The impact of data source on real-world medication adherence and exposure measures: From prescription to sold

Adherence measurement is not fixed but shaped by data source and methodology. Both timestamp choice and gap-handling substantially influence adherence estimates and patient classification. Our dynamic, patient-specific approach to gap adjustment yielded stable metrics, and exposure-based measures remained more source-sensitive. Standardizing and transparently reporting adherence definitions is essential to avoid misclassification. Future work should evaluate the prognostic value of gap-adjusted vs exposure PDC in predicting cardiovascular outcomes to guide more meaningful quality measures and value-based care.