Randomized Controlled Trial of Alert-Based Computerized Decision Support for Optimizing Low-Density Lipoprotein Management

Given the striking failure to achieve evidence-based and guideline-recommended low-density lipoprotein cholesterol (LDL-C) targets and widespread underutilization of effective LDL-lowering therapies, including ezetimibe and PCSK9 inhibitors, even among highest risk patients, a heavy, unmitigated burden of cardiovascular risk persists, is addressable, and demands novel strategies. An alert-based computerized decision support (CDS) strategy addressed successfully a similar unmet need in cardiovascular medicine: the failure to prescribe antithrombotic therapy in atrial fibrillation patients at high-risk for stroke (Piazza G, et al. Eur Heart J. 2020 Mar 7;41(10):1086-1096). This alert-based CDS strategy tripled appropriate antithrombotic prescriptions in this vulnerable patient population. The alert reduced the odds of myocardial infarction (MI) at 90 days by 87% (OR 0.13; 95% CI 0.04-0.45) and cerebrovascular events or systemic embolism at 90 days by 88% (OR 0.12; 95% CI 0.0-0.91). Alert-based CDS effectively overcame barriers to guideline-directed therapy, integrating numerous clinical trials and evidence-based clinical practice guidelines, failure to identify at-risk patients, and educational gaps in strategies for stroke prevention.

The crisis in the failure to achieve guideline-directed LDL-C targets demonstrates similar clinical obstacles, suggesting that a CDS approach could be effective for optimizing lipid management. The following single-center, 400-patient, randomized controlled trial of an EPIC Best Practice Advisory (BPA; alert-based computerized decision support tool) aims to increase guideline-directed utilization of appropriate LDL-lowering therapies in undertreated inpatients and outpatients with atherosclerotic cardiovascular disease (ASCVD), including those with recent acute coronary syndrome (ACS). The study will evaluate this CDS intervention in both patient populations in parallel to gain insight into the impact of different settings on this approach and to better position implementation of this strategy in our own health systems and other clinical settings.

Study Design: A 400-patient U.S.-based single-center Quality Improvement Initiative in the form of a randomized controlled trial focused on the feasibility of implementation of this electronic alert-based CDS (EPIC BPA) based on LDL-C values. The 400 patients will be comprised of 200 in the "Hospitalized Patient Cohort" and 200 in the "Outpatient Clinic Cohort." The allocation ratio will be 1:1 for an electronic alert-based CDS (EPIC BPA) notification versus no notification.

Study Population: Patients eligible for enrollment will be drawn from two populations with atherosclerotic cardiovascular disease (ASCVD), including those with recent acute coronary syndrome (ACS): hospitalized patients (inpatients) and clinic patients (outpatients).

Inpatient Cohort: All patients ≥ 18 years old and admitted to the BWH Cardiovascular Medicine Service with a diagnosis (primary or secondary), medical history entry, or problem list entry of ASCVD (acute coronary syndromes, a history of MI, stable or unstable angina, coronary or other arterial revascularization, stroke, transient ischemic attack, or peripheral arterial disease) will be identified by the EPIC BPA. Those with a most recent (within 1 year) or current LDL-C value greater than 80 mg/dL (allowing for a 10% difference between goal LDL-C and current measurement) who are on a statin but not on ezetimibe and/or PCSK9 inhibitor will be potentially eligible for randomization. Any patients with no recent LDL-C measurement within the year prior to randomization will have an alert issued to their providers with recommendation for checking a lipid panel, which would be standard-of-care.

Outpatient Cohort: All patients ≥ 18 years old and seen in the BWH Watkins Cardiovascular Medicine Clinic with a diagnosis (primary of secondary) or problem list entry of ASCVD will be identified by the EPIC BPA. Those with a most recent (within 1 year) or current LDL-C value greater than 80 mg/dL (allowing for a 10% difference between goal LDL-C and current measurement) who are on a statin but not on ezetimibe and/or PCSK9 inhibitor will be potentially eligible for randomization. Any patients with no recent LDL-C measurement within one year will have an alert issued to their providers with recommendation for checking a lipid panel, which would be standard-of-care.

Intervention: The study team will devise a program to run within the EPIC Electronic Health Record (EHR) that will identify patients with ASCVD who are hospitalized on the BWH Cardiovascular Medicine Service or seen in the BWH Watkins Cardiovascular Medicine Clinic and have a most recent (within 1 year) or current LDL-C value greater than 80 mg/dL and are on a statin but not an PCSK9 inhibitor. In these patients who subsequently randomly assigned to the alert-based CDS strategy, an on-screen electronic alert will prompt the responsible inpatient provider (inpatient cohort) or the cardiologist of record for the clinic visit (outpatient cohort) as follows:

1. All patients whose LDL-C is not less than 70 mg/dL and not taking a maximally tolerated statin dose will prompt a recommendation for statin intensification
2. All patients whose LDL-C is within 20% of 70 mg/dL and who are on a maximally tolerated statin dose but not on ezetimibe or PCSK9 inhibitor will receive a prompt for adding ezetimibe
3. All patients whose LDL-C is greater than 20% of goal, and on a maximally tolerated statin (±ezetimibe) but not on PCSK9 inhibitor will receive a prompt for initiating PCSK9 inhibitor

The alert-based CDS will consist of an on-screen electronic dialogue box that will notify the clinician that the patient is at increased risk for cardiovascular events and is not at LDL-C goal according to current evidence-based clinical practice guidelines. The clinician will have the opportunity to proceed to an order template through which ezetimibe or a PCSK9 inhibitor can be prescribed. The clinician could also elect to learn more about current evidence-based clinical practice guideline recommendations for LDL targets. Finally, the clinician could elect to proceed without ordering more intensive lipid-lowering therapy or reading the guidelines but would have to provide a rationale for not prescribing ezetimibe or a PCSK9 inhibitor.