In its approach to monitoring the safety of human subjects and the data acquired through clinical research, the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) follows the general NIH policy as published in the NIH Guide for Grants and Contracts on June 10, 1998
(http://www.grants.nih.gov/grants/guide/notice-files/not98-084.html).
The NICHD requires a Data and Safety Monitoring Board or Committee (DSMB/C) to provide oversight for large, multi-site Phase III clinical trials. Monitoring of Phase I or Phase II trials depends upon the size of the trial and the research area.
In all cases, communication with relevant NICHD program staff about plans for data and safety monitoring is required.
Phase I Trials
Phase I trials are initial evaluations of safety in human volunteers or in patients affected by a condition/disease. They are usually small studies that focus on the following types of outcomes: the safety and tolerability of a new drug; dosing range studies to find the maximum tolerated dose; pharmacokinetic studies of a new drug; feasibility of a behavioral intervention. Although these studies may enroll fewer than 30 subjects, they often require intense monitoring of subjects because of safety and design issues. Repetitive, intense monitoring is difficult in the setting of a DSMB/C; however, oversight by a small group of independent investigators with relevant expertise is recommended.
Phase II Trials
Phase II trials are pilot trials of safety and efficacy of an intervention or a new indication in a select population. They are larger than Phase I trials, enrolling up to 200-300 patients per treatment arm. When possible, they are masked, e.g., the study subject, the treatment provider or both are unaware of the treatment assignment. To ensure validity of the results, randomization of study subjects to treatment assignment is required.
Because the goal of Phase II trials is to establish safety and efficacy of the intervention, most Phase II trials should be monitored by a DSMB/C. Small "open label" studies, in which interim data analyses will not be performed, may be monitored by an alternative strategy, e.g., by independent investigators who may be masked to the treatment assignment or to the hypothesis being tested.
Phase III Trials
Phase III studies typically enroll several hundred to several thousand participants over a period of months or years. They are trials designed to assess the safety and efficacy of clinical interventions, including drug therapies, medical devices, or management strategies, usually after preliminary data have suggested efficacy of the intervention. They are conducted in patient populations for which the intervention is intended and provide much of the information for the package insert and labeling of a medication.
Classical Phase III clinical trials are randomized, double-masked placebo controlled studies; however, they may also include randomized, double-masked comparisons of new therapies/devices with the existing standard. A DSMB/C is required for all Phase III trials.
DSMB/C Implementation
Role
The role of the DSMB/C is defined prior to initiating a clinical trial. This normally includes review of the protocol before it is implemented by the investigators, review of implementation and progress of the study, and ongoing review of the accumulating data to detect evidence of early, significant benefit or harm for patients while the trial is in progress. This latter review, beyond that provided by the IRB, serves as a means of additional human subjects protection, but does not supplant the regulatory requirement for the investigator(s) to report serious and unanticipated adverse events to the Food and Drug Administration.
The DSMB/C normally has responsibility for defining the frequency of its meetings, based on the study's enrollment rate and the potential risk to participants. It may also establish "stopping rules" based on 1) how great a difference in effects is required to demonstrate clear clinical significance and make the data convincing to clinicians, and 2) statistical techniques designed to preserve the final significance level of the trial.
Membership
Constitution of the DSMB/C depends upon the research question and the funding mechanism. Cooperative agreements that fund clinical trial networks normally have a standing DSMB/C; trials funded by other mechanisms usually are constituted to monitor the individual trial. Membership is tailored to the expertise required to monitor the trial(s). The DSMB/C often has 5 - 7 members, including one or more individuals representing each of the following: clinicians with expertise in the specific disease and/or treatment under study, biostatisticians (usually from the data coordinating center managing the data for the study), an epidemiologist, and an ethicists/patient advocate. Members also may include a trialist and a representative from the Institutional Review Board (IRB). Normally, members are not employees of the NIH or from the same institution as the Principal Investigator(s).
Meetings
The DSMB/C may meet on a regular basis in either "open session" or "closed session." NICHD staff who participate in the trial as investigators do not attend "closed sessions" during which safety and efficacy data are reviewed; however, staff who are representing the interests of NICHD or who are Program Officials attending in an ex officio capacity may attend such sessions.
NICHD staff may attend "open sessions" to provide the DSMB/C with relevant updates in the field under study that might impact the deliberations of the DSMB/C, to provide additional details on the implementation or conduct of the trial, and to answer questions of the DSMB/C.
The DSMB/C recommends continuation, modification or termination of the trial to the NICHD following each meeting. These recommendations are reported to the Principal Investigator(s), who normally accept the recommendations of the DSMB/C, and to the IRB(s) of the participating institutions.
