The primary aim of the CorCal study is to determine if CAC screening is noninferior to PCE for reducing major adverse cardiovascular events (MACE). Secondary aims include superiority for MACE (if noninferiority is achieved), cost, statin-adherence, statin-related events, and proportion of patients requiring statin treatment.

The specific primary objective of CorCal is to compare the effectiveness of 2 proactive cardiovascular primary prevention strategies, based on national guidelines, using the PCE (standard) versus CAC screening, in preventing MACE, including all-cause death, nonfatal myocardial infarction (MI), stroke, or any arterial revascularization among a low to moderate risk population with no current clinical evidence of cardiovascular disease. A secondary objective is to compare outcomes with a proactive approach (ie, MACE in those enrolled in the study and receiving monitored treatment using either method) compared to a passive approach to primary prevention (ie, MACE in invited but unenrolled patients receiving unmonitored “usual care”).

CorCal Outcomes is a randomized, noninferiority, open-label study designed to assess the effectiveness of a computed tomography (CT) coronary artery calcium scan as a guide to lipid-lowering therapy with statins for primary cardiovascular prevention compared to risk factor guidance using PCE. The study was performed at multiple sites in the integrated Intermountain Health care system. The study was approved by the Intermountain Health Institutional Review Board (IRB) and registered with Clinicaltrials.gov (NCT 03439267). The study is being conducted in accordance with the ICH (International Council for Harmonization) Good Clinical Practice guidelines (ICH GCP E6 R2), the Code of Federal Regulations (CFR), the Declaration of Helsinki, the Intermountain Heart Institute Cardiovascular Research standard operating procedures, and all relevant Intermountain Health research policies.

The trial was managed by investigators and study staff located at the Intermountain Medical Center Heart Institute in Salt Lake City, Utah. This study team was responsible for eMR queries of records of the multiple hospitals and clinics of Intermountain Health to identify qualifying patients whose primary care physicians (PCP’s) were willing to have them considered for the study. The central study team also was responsible for sending letters of invitation to these patients, receiving their responses, obtaining informed consent, randomizing them to an assigned management strategy, ordering entry CAC and lipid testing, receiving study test results (ie, CAC and lipid testing, questionnaires), and identifying study outcomes during follow-up. Patients’ PCPs were responsible for implementing the statin treatment recommendations for their enrolled patients and in providing general medical care during the duration of the study.

Figure 1 shows the study’s consort diagram. Patients were identified through electronic query using the inclusion and exclusion criteria described in Table 1 . Patients meeting criteria were randomly assigned to receive a letter inviting them to participate. Letters contained a unique code. The unique codes were associated with preloaded study record numbers in REDCap, the study’s electronic data capture system. This helped to prevent duplicate study records.

CorCal consort diagram.

Invited, accepting patients provided informed consent using a commercially available and fully validated system designed to obtain informed consent electronically. Using this system, patients reviewed study information prior to consenting, answered questions to verify eligibility, and then accessed the approved short form consent form that allowed for an electronic signature to document agreement to participate in the study. This system made it possible to remotely recruit and successfully enroll the large target study population.

Randomization was performed in a 1:1 fashion for consenting patients using a permuted block randomization method. This occurred electronically after consent was obtained.

When a patient was randomized, the REDCap electronic data capture system automatically triggered an email to send to the patient using the address provided on the consent form. This email informed the study participant to which of the 2 arms they were randomized and outlined any necessary testing (ie, CT coronary artery calcium scan and/or lipid panel testing). If no testing was needed (ie, recent lipid panel in PCE assigned subjects), the email indicated that as well.

A lipid panel was requested if not performed within 3 years and available in the eMR prior to study enrollment (both arms). For CAC-assigned subjects, a CT coronary artery calcium scan was ordered and performed according to the standard of practice that had been approved at each of the hospital sites where this procedure was done.

CAC scans were acquired across 8 Intermountain Health geographic sites. Using 64-slice helical CT scanners, ECG gated, high resolution, noncontrast, low-radiation, axial-scanning imaging of the heart, coronary arteries, and proximal great vessels was obtained. CAC scanning and 3-dimensional scoring were completed according to a standardized protocol. Interactive image viewing, volumetric displays, and image analysis were then performed. Age and sex-adjusted calcium percentile scores were determined from the MESA (Multiethnic Study of Atherosclerosis) trial nomograms. CAC reading and reporting were performed by experienced cardiologists and radiologists who were board-certified/eligible. Protocol trained radiology technicians and supervising physicians ensured that studies were of adequate quality, eg, with respect to impacting factors such as heart rate, image artifacts, and weight limitations.

The primary care physicians of study patients in all participating Intermountain Health clinics and/or hospitals were educated about the study and were assigned the responsibility to implement recommended statin therapy in their specific patients. Specifically: (1) primary care leaders (and some groups of primary care physicians [PCPs]) were approached in townhall/roadshow fashion. (2) PCP leaders made an administrative decision to approve the research contact and recommendation approach. (3) A standardized results and statin recommendation letter was drafted (based on PCE or CAC assignment) and sent to the identified PCPs of the enrolled study patient. Each enrolled participant’s results were reviewed by an investigator who filled out the standardized letter and results based on the study protocol. This was sent to the PCP, and a letter with similar results and recommendations was concurrently sent to the participant.

After randomization and completion of lipid profiles and (in the CAC arm) CT coronary artery calcium scans, a recommendation was made by the study investigative team to the patient’s primary care physician via the eMR and to the patient via email, even if no statin was recommended, according to Supplemental Tables I and II. Briefly, a statin was recommended in the PCE (standard of care) arm based on the 2018 Cholesterol Management Guidelines (Supplemental Table I). A moderate-intensity statin was recommended for a 10-year risk of ≥7.5%, and a moderate-intensity statin could be considered for a 5.0 to 7.4% risk. A high intensity statin was recommended for a risk of ≥20%. No statin was recommended for a score of <5%. In the CAC arm, a high-intensity statin was recommended for a score of >100 Agatston Units (AU) (Supplemental Table II). For scores of 1-100 AU, when age/sex matched MESA risk percentile was ≥75%, a moderate intensity statin was recommended. Otherwise, no statin was recommended unless LDL-C was ≥190mg/dL (both arms), in which case a high-intensity statin was recommended.

Recommended statins were rosuvastatin and atorvastatin. High and moderate intensity doses were defined as 20 to 40mg/d and 10mg/d, respectively, for rosuvastatin, and 40 to 80mg/d and 20mg/d, respectively, for atorvastatin. The statin recommendation, the specific selection of the statin used, and the statin dose (within study specifications) were expected to be communicated and prescribed to patients by their primary care physicians.

There were no procedures or evaluations after the initial statin recommendation except for the completion of the statin adherence questionnaire. A study-specific series of questionnaires was developed to assess the subject’s adherence to statin (or no-statin) recommendations, called the Intermountain Statin Adherence Questionnaire or ISAQ (Supplemental Table III). The ISAQ was sent to participating subjects via postal mail or made available for subjects to complete via an electronic system. ISAQ was sent for completion 3 months after the initial statin recommendation, then yearly (within ±60 days of the anniversary date of the recommendation) until study completion. All other follow-up evaluations were performed through a review of each subject’s eMR.

Each subject’s eMR was reviewed on an ongoing basis as needed for the Data and Safety Monitoring Board (DSMB) safety and outcomes updates, which occurred at least annually until at least the end-of-study. This review focuses on baseline characteristics and statin recommendations within and between the 2 arms. All MACE outcomes by study end are adjudicated by a Clinical Events Committee.

Based on a review of internal data, it was anticipated that the proportion of subjects experiencing an event over 4 years would be 11.04% in the PCE arm. Assuming a 10% reduction in events for the CAC group, their proportion experiencing a 4-year event was estimated to be 9.94%. Thus, with an exponential survival function of 0.0292 in the PCE arm, a noninferiority margin of 1.30, and a 10% (2.5%/year) dropout rate over the course of the study, a total of 4,500 patients per arm and an overall sample size of 9,000 subjects was proposed to test for noninferiority between the CAC and the PCE arms. This was based on 90% power at a 0.025 significance level. The number of events required to achieve this power was calculated to be 621. The study end was proposed to be event-driven, which was anticipated to require approximately 4 years. In September 2021, given slower recruitment (ie, due to COVID-19) and a lower than anticipated overall event rate, statistical power was reduced to 80%, and the number of participants targeted for enrollment was reduced to 5,546 (2,773/arm), and the target number of events was reduced to 465.

We estimated that our power to distinguish a cost advantage of CAC over PCE, a key secondary outcome, would be greater than for the primary MACE outcome as follows: to evaluate this key secondary outcome, we estimated costs using internal data associated with performing a CT coronary calcium scan, lipid panels, comprehensive metabolic panels, office visits, statins, and the impact of cardiovascular-related outcomes. Based on data from our Vanguard trial, we anticipated that we would be treating 25% fewer patients in the CAC-arm and would be overpowered with these numbers to show an advantage to CAC-guidance.

The primary endpoint of the study is the composite of all-cause mortality, nonfatal MI, nonfatal stroke, and any arterial revascularization (coronary, carotid, and peripheral). The primary endpoint will be tested for noninferiority comparing the randomized PCE (standard interventional control) group with the randomized CAC-guided (investigational interventional) group. noninferiority will be met if the upper boundary of the 2-sided 95% confidence interval is less than 1.30 when the CAC arm is compared to the PCE arm. The primary outcome will be assessed from the day of randomization to the occurrence of the first event using Cox hazard regression analysis. This analysis will be performed using a modified per-protocol analysis. This approach includes all patients who signed informed consent, were randomized to a study arm, received their CT coronary artery calcium scan and/or a lipid panel (depending on study arm assignment), and received a statin management assignment. Inclusion of these patients in the per protocol analysis will occur regardless of adherence to the recommended treatment recommendation. The intention-to-treat method will be performed as a supporting analysis. If noninferiority is achieved, superiority will be evaluated without further statistical penalty.