Effects of corticosteroids in hospitalized patients with Legionella pneumonia: A retrospective cohort study

Introduction: Legionella pneumophila is an important cause of pneumonia, however there is scant literature assessing the therapeutic benefit of corticosteroids in treatment. We sought to investigate the association between corticosteroid use and in-hospital mortality for patients hospitalized with Legionella pneumonia.

Methods: Data was collected retrospectively from January 2012 to July 2019 at a 705 bed hospital in New York City. Patients were included if they received a positive Legionella test. Exclusion criteria included age <18, concurrent immunosuppression, and HIV diagnosis. We assessed the relationship between corticosteroid use and in-hospital mortality. Statistical analyses were performed in RStudio.

Results: The study included 160 patients, among which 32 (20%) received steroids. Overall mortality was 7.5% (12.5% among steroid recipients, 6.2% among controls). 25% of patients were admitted to the ICU (37.5% among steroid recipients, 21.9% among controls). Adjusted analysis showed steroid recipients did not have significantly different mortality (aOR = 2.56, p = 0.436). Steroid use was not significantly associated with longer LOS (p = 0.22). Steroid use was significantly associated with hyperglycemia (aOR = 2.91, p = 0.018) and GI bleed (OR = 9.0, p = 0.014).

Conclusions: We found that in patients hospitalized with Legionella pneumonia, corticosteroid administration was not significantly associated with longer hospitalization or mortality. All findings held true when adjusting for known predictors of pneumonia severity. Corticosteroid use was associated with increased rates of hyperglycemia and GIB requiring blood transfusion. The results of this study are consistent with guidelines recommending against routine use of corticosteroids in CAP.