BACKGROUND: Though secondary data analyses of large datasets may reduce logistical and financial barriers required to perform significant and innovative work, such research requires specialized skills in data handling and statistical techniques as well as thorough and detailed knowledge of the data sources being used.
OBJECTIVES: To provide an overview of several common types of secondary data, focusing on strengths, weaknesses and examples of how these data may be used for health services research.
RESULTS: Secondary data comprise a broad and heterogeneous category. This review covers several large categories of such data with examples of their use and discussions about their strengths and weaknesses. Sources include administrative data, claims-based datasets, electronic health records health surveys, patient or disease or both registries, quality improvement initiatives, as well as data from existing trials. Linkages of different types of data may expand the scope of questions answerable using secondary data analysis. Specific strengths and weaknesses of each type of dataset are discussed along with examples from the recent urologic literature.
CONCLUSIONS: Choice of the appropriate data source should be tailored to the specific research question as well as the research resources and expertise available. Appropriate decisions about which data to use are the foundation for valid, high-impact research using secondary data.