Continuous tracking of patient’s health data through electronic health records (EHRs) has created an opportunity to predict the impact of healthcare policies. Despite the advances in EHRs, data can be missing or sparsely collected. We developed a simulation model to test treatment guidelines to prevent cardiovascular diseases. We study the treatment benefits and burden based on patients’ medication exposure over time. Our framework consists of using EM algorithms to fit sparse data and a discrete-time simulation model to test guidelines. Our results suggest that the current American College of Cardiology guidelines reduces over-treatment without affecting the risk of having a disease.