Tang and his colleagues compared hospitalised, community and staff Coronavirus disease 2019 (COVID-19) infection rates during the early phase of the evolving COVID-19 epidemic [1] , but an emerging factor is the ethnicity of individuals. In this letter, we report our investigation into the potential associations between ethnicity and COVID-19. Despite only 15% of the UK population being from black and ethnic minority (BAME) backgrounds [2] , recent data has identified that 34% of patients admitted to intensive care (ITU) with COVID-19 belong to BAME groups [3] . The potential difference in outcomes associated with ethnicity for patients with COVID-19 has caused concern within the scientific community, with many calling for further research [ 4 , 5 ].Consequently, we analysed routinely collected data from adult patients at Sheffield Teaching Hospitals (STH) between 01/03/20 and 25/04/20. Ethnicities were categorised into BAME (including Black, Asian-subcontinent, mixed and any other non-White background), White and Not Known. A total of 3018 patients were tested for COVID-19, of whom 1493 were female and 1499 male (26 gender-unknown). Median age for BAME patients was 54 years and for White patients was 71 years.Overall data showed that of the 3018 patients tested, 806(26 • 7%) were positive for COVID-19, including 95 from a BAME and 631 from a White background ( Table 1 ). BAME patients were significantly more likely to test positive than the White cohort (X 2 (1, n = 726) = 4 • 9561, p = • 026). Whilst there was an almost equal gender split in terms of population tested, men were significantly more likely to test positive than women ((X 2 (1, n = 2922) = 16 • 90,189, p < • 0 0,0 01). BAME men were significantly more likely to test positive compared to both BAME women (X 2 (1, n = 296) = 7 • 2608, p = • 007) and White men (X 2 (1, n = 1349) = 6 • 7514, p = • 009). Furthermore, positive BAME patients were significantly younger than White patients, with a median age of 55 years compared to 77 years (Mann-Whitney U = 15,155, p < • 001 two-tailed).Of positive patients, 755 were admitted to hospital for further treatment or were tested whilst inpatient. There was no significant difference between BAME and White groups in terms of overall admissions (X2(1, n = 726) = 3 • 0032, p = • 083). BAME inpatients were, however, significantly more likely to be admitted to ITU compared to White inpatients (X2(1, n = 695) = 23 • 977, p < • 0 0,0 01). After age adjustment, black (2 • 97, p = • 010) and sub-continental groups (2 • 43, p = • 087) had much higher odds than the white cohort (1) for ITU admission. Whilst men accounted for just over half (450/806
