NIH policy (http://grants.nih.gov/grants/guide/notice-files/not98-084.html with additional description at http://grants.nih.gov/grants/guide/notice-files/NOT-OD-00-038.html) requires that grantees have in place procedures for DSM of clinical trials. This is to insure the safety of participants, the validity of data, and the appropriate termination of studies for which significant benefits or risks have been uncovered or when it appears that the trial cannot be concluded successfully. The NIH DSM policy covers clinical trials of all phases for which grant support is sought. DSM plans must be in place before grants supporting such studies can be funded. Applicants must submit a general description of the DSM plan for peer review as part of the grant application and, subsequently, a more detailed plan for review and approval by NCI staff prior to issuing a Notice of Grant Award.
Operational Definition of a Clinical Trial
For purposes of this document, we define a clinical trial operationally as a prospective study involving human subjects designed to answer specific questions about the effects or impact of particular biomedical or behavioral interventions; these may include drugs, treatments, devices, or behavioral or nutritional strategies. Participants in these trials may be patients with cancer or people without a diagnosis of cancer but at risk for it.
In the area of molecular or imaging diagnostics, we consider a study to be a clinical trial if it uses the information from the diagnostic test in a manner that somehow affects medical decision-making for the study subject. In this way the information from the diagnostic may have an impact on some aspect of outcome, and assessment of this impact may be a key goal of the trial. By contrast, studies that do not use information from the diagnostic test in any manner that can affect the outcome of study subjects, but whose objective is only the gathering of data on the characteristics of a new diagnostic approach, are not clinical trials and are not covered by this DSM policy, unless performing the diagnostic test itself imposes some risk on study subjects.
Behavioral clinical trials include interventions whose goals are to increase behaviors (e.g. cancer screening, physical activity, fruits and vegetable intake), eliminate or reduce behaviors (e.g., smoking, sun exposure) and/or improve coping and quality of life (e.g., among cancer survivors) and reduce the negative sequelae of treatment. Interventions may pertain to cancer prevention, early detection, treatment, and suvivorship. Observational studies and those that do not test interventions are not clinical trials.
Requirements for a DSMB
For some time now NCI policy has required that Data Safety Monitoring Boards (DSMB) be in place for all Phase III randomized clinical trials (http://deainfo.nci.nih.gov/grantspolicies/datasafety.pdf). The present document modifies this policy, in that there is no longer a blanket requirement for DSMB in the cases of low-risk behavioral and nutritional trials. As discussed further below, all such trials should include a DSM plan, but this may or may not include a DSMB, depending chiefly on the anticipated level of risk to participants.
Nor does NIH or NCI policy require that formal DSMB's be constituted for clinical trials other than Phase III, though investigators or institutions may wish to do so for certain non-Phase III trials involving particular risk, complexity, likely decisions about early stopping, or the need to obviate conflict of interest.
The Role of Institutional DSM Plans
Cancer clinical trials funded by the NCI are conducted in thousands of institutions nationwide. Many of these institutions -- notably the comprehensive and clinical cancer centers -- have particularly intensive clinical research portfolios that include dozens or hundreds of trials. It makes sense for such institutions to have in place institutional plans for an effective DSM process. An effectively formulated and executed institutional plan should improve both participant protection and trials conduct and should greatly reduce the need to set up new policies ad hoc on a trial-by-trial basis. For most investigator-initiated grant applications supporting clinical trials in an institution with an already approved institutional plan, the investigator should only have to supply the approved institutional plan in the human subjects section of the grant application and describe how it applies to the specific trials.
Tailoring Institutional DSM Plans to Specific Studies
The NCI clinical trials portfolio encompasses a vast array of investigation; examples range across early feasibility studies in treatment, prevention, or diagnosis; nutritional interventions to modulate risk of cancer; gene transfer; and behavioral research relating to cessation of tobacco use. Accordingly, the essential elements for DSM outlined below are described in general terms, and we do not stipulate details of how this process should be carried out. We have used general language to describe the essential content of such plans, leaving to individual institutions and investigators wide discretion in how to carry out these activities in an effective manner.
Clearly, a sensible DSM plan for a particular clinical trial must be based on the medical or health-related context of the particular study and, in particular, the degree of risk to which participants in the trial are exposed. In applying an institutional plan to a particular trial, therefore, the principal investigator will consider whether the institutional plan is sufficiently specific or whether it needs some further tailoring to the conditions of the particular trial. An institution might choose to have one general plan, which investigators tailor to individual trials. Alternatively, the institution might choose to have a plan that is essentially formulated in modules, each of which describes in adequate detail how monitoring will be accomplished for a major class of trials that the institution supports (e.g., early-phase studies in treatment, behavioral studies, bone-marrow transplantation, chemoprevention studies in healthy populations, etc.). Investigators can then apply these plans to particular protocols with little or no change in the description. For purposes of NCI review, as noted above, investigators may append the institutional plan to the human subjects section of their own grant applications and use these institutional documents in their interactions with NCI staff reviewing their plan. Under most circumstances NCI anticipates that a properly prepared institutional plan should suffice both for peer review and for NCI staff review.
Review of DSM Plans by NCI Staff
NCI staff review of institutional or individual DSM plans prior to grant award will focus on the adequacy with which the plan covers the essential elements outlined below. It is not necessary that submitted plans (whether they are institutional or individual) cover all possible aspects of each element down to the last detail. Rather, the plan should describe processes for dealing with these elements such that a reasonable reviewer would conclude that the institution or investigator has a serious, robust process in place for assuring the safety of research participants and the oversight of data integrity.