Contact Information

Organization

THE COMMITTEE ON HUMAN RESEARCH

UCSF GUIDANCE ON RESEARCH TOPICS AND ISSUES

DATA AND SAFETY MONITORING BOARDS (DSMB)
(January 2008)

•	Brief Overview
•	Acronyms/Terminology
•	Purpose of DSMB
•	Basis for this Guidance
•	When Is an Independent DSMB Needed?
•	When Is an Independent DSMB Not Needed?
•	What Do investigators Need to Include in the CHR Application?
•	What if the Sponsor's Plans Do Not Satisfy these Guidelines?
•	Do I Need to Submit DSMB Reports to the CHR?
•	What if I Have Additional Questions?
•	Where Can I Find Additional Information?

Brief Overview

All interventional studies involving more than minimal risk must include a Data and Safety Monitoring Plan (DSMP).  A DSMP is established to assure that each research study has a system for appropriate oversight and monitoring of the conduct of the study to ensure the safety of participants and the validity and integrity of the data.  The DSMP should specify whether or not there will be an independent Data and Safety Monitoring Board (DSMB).

Not all trials require an independent DSMB.  This guidance is published to help clarify when the CHR expects a DSMB to be involved.

Acronyms/Terminology

This guideline uses the term Data and Safety Monitoring Board (DSMB), as preferred by the NIH.  The FDA uses the term Data Monitoring Committee (DMC).

Top of Page

Purpose of DSMB

The primary purpose of an independent DSMB is to protect the research subjects through independent analysis of emerging data from the trial.  This differs from adverse event reporting in that the DSMB can review aggregate and unblinded data as the data accumulate, identify significant issues and trends during the study, and recommend changes in the study including recommending early termination of the study.  The DSMB reviews data for both safety and efficacy.  The protections afforded by this review apply to both current subjects and future subjects if the DSMB identifies the need to modify or even halt the trial.

In addition to the above, an independent DSMB protects the credibility of the trial by virtue of its independence from the study sponsors, and helps to ensure the validity of study results by reviewing data on subject accrual and conducting interim reviews.

Top of Page

Basis for this Guidance

This Guidance is based on and consistent with NIH policies and guidance on requirements for DSMBs.  The NIH requires DSMBs for all Phase III clinical trials, and many of the individual Institutes require broader use of DSMBs.  This is slightly more stringent than FDA guidance, which generally recommends DMCs for controlled trials comparing rates of mortality or major morbidity.  Links to more detailed information from NIH and the FDA are included at the end of this guidance.

Top of Page

When Is an Independent DSMB Needed?

•	The CHR generally expects DSMBs in the following situations:
	•	All Phase III studies require a DSMB, with the exception of low-risk behavioral and nutritional studies. For this discussion, “low-risk” refers to trials where subjects are expected to experience only minor side effects, and interim analyses are not crucial for the protection of the subjects. The involvement of a DSMB may still be requested for low-risk studies if the studies are exceptionally large, long term, and/or involve vulnerable subjects
	•	Phase II clinical trials which are multicenter and randomized require a DSMB, with the exception of low-risk behavioral and nutritional studies.
	•	Phase II studies which are “high risk” require a DSMB.  For this discussion, “high-risk” refers to trials of interventions associated with substantial side effects to subjects (e.g., side effects that could result in serious morbidity or death, or are irreversible), trials of diseases associated with high mortality or morbidity, and trials of highly experimental therapies (e.g., gene therapy). As a general guideline, DSMBs are needed for clinical trials of diseases with high mortality or morbidity, for clinical trials involving high risks, and for large, multicenter clinical trials.

Top of Page

When Is an Independent DSMB Not Needed?

•	The CHR does not generally expect DSMBs in the following situations:
	•	Single-center open-label Phase I and II clinical trials generally do not need a DSMB since the local investigator will have access to all data.
	•	A multicenter, high-risk Phase I clinical trial should not require a DSMB if there are very clear rules for stopping the trial.  For example, the CHR would not ask for a DSMB for a classic open-label dose escalation trial with clear and objective criteria for halting the dose escalation when unacceptable side effects are observed.  A DSMB is likely to be requested if the DSMP lacks objective criteria for continuing or halting the trial.
	•	A DSMB may not be feasible for clinical trials that are expected to accrue too quickly to allow for a DSMB to be constituted and complete data and safety monitoring.
IMPORTANT NOTE:  For some studies involving particularly vulnerable study participants (e.g., children or persons with impaired ability to consent), the CHR may also ask for a DSMB as an additional measure of subject protection.

What Do investigators Need to Include in the CHR Application?

•	If your study utilizes a DSMB, additional details about the membership and function of the DSMB should be included in the CHR application.  In particular, Part 5.B, Data and Safety Monitoring Plan should include information like the following:
	•	Composition:  Describe the expertise represented by the members of the DSMB.  DSMBs generally include members with expertise in biostatistics, clinical trials, and the disease and treatment being studied.  Other areas of expertise such as bioethics may also be useful.
	•	Independence:  Provide a clear statement that DSMB members will not participate in the study as investigators, will not have conflicts of interest regarding the study, the study sponsor, or any study drugs or devices being tested.
	•	Data:  Briefly describe the data the DSMB will review, e.g., data for primary or secondary endpoints (safety and efficacy), data for early termination of trial (stopping rules), or adverse events.
	•	Frequency of Review:  Describe how often the board will meet, whether based on the calendar or accrual targets.  If formal interim analyses are planned, describe when they will occur.
	•	Authority:  Describe the actions the DSMB is authorized to take.  DSMBs should have authority to recommend changes in the study, including discontinuation, if significant trends in safety or efficacy are identified earlier than expected.
	One copy of the DSMB Charter may be submitted but this is not required.

What if the Sponsor's Plans Do Not Satisfy these Guidelines?

The HRPP understands that site investigators may not have authority to make changes in multicenter studies.  However, there have been numerous cases where the concerns or recommendations of local IRBs and site investigators have led to changes in multicenter studies.  Investigators are strongly encouraged to share these guidelines with study sponsors or steering committees before submitting a CHR application.

If plans involving a Data and Safety Monitoring Board diverge from this guideline the CHR application should include a clear and complete rationale.  Please include this information in Part 5.B, Data and Safety Monitoring Plan.  Please call the HRPP office at 476-1814 and ask to speak with an Analyst of the Day for advice.

Do I Need to Submit DSMB Reports to the CHR?

Yes.  DSMB reports must be submitted to CHR within 10 days.  Please use the Reporting Form for Investigator Brochure, Safety Information or Publication, DSMB/DMC Report, Audit Report, and Hold on Study Activity.

What if I Have Additional Questions

If you have questions about whether the CHR will expect you to have a DSMB for your study, or what information about the DSMB to include in your application, please call the HRPP office at 476-1814 and ask to speak with an Analyst of the Day.

Where Can I Find Additional Information?

NIH Policies:             
http://grants.nih.gov/grants/guide/notice-files/not98-084.html
               
http://grants.nih.gov/grants/guide/notice-files/NOT-OD-00-038.html

FDA Guidance:         http://www.fda.gov/cber/gdlns/clintrialDSMB.htm

Susan S. Ellenberg, Thomas R. Fleming, and David L. DeMets.  Data Monitoring in Clinical Trials: A Practical Perspective.  John Wiley & Sons, Ltd, 2003.