Purpose: Current understanding of the relationship between COPD phenotype and health care resource utilization (HCRU) is limited. This real-world study evaluated disease burden and HCRU for COPD subgroups prone to exacerbation as defined by blood eosinophil (EOS) count and multiple inhaler triple therapy (MITT) use.
Methods: This was a large-scale, retrospective, longitudinal, observational cohort study using data from the US IBM Watson Explorys real-world database (GSK Study HO-17-18395). The population of interest comprised patients with COPD ≥40 years of age with ≥2 moderate or ≥1 severe exacerbations (prior year) while on inhaled maintenance therapy, with ≥1 blood EOS count. Data were analyzed during the year prior to index date (last COPD encounter between January 1, 2011 and December 31, 2016). Four subgroups were analyzed based on a combination of EOS counts (<150 and ≥150 cells/μL) and MITT use (receiving or not receiving). Among these groups, clinical characteristics, exacerbations, and HCRU were described. A sensitivity analysis that further stratified EOS into four categories (<150, ≥150-<300, ≥300-<500, and ≥500 cells/μL) was also performed.
Results: The COPD population of interest comprised 34,268 patients. Subgroups with EOS ≥150 cells/μL vs <150 cells/μL had more comorbidities and experienced significantly higher mean numbers of moderate exacerbations (not receiving MITT, ≥150 cells/μL vs <150 cells/μL: 1.93 vs 1.82, P<0.0001; receiving MITT 2.26 vs 2.16, P=0.0062) and COPD-related emergency visits (not receiving MITT, ≥150 cells/μL vs <150 cells/μL: 3.0 vs 2.5, P<0.001; receiving MITT 3.4 vs 3.1, P=0.0011). Increasing EOS category was associated with higher HCRU.
Conclusion: Blood EOS ≥150/μL cells were associated with increased HCRU and higher exacerbation rates compared with EOS <150 cells/μL, irrespective of MITT use. COPD phenotyping using blood EOS could help identify candidates for additional therapies that target eosinophilic inflammatory pathways.
Keywords: COPD; eosinophil; exacerbation; health care resource; phenotype.