Heart is the most important part of the human body, diseases that are related to heart cause a huge thread to human health. In this paper, methods that applied Apache Spark to heart disease related works would be shown and discussed in order to classify these methods and make a conclusion about the innovations and shortcomings of these works. These works are defined into two categories: the ones that adopted traditional machine learning method and the ones that used deep learning methods. By classifying these works into two types, commonalities and similar innovative approaches in the same category of methods can be better observed and summarized, facilitating a clearer comparison of the similarities and differences in the innovative focuses among similar yet distinct methods. By doing so, conclusions were made to show that apart from enhancing operational efficiency and reliability of models for diagnosing and treating heart diseases, current research utilizing Apache Spark in this field also identifies areas for improvement such as expanding sample data representation, speeding up data processing, and addressing concept drift issues with proposed solutions. By addressing these challenges, researchers aim to optimize existing methods using Apache Spark and advanced data analytics techniques to combat heart diseases.