INTRODUCTION
In Cystic Fibrosis (CF) lung disease, inflammation due to bacterial colonization and neutrophil recruitment increases during an Acute Pulmonary Exacerbation (APE) 1. Typical treatment of moderate to severe APE supported by the evidence based CF pulmonary guidelines include intravenous (IV) antibiotics in the hospital setting, maintenance of chronic therapies for lung health, and more frequent administration of airway clearance therapy 2. Despite these treatments, as many as 25% of patients fail to recover to their baseline FEV1 after treatment 3.
Corticosteroid therapy and its potent anti-inflammatory property has long been postulated to be of benefit in CF either by preventing decline in baseline FEV1 percent predicted (FEV1pp) or reversing decline associated with APE. CF pulmonary guidelines for maintenance of lung health report that there is insufficient evidence to recommend routine use of oral corticosteroids in children with CF due to “net-negative” effects 4.
In clinical practice, steroid use remains variable. Across 38 CF centers treating hospitalized pediatric patients with CF, steroid treatment use ranged from 3-61% during APE 5. In the STOP trial, a prospective observational study that assessed APE treatment practices at 11 CF centers, 21% of the 220 enrolled patients were given corticosteroids as adjunct therapy 6. Our pediatric CF center providers have an informal practice of five to seven day “rescue” steroid treatment when standard treatment fails to demonstrate expected improvement in FEV1pp following at least one week of guideline driven hospital based therapy. We hypothesized that use of oral corticosteroids during failed inpatient APE management would increase FEV1pp at discharge and at hospital follow-up visit. We also wanted to understand when and why providers engage in this approach. To evaluate our hypotheses, we conducted a retrospective study examining a cohort of CF patients and recovery to their baseline FEV1pp after hospital admission for APE, FEV1pp at follow up visit, and time to next APE compared to routine treatment. We also employed a propensity score matching scheme to more fairly estimate treatment effect by controlling specific covariates.