Long-term use of antiplatelet drugs by patients with transient ischaemic attack

To determine the degree of long-term non-persistence to antiplatelet drugs in patients with transient ischaemic attack (TIA) and identify determinants of this drug-use pattern.

We used community-based prescription registry data to determine antiplatelet drug use in TIA patients presenting to a Danish neurology department in the period 2006-2010. Non-persistence was defined as failure to present a prescription for antiplatelet drugs within 180 days after the dosage of a previous prescription had run out. We used Cox regression to calculate the hazard ratio (HR) for non-persistence and the corresponding 95 % confidence interval (CI) by potential determinants, including a stroke risk score (ABCD2 score). Adherence during follow-up [80 % medication possession ratio (MPR80)] was calculated for antiplatelets, statins and antihypertensive drugs.

The cohort comprised 594 (84 % evaluated as in-patients) TIA patients. During follow-up (median 1.7 years, interquartile range 0.9-3.0 years), 140 (23.6 %) patients became non-persistent. Non-persistence was associated with younger age (<55 years: HR 1.9, 95 % CI 1.3-2.8) and delay between TIA onset and neurological evaluation (7+ days: HR 2.0, 95 % CI 1.0-4.1). Among admitted patients, a higher ABCD2 score (4+: HR 1.3, 95 % CI 0.8-2.1) was also indicative of non-persistence. Non-persistent users were less adherent to other preventive medication (MPR80: statins 31.8 vs. 75.3 %, p value < 0.001; antihypertensives 64.3 vs. 79.5 %, p value: 0.02) than persistent users.

Long-term antiplatelet non-persistence was most pronounced in patients of younger age, those with delayed evaluation of symptoms and those at greater risk of stroke. It was also associated with a lower adherence to preventive medication in general.