ABSTRACT
This chapter reviews the basic types of epidemiological studies most relevant for secondary data obtained from the electronic health record (EHR). When some exposure is believed to be associated with an outcome, an analytic study is undertaken, and at the heart of epidemiology is choosing the most appropriate study design to answer the research question, acknowledging the limits of data in the EHR. A study defines the population and sampling methodology with the goal of hypothesis testing (known as analytic epidemiology) and can be either experimental or observational. Distinguishing between a sample and a population is important when working with secondary data, as the secondary data will fall into one of these two categories. A research dataset that has been derived from an EHR or health information exchange (HIE) will likely consist of a well-defined population of individuals that shared some common characteristics, such as being a patient in that hospital.
