Suspected COVID-19 in primary care: how GP records contribute to understanding differences in prevalence by ethnicity.

Abstract Background The first wave of the London COVID-19 epidemic peaked in April 2020. Attention initially focussed on severe presentations, intensive care capacity, and the timely supply of equipment. General practice has seen a rapid take up of technology to allow virtual consultations, enabling the management of mild and moderate community cases. Aim To quantify the prevalence and time-course of suspected COVID-19 presenting to general practices during the London epidemic. To report disease prevalence by ethnic group, and explore how far differences by ethnicity can be explained by data in the electronic health record (EHR). Design and Setting Cross-sectional study using anonymised data from the primary care records of 1.3 million people registered with 157 practices in four adjacent east London clinical commissioning groups (CCGs). The study area includes 48% of people from ethnic minorities and is in the top decile of social deprivation in England. Method Suspected COVID-19 cases were identified using SNOMED codes. Explanatory variables included age, gender, self-reported ethnicity and measures of social deprivation. Clinical factors included 16 long-term conditions, latest body mass index and smoking status. Results There were 8,985 suspected COVID-19 cases. Ethnicity recording was 78% complete. Univariate analysis showed a two-fold increase in odds of infection for South Asian and Black adults compared to White. In a fully adjusted analysis, including clinical factors, the odds were: South Asian OR 1.93 (95% CI = 1.83 to 2.04) Black OR 1.47 (95% CI 1.38 to 1.57) Conclusions Using data in GP records Black and south Asian ethnicity remain as predictors of community cases of COVID-19, with levels of risk similar to hospital admission cases. Further understanding of these differences requires social and occupational data.