Ethnic disparities in hospitalisation and hospital-outcomes during the second wave of COVID-19 infection in east London

Abstract

It is unclear if changes in public behaviours, developments in COVID-19 treatments, improved patient care, and directed policy initiatives have altered outcomes for minority ethnic groups in the second pandemic wave. This was a prospective analysis of patients aged ≥ 16 years having an emergency admission with SARS-CoV-2 infection between 01/09/2020 and 17/02/2021 to acute NHS hospitals in east London. Multivariable survival analysis was used to assess associations between ethnicity and mortality accounting for predefined risk factors. Age-standardised rates of hospital admission relative to the local population were compared between ethnic groups. Of 5533 patients, the ethnic distribution was White (n = 1805, 32.6%), Asian/Asian British (n = 1983, 35.8%), Black/Black British (n = 634, 11.4%), Mixed/Other (n = 433, 7.8%), and unknown (n = 678, 12.2%). Excluding 678 patients with missing data, 4855 were included in multivariable analysis. Relative to the White population, Asian and Black populations experienced 4.1 times (3.77-4.39) and 2.1 times (1.88-2.33) higher rates of age-standardised hospital admission. After adjustment for various patient risk factors including age, sex, and socioeconomic deprivation, Asian patients were at significantly higher risk of death within 30 days (HR 1.47 [1.24-1.73]). No association with increased risk of death in hospitalised patients was observed for Black or Mixed/Other ethnicity. Asian and Black ethnic groups continue to experience poor outcomes following COVID-19. Despite higher-than-expected rates of hospital admission, Black and Asian patients also experienced similar or greater risk of death in hospital since the start of the pandemic, implying a higher overall risk of COVID-19 associated death in these communities.

Figure 1

Cox-proportional Hazard analysis included covariates: Age, Sex, Hypertension, Diabetes, Chronic Kidney Disease, Smoking history, Obesity, IMD (quintiles of local study population) for survival over 30 (A) and 90 (B) days in the second pandemic wave ETHICAL cohort.

Figure 2

Predicted survival over 90-days by Ethnicity in a 65-year-old Male living in IMD-3 without Diabetes, CKD, Smoking History, Hypertension or Obesity. Based on second wave Cox-proportional Hazard analysis included covariates: Age, Sex, Hypertension, Diabetes, Chronic Kidney Disease, Smoking history, Obesity, IMD (quintiles of local study population).

Figure 3

Predicted survival by Ethnicity in a 65-year-old Male living in IMD-3 without Diabetes, CKD, Smoking History, Hypertension or Obesity based on 12-month follow-up of the first pandemic wave ETHICAL population in a Cox-proportional Hazard analysis stratified by ethnicity. Included covariates: Age, Sex, Hypertension, Diabetes, Chronic Kidney Disease, Smoking history, Obesity, IMD (quintiles of local study population). Excess mortality associated with Asian ethnicity has persisted over follow up while early survival disadvantage associated with Black ethnicity has attenuated with longer follow-up. Other/Mixed group omitted for clarity.

Figure 4

Numbers of first admissions to Barts Health hospitals in residents of the London Boroughs of Tower Hamlets, Newham, and Waltham Forest aged ≥ 16, grouped by age and ethnicity for the first and second waves of COVID-19 (panels A & B) or any acute admissions during 2013–18 (panel C). Age and Ethnicity distribution of the Tower Hamlets, Newham, and Waltham Forest population in the 2011 UK Census is shown for comparison (panel D).