Association between immunosuppressive medications and COVID-19 hospitalisation and death: a retrospective cohort study

Importance: Immunocompromised status is a risk factor for severe SARS-CoV-2 infection. Little is known about how systemic corticosteroid dose and concurrent use of immunosuppressants are associated with COVID-19 outcomes.

Objective: To assess the association between corticosteroid dose/duration and concurrent immunosuppressant use on COVID-19 hospitalisation and death in the era of COVID-19 vaccinations.

Design: This is a retrospective cohort study using a deidentified insurance claims database from 1 July 2020 to 30 June 30, 2022, with the risk period starting on 1 July 2021. Impact of corticosteroid exposures and concurrent use of other immunosuppressants was assessed with attributable risk analysis and Cox regression that included COVID-19 vaccination status and time-updated dichotomous immunosuppressive medication exposures.

Participants: There were 10 109 596 eligible patients enrolled during the risk period, each with at least 365 days of continuous enrolment prior to 1 July 2021.

Exposures: Systemic corticosteroids, disease-modifying antirheumatic drugs (DMARDs), tumour necrosis factor-alpha inhibitors (TNFis) and other immunosuppressive drug categories.

Main outcomes: Incidence rate ratios and hazard ratios for COVID-19 hospitalisation and death.

Results: Corticosteroids were prescribed to 1 379 049 (13.6%) of 10 109 596 individuals. After adjustment, corticosteroids were associated with an increased risk of COVID-19 hospitalisation (HR: 5.40; 95% CI 5.27 to 5.53; p<0.0001) and death (HR: 5.90; 95% CI 5.59 to 6.22; p<0.0001). Among individuals exposed to corticosteroids without a record of COVID-19 vaccination, risks for COVID-19 hospitalisation and death were increased by 3- and 14.5-fold. The population attributable risk of corticosteroid use for COVID-19 hospitalisations was 13.9% (95% CI 13.5 to 14.3%). There was a significantly increased risk of COVID-19 hospitalisation associated with the use of corticosteroids plus DMARDs (HR: 1.55; 95% CI 1.42 to 1.70; p<0.0001) or plus TNFis (HR: 1.60; 95% CI 1.15 to 2.22; p=0.005).

Conclusions: Corticosteroids are associated with greater risk of COVID-19 hospitalisation and death, especially among unvaccinated individuals. Concurrent use of DMARDs and TNFis with corticosteroids confers greater risk.