Ischemic Heart Disease (IHD) and rheumatic diseases are both pervasive, expensive, and result in grave health consequences. Ischemic Heart Disease (IHD) affects an estimated 15.4 million Americans 20 years of age-representing 6.4% of the adult population. The direct and indirect cost of Ischemic Heart Disease has been estimated at $195.2 billion, with a doubling of cost projected by 2030. Similarly, the direct cost to the U.S. workforce for rheumatoid arthritis alone approaches $5.8 billion yearly.
Widely-accepted national evidence-based guidelines support the use of cardio-protective medications to reduce the risk of adverse consequences resulting from Ischemic Heart Disease and Disease Modifying Anti-Rheumatic Drugs (DMARDs) to reduce the risk of adverse consequence in rheumatic diseases. For example, numerous rigorously conducted randomized trials show that statins improve outcomes and reduce mortality in patients with established cardiovascular disease (i.e., secondary prevention), including those undergoing Percutaneous Coronary Intervention (PCI). The use of statins and beta-blockers has been repeatedly demonstrated to be cost-effective in lowering Cardiovascular Events (CVE) event rates, in part by their effects on cholesterol, and blood pressure, respectively. Accordingly, the most recent Veteran Affairs (VA) performance measures and American Heart Association guidelines encourage the use of statins in patients with atherosclerotic disease; beta-blockers in subjects with left ventricular systolic dysfunction (ejection fraction 40%), prior Myocardial Infarction (MI), or blood pressure of 140/90 or greater; and clopidogrel following any acute coronary syndrome (ACS) or Percutaneous Coronary Intervention with stent. The rheumatology literature provides similar evidence for the benefit of Disease Modifying Anti-Rheumatic Drugs (DMARDs) in rheumatic diseases, and guidelines strongly endorse their use.
Unfortunately, non-adherence to medications is common, and increases the risk of poor outcomes. Our 2011 national preliminary data from VA Cardiac Catheterization Laboratories (CCLs) demonstrate that over 6300 patients experienced at least one refill gap of >= 7 days for statins in the year following Percutaneous Coronary Intervention. The mean Proportion of Days Covered (PDC) for these patients was only 75%-below the Proportion of Days Covered threshold of 80% that typical defines adherent patients, based on the empiric evidence for effectiveness of medications at this cut-point. Non-adherent patients were present at all Cardiac Catheterization Laboratories (CCLs) without substantial variation in mean Proportion of Days Covered (PDC) by center, suggesting a global problem.
Systematic problems underlie and contribute to non-adherence to medications. Usual care of Ischemic Heart Disease (IHD) and rheumatic disease patients is encumbered by systematic deficiencies including: passive monitoring (contact with patients only when initiated by the patient) and inefficiency (time-consuming patient-by-patient approach, rather than through population management). The proposed intervention addresses both the complex patient-specific factors (emphasizing forgetfulness and carelessness) and the systematic inadequacies using a multi-modal, escalating approach in two clinical scenarios: Ischemic Heart Disease and rheumatic disease.
1.To assess the effectiveness of a multi-faceted patient-centered intervention versus usual care in improving medication adherence as measured by proportion of days covered (PDC, primary outcome). This will be tested among Ischemic Heart Disease (IHD) patients for statins, beta-blockers and clopidogrel in the year after Percutaneous Coronary Intervention (PCI) and among rheumatology clinic patients chronically prescribed Disease Modifying Anti-Rheumatic Drugs (DMARDs). Hypothesis: The Proportion of Days Covered (PDC) for patients in the intervention arm will exceed the Proportion of Days Covered (PDC) for the usual care arm by a 10% absolute difference.
2.(Secondary outcome): To determine the effectiveness of a multi-faceted patient-centered intervention versus usual care in reducing secondary Cardiovascular Events (CVEs; myocardial infarction [MI], repeat revascularization [Percutaneous Coronary Intervention (PCI) or coronary bypass graft], and all-cause mortality) among Ischemic Heart Disease (IHD) patients at 12 months post- Percutaneous Coronary Intervention (PCI) and progressive erosive disease demonstrated on plain film radiographs in patients with rheumatic diseases (i.e. "radiographic progression"). Hypothesis: The rate of Cardiovascular Events (CVEs) and radiographic progression will be 5% relatively lower for patients in the intervention arm compared with usual care.
3.(Secondary outcome): To establish the cost to implement and maintain the intervention, above the cost of usual care, as well as the incremental cost effectiveness (Incremental Cost Effectiveness (ICE); e.g. cost to achieve at 10% improvement in Proportion of Days Covered (PDC); cost per Cardiovascular Events (CVE) prevented). Hypothesis: This aim does not posit a hypothesis as the objective is descriptive. The available funding for this project limits this outcome to Ischemic Heart Disease (IHD) patients (no rheumatic disease patients will be analyzed according to cost).
5.0 Study Procedures
5.1 Study Design
This is essentially two concurrent 4-year multi-site cluster stepped-wedge Randomized Controlled Trial (RCTs) that tracks each subject for 12 months. The Randomized Controlled Trials (RCTs) are executed simultaneously in two conditions (Ischemic Heart Disease (IHD) and chronic rheumatic disease) at the same Veteran Affairs Medical Centers (VAMCs), with enrollment periods ("observation phases") of 6 months' duration. At the 4 intervention sites, which will be rolled-out through 4 phases, subjects undergo prospective monitoring for med refill gaps, with deployment of a multifaceted, tailored program. The program assists veterans using Short Messaging System (SMS), and if necessary, escalates to Interactive Voice Response (IVR), and then to research interventionalists. The intervention will determine the specific reasons for non-adherence for each veteran and then, based on the individual's needs: facilitate drug refills, counsel patients on correct medication usage, simplify drug regimens, provide low literacy instructional materials, facilitate communication with providers for adverse drug reactions, and engage in conversation about changing non-adherent behavior Twelve control sites, matched according to Percutaneous Coronary Intervention (PCI) volumes & baseline adherence (median Proportion of Days Covered (PDC)), will deliver usual care.
Our investigation employs modifications to the classic stepped-wedge trial design by collecting data from concurrent controls/usual care sites (as well as the rheumatology clinics at these sites) in addition to the typical pre-intervention/usual care observation periods. Because the infrastructure already exists to collect pharmacy data from patients treated at all Cardiac Catheterization Laboratories (CCL's), we are able to obtain this additional control data. Our statistical simulations suggest this will enhance the power of our study.
For the first clinical scenario (Ischemic Heart Disease), patients discharged after Percutaneous Coronary Intervention (PCI) at any one of the four intervention sites who fill at least one prescription from one of the three qualifying medication classes (statins, beta-blockers, thienopyridines) will be monitored for medication refill gaps. The monitoring will occur daily through automated queries of pharmacy dispensing data, leveraging the considerable infrastructure of the VA's Clinical Assessment, Reporting, and Tracking (CART) Program. We have already developed these queries to identify when patients will need to refill their medications, based on the most recent medication refill date and the number of days supplied, but have not previously incorporated this in a real-time intervention. In the current study, patients with a refill gap (4 days for clopidogrel, 7 for all other medication classes) will be contacted in a manner that allows for simple refilling of their medication, thereby addressing a common cause of missed doses. The intervention modality will escalate through increasingly more intensive components, including Short Messaging System (SMS) (only if patient opts in to allow Short Messaging System (SMS) during an initial Interactive Voice Response (IVR) call), then Interactive Voice Response (IVR), then research interventionalist assistance (clinical pharmacist), then notification of Primary Care Provider or Patient Aligned Care Team. Further, in order to make study subjects' primary care team aware of a subjects' participation in the study, a note will be included in the participants' electronic medical record indicating their participation in the study.
For the second clinical scenario (rheumatic diseases), patients followed in rheumatology clinics with diseases such as rheumatoid arthritis and spondyloarthritis who have recently filled at least one Disease Modifying Anti-Rheumatic Drugs (DMARD) prescription will be monitored for medication refill gaps. The monitoring will occur weekly through automated queries of pharmacy dispensing data, as developed by analysts associated with the local pharmacy service. Since all Veteran Affairs Medical Centers (VAMCs) with Cardiac Care Laboratories also have rheumatology clinics, the rheumatology intervention will be conducted simultaneously with the IHD-intervention and at the same VAMC sites, based on the randomization assignment used for the Ischemic Heart Disease (IHD)-intervention. This approach to randomization is necessary because it simplifies an already complicated randomization procedure and allows for more efficiency by hiring interventionalists/clinical pharmacists at fewer sites than if the Ischemic Heart Disease (IHD) and rheumatology site assignments are performed independently. The intervention approach (Short Messaging System (SMS), Interactive Voice Response (IVR), and pharmacist) will be identical to that of the Ischemic Heart Disease (IHD) intervention, but with content adapted to Disease Modifying Anti-Rheumatic Drugs (DMARDs).
To protect confidentiality, whenever possible, study records will identify patients only by unique study number. The only exception will be the tracking database that will include the patient's name, Social Security Number (SSN), and telephone number which will be used to obtain pharmacy data from the Corporate Data Warehouse (CDW) following Percutaneous Coronary Intervention (PCI) and to validate Cardiovascular (CV) events in the medical record. Access to the database and study records will be limited to research staff. All paper records will be maintained in locked file cabinets within locked offices. Electronic data files will be encrypted/password-protected and user-restricted on computers maintained in a secure environment behind the Veterans Affairs (VA) firewall per Veteran Affairs (VA) security regulations. All data will be maintained in accordance with the VHA Records Control schedule 10-1 (v.January 2016).
5.2 Recruitment Methods
This study will employ an approach to consent similar to SDP-179 Hybrid Effectiveness-Implementation Study to Improve Clopidogrel Adherence which has previously been approved by Central Institutional Review Board (CIRB) (CIRB Protocol #12-12) The study will enroll all subjects undergoing Percutaneous Coronary Intervention (PCI) or using Disease Modifying Anti-Rheumatic Drugs (DMARDs) at 16 Veteran Affairs Medical Centers (conservatively estimated at n~300 subjects from intervention periods at 4 sites and n~2100 at control sites for each intervention: Ischemic Heart Disease (IHD) and rheumatic diseases; n=4800 subjects total for both Randomized Controlled Trials). These Medical Centers are all tertiary referral hospitals within their respective VISN and provide a full spectrum of cardiac and rheumatologic care services.
Recruitment/Consent for clusters (Cardiac Catheterization Laboratories (CCLs)): The National Program Director for Cardiology, Director of the National Clinical Assessment Reporting and Tracking Program (CART Program), Chair of the VA Rheumatology Field Advisory Committee, and President of the VA Rheumatology Consortium all strongly endorse this proposal and have committed to assist with recruitment of sites. Over sixty Cardiac Catheterization Laboratories (CCLs)/rheumatology clinics are available from which to select the 16 study sites, Cardiac Catheterization Laboratories (CCL) and Rheumatology Clinic Directors will provide formal "site-level consent" --i.e. commitment in writing to institute the protocol--that submits their catheterization laboratory to randomization.
Recruitment/Consent for individual veterans: For the Ischemic Heart Disease (IHD) intervention, following their Percutaneous Coronary Intervention (PCI) and prior to discharge from the Cardiac Catheterization Laboratories (CCL), patients at intervention sites will be informed of the study. This will be achieved through a written Information Sheet that is distributed to all patients; it will consist of an IRB-approved document explaining the purpose of the study and intervention details. Individuals at the intervention site will be given the information sheet and will be called by a member of the study team within a few days to see if the potential participant is interested in participation in the study. If the individual is interested in participating then the elements of consent will be reviewed at that time with verbal consent if the patient agrees to participate (Waiver of Documentation of Informed Consent and Waiver of HIPAA are requested). For the rheumatic disease intervention, patients will be informed of the study by receiving a written Information Sheet at the end of their regularly scheduled clinic visits and will be called by a member of the study team to undergo the same verbal consent process used in the Ischemic Heart Disease (IHD) intervention group if there is interest in participating.
For the control group sites, a waiver of HIPAA authorization and a Waiver of Informed Consent for access and use of the data will be requested.
Unit of randomization: In order to prevent cross-contamination of the intervention, a clustered design is being used with randomization at the level of Cardiac Catheterization Laboratories (CCL), rather than at the patient level. Rheumatology Clinic assignment follows Cardiac Catheterization Laboratories (CCL) assignment; that is, rheumatology clinics at each site allocated to the same treatment arm (intervention or control) as the Cardiac Catheterization Laboratories (CCL) at that site.
Procedures: Cardiac Catheterization Laboratories (CCLs) Directors and Rheumatology Clinic Directors will agree in writing that their Cardiac Catheterization Laboratories (CCLs) (and the associated rheumatology clinic at each VAMC) will be randomized to intervention or control/usual care. A two-stage randomization procedure will be employed. First, consented Cardiac Catheterization Laboratories (CCLs) will be grouped with the most similar sites, in terms of baseline Proportion of Days Covered (PDC) and Percutaneous Coronary Intervention (PCI) volumes. Using the median Proportion of Days Covered (PDC) and median Percutaneous Coronary Intervention (PCI) volume for all consented Cardiac Catheterization Laboratories (CCLs), each site will be assigned to one of four groups: high Proportion of Days Covered (PDC)/high Percutaneous Coronary Intervention (PCI) volume, high Proportion of Days Covered (PDC)/low Percutaneous Coronary Intervention (PCI) volume, low Proportion of Days Covered (PDC)/high Percutaneous Coronary Intervention (PCI) volume and low Proportion of Days Covered (PDC)/low Percutaneous Coronary Intervention (PCI) volume. From each group of 4 consented Cardiac Catheterization Laboratories (CCLs), one site will be randomized to intervention and three will be randomized to control. In the second stage, the four intervention sites will undergo randomization to determine the order that sites roll-out the intervention, as is typical of stepped-wedge designs. The Rheumatology Clinic randomization will follow the assignment of the Cardiac Catheterization Laboratory at their local site.
Sequence Generation: Randomly permuted blocks with four centers per block (matched according to annual Percutaneous Coronary Intervention (PCI) volumes and center Proportion of Days Covered (PDC)) will be generated using computer-derived assignments.
5.3 Informed Consent Procedures
For the purposes of this study we are submitting a waiver of informed consent and instead depending on an Opt Out procedure, as described above and as previously employed for CIRB #12-12.
We have enrolled a total of 381 participants from the four intervention sites (237 from Site #1 Durham, 115 from Site #2 San Francisco, 17 from Site #3 Baltimore, and 12 from Site #4 San Juan). Of those in the study we have detected a gap in medication refills in a total of 216 individuals (151 from Site #1 Durham, 56 from Site #2 San Francisco, and 9 from Site #3 Baltimore).
All analyses will be held until all 4 intervention sites are finished with recruiting patients.
Execution of the intervention will have important ramifications for veterans' wellness. First, the intervention may directly improve veterans' quality of care, leading to better outcomes. Second, it will reinforce the importance of managing their Ischemic Heart Disease (IHD) and rheumatic diseases- conditions of widespread relevance to veterans. Third, the intervention will be responsive to veterans by assisting them in a proactive manner. Our system has the potential to make Veterans feel heard by proactively addressing adherence barriers, rather than placing the responsibility and burden entirely upon the patient to overcome obstacles and navigate the health care system. Fourth, the Ischemic Heart Disease intervention provides a template for effective management of numerous similar clinical situations in which a procedure identifies high-risk individuals and requires med adherence to prevent recurrence or exacerbation of a serious condition. For example, in patients with an endoscopy that reveals a peptic ulcer; monitoring proton pump inhibitors adherence might avert recurrent bleeding ulcers. Thus, this proposal will hone our understanding of the risks for long-term untoward outcomes following procedures, using some of the most common procedures for some of the most common diseases within the VA.
- Daniel CM, Davila L, Makris UE, Mayo H, Caplan L, Davis L, Solow EB. Ethnic Disparities in Atherosclerotic Cardiovascular Disease Incidence and Prevalence Among Rheumatoid Arthritis Patients in the United States: a Systematic Review. ACR open rheumatology. 2020 Sep 1; 2(9):525-532.
- Raghavan S, Liu WG, Berkowitz SA, Barón AE, Plomondon ME, Maddox TM, Reusch JEB, Ho PM, Caplan L. Association of Glycemic Control Trajectory with Short-Term Mortality in Diabetes Patients with High Cardiovascular Risk: a Joint Latent Class Modeling Study. Journal of general internal medicine. 2020 Aug 1; 35(8):2266-2273.
- Hirsh J, Wood P, Keniston A, Boyle D, Quinzanos I, Caplan L, Davis L. Universal Health Literacy Precautions Are Associated With a Significant Increase in Medication Adherence in Vulnerable Rheumatology Patients. ACR open rheumatology. 2020 Feb 1; 2(2):110-118.
- Yazdany J, Caplan L, Fitzgerald J, Schmajuk G. Editorial: The Evolving Art and Science of American College of Rheumatology Guidelines. Arthritis & rheumatology (Hoboken, N.J.). 2019 Jan 1; 71(1):2-4.
- McCulley CB, Barton JL, Cannon GW, Sauer BC, Teng CC, George MD, Caplan L, England BR, Mikuls TR, Baker JF. Body mass index and persistence of conventional DMARDs and TNF inhibitors in rheumatoid arthritis. Clinical and Experimental Rheumatology. 2019 May 1; 37(3):422-428.
- Baker JF, Sauer B, Teng CC, George M, Cannon GW, Ibrahim S, Cannella A, England BR, Michaud K, Caplan L, Davis LA, O'Dell J, Mikuls TR. Initiation of Disease-Modifying Therapies in Rheumatoid Arthritis Is Associated With Changes in Blood Pressure. Journal of clinical rheumatology : practical reports on rheumatic & musculoskeletal diseases. 2018 Jun 1; 24(4):203-209.
- Raghavan S, Liu WG, Michael Ho P, Plomondon ME, Barón AE, Caplan L, Joynt Maddox KE, Magid D, Saxon DR, Voils CI, Bradley SM, Maddox TM. Coronary artery disease severity modifies associations between glycemic control and both mortality and myocardial infarction. Journal of diabetes and its complications. 2018 May 1; 32(5):480-487.