5.88); people with chronic kidney disease (CKD) (aOR 1.88, 95%CI:1.29-2.75) and increasing body mass index (aOR 1.02, 95%CI:1.00-1.03). People in the least deprived deprivation quintile had lower odds of a positive test (aOR 0.49 95%CI:0.36-0.65) as did current smokers (aOR 0.53, 95%CI:0.38-0.74). Interpretation: A positive COVID-19 test result in primary care was associated with similar risk factors for severe outcomes seen in hospital settings, with the exception of smoking. We provide early evidence of potential sociodemographic factors associated with a positive test, including ethnicity, deprivation, population density, and CKD. Funding Statement: Public Health England provides the core funding for RCGP RSC, no specific funding was provided for this analysis.Declaration of Interests: The authors have no competing interests. SdeL is the Director of the Oxford RCGP RSC, RB, JS, FF, EK and GH are part funded by PHE; and CO and AC by a Wellcome Biomedical resources grant (212763/Z/18/Z). JD is funded by Wellcome Trust (216421/Z/19/Z).Ethics Approval Statement: This study was approved by the RCGP RSC study approval committee and was classified as a study of “usual practice”. Therefore, no further ethical approval was required.