report sponsored in 1967 by the National Heart Insti-
tute, now the National Heart, Lung, and Blood Institute
(NHLBI). Dr Bernard Greenberg chaired this task force,
and its report, known as the Greenberg Report, was
eventually published in 1988 in Controlled Clinical Tri-
als.
1
The task force developed an organizational struc-
ture by which the NHLBI would manage and conduct
several large phase III RCTs then being planned. One
feature of the proposed structure was an external inde-
pendent advisory body, called a Policy Advisory Board,
which was to provide guidance and input to the NHLBI
about the conduct and progress of the forthcoming tri-
als. The first trial to adopt the recommendations of the
Greenberg Report was the Coronary Drug Project
(CDP), which studied cholesterol lowering in men who
had had a recent heart attack and which began to enroll
patients in 1966. As the CDP progressed, a subcommit-
tee of the Policy Advisory Board emerged to focus on
monitoring the safety of the trial and the efficacy of
treatment. The subcommittee functioned essentially as
an early DSMC; its activities and decision making were
summarized later in an article by Dr Paul Canner.
2
Soon
after, in the late 1960s and early 1970s, as other NHLBI
trials were launched to study hypertension, blood dis-
eases, and pulmonary diseases, the DSMC model was
repeated and became the standard for that institute. In
the early 1970s the National Eye Institute began the Dia-
betic Retinopathy Study and it also used a DSMC for its
investigation. Over time, DSMCs became more widely
used in trials sponsored by the National Institutes of
Health, which now requires DSMCs for all its phase III
trials.
3
In recent years academia and industry have
increasingly partnered to conduct large-scale, pivotal
phase III trials for new therapies; in this environment,
DSMCs have more and more frequently been integrated
into the organizational structure for trials studying
effects on mortality rates and irreversible morbidity out-
comes.
Although some accounts of the experiences of DSMCs
in specific trials have been published, the DSMC as an
actual functioning entity has not, despite having a his-
tory of more than 40 years, been widely discussed in the
literature. Consequently, its proliferation in contempo-
rary studies is often accompanied by confusion about its
role, scope, organization, and the rules by which it
operates. The articles in this series have addressed many
practical questions that any sponsor of an RCT must
consider when forming such a committee. For many of
The randomized clinical trial (RCT) has become a
principal research tool in the evaluation of new medical
products, including drugs, biologics, and devices. A crit-
ical component of the the RCT in studies that investi-
gate mortality or irreversible morbidity is the Data and
Safety Monitoring Committee (DSMC). The American
Heart Journal has now published all seven articles in a
series that presents the proceedings of a conference on
DSMCs that was sponsored by the Duke Clinical
Research Institute (DCRI) and the American Heart
Journal. It was our privilege to serve with Dr Robert
Califf and Dr Kerry Lee as cochairpersons for the confer-
ence, and it is our hope that this series on the philoso-
phy and practice of DSMCs will be widely read.
As these articles make clear, the DSMC is a vital part
of the organizational components of a clinical trial,
which includes the Steering (or Executive) Committee,
Data Management Center, Statistical Analysis Center
(sometimes the latter two are combined into a Data
Coordinating Center), and participating clinical centers.
The primary charge of a DSMC is to monitor the ongo-
ing trial for safety and for early convincing evidence of a
treatment benefit in the therapy under investigation, in
which case it may recommend early termination of the
study. Members of the DSMC may also provide sugges-
tions about the trial protocol and offer views on emerg-
ing design issues, data quality, and patient recruitment.
They have varied responsibilities to a number of con-
stituents including the patients who enter the trial (as
well as future patients), the clinical investigators and
their Executive Committee, the trial sponsor, and, indi-
rectly at least, the regulatory agencies in the clinical trial
field. With a long and highly successful track record,
DSMCs have played a decisive role in pivotal trials in
cardiology, cancer, ophthalmology, AIDS, and many
other diseases.
The origins of the DSMC can be traced to a task force
From the
a
Department of Statistics and Biostatistics, University of Wisconsin, Madi-
son, Wis, and the
b
Division of Cardiology, Department of Medicine, McMaster Uni-
versity, HGH McMaster Clinic, Hamilton General Hospital, Hamilton, Ontario,
Canada.
Submitted July 8, 2000; accepted December 29, 2000.
Reprint requests: David L. DeMets, PhD, Department of Statistics and Biostatistics,
University of Wisconsin, 600 Highland Ave, K6/446, Madison, WI 53792-4675.
E-mail: demets@biostat.wisc.edu
Am Heart J 2001;141:548-9.
Copyright © 2001 by Mosby, Inc.
0002-8703/2001/$35.00 + 0 4/1/113837
doi:10.1067/mhj.2001.113837
The Data and Safety Monitoring Committee:
Some final thoughts
David L. DeMets, PhD,
a
and Salim Yusuf, MD
b
Madison, Wis, and Hamilton, Ontario, Canada