Assessing Factors Associated with COVID19 Risk in Asthma

Abstract

RATIONALE Several factors may be associated with the apparent reduced risk of severe COVID-19 disease in asthma patients. These include a proposed protective effect from inhaled corticosteroids (ICS) or from a type-2 immune airway response. Large UK population cohorts have found conflicting results, but each had missing information on asthma phenotype, including exacerbation history or markers of type-2 inflammation. Protective factors may be determined by comparing the outcomes of asthma patients to other patients using ICS or patients that have similar inflammation. This will be done using a national primary care database linked to secondary care data and Public Health England (PHE) COVID-19 data. METHODS From a primary care database covering around 15% of the UK population, we drew three cohorts of adults (>17 years) with asthma, chronic obstructive pulmonary disease (COPD) or allergic rhinitis, and a general population cohort matched by age, gender and primary care practice. COVID-19 infection was defined using three sources: primary care records (confirmed and suspected cases), hospital admissions or PHE data from their surveillance system. Preliminary analysis calculated the proportion of patients diagnosed as suspected or confirmed COVID-19 in primary care between 1st February and 22nd June 2020. We used multivariable logistic regression to assess associations between risk factors and COVID-19. RESULTS We identified 729,045 patients with asthma, 280,892 with COPD, 137,312 with allergic rhinitis and 1,138,018 in the general population cohort. 2.2% of the asthma cohort, 3.5% the COPD cohort, 2.4% of the allergic rhinitis cohort and 1.1% of the general population, were identified as having either suspected or confirmed COVID-19 in primary care (approximately a quarter were confirmed). In the asthma cohort, the following were found to be independently significantly associated with having suspected/confirmed COVID-19;older age, being male, diabetes, cardiovascular disease, obesity, sleep apnoea, bronchiectasis, COPD, ILD, chronic renal failure, cerebrovascular disease, dementia and liver disease (all p<0.0001). A history of past or current smoking and a high eosinophil count were not significantly associated. CONCLUSIONS These preliminary findings suggest patients with asthma are more likely to consult primary care about COVID-19 than a matched general population, but less likely than COPD patients, and that type-2 inflammation may not be associated with an increased risk of COVID-19 consultation with a primary care physician. Analysis comparing to the other cohorts and the general population, and using data from secondary care and PHE, is ongoing.