Abstract
Background
A major goal of the National Institutes of Health’s Clinical and Translational Science Award program is to facilitate clinical research and enhance the transition of basic to clinical research. As such, a number of Clinical and Translational Science Award centers have developed services to facilitate the conduct of clinical research, including support with fulfilling regulatory requirements.
Methods
The University of Kentucky sought to establish an institutional semi-independent monitoring committee to provide oversight for clinical research studies per National Institutes of Health requirements and recommendations. Our semi-independent monitoring committee was initiated in 2010.
Results
Since the inception of our semi-independent monitoring committee we have restructured its operations and protocols to improve efficiency. This article discusses our experiences with semi-independent monitoring committee creation and growth.
Conclusion
This article summarizes our experience in creating and maturing an institutional data monitoring committee.
Keywords
Introduction
In 2011, the University of Kentucky was awarded a Clinical and Translational Science Award titled “The Center for Clinical and Translational Science.” The underlying theme was to address chronic health disparities in rural and underserved populations throughout Kentucky and the Central Appalachian region, and to foster innovative team science. A major goal of the Clinical and Translational Science Award programs in general, and our Clinical and Translational Science Award in particular, is to build and grow clinical and translational research. There are a number of well-described challenges to conducting clinical research, including regulatory requirements. One of our primary objectives is to implement standard operating procedures and resources to help faculty transition their research ideas to clinical and translational steps. Several actions we initiated included regulatory assistance with Institutional Review Board (IRB), Food and Drug Administration, and Investigational New Drug submissions and reporting, and implementation of an institutional, semi-independent data safety monitoring committee to provide oversight and safety monitoring for institutional clinical studies that do not meet requirements for truly independent data and safety monitoring.
In 1998, the National Institutes of Health (NIH) started to require data safety monitoring boards for phase 3 multicenter clinical trials. Subsequently, federal agencies including NIH, as well as local IRBs, have started requesting data monitoring boards for additional types of clinical studies. Typically, a data monitoring board (rather than a data and safety monitoring plan, which is required for any study involving humans) is suggested when a study involves high risk interventions, vulnerable populations, or phase 3 studies (Table 1). For trials meeting these criteria a truly independent data monitoring board is required; however, not all studies meet the threshold for an independent data monitoring board. Our goal was to create an institutional semi-independent data monitoring committee (IDMC) to serve investigators needing data and safety monitoring but not meeting the threshold of needing a truly independent data monitoring board.
Categories of studies for data safety monitoring board review.
There is significant literature on what a data monitoring board “should” do,1–5 including literature on challenges they face, particularly in regard to decisions about terminating studies early.6,7 In addition, there are several text books and government documents that address their operations.8–12 However, despite these resources we faced many challenges with the initiation and operation of a new IDMC, in part as operating an institutional IDMC has differences and challenges with regards independence, compared to a true independent data monitoring board. With the goal of supporting and guiding IDMC development at other institutions, whether it be institutional or by individual investigators, we share our story and protocols.
Institutional data monitoring committee development
Meeting investigator needs
Prior to the creation of our IDMC, investigators formed and managed their own data monitoring committees if needed. Thus, one challenge during our IDMC development was to form a committee that would broadly serve our research community. Some investigators expressed concerns that a core committee may not have the composition or expertise to address particular issues relevant to each individual study. In addition, there were concerns that an IDMC may limit investigator interactions, or not be conducive to active ongoing protocols. We acknowledged these concerns, and structured our committee to minimize inflexibility and maximize investigator interactions, as outlined below.
Limiting excessive use
Shortly after creating the IDMC, we realized we needed to address another challenge: limiting unnecessary use. Many investigators, especially those relatively new to human research, “bend over backwards” to demonstrate their compliance with regulatory requirements during grant applications, possibly in part due to time pressures. All studies involving humans require a data and safety monitoring plan; this can be quite challenging, lengthy, and time consuming to write, and may require inclusion of additional personnel to meet funding agency and/or IRB approval. Thus, once an institutional IDMC was available, we found that investigators were often defaulting to a simple statement that “the IDMC will be used to monitor this project” in research proposals. We heard that investigators felt that stating they would use an IDMC both saved valuable text space and improved their likelihood of favorable review of their human studies protocols, even if they didn’t technically meet criteria for an IDMC. Furthermore, since the IDMC does not charge investigators for the service, there were no fiduciary barriers to limit IDMC requests. While some of these studies were appropriate for IDMC use, for others an IDMC was not necessary. Without any pre-approval or processes from the IDMC for accepting studies, we found ourselves at risk of drowning in an onslaught of studies. To limit excessive demand, we have implemented a process whereby an investigator can request IDMC oversight: the study investigator meets with the IDMC chair who reviews the protocol to determine if the study truly needs a IDMC, and if there is a more appropriate alternative given the IDMC’s present workload and the expertise needed for the study. If the study is deemed appropriate for the IDMC then the investigator is provided with a letter that can be included in the grant application. In addition, our IRBs will not accept a statement that the IDMC will be used without a letter from the IDMC indicating such.
Goals and responsibilities
The IDMC’s responsibility is to assure the safety of research participants, efficacy and appropriateness of study interventions, relevance of the study questions, and integrity of the accumulating data throughout the life of a research project. As our IDMC is funded and supported by our institutional Clinical and Translational Science Award, the goal of our IDMC is to support affiliated researchers, and manage studies carried out under the auspices of the Clinical and Translational Science Award, including multi-site studies.
Institutional data monitoring committee membership
The IDMC has a standing membership comprised of a small group of faculty including at least two physician-scientists with clinical research experience (one of whom is designated the Research Subject Advocate), a research pharmacist, a biostatistician, and a Safety Officer (usually an administrator with regulatory expertise) who comprise the core multidisciplinary team. This comprises four voting members (the Safety Officer is a non-voting member). An executive secretary records minutes and is the point of contact for all studies. This individual attends all IDMC sessions but in a “mute” role thus does not participate in discussion and does not have a vote, which is our strategy to minimize any potential conflicts of interest that this individual may have with any study or the institution. We found it useful to have at least one member with IRB experience. Ad hoc members are appointed as needed to provide specialized expertise related to a specific research study. Ad hoc members have voting rights only for the specific study they are reviewing. The statistician is a critical member of the IDMC and needs to be completely independent from the trials. While our practice may differ from others, we have found it useful to appoint a primary statistician as a member of our IDMC, with an alternate readily available for potential conflict. The statistician is the liaison between the IDMC and the data, and provides analyses of emerging evidence as needed. We have three statisticians who each have effort allocated to (and paid by) our Clinical and Translational Science Award. One of them is appointed to the IDMC, and the other two serve as alternates. We do this, in part, due to limited resources; for a truly independent data safety monitoring board, there should be an independent statistician or data analysis center that liaises between the independent data safety monitoring board and the study. 4 We have found that keeping our membership small but with each member assigned a specific focus is practical, nimble, and provides adequate scientific and human protections review, with the assistance of an ad hoc expert when needed. To date, we have not included either an ethicist or any lay members. While the expectation is that all decisions are made by consensus; we have elected to include a voting process with a majority approval. Liability issues for IDMC members is a potential significant issue.4,13 Institutional legal counsel has reviewed and approved the IDMC charter and approved indemnification for IDMC members for review of institutional studies.
Institutional data monitoring committee chair
The chair has a critical role in the integrity of the IDMC process and thus the selection of this individual needs to be considered carefully. Ideally, the IDMC chair will have extensive personal experience and background in clinical research, with expertise in regulatory issues and proper training in the data safety monitoring board process, perhaps by prior participation in a true independent data safety monitoring board. However, a novel IDMC may be limited in the options for a chair. Our experience to date has been to identify a faculty physician-scientist with the clinical research experience and interest in IDMC operations and provide “on site” training by having that individual participate in IDMC meetings for several cycles prior to transition to the chair role. However, we recognize that an IDMC chair should not be restricted to physician-scientists, and individuals with other expertise such as statistics, ethics, and so on would be appropriate to serve as chair should they have the appropriate training and experience.
Conflict of interest policy
Any potential conflicts of interest whether actual or perceived are reported to the IDMC chair. These conflicts can include financial, professional, intellectual, proprietary, and other actual or potential conflicts. One of the more common conflicts of interest that has arisen is a conflict regarding reporting structure and responsibilities (such as if a study investigator is the supervisor of an IDMC member, or vice versa, or even if a study investigator and an IDMC member are appointed within the same administrative unit). This is perhaps not surprising when the IDMC members and study investigators all belong to the same institution. As our IDMC members are active researchers themselves, a member may be in conflict if they are a contributor to the study under review. In these or other circumstances involving actual or perceived conflict, we have simply identified an alternate to substitute for the conflicted IDMC member for that particular study and the conflicted member recuses themselves from all matters related to the conflicted protocol. To date, we have always found a suitable alternate faculty to serve in this ad hoc role; however, at a smaller institution or for a trial arising from a small unit, it is possible that there may not be an appropriate alternative. In this case, seeking an alternative member from a different institution may be a solution. An additional potential conflict that is more difficult to address is that of institutional loyalty of the members. IDMC recommendations can influence an institution’s reputation and/or have financial implications. The IDMC chair and members need to be aware of this risk, and consider alternatives (such as referring a trial to an external, independent data safety monitoring board or bringing in an ad hoc reviewer external to the institution) when such conflict could occur.
Institutional data monitoring committee operations
The IDMC is managed and operated through the Clinical and Translational Science Award Regulatory Knowledge and Support core, with 0.25 full-time effort of administrative support from a clinical research associate. The IDMC webpage is the first point of contact and outlines the goal of our IDMC and steps to apply. An initial study assessment is performed either during study preparation or after IRB approval is obtained. The first step is to determine if the study needs an IDMC, or would be more appropriate for either a data safety monitoring plan or a true independent data safety monitoring board; this is particularly common when investigators submit requests during the grant preparation stage. If the IDMC declines the study, the investigator is provided guidance on developing a data safety monitoring plan or is referred to an independent data safety monitoring board, as appropriate. A major goal of the initial assessment is to determine if the core IDMC membership has adequate expertise for the study, or if an ad hoc member will be needed.
Accepted studies are provided with the IDMC forms used for periodic study evaluations (Figure 1). These forms provide standardization to facilitate the IDMC review of participant recruitment, accrual and retention, assessment of participant safety, performance of study site(s), quality, and review of data.8,9,14 At present, our IDMC meets 3 times a year. Each study is required to have the study investigator or a designated study member (usually a senior research coordinator) attend the IDMC meeting to address any questions on data provided. For each study the IDMC review process includes both an open session (the study principal investigator or personnel are present to answer any queries from the IDMC) followed by a closed session (whereby the study investigator is excused, and the IDMC discusses any efficacy and/or safety concerns in a confidential setting). The IDMC reports back to the investigator (in writing) on the safety and progress of the study; recommendations include continuation unchanged, modification of protocol, suspension, or study termination. After an investigator reports study completion, a close-out review is performed and is the only step by which a protocol is removed from the IDMC’s oversight.
Institutional data monitoring committee study review forms. (a) Study overview form. (b) Adverse event log. (c) Protocol violation log. Investigators complete/update these forms for each IDMC review. Definitions of terms listed in Table 2.
Operational challenges
Workflow processes and forms
A major challenge has been to develop a workflow that maximizes efficiency and provides needed information without excessive documentation time. We have found it most practical to provide individual training on the use of the forms (Figure 1(a)–(c)) with the investigator or study coordinator to ensure all details needed for IDMC review are understood. This training includes individual review on definitions of IDMC terminology and reporting (Table 2). All adverse events are reported in forms that depict both investigator/study staff assessment of causality (using terms definitely not, probably not, possibly, probably, definitely) and intensity (using terms mild, moderate, severe, death). Data are presented in text format, as well as by tables and graphs as appropriate for the individual study.
Institutional data monitoring committee form definitions.
PI: principal investigator; DSMB: Data and Safety Monitoring Board.
Multi-site protocols
Another operational challenge has been utilizing our IDMC for multi-site studies. To date, we have taken on only multi-site studies that are funded by NIH grants; with the increase in multi-principal investigator awards from NIH, we found a number of investigators had received grant funding for trials that included more than one institutional site, but we were often the only institution that offered an IDMC. Legal risks and responsibilities and indemnification issues related to multi-site studies are beyond the scope of this report and will not be discussed here.15,16 Operational challenges for the IDMC include “signal to noise ratio,” or the risk that adverse events at one site may go unnoticed by the IDMC due to the volume of data reviewed from multi-site trials. Our approach to minimize this risk was to require each site to report the data from their own site individually (thus treating each site as a separate entity) as well as reviewing a collective report from all sites; the collective data report is provided by the lead investigator or study statistician. We utilize teleconferencing to communicate with the investigator/study staff from each site. We charge a nominal fee to external sites for the extra work involved, but there is no charge for any local site or study. The IDMC reports are provided on each site separately and for the collective study; the site investigators share these reports with their IRBs.
Workload challenges
After a number of years, we came to the recognition that the IDMC workload needed to be monitored and capped. Our IDMC is small, which means a few members are responsible for reviewing all the studies, with ad hoc members providing study relevant expertise for studies as needed. We have determined that 15–20 active studies (acknowledging variable size and complexity) is a reasonable cap. We have just reached this point and are developing a second IDMC. We are facing the decision of reallocating existing studies to have each IDMC have an area of focus, or to simply assigning new studies to the second IDMC which will grow over time. One advantage of having each IDMC have its own focus is that the membership can specifically include faculty with experience to that area. Another advantage is that each IDMC would then be composed in part from existing members with experience, and in part from new members. However, the disadvantage of splitting the current committee and having area-focused IDMCs is that it risks disrupting the strong working relationship/teamwork ethic of the current committee, and a topic-focused IDMC may lose the breadth of scope that a diverse membership can have.
Institutional data monitoring committee faculty recruitment and effort
An ongoing challenge for the IDMC as well as other regulatory units is finding faculty with the appropriate skill set and finding time and/or funding for their effort. We do not provide salary support for IDMC members. To date, we have been successful in soliciting participation by focusing on two, non-fiduciary benefits: IDMC service is reviewed favorably by the promotions and tenure committee, and promoting faculty engagement. For many faculty, the ability to participate in high-level decisions across a broad scope is in itself rewarding; the small membership encourages camaraderie and mutual support. Finally, participation on the IDMC is well respected at our institution. Our IDMC is structured for episodic involvement without excessive time commitment. However, it is important to accommodate turnover of members as needed for their own professional needs elsewhere, and to keep the committee engaged and fresh. We recommend appointing members to terms (3–5 years) and rotating off members in a staggered approach. The ad hoc members that we call upon for particular areas of expertise are potential future members, and we ensure that we maintain good communication, consideration, and appreciation for their work. Study investigators who have used the IDMC themselves are also potential future IDMC members. When a new IDMC member is recruited we invite them to participate in an IDMC review meeting with the individual they are replacing so that they are mentored through the process.
Summary and conclusion
With growth and expansion of clinical research and growing regulatory requirements, more institutions or research units will need to develop core IDMCs to meet demands and maintain high standards for subject safety and research quality. While operations will by necessity be different between different institutions and organizations, the University of Kentucky IDMC now runs efficiently meeting many investigators needs. Our structure and scope allow for scalability and adaptation to needs as the research focus of the institution grows and changes. We have maintained strong IDMC membership despite lack of direct funding support for faculty participation. We have developed a process that facilitates required evaluations of study data without excess burden on either IDMC members or study staff. We have recently expanded to take on multi-site studies, and are now developing a second IDMC due to growth in demand.
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institutes of Health (NIH) National Center for Advancing Translational Sciences through grant numbers UL1TR000117 and UL1TR001998. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or the US Government.
