Abstract
Background:
A data monitoring committee (DMC) has special responsibilities for protecting the safety of clinical trial participants. Few guidance documents are available that address the operations and mechanics of establishing, serving on, or reporting to a DMC. This article provides a practical guide to sponsors, institutions, and individuals responsible for, or serving on, a DMC.
Methods:
A workgroup of professionals from academia and not-for-profit and commercial organizations that included investigators, statisticians, patient advocates, and ethicists met to define the essential elements of planning, coordinating, and populating a DMC. All members of the group have formed, served on, advised, or worked with DMCs.
Results:
The group outlined the objectives and mechanics of running a DMC, including operational and practical considerations, membership characteristics, roles, members’ liability, and indemnification. Further, it delineated the roles and responsibilities of each DMC member.
Conclusions:
The group recommended practices for each phase of the DMC process from inception through execution of a clinical trial, with appropriate considerations for confidentiality. The group’s practical guidance should assist in comprehensive oversight of appropriate clinical trials and should help DMC members execute their obligations with greater assurance.
Keywords
Introduction: Objectives and Background
A data monitoring committee (DMC) i is an independent panel of experts whose principal function is to review the conduct and data of an ongoing clinical study, typically one with a randomized and often blinded design, with particular emphasis on the safety of study participants. The DMC is composed of medical experts, biostatisticians, ethicists, and others with experience in the conduct of clinical trials and the therapeutic area of the study. Unlike other individuals and committees responsible for execution of a trial, members of the DMC are typically unblinded to aggregate or individual data by study arm during the conduct of the trial. The DMC provides independent advice, usually to the sponsor or its designee, on the continuing scientific validity and safety of the trial and the efficacy of the therapy under investigation. Many DMCs are given additional quality control responsibilities including, but not limited to, reviewing procedural aspects of the trial such as verifying that appointments, laboratory tests, and procedures are accomplished within designated time “windows”; examinations (eg, pathology of biopsies) are carried out by properly certified professionals; and assessing whether certain attributes (eg, gender, race, body mass index, disease severity) are balanced. DMCs that are responsible for implementing interim analyses typically check that endpoints are being reported in a timely fashion and adjudication of endpoints appears to be proceeding quickly and accurately. The scope of a DMC might not be limited to the monitoring of a single trial; some DMCs review a set of trials that are part of a drug development program or a clinical research network.
DMCs differ from other committees that provide oversight to clinical trials (see Table 1). As often the only committee that has access to unblinded data in a double-blind trial, the DMC plays an important role in monitoring patient safety. The DMC communicates its recommendations to the sponsor or its designee who is responsible for making a final decision on whether to continue or stop the trial or modify the protocol. Thus, the appropriate constitution and organization of a DMC enhances the execution of safe clinical trials. 1
Responsibilities of Research Ethics Committees/IRBs, CECs, and DMCs.
Abbreviations: CEC, clinical endpoint committee; DMC, data monitoring committee; IRB, institutional review board.
Methods
In 2013, the Multi-Regional Clinical Trials Center of Brigham and Women’s Hospital and Harvard (MRCT Center) embarked on a program to train international clinical trialists, statisticians, and regulators on the principles, processes, expectations, and responsibilities of DMCs, with particular attention to individuals in developing countries for trials occurring in their country. In the course of this effort, trainees requested a short practical guide for individuals who are establishing, serving on, or reporting to a DMC to supplement their training. 2 We assembled professionals from various fields to develop a practical, role-based guide for DMC operations and decision making.
A workgroup of 24 members defined the essential elements of planning, coordinating, and populating a DMC based on existing materials and their own experience. All members of the group had formed, served on, advised, or presented to DMCs. Materials were further refined iteratively. An initial set of training materials was piloted in 2013. Questions arising during subsequent training sessions led to modifications of materials that lacked clarity after each training session. Training sessions have been held in the United States, Korea, India, Japan, China, Thailand, and South Africa, with a total of more than 300 participants.
Results
Practical Considerations in Constituting a DMC
All interventional clinical trials require some form of monitoring to assess safety and quality of data but not all require a DMC. ii Trials sponsored by United States National Institutes of Health (NIH) require a data and safety monitoring plan that outlines the process for safety monitoring. Many early phase trials, many social and behavioral studies, and some short-term trials of low risk to participants do not require a DMC. The classes of trials likely to utilize a DMC include (1) large randomized multicenter trials, (2) blinded interventional trials, (3) all NIH-supported phase III interventional trials, (4) trials that expose the participants to a high-risk intervention or an intervention that has the potential for serious toxicity, (5) trials with outcomes of mortality or clinically deleterious morbidity (eg, heart attack, stroke, tumor progression), and (6) some trials conducted in a vulnerable study population (eg, children, pregnant women) or with the inclusion of participants who are unable to make informed decisions, who are at particularly high risk, or who are not literate. 3 In addition, a DMC may be useful where the sponsor may wish to stop the trial early if the interim data demonstrate either strong evidence of benefit or convincing evidence of lack of benefit. Statistical methods allow interim analyses to assess benefit in a way that preserves the statistical operating characteristics of the trial (ie, maintains nominal type I error rate and power).
The organization and membership of a DMC will often depend upon the nature of the trial, the DMC charter, and the expertise necessary for appropriate oversight of the specific trial. A DMC typically must have individuals with certain areas of expertise, including (1) the clinical disease under study, (2) interpretation of data from clinical trials, (3) ethical concerns relevant to the trial population, (4) statistical analysis and interim monitoring, (5) evaluation of safety, including of serious adverse event (SAE) reports, and (6) the regulatory environment in which the trial is conducted. DMC statisticians should be conversant in the specific planned statistical analysis of the trials they monitor. Occasionally, additional expertise is required. Examples include a safety concern that requires specific knowledge (eg, a neurologist to evaluate neurologic adverse events in the study of an oncology drug), knowledge of specific diagnostic techniques or biomedical engineering, or specific understanding of cultural and ethnic issues. Some DMCs include a patient advocate, that is, someone who has had the disease or who has a close family member with the disease. While any one member may be proficient in more than one area, the DMC membership collectively should have the expertise to review and address issues predicted to arise; ad hoc members may be added for unanticipated challenges.
The Sponsor-DMC Partnership
In addition to having functional expertise, the DMC must work well with the sponsor (or its designated liaison to the DMC). The DMC should review and abide by its charter, document their deliberations, and communicate effectively with the sponsor. In addition, DMC members must have sufficient time to, and dedicate the time to, review meeting materials in advance and to attend DMC meetings. The DMC is an advisory board, and as such, should explain the reasons for their recommendations. If a recommendation is not to make a change, the “explanation” should be succinct and not include information that could compromise the integrity of the trial.
Sponsors should form a DMC with whom they can work but one that will have sufficient stature to provide independent recommendations. iii Further, sponsors should anticipate both human and financial resources required to support the DMC.
DMC Communication and Documentation
The DMC charter (see below) should include a communication plan, specifying to whom the DMC reports and makes its recommendations. If the sponsor is a public agency (eg, the NIH or the Veteran’s Affairs Administration in the United States, or the Medical Research Council in Great Britain), the DMC will generally report directly to that entity. On the other hand, if the sponsor is a commercial entity (eg, a pharmaceutical company), the reporting structure will depend on the trial and be specified in the charter. In many cases, the DMC reports directly to the commercial sponsor or the sponsor’s designee. In other cases the DMC may report to an independent steering committee that has been granted authority by the sponsor to make decisions based on the DMC’s recommendations. Communications between the DMC and the sponsor (or designee) may be privileged, thus encouraging direct and honest, but appropriately limited, exchange. 4 iv
The DMC chair should discuss expectations for communication and emphasize the imperative of confidentiality at the first meeting of the committee. The DMC, typically through the DMC chair, should only communicate with individuals as specified in the charter. DMC members should avoid ex parte communications and casual statements. v The chair should also discuss expectations regarding communication among the DMC members themselves and whether they may or may not independently consult with each other or members of the independent statistical group outside of convened committee meetings. The DMC chair should decide whether there should be written communications and to whom they should be addressed.
Communication between a DMC and other parties requires a balance of openness and confidentiality. The DMC must make its recommendations clearly and in a way that preserves the integrity of the trial. Thus, any communication should provide sufficient, but minimal, information for the sponsor (or designee) to consider the DMC’s recommendations. Shared DMC meeting minutes are terse and often refer only to information that is available in open session vi (eg, “accelerate recruitment”), and its recommendations are concise and succinct (eg, “The DMC reviewed data from the trial and other relevant information. On the basis of that review, the DMC sees no cogent reason to modify the trial.”). Or, for instance, if there is an imbalance of safety events across sites or suspicion of unblinding at a certain site, the DMC might recommend that the sponsor re-educate all sites on proper safety reporting and blinding procedures without disclosing the specifics of the concern. Some DMCs create closed session minutes that remain confidential until the trial is complete and is unblinded. Others create summary notes instead of minutes. Generally, neither the minutes nor the notes specify which members said what.
Frequently, research ethics committees/institutional review boards (IRBs), and sometimes regulatory agencies, wish to receive a letter signed by the DMC chair attesting to the fact that the DMC has met and summarizing the most important recommendations. In many studies, the chair then writes two letters—one very short document that the sponsor sends to the research ethics committees/IRBs and regulatory agencies (“The DMC met on XX date and sees no cogent reason to modify the trial at this time”) and a longer letter that amplifies and details the nature of any specific recommendations for the sponsor. Unless what the DMC recommends leads to a major change in the protocol or to stopping the study, no communication should divulge any information that could unmask the blind or the outcomes in the trial. DMC minutes should document the sponsor’s responses to previous DMC’s recommendations. In addition to bringing closure to outstanding items, such written documentation attests that outstanding issues have been addressed. It may be of particular value in response to subsequent external requests.
Whether the DMC reports its recommendations directly to the sponsor, or the sponsor has given responsibility to decide on DMC recommendations to a separate group, there may be circumstances where the DMC’s recommendations are rejected. These situations should be quite rare, particularly when the recommendation is in response to what the DMC perceives as risk of harm to study participants, and will typically reflect a fundamental disagreement in interpretation of the data. vii The DMC should be informed of this decision but has no authority to overturn it; often, such disagreement represents an opportunity for the sponsor to clarify data or present additional data or explanation, and the DMC might modify its recommendation. In rare circumstances when disagreement persists, the DMC might consider taking the rather drastic step of resigning from the study; even in this event, DMC deliberations remain confidential.
DMC Membership and Liability
Because DMC members may be individually and jointly liable for committee recommendations, all members should be comfortable with the insurance provisions that cover their participation. Sponsors should describe the financial liability that members might incur if they agree to serve, and corporate sponsors should include provisions for indemnification and legal costs in their contracts with DMC members. 4 In these provisions, the sponsors should include procedures to minimize potential conflicts of interest between sponsors and DMC members. viii Many government sponsors, however, do not possess such insurance; therefore, DMC members may wish to seek their own professional insurance and ascertain whether these insurance premiums are reimbursable. 5
Process: Practical Suggestions for Each Phase of the DMC
Ideally, sponsors should convene the DMC and hold an inaugural (“kickoff” or “organizational”) meeting prior to enrolling the first participant in the trial. Such a meeting should include not only the DMC members but also the reporting statistical group. If a kickoff meeting prior to the start of the study is not possible, the first meeting should be held as early as feasible after the trial commences. All members of the DMC (excluding ad hoc members) will be involved in the following phases:
A. Development and review of the charter
Every DMC should have a written charter that delineates its responsibilities, the roles of its individual members, and the standard operating procedures for the conduct of the committee (see Table 2). Prior to the first meeting, the sponsor or designee should provide members with a draft charter for review. ix The DMC should review the draft at its first meeting. The charter should empower the DMC to evaluate the trial data effectively. Appropriate safeguards should be in place with respect to access to and communication of ongoing results of the trial. x Questions or clarifications concerning the charter should be addressed at the first DMC meeting.
The Basic Elements of a DMC Charter.
Abbreviations: DMC, data monitoring committee; IRB, institutional review board.
B. Kickoff meeting
DMCs should hold a kickoff organizational meeting in which all DMC members participate. The meeting should include discussion of the rationale of the trial, the information known about the intervention under study thus far, finalization of the charter, the sponsor’s rationale for the DMC, and expectations of the DMC. The sponsor should describe prior relevant interactions with regulatory bodies. For multiregional trials, the sponsor should explain special regulatory requirements in the countries where the trial is being held. xi The kickoff meeting allows the opportunity to explore what actions the DMC might take in hypothetical scenarios.
C. Periodic and follow-up meetings
All DMC members should be prepared to attend periodic and ad hoc meetings, many of which are held by teleconference or videoconference. Prior to any meeting, DMC members should have reviewed all distributed information (eg, meeting minutes, the DMC report, serious adverse events) (see Table 3). DMC members should be prepared to discuss potential issues at the start of the open session if the issues are ones that can be discussed with the sponsor present, and at the closed session for issues that should not be discussed with the sponsor present. DMC meetings are focused on problems; the minutes should capture issues and their resolution to facilitate preparation of future meetings. Procedures for calling for special follow-up meetings should be defined. Review and discussion of the study data will allow the DMC to assess the integrity of the trial, the safety of research participants, and the appropriateness of continued random assignment.
Study Data Provided to a DMC for Review.
Abbreviation: DMC, data monitoring committee.
D. Ad hoc or emergency meetings
DMC members must recognize that they may have to attend occasional ad hoc meetings. While rare, these meetings constitute a core responsibility. There is often little time to prepare for an emergency session; the questions and supporting analysis are discussed during the meeting. Emergency sessions are convened when significant new information may potentially cause the DMC to recommend modification or curtailment of the clinical trial or program. xii
E. Meeting format: open, closed, and executive sessions
Open sessions
The first kick-off meeting often has only an open session with representatives of the sponsor, the steering committee, the reporting statistical team,
xiii
and all members of the DMC present. In open sessions at subsequent meetings, the sponsor informs the DMC about the trial status and describes questions it would like the DMC to address. Full disclosure of the progress of the clinical trial program helps the DMC to perform its job effectively. The sponsor should provide up-to-date information about serious adverse events with particular attention to events that occurred since the study documentation was prepared (eg, events that occurred in the period between report preparation and the meeting itself). In addition, at this session, the DMC and sponsor may review an
Some DMC meetings end with a final open session during which the DMC presents its recommendations to the sponsor. If the DMC recommends a modification to the protocol, the sponsor needs to understand the rationale behind the recommendation, decide whether to accept the recommendation, and, if they do, move expeditiously to implement the change. xv
Closed sessions
For industry trials, the participants of the closed session portion of the meeting include the DMC and the unblinded and independent reporting statistical team. For public-sector trials, the trial statistician and some sponsor representatives may attend as well. xvi During this session, the reporting statistical team presents trial results concerning safety by study arm. It will also present efficacy outcomes by study arms if an interim look at unblinded data is planned or if the DMC needs data on efficacy in order to assess the balance of risks and benefits. xvii
Executive session
For industry, only the DMC attends the executive session portion of the meeting. xviii The DMC may use this session to discuss the quality of the report and data being presented and to formulate its recommendations to the sponsor of the study. xix
Discussion: Practical Suggestions for Roles and Responsibilities Related to DMCs
The qualifications and responsibilities for DMC members must be based on the type and scope of the clinical trial. Roles should not be confused with individuals—some people are capable of filling more than one role and, conversely, some roles may require multiple people. We have segregated the roles into four classes: sponsor, the reporting statistical team, the administrator for the DMC, and the DMC itself. In the last class, we also describe roles of specific members of the DMC. For each class, we describe the role definition, qualifications, and involvement in the functioning of the DMC (see Tables 4 and 5).
Abbreviation: DMC, data monitoring committee.
*Patient advocate is not a required member of most DMCs.
aActions that have ethical implications include many considerations in which the interests of one group of stakeholders (eg, sponsors who wish to extend a trial for the collection of secondary endpoints) are in actual or potential conflict with those of another group (eg, participants who would benefit from termination of the trial and knowledge of the primary outcome). Patient advocate is not a required member of most DMCs.
Roles of DMC Teams During Charter Development and Kick-Off Meeting, and in Subsequent Meetings.
Abbreviation: DMC, data monitoring committee.
The sponsor, the reporting statistical team, and the DMC each have specific and different responsibilities in order to protect the subjects in the study and to ensure that the study is conducted rigorously. DMCs are constituted for general safety surveillance, formal interim efficacy analyses, and central risk monitoring; there is flexibility in implementation. Regardless of its structure, an effective DMC must be a true partner of the clinical trial sponsor and the investigators. As in any partnership, the relationships have multiple interdependencies as well as potential conflicts, which must be identified and navigated successfully.
The independence of the DMC is central to the committee’s operation. From the DMC’s perspective, “independence” means that its remit is sufficiently flexible to execute its charge: to protect the safety of the participants and to ensure scientific validity. xx “Independence” for the sponsor implies that sponsors not influence the decision making of the DMC and freedom from conflicting interests in the DMC members themselves—for example, DMC members should not be selected for their future marketing potential. Sponsors typically compensate DMC members for their time, effort, and expertise; therefore, awareness of potential conflicts, other than this singular financial relationship, is particularly important. DMC members must disclose potential conflicts of interest before joining the DMC and at the start of each meeting. Examples of potential conflicts of interest include salary, consultancy, stock in either the company whose product is being tested or a competitor of that company, professional relationship with study investigators, and public positions indicating lack of equipoise. Some of these conflicts may be managed rather than disqualifying a prospective DMC member, but they must be disclosed transparently. “Independence” for the reporting statistical team suggests that their primary role and obligation is reporting to the DMC, not to the sponsor that provides its financial support. Unlike the trial statistician, members of the reporting statistical team are not typically involved in the final statistical analysis for the study. xxi The initial scope of work should include not only the analysis of expected adverse events and efficacy, but anticipate additional analyses that the DMC may request.
The US FDA issued a draft guidance (Safety Assessment for Investigational New Drug Application Safety Reporting, December 2015) that recommends sponsors institute a Safety Assessment Committee to review accumulating safety information in an unblinded fashion regularly across all trials in a development program. 7 This committee, if the guidance is enacted as currently envisioned, differs from a DMC in that its role is to recommend whether or not the sponsor should submit an Investigational New Drug (IND) safety report to the FDA. In contrast, the DMC’s recommendations are focused on modifications of a trial. Most DMCs are not currently constructed to serve as safety assessment committees. If this guidance is finalized, it is likely that the function of these two committees will be complementary.
Conclusions
Over the past two decades, clinical trials have become increasingly global 8 ; the composition and demographics of DMCs have not changed concomitantly. The pool of experienced DMC members outside of the US and EU has remained small. Country-specific regulations are beginning to appear. For example, the China FDA (CFDA), in its final “Guidance for International Multicenter Clinical Trials (IMCT),” suggests that studies with more than 20% of Chinese patients include experts from China in their DMC. 6
While it appears to be important to include DMC members who appreciate the needs and expectations of diverse participants in the trials, it is usually not feasible to represent all regions in the membership of the DMC. That said, greater representation from participating nations is an important and achievable goal. 8 The MRCT Center has been training DMC members since 2012, selecting midcareer physicians, statisticians and ethicists with strong clinical trial experience to participate in a one-day training program with leading clinical trialists, physicians, and statisticians. This guidance, intended for DMC members newly embarking on their responsibilities, provides practical knowledge about the day-to-day responsibilities, terminology, and roles required for them to fulfill their duty to trial participants.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.