Background: Despite intensive pharmacotherapy, a considerable number of patients with severe asthma have inadequate disease control. Patients with severe asthma who experience exacerbations consume significant health care resources.

Objective: To assess health care resource utilization and associated costs among patients with persistent severe asthma who experienced exacerbations compared with patients with persistent but nonsevere asthma.

Methods: This retrospective analysis of a national administrative claims database identified patients aged ≥ 12 years who had at least 1 medical claim with an asthma diagnosis in 2012 and had continuous medical and pharmacy coverage under a commercial or Medicare Advantage plan from January 1, 2012, to December 31, 2013. Patients were assigned to 1 of 2 mutually exclusive cohorts-persistent asthma (PA) or severe asthma (SA)-according to an established algorithm based on asthma-related health care resource use and pharmacy claims for controller medication. SA patients were required to meet PA criteria and also have evidence of ≥2 asthma exacerbations in 2012. Asthma-related health care resource utilization and costs were computed from asthma medication use (rescue and controller therapy) and medical claims with an asthma diagnosis in the primary position in 2012 and 2013. Adherence to controller therapy was assessed over 365 days by using the proportion of days covered (PDC), starting with the first claim for controller therapy in 2012. Differences between the PA and SA cohorts were analyzed by t-test for continuous variables and chi-square test for categorical variables. Asthma-related costs in 2013 were also analyzed using a generalized linear model with a gamma distribution and log link, adjusted for patient demographics (age, gender, region, and insurance type) and Quan-Charlson comorbidity score.

Results: A total of 65,359 patients were included: 63,597 (97.3%) PA patients and 1,762 SA patients (2.7%). Compared with the PA cohort, the SA cohort was older (mean age = 50.8 years vs. 46.5 years, P < 0.001) and had higher mean comorbidity score (1.47 vs. 1.31, P< 0.001). The mean count of all asthma medications fills was 2.2-fold (2012) and 2.1-fold (2013) higher in the SA cohort, compared with the PA cohort (P< 0.001). Mean PDC for all oral and inhaled controller therapy was also higher in the SA cohort compared with the PA cohort (0.80 vs. 0.65, P< 0.001). SA patients had a significantly greater mean count of asthma-related hospitalizations, emergency room visits, and ambulatory visits in 2012 and 2013 (P< 0.001). Unadjusted mean annual asthma-related costs in the SA versus PA cohorts were $6,496 versus $2,739 (P < 0.001) in 2012 and $5,174 versus $1,775 (P< 0.001) in 2013. Higher asthma-related costs were driven by greater mean annual asthma medication costs in 2012 ($4,545 vs. $1,738, P< 0.001) and 2013 ($4,068 vs. $1,348, P< 0.001). Adjusted mean annual asthma-related costs in 2013 were $3,336 greater (cost ratio=2.878, P< 0.001) in the SA cohort, and adjusted mean annual asthma medication costs were $2,672 higher (cost ratio=2.982, P< 0.001) in the SA cohort.

Conclusions: Patients with SA who experienced 2 or more exacerbations had 2.1-fold greater use of controller medications across both study years and were more adherent to controller therapy than patients with PA. Despite more intensive pharmacotherapy, SA patients incurred 2.9-fold higher adjusted asthma-related costs and 3-fold higher adjusted asthma medication costs than PA patients. Patients with SA consistently demonstrated a higher rate of health care utilization.

Disclosures: Funding for this study (HO-14-14443) was provided by GlaxoSmithKline (GSK). All listed authors meet the criteria for authorship set forth by the International Committee for Medical Journal Editors. Albers, Forshag, and Yancey are employees of GSK and hold stock in GSK. Dalal, Nagar, and Ortega were employees of GSK at the time this research was conducted. Chastek and Korrer are employees of Optum, which received consulting fees from GSK for research related to this study. Study concept and design were contributed by Chastek, Nagar, and Dalal. Korrer took the lead in data collection, along with Chastek, and data interpretation was performed by Chastek, Ortega, Forshag, and Dalal. The manuscript was written by Chastek and Dalal and revised by Albers and Yancy, assisted by the other authors.