Transcript Cluster Randomized Trials

Cluster Randomized Trials
Lainie Friedman Ross, MD, PhD
Carolyn & Matthew Bucksbaum Professor of
Clinical Ethics
University of Chicago
© Ross, 2013
What is a cluster randomized
trial (CRT)?
A cluster randomized controlled trial is a
type of trial in which groups of subjects (as
opposed to individual subjects) are
randomized.
So, we randomize different communities, clinics, or
cities to either get or not get a particular
intervention (or for one cluster to get treatment A
and one to get treatment B.
In contrast with a traditional clinical trial in
which we randomize individuals.
© Ross, 2013
Who is the research subject?
 Cluster randomized trial to improve adherence to
guidelines for hypertension drug prescribing
 Providers were randomized to receive
 general guideline education
 General guideline education plus patient-specific reminders
 Outcomes
 Overall compliance with guidelines
 Adequacy of patient’s BP as ascertained by review of patients’
charts
 Questions:
 Who are the subjects?
 Whose consent is needed?
© Ross, 2013
Who is the research subject?
 COMMIT: Community Intervention Trial for Smoking Cessation
 22 cities in US and Canada randomized to
 Either a community-level antismoking intervention delivered through mass
media, health care professionals and worksites or
 No intervention
 Investigators interacted with the community by forming a board of
community representatives.
 Individual respondents not aware they were involved in a trial, although
they provided verbal consent to complete a telephone survey
 Questions
 Who are the subjects?
 Whose consent is needed?
 What role can/should community board play?
 Are Community Reps morally necessary?
 Are they morally sufficient?
© Ross, 2013
Randomized Control Trial (RCT) vs.
Cluster Randomized Trial (CRT)
 What question does the trial seek to answer?
 RCT: is drug A superior to drug B
 CRT: Questions are typically more complex and inherently pragmatic than those
in conventional RCTs
 the effectiveness of multicomponent strategies intended to promote health or to prevent
disease,
 to encourage the adoption of uncontentious processes of care; and
 to evaluate the efficacy of certain explicit algorithms designed to influence the delivery of
care.
 When do you want to use CRT for pragmatic and logistical reasons?
 Intervention at community level (*fluoride in the water)
 Specialized resources (specially programmed devices such as computer based
clinical decision support tools)
 To minimize contamination bias which occurs…
 When providers assigned to deliver the innovative intervention also treat individuals
randomized to the control group; or
 When providers consciously or unconsciously adopt aspects of the comparator
intervention because of proximity to, or informal interaction with, others delivering the
intervention; or
 When patients modify their behavior after interacting with patients receiving the
innovative interventions
© Ross, 2013
Problems with CRT
 Interventions intended to influence at one level whereas
outcomes measured at another level
 Attempt to change doctors behaviors; determine if successful by
measuring patient outcomes.
 The level of outcome data analysis—best termed the level of
inference-determines what data must be obtained, about
whom, and the statistical analyses employed (and whose
consent is necessary).
 CRTs reduce the statistical efficiency relative to RCTs that
randomize the same number of individuals
 The threshold for determining the magnitude of an effect
necessary to have public heath significance may be quite
different than the threshold for “clinical significance”, a
consequence of the prevention paradox:
 Low intensity interventions that are less efficacious but can be delivered
to a large number of people may have a more pervasive impact
© Ross, 2013
Weijer et al. Ethical issues posed by
cluster randomized trials (CRTs) in
health research. Trials 2011; 12: 100
1) Who is a research subject?
2) From whom, how and when must informed consent be
obtained?
3) Does clinical equipoise apply to CRTs?
4) How do we determine if the benefits outweigh the
risks of CRTs?
5) How ought vulnerable groups be protected in CRTs?
6) Who are gatekeepers, and what are their
responsibilities?
© Ross, 2013
References
 Eldridge SM, Ashby D, Feder GS. Informed patient consent to participation in cluster
randomized trials: an empirical exploration in primary care. Clinical Trials 2005; 2: 91.
On the web at: http://ctj.sagepub.com/content/2/2/91.long
 Mann H, Reyes M. Identifying the Human Research subject in cluster randomized
Controlled Trials. IRB: Ethics and Human Research 2008; 30 (5): 14-18
 Medical Research Council (UK). Cluster randomised trials: Methodological and ethical
considerations. 2002. On the web at:
http://www.mrc.ac.uk/Utilities/Documentrecord/index.htm?d=MRC002406
 Weijer C, Grimshaw JM, Eccles MP, et al. (2012) The Ottawa Statement on the Ethical
Design and Conduct of Cluster Randomized Trials. PLoS Medicine. 9(11): e1001346.
On the web at:
http://www.plosmedicine.org/article/info%253Adoi%252F10.1371%252Fjournal.pmed.10
01346
 Weijer C, Grimshaw JM, Taljaard M, et al. Ethical issues posed by cluster randomized
trials in health research. Trials. 2011; 12: 100. On the web at:
http://www.trialsjournal.com/content/12/1/100. This is the 1st in a series of articles in
Trials by a group of Canadians who had a grant to study the issues.
© Ross, 2013