The Data Science of Health Informatics

In this module, we will begin by introducing and defining databases, and placing the role of databases within the context of clinical informatics. We will continue by introducing the common health data types such as demographics, diagnosis, medications, procedures, and utilization data. We will finish this module by reviewing the emerging health data such as lab orders/results, vital signs, social data, and patient-generated data.

In this module, we review the data specifications extracted from insurance claims and electronic health records. We will then discuss the common challenges in using health data, specifically issues with data quality, data interoperability, and data system architectures. Finally, we will describe the “Big Data” challenges of health data and explain some of the data problems that may hinder analytical efforts.

With this understanding of the data available, it’s time to see how to turn questions you and your colleagues will have into queries the database can understand. Besides getting rules of thumb for doing this translation, you will also be introduced to three online tools available to test some of these skills. You will also watch an interview with Sam Meiselman, course instructor and the data manager in charge of the Johns Hopkins Enterprise Data Warehouse, who has to use these skills on a daily basis.

To send home the recurring message on the challenges and art of translating questions into queries, you will see interviews with two professionals: One who comes from the data management side of the equation, and one who comes from the domain. They will give you perspectives that are both similar (the need to understand the problem for which the data are being retrieved) and different (the multiplicity of data available vs the richness of the domain problem).