Dispensing data captures individual-level use of aspirin for cardiovascular disease prevention, despite availability...

To assess the level of agreement in aspirin use measured by self-report and dispensing data.

We assessed preventive cardiovascular medication use (prescription-only statins and blood pressure-lowering therapy; and aspirin-also available over-the-counter) at baseline in participants in the New Zealand IMPACT trial for whom these medications were prescribed by their general practitioner. A trial nurse not involved in their ongoing health care obtained participants' self-reported aspirin use data. We obtained dispensing data from the national pharmaceutical dispensing database and assessed agreement between the two measures using kappa coefficients.

Of the 513 trial participants, 36% were women, 50% were of Māori ethnicity, and 45% had a history of cardiovascular disease. The level of agreement between self-reported aspirin use and dispensing data was substantial (kappa 0.75, 95% CI 0.69 to 0.82). The level of agreement in aspirin use measured by these two sources of data was similar to that for statin and blood pressure-lowering therapy use, for all participants combined, for subgroups according to ethnicity (Māori and non-Māori) and history of cardiovascular disease.

Despite its availability over-the-counter, aspirin use in patients for whom cardiovascular medications are indicated can be assessed accurately from dispensing data.