Use of corticosteroids in nonsevere community acquired pneumonia a propensity score matching study

Abstract Community acquired pneumonia (CAP) is a major cause of morbidity and mortality. Although corticosteroids have shown benefits in severe CAP, their usefulness in non severe cases remains uncertain. This study assessed the association between systemic corticosteroids and 30-day mortality in adults hospitalized with non severe CAP. A multicenter retrospective cohort was conducted including adults hospitalized with non severe CAP between 2010 and 2020 in Colombia. Patients whose index episode was nosocomial pneumonia, as well as those with immunosuppression or connective tissue disease and those requiring vasopressors or mechanical ventilation, were excluded. Propensity score matching (1:1) was applied, and survival was analyzed using Kaplan Meier curves, including evaluations by corticosteroid type, dose and duration. A total of 2821 patients were included, with an overall 30-day mortality of 7.6%. Of these, 507 patients (18.0%) received corticosteroids, with a mortality rate of 10.4%, whereas the remaining 2314 patients who did not receive corticosteroids had a mortality rate of 7.0%. The propensity score-matched cohort included 448 patients (224 in each group). In the matched analysis, 30-day mortality was 14.3% among patients who received corticosteroids compared with 10.7% among those who did not. Before matching, patients treated with corticosteroids were older and had a higher burden of respiratory symptoms and comorbidities. After matching, baseline characteristics were well balanced between groups. After propensity score matching, no significant association was observed between corticosteroid use and 30-day mortality (average treatment effect on the treated [ATET] 0.031, 95% CI − 0.040 to 0.102; p = 0.389). The ATET estimates the absolute risk difference; an ATET of 0.031 corresponds to an absolute difference of 3.1 percentage points in 30-day mortality. Additionally, no significant differences in survival were found according to corticosteroid type, dose, or treatment duration. In adults with non-severe CAP, the use of systemic corticosteroids did not demonstrate a statistically significant difference in 30-day mortality.