TL14: Preparing Effective Interim Reports for a Data Monitoring Committee
Robin Bechhofer, University of Wisconsin  *Melissa K Schultz, University of WIsconsin 

Keywords: DMC, interim reports

In many clinical trials, an independent Data Monitoring Committee (DMC) is responsible for reviewing accumulating data and safeguarding participant safety and trial integrity. The DMC relies on multiple sources of information, primarily a DMC Report on emerging trial data received at regular intervals from a reporting statistician or statistical group.

Providing clear and effective reports is important for allowing DMC members to fulfill their responsibilities, but it is not a simple task, and good advice on best reporting practices may be hard to find. Because of the confidential nature of DMC operations, sponsors and CROs may have a limited understanding of how DMCs actually use these reports, and DMC members themselves may find it difficult to articulate their needs before seeing actual displays of real data. Often the responsibility falls on the reporting statistician to take an active role in designing effective reports for a DMC.

Our group works as an independent analysis center supporting DMCs monitoring clinical trials; our focus is on preparing comprehensive graphics-based reports, with flexibility to adapt the content in response to requests by DMC members or changes in the trial or data collection. We have encountered many challenges unique to interim reports, in contrast to CSRs which are based on the sponsor's pre-specified SAP for analyzing clean, complete data.

Some considerations include: tips for working with dirty data (with inconsistencies, missing values, bad dates, etc.); how to schedule data transfers (balancing the need for timely data with adequate time to generate and QA the report); choosing among graphics, tables, and listings for data displays; deciding whether to provide the DMC with a report that is comprehensive vs. focused on key data (understanding the DMC's need to balance risk/benefit); maintaining flexibility in preparing reports vs. relying on fully pre-specified or pre-programmed analyses; and whether the DMC should have knowledge of the treatment arm identities (fully or partially unmasked).

We are interested in discussing these issues and hearing about others' experiences, good or bad, with DMC reports. In particular, for attendees who have served on a DMC or as a reporting statistician preparing or presenting interim DMC reports, what attributes have you found to be most (or least) helpful? Do you have recommendations for best practices, or for ways to disseminate and share such ideas with a broader audience?