Interim Reports for Data Monitoring Committees of Clinical Trials
Scott Diegel, University of Wisconsin - Madison, Department of Biostatistics and Medical Informatics, SDAC  Melissa K Schultz, University of WIsconsin  *Ryan Zea, University of Wisconsin - Madison, Department of Biostatistics and Medical Informatics, SDAC 

Keywords: Interim analysis, DSMB, DMC, clinical trials, data visualization

Data Monitoring Committees (DMCs) are charged with the task of safeguarding both trial participant safety and the integrity of the clinical trial for which they serve. This role requires making sense of a large amount of wide-ranging, complex, and incomplete information generated from multiple sources of data in the trial. An interim DMC report provides a comprehensive look at the trial, as a whole, as the information emerges and often plays a vital role in the decisions of the DMC.

An effective interim report presents the accumulating trial data as easily understood summaries, while providing as complete a picture of the current state of the trial as possible. This daunting task is made even more challenging by the confidential nature of DMC operations, leaving sponsors and contract research organizations with limited knowledge of how DMCs actually use interim reports and what constitutes a “good” report. Flexible reporting plans are required in order for DMCs to be able to respond to emergent, unexpected trends in the data.

The University of Wisconsin–Madison Statistical Data Analysis Center (SDAC) works as an independent analysis center providing DMCs with interim reports based on accumulating data from ongoing trials. We address issues unique to interim reporting such as inconsistencies between data sources, bad dates, and balancing timeliness versus cleanliness, while maintaining the flexibility to respond to DMC requests. Our reports focus on comprehensive graphical representations of data so meaningful signals are relatively easy to detect. Ancillary backup tables and supplemental listings are provided, detailing the information found in the graphics.

This poster presents some of our best practices in methodology and procedures in hopes of opening up conversations among reporting statisticians, sponsors, and DMC members. Using simulated trial data, we present several sample graphical displays representative of the wide range of pages seen in an actual DMC report.