Transcript What is a Clinical Trial

What is a Clinical Trial
(alpha version)
John M. Harris Jr., MD
President
Medical Directions, Inc.
Learning Objectives
 The
US Government’s definition of
a clinical trial
 Important terms that describe
clinical trials
 The four phases of clinical trials
 Controls in clinical trials
 Randomization in clinical trials
 Blinding in clinical trials
The US Government’s
Definition of a Clinical Trial
US Government’s Definition


Practical definition of the term, “clinical
trial” that clarifies the difference between
clinical trials and other types of human
research
If a research project is funded or
regulated by the US government, then it
must also meet a number of legal and
regulatory requirements that do not
apply to other types of research
US Government’s Definition



It is a prospective study
It is in humans
It is designed to answer questions about
an intervention
–
–

a biomedical intervention
a behavioral intervention
A prospective biomedical or behavioral
research study of human subjects, that
is designed to answer specific questions
about biomedical or behavioral
interventions.
What is NOT a Clinical Trial?
A
study that seeks to understand
how a disease affects people is not a
clinical trial unless participants are
deliberately given either the disease
or a treatment for it.
Drug abuse research projects that
are not clinical trials




Primate (monkey) testing of a new HIV therapy
intended for humans
Ethnographic research to understand information
gathering patterns of injecting drug users
A program evaluation of a government initiative
to reduce cocaine use by unwed mothers
A descriptive (epidemiologic) study of drug abuse
patterns among persons receiving naltrexone for
alcohol dependency
Examples of activities that are
clinical trials

Suppose you supervise four drug abuse
clinics
– female patients may not be attending drug
abuse program due to child care problems
– provide supervised child care and a hot meal in
two clinics for 6 months
– Is this a clinical trial?

Yes - prospectively testing an intervention
(food and child care) in humans.
What do you have to worry about?
 Ethical
issues
 Scientific
issues
 Successful
management of ethical
and scientific issues, in addition to
any local regulatory requirements
Important Terms That
Describe Clinical Trials
Four Major Features to
Clinical Trials

The stage of the study intervention (study
phase)

Whether the trial is a controlled study

Whether participants are randomized

Whether participants are unaware of (or
“blinded” to) treatments and results
Key terms to describe the four
basic features of a clinical trial
FEATURE
Study Stage
Controls
Randomization
Blinding
TERMS USED
Phase I
(New
intervention)
Phase II
(Intervention is
promising)
Uncontrolled
Non-randomized
Unblinded
Singleblind
Phase III
(Real world
evaluation )
Placebo
Phase IV
(Intervention is
in use)
Active Control(s)
Randomized
Double-blind
Triple(+)-blind
The Four Phases of
Clinical Trials
The Four Phases of
Clinical Trials




Phase I – The intervention is new and is being
tested in a small group of people for safety
Phase II – The intervention looks promising and
is being evaluated in a larger group of people for
safety and efficacy
Phase III – The intervention appears safe and
effective and is being evaluated in a real-world
study
Phase IV – The intervention is now in use and is
being monitored for safety
How clinical trials can be
interpreted


The US NIH definitions of a clinical trial
used for human subjects research give
estimates of the number of people
involved and specifically state that all
phase III trials require comparison studies
The US NIH definitions of a clinical trial
used to register clinical trials do not
mention the number of persons involved
in the different phases but note that phase
II studies involve comparisons (controls),
while phase III studies may not
Controls in Clinical
Trials
What is a clinical control?
 What
has been done in the past?
 Current treatments (standard
treatment)?
 A similar-looking but ineffective
treatment (placebo)?
 Whenever
an intervention is
compared to something else, the
alternative is called a “control”.
Placebo Controls


Designed to duplicate the experience of the
intervention without providing the active
component
Recognizes that many health outcomes are
dependent on the patient’s willingness to
participate and the patient’s and physician’s
perception of benefit

Can provide clear improvement in symptoms (the
“placebo effect”)

Can more accurately determine the extent to
which the intervention alone affects treatment
Interventions
 The
control contains an active
component and is, therefore, an
active control
 Active
controls are frequently done in
phase III trials where the intent is to
understand the benefits of a new
treatment in a real-world setting
Randomization in
Clinical Trials
Why is randomization necessary

How does the researcher decide who gets
which treatment?
– every person will get the new treatment
– some persons may be willing to accept the new
treatment while others are not.

Whenever there is a good possibility that
the control group and the intervention
group are not identical, there is also a
good possibility that the results will differ
because of some factor totally unrelated
to the treatment.
What is Randomization?



Randomization -the assigning of persons
to the control group or the intervention
groups based on chance.
An essential feature of good clinical trial
design whenever there are one or more
control groups.
The process of randomization must be
totally unrelated to the variables of
interest.
Common randomization schemes

Complete randomization

Block randomization

Urn randomization

Stratified randomization

Cluster randomization
Complete randomization



Individuals are assigned to control and
intervention groups based on a random
event that is proportional to the size of the
groups
The event could be a coin toss for two
groups of equal size or a more complex
computer generated algorithm
Resulting group sizes may not be equal
(or in the desired ratio) because the
process is truly random
Block randomization



Also known as permuted block randomization
Ensures that the number of subjects assigned to
treatment and control groups is balanced
Individuals are placed into small groups (blocks)
and then randomly assigned to control and
intervention groups.
– Example:




Two groups (A & B) and block size is four
Two persons are randomly assigned to group A
Remaining two in that block are assigned to group B
Creates the potential for bias because the
assignment of certain block members is
determined by the assignment of others.
Urn randomization


Form of “adaptive randomization” whose
intent is to balance the number of
individuals in each group without
specifying that an individual must be
assigned to a specific group
The chances that an individual will be
assigned to a treatment or control group
are allowed to vary based on prior
assignments
Stratified randomization



Used to guarantee that there is a balance
of certain characteristics in treatment and
control groups
Individuals are separated into subgroups
(strata) based on a predetermined feature
Randomly assigned to control and
intervention groups using one of the
methods mention in previous slides
Cluster randomization




Created in an attempt to deal with issues of
contamination within subgroups
Individuals are placed in subgroups
The entire subgroup (cluster) is randomly
assigned to be part of a control or intervention
group
Increasingly popular but requires an adequate
number of clusters and careful design
Blinding in Clinical
Trials
The purpose of blinding



Many clinical trials lack well-defined,
objective endpoints
Such measures are subject to intentional
manipulation and unintentional bias
Researchers may seek to disguise whether
the subjects are receiving the intervention
or the control treatment
What is Blinding?
 The
term “blinding” is used to
describe the techniques used to
disguise treatments, and sometimes
outcomes
 Different
types of blinding
–Unblinded or Open Label
–Single blind
–Double blind
–Triple blind
Summary




US Government’s Definition of a Clinical Trial as a
prospective study, that deals with human subjects
and designed to answer questions about an
intervention
There are important features and terms that help
describe clinical trials
The four phases of clinical trials help define the
intent of the research project than its scientific
details
Controls, randomization, and blinding are tools
used in in clinical trials to reduce data manipulation
and bias