Quantifying Gains in Data Quality for Sampling Plans Used in Clinical Trial Monitoring

Abstract

Clinical study monitoring is the first level of quality assurance checks on clinical trials data. The clinical study monitor goes to the study site and compares the case report forms to the source documents. Because some clinical trials are very large, the monitor may not be able to do a 100% check of all case report forms versus their corresponding source document. This paper will examine three possible sampling plan scenarios for monitoring case report forms in large studies.

Three sampling plans for monitoring quality assurance checks will be discussed. The first two plans are variations on simple random sampling plans. The other plan involves what is known in the quality control literature as a continuous sampling plan. The advantages of using these particular sampling methods are: (1) they are easily performed by monitoring personnel; (2) the cost/time necessary to perform the check along with the gain in quality of the data set can be computed; and (3) the sampling plan can be constructed to attain data quality goals set by the monitoring staff. This paper will explore the relationship between the costs of quality checks on clinical data and the gain in data quality produced by their use.