Percutaneous coronary intervention (PCI) is a common invasive cardiovascular procedure performed in the VA with over 13,000 procedures in FY10. Clopidogrel is a critical adjuvant therapy following PCI with stent placement and is generally recommended for up to 1 year following the procedure. Despite the evidence supporting clopidogrel use, studies both outside and within the VA suggest that poor adherence to clopidogrel is common. However, prior interventions targeting non-adherence have not specifically focused on clopidogrel adherence among PCI patients.

There are many potential reasons for early clopidogrel discontinuation that involve patient and healthcare system factors. Patients reported the following reasons for discontinuing clopidogrel within 1 month after drug-eluting stent (DES) implantation: 1) misunderstanding the intended treatment duration; 2) conflicting recommendations about intended duration; 3) cost of the medication; and 4) patients' own decision to stop. In contrast, patients who continued to take clopidogrel reported the following as helpful: 1) communication such as letters from their physician; and 2) receiving specific instructions on clopidogrel use. These findings suggest that there are specific interventions that can be implemented to improve clopidogrel adherence.

Multi-modal interventions that incorporate frequent follow-up, especially with pharmacists and use interactive voice response (IVR) technology have improved medication adherence. IVR technology is a computer-based telephone system which initiates calls, receives calls, provides information, and collects data from users. IVR is currently a mainstay in the VA where patients frequently interact with these automated systems to get clinic appointments and/or refill prescriptions. IVR as part of multi-modal interventions have been well received by patients, increased adherence to medications (e.g., statins), and improved clinical outcomes (e.g., blood pressure, diabetes symptoms, health status). In addition, the investigators have successfully used IVR as part of a multi-modal, multi-site intervention including pharmacists to improve blood pressure levels among hypertensive patients. Accordingly, the investigators have designed the intervention to improve clopidogrel adherence that builds on the investigators' prior work and other successful adherence interventions from the literature.

The investigators propose a hybrid effectiveness-implementation study of a multi-faceted intervention to improve clopidogrel adherence at VA PCI centers. The investigators will use the VA's Cardiovascular Assessment Reporting and Tracking (CART-CL), a uniform cath lab procedure reporting tool at all VA cath labs. The intervention consists of 4 components: a) an alert from CART-CL will be sent to an inpatient pharmacist prior to discharge that a patient has received a stent; b) a pharmacist will bring clopidogrel to the patient's bedside prior to hospital discharge as well as educate the patient on the importance of and adherence to clopidogrel following PCI; c) interactive voice response (IVR) calls will be made to patients prior to the time of clopidogrel refill to remind patients and to facilitate refills during follow-up; and d) a Patient Aligned Care Team (PACT) member will contact patients who delay filling clopidogrel.

I. Hypotheses and Specific Aims

The main objective is to conduct a type 1 hybrid effectiveness/implementation study to test the effectiveness of a successfully-piloted, evidence-based, multi-faceted intervention to improve patient adherence to clopidogrel following PCI. The proposed study will test the hypothesis that a successfully-piloted, evidence-based, multi-faceted intervention targeting Veterans following PCI procedure improves adherence to clopidogrel, reduces bleeding, myocardial infarction, stroke, and mortality among these patients, and is cost-effective. The proposed intervention will be based on the Chronic Care Model and will build on the investigators' pilot work as well as will leverage the VA's CART-CL, a uniform cath lab procedure reporting tool at all VA cath labs. The study will be evaluated using the REAIM framework consisting of the following components, reach, effectiveness, adoption, implementation and maintenance. The intervention will adapt elements of prior successful intervention, including patient education, collaboration between cath lab clinicians and impatient/outpatient pharmacy teams, and telemonitoring via interactive voice response (IVR) technology. This type I hybrid effectiveness/implementation study will be tested at 16 sites randomized to intervention with an average of 90 patients per site per 6 month period versus usual care.

Specific Aims:

1. To evaluate current practices at VA PCI facilities (n=20) to enhance adherence to clopidogrel, both at hospital discharge and during longitudinal follow-up.
2. To implement the multi-faceted intervention at 16 total sites through 4 roll-out phases (4 sites during each roll-out) in a randomized stepped wedge trial design.
3. To assess barriers and facilitators to intervention implementation during each roll-out phase through semi-structured interviews and incorporate lessons learned from each roll-out phase into subsequent roll-out phases.
4. To determine the effectiveness of a successfully-piloted, evidence-based, multi-faceted intervention versus usual care for improving clopidogrel filling at hospital discharge and adherence to clopidogrel (primary outcomes).
5. To determine the effectiveness of a successfully-piloted, evidence-based, multi-faceted intervention versus usual care for reducing the combined cardiovascular endpoints of bleeding, myocardial infarction, stroke, and mortality (secondary outcomes).
6. To assess the incremental cost-effectiveness of the successfully-piloted, evidence-based, multi-faceted intervention compared with usual care.

II. Background and Significance:

The investigators have successfully piloted the intervention to improve clopidogrel adherence following PCI. The investigators developed a software application integrated within CART-CL that assessed whether patients fill their clopidogrel prescription at hospital discharge. If clopidogrel was not filled, patients were contacted to identify barriers to medication filling. During follow-up, automated telephone calls were sent to patients to educate them about the importance of medication adherence and to remind them to refill clopidogrel prior to the refill due date. The software application successfully identified all 15 patients enrolled in the pilot who underwent PCI and whether they filled clopidogrel at hospital discharge at 2 VAMCs. In qualitative interviews about the automated calls, all patients indicated that the refill reminder messages were helpful. A majority of participants indicated that they felt the education/support from the messages or participation in the study was helpful or would be helpful to others. A few patients shared that the messages also supported refilling other medications and reminding them to ask clarifying questions about other medications. Following the successful pilot, the investigators' next step is to test the effectiveness of the intervention and study the implementation process across a range of VA facilities.

Building on the prior work and evidence from the literature, the investigators have developed the intervention informed by the Chronic Care Model (CCM) which is a framework that uses clinical information systems to facilitate evidence-based quality improvement. The investigators will leverage CART-CL, IVR technology, patient self-management, and team-based care to foster efficient, productive interactions between activated patients and proactive clinical teams to improve clopidogrel adherence. Further the components of the intervention directly address many of the reasons that patients have highlighted as leading to early clopidogrel discontinuation.

This study will refine the current state of knowledge on improving medication adherence in multiple ways. First, it combines multiple interventions that have been separately shown to be effective in improving medication adherence and addresses causes of clopidogrel non-adherence identified by patients. Second, the intervention focuses on a novel setting (i.e., patients discharged following PCI and transitioning to outpatient care) in contrast to prior adherence interventions that have focused only on patients with stable chronic diseases (e.g., hypertension). Prior work by the investigators' group and others suggest that the immediate post-ACS period, and more generally transitions from the inpatient to the outpatient setting, are particularly 'vulnerable periods' for medication adherence. Yet prior interventions have not specifically targeted this setting. Third, the targeted medication in this study has demonstrated short-term benefits and where non-adherence can have immediate adverse outcomes (e.g., stent thrombosis). Prior studies have focused on medications (e.g., hypertension medications) where non-adherence leads to problems longer term rather than the short term. Fourth, the study utilizes existing resources (i.e., cardiac data systems integrated with the VA electronic health record, VA pharmacists and patient-aligned care teams) to implement the intervention, improving the feasibility of broader implementation. Prior adherence interventions have generally required significant additional resources and the majority of quality improvement interventions are not continued following the end of the research project. This proposal will extend the current state of knowledge on medication adherence by demonstrating that a successful intervention will need multiple evidence-based components, and designed with plans for implementation in mind. Finally, this study is being conducted as a type I hybrid implementation study and therefore includes extensive study of the implementation process, which will yield both contributions to implementation science and facilitate wider dissemination if the intervention is found to be effective.

Furthermore, the study will make two important contributions to implementation science. First, the investigators will integrate a structured survey, the organizational readiness to change assessment (ORCA), for use in the formative and summative evaluations (described further in the evaluation plan). The ORCA has previously been psychometrically validated and found to predict implementation effectiveness. However, it has not been previously tested as a tool to support implementation. Findings from this study will help us understand if and how the ORCA can be used to support other implementation projects, and will contribute to users' materials for the survey. Second, the investigators are using an innovative study design, a randomized stepped wedge design, and building-in iterative formative evaluation to guide implementation of the intervention at sites in subsequent cohorts. As far as the investigators know, this kind of iterative rollout design with planned formative evaluation has not been used in an effectiveness/implementation hybrid study. If it works as planned, and provides useful formative evaluation findings from early cohorts, it may represent an important adaptation of the hybrid design.

In terms of VA patient care, if the intervention is successful, it will increase adherence to clopidogrel by helping patients take their anti-platelet medication routinely as prescribed, the quality of cardiovascular care for Veterans since adherence to clopidogrel has been associated with reductions in cardiovascular morbidity and mortality following PCI, and the efficiency of care by using telephone calls and telemonitoring for communication with patients rather than clinic visits. This study will leverage the automated pharmacy system and tele-monitoring technology for which the VA is a nationally recognized leader. Further, the study will address an important gap in knowledge (i.e., how to improve adherence to clopidogrel medications following PCI) and the findings can inform future interventions to improve adherence to other chronic cardiovascular medications. Finally, since the study is designed to be implemented using existing personnel/staff and current national platforms (i.e., CART-CL and PACT teamlets), it will be generalizable to other VA Medical Centers and Veterans.

Inclusion Criteria:

At the PCI sites, the investigators will include

• all patients undergoing PCI with either a bare-metal (BMS) or drug-eluting stent (DES) and are prescribed clopidogrel regardless of the intended treatment duration
• other potential anti-platelet medications (thienopyridines) used following PCI to accommodate changes in practice (e.g., prasugrel or ticagrelor or ticlopidine).
• all patients undergoing PCI and receiving clopidogrel at the randomized sites, regardless of gender, ethnicity or race. Based on data from the national CART Program, the investigators anticipate ~23% minorities (African American 16.8%, Hispanic 4.4%, Asian/American Indian 1.4%) and 3.1 women will be included in the study.

Exclusion Criteria:

The investigators will exclude

• sites with low PCI volume,
• less than 20 PCI procedures performed during the last fiscal year (n=3),