Metrics and Best Practices in Clinical Data Management: Conclusions of a DIA Roundtable Workshop *

Abstract

Metrics are statistics that characterize the quality, productivity, speed (performance), or cost of one or more clinical data management activities. Even though clinical data management metrics have been used for decades, little effort has been made to standardize definitions of metrics or publish their values (“practices”). Participants in a Drug Information Association (DIA) roundtable conference on metrics, organized by Ron Fitzmartin, began discussions that could lead to standardization and publication of metrics and practices. The participants discussed a wide variety of clinical data management and biostatistical metrics, assigned subjective priority (importance) levels, created descriptions for many, and estimated current and best practices for some. Noting that clinical data management procedures vary widely, that actual procedures profoundly affect practices, and that little is known publicly about current or best practices, the participants were reluctant to place great confidence in their estimates of best and current practices.