How many blood eosinophil measurements are needed to identify COPD patients eligible for biologic therapy? A multi-source decision analysis simulation

Objective Biologic therapies for eosinophilic COPD require a blood eosinophil count (BEC) ≥300 cells/µL for eligibility. Landmark trials used a single screening measurement, yet BEC exhibits substantial within-patient variability. The study aimed to evaluate the optimal number of measurements to reliably detect the eosinophilic phenotype in COPD. Methods We developed a decision-analytic simulation model using pooled estimates of BEC variability from independent COPD cohorts, with intraclass correlation coefficients (ICCs) pooled using Fisher's z transformation. Persistence data from five cohorts - ECLIPSE, COSYCONET, CHAIN, BODE, and a Swedish cohort - comprising a total of 3347 patients, were used to estimate the probability of detecting BEC ≥300 cells/µL at least once according to the number of serial measurements. Results Using the ≥300 cells/µL threshold, the pooled COPD population was classified as persistent eosinophilia, 9.2%; intermittent eosinophilia, 32.5%; and never eosinophilic, 58.3%. Among patients with intermittent eosinophilia, one measurement detected BEC ≥300 cells/µL in 43% of cases. Two measurements at 6-month intervals increased detection to 67%, three measurements over 12 months to 81%, and five measurements over 24 months to 94%. Weighted pooled ICCs declined from 0.784 at ≤3 months to 0.570 at ≥36 months, supporting greater variability over longer intervals. Conclusions Intermittent eosinophilia affects approximately one-third of patients with COPD and may be missed by single-BEC screening. In this model, three measurements over 12 months achieved ≥80% detection. Integrating historical and prospective BEC values may improve biologic eligibility assessment without delaying treatment decisions.