(127) Risk factors for severe outcomes among patients hospitalized with COVID-19 in US Medicare between December 2020 and March 2023

Background: Studies have suggested systemic corticosteroids delay SARS-CoV-2 clearance, especially when given earlier in the course of infection, but data on potential harmful outcomes of such use in COVID-19 patients are limited.

Objectives: To evaluate the association between new systemic corticosteroid use and invasive mechanical ventilation (IMV) or endotracheal intubation (ET), intensive care (IC), or death related to COVID-19 hospitalization in the time period when COVID-19 vaccines were available.

Methods: Using Medicare Fee for Service data, community dwelling beneficiaries, aged 66 years or older, with an incident inpatient claim indicating COVID-19 (primary discharge diagnosis), between December 27, 2020, and March 31, 2023 (vaccine era) were identified. Beneficiaries with use of systemic corticosteroids during days -105 to -15 prior to hospital admission (index date) were excluded. A risk factor analysis was conducted examining new use of oral or injectable corticosteroids in the 14 days prior to a COVID-19 hospital admission, as well as several other potential risk factors (demographics, dominant COVID-19 strain at study entry, comorbidities, healthcare utilization, medication use, and COVID-19 vaccination), for IMV or ET, or IC during the index COVID-19 hospitalization and death. Adjusted odds ratios (OR) and 95% confidence intervals (CI) were generated for associations between potential risk factors and outcomes.

Results: Among 165,574 eligible beneficiaries, 85.3% were white, 33.1% had a record of receiving a COVID-19 vaccine at any time prior to the index date, and 20,778 (12.5%) received a systemic corticosteroid in the 14 days prior to their COVID-19 hospitalization. In the 365 days prior to the index date, 40.8% had a record of diabetes, 26.1% of obesity, 81.2% of hypertension, and 19.7% of COPD. Systemic corticosteroid use in the 1-14 days prior to index hospitalization was associated with an increased likelihood of IMV or ET (OR 1.51; 95% CI 1.44-1.59), ICU admission (OR 1.44; 95% CI 1.39-1.50), and death (OR 1.51; 95%CI 1.45-1.58). Other risk factors that increased odds for each outcome were obesity, male sex, Asian race, Hispanic ethnicity and use of immunosuppressants and immunomodulators 1-90 days prior to index hospitalization.

Conclusions: Use of systemic corticosteroids in the 14 days prior to COVID-19 hospitalization was associated with worse outcomes for patients with COVID-19. Of note, however, corticosteroid use may also signify more severe COVID-19 which itself can be associated with worse outcomes. Studies to further investigate this potential association are warranted. As expected, healthcare claims data are limited in capturing COVID-19 vaccine administration.