Presently, the process of extraction and dissemination of data subsets for research from clinical data warehouses is cumbersome and error prone. Furthermore, large-scale research projects often involve multiple users of the same data extract; each of these users may be authorized to access different data elements and specific subsets of the data extract. Once initial data extraction has been done for a research project, capability to transform the data for individual users and track which data are being accessed by which users in a secure environment is lacking in existing systems. This paper describes several methods that the authors are integrating into a system designed to provide secure, flexible, and auditable support for supplying users with data subsets. The methods implement secure, redactable, and auditable mechanisms for data extraction and dissemination. This paper describes the architecture along with an initial proof-of-concept implementation. Preliminary performance measurements show that the approach manages clinical data in redactable and auditable form with reasonable overheads.