3541 The association of corticosteroid use with inpatient mortality in acute exacerbation of idiopathic pulmonary fibrosis

Abstract

OBJECTIVES/SPECIFIC AIMS: Objective: To assess the impact of corticosteroid therapy on in-hospital mortality in IPF patients admitted with acute respiratory failure. METHODS/STUDY POPULATION: Methods: Patients with IPF were retrospectively identified in the University of California San Francisco medical center’s electronic health records from January 1, 2010 to June 1, 2018. Cases with IPF were defined as age 50 years or older, having at least two codes one month apart for idiopathic fibrosing alveolitis or post-inflammatory fibrosis (ICD-9 516.3, 516.31 or 515.0 or ICD-10 codes J84.9, J84.10, J84.111 or J84.112), and a subsequent hospitalization for acute respiratory failure or acute respiratory symptoms. The prevalence of pre-selected co-morbidities, clinical events (ICU admission, mechanical ventilation, lung transplantation) and clinical outcomes were assessed. A propensity score model for corticosteroid use was constructed using a multivariable logistic regression with inclusion of corticosteroid-associated demographic and baseline variables (univariate p-value < 0.25). A marginal structural model (MSM) was used to address time-dependent confounding and mediating effects of ICU admission and mechanical ventilation by applying inverse probability weighting for receipt of corticosteroid treatment. Secondary outcome analysis was performed on patients who survived hospital admission. RESULTS/ANTICIPATED RESULTS: Results: A total of 132 patients with IPF and an acute respiratory admission were identified. 48 patients (36%) received corticosteroids during their admission. Applying inverse weighting to time-dependent co-variates (ICU admission and invasive mechanical ventilation) in a MSM, corticosteroid therapy was not associated with risk of in-hospital mortality (odds ratio 1.82; 95% CI, 0.47-6.99; p = 0.39). After adjusting for corticosteroid therapy using a propensity score, corticosteroid therapy remained unassociated with in-hospital mortality (odds ratio 1.53, 95% confidence interval [CI] 0.37, 6.29; p = 0.55). There was no difference in discharge disposition or time to hospital readmission by corticosteroid treatment. There was a possible increase in time to death following discharge in patients receiving corticosteroids (Figure). DISCUSSION/SIGNIFICANCE OF IMPACT: Conclusions: This study suggests that treatment of acute exacerbations of interstitial lung disease with corticosteroids does not improve short-term outcomes, including in-hospital mortality, all-cause non-elective re-hospitalization or death within 6 months of discharge. Further research in larger cohorts is needed to more definitively assess this relationship.