AN ELECTRONIC MEDICAL RECORD BASED ALGORITHM TO TAILOR CARDIOVASCULAR DISEASE PREVENTION USING LIPOPROTEIN(A), APOLIPOPROTEIN B, CHOLESTEROL AND MYOCARDIAL INFARCTION DIAGNOSIS: ABCDS PREVENTION PROGRAM

Abstract

Therapeutic Area: CVD Prevention – Primary and Secondary; ASCVD/CVD Risk Assessment; Preventive Cardiology Best Practices Background: According to the 2022 American Heart Association Heart Disease and Stroke Statistics, coronary heart disease remains the leading cause of death attributable to cardiovascular disease (CVD). Opportunity exists to utilize electronic medical records (EMRs) and biomarkers to facilitate early identification of patients at high risk for CVD. Additionally, automatic or opt-out orders are EMR-based tools that have the potential to improve referral rates to prevention programs. The role of cardiovascular biomarkers and electronic medical records (EMRs) in optimizing identification and referral of patients at risk for CVD are explored in the ABCDs PREVENTION program. Methods: A multidisciplinary team of cardiologists, internists, engineers, and clinical informaticists defined the logic for the guideline based ABCDs PREVENTION tool. The EMR algorithm used the cardiovascular risk biomarker thresholds of lipoprotein(a) > 70 nmol/L, apolipoprotein B > 90 mg/dL, low-density lipoprotein cholesterol > 150 mg/dL, and triglycerides > 200 mg/dL, and/or a diagnosis of ST-elevation myocardial infarction (STEMI) or non-ST-elevation MI (NSTEMI) based on ICD-10 codes to generate automatic referrals to (1) cardiac rehabilitation (CR), (2) the advanced lipid disorders clinic, and/or (3) Corrie Cardiovascular Health Program (Figure 1). Results: In a test environment, the algorithm was applied to 27 patients identified by the clinical team with STEMI or NSTEMI. The algorithm was 90% successful in triggering automatic referrals to CR and Corrie. Fail rates can be attributed to our current algorithm not detecting some ICD codes related to NSTEMI. The automatic referral to lipid disorders clinic based on abnormal lipid biomarkers is now live and undergoing automation optimization to validate accuracy. Conclusion: Building an EMR-based algorithm to individualize CVD prevention using cardiovascular risk biomarkers and diagnoses may enable early identification and intervention on high-risk patients. Future directions include applying the algorithm to clinical decision support tools as well as an automated order set to increase referral rates to evidenced-based programs focused on primary and secondary CVD prevention. Ultimately, use analysis will determine if the algorithm improves referral rates to CR, lipid clinic, and the Corrie Cardiovascular Health Program to improve access to these evidence-based services.