Inhaled Corticosteroids and COVID-19 Risk and Outcomes — Findings from a Population-based Cohort Within the SCIFI-PEARL Project

Background: Inhaled corticosteroid (ICS) therapy is an essential therapeutic modality in the optimal management of respiratory disease especially chronic obstructive pulmonary disease (COPD). Few population-based studies exist that have critically evaluated associations between ICS therapy and COVID-19 outcomes; consequently, controversies exist that necessitate further investigations for additional clarity. Such population-based findings with broad generalizability are important, especially for COPD patients who regularly use ICS therapy and are considered at high risk of severe COVID-19.

Objectives: To investigate whether ongoing regular exposure to ICS affects risk, severity, or survival in SARS-CoV-2 infection, using a nationwide linked Swedish population register database.

Methods: From Jan to Dec 2020, we studied two defined Swedish adult populations ≥40 years (general population, and COPD patients) in three different study cohorts (with corresponding index dates): 1. overall cohort (index date 1 Jan 2020), 2. individuals with COVID-19-onset (diagnosis date), and 3. hospitalized COVID-19 patients (admission date). Regular exposure to ICS in all cohorts was defined as ≥3 ICS prescription in the year before the index date. Hazard ratios (HRs) for studied outcomes (COVID-19 onset, hospitalization, ICU admission, or death) related to ICS exposure were estimated using Cox regression models, and confounding was controlled by propensity score methods applying Average Treatment effect in the Treated (ATT) weighting.

Results: After ATT weighting, in the overall cohorts of both the general population and COPD patients, ICS therapy was not associated with an increased onset of COVID-19, hospitalization, ICU admission, or death (all non-significant HR within 1.03-1.49 (general population), and 0.32-1.05 (COPD patients)). In patients who acquired COVID-19, ICS therapy was not associated with hospitalization, ICU admission, or death in both adult populations with a similar pattern non-significant HRs (within 1.0-1.45 (general population), and 0.94-1.12 (COPD patients)). A slight possible increase for ICU admission was observed in the general population hospitalized for COVID-19 who received ICS therapy (H.R. 1.22 [1.05, 1.42]), however, COPD patients did not have an increased risk.

Conclusions: Our findings show that during the pre-vaccination pandemic period neither patients in the general population nor COPD patients receiving regular ICS therapy were at an increased risk or protected from acquiring, being hospitalized or ICU admitted for, or dying of COVID-19. Our findings are thus not supportive of the hypothesized protective role of ICS therapy in COVID-19. Consequently, patients on ICS should continue their therapy as otherwise indicated.