An estimated 47% of Americans 65 years of age and older take statins, which are highly effective in lowering low-density lipoprotein (LDL) cholesterol, preventing atherosclerotic cardiovascular disease (ASCVD), and reducing all-cause mortality. Unfortunately, ~50% of patients prescribed statins do not obtain these critical benefits because they discontinue use within 1 year of treatment initiation. There, Statin discontinuation has been identified as a major public-health concern due to increased morbidity, mortality, and healthcare costs associated with ASCVD. In clinical practice, statin-associated symptoms (SAS) often result in dose reduction or discontinuation of these life-saving medications. Currently, physicians employ reactive strategies to manage SAS concerns after they manifest, such as offering an alternative statin treatment plan (e.g., reducing dosage or changing medication) or a `statin holiday'. However, with numerous statin treatment strategies available and no means of optimizing their match to a given patient, physician decision-making is based on minimal patient data elements. Moreover, using a single patient's data to identify the optimal statin regimen and treatment plan is inadequate to ensure that the harms of statin use are minimized and the benefits are maximized. A decision- support system, by contrast, can use a vast number of variables from a large number of patients (?big data?) to match an optimal statin treatment plan to an individual patient prospectively. We propose to use complex patient information to develop and test an effective predictive model and tool, the personalized statin treatment plan (PSTP) platform, which could be used by physicians to optimize personalized statin treatment to minimize harms (SAS and statin discontinuation) while at the same time maximizing benefits (LDL reduction). The proposed study leverages data from the OptumLabs Data Warehouse (which includes more than 20 years of insurance claims and electronic health records data from more than 150 million patients across the United States), as well as clinical trial simulations, for model development, validation, and evaluation. We will address the following specific aims: 1) Develop and validate a deep-learning model to predict both SAS and statin discontinuation using linked insurance claims and EHR data; 2) Develop and validate the PSTP platform to identify statin treatment plans that optimize both LDL reduction (benefit) and SAS and statin discontinuation (harm) for a given patient profile; and 3) Evaluate the PSTP relative to current and guideline-driven practices using CTS to assess clinical benefits and harms. The proposed study will produce a precision-medicine tool to empower physicians to make proactive clinical decisions regarding statin treatment planning (i.e., selecting the statin drug and dosage optimized for a particular patient to maximize LDL reduction and minimize statin discontinuation and SAS) before any statin is prescribed. The implementation of such a tool would substantially improve public health by reducing statin discontinuation and sub-optimal clinical outcomes inherent in current reactive statin-prescribing strategies based on non-personalized statin treatment plans.
About 50% of patients prescribed statins do not obtain critical benefits in lowering low-density lipoprotein cholesterol, preventing atherosclerotic cardiovascular disease, and reducing all-cause mortality because they discontinue use within 1 year of treatment initiation largely due to statin-associated symptoms. Statin discontinuation has been identified as a major public-health concern due to increased morbidity, mortality, and healthcare costs associated with atherosclerotic cardiovascular disease. The proposed study will produce a personalized (?precision?)-medicine tool using big data to empower physicians to make proactive clinical decisions for prescribing statins that will maximize LDL reduction and minimize statin discontinuation and statin-associated symptoms.
