Introduction: Adherence with controller medication is a major challenge in asthma management. Thus, a reliable method of measurement is mandatory to assess adherence.

Aim: To examine the test-retest reliability on adherence with inhaled corticosteroids in adults with asthma using, a self-reported adherence score (Foster score).

Methods: Patients with asthma and >1 routine follow-up appointment at a university hospital outpatient clinic reported Foster scores. The objective Medication Possession Ratio (MPR) was calculated based on pharmacy redemption data and physician-prescribed doses of inhaled corticosteroids. The difference between Foster score and MPR at the first and second visit was assessed using a Bland-Altman plot, outcomes reported as limits of agreements and bias. Foster scores from both visits were used to calculate an intraclass correlation coefficient (ICC).

Results: Self-reported adherence with asthma controller medication measured by Foster score was significantly higher than the objective MPR (p < 0.0001). The Bland-Altman plot for MPR and Foster score at the first and second visit showed upper and lower limits of agreement of 83.5 - (-1.6) and 80.9 - (-6.9) and bias was 41.0 and 37.0, respectively. Of the included patients, 93.1% reported identical Foster scores between visits, resulting in an excellent ICC of 0.92. Absolute median difference between Foster scores and MPR at first and second visit was 8.7 percentage points (p = 0.049).

Conclusion: Foster score shows an excellent ICC; however, its poor agreement with objective measures of adherence suggests that clinicians should not rely on Foster