RANDOMIZED CLINICAL

TRIALS
Mario Merialdi
phone 791 3387
Office X 39
[email protected]
Rationale
To evaluate the efficacy of a drug or a
procedure in the treatment or prevention of
disease
We compare a new drug with the one
currently in use or, if no such drug exits, to
a placebo (inert substance)
The groups
The treatment group is given the drug being
tested
The comparison group is given the drug in
current use or a placebo
Experiment
All the participants in the study are assigned
to either the treatment group or the
comparison group
Then the participants are followed over the
same time period to determine if there are
differences in the outcome of interest
between the treatment and control group
The investigator has direct control over the
assignment of subjects to study groups (this
is the main difference from observational
studies)
Not a new idea
Ambroise Par at the battle of
Villaine (1537)
I raised myself very
early to visit them,
when beyond my hope
I found those to whom
I had applied the
digestive medicament,
feeling but little pain,
their wounds neither
swollen not inflamed,
and having slept
through the night...
...The others to whom
I had applied the
boiling oil were
feverish with much
pain and swelling
about their wounds.
Then I determined
never again to burn
thus so cruelly the
poor wounded by
arquebuses.
Randomization
Random assignment (by chance) of
individuals to the treatment and comparison
groups
Why randomization?
We want the treatment and comparison
groups to be comparable in all respects
except the one being studied
We try to ensure as much as possible that
the distribution of all factors and population
characteristics, except for the therapy being
studied, is based on chance and not to some
other factor such as patient or investigator
preference (bias).
Randomization and trial results
The process of randomization will provide
comparable groups for most factors so that
differences in outcomes at the end of the
trial can be attributed to the intervention
being tested
Preventing biased assignment is important
as it permits a more definitive interpretation
of the trial results.
Randomized trials are not only
about treatment
Therapeutic trials evaluate the efficacy of
drugs or procedures to cure the disease
Preventive trials evaluate the efficacy of
drugs or procedures to prevent a disease
(e.g. Calcium supplementation for the
prevention of preeclampsia)
Intervention trials evaluate the efficacy in
treating the risk factors of a disease by an
intervention (somewhere between the two)
Procedures
The protocol
Contains the objective and specific
procedures to be used in the trial (scientific
rationale, methods for patient selection and
allocation, validity of clinical and
laboratory procedures, administration of the
intervention)
Must be written before the start of the trial
(justification for a new trial, guarantee of
the quality and integrity of the study,
scientific and ethical approval of the
proposed study, funding, publication)
The manual of operations
Detailed description of all the procedures
presented in the protocol
Intended for the study personnel
implementing the study activities
It is the guide for what the investigator is
supposed to do
Selection of study subjects
The investigator decides, on the basis of the
purpose of the study and the difficulty of
recruitment, the characteristics of the
population to be studied
Eligibility criteria
Exclusion criteria
Internal vs. External validity (typical
problem in epidemiology)
Sample size
A difference to be a difference must make
a difference
The study must have a sufficient sample
size to have adequate statistical power so
that if the treatment being studied is
effective, its efficacy will not be mistakenly
missed in the trial
Possible outcomes of a trial
Treatment is
effective
Treatment is not
effective
Results show
treatment is
effective
Correct
conclusion
False positive
(Type I error)
Results show
treatment is
not effective
False Negative
(Type II error)
Correct
conclusion
Factors determining the sample
size calculation
With a sample size calculation the
investigator determines the number of
patients required to detect a designated
treatment difference in a specific outcome
with specified levels of Type I and Type II
protection
Informed consent
The study subjects must be clearly and
understandably informed of the risks and
possible benefits of the trial intervention
Informed consent must be obtained and
documented for all study subjects
Requested by scientific institutions (Ethical
review board), funding agencies, and
journals
Random allocation
Eligible subjects who consent to participate
in the study are recruited and randomly
assigned to the treatment and comparison
groups
Random allocation can be done in different
way (computer generated list of subject
numbers)
Stratification for some factors known to
influence the outcome (more needed for
small studies).
Follow up
Follow up patients from recruitment to the
end of the study
Main purpose: assess how many patients
will develop the outcome of interest
Compliance or Adherence
We must be sure that the treatment group is
actually receiving the intervention
Compliance must be constantly assessed
during the study
Different strategies can be used:
count unused tablets
biological methods
Preventing bias in the assessment
of the effect: Masking
Knowledge of whether the participant was
in the treatment or comparison group can
influence the perception of he benefit by the
subject or the observation and care by the
investigator: biased assessment of the effect
Single masking (subject)
Double masking (subject and observer)
Triple masking (subject, observer, and data
reviewer)
Analysis of the results. Are the
two groups comparable?
Are there differences in the characteristics
of the two groups at baseline?
If there are no differences it means that
randomization resulted in the formation of
comparable and evenly balanced groups
Analysis of the results. Was the
intervention effective?
Incidence rate for the development of the
disease are calculated among the subjects in
the treatment and control group
If the incidence rate of the disease is greater
among the persons in the control group than
among the ones in the treatment group, we
have evidence that the intervention
decreased the risk of disease
What is an incidence rate?
Number of cases of a disease occurring in
one of the study groups
___________________________________
Number of persons in the study group
Incidence rates of preeclampsia
Incidence rate = cases/number of subjects
Incidence rate in calcium group 119/4250 =
0.028
Incidence rate in placebo group 170/4250 =
0.040
The incidence rate is greater in the placebo
than in the calcium group
We have evidence that calcium
supplementation is effective in reducing the
risk of preeclampsia
How strong is the treatment
effect?
The relative risk measures the strength of
the treatment effect
Incidence rate in the treatment group /
incidence rate in the control group
In our example 0.028 /0.040 =0.7
How to interpret relative risk?
If relative risk is = 1 there is no evidence of
an effect of the treatment between (the
incidence rate in the treatment group is the
same as in the control group)
If relative risk is < 1 the treatment is
beneficial (decreases the risk of disease)
If relative risk is > 1, the treatment is
associated with an increased risk of disease
Other way to assess treatment
effectiveness
Comparison between average values of an
outcome between the two groups (e.g.
average blood pressure levels)
If the study is longitudinal we can perform a
survival analysis or use longitudinal data
analysis
A major source of difficulty
Maintaining the follow up of the individuals
in the study (assure compliance with
treatment and study procedures, minimize
drop outs and lost to follow-up)
It is important to trace as many subjects as
possible
Try to get information on the individuals
lost to follow up and look if they are
somehow different from the individuals
who stay in the cohort
Other important issues
Adequate sample size (can be very big if the
outcome is rare)
Ethical considerations (is it correct to
withhold a treatment that is potentially
beneficial?)
Expensive
Generalizability of the results (multicentre
trials)
How randomized clinical trials
differ from observational studies?
The investigator has direct control over the
assignment of subjects
Random allocation of study subjects to
treatment and control group
Preventing biased assignment and outcome
assessment permits a more definitive
interpretation of the results
Important
Randomized clinical trials provide strong
epidemiological evidence but always
consider all the criteria of causality.