Source Data Verification by Statistical Sampling: Issues in Implementation

Efficiency of the drug development process is a continuing concern for pharmaceutical companies, governments, regulatory authorities, and patients. While much time and effort have been spent on developments in genetics and on sophisticated statistical designs, there has been less concern about the processes that govern the running of clinical trials. In this article, I describe a statistical method for source data verification whose implementation can have a large impact on the workload of trial monitors. I investigate the consequences of a less stringent form of source data verification on the quality of data and the inferences that can be drawn from the data.