How to Run DMC? It’s Tricky – FDA’s New Draft Guidance Provides Updated Recommendations on How to Best Use Data Monitoring Committees in Clinical Trials

March 7, 2024By Mark A. Tobolowsky & Véronique Li, Senior Medical Device Regulation Expert & David B. Clissold

The trio of CDER, CBER, and CDRH released a new draft guidance titled “Use of Data Monitoring Committees in Clinical Trials” that revises the 2006 guidance “Establishment and Operation of Clinical Trial Data Monitoring Committees” and, when final, will replace the 2006 guidance.  The new draft guidance was published in recognition by FDA that Data Monitoring Committees (DMC) (aka Data and Safety Monitoring Boards (DSMB), Data and Safety Monitoring Committees (DSMC), or Independent Data Monitoring Committees (IDMC)) are increasingly being utilized by sponsors to implement adaptive trial designs, to review aggregate data for safety reporting, and to oversee an entire clinical development program rather than a single trial.  Moreover, DMCs are being used in trials of modest size and in the context of increased globalization of medical product development.  DMC charters have also grown longer and more detailed.  The new draft guidance is generally reflective of these developments, and we wanted to highlight several of the changes in this blog post.

One notable change in the new draft guidance is the language regarding when a DMC is recommended, reflecting the much broader adoption of DMCs since 2006.  In 2006, FDA stated that DMCs were not recommended “for most clinical studies,” particularly those in early product development, short-term studies, or studies addressing less severe outcomes.  Instead, FDA recommended that sponsors “limit the use of a DMC” to clinical studies in which safety concerns may be unusually high, such as studies that compare rates of mortality or major morbidity or involving a potentially fragile or high-risk population, or large, multi-center studies of long duration.  The new draft guidance maintains some of these recommendations (that is, a DMC is warranted when subjects in the study are at risk of serious morbidity or mortality, the study enrolls vulnerable populations, and for longer-term studies), but also highlights factors that might suggest the value of DMCs in new roles, such as where causation of adverse events may be difficult to assess without a review of unblinded data or where there is limited experience in a therapeutic area.  Along those lines, the new draft guidance does not contain any language regarding DMCs not being appropriate for early phase studies.

Other Oversight Groups

Both documents described the role of other oversight groups, in addition to DMCs, that may be involved in a clinical trial in similar, sometimes overlapping, roles such as IRBs, trial steering committees, and site monitors.  In another update, the recent draft guidance added “entities reviewing safety data” and adaptation committees.  FDA notes that the distinctions between responsibilities for these groups and DMCs should be clearly defined, particularly with respect to access to unblinded information.  For example, an entity reviewing safety data may need to review unblinded safety data to recommend whether or not the sponsor should submit a report to FDA.  However, in contrast to DMCs, such an entity should remain blinded to efficacy data.  FDA acknowledges that the threshold for a DMC to recommend termination or significant modification based on safety concerns may be higher than the threshold for reporting potential serious risks to FDA; this is reflective of the fact that there may be situations where oversight groups observe a serious risk and recommend that the sponsor report it to FDA, but the DMC may still recommend the trial continue based on its overall assessment of unblinded safety and efficacy data.  In such cases, the draft guidance illustrates how employing a DMC may allow a sponsor to proceed with a trial where it might otherwise have been terminated.

Both guidances note that the most common purpose of a DMC is to monitor clinical trials for safety, usually by conducting unblinded interim analysis of data from an ongoing clinical trial.  The new draft guidance acknowledges that an interim analysis may be appropriate in other circumstances, particularly when “[i]mplementing a predefined adaptive feature” in a study, such as to increase trial size or to introduce prognostic enrichment.  Adaptive study designs are becoming increasingly prevalent, and sponsors often use adaptation committees to perform such interim analyses. If the adaptations are to be done on the basis of unblinded data, the new draft guidance emphasizes the importance of prespecification and preservation of trial integrity.  The new draft guidance states that while a DMC could be assigned the role of recommending to the sponsor that a specific adaptive design element be implemented (if specified in the DMC charter), that might best be reserved for relatively straightforward adaptive designs with simple adaptation algorithms.  Use of a separate adaptation committee might enable the inclusion of more relevant expertise and allow the DMC to focus on its primary responsibilities – subject safety and trial integrity.

DMC Composition

Regarding the membership of the DMC, the new draft guidance is largely similar, but substantially less specific on its recommendations.  Both documents discuss how membership should include individuals without conflicts of interest and with expertise in trial conduct, relevant clinical specialties, a qualified biostatistician, and a medical ethicist (in studies with greater risk).  They also both stressed the value of previous DMC experience.  However, notably absent from the new draft guidance are some specific recommendations from the 2006 guidance regarding number of members, types of scientists, and representatives of the relevant patient population and from different gender and ethnic groups.  Another update in the new draft guidance is a recommendation that for the DMC chair, in addition to previous DMC experience (which was in the 2006 guidance), familiarity with FDA regulatory requirements is “typically critical.”

Increased Connections Between a DMC and FDA?

Both documents note that FDA may request that the sponsor submit the DMC charter to FDA for review before the performance of any interim analyses, and ideally before the initiation of the trial. Both documents also note that FDA may request copies of the DMC meeting records when the study is completed, and access to the electronic data sets used for each set of interim analysis.

However, the new guidance suggests that FDA may be recommending, or perhaps acknowledging, a greater degree of its involvement with DMCs.  For example, the new draft guidance recommends that the DMC should have “procedures for adding or removing members when appropriate or for disbanding the DMC, including procedures for informing FDA and disclosing to FDA the rationale for these changes.”  As another example, both documents note the possibility that if FDA has safety information relevant to an ongoing study, it may request that the sponsor confirm that the DMC is aware of that safety data and is taking it into consideration, or “request that the sponsor arrange for FDA to communicate with, or even meet with, the DMC.”  The 2006 guidance noted that these would be “rare cases” limited to “specific issues of urgent concern.”  FDA has dropped such restrictive language in the new draft guidance, and instead intends to refer “relevant and important” information.  As a final example, current regulations require a sponsor of an investigational medical device to report to FDA (and others) “unanticipated” adverse events, which are defined in part as “serious,” and sponsors of investigational drugs to report to FDA serious and unexpected suspected adverse reactions.  The new draft guidance maintains a recommendation from the 2006 guidance that “sponsors inform FDA about all DMC recommendations related to the safety of the investigational product, whether or not the adverse events that led to the recommendation meet the definition of serious.”  The draft guidance therefore continues to “recommend” sponsors go beyond what is required in its regulations in its communications to FDA based on DMC recommendations, highlighting the tension between what FDA needs and what it wants.  Taken together, such changes suggest that FDA may intend to play a larger role in its interactions with the DMC, either directly, or through the study sponsor.

Safety Monitoring

Regarding DMC responsibilities, the two guidance documents are fairly similar on safety monitoring.  In addition to any identified adverse events of particular concern, the 2006 guidance stated that the DMC should be provided with interim summaries of adverse events by treatment arm; the new draft guidance limits this to interim summaries of serious adverse events.  Similarly, the 2006 guidance stated that DMCs will not usually review in detail every adverse event reported, or even every serious adverse event; the new draft guidance is more agnostic on whether all serious adverse events should be reviewed (“the committee may elect to review all or just certain serious adverse events”).  The new draft guidance notes the practicalities of a DMC for short-term trials are not as well established as those for long-term trials.  It cautions that sponsors who consider DMCs for safety monitoring of short-term trials should ensure processes are in place to allow for timely DMC evaluations or oversight on data and safety.  Sponsors of short-term trials, such as those where enrollment is expected to be completed quickly or with short follow up, may find that establishing DMCs to review interim data/analysis may be “impractical and of little value.”


Both guidances acknowledge that DMCs should have access to unblinded interim data and analyses, including analyses of effectiveness, and emphasize procedures to ensure that the sponsor and investigators in the study remain blinded.  The new guidance specifically recommends that the independent statistician performing those analyses for the DMC should be “clearly firewalled and have no role in modifications of the trial conduct.”  The 2006 guidance also contained an admonition that prematurely terminating studies of less serious outcomes based on effectiveness is “rarely appropriate;” this is absent in the new draft guidance.  Although the sponsor must keep in mind the impact of an early termination, this change appears to be reflective of an acknowledgement that it is perhaps more than just “rarely appropriate” for DMCs to recommend early termination for effectiveness when data are compelling and a false positive risk is acceptably low, even where outcomes are less serious.


Sponsors considering the use of DMCs (or DSMBs, DSMCs, IDMCs) should carefully consider the recommendations and advice provided in this draft guidance, even prior to its finalization. As reflected in the changes from the 2006 guidance, the new draft guidance describes how DMCs are increasingly used for a variety of critical functions as an independent body with access to unblinded data from the ongoing trial.  It is also reflective of evolving FDA perspectives, in addition to evolving regulatory science.  In addition to adjusting its recommendations in acknowledgement of expanding DMC roles, the new draft guidance is also more focused on the potential utility of DMCs, with most of the language removed describing where they should not be employed (early phase trials, etc.).  This blog post focuses on a few of the specific changes from the 2006 guidance to highlight some changing perspectives, but sponsors should review the new draft in detail to fully understand FDA’s recommendations.