Agreement between pharmaceutical claims data and patient-reported medication use after stroke

Abstract

Objectives: To assess the agreement between pharmaceutical claims data and patient-reported medication use after stroke.

Methods: Claims data from the Pharmaceutical Benefits Scheme were used to estimate medication use for a subset of participants registered in the Australian Stroke Clinical Registry in 2016. Estimates on medication use were validated against patient-reported responses (considered the reference standard).

Key findings: For antihypertensive and lipid-lowering medications, the sensitivity of claims data was excellent (85-87%) and the specificity was good (73-78%). Whereas for antithrombotic medications, sensitivity was modest (61%), but specificity was excellent (85%).

Conclusions: Pharmaceutical claims data can be used to infer medication use after stroke with mostly good to excellent sensitivity and specificity compared with the patient report.

Keywords: administrative claims; healthcare validation study; pharmacoepidemiology; registries; stroke.