Development and Application of Pharmacological Statin-Associated Muscle Symptoms Phenotyping Algorithms Using Structured and Unstructured Electronic Health Records Data.

Statins are commonly prescribed cholesterol-lowering medications in the US, but some patients may experience statin-associated muscle symptoms (SAMS) that can reduce their benefits. In this study, we developed and tested a simple algorithm using electronic health records (EHRs) to identify cases of SAMS. We retrieved data from statin users in the Minnesota Fairview EHR system and manually identified a gold standard set of SAMS cases and controls using a clinical tool. We developed machine learning and rule-based algorithms that considered various criteria, such as ICD codes, statin allergy, creatine kinase elevation, and keyword mentions in clinical notes. The best performing algorithm, called the combined rule-based (CRB) algorithm, achieved similar performance to machine learning algorithms in identifying SAMS cases. When applied to the larger statin cohort, the CRB algorithm identified a prevalence of 1.9% for pharmacological SAMS, and identified selective risk factors such as female gender, coronary artery disease, hypothyroidism, and use of immunosuppressants or fibrates. The developed algorithm has the potential to help create SAMS case/control cohorts for future studies such as building models to predict SAMS risks for patients.