An Introduction to Real-world Data and Tips for Analysing It

Medical big data, also referred to as 'real-world data' (RWD) is defined as "data related to patient health status and/or health care delivery collected routinely from a variety of sources". This includes data from disease and drug registries, electronic health records, claims and billing data and census data collected from clinicians, hospitals, and payers. Observational studies using RWD collected during general clinical practice are considered complementary to randomized control trials. However, since this design does not allow the random assignment of patients, causal inference analyses are required. Researchers should study the protocol properly before considering the combination of study design, the characteristics of data source, calculation of the appropriate sample size and the validity of outcomes. Data definition using data code should also be considered. Furthermore, the reliability of the source studies must be considered and discussed when the article is written. This review aims to outline the methods for performing reliable observational studies using RWD.