Comparisons of healthcare utilization with different methods to define eosinophil elevation in the us asthma population

Document Type

Conference Proceeding

Publication Date


Publication Title

Am J Respir Crit Care Med


Rationale: Blood eosinophil elevation has a direct correlation with increased resource use and cost in asthma. However, little guidance has emerged to define how to evaluate eosinophil elevation in real-world studies.

Objective: To compare healthcare resource use and direct medical costs between different operational definitions of eosinophil elevation in the US asthma population. Methods: In this retrospective cohort study, data were extracted from a Midwest-focused health system component of EMRClaims+®, which included patient EMR data linked with insurance claims (2012-2016). Patients ≥18 years old on the date of asthma diagnosis recorded in EMR (ICD 9 493.xx or ICD 10 J45), having at least 12 months of enrollment after the asthma date (“assessment” period) and 12 months of enrollment after assessment period (“follow-up” period) were included. Patients were required to have at least one eosinophil test conducted during the assessment period. Eosinophil “elevation” was defined using 2 thresholds: ≥150 cells/μL and ≥300 cells/μL. Five discrete operational definitions based on timing and consistency of test results on a per-patient level were used, elevation noted in: 1) at least one test at any time during assessment period; 2) mean test result during assessment period; 3) the most recent test to follow-up period; 4) any test within 3 months before follow-up period; and 5) any test within 6 months before follow-up period. Demographics, comorbidities, resource use (hospitalizations, ER visits, outpatient visits) and cost during follow-up were compared between these definition cohorts.

Results: The study included 2126 patients. Using Definitions 1-3, no significant differences were observed between elevated and not elevated eosinophil counts for follow-up resource use or cost. Using Definition 4, there was a significantly greater percentage of patients with follow-up hospitalizations in the elevated eosinophil group (≥150 cells/μL: 16% vs. 10%, p=0.001; ≥300 cells/μL: 17% vs. 11%, p=0.011). Similar result was noted for Definition 5 but only at the ≥150 threshold (≥150 cells/μL: 15% vs. 10%, p=0.0004). The elevated eosinophil group had greater mean number of hospitalizations with definitions 4 and 5 at ≥150 cells/μL (0.23 vs. 0.13, p=0.0004; 0.20 vs. 0.13, p=0.0004, respectively) as well as greater mean hospitalization cost ($2,254 vs. $1,442, p=0.001; $2,134 vs. $1,332, p=0.001, respectively).

Conclusion: Patients with eosinophil counts ≥150 cells/μL from individual blood test results taken within the most recent 3- to 6-month period accounted for greater annual resource use and cost, indicating the importance of eosinophil elevation definition.