RESUMEN
BACKGROUND: Corticosteroids may be beneficial in a subset of patients with coronavirus disease 2019 (COVID-19), but predictors of therapeutic response remain unknown. C-reactive protein (CRP) is a routinely measured biomarker, and reduction in its levels after initiation of therapy may predict inpatient mortality. METHODS: In this retrospective cohort study, the charts of patients who were admitted to Montefiore Medical Center between March 10, 2020, and May 2, 2020 for the management of COVID-19 were examined. Of all patients who met inclusion criteria, patients who received corticosteroid treatment were categorized as CRP responders (≥50% CRP level reduction) and CRP nonresponders (<50% CRP level reduction) based on change in CRP within 72 hours of corticosteroid treatment initiation. The outcomes of interest were two-fold: (1) CRP response after treatment with corticosteroid, and (2) differences in mortality among patients with CRP response compared those without. RESULTS: Of 2,707 patients admitted during the study period, 324 received corticosteroid treatment. Of patients who received corticosteroid treatment, CRP responders had reduced risk of death compared with risk among CRP nonresponders (25.2% vs 47.8%; unadjusted odds ratio [OR], 0.37; 95% CI, 0.21-0.65; P <.001). This effect remained strong and significant after adjustment for potential confounders (adjusted OR, 0.27; 95% CI, 0.14-0.54; P <.001). CONCLUSION: Reduction in CRP by 50% or more within 72 hours of initiating corticosteroid therapy potentially predicts inpatient mortality. This may serve as an early biomarker of response to corticosteroid therapy in patients with COVID-19.