Ask the Experts: Setting the Data Monitoring Committee Up for Success
This monthly feature presents a variety of questions from clinical trial professionals with answers from WCG’s expert staff. This month features Matt Downs, a statistical scientist with WCG Statistics Collaborative.
Question: How important is the first data monitoring committee (DMC) organizational meeting and when should it occur?
Answer: For a DMC, that first organizational meeting sets the stage for the DMC’s future data reviews. It should occur prior to recruitment and have a robust agenda.
I like to use the analogy of a boat in a port: before you set sail, you want to make sure you’re headed in the right direction. You’re going to plot your course and check the navigational devices before heading out to the ocean. That’s why the organizational meeting needs to take place before recruitment begins. This meeting is the last chance for the DMC to identify potential trial obstacles and offer feedback while its members are still naïve to the study data. Any requests after the DMC begins to review data will make sponsors wonder, “What’s really going on? What’s behind that question?” It’s much cleaner to be able to have those discussions before the DMC has seen interim data.
There’s another reason to meet before recruitment begins: some potential DMC members will refuse to serve because holding the organizational meeting post-recruitment suggests that the sponsor doesn’t have its safety monitoring process fully established.
Q: How should I begin planning the meeting and how should it be held?
A: To get started, you’ll need an agenda that includes a review of past trials, the study design and the template for the interim reports the DMC will review. And, although it may change because of enrollment, you want to give thought to scheduling the date of the first data review meeting.
Ideally, this first meeting should be in person. Yes, that makes scheduling difficult, but keep in mind most DMCs will be working together for years. It’s useful to meet the other committee members and sponsor representatives — and the independent statistical team that will be preparing interim reports.
Q: What are the most important questions to address with the DMC?
A: It’s important to hash out the following with the DMC early on:
- Will the DMC have access to efficacy data? This is one of the stickiest issues that comes up. It’s best to address it at the organizational meeting. The DMC members must have the ability to access efficacy data, even when reviewing only safety, because they need the ability to assess risk vs. potential benefit;
- How current will the data be? Too often, this topic gets overlooked, but DMCs need to know — if for no other reason, to control expectations;
- What is the level of masking? Determine the level of masking the DMC expects in its closed reports. Interim DMC reports often semi-mask the treatment groups with labels such as “Group A” and “Group B.” Sometimes, committee members will feel that the DMC should only know the true identity of the treatment upon request if needed given the emerging data. However, we generally recommend that DMC members should be aware of the treatment group identities corresponding to the semi-masked labels at the DMC’s very first report;
- How will the DMC handle futility/overwhelming efficacy? State in the charter whether one of the DMC’s charges is to consider a recommendation for trial termination due to futility or overwhelming efficacy. If this isn’t part of the DMC’s charge, the sponsor should explain why during this first meeting; and
- Can the DMC hold unscheduled meetings? The charter should give the DMC the authority to conduct unscheduled meetings without notifying the sponsor. For instance, if the DMC sees a potential emerging safety signal, it may choose to meet sooner than the next planned meeting.
It’s also important to conduct a thorough review of the protocol. DMC members will often freely give you their opinions on your study design. It is also a time to review prior studies for this and other indications. What do you know about the existing safety profile of the drug? This gives committee members a sense of what to expect and the ability to recognize if there is a new safety issue that wasn’t uncovered from prior studies.
Q: How should communication flow between the DMC and the sponsor?
A: We encourage all study communication between the DMC and sponsor to go through the Statistical Data Analysis Center (SDAC), which can make sure all messages are appropriately routed. Think of it as a firewall. The SDAC goes by various names, but the task is the same: it prepares and presents the safety and efficacy reports for the DMC’s reviews at their interim meetings.
An SDAC statistician … will attend the DMC meetings and have access to unmasked data. As an independent organization, it has no vested interest in the outcome of the trial. Usually, at least two other statisticians will be participating: the DMC statistician, who is a member of the DMC and typically sees unmasked data, and the trial statistician, who typically has access only to masked data that are pooled over treatment group.