Long-Term Use of Cardiovascular Medications: Identifying and Encouraging Optimal Duration

In the face of increasingly widespread availability and uptake of medications to treat and prevent cardiovascular disease, this dissertation is motivated by two broad research questions regarding the duration of cardiovascular therapy. First: how long should patients continue therapy when there are risks and benefits that may change over time? Second: how can continuation be encouraged when therapies are known to be safe and effective? In Aim 1, we compared exposure definitions of antiplatelet discontinuation versus continuation at 12 months after a drug-eluting coronary stent and associations with ischemic and bleeding events. We found that increasing restrictions on the definition of therapy continuation yielded results consistent with those from the Dual Antiplatelet Therapy trial, in particular through greater compliance with assigned exposure status during follow-up. Our results also suggest the potential for residual confounding by unmeasured characteristics, which for ischemic events may exaggerate effects of continuation, while for bleeding events may attenuate such effects, particularly if providers are monitoring and appropriately discontinuing therapy among patients at higher risk of bleeding. In Aim 2, we compared the long-term effectiveness and safety of prasugrel and clopidogrel. Among patients with acute coronary syndrome at the time of the coronary stent procedure, prasugrel use suggested ischemic benefit without increased risk of bleeding. In contrast, in patients with stable ischemic heart disease only, prasugrel use was not associated with ischemic benefit but was associated with increased risk of bleeding. For both groups, ischemic endpoint results are consistent with trial findings; for the bleeding endpoint, results may reflect differences in baseline bleeding risk. In Aim 3, we evaluated the impact of a pharmacy-based adherence improvement program on cardiovascular and healthcare outcomes. We found that adherence to cardiovascular medications improved by a small but significant amount among synchronized patients, with the largest improvements observed among patients with lowest baseline adherence level. Healthcare resource use decreased significantly, however cardiovascular clinical outcome rates did not differ between synchronized and control patients